Cuprins Tehnomus XVI

TEFAN CEL MARE UNI VERSI TY of SUCEAVA
FACULTY of MECHANI CAL ENGI NEERI NG, MECHATRONI CS AND

MANAGEMENT
TECHNOLOGY AND MANAGEMENT DEPARTMENT
TEHNOMUS
NEW TECHNOLOGI ES AND PRODUCTS I N MACHI NE MANUFACTURI NG
TECHNOLOGI ES
2011
Copyright Technology and Management Department
FACULTY of MECHANICAL ENGINEERING, MECHATRONICS AND MANAGEMENT
STEFAN CEL MARE UNIVERSITY of SUCEAVA
Editura Universitii tefan cel Mare din Suceava
Str. Universittii nr.13, 720229, Suceava, Romania
Telefon: 0230-216147 int. 273
Editura este acreditat de
Consiliul Naional al Cercetrii tiinifice din nvmntul superior CNCSI S
COD 115
Authors instructions
The full-length paper for publication should not be less than four pages and not more than eight, in
English. Please follow the conference format. A maximum of two papers are accepted, by either a main
author or a unique author. Prospective authors are kindly invited to submit full text papers including results,
tables, figures and references. Full text papers will be accepted by email tehnomus@fim.usv.ro.The
Organizing Committee can accept only*.doc or *docx format.
All the papers will be evaluated by two reviewers. Selected papers will be published in the
TEHNOMUSNew Technologies and Products in Machine Manufacturing TechnologyJ ournal.
Web addresshttp://www.tehnomusjournal.fim.usv.ro
Contact:
Faculty of Mechanical Engineering, Mechatronics and Management
Department of Technology and Management
Stefan cel Mare University of Suceava
Str. Universittii nr. 13, Corp B,
720229 SUCEAVA, ROMANIA
Phone: (+40) 230 216147 int. 260
E-mail: tehnomus@fim.usv.ro
Journal web site: www.tehnomusjournal.fim.usv.ro
Faculty web site: www.fim.usv.ro
University web site: www.usv.ro
CONTENT:
1. COMPARI SON OF A MACHI NE OF MEASUREMENT WI THOUT
CONTACT AND A CMM(1) : OPTI MI ZATI ON OF THE PROCESS OF
METROLOGY.
WOLFF Valery, TRAN DINH Tin
9
2. EFFI CI ENT DEVELOPMENT OF HI GH PERFORMANCE POLYMER
PARTS USI NG ADVANCED SI MULATI ON METHODS
Achim Frick, Timo Dolde, Dorota Sich
15
3. SUPERI OR VALORI SATI ON OF THE METALLI C SCRAPS FOR
OBTAINING BI METALLI C STEEL-BRONZE PARTS
Anisoara Ciocan, Florentina Potecasu
21
4. STUDY FOR THE VARI ATI ON OF HARDNESS AND RESI LI ENCE
CHARACTERI STI CS AFTER HEAT TREATMENT FOR STAI NLESS
STEELS
Manuela Cristina Perju, Dragos Cristian Achitei, Carmen Nejneru, Anca Elena
Lrgeanu, Mihai Axinte, Ion Hopulele
27
5. AN I NTELLI GENT SYSTEM SEARCH FOR I NDOOR COLOR
RECOGNI TI ON: A COMPARI SON
Mrio Antnio Ramalho
31
6. MATERI AL SI MULATI ON OF POLYMERS CONTAI NI NG SPHERI CAL
STRUCTURES
Achim Frick, Melissa Matzen, Timo Dolde
37
7. MONTE CARLO PI VOTAL CONFI DENCE BOUNDS FOR WEI BULL
ANALYSI S, WI TH I MPLEMENTATI ONS I N R
J urgen Symynck, Filip De Bal
43
8. MODERN GREEK LANGUAGE FREQUENCY COUNTS FOR TEXT
ENTRY DEVI CES
Ilias Sarafis, Anastasios Markoulidis
51
9. ERRORS OF MECHANI CAL TREATMENT PROVOKED BY THERMAL
DEFORMATI ONS OF TECHNOLOGI CAL SYSTEM
V.I.Savulyak, S.A.Zabolotniy
59
10. I RON REMOVAL FROM WASTEWATER USI NG CHELATI NG RESI N
PUROLI TE S930
Petru Bulai, Elena-Raluca Cioanca
63
11. ANALYSI S OF CONDI TI ONS FOR OBTAI NI NG A KEYHOLE REGI ME
FOR LASER WELDI NG
Remus Boboescu
69
12. ASPECTS REGARDI NG THE QUALI TY OF CAST I RON PRODUCED I N
S.C.FONTUR S.A. SUCEAVA
Gramaticu Mihai; Cosmaciuc Vasile, Costiug Gina, Hritcu Ionut Marius
77
13. OPTI MI ZATI ON OF THE PLASTI C I NJ ECTI ON PROCESS THROUGH
THE MODI FI CATI ON OF THE PROCESS FUNCTI ONAL
PARAMETERS
Teodor - Daniel Mndru, Ciprian Dumitru Ciofu, Dumitru Nedelcu
87
14. A PROBABI LI STI C DYNAMI C MODEL ASSOCI ATED TO A J ACKSON
NETWORK, WITH SERIES QUEUES; THE SOLUTI ON OF THE
ASYMPTOTI CALLY STABLE STEADY STATE
Popa Marin, Drgan Mihit, Popa Mariana
93
15. TECHNI CAL SOLUTI ONS TO DESI GN AN EQUI PMENT FOR I CE
PARTS RAPI D PROTOTYPI NG
Nicolae Ionescu, Aurelian Visan, Alexandru Savin, Mihai Trifnescu
101
16. COMPOSI TE SHAPE MEMORY ALLOYS USED I N ENERGY
DI SSI PATI ON APPLI CATI ONS
I. Cimpoesu, S. Stanciu, A. Enache, D. Dan, V. Manole, P. Paraschiv
107
17. I RESEARCH ON THE CORROSI ON OF CONSTRUCTI ON MATERI ALS
FI SSURE MACHI NERY HYDROCARBONS I N CONTAMI NATED
ENVI RONMENTS WI TH STAFF FROM CORROSI VE BY CRUDE OI L
DI STILATI ON
Morosanu Marius , Ovidiu Georgescu
113
18. SOURCES OF POLLUTI ON I N SI DERURGY AND TECHNI QUES FOR
REDUCI NG NOXI OUS EMISSI ONS
Elisabeta Vasilescu , Ana Doniga , Marian Neacsu
117
19. THEORETI CAL CONSI DERATI ONS ABOUT THE PRECI SI ON
I NCREASE OF CONI CAL DRAWI NG PI ECES
Lucian V. Severin, Traian Lucian Severin
123
20. METHOD FOR THE CONTROL OF THE GEOMETRI C AND
CONSTRUCTI VE ELEMENTS OF THE SPI RAL-FACE AND END
MI LLI NG CUTTER
Strjescu Eugen, Pavlov Olimpia, Dogariu Constantin
127
21. ZI NC COATI NGS ON STEEL SUBSTRATE ATTAI NED BY DI FFERENT
ELEMENTS ADDED
Radu Tamara, Vlad Maria
133
22. ACCELERATI ON TEST MACHI NE
Marc J uwet, Koert Bruggeman, Filip De Bal
139
23. DEVELOPMENT OF WASTE MANAGEMENT SYSTEMS I N AN
I NTEGRATED SHI PYARD
Daniela Buruiana
145
24. METHOD FOR DETERMI NATION OF THE DI STANCE BETWEEN
ELECTRODES TO ELECTROHYDEAULI C UNDERWATER THROUGH
SCI NTI LLAS DISCHARGE
Dumitru Iacob
151
25. SUPERI OR WASTE RECOVERY I N THE METALLURGI CAL
I NDUSTRY
Elisabeta Vasilescu , Ana Doniga, Alexandru Chiriac
155
26. I MPACT OF STREETS LANDSCAPE ON PEOPLE ORIENTATION
AND ROAD SAFETY I N THE CI TY
A. Polyakov, V. Shvets, O. Veremiy, M.Grabenko
161
27. I MPLEMENTI NG A SPC I NTEGRATED SYSTEM TO I MPROVE
MANUFACTURI NG PROCESSES I N AUTOMOTI VE I NDUSTRY
L., Lobont, C. V., Kifor, C., Oprean, O., Suciu, Lucian Blaga
165
28. A NEW APPROACH OF THE MAI N LANDI NG GEAR EQUATI ONS
Daniel BOSNICEANU
173
29. RESEARCH ON HOLLOW CATHODE EFFECT AND EDGE EFFECT
AVOI DANCE I N PLASMA NI TRI DI NG TREATMENT
Axinte Mihai, Nejneru Carmen, Perju Manuela Cristina, Cimpoesu Nicanor,
Hopulele Ion
181
30. AN ALGORI THM FOR SOLVI NG QUALI TY PROBLEMS, USI NG
QUALI TY TOOLS
Marius Besu, Mircea Ciobanu
185
31. ABOUT NATI ONAL STRATEGY OF ROAD SAFETY I N PRESENT AND
FUTURE
Ciubotariu Danut, Neculaiasa Vasile
191
32. RESEARCHES CONCERNI NG THE CORRELATI ON STRUCTURE OF
PROPERTI ES OF STEELS FOR METAL CONSTRUCTI ONS
STRONGLY REQUESTED
Stela Constantinescu, Maria Vlad
195
33. DETERMI NATI ON OF THE EFFORTS THAT OCCUR I N THE CASE OF
PERFORMI NG I NNER THREADS WI TH FORMI NG ROLLERS
Ion Cristea , Crina Radu
203
34. CONSI DERATI ONS REGARDI NG THE USE OF CHUCK COLLETS I N
MECHANI CAL SYSTEMS
Traian Lucian Severin, Vasile Rata
207
35. THE CORROSI ON OF TUBULAR FURNACES I N PLANTS.
ATMOSPHERI C DI STI LLATI ON OF CRUDE OI L AND I N VACUUM
FOR FUEL OI L
Morosanu Marius, Ovidiu Georgescu
211
36. APPLI CATI ONS OF FULL FACTORI AL DESI GN EXPERI MENTS FOR
LASER WELDI NG
Remus Boboescu
215
37. U.S. QUENCHI NG AND DI MENSI ONAL STABI LI TY I N TI ME OF
100Cr6 STEEL
Nicolai Bancescu, Constantin Dulucheanu, Traian Lucian Severin
223
38. I NFLUENCE OF POLYMER CONCENTRATI ON ON THE
PERMEATI ON PROPERTI ES OF NANOFI LTRATI ON MEMBRANES
Balta Stefan, Bodor Marius, Benea Lidia
227
39. STATI C DEFORMATI ON OF A WORKPI ECE FI XED I N UNI VERSAL
CHUCK AND LI FE CENTRE
Laurentiu Sltineanu, Lucian Tbcaru, Margareta Coteat, Teodor Tilic, Mihai
Boca, Irina Grigoras (Besliu),
233
40. RESEARCH REGARDI NG THE USE OF QUALI TY TECHNI QUES AND
I NSTRUMENTS I N VI EW OF MAI NTAI NI NG AND I MPROVI NG
QUALI TY MANAGEMENT SYSTEMS
Liliana Georgeta Popescu, Mihai Victor Zerbes, Marilena Blaj, Radu Vasile
Pascu, Roland Blaj
239
41. PROPERTI ES OF HARD ALLOYS SI NTERED FROM METALLI C
CARBI DES
Constantinescu Stela
245
42. RESEARCH ON THE CORROSI ON BEHAVI OR OF ZI NC COATI NG BY
MEASUREMENT OF POLARI ZATI ON RESI STANCE
Radu Tamara, Istrate Gina Genoveva
251
43. MATHEMATI C MODEL FOR OPTI MI ZATI ON OF ZI NC-NI CKEL
ALLOY CO-DEPOSI TI ON PROCESS
Violeta VASILACHE, Marius BENTA
257
44. MODULAR DESI GN FOR A FAMI LY OF MECHANI CAL
ANTHROPOMORPHI C POLY-MOBI LE GRI PPERS WI TH 4 FI NGERS
FOR ROBOTS
Ionel Staretu
261
45. THE I NFLUENCE OF ENVI RONMENT TEMPERATURE VARI ATI ON
ON THE STRENGTH CHARACTERISTI CS OF COMPOSI TE
MATERI ALS TYPE ALUCOBOND"
Constantin Dulucheanu, Traian Lucian Severin, Nicolai Bancescu
267
46. MATHEMATI CAL MODELS FOR SMOOTHI NG PROCESS BY
MECHANI CAL SHOCKS OF CYLI NDRI CAL SURFACES
Raluca Tanasa, Traian Gramescu
271
47. EXPERI MENTAL RESEARCH ABOUT I NFLUENCE OF I NCLI NATI ON
ANGLE DI RECTI ON UPON THE SURFACE ROUGHNESS I N BALL
END MI LLI NG OF OLC45 (C45) MATERI AL
Pasca Ioan, Lobontiu Mircea
277
48. PARTI CLES DI MENSI ONAL ANALYSI S AND MI CROSCOPI C
CHARACTERI ZATI ON OF HYDROXYAPATI TE POWDER
Aurora Anca Poinescu, Rodica Mariana Ion
283
49. OVERVI EW OF LATEST MI NERAL CARBONATI ON TECHNI QUES
FOR CARBON DI OXI DE SEQUESTRATI ON
Bodor Marius, Vlad Maria, Balt Stefan
289
50. DEVELOPMENT OF MEANS AND A PLATFORM FOR RESEARCH
Liliana Georgeta Popescu, Mihai Victor Zerbes, Radu Vasile Pascu, Lucian
Lobont, Livia Dana Beju
297
51. I DENTI FI CATI ON AND QUANTI FI CATI ONS OF CHEMI CAL
ELEMENTS AND MI CROSCOPI C CHARACTERI SATI ON OF THE
COPPER BASE ALLOY AT DI FFERENT TEMPERATURE OF MOULDS
Cristina Camber Iordache, Maria Vlad, Simion Balint, Vasile Basliu, Mihaela
Gheorghe
303
52. MANGEMENT OF TECHNI CAL DATA I N THE STUDY OF THE
BRAKI NG SYSTEM OF CARS
Ciubotariu Danut, Neculaiasa Vasile
309
53. COMPOSI TE SHAPE MEMORY ALLOYS FOR METALLI C
STRUCTURES
A. Enache, I. Cimpoesu, S. Stanciu, I. Hopulele, R. M. Florea, N. Cimpoesu
315
54. STUDY OF THERMOMECHANI CAL FATI GUE FOR SHAPE MEMORY
ALLOYS TYPE CuZnAl
Achitei Dragos Cristian, Hopulele Ioan, Perju Manuela Cristina, Mihai Axinte
321
55. STUDY REGARDI NG THE AUDI T OF MANAGEMENT PRI NCI PLES
Costel Mironeasa, Silvia Mironeasa, Georgiana Gabriela Codin
325
56. I NTEGRATI ON OF SI X SI GMA AND LEAN PRODUCTI ON SYSTEM
FOR SERVI CE I NDUSTRY
Militaru Emil
329
57. OPTI MAL PERTURBATI ON FOR MI XED CONVECTI ON HEAT
TRANSFER I N RECTANGULAR CHANNEL
Rachid SEHAQUI
335
58. STUDY ABOUT VI RTUAL AND ACTUAL MANUFACTURI NG
PROCESS WI TH THE ROBOT
Traian Lucian SEVERIN, Romeo IONESCU
341
59. CONSI DERATI ON OF REFLECTI ON COEFFI CI ENT AS A FUNCTI ON
OF FREQUENCY FOR WEDGES OF DI FFERENT MATERI ALS
Liliana Petre-Cainiceanu, Constantin D.Stanescu, Tudor Burlan-Rotar
349
TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies
9
COMPARI SON OF A MACHI NE OF MEASUREMENT
WI THOUT CONTACT AND A CMM
(1)
: OPTI MI ZATI ON OF THE
PROCESS OF METROL OGY.
WOLFF Valery
1
, TRAN DINH Tin
2
1
IUT Lyon 1, University of Lyon, valery.wolff@univ-lyon1.fr
2
University of Lyon 1, tdtinbd@yahoo.com
Abstract: In the field of the control of manufactured products, the process of measurement usually
proceeds on a type of machine (for example CMM) and the totality of measurement is then carried out.
We propose here to study the capacities of 2 types of measuring machines in order to guide the operator
towards an optimized choice of the process of measurement. The taking into account of the capability of
the machine associated with the feasibility (facility of use) of measurement will be also supplemented by
the taking into account the time necessary to realize measurement. A better knowledge of the limits of
each machine will allow an optimization of the process of control.
Keywords: metrology, control, CMM, no-contact measuring system, industrial vision
1 Introduction
The three-dimensional metrology is one of the
main problem for many industries for achieving
dimensional and geometric product control. The
control of technical parts for the plastic industry,
automotive or electronics is more and more using
the optical measuring systems without contact. The
other way is to use a CMM (coordinate measuring
machine) for a more classical control.
We have chosen to work with these two types
of machine to develop our study. We try to define
a methodology to optimize the use of these
different types of measuring machines.
2 Problematic
There are three main steps during the
realization of manufactured products: the stages of
design, manufacture and control.
The designer defines the functional
dimensioning of the parts by using the GPS
standards.
The traditional method, for the control of the
parts in metrology, consists in defining a
measurement plan in adequacy with the measuring
instruments available (We have to separate, for
example, the measurement of the roughness of a
surface, and the 3D control of geometrical
specifications).
We can note that an operator usually prefers to
use only one machine when it is possible. For
example, in the case of the use of a CMM, the
entirety of measurement will be carried out on the
machine (dimension, form, and position
specifications).
The idea we want to develop here, relates to the
possibility of a distributed control: using different
measuring instruments for different specification.
Is it preferable, in term of cost, to carry out the
totality of a control on the same machine? Or is it
better to carry out measurements on various
machines?
This question is the base of the process
planning in manufacturing activity. But, it is a non
studied point for the control of the parts.
3 Study
The context of the study is the industrial vision
systems [2] and we are using the optical methods
for dimensions measuring. In our study, the system
is equipped with an optical device (camera) giving
to the operator an image of the measured part. The
optical machine is a TESA V300 (Figure 1). It is a
manual machine with triple lighting: diascopic,
episcopic and LED. Each axis (X/Y/Z) has a
resolution of 0.05 m.
10
Figure 1 : Tesa Visio 300
3.1 Measurement method
For this study, we use a reference standard ring.
So, we can compare the results of the measurement
to the theoretical value. This is a standard datum
ring with diameter accuracy less than 10
-3
mm.
The TESA Visio provides 3 possibilities:
manual selection of each point, automatic
acquisition for each point, global acquisition of a
profile (automatic sector detection).
The first step of the study was to measure
diameter dispersion by using different ways to
capture a geometrical element on a standard ring.
The obtained results [8] show that the minimum
diameter dispersion is given by the manual method
(Figure 2).
Figure 2 : accuracy of measurement methods
In this article, we have chosen the manual
mode. It is the best way to get accurate results for
our experiments.
By measuring the reference standard ring, we
can obtain several values: the diameter, the XY
position of the center and the circularity default
(form). The figure 3 shows the definition of the
circularity.
Figure 3 : circularity default
The real line (we can call it: real circle) must
be included into an area: the Tolerance Zone. In
the case of the circularity, the tolerance zone is
defined by two circles with the same center, and
with a radius difference lower than the value of the
specification.
If we are using the Tesa Visio machine in a
classical way, we can obtain each circle we
measure with only three data: diameter, form and
position.
But, for the form information, we have only a
global result without any possibility to have a
detailed measure. For example, we choose to
acquire 20 points for a circle. We could expect to
have 20 pairs of XY coordinates. In the Tesa Visio
300, the machine calculates a theoretical best fit
circle associated to the 20 points and gives only the
final result: the default of form (circularity).
Because we want to have a detailed study of the
results of each measurement, we have chosen to
measure all points independently. After that, we
use an Excel Sheet to calculate the same
information we could have obtained with the Tesa
Visio. The method of calculation is presented into
the next paragraph.
3.2 The calculation method
In the mechanical engineering field, the
metrology often needs to calculate mathematical
elements associated with real elements. In the GPS
(geometrical product specification) concept, the
term of skin model represents the real surfaces
(figure 4).
The most common method used to associate a
theoretical element to a skin model is the method
of least squares [1]. This method allows
estimating the numerical values of the form.
15 15,005 15,01 15,015 15,02 15,025 15,03 15,035 15,04
1
2
3
Automatic
Global profile
Manual acquisition
Theoretical
value
Tolerance
tol
11
$
Figure 4 : skin model and mathematical
association
It exists with several variations: its simpler
version is called ordinary least squares (OLS), a
more sophisticated version is called weighted least
squares (WLS), which often performs better than
OLS because it can modulate the importance of
each point in the final solution. Recent variations
of the least square method are alternating least
squares (ALS) and partial least squares (PLS). [1]
The oldest (and still most frequent) use of OLS
is linear regression, which corresponds to the
problem of finding a line (or curve) that best fits a
set of points. In the standard formulation, a set of
N pairs of coordinates {x
i
, y
i
} is used to find a pair
of parameterso, ande (figure 5)
i i i
x y e = o e
Figure 5 : least square method for a line
The least square method defines the parameters
o
and
e
as the values which minimize the sum of
the distances e
i
(hence the name least squares)
between the measurements and the model. What
leads to minimize the function:
2
) (
=
i
i
e c (1)
This is achieved using standard techniques from
calculus, namely the property that a quadratic (i.e.,
with a square) formula reaches its minimum value
when its derivatives is equal to zero. This gives the
following set of equations (called the normal
equations):
0 =
c
c
o
c
and
e
c
c
c
=0 (2)
Solving these 2 equations gives the least square
estimates of o, ande.
The OLS method can be extended to more than
one independent variable (using matrix algebra)
and to non-linear functions.
In this article, we are calculating the default of
form of the datum ring. The result of the
experiment (coordinates of points) to determine the
circle requires to use the polar coordinates.
Figure 6 : least square method for a circle
The figure 6 shows the parameters we use to
determine a theoretical circle associated to a real
set of points.
-
Z
i
: distance is measured from the nominal
circle
-
e
i
: error after calculation
-
A
R: radius variation of the experimental
circle compare with the nominal circle
-
u, v: distance between the two centers
(small displacement)
Based on the figure 6, e
i
is calculated as
follows:
R v u z e
i i i i
A = u u sin cos (3)
After calculation, we replace u, v and AR into
the equation (3). The values (e
max
- e
min
), determine
the default of form of the circle.
3.3 The design of experiment
In the aim to compare the two machines we
have, the experimentation was carried out in a
Nominal element Associatedmodel
Skin model
u
i
12
X Y u 0,00
8,504 -0,038 v 0,00
8,128 2,512 dr 0,002
0,010
eimin 0,005 -
8,066 -2,678 eimax 0,005
X Y u 0,001
8,504 0,137 v 0,001
8,126 2,503 dr 0,002
0,010
eimin 0,005 -
8,068 -2,685 eimax 0,005
similar way on the Tesa Visio 300 and the tri-
mesure CMM.
We used the manual mode for TESA Visio 300
with the setup parameters as follows: down light,
zoom at 50%, and manual acquisition of the points.
The CMM allows to measure circles and to export
all points after the measurement operation.
We measured the datum ring several times. For
each machine, we get the points from five
measures. Each measure, is composed of twenty
points with Xi, Yi coordinates.
The figure 7, shows the position of the centre of
a measured circle obtained by the calculation of the
two parameters u and v. If we want to compare
(with a graph) our results on the circularity default,
we need to have all of the points x
i
,y
i
expressed
into the same datum coordinate system.
Figure 7 : nominal (dot) and optimized (red line)
circles
On the CMM machine, its easy to obtain all
coordinates with a defined origin. The software
allows defining an origin before to save all the
coordinates (x
i
, y
i
). In this case, each circle is used
as the origin of the system. The calculation of the
least square method gives u =v =0 (table 1). The
only one parameter is
A
R.
For TESA V300 machine, we measured 8
points on the circle to determine the center of the
circle. This center is then defined as the origin of
the system. This is necessary before to obtain 20
points by second measurement operation. In this
case we can have small displacement: u and v have
small values (Table 1). This is one of the
assumptions of the least square method.
After performing the experiments, we calculate
with an excel sheet the parameters of each circle
composed of 20 points.
3.4 The statistical approach
We made five measurements for each type of
machine. For more accurate results during the
analysis, we took the average value with two
methods (Figure 8).
The first approach: Each set of 20 points
measurement determine a circle by using the least
squares method. Each measurement allows
obtaining a circularity default. We will then
calculate the average value of the circularity
parameter.
The second approach: After performing the
experiment, we calculate the average (x
i
, y
i
)
coordinate for each point of each measurement.
The least square method is used only after this
step. We determine the parameters of a circle
calculated with the average set of points.
Figure 8 : statistical method
The aim of these two approaches was to
determine the best methodology (in term of time of
measurement and calculation).
4 Results and analysis
The quality of the measured circle (circularity
default) has the same average value for the two
machines we used (0,007 mm on the CMM and
0,010 mm on the Visio300).
Table1: CMM and Visio300 datum system
CMM
Visio 300
13
The repeatability of the measure between each
set of 5 measurements allows us to say that the
result can be studied (figure 9).
Figure 9 : graphical results
The most important result we have noticed
relates to the graphics on figure 10 (case of the
Visio300).
The graphical observation shows that the area
where we can obtain the outer and the inner point
of each measure is at a different place. For the
CMM machine, the results are good at the
positions: 0, 90, 180, 270 and for the TESA
Visio300 machine, the best results are at points 45,
135, 225, and 315.
This can be explained by the way we acquire
the points.
On the CMM machine, the move of the sensor
is controlled by a numerical system with two X
and Y axis. The direction of the move gives more
reliability if we use only one movement at the
same time: X only or Y only.
Figure 10 : best result area for the Visio300
Figure 11 : visio300 acquisition
The combination of the two axis for a
simultaneous move involve a different dispersion
due to the detection of the measured surface by the
sensor.
In the case of the Visio300 machine, it is more
difficult for the operator to be precise when he try
to measure a point with the tangent position than
with a clear intersection of the cursor and the circle
(Figure 11).
5 Conclusion
This article is a first step for a more global
study. It shows that the choice of the machine we
want to use is maybe linked to the capability of the
measurement machine (average value of the
circularity in our case), but also linked to the time
we want to spend for the measurement operation.
The choice of the points (number and position),
and the way to acquire each point is a subject that
will be experimented in a next study.
We will also extend the study by taking into
account the velocity of the machine (in particular
for the CMM machine).
References
[1] Herve Abdi, Least Squares, The University of
Texas at Dallas, Richardson, TX 750830688,
USA.
[2] P. Bourdet, C. Lartigue, A. Contri. Les capteurs
3D tat de lart, problmatique lie la prcision
des systmes de numrisation 3D." 3me congrs
Numrisation 3D/Human Modeling, Paris, 27-28
mai 1998.
[3] J.-L. Charron. Mesure sans contact Mthodes
optiques (partie 1). Technique de lIngnieur,
2004.
[4] J.-L. Charron. Mesure sans contact Mthodes
optiques (partie 2). Technique de lIngnieur,
2004.
small dispersion
large dispersion
CMM
Visio300
0, 90, 180 or 270 45, 135, 225 or 315
14
[5] R. Christoph, H. Neumann. Multisensor
coordinate Metrology: Measurement of form,
size, and location in production and quality
control. Verlag modern industrie, 2004.
[6] D. Gava. Vision conoscopique 3D : calibration et
reconstruction. Thse de doctorat, Universit
Ren Descartes Paris V, 1998.
[7] C. Mehdi-Souzani. Numrisation 3D intelligente
dobjets de formes inconnues base sur des
critres de qualit. Thse de doctorat, Ecole
Normale Suprieure de Cachan, 2006.
[8] Valery Wolff, Arnaud Lefebvre, Dimitri Pachel,
Jasper Thijs, Capability of measuring machine:
Case of an optical measuring machine without
contact, Proceedings of IDMME Virtual
Concept 2010 Bordeaux, France, October 20
22, 2010.
[9] H. Zhao. Multisensor integration and discrete
geometry processing for coordinate metrology.
Thse de doctorat, Ecole Normale Suprieure de
Cachan, 2010.
15
EFFI CI ENT DEVEL OPMENT OF HI GH PERFORMANCE
POLYMER PARTS USI NG ADVANCED SI MULATI ON METHODS
Achim Frick, Timo Dolde, Dorota Sich
HTW Aalen, University of Applied Sciences, Institute Polymer Science and Processing,
Beethovenstr.1 73430 Aalen, Germany, PHONE: +49/7361/576-2169,
e-mail: Achim.Frick@htw-aalen.de
Abstract: The optimization of part development process should lead to an improved time to market
performance, raise cost and resource efficiency and at the same time it has to ensure product quality and
reliability. The quality and performance of polymer parts, in contrast to metal parts, depend not only on
selected material but additionally they are determined by processing conditions during manufacturing e.g.
injection moulding process. Thus, the optimization of development process of polymer parts is of high
importance.
The nowadays available simulation techniques are necessary tools in the virtually based development
process and by the creation of a reliable simulation results the development process can be considerably
accelerate and the competitive quality parts can be in short time realized.
This article shows the advantages of application and combination of the current available simulation
techniques in the development of single stage gear made of plastic. The evaluation and verification of the
simulation results will be presented and discussed.
Keywords: CAE, FEM, Structure Optimization, Filling Simulations, Rapid Prototyping, Virtual
Development
Figure 1: Flow chart of the virtual development process of polymer products
16
1. Introduction
For the development of highly loaded parts, the
selection of a suitable material is of the first
importance while it determines the final product
properties and the whole development process. The
development process depends on the chosen
material class, namely for polymer, in comparison
to metal, it needs special polymer adapted methods
(e.g. polymer design, filling simulations). Thus, the
complexity of development process of plastic parts
is much higher than steel parts. Nevertheless,
polymer materials, due to their lower density in
comparison with metals, show a special potential
in the field of lightweight construction. They offer
also flexibility in the processing and design.
Additionally, polymers offer the chance to
combine different properties (multi-functionality,
multi-component parts, tailor made polymers
e.g. fibre reinforcement), which can lead to some
advantages in comparison to steel.
For a complete use of the potential of polymeric
materials, the virtual material development and
simulation have to be involved in the polymer part
development. The flowchart [Figure 1] shows the
necessary steps in the development process of
polymer parts and the necessary interfaces to
simulation and development of the material.
Consecutively the steps in the flow chart will be
shown in details on the example of a gear wheel
made of engineering thermoplastic -
polyoxymethylene (POM). Additionally, the aim is
to replace the steps prototyping and prototype
evaluation, which include the cost expensive and
time consuming manufacturing and testing of
hardware prototypes, by virtual assessment
methods.
2. Development of a highly loaded polymer
part
According to the flow chart [Figure 1] the first
step in the development process is the creation of
the product idea. Principally in the creation of
product idea following aspects have to be taken
into consideration: function, economy and ecology.
2.1 Product idea
A single-stage spur wheel gear made of
polymeric material is intended for application in
electrically driven vehicles. A single stage gear
needs a high gear transmission ratio between the
pinion and the gear wheel, and therefore the gear
wheel is big-sized. The application of polymeric
materials enables the lightweight constructions
what results in the reduction of both, the rotating
masses and the mass of the car. The lightweight
leads to a higher acceleration, energy saving and
thus less CO
2
emission. Another important aspect
for using a polymer material is that for the large
scale production injection moulding is competitive
method due to the short cycle time and the ready
mould part without post handling. Hence, a
significant cost reduction can be achieved in
comparison to a gear wheel made of steel.
The material of choice for the gear wheel
construction is polyoxymethylene (POM). A gear
wheel made of POM can provide low friction and
low noise emission as well as the damping
behaviour. In order to reduce the mass of vehicle,
the polymer gear wheel shall substitute the safety
clutch between the electrical engine and the gear.
Due to sufficient impact toughness, POM is the
suitable material for this application.
2.2 Concept
To concept the product it is necessary to
analyze the problem and clearly define the
requirements. Thus, all boundary conditions of the
electrical vehicle and the polymer gear are
specified:
- Rear wheel drive
- Max. velocity =120 km/h
- Worst load case: Blocking of the electrical
engine
- Consequence of the blocking: Blocking of the
rear wheels which leads to deceleration of the car
- Dynamic axle load during the deceleration:
300 kg on the rear axis
- Friction between the tiresand the street: =0.8
- Diameter of the wheel: 0.62 m
- Rotational speed of the electrical engine:
3000 rpm
- Material pre-selection: Polyoxymethylene(POM)
- Manufacturing process: Injection moulding
Based on the basic information, the transmission
and the worst case loading with respect to torque
as well as the required gear wheel dimensions can
be calculated as follows.
17
2.2.1 Calculation of the transmission [1]
Circumference of the tirein [m]:
2nr =0.6n =1.884m
Velocity in[
m
s
]:
: = 120
km
h
3.6=33.33
m
s
Rotationsper second:
33.331.884 =17.69r/ s
Rotations per minute:
17.6960 =1061.46rpm
Max. rpm engine: 3000rpm
Selected transmission: i =3:1
2.2.2 Calculation of the maximum torque
Max. friction force:
F
]
=p F
n
=0.8300kg 9.81
m
s
2
=2354N
Max. torque:
I
mux
=F
]
r
dn
=23540.3 =706Nm
Max. torque, gear wheel:
I
mux
=706Nm
Max. torque, pinion:
I
mux
=7063 =235.33Nm
2.2.3 Conception of the gear
The basic concept of the gear is shown inFigure 2.
Dataof the gear:
Transmission: i =3
Pressure angle: o =20
Module: m=12
Pinion: 10 teeth
Gear wheel: 30 teeth
Pitch circle pinion: p
c
=1230 =360
Pitch circle gear wheel: p
c
=1210=120
Centredistance: C =245 mm
Figure 2: Basic concept of the single stage gear
2.2.4 Calculation of tooth root stress
Parameters:
E-Modulus of POM: E =2700 MPa[2]
Temperature: 40C
Load cycles: 10
5
Safety factor: s =1.4
Dynamic strength of the root [3]:
o
mux
=
c
]N
s
=46.4HPo
Circumference force[2]:
F
u
=
21
d
= 3922.22N
Tooth root stress[2]:
o
P
=
P
u
bm
K
A

P

s

c
]N
s
=46.4HPo
width of the tooth: w=40 mm
2.2.5 Calculation of surface pressure [2]
Max. surface pressure: o
HN
=100HPo
Temperature: 40C, in oil
Load cycles: 10
5
Safety factor: s =1.4
o
HNmux
=
c
HN
s
=71.4HPo
Surface pressure:
o
H
=Z
H
Z
L
Z
s
_K
A
z
1
+z
2
z
2

P
u
d
c
b
o
HN
s
=71.4HPo
o
H
=64.26 71.4HPo
2.3 Basic Design
On the basis of the calculated, required
dimensions of the gear wheel, the construction of
the basic design can be done. With the 3D-CAD-
program ProEngineer, 3-D models of the gear
wheel [Figure 3] and the pinion [Figure 4] are cons-
tructed. In addition, the design of the gear shaft
hub is realized. By assembling the single parts, the
fitting of the parts can be proven.
The models will be used in the dynamic impact
calculation with LS-Dyna. For this purpose only
the gear teeth are important and that is why the
first draft of the wheel and the pinion contains no
ribs.
Advantages of 3-D-CAD systems:
With virtual assemblies, the functionality of the
system can be ensured.
18
Iterative course of action:
Depending on the results of the dynamic strength
calculation, the basic design can be improved until
the necessary strength is achieved.
Figure 3: Basic design of the gear wheel
Figure 4: Basic design of the pinion
2.4 Dynamic strength calculation
The FEM-calculation of the gear wheel
impacting the pinion is executed with the explicit
FEM-solver LS-Dyna. For the simulation of an
abruptly stopping engine, the pinion is stopped
after a rotation of 180. Like in the reality, the gear
wheel is loaded with the momentum from the
sliding wheel of 607 Nm. In consequence of the
abrupt stopping, the teeth of the gear are impacting
each other. But also the hub-rim-connections are
highly loaded [Figure 5].
Advantages of dynamic simulations:
Load peaks which results from dynamic loads can
be calculated. In addition, the functionality is
proven by examination of the sliding behaviour of
the teeth.
If necessary, the results of the dynamic FEM-
calculation must be taken into consideration for the
improvement of the basic design. On the example
of the gear, the teeth are strong enough, only the
rim-hub-connection of the pinion requires
improvement [Figure 5].
Figure 5: Dynamic simulation of the gear wheel
impacting the pinion (Ls-Dyna)
2.5 Structure Optimization
The aim of a structure optimization is the
improvement of the properties. These properties
could be e.g. the eigenfrequencies, the stiffness of
the part and its weight.
On the example of the gear, the aim is to reach
the lowest possible weight, under the condition,
that the gear wheel will still resist the worst case
load. The optimization is realized with the
optimization program "Optistruct" from the Altair
Company. It is an implicit solver and the
optimization works with a static FEM-calculation.
The course of action is that the FE-meshed gear
wheel is loaded with the circumference force on
the teeth. During the calculation, the program
identifies elements with low stresses and
deformations. These elements will be deleted and
the calculation runs again. After some steps of
iteration, the program offers a proposition of an
improved shape of the part. In addition, there are
some tools for defining boundary conditions. For
plastic parts, the opportunity to propose shape
without undercut which guarantees good release
properties is really important tool. Another
important tool is the pre-setting of a constant wall
thickness. This new shape can be converted back
into a CAD-part, in this case a ProEngineer file. In
case, that the result, e.g. the weight reduction, is
not satisfying, there is the possibility to switch
19
back to the material simulation and try to get e.g a
material with a higher stiffness.
Advantages of structure optimization:
The structure optimization enables great progresses
in the shape of a part, because it identifies the
unloaded sections and carries out much iteration.
This is much more efficient than a manual
improvement of the shape.
On the one hand, the results from the optimization
can require improvement of the materials, but on
the other hand, the results of the filling study can
require changes of the design, so that the
optimization has to be processed again.
2.6 Final Design
Based on the optimized CAD-part, the final
design is reached by adding the necessary details to
the construction. These details are e.g. radiuses,
chamfers and relief grooves. In the example, this
step is realized with ProEngineer. For the next step
in the development process, the simulation of the
injection moulding process, some versions of the
gear wheel with different sprues should be
designed. For the gear wheel, the best location for
injection is from the hub, maybe with single sprues
or with a film gate.
Figure 6: Final design of the gear wheel
2.7 Filling study
The injection moulding simulation is used for
the assessment of the draft design concerning the
ability to reach good material and morphological
properties. The program which is used is
"Moldflow", a product of the Autodesk Company.
The purpose of Moldflow is a complete virtual
testing of the production process (injection
moulding). It is possible to edit all necessary
parameters for the simulation.
- Selection of the material with all relevant
properties:
Matrix material
Filling material (fibres, spheres)
Viscosity of the material
Thermal properties of the material
- Adjustment of the tool temperature
- Setting of injection and holdingpressure
- Setting of switchover point
- Setting of holding time
The results of the Moldflow simulation are
information about the following parameters:
- Cooling rate of the part
- Building of the morphology
(amorph, semi-crystalline etc.)
- Process time
- Fibre orientations
- Appearance of voids [Figure 7]
- Weldlines [Figure 8]
- Mould shrinkage
- Deformation of the part triggered by different
cooling rates and the resulting differences in
morphology
Figure 7: Voids in the gear wheel
Figure 8: Weld lines in the gear wheel
20
Interpretation of the filling study:
The voids signify a weakening of the material.
With the knowledge about their appearance in the
part, it can be assessed, if they are critical or not.
The weld lines [Figure 8] mean considerable
weakening of the part, particularly in combination
with fibre reinforced materials, because the
orientation of the fibres at the weld lines are
disadvantageous. Furthermore, the weld lines in
the gear wheel are localized in the critical area
around the highly loaded hub.
Advantages of filling study:
The filling study enables the optimization of the
shape, the mould, the material and the process
parameters in a way, which cannot be achieved
easily with normal methods. It means an enormous
cost reduction compared with the manufacturing of
testing-moulds. Moreover, it allows assessing of
the morphological properties of the manufactured
part. Normally, the analysis of the morphological
properties would require the expensive analysis of
hardware prototypes with a computer tomography.
According to the flow chart [Figure 1], the results
of the filling study can require an iterative back
step to the material simulation and the final design
of the part.
4. Outlook
The subject of prospective researches will be
the analysis of the influence of the voids on the
mechanical strength of polymer parts. Therefore,
the real parts with voids will be scanned in the
computer tomography. Using the methods of
reverse engineering, a CAD-model of the part with
the voids will be created. Based on this virtual part,
all the necessary FEM-calculations will be
executed. For the purpose of evaluation and
verification, the simulation result will be compared
with the results of mechanical tests on the physical
parts.
5. References
[1] Grote, K.-H., Feldhuse, J., Dubbel, Springer
Verlag, 2007.
[2] Ehrenstein, G.W., Mit Kunststoffen
konstruieren, Hanser-Verlag, Munich, 2007.
[3] Hermann R., Matek, W., et al.,
Maschinenelemente, Verlag Vieweg und
Teubner, 2007.
21
SUPERIOR VALORISATION OF THE METALLIC SCRAPS
FOR OBTAINING BIMETALLIC STEEL-BRONZE PARTS
Anioara Ciocan
1
, Florentina Potecau
2
1
Dunrea de J os University of Galati, aciocan@ugal.ro
2
Dunrea de J os University of Galati, mihaela_potecasu@yahoo.com
Abstract: The possibility of bimetals producing by direct melting of the bronze machining chips into
steel support is considered. The aim is the manufacturing of the bimetal that is formed by joint of a
bronze layer on steel support. By this way the recycling of the metallic wastes as final products with
value added is possible. In this paper the flow sheet for the bimetal manufacturing by this method was
showed. Also the parameters that influenced the quality of bimetal were analysed. The quality of the
bimetal was analysed in term of contact zone quality. The interface zone of two alloys was investigated
by methods of metallographic and microscopic analyses.
Keywords: bimetals, recycling, bronze chips, steel
1. Introduction
In many industrial applications, the working
surfaces of the parts are simultaneously exposed to
a very different kind of stresses. In these types of
working regime, the bimetallic materials can be
used. These materials are obtained using the
depositing process of one material, as layers with
different thicknesses, on top of another material
considered as a base. These metallic materials, also
named bimetallic can satisfy some requests of the
working conditions impossible to be obtained
using one single metallic material. The covered
steels with non-ferrous alloys layers are an
example and the bimetallic pieces are obtained by
the different methods production. The bimetals are
obtained by metallic layers deposition on the
surface of the support parts. The joining of the
metallic layer on the solid support is determined by
diffusion and thermal processes [1]. The metal
deposition most is makes on one or both surfaces
of the solid support. The thickness of covering
layers is 8 20% of the support alloy. The first
dates about the bimetals are early mentioned. In
the published papers is show that 1858 is the year
when the first patent (USA Patent) for the bimetal
obtaining is mentioned. For the Germany the
researches for the production of the cladded steel
sheets begins in the seventh decade of the XIX
century and a patent was realized. The first special
industrial application for bimetals is dated from
1930. This is from USA and is referred at
utilization of some steel sheets with nickel cladded
for construction of the tank wagon for the chemical
products transports. The production of the great
quantity of the carbon steel sheets cladded with
stainless steel starts around 1938 [2].
One method for obtaining of the bronze layer
added on the steel support for the bimetal
manufacturing is by welding process. For this is
necessary the remelting of the bronze wastes and
moulding in the form of bars. The gas-shielded arc
welding process with wolfram electrode can be
used for welding bronze on the parent steel [3, 4].
During the bimetal obtaining by this method a lot
of problems are associated: the lost of metal is
greater; the costs of labour and energy or materials
are increased as well as the problems of
environment protection are higher.
The present paper presents the researches for
bimetal manufacturing from waste products. To
create a bronze layer on the steel surface, bronze
machining chips were used. In this case the chips
waste of the bronze with aluminium complex
alloyed with iron, nickel, manganese were used.
2. Experimentals and materials
In the experimental work, steel samples as
supports were utilized, Figure 1.
22
Figure 1: Experimental steel sample used that steel
support into experimental
In accordance with the complex work of this
bimetal that be used as bearing parts (ability to
carry heavy loads under friction conditions without
excessive wear and resistance to corrosion) we
used selected bronze chips in accordance with
certain chemical composition, Table 1 [5, 6].
Table 1: Chemical composition of the bronze
Elements,
[wt%]
Al Ni Fe Mn Cu
Range
expected
14-
15
4-
4.5
4-
4.5
0.8-
1.2
bal.
Chips 14.7 4.4 4.1 1.1 bal.
The fluxes with multiple actions were used:
borax as neutral cover fluxes, glassy fluid cover
flux, graphite as reducing flux containing
carbonaceous materials and, copper-phosphorus
metal mixed with bronze chips that self-fluxing on
copper-base metals [7].
A resistance furnace was used for heating the
samples, Figure 2.
Figure 2: Furnace used to melting the bronze chips into
steel support
The quality of bimetal is discussed in term of
contact surface for both alloys (steel and bronze
with nickel, iron and manganese). The structural
aspects associated with heating process were
presented. An Olympus microscope
Metallographic was used. Optical microscopy
included standard methods for preparing samples.
2. Results and discussions
Although the method proposed for bimetal
manufacturing is cheap and simple this involves
any problems. The main are associated with the
quality and the properties of the bronze and the
bronze chips. Other are associated with processes
that accompanying the copper alloys melting and
joining with steel support.
The first condition for a quality process is
referred on the properties of the recycled material.
The success of any melting system depends on the
physical characteristics and nature of the
feedstock, therefore, the integration of durable,
efficient pre-treatment is crucial in achieving full
processing efficiency and hence, a high metal
recovery. This is particularly important when
recycling the machining chips, which have a high
surface area per unit volume. Even small residues
of water/oil soluble fluid will have a significant
impact on metal recovery. Such contaminants
interfere with the bronze and may lessen the joint
strength or cause failure. For this reason the first
step of the flow sheet is the pre-treatment.
For good bronze melting into steel support are
necessary small and uniform chips. The machining
chips have diverse forms. The types of chips are
categorised or subdivided into following
categories, Figure 3.
To obtain the uniform chip size, these are
crushed. The cleaning can be considered to be a
two-stage process. In the first process of the stage
the centrifuge separation to minimized water and
coolants is applied. Also the magnetic separation to
eliminate ferrous parts is sometimes necessary.
Second process of this cleaning stage is based on
thermal processes. The object is to remove the
organic compounds from the surface of the chips
by converting them into a gaseous state. This
process requires a low temperature. Any water that
is present within the chips will be vaporized. Then,
at a higher temperature is removed the carbon-
based deposit that remains on the surface of the
chips.
The second condition for a quality process is
referred on the physical and chemical processes
that are developed in accordance with thermal
conditions that are occurred at the melting of the
copper alloys chips and at the heating steel
support.
23
a.
b.
Figure 3: Chip forming classification: a - not
desirable; b good [8]
Copper is considered a half noble metal but with
a high solubility for oxygen in the liquid state.
Generally an oxygen and hydrogen pick-up can
lead to very negative effects on mechanical and
physical properties of copper and copper alloys.
These gases have a high solubility in liquid copper
alloys that decreases sharply during solidification.
This can lead to bubble formation, i.e. porosity in
the solid material. Oxygen can also form cuprous
oxide (Cu
2
O) above its solubility level that
immediately reacts with the moisture of the air
forming water vapors. Dissolved hydrogen and
oxygen (or Cu
2
O) will react with water under
extreme pressure in the lattice and will form cracks
and lead to embrittlement [9].
Borax as neutral cover fluxes are used to reduce
metal loss by providing a flux cover. This melts at
copper alloy melting temperatures to provide a
fluid slag cover (borax melts at approximately
740
0
C). The glassy fluid cover flux is used also.
This flux agglomerates and absorbs nonmetallic
impurities from the input material (oxides,
machining lubricants, and so on). A reducing flux
containing carbonaceous materials such as graphite
is used. Its principal advantage lies in reducing
oxygen absorption of the copper and reducing melt
loss. Carbonaceous flux material should always be
used with copper alloys to avoid gaseous reactions
with sulfur or hydrogen from contained moisture.
Use the copper-phosphorus metal mixed with
bronze chips that are considered self-fluxing on
copper-base metals is utilized. Some of these
fluxes are mixed with bronze chips and some
covered the input materials after their preliminary
heating. Also the steel samples were heated before
the filling with bronze chips and fluxes. The
amount of flux used was established in accordance
with their effect on the porosity and mechanical
properties of bronze alloy. This is more that 5% of
metal charge.
Figure 3: Preparation of the samples before
introduction in the heating furnace
The heating temperature ensures only the
melting of the bronze chips and also for the bronze
bounding on the surface of the steel. The heating
temperature is very important for obtaining a good
adhesion of the bronze layer on steel surface. This
was established in accordance with the thermal
process that is developed in the samples. The
thermal regime should ensure the melting of the
bronze, the superheating of this melt for
developing the diffusion zone at interface with
steel support. Certainly, this was correlated with
the Cu-Al binary diagram and with the influence of
other elements (iron, nickel and manganese) that
are present in the aluminium bronze composition.
After heating the samples were maintained at
optimum temperature and then were slowly cooled
together with the furnace.
The experiments were carried at different work
temperatures between 1200 and 1300
0
C. The
experimental shows that heating of the samples at
1200
0
C cannot ensure a complete melting of the
bronze chips, Figure 4.
The choice of 1300
0
C as optimum temperature
is confirmed by the experimental samples, Figure
5.
Figure 4: Bronze chips incomplete melted that were
heated at 1200
0
C
24
Figure 5: Good adherence of bronze on the steel
surface for the bimetal sample obtained at 1300
0
C,
slowly cooled with furnace (surface after exposure to
metallographic attack)
At surface of the support steel, the aluminium
bronze with nickel, manganese and iron was
adhered to obtain the bimetal. The appearance of
the joint zone shows that the process was
conducted with optimum parameters. The surface
layer had no defects such as oxide films, or
porosity, Figure 6 and 7.
The bonding takes place during heating and
melting processes. In this way, with participation
of diffusion processes good bimetals can be
manufactured.
Figure 6: Aspect of bimetal at interface (100X without
exposure to metallographic attack)
Figure 7: Microstructure of bronze and steel support
after of exposure to metallographic attack (100X)
4. Conclusions
Usually, the recycling of the chips presents
many difficulties [10, 11]. Firstly, the metal chips
can be classified in the category of hazardous
waste because have residues of water/oil soluble
fluid with significant impact on metal recovery.
The generation of pollutants are considerable. In
the method proposed some of these negative
problems are suppressed. By other hand the metal
losses at remelting are higher. The direct melting
of the bronze chips into steel supports for the
bimetal manufacturing is possible. It is a cheap and
simple solution to transform the machinery wastes
as final products that can satisfy the requirements
in industrial applications. The important factors
that influenced the manufacturing process are the
dimension of the chips and the cleaning. Also, the
heating and the melting parameters in accordance
with the physical and chemical properties of
materials most correlate. Especially, all of these
have particularity in respect to the materials
utilised, the manufacturing processes and the
bimetal applications.
References
[1] Popovits, D., Subu, T., Bimetale, Editura
Facla, Timioara, 1982.
[2] Radu, T., S. Constatinescu, S., Balint, L.
Materiale metalice rezistente la coroziune,
Editura F.M.R., Bucureti, 2004.
[3] Ciocan, A., Bratu, F., Chemical and structural
changes for bimetallic materials obtained by
the welding process, The Annals of Dunarea
de J os University of Galati, no2, 2005, p.15-
22.
[4] A. Ciocan, A., Potecasu, F., Drugescu, E.,
Constantinescu, S., Characterization of the
Diffusion Zone Developed in a Bimetallic
25
Steel-Bronze at Interface, Materials Science
Forum; ISSN 0255 5476, Trans Tech
Publications LTD Switzerland, UK, USA III,
ISSN 0255 5476, 2010, Vols. 636-637, pp
556-563.
[5] Campbell, H., Aluminum bronzes alloys.
Corrosion resistance guide, Publication 80,
Copper Development Association, U.K., pp.1-
27, J uly 1981, www.cda.org.uk.
[6] AluminiumBronze alloys for industry, CDA
Publication No.83, 1986.
[7] Ciocan, A., Scraps Recycling. Technology for
Processing and Valorization of Copper Scraps,
Editura Fundaia tiinific Metalurgia
Romn, Bucureti, 2004; ISBN 973-8151-29-
5.
[8] Standard test method of evaluating machining
performance of ferrous metals using an
automatic screw/bar machine, ASTM E618,
American Society for Testing and Materials,
Philadelphia, USA.
[9] Oprea, Fl., Teoria proceselor metalurgice,
EDP, Bucureti, 1966.
[10] Gronostajski, J ., Chmura, W., Gronostajski, Z.,
Phases created during diffusion bonding of
aluminium and aluminium bronze chips,
J ournal of Achievements in Materials and
Manufacturing Engineering, Volume 19
ISSUE 1 November 2006, p.32-37
[11] Copper Development Association, Cost-
Effective Manufacturing Machining Brass,
Copper and its Alloys, Publication TN 44,
October 1992 Revised for inclusion on CD-
Rom, May 1994
26
27
STUDY FOR THE VARI ATI ON OF HARDNESS AND RESI LI ENCE
CHARACTERI STI CS AFTER HEAT TREATMENT FOR STAI NLESS
STEELS
Manuela Cristina Perju
1
, Drago Cristian Achiei
1
, Carmen Nejneru
1
, Anca Elena
rgeanu
1
, Mihai Axinte
1
, Ion Hopulele
1
1
Technical University Gheorghe Asachi of Iai-Romnia, E-mail address: cryss_ela@yahoo.com
Abstract: The paper presents tests of the stainless steel. This material was subjected to heat
treatments, afterwards conducted impact bend tests and hardness measurements. Using the data was
drawn graphic variation of hardness and resilience depending on the condition of structural material
(hardened, annealed, solution treated and aged).
Keywords: hardness, resilience, heat treatment
1. Introduction
Metallic materials structure researches are
obvious by the fact that structure is very important
determining the properties, the behavior of
materials in service, and their use areas. In modern
metallurgical technology, based on thorough
knowledge of the structure, it can be obtained new
metallic materials, with mechanical and chemical-
physical features precise, requested the use of these
materials [5].
Material behavior is determined by its reaction
on stress. Material property is defined as measured
behavior on given test. By the exterior stresses
there are three properties categories:
- Mechanical properties, which reflects the
material behavior on the mechanical forces.
- Physical properties which measures the
material behavior, under the temperature field,
electrical or magnetic field effect.
- chemical properties which characterizes the
material behavior in an more or less aggressive
environment [6].
The components behavior regarding mechanical
stresses produced by the exterior forces depends on
the certain specific material properties, called
mechanical properties.
Usually, mechanical properties for a metallic
material is determined by specific tests, that means
sample testing (some parts with well defined
configurations and dimensions), in adequate
conditions for the needed properties, [8]. With
mechanical test aid, qualitative data on the material
behavior is obtained. Also with mechanical test aid
we can obtain the physical quantities, called
mechanical characteristics, which can be used as
quantitative parameters for mechanical properties.
2. Experimental results
For the experiment stainless steel was used.
Standard samples were heat treated to modify their
properties. Afterwards the samples were tested by
impact bending and hardness.
2.1 Austenitic stainless steel analysis
Austenitic stainless steel is used in different
domains (fig.1), [10].
Figure 1: The use of austenitic stainless steel.
Energy
production
Transports
Environment
protection
Chemical
industry
Medicin and
pharmacy
Machine
building
Food
Industry
Usestainless
steel
28
Austenitic stainless steel, X5CrNi 18-
10(AISI 304), STAS EN ISO 3651-2 was subject
to chemical composition determination, presented
in table 1.
Table 1: Steel chemical composition, %
C Si Mn Cr Mo Cu Ni
0,025 0,49 1,6 18,38 0.43 0.49 8.12
Foundry Master Spectrometer qualitative
analysis was used.
2.1.1 Heat treatment application
In order to modify the resilience and hardness
properties, a solution treatment was applied [3, 4],
followed by aging heat treatment, using the Uttis
heat treatment furnace.
Solution treatment temperature was 1000C,
heating the samples with the furnace, maintaining
30 minutes, and the cooling made in warm water at
60C. Heating rate was 10C/minute.
Figure 2: Solution treatment and aging treatment
diagramat 400C.
Figure 3: Solution treatment and aging treatment
diagramat 600C.
After the solution treatment, followed the aging
heat treatment, at two temperatures:
- aging heat treatment, at 400C, maintaining 2
hour, followed by air cooling. Aging heat
treatment diagram in figure 2.
- aging heat treatment, at 600C, maintaining 2
hour, followed by air cooling. Aging heat
treatment diagram in figure 3.
2.1.2 Hardness testing
The tests for hardness were conducted on
Wilson Wolpert hardness test instrument. The
Brinell method was used. The pressing force was
187,5 daN, with steel ball 2,5 mm diameter.
The samples hardness was tested before the
heat treatments; data in table 2. For a better
precision on hardness values an arithmetical
average for three tests was calculated.
Table 2: Untreated and heat treated stainless steel
hardness values
Stainless steel
Untreated
Hardness, HB
174,1 177,0 175,3
Average: 175,46
Stainless steel
Solution treatment
Hardness, HB
163,3 164,9 161,5
Average: 163,23
Stainless steel
Aging heat
treatment at 400C
Hardness, HB
208,0 221,6 216,7
Average: 215,43
Stainless steel
Aging heat
treatment at 600C
Hardness, HB
188,2 193.4 191,6
Average: 191,06
175,46
163,23
215,43
191,06
0
50
100
150
200
250
HB
unt r eat ed solut ion
tr eated
aged 400 aged 600
Comparative analysis of hardness
Figure 4: Comparative analysis for stainless steel
hardness untreated and heat treated.
29
Comparative hardness analysis is presented in
figure 4.
2.1.3 Impact bending test
To achieve impact bending tests were used
standardized samples, parallelepiped shape
dimensions: 100x100x550 mm: samples with V
shape indentation.
Testing instrument is Charpy-type pendulum
hammer. The test is conducted as follows:
- positioning the sample on the fixed bearing
device, so as to be hit right behind the indentation;
-hammer rises releasing its action space;
-positioning the pointer at its maximum value;
-fall firing;
-after sample breakdown the hammer is stopped
with a computer command;
-analyzing the samples.
After the impact bending, the untreated
stainless steel and the heat treated one were subject
to macroscopic analyses; this is for highlighting
the fracture character. The fracture character could
be brittle or ductile [1, 7, 9]:
-the brittle section looks crystalline grain and
bright.
-the ductile fracture section (tenacious), looking
fibrous, mat.
Austenitic steel, beside the corrosion resistance,
is characterized by high resilience, tenacity and
strength.
The untreated stainless steel, X5CrNi18-10
(AISI 304), was called group 1 (the group
represents the average from three samples tests, at
impact bending test) (figure 5), and the average for
the two tests gives a high value for resilience
KV=357,81 J .
Figure 5: Macrostructure for untreated stainless steel
tested at impact bending.
Austenitic stainless steels are generally
characterized by low yield and good cold plastic
deformation properties. They present a stronger
stress hardening on cold plastic deformation due to
martensite formation in metastable austenite [7].
The fracture structure is determined by the
existence of hollows or indentations. They
characterize the material ductile fracture and they
have a concave structure resulted from vacancies
fracture, which are initiated and grows during the
plastic deformation process. From a structural
point of view is presented as a solid solution with
precipitates, chemical compounds like complex
carbides, and intermetallic compounds.
According to the corresponding ternary
diagrams Cr carbides are of the form: Cr
3
C
2
, Cr
7
C
3
,
Cr
23
C
6
, Mo compounds like: Fe
2
MoC, Fe
2
Mo,
Mo
2
C, and Mn compounds like: Mn
3
C, Mn
5
C,
Mn
7
Cr
3
.
In figure 6 is presented the group 2 fracture.
Group 2 represents the solution treated material
followed by artificial aging, at 400C. This
material was subject to impact bending test,
obtaining a lower value for the resilience
KV=193,84 J .
Solution treatment generally seeks plasticity
values needed for processing by cold forming and
aims at removing the secondary phase crystal
structure (complex carbides of complex elements).
Resilience decreasing is possible by increasing
hardness from aging treatment.
Figure 6: The heat treated stainless steel
macrostructure solution treatment 400C aging
subjected to resilience test.
The stainless steel samples, solution heat
treated and aged at 600C resulted with a very high
toughness, (group 3), having a maximum resilience
30
at 375 J . Samples were partially fractured (figure
7) when hitting with the force of 300 daN.
The reason for this is given by artificially aging
at 600C, when the temperature exceeds the
solubility variation line, so that the precipitates
remain dissolved in material, even on cooling in
air. The obtained structure will be a supersaturated
solid solution, soft and plastic, with a very high
resilience.
Figure 7: The heat treated stainless steel
macrostructure solution treatment 600C aging
subjected to resilience test.
357,81
193,84
375
0
100
200
300
400
[ J]
lot 1 lot 2 lot 3
Comparative analysis of resilience
Figure 8: Comparative analysis of resilience.
Comparative analysis of resilience is presented
in figure 8.
3. Conclusions
1) The heat treatment importance is given
through their application.
A structural transformation occurs in materials,
which alters the properties and gives them
opportunities for rational use. Metallic materials
subjected to a correct heat treatment have a relative
low cost and higher durability.
2) After the hardness testing it was found that
the hardness value for austenitic stainless steel is
175,46 HB, and after the solution treatment the
hardness value decreased (163,23 HB), corresponding
to this treatment. Because the supersaturated solution
decomposition occurs during maintaining, at the
beginning the hardness grows gradually, as the
precipitation of crystals related to the matrix. Since
the consistency fracture between the precipitates
and the matrix, and the extent of their growth by
coalescence, the hardness decreases.
3) The stainless steel hardness values, artificially
aged, are: at 400C, 215,43 HB, and 600C, 191,6
HB, thus obtaining a major cases difference.
4) The impact bending test, on austenitic
stainless steel gives a resilience value KV=193,84
J for the heat treated group, followed by aged
treatment at 400C. The conclusion was that
samples from this group have a much better value
than other groups from this class of steel.
4. References
[1] u Gheorghe, Proprietile materialelor
metalice, Note de curs, 2007-2008.
[2] Geru N., Teoria structural a proprietilor
metalelor, Editura Didactic i Pedagogic,
Bucuresti, 1980.
[3] Hopulele Ion, Alexandru Ion, Glu Dan-
Gelu, Tratamente termice i termochimice, vol
I, 1983.
[4] Hopulele Ion, Alexandru Ion, Glu Dan-
Gelu, Tratamente termice i termochimice, vol
II, 1984.
[5] Mitilea Ion, Budu Victor, Studiul metalelor,
Editura Falca, Timioara, 1987.
[6] Vermean G., Tratamente termice. ndrumtor,
Editura Dacia, Cluj-Napoca, 1987.
[7] Schumann, H., Metalurgie fizic, Editura
Tehnic, Bucureti, 1962.
[8] Shackelford F. J ., Introduction to materials
science for engineers, Macmillan Publishing
Company, New York, 1991.
[9] Zaharia Luchian, Teoria deformrii plastice,
Editura Gh.Asachi, Iai, 2001.
[10] http://zeus.east.utcluj.ro/~hvermesan/doctorat/
cap1-1.html
31
AN I NTELLI GENT SYSTEM SEARCH FOR I NDOOR COL OR
RECOGNI TI ON: A COMPARI SON
Mrio Antnio Ramalho
1
1
TU Lisbon, I ST, I DMEC,
Av Rovisco Pais, 1049-001 Lisboa, Portugal
mar@dem.ist.utl.pt.
Abstract: This work addresses a map-based method for the localization of mobile robots (wheeled or
legged) in indoor environments, using monocular vision. This method is comprised of four stages: image
acquisition, image feature extraction, image and model feature matching and camera pose computation. A
special attention is given to the matching phase. In particular we research the advantage of including a
retinex pre-processing stage in neural networkthe color classification schemes. Results presented suggests
that this is correct.
Keywords: Image Processing Retinex Neural Networks
1. Introduction
Autonomous guided vehicles are increasingly
common these days and will be more in a near
future. Effectively their use has been extended to
domestic applications, such as cleaning robots or
for assistance of disabled persons, but also in
public buildings, such as museums, serving as
visitor guides. These applications all have in
common the fact that they are indoor
environments. And many more applications can be
envisaged. In these environments, the robot must
be autonomous, being able to fulfill a navigational
task, which involves reaching a goal with no
human intervention, while at the same time
avoiding obstacles and people..
There are mainly two types of AGVs, free-path
or free-ranging (1). Fixed path guidance refers to a
physical guide path (e.g., wire, tape, paint) on the
floor that is used for guidance, and free-ranging
guidance has no physical guide path, thus it is
easier to change the vehicle's path (in software),
but absolute position estimates (optical or laser
guided) are needed to correct dead-reckoning error
(2)
A human do not use neither a path way on the
floor for navigation (eventually for orientation),
neither requires a laser to sense the area in front.
Rather than that it uses vision, from which he can
extract the relevant information for the short path
in front, and memory for getting orientation and
reference points.
We believe that a robot can work with this
information only.
To be autonomous, the robot must be able to
compute and update position and orientation
relative to a fixed global frame.
This work addresses a map-based method for
the localization of mobile robots (wheeled or
legged) in indoor environments, using monocular
vision. This method is comprised of four stages:
image acquisition, image feature extraction, image
and model feature matching and camera pose
computation. A special attention is given to the
matching phase.
Results are presented for a real indoor
environment, suggesting being adequate towards
the envisaged applications
32
2. Previous work
The problem of indoor localization has been
addressed by a variety of techniques. Among them
map based approaches techniques are often used.
Typically there are three main types of map
representations: topological maps, grid maps and
geometric models. The environment model can
either be pre-stored or built online simultaneously.
When performing map-based localization, an
usual approach is to match the local representation
of the environment, with the model map.
The matching phase is a critical step, since it
involves determining the correct correspondence
between image features and model features. To
address the problem of map-based robot
localization using visual sensors, Kosaka et al (3)
proposed a method where robot orientation and
position were updated from initial conditions,
according with navigational commands. The new
robot pose is used to project model features on the
image plane and compare them with the actual
image being viewed. Uncertainty on the new robot
position and orientation is propagated to predict
uncertainty zones in the image, where a search for
corresponding features is required. This method
requires a high sampling rate and not every image
acquisition may provide enough information to
accurately update the robot pose. Talluri et al (2),
introduced the notion of Edge Visibility Regions.
These are used to restrict the matching process to a
small subset of features in the global model.
However it only focuses on horizontal features,
representing rooftop edges.
In this work a map-based localization method
using monocular vision is researched. The focus
will be on the critical matching phase, between
image and model features. The approach
researched is based on color correspondence. The
environment model is constructed offline and pre-
stored. The map is of the geometric type and
contains the main parts of the environment, which
are not supposed to move, such as wall and door
edges.
3. Indoor Map Building
Indoor environments considered in this work
are man-made and semi-structured. These are
typically delimited by flat surfaces, such as walls,
ground and ceiling. The intersections between
these flat surfaces tend to be straight lines, the
edges.
Figure 1 Interior scene with light gradient
The built model contains the main parts of the
environment that are not supposed to move, such
as walls and door edges, and is of the geometric
type, that is, it contains the lengths of each edge
and its position relative to a global fixed frame.
Figure 2 Interior scene model in VRML
4. Color Models
The basic system is color is the RGB (red-
Green and Blue), used by the majority of the
cameras. A more suitable system for computer
processing is the HSV (Hue saturation-
brightness). The later correspond to an axe rotation
from the RGB, making the lightness value to be
defined by the extreme points (black and white).
In computer vision segmentation refers to the
process of partitioning a digital image into multiple
segments (sets of pixels). The goal of segmentation
is to simplify and/or change the representation of
an image into something that is more meaningful
and easier to analyze (5). Image segmentation is
typically used to locate objects and boundaries
(lines, curves, etc) in images.
33
More precisely image segmentation is the
process of assigning a label to every pixel in an
image, such that pixels with the same label share
certain visual characteristics.
In the particular case of this work, image
segmentation refers to the process of identifying
components in the image. And one logical
approach is to search for color areas, which in
interior environments can be many times linked
with a particular type of objects (e.g. doors are
coherent in color).
This approach seems logical; the
implementation is a little more difficult, as the
illumination gradients make the colors vary, and in
particular the triplet of values. This is not evident
all the times due to the quality of the human
visual system. Color constancy mechanisms
compensate this characteristic. Unfortunately these
are not reproducible in computer vision. So we
look for a method to correct the brightness that
allows that similar color appears similar in the
image.
4.1.Retinex model of vision
In the early 70s Land and McCann (5), (6)
introduced the retinex model for the computation
of lightness. Since then, variants have been
presented mainly to improve the computational
efficiency of the model. Among them, Funt et al
(7) (8) have presented a MatLab implementation
which is used through this paper.
Retinex calculations aim to calculate the
sensory response of lightness. According to the
authors, For testing the retinex model it is crucial
that the data be calibrated in the sense that the
image digit values must be a logarithmic function
of scene radiance and they must be represented
with sufficient precision. This is a problem, as we
use, and intend to use standard cameras.
Nevertheless we are investigating the ability to
homo
Retinex algorithms present this characteristic,
and thus Funt et al algorithm was tested, presenting
a poor performance. However during the tests
performed it seems to perform well in some images
where the color didnt have much variation.
A small variation from the strategy proposed
gives much better results. Effectively, retinex
algorithm is being using for reassigning of the
brightness component of the image. We split the
image in the HSV sub-matrices, apply the method
to the v-image (as monochromatic), and
reconstruct the image with the H, S and, corrected
V.
4.2.Artificial Neural Networks
Artificial Neural Networks (ANN) are
structures that mimic the human brain and who
have the capability to learn. They have been used
for edge detection (10)and edge detection
arbitration, among other applications. (10) (11)
Among the several approaches that could be
used in implementing this process, an artificial
neural network seems to be particularly suited to
this job.
Effectively, artificial neural networks can
handle incomplete or corrupted sets of data thus
they can be applied to the recognition of images,
can be used to infer the position of missing edges
or misplaced edges based on the knowledge
applied by the different edge detection techniques
described or color identification. To this purpose a
multi-layer back-propagation neural networks are
being used. These types of networks are capable of
reproducing an input/output relation, learned from
a repetitive exposition to a set of examples. They
also have an inherently parallel structure allowing
for a parallel implementation.
4.3.Results
Examples of the execution of the algorithm
are presents in the following pictures. Figure 3
until 8 are from interior scenes, corridors at
Technical University of Lisbon, under normal
light. Images were taken with a digital camera,
Olympus 750UZ, with 4 Mpixels, and resized.
They are originally jpeg images from the camera.
No further processing was done.
The processed images present more objects,
darker areas are visible, without burn the more
brilliant areas.
Figure 3 Interior Scene - Original
34
Figure 4 Interior Scene Processed
Figure 5Interior Scene - Original
Figure 6 Interior Scene - Processed
ANN are trained with a sample of the colors.
The colors are manually handpicked. The network
is then applied to the original image and the
Retinex processed image
Visually, colors seem to be more
homogeneous, without shifting from the original
hue. Details in dark areas are visible. Tones seem
to be more coherent.
Examples are presented in Figure 7 through
11. The Retinex processed images will present
more colors detected and better detail. By the
opposite original color images are less consistent
in the results.
Figure 7-Corridor after retinex (Figure4) processed
by the ANN
Figure 8- Corridor (Figure4) processed by the ANN
Figure 9 Office corridor after retinex processed by
ANN
Figure 10 Office corridor processed by ANN
35
5. Conclusions and Work under Development
This work presented studies towards a
collaborative frame-work for the localization and
cooperation of mobile robots in indoor
environments. We studied one method for color
identification for a segmentation of interior scenes.
The integration of the system into a robust color
identification algorithm, e.g. Neural Networks, was
studied. It can be seen from the examples that the
preprocessing with the retines algorithm will
increase the ann function. Results are very
promising.
6. Acknowledgments
This work was supported by Fundao para a
Cincia e a Tecnologia, through IDMEC under
LAETA
7. Bibliography
[1]. WikiMehda . http: //www.wikimheda.org
/wiki/AGV. [Online] [Cited: 09 21, 2010.]
http://www.wikimheda.org/wiki/AGV.
[2]. Industrial Trucks. About the MHE
Taxonomy. [Online] [Cited: 09 21, 2010.]
http://www.ise.ncsu.edu/kay/mhetax/TransEq/In
dusTr/index.htm#Light%20load%20AGV.
[3]. Kosaka, A. and Kak, A.C. Fast Vision-
Guided Mobile Robot Navigation Using Model-
Based Reasoning and Prediction of
Uncertainties. Computer Vision, Graphics and
Image Processing - Image Understanding.
1992, Vol. 56, 3, pp. 271-329.
[4]. Talluri, R. and Aggarwal, J .K. Mobile
robot self-location using model-image feature
correspondence. IEEE Trans. Rob. Autom.
2005, Vol. 12, 1.
[5]. Shapiro, L.G. and Stockman, G.C.
Computer Vision. New J ersey : Prentice-Hall,
2001.
[6]. McCann, Edwin Land and J ohn. Method
and Systemfor Image reproduction BAsed on
Significant Visual Boundaries of Original
Subject. 3 533 3601971. US Patent.
[7]. Edwin Land, J ohn McCann. Lightness
and Retinex Theory. J ournal of the Optical
Society ogf America. 1971, Vol. 61, 1.
[8]. Briean Funt, Florian Ciurea, J ohn
McCAnn. Retinex in MatLab.
http://www.cs.sfu.ca/~colour/publications/IST-
2000/retinex_mccann99.m. [Online] [Cited: Sep
15, 2010.]
[9]. Brian Funt, Florian Ciurea, and J ohn
McCann "," Proceedings of the IS&T/SID
Eighth Color I maging Conference: Color
Science, Systems and Applications, 2000, pp
112-121. Retinex in Matlab. Proceedings of
the IS&T/SID Eighth Color Imaging
Conference: Color Science, Systems and
Applications. 2000, pp. 112-121.
[10].Integrating Parallel edge detection Schemes
using Neural Network Arbitration. Ramalho,
M and Curtis, KM. s.l. : ICSPAT'93, USA,
1993.
[11].Edge detection using neural network
arbitration;. Ramalho, Mrio and Curtis, K.
s.l. : IEE Conf. Pub. 1995, 514 (1995),, 1995.
[12].Ramalho, M and Curtis, KM. Neural
Network Arbitration of Edge Maps. [book
auth.] A, J ane, M and Marini, D. (Eds),
Gloria. Transputer Applications and
Systems'94;. s.l. : IOS Press, Netherlands,
1994.
[13].Borenstein, J ., Everett, H.R. and Feng, L.
Navigating Mobile Robots: Systems and
Techniques. s.l. : Wellesley, Mass.:AK Peters,
1996.
[14].Borenstein, J ., Everett, H.R. and Feng, L.
Navigating Mobile Robots: Systems and
Techniques. Wellesley, Mass : AK Peters,
1996.
[15].Aider, O.A., Hoppenot, P. and Colle, E. A
model-based method for indoor mobile robot
localization using monocular vision and
straight-line correspondences. Robotics and
Autonomous Systems. 2005, Vol. 52, 2, pp.
229-246.
[16].Kim, J .S. and Kweon, I .S. Estimating
Intrinsic Parameters of Cameras using Two
Arbitrary Rectangles. Proc. 18th Int. Conf.
Pattern Recognition. 2006, Vol. 1, pp. 707-
710.
[17].Simsarian, K.T., Olson, T.J . and
Nandhakumar, N. View invariant regions
and mobile robot self-localization. IEEE
Trans. Rob. Autom. 1996, Vol. 12, 5, pp. 810-
815.
[18].Valente, Filipe. Movement and Image
Prediction. Lisbon : MSc Thesis, IST,TU
Lisbon, 2009.
36
37
MATERI AL SI MUL ATI ON OF POLYMERS CONTAI NI NG
SPHERI CAL STRUCTURES
Achim Frick
1
, Melissa Matzen
2
, Timo Dolde
3
1,3
Aalen University, Inst. Polymer Science and Processing iPSP, Aalen (Germany),
2
University of Bremen, Bremen (Germany);
Email: Achim.Frick@htw-aalen.de
Abstract: One of the keys for successful polymer product development is proper material selection. It
is recommended to apply tailor sized polymers in order to realize best product performance with respect
to the requirements. Material simulation is a possibility to predict material data from modified polymers,
e.g. particle filled polymer, for strength calculation purpose, which is eminent part of the product design
process. A method to simulate stress-strain behaviour of polymers containing spherical structures is
shown for TPU material.
Keywords: Polymer, Morphology, Material Simulation, Computer Aided Product Development
1. General I ntroduction
A successful polymer product requires a proper
material selection. Therefore it is necessary to
integrate both a material based strength calculation
and structural optimization as the two main
elements of a computer aided product developing
process, Figure1. Material simulation is a possible
tool to provide material data from any polymer
modification for strength calculation.
Figure 1: Product development process fromthe idea to
the final production
Material properties of thermoplastic polymers
depend on the one hand on their thermal history,
due to the processing conditions during moulding,
as well as on a possible tailor sizing by adding e.g.
fillers or reinforcement to the polymer matrix.
To study the possible consequences of such
variations for the strength behaviour of the
polymer a material simulation is recommended. It
provides a fast and helpful method for advanced
product development.
2 I nvestigated Material
Segmented thermoplastic polyurethanes TPU
belong to the class of TPE and they show soft and
rubberlike behaviour at RT. Nevertheless they
perform comparably strong, are oil resistant and
behave superior abrasion resistant, which
recommend them for e.g. sealing applications. But
the properties of TPU are highly influenced by
their morphology developed during the melt
processing and subsequent solidification process.
The stress-strain behaviour of TPU is observed to
be strongly dependent on the present crystalline
domain structure, which shows to be spherical
shaped. Also for semi crystalline polymers the
spherical crystalline fraction determines the end
use properties of the material. Concerning that a
material simulation can provide to better material
understanding basically and can also suggest
optimal morphological structure for the polymeric
material in order to perform best in
application.Thermoplastics can further contain
fillers such as glass balls or minerals particles to
enhance the strength and stiffness of the basic
38
material. Material simulation gives a possibility to
predict the material behaviour of modified
polymers for strength calculation purpose. It can
provide information for improved material
selection and qualification for a demanded product
under development.
Material simulation requires material
parameters for input in order to model a polymeric
material as a part of product development process.
There are two possibilities to obtain the
parameters. Either a theoretical based material
system can be generated or the material data are
received by analyzing a real polymer sample.
3 Modelling
3.1 GENERATED STRUCTURE
To study theoretically on a possible influence
of spherical inclusions in a polymeric material
matrix a determined distribution of the spheres
with a chosen radius, Figure 2, is input into the
software, e.g. Digimat from Xstream-
Engineering, and a representative volume element
(RVE) is generated. The software then provides
the stress-strain curve of modeled material. It is a
way to simulate material behavior depending on its
spherical inclusion fraction and inclusion size and
distribution, where the inclusions may represent
spherulitical crystalline structure in semi
crystalline polymers or any filler with an aspect
ratio near 1.
Figure 2: Virtual material simulation based on
generated structure
A virtual section view of this RVE can be taken
to visualize and study the morphology and it also
can be compared with a micrograph taken from a
real cross section of a material sample micro
structure, in order to validate how realistic the
modeled morphology is.
3.2 DERIVED STRUCTURE
To approach on the evidence of a certain
morphological structure on the mechanical
behavior of a polymeric material sample it is
helpful to model this real existing material
morphology by means of material simulation in a
reverse procedure.
As a starting point a light microscopy (LM) or
even transmission electron beam microscopy
(TEM) micrograph is observed and the distribution
of the detected spherical inclusions, e.g. spherulits,
is identified, seeFigure 3.
Figure 3: Real material modelling based on the derived
structure
Using these information received a RVE is
generate and is basis for further material
simulation procedure, in order to establish the
stress-strain behaviour of the modelled material
structure.
The challenge hereby is on one hand to analyse
properly the inspected morphological structure in
the 2D micrograph and on the other hand to
reconstruct the size and dispersion of the observed
inclusions in the regarded volume.
3.2.1 Image analysis of micrographs
A computer assisted image analysis of
morphological micrographs seems problematic
regarding a proper detection and reproducible
results.
As a basic requirement for computer assisted
determining the diameter of a spherical structure
embedded in a grey environment it needs to have a
sufficient contrast between the both structures next
each other. The grey values of each of them must
differ strongly in their brightness. Otherwise an
automatic detection is not possible, whether due to
insufficient contrast of the structure intended to get
measured or due to random variation of the
39
structure in geometry and deviation from the
predefined geometry.
If an automatic evaluation of the micrograph of
a material morphology is not feasible and the
inspection is done visually then there is no
necessity for any contrast optimization and
correction of inhomogeneous grey values
distribution in the image due to the varying section
thicknesses of the previous inspected microtome
section under transmission light microscope. To
avoid inconsistencies in the measuring results by
visual evaluation of morphologies it is recommend
involving always the same person for inspection
task.
3.2.2 Relation between measured circle radius
and the radius of the sphere
When observing on a cross section of a material
filled with spherical inclusions randomly
dispersed, and even all the inclusions have same
diameter, the circle shaped cross sectional areas of
those inclusions show different diameters.
The mathematical calculation of the diameter
distribution of spherical inclusions in a matrix
material from the estimated radius distribution of
the measured circle shaped areas is mathematically
an ill-defined problem [3], which needs to be
solved. It is quite ambitious and requires expressed
mathematical skills. It is like tomato salad
problem and to estimate the diameter of a tomato
by measuring the diameter of its slices. Due to the
difficulties to solve this mentioned problem some
approximation methods have been suggested.
Wicksell [4] was the first who described in 1925 in
a treatise titled corpuscle problem the relation
between a measured radius of an arbitrary section
of a sphere and its real radius. Bach [5] adopted in
1958 the problem and gave refined solution
regarding the size distribution of inspected
spherical sections in translucent slices of finite
thickness.
The approach of Bach [5] uses the second
Volterra integral to describe the relationship
between the size distribution of spherical radius
and circle radius, see equation 1,
g(R
CIrcIc
) =2 F R
CIrcIc
JR
CIrcIc
_
0(R
Sphcrc
)JR
Sphcrc
R
Sphcrc
2
-R
CIrcIc
2
R
CcIe
+o 0(R
CIrcIc
)
(1)
It means:
G(R
Sphcrc
) Sphere radius distribution
g(R
CIrcIc
) Circle radius distribution
F Cross section area
R
Sphcrc
Sphere radius
R
CIrcIc
Circle radius
Sample thickness
The calculation method according to Fullman
[6] is simple compared to the suggested method of
Bach. Fullman assumed that the sphere radius is
identical to the average radius of Gaussian
distributed circle radii, Figure 4, which are detect.
Figure 4: left: the sphere size distribution (distinct);
right: detected cross section radius distribution
According to that assumption, Fullman
concluded the sphere radius R
xphere
to be equal
the circle radius R
c|rc|e
.
R
ccIc
= R
sphcc
(2)
Mathematically, the Bach method is relatively
elaborative, but it provides excellent results. The
solution of Fullman, which bases on the simplest
approach to determine the relationship between the
radius of a two dimensional circular section and
the three dimensional sphere radius, provides at
minor calculation effort also useable results.
3.2.3 Relation between measured cross sectional
area and the volume
To simulate a filled material it is essential to
know on its composition (filler volume fraction) as
well as on the inclusion size and distribution. So
the content volume of filler has to be estimated
from a micrograph of the morphological structure
of a sample, Figure 5. It is necessary to estimate
the 3D volume of an inclusion from the diameter
of its two-dimensional (2D) cross sections. Figure
5 depicts the problem.
According to [7] the sum of circular areas
_ F
n
=1
is equivalent to the total volume of the
spheres _ I
n
=1
. Thus the following equation 3
expressing the relationship between the area
fraction and volume fraction is valid.
Sphere radius [m] Circle radius [m]
40
_ F
n
=1
F
=
_ I
n
=1
I
(3)
Figure 5: Left: micrograph of the cross sectional area
of samples with inclusions; right: reconstructed volume
(F
1
: circular area, V
1
: sphere volume)
By means of image analysis the cross section of
a material containing spherical inclusions both, the
sphere radius distribution of the inclusions, as well
as the volume fraction of the spheres, can be
estimated.
4. TPU Material Modelling
Material modelling requires first a
mathematical equation which is adapting to the
specific material deformation behaviour and thus
fits to its stress-strain behaviour best. For instance
Mooney-Rivlin equation can be used for modelling
elastomeric materials such as TPU. The
mathematical concept behind this model is a hyper
elastic approach.
In case of segmented thermoplastic
polyurethanes TPU the macromolecular chains
exist from sequentially linked hard and soft
segments. The hard segments can build rigid
domains in the soft segment matrix, depending on
a possible phase separation between both, which is
ruled by the melt processing conditions during
moulding. If the material will be processed at high
melt temperature the hard segments does not
segregate and shape distinct domains, the TPU
therefore shows almost transparent. In such case
there is no expressed crystalline super structure in
the transmission light micrograph visible. The
material is considered to be amorphous in its
morphological structure.
Figure 6 shows a measured stress-strain curve
of an amorphous TPU sample (nominally 94 Shore
A durometer hardness) processed at a melt
temperature of 250C and its best fit curve
considering Mooney-Rivlin equation. The material
model describes the measured data quite well,
especially at higher strain more than 40%.
Based on the fit curve, considered as the
material model representing the amorphous matrix
material, it is now possible to study the
consequences of a certain crystalline volume
fraction in the amorphous matrix on the
mechanical properties of the considered TPU by
simulation. The material simulation of partially
crystalline TPU has to approach the stress-strain
behaviour of a RVE containing an amount of hard
and spherical shaped inclusions, which represents
the crystalline structure.
Figure 6: Modelling stress-strain behaviour of
amorphous TPU by Mooney-Rivlin approach (fit curve
calculated with software ABAQUS 6.7)
The further simulation of partially crystalline
material requires the mechanical properties, e.g.
the Youngs modulus of the pure hard segments of
TPU. Unfortunately the mechanical properties of
the pure hard segments or even of the hard
domains cannot be determined easily because it is
almost impossible to approach those separately and
simply. Therefore it needs to find approximate
material data for crystalline structure.
It is well known that a homo-polymeric
polyoxymethylene (POM-H) represents a strongly
crystalline polymer and has crystalline content up
to 90% [8, 9, 10]. Thus in a first order approach
the TPU hard segments are considered to have a
similar stiffness than POM-H. The Youngs
modulus is taken to be 3000 MPa for the hard
segments.
4.1. SIMULATION OF CRYSTALLINE
CONTENT
In order to simulate the influence of hard
segment domains in amorphous TPU material
representative volume elements (RVEs) of TPU
were generated containing spherical shaped
0 20 40 60 80 100 120 140 160
0
2
4
6
8
10
12
14
S
t
r
e
s
s

[
M
P
a
]
Strain [%]
Material model Mooney-Rivlin
tested material
1
1
F
1
V
1
1
1
1
41
crystalline fraction content. The RVE generation
procedure was as described before and RVE
created using Digimat software. The diameter was
considered 10 m for simulating crystalline
structure when the volume fraction content varies
from 0 to 30 vol%.
Figure 7 shows the influence of growing
content of crystalline structure on the stress-strain
behaviour of TPU, the material becomes
significantly stiffer and its strength rises. It
becomes obvious that the TPU must perform in
service application quite different depending on its
present crystalline domain fraction. The stress-
strain behaviour of this material depends hardly on
the melt processing conditions during
manufacturing process.
Figure 7: Simulated stress-strain curves for TPU with
different vol.% crystalline domain content (d: sphere
diameter of inclusion)
Figure 8: Simulated stress-strain curves for TPU,
amorphous state and with 20 vol.% crystalline domain
content (mono and bi-sized inclusions, d: sphere
diameter of inclusion 1, D: sphere diameter of
inclusion 2)
Assuming the crystalline domain fraction in
total to be 20 vol% and it exists from 2 different
diameter sized inclusions (10m and 30 m) it is
calculate by simulation, that the bi-modal
crystallinity gains better material strength
behaviour than a mono-modal one, Figure 7
5. Summary
Material simulations allow to model polymeric
material with incorporated inclusions, such as
crystalline domains or fillers, in order to study the
resulting mechanical behaviour. The data received
can be used for within the product development
process and also for material science purposes.
The simulation possibility was approached on
segmented TPU material.
References
[1] A. Frick, M. Mikoszek: Predispositions to failure
of Polyurethane Sealings: Relation between
Morphology and Visco-elastic Properties, p. 413-
414, full paper: S1205_A0273 in Gomes, J .F.S.;
Meguid, S.A. (editors): Integrity, Reliability and
Failure: challenges and Opportunities, Edition
INEGI, 2009 (ISBN 978-972-8826-22-2)
[2] A. Frick, M. Mikoszek, M. Kaiser: Qualitt von
TPU-Dichtungen: ber den Einfluss der
Morphologie auf die Gebrauchseigenschaften in
Dichtungstechnik J ahrbuch 2010, S. 356-372
[3] A. Louis: Inverse und schlecht gestellte Probleme;
Stuttgart; Teubner Verlag; 1989.
[4] S. Wicksell: The corpuscle problem, A
mathematical study of a biometric problem;
Biometrika; 1925, A-17; S. 8499
[5] G. Bach: ber die Grenverteilung von
Kugelschnitten in durchsichtigen Schnitten
endlicher Dicke; Diss; Universitt Gieen; 1959;
S. 2ff.
[6] R. Fullman: Measurement of approximately
cylindrical Particles in Opaque Samples;
Transactions AIME;1953; A-197; S. 1267ff.
[7] W.Rostoker, J .Dvorak: Interpreatation of
metallographic structures; Academic Press Inc.;
1965; S. 198.
[8] G. Becker, L. Bottenbruch, D. Braun: Technische
Thermoplaste: Polycarbonate, Polyacetate,
Polyester, Celluloseester; Mnchen; Hanser
Verlag; 1992; S. 333.
[9] [Online] Wapedia: 5.07.2010;
http://wapedia.mobi/de/Kristallisation_(Polymer).
[10] G. Ehrenstein, G. Riedel, P. Trawiel: Praxis der
thermischen Analyse von Kunststoffen; Mnchen;
Hanser Verlag; 2003; S. 74.
0 20 40 60 80 100 120 140
0
5
10
15
20
25
30
S
t
r
e
s
s
[
M
P
a
]
Strain [%]
d=10; 0 Vol-%
d=10; 5 Vol-%
d=10; 10Vol-%
d=10; 15Vol-%
d=10; 20Vol-%
d=10; 30Vol-%
d=10; 40Vol-%
0 20 40 60 80 100 120 140
0
5
10
15
20
25
30
S
t
r
e
s
s
[
M
P
a
]
Strain [%]
d=10; 0 Vol-%
d=10; 20Vol-%
d=10; 15Vol-%; D=30; 5 Vol-%
42
43
MONTE CARLO PI VOTAL CONFI DENCE BOUNDS FOR WEI BULL
ANALYSI S, WI TH I MPLEMENTATI ONS I N R
J urgen Symynck
1
, Filip De Bal
2
1
KAHOSint-Lieven,jurgen.symynck@kahosl.be
2
KAHOSint-Lieven, filip.debal@kahosl.be
Abstract: To gain expert insight in the inner workings and pitfalls ofcommercial lifetime analysis
software, the authors created an open source alternative with asubset of analysis tools and made it freely
available as a package for the R statistical software, called the Weibull Toolkit for R. This articlefocuses
on creating pivotal confidence bounds using Monte Carlo simulation for B-lives from a Weibull model.
These bounds were suggested by Lawless (ref. [1]) and are recommended for small sample sizes of
(nearly) complete databy Abernethy and Fulton(ref. [2], [3]). Fully functional and annotated R code is
presented, derived from the toolkits codebase. For the latest version of this document, check ref. [4].
For more in-depth treatment of the Weibull analysis with R, check ref. [5].
Keywords: Weibull, R, pivotals, Monte Carlo, confidence bounds.
1 Introduction
1.1 The FATI MATProject
FATIMAT (FATigue In MATerials) (ref. [6])is
a completed PWO project (ref. [7]) supported by
KAHOSint-Lieven (ref. [8]) that investigated
cheaper alternatives to servo hydraulic dynamic
testing machines, like servo pneumatic (ref. [9])
and electromechanical spindle actuator machines.
Products can be tested in various ways: for
example an engine mount for fixing car engines to
a chassiscan be tested by cyclically compressing it
between 10 [mm] and 20 [mm] on a Zwick/Roell
EZ020 high speed electromechanical spindle
actuator ata load frequency of 3 [Hz]. After some
time, the specimen will show signs of wear-out (in
this case oil leakage) that are detected by the
machine. The test is then halted and the failure
time is recorded. This test is repeated (under
identical loads and circumstances!) for a number of
specimens.
One of the subprojects of FATIMAT was to
analyze these failure data using the Weibull
lifetime distribution for drawing conclusions on the
general reliability of the specimens.Commercial
packages such as ReliasoftsWeibull++(ref. [10])
and superSMITHWeibull (ref. [3]) were evaluated,
but for gaining expert insight the authors created
their own software: the WeibullToolkit (ref. [11])
for the statistical software R (ref. [12]).
1.2 The Two-Parameter WeibullDistribution
The two-parameter Weibull model is widely
used in the field of reliability engineering, because
it allows useful analysis with extremely small
sample sizes (two failures or less). The Weibull
model covers many other distributions (either
exactly or approximately) like the(log-)normal and
exponential.Also, an informative graphical plot
can be created that helps to convey the analysis
results to non-statisticians like engineers and their
managers. In many cases, a two-parameter
Weibull model is sufficient for accurately
describing failure data. Its cumulative distribution
function (c.d.f.) is given by:
F(t) = 1- e
-[
t
q
(1)
The shape parameter indicates the type of
failure: <1 is a sign of infant mortality while>1
is a sign of wear out failures. = 1 suggests a
constant failure rate, not related to lifetime. The
scale parameter , also called the characteristic
liferepresents the age at which 63.2 [%] of the
specimen have failed.
1.3 The R Statistical Programming Language
(from the Rhomepage, ref. [12]:) R is a
language and environment for statistical computing
and graphics. It is a GNU project which is similar
44
to the S language and environment which was
developed at Bell Laboratories (formerly AT&T,
now Lucent Technologies) [] R provides a wide
variety of statistical [...] and graphical techniques,
and is highly extensible.
R can be downloaded with no cost from its
homepage and can be installed on most computers.
It is essentially a console-like application where
the user enters commands at the prompt. Multiple
commands can be scripted and stored in a plain
text file, making complete applications possible.
Some graphical interfaces for R, and some
dedicated R code editors like Tinn-R (ref. [13]) are
available, making R easier to use.
1.4 The WeibullToolkit in R
The WeibullToolkit is a package for the R
statistical programming language. It was initially
conceived for doing analysis on the very simple
reliability problem of complete, uncensored data,
but its feature set and capabilities are continuously
updated.In this paper, its code is demonstrated
using simplified toolkit functions: just copy and
paste the code in an open R console.More details
on the Weibull toolkit can be found under ref. [5]
and [14].
The WeibullToolkit is hosted online at
Sourceforge (ref. [11]) and can be downloaded as
an installable package. The source code can be
browsed online (ref. [15]) or downloaded using the
Git fast version control system (ref. [16]).
2 Entering Data in the R Console
Let us assume that eight engine mounts were
tested under the previously mentioned conditions
(cyclic compressive sine wave load alterations
between 10 [mm] and 20 [mm], 3 [Hz], same
environmental temperature and humidity, ).The
recorded observations are 149971, 70808, 133518,
145658, 175701, 50960, 126606 and 82329 cycles.
All specimen were tested to failure (in this case:
failure by oil leakage) creating a so-called
complete dataset. The data is imported in R by
opening the R console and entering the following
code at the prompt:
### Ent er i ng f ai l ur e dat a ###
d <- dat a. f r ame(
t i me=c( 149971, 70808, 133518,
145658, 175701, 50960,
126606, 82329) , event =1)
This creates a table-like variable namedd of the
dataframeclass, with two columns: d$t i me and
d$event .The latter represents the event at the
time of observation: here, all the specimens failed,
corresponding with event 1 or died. J ust typed
followed by <ENTER>at the prompt to display
the contents of d(note that help is available for
dat a. f r ame and all other functions by typing
?dat a. f r ame followed by <ENTER> at the
prompt).
2.1 Censored Data
In many cases, mechanical-dynamical testing
means only a few specimens are tested (3-8) until
failure, after which the failure time is recorded.
This creates a complete or uncensored dataset.
Sometimes an upper test duration limit is enforced,
after which unfailed specimen are taken from the
machine and labelled as survived. This kind of
censored data is called right censored data or
suspended data.
The WeibullToolkit can handle this type of
censored data, but for simplicity, this (important)
subject is not treated in this article and the
presented code does not support it; check out ref.
[5] for a more detailed treatment.
3 CreatingaWeibullPlot
The goal of the Weibull plot is to provide a
useful 2D representation of the observations. On
the vertical axis, the unreliability of the
specimens is found while on the horizontal axis
one finds the observation time. The double
logarithmic scale of the Weibull plot's vertical axis
together with the horizontal axis logarithmic scale
makes the Weibullc.d.f. (eq. [1])appear as a
straight line (fig. [1]).
3.1 Median Rank Regression
To create the straight line representing the
samples and , a vertical plotting position is
assigned to the ordered observation times
(ranking).These ranksare equivalent to the
unreliability of the specimens population. Then,
by means of simple linear least-square regression
on transformations of the observations and median
ranks, the Weibull parameter estimates are
calculated from the data points.Several methods
exist for the rank assignment (Hazens, Bernards,
mean ranks) but here the inverse of incomplete
beta function is used, which is considered the best
method.
45
For small and moderate sample sizes (2-100)
with few or no censoring it is always best practice
to determinethe estimates usingmedian rank
regression (MRR) in favour of other techniques
like maximum likelihood estimation (MLE) (ref.
[2]).Continue by running the following code to
assign median ranks to the previously entered
failure data (note that for this uncensored data the
d$event column is not used or needed,
simplifying the ranking process):
### Ranki ng f ai l ur e dat a ###
mr ank. ob <- f unct i on( j , f ) {
r <- qbet a( 0. 5, j ,
f - j +1) ; r }
mr ank. dat a <- f unct i on( d) {
n <- nr ow( d)
d$r ank <- r ank( d$t i me, " f i r st " )
d <- d[ or der ( d$r ank) , ]
d$mr ank <-
mr ank. ob( d$r ank, n) ; d}
pr i nt ( d <- mr ank. dat a( d) )
The ranking is done by theqbet a( ) function,
the inverse of the incomplete beta function and part
of the standard libraries of R. The dataframed now
holds the median ranks for each observation in the
d$mr ank column, as shown below.
t i me event r ank mr ank
6 50960 1 1 0. 08299596
2 70808 1 2 0. 20113119
8 82329 1 3 0. 32051897
7 126606 1 4 0. 44015520
3 133518 1 5 0. 55984480
4 145658 1 6 0. 67948103
1 149971 1 7 0. 79886881
5 175701 1 8 0. 91700404
>
After ranking the observations, a line can be
fitted through the data points, calculating the
Weibull parameter estimates.Continue by running
the following code to fit a line through the data
points:
### Medi an Rank Regr essi on ###
F0i nv<- f unct i on( p) {
l og( l og( 1/ ( 1 - p) ) ) }
f wb <-
l m( l og( d$t i me) ~F0i nv( d$mr ank) , d)
bet a <- 1/ coef ( f wb) [ 2]
et a <- exp( coef ( f wb) [ 1] )
pr i nt ( past e( "bet a=", si gni f ( bet a) ) )
pr i nt ( past e( "et a =", si gni f ( et a) ) )
Thel m( ) function fits a straight line by fitting
l og( d$t i me) on a transformation of the median
ranks.The F0i nv( ) transformation, together with
the logarithmic transformation of the observations
allow treating the soughtWeibullc.d.f. as a straight
line, making a line fit possible (the transformations
can be derived from eq. [1] and will result in
F(t) = [ log
c
(t) - [ log
c
(p), which is of the
same structure as the standard line equation
y = o x +b).
Note that here it is the observations that are
fitted on the ranks (X-on-Y regression), and not
vice-versa (Y-on-X, as is standard in most
implementations of least square fitting).It is good
practice to fit the variable with the most variability
(the observations) on those with less variability:
the values of the median ranks are exactly defined
and do not depend on the actual values of the
observations, only on their positions in the list of
ordered lifetimes, and the total number of
observations!
Cal l :
. . .
[ 1] "bet a= 2. 58128"
[ 1] "et a = 132512"
>
3.2 The WeibullPlot
The WeibullToolkit automates all of the above
steps. For convenience however,the stand-alone
code for a simplified version of the Weibull plot is
presented here. Because R lacks native support for
automatically transforming the vertical axis to the
double logarithmic scale, this must be done
manually with theF0i nv( ) function (logarithmic
transformations are supported by the l og=x
parameter).It will be used frequently with most
graphical functions in this article.Continue by
running the following code to produce aWeibull
plot for the given example:
### Si mpl i f i ed Wei bul l pl ot ###
opt i ons( sci pen=10) # no sci ent i f i c
# not at i on
pl ot ( y=F0i nv( d$mr ank) , x=d$t i me,
l og="x", axes=F, l wd=2, cex=1. 2,
mai n=" Engi ne mount cycl i c t est ",
xl i m=c( 5000, 500000) ,
yl i m=F0i nv( c( 0. 01, 0. 99) ) ,
xl ab=" t i me", yl ab=" Unr el . [ %] ")
46
cur ve( add=TRUE, l wd=2,
bet a*l og( x) - bet a*l og( et a) )
ygr i d <-
c( 1: 9, seq( 10, 90, 10) , 91: 99)
axi s( 2, F0i nv( ygr i d/ 100) ,
ygr i d, l wd=2) ; axi s( 1, l wd=2)
abl i ne( v=c( 5000, seq( 1e4, 1e5, 1e4) ,
seq( 1e5, 5e5, 1e5) ) ,
h=F0i nv( ygr i d/ 100) )
abl i ne( l t y=2,
h=F0i nv( ETA <- 1- exp( - 1) ) )
Figure 1:SimplifiedWeibull plot, showing the median
rank regression (MRR) fitted line, representing the
Weibull parameters and .
After plotting the median ranks versus the
observations, the cur ve( ) function draws a
straight line withbet a =[
`
andet a =p . Note that
an identical line can be plotted by substituting
cur ve( . . . ) with the following code, where in
the argument of F0i nv( ) one recognises the
Weibullc.d.f. from eq. [1].
### Al t er nat i ve cur ve( ) met hod ###
cur ve( add=TRUE, l wd=2,
F0i nv( 1- exp( - ( x/ et a) ^bet a) ) )
Grid lines are plotted at horizontal and vertical
plot position (vertical positions are listed in
ygr i d) . The 63.2 [%] (dashed) unreliability line
is also plotted: the point of its intersection with the
fitted line provides the characteristic life of the
distribution, .
4 B-life
When the Weibull plot is created, predictions of
the failure behaviour of the specimenspopulation
can be made.
The B-life is the age at which a certain
percentage of the investigated populationis
expected to fail (based on the analyzed sample!).
For example, the B10-life for the population of
engine mounts can be read from the plot and
isapprox. 55500 cycles (the number following the
capital B indicates the unreliability percentage).
The B1-life is approx. 22300 cycles. J ust draw a
horizontal line at the 10 [%]unreliabilityand find
the intersection with the fitted (straight) line.Then,
read the age from the horizontal axis.Run the lower
code for marking the B10 life:
### Mar ki ng t he B10 B- l i f e ###
abl i ne( l wd=2, l t y=2,
v=55500, h=F0i nv( 0. 1) )
In R, the B-lives are very easily calculated by
means of the qwei bul l ( p, bet a, et a)
function which is part of the R standard libraries.It
calculates the p
th
quantile from the Weibull model
described by [
`
and p . Executing
qwei bul l ( c( 0. 1, 0. 01) , bet a, et a) calcul
ates the B10 and B1 life, respectively:
[ 1] 55415. 93 22299. 16
>
5 Confidence in Predicted B-lives
5.1 Definition of Confidence I nterval
It is evident that B-livesbased on Weibull
parameter estimates from large sample sizes are to
be taken more seriously than those based on two or
three observations; small samples contain very
limited lifetime information.To get an indication of
the confidence that one should have on an
estimated B-life, a confidence interval can be
calculated. A90 [%]confidence interval for a B-
life has the following meaning:
When the B-life would be estimated over and
over again from samples similar to the original
one, then the real, unknown B-life of the
specimens population will, with a 90 [%]
frequency, be situated inside the 90 [%] confidence
interval.
A 90 [%] confidence interval is limited by two
bounds, who can also be described as at the
lower side thelower confidence bound of a
47
95 [%] confidence interval with no upper limit
(100 [%]), and at the upper side the upper
confidence bound of a 95 [%] confidence interval
with no lower limit (0 [%]).
This means that the real, unknown B-life
exceeds the lifetime at the lower confidence
boundwith a 95 [%] frequency. Also: the real,
unknown B-life will,with a 5 [%] frequency, be
smaller than the lifetime at the lower confidence
bound. The latter shows that there is still a small
chance that the actual B-life turns out to be worse
than anticipated by the confidence bounds!
5.2 Usage
According to the method demonstrated in this
article the real, unknown B10 life for the provided
example lies between approx.24500 and 87000
cycles, with 90 [%] confidence level. The B1-life
(estimated at 22300 cycles) liesin a big 4700
50200 cycles interval, with 90 [%] confidence
level.
Generally, confidence levels of 90 [%] are used
and silently implied. In the automotive industry, a
confidence level of 50 [%] is often used, meaning
that the confidence concept isnot used at all but B-
lives are read straight from the fitted line. In this
case they rely on large sample sizes for reliable
predictions. Much higher confidence levels are
used (90 [%] to 99.9 [%] or more) in the medical
and aircraft industry.
The presented confidence limits widen
dramatically when the sample size decreases.
Although still valid, it is clear that the interval will
be so wide that it could be of little practical use:
increasing the sample size is the only remedy
against wide confidence intervals.
5.3 Calculating Confidence Bounds
To obtain a lower bound for a 90 [%]
confidence bound for a B10 life, one must find the
5th percentile of the distribution of the B-lives at
the given (10 [%]) uncertainty. This distribution is
not easily described and it is, except for complete
or Type II censored data (where all the unfailed
specimens are censored at the time of the latest
failure) difficult or impossible to calculate
analytically (ref. [1]). Some methods approximate
this distribution by a well known distribution or
apply transformations to the B-lives.
Confidence bounds come in a great variety:
beta-binomial bounds, Fishers matrixbounds,
likelihood ratio bounds, Monte Carlo pivotal
bounds and bootstrap bounds are the most popular.
The first are very easily calculated (even by hand)
but cannot be extrapolated to lower unreliabilitys;
precisely where the reliability engineer needs
them. Abernethy (ref. [2]) concludes that Monte
Carlo pivotal bounds are best practice when
median rank regression is used, given that the
needed computer power is available. For larger
sample sizes (>400), likelihood ratio bounds for
MLE or MLE-RBA based Weibull estimates are
faster to calculate. This article describes how to
calculate the Monte Carlo pivotal bounds.
5.4 Straightforward Monte Carlo Bounds
It would make sense to calculate the confidence
interval of a B10 life using the following Monte
Carlo based method:
Calculate the parameter estimates [
`
and p
from the original sampleI = t
1
,t
2
, ,t
n
with
sample sizen, as explained earlier.
Create a lot (2000 <=R<=5000)of synthetic
samples (samen) by randomly generating
synthetic observations based on [
`
and p ,
resulting in I
1.R
-
= (t
1
-
,,t
n
-
)
1
, (t
1
-
,,t
n
-
)
2
,
,(t
1
-
,,t
n
-
)
R
. In R this is accomplished by
repeating the
r wei bul l ( n, bet a, et a) functionRtimes.
Find the synthetic Weibull parameters
estimates[
`
-
and p
-
for all the synthetic
samples, [
`
1.R
-
= [
`
1
-
, [
`
2
-
, ,[
`
R
-
andp
1.R
-
=
p
1
-
, p
2
-
, ,p
R
-
.
Calculate the B10 life for each synthetic
Weibull plot, resulting in B10
1..R
-
= B10
1
-
,
B10
2
-
,, B10
R
-
by repeating
qwei bul l ( 0. 1, bet a, et a) for all [
`
-
and
p
-
.
Calculate the 5
th
and 95
th
percentile (for a
90 [%] confidence interval) from the empirical
distribution of B10
1.R
-
using the
quant i l e( ) function. These values
represent the lower and upper confidence
bounds.
The above straightforward method turns out to
provide too optimistic (narrow) intervals,
especially for small sample sizes (n<20). The
reason is that the distribution of the synthetic B10
lives (from which one calculates the 5
th
and 95
th
percentiles) depends too much on the real,
unknown values of and . The above
calculations are based on estimates[
`
-
and p
-
, who
themselves become less accurate with smaller
sample sizes. The deviations of these estimates
48
from theirreal, unknown values should be reflected
in the range of the confidence intervals, but they
are not. When bigger sample sizes are used, the
estimates are closer to their true values, making
theseconfidence bounds more realistic, however.
5.5 Monte Carlo Pivotal Confidence Bounds
A better approach is to derive the 5
th
and 95
th
percentiles from a pivotal quantity: this is a
quantity that does not depend on the underlying
true and unknown distribution parameters and .
Lawless (ref. [1]) supplies a pivotal quantity for
determining B-life distributions:
Z
p
=
u-
p
b
`
(2)
Where u = log
c
(p ), y
p
= log
c
(t
p
) and
b
`
= 1/ [
`
. t
p
isthe time where the unreliability is
p = F(t; [, p) from eq. [1]; the Weibullc.d.f. for
the real, unknown parameters and . Otherwise
said, t
p
= F
-1
(p; [, p) which is the inverse of the
Weibull c.d.f..Practically,t
p
is the B(100*p)-life
where a confidence bound is to be calculated for.
By definition, Z
P
does not depend on unknown
parameters, meaning that its distribution formstays
the same regardless of the actual values of and.
The pivotal can therefore be calculated by
arbitrary setting = = 1 (or u = 0 andb = 1) for
further calculations.
Recall: t
p
= F
-1
(p; [, p) (3)
Let w
p
= log
c
(t
p
), and with = =1:
w
p
= log
c
(F
-1
(p;1,1)) = log
c
[F
0
-1
(p) (4)
Eq. [4] can now be calculated for it represents the
inverse of the standard Weibullc.d.f. (eq. [1]) and
results in:
w
p
= log
c
[log
c
[
1
1-
(5)
To determine the empirical distribution of Z
p
,
Monte Carlo methods are used again:
Create a lot (2000<=R <=5000) of synthetic
samples (same size as the original sample) by
randomly generating synthetic observations
from the standard Weibull model ( = =1).
Find the synthetic standard Weibull parameter
estimatesu
0
-
= log
c
(p
0
-
)and b
0
-
= 1/ [
0
-
for all the synthetic samples,
givingu
0
(1.R)
-
andb
0
(1.R)
-
.
Calculate all pivotalsz
p
-
using a combination
of eq. [2] and [5]:
Z
p
= z
p
-
(1.R)
=
u
0
(1.R)
-
-w
p
b
0
(1.R)
-
(6)
The empirical distribution of these pivotals can
now be calculated. To calculate the bounds for a
90 [%] confidence interval for the B10 life, on
takes the q=5
th
and q=95
th
percentile of Z
(p=0.10)
and calculates the corresponding B-life with a
rearranged eq. [2]:
y
(p,q)
= log
c
(p ) - Z
(p,q)
[
`
/ (7)
t
(p,q)
= e
(p,q)
(8)
Continue with the next code block to load the
pi vot al s( ) function:
### Pi vot al s f unct i on def . ###
MC<- f unct i on( n) {
st d <- dat a. f r ame(
t i me=r wei bul l ( n, 1, 1) ,
event =1)
st d <- mr ank. dat a( st d)
f wb <- l m( l og( st d$t i me) ~
F0i nv( st d$mr ank) , st d)
c( u0_hat =coef ( f wb) [ 1] ,
b0_hat =coef ( f wb) [ 2] ) }
pi vot al s<- f unct i on( r , R, unr el ) {
pi v <- as. dat a. f r ame(
t ( r epl i cat e( R, MC( r ) ) ) )
wp <- F0i nv( unr el )
Zp <- f unct i on( wp) {
( ( pi v$u0_hat - wp) / pi v$b0_hat ) }
pi v <- cbi nd( pi v, sappl y( wp, Zp) )
names( pi v) <- c( " u0_hat ",
" b0_hat " , si gni f ( unr el ) )
pi v}
The MC( ) function calculates and returns one
pair of u
0
-
andb
0
-
by fitting a Weibull line trough
a synthetic sample from the standard Weibull
distribution. The pi vot al s( ) function repeats
MC( ) for R times (usually R=2000),applies eq.
[6] with sappl y( ) on all u
0
-
andb
0
-
and binds
the pivotalsto the pi vdataframein an extra
column.
49
Note that wp<- F0i nv( unr el ) gives the
same results as wp<-
l og( qwei bul l ( unr el , 1, 1) ) , showing the
relation with the standard Weibull model and the
calculation of B-lives.Proceed by calculating the
pivotals for determining the B10 lifesconfidence
interval, based on the exemplary samplewith size
n=8:
###B10 pi vot al s cal cul at i on ###
pi v <- pi vot al s( 8, 2000, 0. 1)
head( pi v, 3) ; t ai l ( pi v, 3)
In the third column of the pi vdataframe, one
can find the 2000 pivotal quantities (only a few are
displayed with the head( ) and t ai l ( )
functions).Continue by calculating the pivotal
percentiles and the actual confidence intervals:
### B10 pi vot al conf . bounds ###
t p_l ow <- exp( l og( et a) - quant i l e(
pi v[ , ' 0. 1' ] , 0. 95) / bet a)
t p_upp <- exp( l og( et a) - quant i l e(
pi v[ , ' 0. 1' ] , 0. 05) / bet a)
pr i nt ( past e(
" B10 90[ %] CI = ( " ,
si gni f ( t p_l ow) , " , " ,
si gni f ( t p_upp) , " ) " ) )
poi nt s( pch=3, l wd=2, cex=2,
x=c( t p_l ow, t p_upp) ,
y=r ep( F0i nv( 0. 1) , 2) )
Eq. [8] is executed twice for the 5
th
and 95
th
percentile. Finally, the pivotal confidence interval
for the B10 life is calculated and displayed, and
added to the plot:
[ 1] "B10 90[ %] CI =
( 23931. 3 , 86688. 7 ) "
>
Note that recalculated bounds will never be
exactly identical because of the Monte Carlo
variability. For lower B-lives like B1, many
replications like R=5000 or more may be
necessary. If repeatable results are needed, one
can runset . seed( 1) before all code to lock the
random number seed value.
Repeating the above steps for a number of
unreliability levels and connecting the points so
that a curve emerges is the next logical step; these
are the pivotal confidence bounds as plotted in the
WeibullToolkit:
### Pl ot pi vot al conf . bounds ###
pi vot al _CB<- f unct i on( pi v, CL) {
unr el <- as. numer i c( names(
pi v[ , c( - 1, - 2) ] ) )
r df <- dat a. f r ame( unr el =unr el ,
r ow. names=unr el )
Tp <- f unct i on( Zp, conf ) {
exp( l og( et a) - quant i l e(
Zp, conf ) / bet a) }
r df <- cbi nd( r df ,
l ower =sappl y( pi v[ , c( - 1, - 2) ] ,
Tp, 1- ( 1- CL) / 2) ,
upper =sappl y( pi v[ , c( - 1, - 2) ] ,
Tp, ( 1- CL) / 2) )
r df }
pi v<- pi vot al s( 8, 2000, ygr i d/ 100)
CB <- pi vot al _CB( pi v, 0. 90)
l i nes( l wd=2,
CB$l ower , F0i nv( CB$unr el ) )
l i nes( l wd=2,
CB$upper , F0i nv( CB$unr el ) )
Figure 2:Weibull plot with Monte Carlo pivotal
confidence bounds for a 90 [%] confidence level.
The B10 confidence intervals can now be read
from the graph in the same way as the regular B10
lives.
6 Conclusion
The article demonstrates the inner workings of
the Weibull Toolkit for R, an open source
reliability and lifetime data analysis package.
After demonstrating R code for calculating the
Weibull parameters for complete lifetime data
50
using the median rank regression method,Monte
Carlo pivotal confidence bounds are discussed and
calculated. Where appropriate, alternative
calculation methods are briefly mentioned and
discussed. The calculation of the pivotal quantities
is clarified in detail and applied in functional R
code, culminating in a simplified version of the
Weibull plot as generated by the Weibull toolkit.
7 References
[1] J erald F. Lawless, Statistical Models and
Methods for Lifetime Data, 2
nd
edition,
Wiley-Interscience, Hoboken N.J ., 2003.
[2] Robert B. Abernethy, The New Weibull
Handbook, 5
th
Edition, 2008.
[3] http://www.barringer1.com/wins.htm
[4] http://mechanics.kahosl.be/fatimat/index.php
/downloads-and-information/40/198
[5] J urgenSymynck, Filip De Bal, The Weibull
model in fatigue and reliability analysis: an
introduction & implementation in R, KAHO
Sint- Lieven, 2011
http://mechanics.kahosl.be/fatimat/index.php
[6] http://mechanics.kahosl.be/fatimat
[7] http://www.kahosl.be/site/index.php?p=/nl/p
age/153/
[8] http://www.kahosl.be
[9] http://www.youtube.com/watch?v=g2IczVy
uQkQ
[10] http://www.reliasoft.com
[11] http://sourceforge.net/projects/weibulltoolkit/
[12] http://www.r-project.org/
[13] http://www.sciviews.org/Tinn-R/
[14] J urgenSymynck, Filip De Bal, Weibull
analysis using R, in a nutshell,New
Technologies and Products in Machine
Manufacturing Technology, Stefan cel Mare
University. of Suceava, 2010
http://mechanics.kahosl.be/fatimat/index.php
[15] http://weibulltoolkit.git.sourceforge.net/git/g
itweb-index.cgi
[16] http://git-scm.com/
http://code.google.com/p/msysgit/
8 Further reading
Chi-Chao Lui, A Comparison Between The
Weibull And Lognormal Models Used To
Analyse Reliability Data, dissertation from
University of Nottingham, 1997.
William Q. Meeker and Luis A. Escobar,
Statistical Methods for Reliability Data,
Wiley-Interscience, New York, 1998.
http://www.weibull.com/
Reliasofts Reliability subjects website
http://cran.r-
project.org/web/packages/survival/
The homepage of the survival R package for
general survival and reliability analysis.
http://cran.r-project.org/web/packages/boot/
The homepage of the boot R package, for
bootstrapping functions.
51
MODERN GREEK LANGUAGE FREQUENCY COUNTS FOR
TEXT ENTRY DEVI CES
Ilias Sarafis
1
, Anastasios Markoulidis
2
1
Technological Educational Institute of Kavala Agios Loukas, GR65404, Kavala, GREECE
isarafis@teikav.edu.gr
2
Technological Educational Institute of Kavala Agios Loukas, GR65404, Kavala, GREECE.
Abstract: In this paper we present and analyze tabulated case-sensitive single letter, digram and trigram
frequency counts from a 450 Kword modern Greek corpus. Special characters, such as space, comma,
paragraph change and full stop, used for text input, are also included. Data analysis revealed moderate
correlation between English and Greek frequency counts, justifying a different approach when studying
Greek keyboards. A keystroke-level model using the digram frequency data and applying Fitts law was
employed to predict entry speed rates for Qwerty, Fitaly, and Telephone virtual keyboards, both in
English and Greek layouts. The results of this paper could be used when designing new software or
hardware keyboard layouts and in order to improve their efficiency and text entry speed.
Keywords: digram frequencies; linguistic data; modern Greek; text input devices; virtual keyboard
1. Introduction
Moving towards mobile computing and
portable devices has brought to the fore, the need
for smaller, more versatile, and space efficient text
entry methods. The use of T9 (http://www.t9.com,
Tegic Communications 1998, J ames and Reischel
2001) on mobile phones for SMS editing is a
characteristic example of a smart and effective way
to enter text on a keyboard initially designed for
numeric input. Only 9 alphanumeric keys (12
including space and special characters), and a
disambiguation algorithm and lexicon small
enough to fit into mobile phones memory, provide
millions of users the ability to communicate
quickly using text messages.
Apart from hardware mini-keyboards (Green et
al. 2004, Clarkson et al. 2005) there are also many
virtual (soft-type) keyboards, which have been
studied and applied on mobile devices (Masui
1998, Goldstein et al. 1999, Zhai et al. 2005).
Research for better keyboard layouts and
disambiguation techniques is an on-going process,
emphasizing both in better modelling algorithms
and in designing smaller and more user-friendly
layouts with higher text entry rates.
Much effort was made to develop theoretical
models, in order to evaluate the performance of
such virtual keyboards, in terms of maximum
words-per-minute. These models mainly use the
well-known Fitts' law (Fitts 1954) and keystroke-
level timing together with linguistic data (digram
frequencies) (Mackenzie et al. 1999, Soukoreff
2002).
Concerning research on Greek language
keyboards, extended literature review has revealed
two matters: first, no published research exists
about optimized and efficient Greek language
keyboard layouts, not even actual performance
measurements of existing Greek keyboards.
Second, there are no linguistic data published for
Greek digram and trigram frequencies, but only for
letter frequencies, despite the fact that there is a
large (47 Mwords) corpus available.
Thus, our research effort was focused, on one
hand to compute and publish Greek digram and
trigram frequencies and on the other hand to apply
a keystroke-level model using Fitts law to a
number of Greek keyboards. One of the modelled
keyboards is a novel design, based on FITALY
English layout (http://fitaly.com/, Textware
Solutions 1996) and applying the philosophy of
minimizing key distances between the most
frequent digrams.
52
2. Modern greek linguistic data
A large (47 Mwords) corpus is available for
Greek language, by the Institute of Language and
Speech Processing (ILSP - website
http://hnc.ilsp.gr). This corpus is known as
"Hellenic National Corpus" (HNC) and it is not
publicly available as a whole, rather through a web
interface for searching words and lemmas.
Mikros et al. (2005) have published quantitative
characteristics based on HNC corpus, mainly for
linguistic purposes. Thus, their data do not include
frequencies for special characters, such as space,
comma, paragraph change and full stop, used for
text input. Moreover, they do not provide data for
digram and trigram frequencies.
In our study, we decided to use corpora from
two nationwide Greek newspapers, "Ta Nea" and
"Macedonia". The corpora are available through
the Center for Greek Language (http://www.greek-
language.gr/). We have selected the "articles"
section and created a combined corpus having
453.407 words, 2.574.047 characters without
spaces, 3.018.705 characters with spaces and 8.807
paragraphs.
Although our corpus is much smaller than
HNC, it is certainly larger than the limited 20
Kwords corpus (Mayzner and Tresselt 1965), used
by keystroke-level models for English language
(Mackenzie et al. 1999), thus the results can be
considered reliable enough. Further on, we
computed Pearson product-moment correlation
coefficients in order to find out if there is a
significant statistical difference between HNC and
our corpus. We compared single letter frequencies
and ranking from the two corpora and we found to
be highly correlated (p = 0.9996 & p=0.9991).
Therefore, one could assume that the digram
frequencies and other linguistic data we present
here would be reliable enough, at least for
keystroke-level modelling.
2.1. Word-length and single letter frequencies
Word length distribution for the studied corpus
is presented in Figure 1. The average word length
of 5.551 characters is very close to that reported
for HNC (5.33 total and 5.56 for "miscellaneous"
media type). We have also tested three, randomly
selected subsets of the 450k corpus, one with 20
Kwords, one with 70 Kwords, and one with 180
Kwords. The word length for these subsets showed
no significant differences (5,529, 5,552 and 5,557
characters respectively). The word length
distribution shown in Figure 1, is also consistent
with the one reported for HNC.
Single letter frequencies for lower-case, upper-
case, accented and non-accented letters, are shown
in Table 1. Notice that in Greek language there are
two types of accent: (a) the "tonos" ( ' ) that can go
over all the seven vowels and (b) the "dialitika" (
) that go over "" or "" in certain cases. In order to
type an accented letter on a keyboard, one should
first press the accent key, prior to the letter key.
There are also cases where "tonos" and "dialitika"
are combined, e.g. and . In these cases also, one
should first press the proper accent key, prior to
the letter key.
Upper case letters relative frequency is 2.85%,
the seven vowels count for 54.83% of the corpus
and accented vowels are the 14.83% of the total
letters. Correlation of the single letter frequencies
list between the 450k corpus and the subsets of
20k, 70k and 180k corpora, was high (0.9995
0.9997 and 0.9999 respectively).
By comparing upper and lower-case counts we
found a Pearson correlation of 0.7697 which is
moderate but higher than the reported for English
language (0.6337) (J ones and Mewhort 2004).
Correlation between upper-and lower-case vowels
is 0.7953 (accented 0.4969 & non-accented
0.8249), between upper- and lower-case
consonants 0.7350 and between accented and non-
accented vowels 0.8564.
In order to compare the Greek with the
English frequency counts, we used the standard
QWERTY character mapping, as in Table 2. We
have used English linguistic data published by
J ones and Mewhort (2004). We can observe that
there are 13 totally identical letters and 7 letters that
are similar (Delta, Fi, Gamma, Lamda, Pi, Ro,
Sigma).
Using this mapping, correlation between the two
lists of single letter frequency counts gives a
p=0.8157. This result reveals the need for a
different approach when designing and evaluating
Greek keyboard layouts. Any proposal for
optimized Greek layouts could not be based on
work done for English ones.
After reordering the mapping to achieve a higher
correlation (p=0.9703), we concluded to the results
shown in Table 3, referred as "Greek reordered"
layout. It is obvious that such a mapping would be
totally confusing for the user of hardware
keyboards, where two letters (English and Greek)
are printed on the same key
53
2.2 Special Characters
Special characters are of significant importance
for keyboard design and efficiency. As shown
in Table 4, special characters, count for a total
of 17.98% of the corpus. Space only, counts for
14.74% of the total characters, in comparison to
18.65% reported for English language
(Mackenzie et al. 1999). Other significant
special characters are the full-stop (0.75%),
comma (0.91%) and paragraph change (0.29%).
54
2.3 Digram and Trigram Frequencies
Digram and trigram frequencies are useful in many
cases. From anagram solutions (Dominowski and
Duncan 1964) and word recognition to keyboard
modeling and even for developing disambiguation
algorithms (Maragoudakis et al. 2002). An
Appendix that contains case sensitive bigram
frequencies, including space, full stop and comma,
together with a list of the 500 most common
trigrams (case independent), is available by the
authors, upon request.
We have tested the correlation between the
following lists: (a) single letter counts, (b) letter as
predecessor counts and (c) letter as successor
counts. We found high consistency between them
that was over 0.9997 in the worst case. There is
moderate agreement between single letter counts
and the frequency counts of a letter at the
beginning or the end of the word (0.6992 & 0.6838
respectively).
High inconsistencies are observed between the lists
of (a) the first letter and (b) the last letter of a word
(p=0.2377), even after ignoring the final , which
can only be found at the end of the word. This
could affect the positioning of the "space" key,
since space is the dominant character, both before
(97.9%) and after a word (89.8%). Other characters
following the end of a word are paragraph change
(i.e. the "Enter" key), full stop and comma.
We have tested the correlation between case
sensitive bigram counts and compared them with
the values reported for English language (J ones
and Mewhort 2004). Correlation between
upper/upper and upper/lower digrams exhibits poor
consistency at 0,4316 (English 0,372). Correlation
between upper/upper and lower/lower is also low
at 0,3526 (English 0,665) and there is moderate
agreement between upper/lower and lower/lower
digrams at 0,5945 (English 0,587). The total
number of instances for each bigram type is:
upper/upper: 9.919 occurrences upper/lower:
52.152 occurrences lower/lower: 1.962.339
occurrences lower/upper: 28 occurrences.
3. EVALUATI ON OF GREEK SOFT
KEYBOARDS
Wehave used a keystroke-level model proposed
by Mackenzie et al. (1999), to predict the
performance of three different Greek keyboard
layouts and compare them to their English
equivalents. The model is valid for one-hand typing
on soft (virtual) keyboards.
At first, a spatial layout of each keyboard is
drawn and the x-y coordinates of the keys are
derived. Hence, the distance between every two
keys in the bigram table can be calculated:
where i is the predecessor and j is the successor
in a bigram.
By applying Fitts' law (Fitts 1954), which models
the time a human needs to move from one key to the
other, we could calculate the movement time for
each possible bigram
where RT is the reaction time a human needs to
locate a specific key on the keyboard, BW is the
human motor bandwidth as defined by Fitts, Aij is
the distance between the centers of the two keys, as
in (1) and Wj is the width or the size of the target
key.
Based upon the above prediction, we can then
summate the movement time of all bigram
combinations, after weighting them with the
probability Pij of occurrence of a digram with first
letter i and second letter j (Mackenzie et al. 1999):
Equation (3) gives the mean movement time in
seconds, between two characters. To convert this
to words-per-minute (WPM) we use:
where CPW is the mean characters per word for
a specific language. In order to have comparative
results, we used CPW=5 as proposed by Mackenzie
et al. (1999), both for English and Greek language.
In our calculations we have used the dimensions
of a common hardware keyboard with key width W
=18mm but with zero clearance between keys. We
also focused on maximum entry speed prediction,
thus we used RT=0 (expert user) and the value
BW=4,9 proposed by Mackenzie et al. (1999). For
the key repeat time (bigrams with same letters), we
used the 0,153 seconds value, also proposed by
Mackenzie et al. (1999). English bigram data were
55
downloaded from
http://www.dynamicnetservices.com/~will/academi
c/bit95.tables.html.
All the above model equations and bigram tables
were entered into a spreadsheet and produced the
final wpm results for each of the following
keyboard layouts.
3.1 QWERTY layouts
We used typical QWERTY layouts with
English-to-Greek mapping as in Table 2. For
English language we used case independent
bigram tables (a 27X27 matrix including the
space character). For the Greek layout, we applied
in addition, an analytical model with case-
sensitive bigrams with accented letters, including
full-stop and comma, as well as the space
character. The analytical bigram table is given in
the Appendix and leads to a 66X66 matrix with
bigram probabilities.
In the analytical model, we assumed that if the
second bigram letter is in upper case (e.g. at the
begin of a sentence), a shift key should be pressed
before the letter key, thus increasing the travel
time. In the same manner, the accent key should
be pressed before any accented letter that is
successor in the bigram.
The Greek QWERTY layout for the analytical
model is shown in Figure 2. In order to calculate
the distance between the space key and the letter
keys, we divided the space key length to three
equal parts and used the distance from the part
that was closer to each key.
We have also tested the EN-GR mapping in
Table 3, which results from a reordering of letter
mapping that gives a maximum correlation
between English and Greek single letter
frequency counts.
The predicted time for QWERTY layouts, for
expert users (RT=0), was as follows:
English layout simplified: 31,23 wpm
Greek simplified model: 29,91 wpm
Greek reordered-simplified: 30,75 wpm
Greek analytical model: 27,73 wpm
As we can see, reordering the Greek keys
results in a typing rate close to the English
QWERTY. The lower speed rate for the analytical
model is mainly due to accented and upper case
letters that require an extra key-press before the
letter key.
3.2 FI TALY layouts
FITALY layout shown in Figure 3, is a
commercially available software keyboard layout
(http://fitaly.com/, Textware Solutions 1996), that
was designed for minimizing travel distance
between the most frequent letters. According to
Textware, a 56,7% of total keystrokes occurs in
the central area of the keyboard (letters T-A-N-E-
O-R and the two space bars)
We have computed speed rates for the
following FITALY-like Greek keyboards: (a) one
using the typical QWERTY mapping as in Table 2
and (b) one with reordered keys as in Table 3
(Figure 4). We have tested layouts using the
simplified bigram tables (case independent without
comma, full stop and accent keys) and a Greek
reordered layout using the analytical bigram tables
of the Appendix.
After modifying the spreadsheet to calculate
optimum distances from the two space keys, we
obtained the following results for expert users:
English layout simplified: 40,57 wpm
56
Greek simplified model: 37,22 wpm
Greek reordered-simplified: 3 9,67 wpm
Greek reordered-analytical: 34,81 wpm
As expected, FITALY layout is much more
efficient than QWERTY, both in English and in
Greek languages. Entry speed rate improvement is
29,9% for English layout, 24,43% for Greek
simplified model, 29% for Greek reordered-
simplified model and 25,53% for the analytical
Greek model.
3.3 Telephone pad layouts
The telephone keyboard is used extensively for
exchanging short text messages via mobile phones.
As there are three or more letters assigned to each
key, text entry is done (a) either in a multi-tap
mode, or (b) using a disambiguation algorithm.
In multi-tap mode the user has to press a key one
or more times until the desired letter appears. I.e.
one should press three times the number 2 key, in
order to enter the letter "C". In word
disambiguation mode, each letter key is pressed
once and the algorithm selects and suggests the
possible word combinations from a built-in
lexicon. In case of more than one ambiguous
words, the user can select the desired word from a
list of suggestions.
As an example, to enter the word "HELLO" the
key sequence is "4-3-5-5-6" in word
disambiguation mode. In multi-tap mode this
would be "4-4-3-3-5-5-5-1 second pause-5-5-5-6-
6-6". Notice that in multi-tap mode a delay of one
second is needed if the subsequent keys are the
same.
It is obvious that multi-tap mode is highly
inefficient in terms of text entry speed.
Furthermore, modern mobile phones and devices
have enough memory and power to implement a
disambiguation algorithm. Thus we focused on
telephone keypads incorporating such an
algorithm. In addition, we assumed an "ideal"
algorithm and we have not taken into account the
time a user needs to select the desired word from a
list of suggestions.
With the assumptions above, modelling the
telephone keyboard (Figure 5) for maximum entry
speed (the case of an expert user), produced the
following results:
English layout simplified: 40,86 wpm Greek
simplified model: 41,30 wpm
Accent is automatically entered when typing in
lower case with disambiguation. Thus, an
analytical model for telephone keypads was not
computed. Further on, since the telephone keypad
follows a simple logic of alphabetically ordered
keys, a reordered Greek layout for this type of
keyboard would be meaningless and thus was not
tested.
4. CONCLUSI ON
One main difference between English and
Greek language is the presence of accented letters.
The high frequency (~15%) of these characters
should be taken into account when designing,
modeling and evaluating keyboard layouts, since
accent is an extra key that should be pressed prior
to the letter key. An exception is telephone pads,
where there is no provision for special accent key
Our predictions using an analytical model that
takes accent and letter case into account, showed a
7,3% (QWERTY) to 12,2% (FITALY-like)
decrease in entry speed, comparing to a simplified
model.
Another key finding is the moderate agreement
between English and Greek single letter counts,
when using the typical QWERTY mapping. After
reordering the keys so that the two frequency lists
correlate, we predicted a slight 2,8% increase in
typing speed for QWERTY-like keyboard, but a
noticeable 6,6% increase in the FITALY-like
layout.
Though reordering the keys on hardware
QWERTY keyboards might be confusing for expert
users, novel software keyboards could benefit from
it.
Finally, a very promising keyboard layout is the
telephone pad, being very efficient, both in terms of
space and in terms of entry speed rates, when
disambiguation algorithms are used. A vast majority
of the population is very familiar with it and
especially young people, which in many cases make
use of a mobile phone before they try any other
mobile device.
REFERENCES
[1]. Clarkson, E., Clawson, J ., Lyons, K., and
Starner, T., 2005. An Empirical Study of
Typing Rates on mini-QWERTY
[2]. Keyboards. In: ACM CHI 2005 Conference
on Human Factors in Computing Systems,
April 2-7 2005, Portland Oregon
[3]. USA. ACM Press, 1288 - 1291 Dominowski,
R.L., and Duncan, C.P., 1964. Anagram
57
solving as a function of bigram frequency.
Verbal Learning & Verbal
[4]. Behavior, 3, 321-325. Fitts, P.M., 1954. The
Information Capacity of the Human Motor
System in Controlling the Amplitude of
Movement.
[5]. Experimental Psychology, 47, 381-391.
Goldstein, M., Book, R., Alsio, G. and Tessa,
S., 1999. Non-keyboard QWERTY Touch
Typing: A Portable Input Intereface
[6]. For The Mobile User. In: ACM CHI 1999
Conference on Human Factors in Computing
Systems, 15-20 May 1999
[7]. Pittsburgh PA USA. ACM Press, 32-39.
Green, N., Kruger, J ., Faldu, C. and Amant,
R.S., 2004. A Reduced QWERTY Keyboard
for Mobile Text Entry. In: ACM
[8]. CHI 2004 Conference on Human Factors in
Computing, 24-29 April 2004 Vienna Austria.
ACM Press, 1429 - 1432. James, C.L., and
Reischel, K.M., 2001. Text Input for Mobile
Devices: Comparing Model Prediction to
Actual Performance.
[9]. In: SIGCHI 2001 Conference on Human
factors in Computing, 31 March - 4 April
2001 Seattle WA USA. ACM Press,
[10]. 365-371. J ones, M.N., and Mewhort,
D.J .K., 2004. Case-sensitive letter and bigram
frequency counts from large-scale English
corpora.
[11]. Behavior Research Methods, Instruments,
& Computers, 36 (3), 388-396. Mackenzie,
I.S., Zhang, S. X., and Soukoreff, R. W.,
1999. Text entry using soft keyboards.
Behavior & Information
[12]. Technology, 18 (4), 235-244.
Maragoudakis, M., et al., 2002. Improving
SMS Usability Using Bayesian Networks.
In: I.P. Vlahavas and CD.
[13]. Spyropoulos, eds. Lecture notes in
artificial intelligence (LNAI, Vol. 2308.
Springer-Verlag, 179-190. Masui, T., 1998.
An Efficient Text Input Method for Pen-based
Computers. In: ACM CHI 1998 Conference
on Human
[14]. Factors in Computing, 18-23 April 1998
Los Angeles California USA. ACM Press,
328-335. Mayzner, M.S., and Tresselt, M.E.,
1965. Tables of single-letter and digram
frequency counts for various word-length and
[15]. letter-position combinations. Psychonomic
Monograph Supplements, 1 (2), 13-32.
Mikros, G., Hatzigeorgiu, N., and Carayannis,
G., 2005. Basic Quantitative Characteristics of
the Modern Greek Language
[16]. Using the Hellenic National Corpus.
Quantitative Linguistics, 12 (2), 167-184.
Soukoreff, R.W., 2002. Text entry for mobile
systems: Models, measures, and analyses for
text entry research. Thesis (MSc).
[17]. York University. Tegic Communications,
1998. Reduced keyboard disambiguating
computer. US Patent no. 5,818,437. Textware
Solutions, 1996. Method for designing an
ergonomic one-finger keyboard and
apparatus thereof. US Patent no.
[18]. 5,487,616. Zhai, S., Kristensson, P.O., and
Smith, B.A., 2005. In search of effective text
input interfaces for off the desktop computing.
[19]. Interacting with Computers, 17, 229-250
58
59
ERRORS OF MECHANI CAL TREATMENT PROVOKED BY THERMAL
DEFORMATI ONS OF TECHNOLOGI CAL SYSTEM
V.I .Savulyak
1
, S.A.Zabolotniy
2
Dr.Sci.Tech., vsavulyak@mail.ru,
Cand.Tech.Sci., zab82@mail.ru
State Technical University Vinnitsa, Ukraine
Abstract: The thermal deformations of a detail and boring bar during the cutting out precise slots
were investigated by the means of modeling of compassing the method of final elements. The main
factors, which define the magnitude of thermal errors and their influence on distortion of the profile of a
detail were also established.
Keywords: keywords list
Introduction
One of the most widespread processes of
casting machine detail slots remains the process of
cutting of boring bar. The preciseness of slots is
determined by the complex operation of a number
of factors, which canse the emergence of
systematic and casual errors of treatment.
The important factor of forming of macro-
geometry of slots are the errors which arise due to
the action of heat the springs up in the place
cutting [1-4]. The most actual is the taking into
account the thermal deformations of the cutting
instrument as well as the surface of a detail during
the final operations.
It is very difficult to calculate most analytically
the defined errors, for the details having complex
configuration or being heterogeneous as for their
physical and mechanical peculiarities for example
having functional coatings. For these cases it is
very advisable to take into consideration the
calculation of errors using the models based on the
method of final elements and the realizations of
them by the means using a computer.
Organization of a Task of I nvestigation
To the thermal deformations of the
technological system, which influence the arising
of treatment errors, they regard the deformation of
machine mechanisms, a cutting edge of the
instrument (a cutter, boring bar, etc) as well as a
surface coat of the being treated allotment of a
detail in the place of cutting under the influence of
thermal currents.
In this paper the influence the current of heat,
that appears during the cutting a slot in the details
having different configuration.
The profile distortion is extremely undesirable
while forming of cutting of the slot surfaces, which
are the bases for the roller-bearings or those that
operate conjugating with pistons, plungers etc
[5-8].
The modeling was carried out with using the
method of final elements. The algorithm of
calculations made on a computer are described in
the paper [9]. The preliminary results of modeling
have shown that the thermal deformation of the
detail surface being treated mechanically directly
depends on general configuration of a detail.
In this paper the results of modeling of cutting
of precise slots in the details having different
thickness of a side are also brought into account.
Taking into consideration, that the distortion of
a slot profile is the result of the combined action of
such factors as: thermal deformation of a detail
caused by a source of a thermal stream, that was
arised by the process of cutting having intensity of
q
d
and moving along the slot axis with the sped of
feeding (fig.1, a) and the thermal deformation of a
instrument under the influence of the thermal
stream having the intensity of q
i
(fig.1, b).
60
For the simplification of the model such
assumptions were made: thermal streams into the
detail q
d
and the instrument q
i
are constant in time:
the whole period of treating is divided into the
steps periods during the next in turn step the
certain new allotment of the detail is heated by the
thermal stream caused by the process of cutting
and if is being known that during each next step
thermal influences of previous thermal inserting
were taking into account.
Figure 1: Thermal streams in the slot cutting
process: a) a scheme of putting a thermal streaminto
a detail operation; b) a schema of putting a thermal
streaminto a cutter into operation.
The quantities of thermal streams, that
influence the instrument and the detail, are
calculated with the help of the well known
dependences [1]. For example, in this paper the
results of investigations used for cutting based on
0,5 mm/ feeding and 100 mm/min cutting sped are
clearly shown.
Having aimed on determing the range of
influence of the thermal heat capacity of a detail,
that occur in the process of slot cutting, modeling
for two types of details were carried out: 1 details
having relatively small thickness of sides, that
makes possible of increase the temperature out the
external surface during the cutting process; 2
details having relatively big thickness of sides, that
practically doesnt allow to change the temperature
on the external surface and that is quite typical for
frame details.
If we call the detail of the first type a bush
which is modeled by a cylinder having such
parameters: the slot diameter d=60mm; the
external diameter D=100mm; the length L=40mm;
the material steal 45. Lets call the details of
other type a frame which is modeled by a cylinder
having such parameters: the slot diameter
d=60mm; the external diameter D=300mm; the
length L=40mm; the material steal 45.
The results and their discussion
The positional of the thermal source q
d
on the
each step of thermal loading are corresponding to
the position of the instrument during the slot
treating. The character of the slot profile distortion
caused by the thermal deformations of the detail
will be determined by the movement of some
points of material in the zone of cutting. Figure 2
and 3 show the radial shift of slot surface points in
the process of treating under the influence of the
thermal stream.
The investigation of the details having slots has
shown that the macrogeometry of the treated slots
depends on the their configuration and the thermal
heat capacity of their sides.
For the models of a bushing type, i.e. when the
thickness of the side is relatively small, the
extention of the slot sizes due to the heating caused
by cutting process is obviously seen and that leads
to the diminution of metal layer, which is being
cut. On account of this fact, the slot diameter of the
cut detail appears to be smaller after cooling than it
was calculated during the desposing the instrument
to a certain size.
Absolutely another result is observed while slot
cutting of frame details, and it is very characteristic
or them that the thickness their slot sides is rather
significant correlated to their diameter. As the
thermal heat capacity and the hardness of their
slots sides are rather big, external layers of details
are almost never become deformed, and thats why
significant strain pressure is developed in the zone
of cutting, and the deformation of the slot surface
is directed to its axis.
As a result at this, the bigger than necessary
layer of metal is cut off and that leads to the
increasing of a slot after its treating and then
cooling.
Residual distraction of the slot profile is the
result of the total sum of the thermal shifting of
slot surface points and of the instrument top. The
distortion character of longitudinal slot profile
achieved in the result of superposition of shifting is
shown in figures 2 and 3 applied to the details such
as the bushing and frame relatively.
The main component of the thermal
deformation of a instrument is its lengthening in
the direction perpendicular, to the axis of the
treated slot which appears under the influence of
the thermal stream, and has an effect on the
operating part of the cutting edge.
61
The distortion of the slot profile in the process
of treating owning to the lengthening of a
instrument is shown in figure 2 and 3 and on curve
1. It is clearly seen that the maximum distortion of
the slot radins caused by the tool instrument
deformation makes up 0.022 mm.
Figure 2: Thermal deformations of the bushing in
the process of cutting: 1 cutter top shifting, 2 slot
surface points shifting in the zone of cutting, 3 slot
profile distortion after treating and cooling.
Figure 3: Thermal frame deformations in the
process of cutting: 1 cutter top shifting, 2 slot
surface points shifting in the zone of cutting, 3 slot
profile distortion after treating and cooling.
The maximum slot distortion for the details of
the bushing tupe is twice bigger than the radial
shifting and makes up 0,138 mm, and for the frame
details it makes up 0,069 mm.
Conclusion
1. It is expedient to calculate the quantity of the
error caused by the thermal deformation of the
instrument-detail system using the models based
on the method of the final elements.
2. The character and parameters of slot profile
distortions appeared in the process of cutting
depend on the construction and sizes of the detail
as well as the sizes and physical and mechanical
qualities of the cutting instrument.
3. To reduce the errors of mechanical treating
appeared in the process of cutting, it is expedient to
use the system of adaptive guidance together with
adjusting the instruments accordingly to the results
of model calculations.
References
[1] Reznikov A.N., Reznikov A.N., Thermal
processes in technological systems,
Machinebuilding, Moscow, 1990.
[2] Sylyn S.S. Limilarity method of metal cutting ,
Machinery, Moscow, 1979.
[3] Bobrov V.F., Theory basis of mental cutting,
Machinery, Moscow, 1995.
[4] Yascherytsin P.I., Feldstein .,
Kornyevych MA, Theory of cutting, Novoe
Knowledge, Minsk, 2005.
[5] Petrakov Y.V. Automatical management of
processes of treating materials by cutting,
UkrDNIAT, Kyiv, 2004.
[6] Kosilova A.G., Meshcheryakov RK Reference
book of technologist- and Machine-building
engineer. Vol.1 Machinery, Moscow, 1985.
[7] Kurchatkin V.V., Telnov N.F., Ivanov K.A.
Reliability and maintenance of automobiles,
Kolos, Moscow, 2000.
[8] TChernoivanov V.I. Organization and
Technology Restoration of automobiles :
HOSNYTY, Moscow, 2003.
[9] Savuliak V.I., Zabolotny S.A, Shenfeld V.Y.
Thermal fields and deformations during
restoration details of transport techniques
Bulletin DREAMS them. Vladimir Dal., Luck,
2009.
62
63
I RON REMOVAL FROM WASTEWATER USI NG CHELATI NG
RESI N PUROLI TE S930
Petru Bulai
1
, Elena-Raluca Cioanca
1
tefan cel Mare University, Faculty of Mechanical Engineering, Mecatronics and Management, Department of
Technology & Management, no. 13, Universitii, 720229, Suceava, Romania
bulaipetru@yahoo.com
cioancaraluca@yahooo.com
Abstract: The work presents the sorption characteristics of Iron (II) on iminodiacetic resin Purolite
S930 in various operating conditions such as initial pH, copper concentration, contact time, temperature,
ionic form of the resin and resin dose. The percent of Iron (II) removal has a maximum at pH 5.0, and
increases with the increasing of resin dose, of the contact time and decreases with increasing initial
concentration of solution.
Keywords: iron, sorption, chelating resin, wastewater
1. Introduction
In water (drinking water, water supplies and
wastewater) the Iron causes problems, such as
giving reddish color and odor [3].. In general in
water, iron exists in soluble form as Iron (II) and is
oxidized to ferric iron according to eq. 1 [3].
2Fe
2+
+(1/2) O
2
+H
+
2Fe
3+
+H
2
O (1)
The state of iron in water depends on the
hydrogen concentration and the redox potential.
With the increasing of the hydrogen concentration,
dissolved iron (Fe
2+
or Fe
3+
) hydrolyzes to form
precipitates. The ferrous ion hydrolyzes to produce
the array of mononuclear species FeOH
+
to
Fe(OH)
4
2
between pH 7 and 14 [10].
There are several methods for removal of iron
and other heavy metals, used in water purification
processes (drinking water treatment and
wastewater treatment. All techniques have their
advantages and limitations in application [4].
The most important techniques are chemical
precipitation, coagulationflocculation, flotation
[6], [4], membrane filtration: ultrafiltration,
nanofiltration [1] and reverse osmosis;
electrochemical treatment techniques: electro
dialysis, membrane electrolysis and
electrochemical precipitation, electroextraction [9];
sorption treatment techniques: ion exchange [7].,
adsorption [12], biosorption [11].
As other heavy metals the presence of iron in
the environment is a major concern due to their
non-biodegradability, bioaccumulation tendency,
persistence in nature and toxicity to many life
forms. Therefore, treatment of wastewaters
containing heavy metal ions before discharge is an
important component of water pollution control
and becomes more important with the increasing of
industrial activities
Iron (II) ions have a high solubility in the
aquatic environment and can be absorbed by plants
and living organisms. The maximum acceptable
concentration of Iron (II) in drinking water
recommended by World Health Organization
(WHO) is 0.2 mg/L. The limit of Iron (II) into
wastewater 5 mg/L according to Romanian
regulation (NTPA001/2005).
Compared with other usual methods, ion
exchange provides some advantages and is one of
the most widely used techniques for treatment of
wastewaters in many chemical process industries
(Macoveanu et al., 2002). The efficient removal of
metals ions requires an exchange material which
can function in severe environments, exhibit a high
affinity for the metal of interest and can be readily
64
regenerated. Due to their physical and chemical
properties chelating resin are generally efficient in
the removal of heavy metals ions from waste
effluents.
2. Experimental
2.1. Materials
In the experiments was used the S930
chelating resin obtained from Purolite International
Limited (Hounslow, UK). The main physical and
chemical properties of the resin are presented in
Table 1.
The stock solution of iron (II) (500 mg/L) was
prepared from analytical-reagent grade copper
sulphate (FeSO
4
.7H
2
O) in distilled water. Required
pH adjustments were performed with sulphuric
acid.
2.2. Sorption experiments
To avoid the iron precipitation in alkaline
solution the hydrogen resin form was converted
into sodium resin form. For the conversion of the
resin was used 10% HCl solution. The process was
followed by washing with distilled water until the
pH of the effluent dropped to neutrality. The
hydrogen resin form has been dried at 60 C using
an oven.
Sorption of iron (II) ions on Purolite S930
in hydrogen (S930-H) form was carried out in
batch experiments using 50.0 mL of iron (II)
solutions with different initial concentrations
(10300 mg/L) that where added to
Erlenmeyer flask already containing 0.05 g of
dry resin. The initial pH of the solution was
adjusted by using diluted solutions of H
2
SO
4
or 10 mL of acetate buffer solutions and
measured with WTW pH/cond 340i. The
flasks where mechanically shacked at several
fixed temperatures and at the rate of 120
cycles min
1
using Orbital Shaking Incubator
GFL 3031. After equilibrium (24 hours), the
resin and solution were separated by filtration
and the iron content of the solution and, also,
the final pH of solution (pH
e
) were measured.
In the experiments concerning the effect of the
resin dose, a range of resin samples from 0.01
to 0.2 g was used. For contact time
experiments, the procedures were the same as
those presented in equilibrium experiments but
the samples were analyzed after a specified
period of contact time.
The concentration of Fe(II) in solutions was
measured using a spectrophotometric method with
1,10-phenantroline and hydroxilaminochlorohidrat
=510nm) using Hach DR/2000
spectrophotometer [8]. Because in the presence of
O
2
from air a part of Fe(II) becomes Fe(III),all
experiments measured the concentration of total
iron as Fe(II) [7].
The sorption of Iron (II) by Purolite S930-H
resin was quantitatively evaluated by percent of
Iron (II) removal R (%) and by amount of Iron (II)
retained on resin, q(mg/g):
0
0
100
e
C C
R
C
(2)
0
( )
e
C C V
q
m
(3)
where C
0
and C
e
are the initial and the
equilibrium concentration of Iron (II) in the
solution (mg/L), V is the volume of solution
(L) andmis the mass of the resin (g).
Table 1. Characteristic proprieties of the chelating resin used*
Polymer matrix structure Macroporous styrene divinylbenzene
Functional groups Iminodiacetic acid
Ionic Form (as shipped) Na
+
pH range (operating): H
+
form, Na
+
form 2- 6; 6- 11
Maximum operating temperature 70C
Particle size range +1.0mm <10%, -0.3mm <1%
Total exchange capacity 1,9 meq/mL
* Manufacturer supplied.
65
3. Results and discussion
3.1. Effect of solution pH
Sorption of Iron ions on chelating resins
Purolite S930 involve an ion exchange process in
conjunction with the formation of a coordinate
bond. The ability to bind metal ions through amine
groups lead to the grater selectivity for heavy
metals from wastewaters. The formation and
stability of the chelates with many metal ions are
very dependent on the solution pH. Because the
functional iminodiacetate groups of the
Purolite S930 are weak acids, they are very
selective for hydrogen ions (their dissociation
is dependent on pH) [2].
In order to investigate the influence of initial
solution pH on the Iron (II) removal by S930 resin,
experiments were carried out using a resin dose
(S930-H) of 1g/L and the initial iron concentration
of 300 mg/L at 293.15 K. The initial solution pH
was adjusted using H
2
SO
4
diluted solutions. The
correlation between the initial pH of solution and
Iron (II) equilibrium concentration and the percent
of Iron (II) removal, respectively, is illustrated in
Figure 1.
Figure1: Effect of solution acidity on Iron removal
onto S930 resin in H
+
form:
C
0
=300 mg/L, T =293.15 K, time =24 h, resin
dose =1g/L
The results reveal that in strong acidic media
(pH=2) most carboxylic functional groups are
protonated but with increase of pH up to 4-5 they
start to dissociate and can react with Iron ions. The
equilibrium Iron concentration decreases from the
initial value of 300 mg/L to 296.82 mg/L and
235.27 mg/L for initial pH 2 and pH 5,
respectively.
Figure 2 illustrate the variation of sorption
capacity of S930-H resin for Iron (II) ions as a
function of the initial pH of solutions.
Figure 2: Effect of solution acidity on loading
capacity of Purolite S930 resin in H
+
form:
C
0
=300 mg/L, T =293.15 K, time =24 h, resin
dose =1g/L
The loading capacity decrease significantly
from 64 mg Iron (II) per gram resin for pH =5
to 3.14 mg per gram resin for pH =2. Figure 3
present a correlation between initial pH of the
solution and equilibrium pH.
Figure 3: Comparison between initial hydrogen ions
concentration ( pHi) and equilibriumhydrogen ions
concentration ( pHi):
C
0
=300 mg/L, T =293.15 K, time =24 h, resin dose
=1g/L, S930 resin in H
+
form
Figure 3 shows that equilibrium pH are lower
than initial pH and the difference from the initial
pH and equilibrium pH decrease with decreasing
of initial pH. The decrease of equilibrium pH can
be explained by hydrogen ions releases in solution
from ion exchange process (Eq. 4.).
2
2
2 930 ( 930) 2 S H Fe S Fe H

(4)
3.2. Effect of metal ion concentration
Metal ions sorption onto resin is strongly
influenced by the initial metal ion concentration.
The sorption data for Iron (II) ions on S930 resin in
hydrogen form, namely removal percent and
equilibrium concentration, as a function of initial
66
concentration of iron are showed in Figure 4 and
figure 5, respectively. Experiments were carried
out in solutions of pH=5 adjusted with sulphuric
acid, at a resin dose of 1g/L and temperature
293.15 K.
Figure 4: Effect of metal ion concentration on the
Iron (II) removal percent onto S930-H:
pH=5, time =24 h, T =293.15 K, resin dose =1g/L
Figure 5: Effect of metal ion concentration on the
Purolite S930-H loading capacity for Iron (II): pH=5,
time =24 h, T =293.15 K,
resin dose =1g/L
Figure 4 reveals that with the increasing of Iron
(II) concentration the percent of removal decrease
but even for an initial concentration up to 25 mg
Iron (II)/L, the percent removal was of 80%.
Figure 5 shows that the Purolite S930-H loading
capacity decreases with increasing of initial ion
concentration.
3.3. Effect of contact time
The effect of the phases contact time on Iron
(II) sorption onto hydrogen ionic form of the resin
Purolite S930 is illustrated in Figure 6. The
experiments were conducted using solutions of 100
mg Iron /L at temperature 293,15K; the initial pH
=5 was adjusted using H
2
SO
4
diluted solutions.
Fig.6. Effect of contact time on the Iron (II) removal
on S930-H resin:
C
0
=100 mg /L, pH =5, T =273.15 K, resin dose =
1g/L
The results reveal that within first 10 hours
40% sorption was occurred. The equilibrium was
considered attained (reached) after 24 hours; a
further increase in contact time has a negligible
effect on the percent removal. The faster initial
sorption rate may be explained by the greater
number of resin sites available for the sorption of
metal ions.
4. Conclusions
The present study shows that Purolite S930 is
an effective sorbent for the removal of Iron (II)
ions from aqueous solutions. The percent of Iron
(II) removal has a maximum at pH 5.0 and
increases with the increasing of resin dose, of the
contact time and decreases with increasing of
initial concentration of solution.
5 References
[1] Bernat X., Fortuny A., Stber F., Bengoa C.,
Fabregat A., Font J ., Recovery of iron (III)
from aqueous streams by ultrafiltration,
Desalination, 221, 413418, 2008.
[2] Bulai P., Balan C.D., Bilba D., Macoveanu M,
Study of the copper (II) removal fromaqueous
solutions by chelating resin Purolite S930,
Environmental Engineering and Management
J ournal, 8(2), 213-218, 2008.
[3] Cho B.Y., Iron removal using an aerated
granular filter, Process Biochemistry, 40 (10),
3314-3320, 2005.
[4] Kurniawan T. A., Chana G.Y.S., Loa W.,
Babel S., Physicochemical treatment
techniques for wastewater laden with heavy
metals, Chemical Engineering J ournal, 118,
8398. 2006.
67
[5] Petrov S., Nenov V., Removal and recovery of
copper fromwastewater by a complexation-
ultrafiltration process, Desalination, 162, 201-
209 2004.
[6] Polat H., Erdogan D., Heavy metal
removal from waste waters by ion
flotation, J ournal of Hazardous Materials.,
148, 267273, (2007).
[7] Popa C , Bulai P., Macoveanu M., The study of
iron (II) removal from34% calciumchloride
solutions by chelating resin Purolite S930,
Environmental Engineering and Management
J ournal, 9(5), 651-658, (2010).
[8] Popa G., Moldoveanu S., Quantitative
Chemical analysis using organics reagents, (in
Romanian), Technical issue, Bucureti,
Romania, 1969.
[9] Smara A., Delimi R., Poinsignon C.,
Sandeaux J ., Electroextraction of heavy
metals fromdiluted solutions by a process
combining ion-exchange resins and
membranes, Separation and Purification
Technology., 44, 271277, 2005.
[10] Tekerlekopoulou A.G., Vasiliadou I.A.,
Vayenas D.V., Physico-chemical and
biological iron removal frompotable water,
Biochemical Engineering J ournal, 31 [1], 74-
83, 2006.
[11] Tuzen M., Uluozlu O. D. , Usta C. , Soylak M.,
Biosorption of copper(II), lead(II), iron(III)
and cobalt(II) on Bacillus sphaericus-loaded
Diaion SP-850 resin, Analytica ChimicaActa,
581, 241246, (2007).
[12] Veli S., Alyuz B., Adsorption of copper and
zinc from aqueous solutions by using natural
clay, J ournal of Hazardous Materials, 149,
226233, 2007.
68
69
ANAL YSI S OF CONDI TI ONS FOR OBTAI NI NG A KEY HOLE
REGI ME FOR LASER WEL DI NG
Remus Boboescu
1
1
University Politehnica Timioara, remus_boboescu@yahoo.com
Abstract:. For laser welding using high intensity of laser beam produce the keyhole phenomenon in
welding pool. For keyhole welding regime there is a particular of the weld cross-section shape aspect
given by the heat affected zone area. It analyzes the laser welds made with Nd: YAG laser on low alloy
steel plates using irradiation in continuously regime. The power, welding speed and defocusing are
varied. Experiments were conducted after the full factorial experimental design 2
2
. This showed the effect
of laser power for setting keyhole welding regime.
Keywords: laser keyhole welding, full factorial design, laser beamdefocusing, heat affected zone.
1. Introduction
Laser welding has two distinct regimes of
conduction welding and keyhole welding regime
which differ in the weld penetration and weld cross
section shape.
Conduction welding regime. This regime is
called conduction limited welding. The solid-liquid
interface heat transfer mechanism is exclusively by
conduction. Heat is conducted directly from the
heat source that occurs due laser radiation from the
workpiece surface within the material. Conduction
welding regime is characterized by irradiation
sufficient to produce piece surface melting and
vaporization but not enough to produce
vaporization in the depth of material. For
conduction welding regime values of F ratio
between the weld width and weld depth are higher
than 1, 1 F . As an indicative value for the weld
depth in conduction regime is 1.5 mm [1]. Defects
that are obtained in conduction welding regime are
cracks on the weld surface.
Keyhole welding regime. This welding regime
involves the keyhole phenomenon in weld pool
which involves the values of laser beam intensities
that exceed threshold needed to produce a vapor
front propagation in the material. Heat transfer
mechanism at solid-liquid interface involves both
conduction and convection due to the melt
movement in the weld pool.In weld cross section
there is a welding defects gas bubbles, the presence
of this pore led to the designation of keyhole
regime (key hole). F ratio between weld width
and weld depth has a smaller value than 1, 1 F .
Defects in welds are obtained for the keyhole
welding regime are gas porosity due to dissolution
into the melt a gas bubbles and vacuum large areas.
The presence of the Keyhole and vapor within the
material made pushing melting front (solid-liquid
interface) inside the material. This provides
increased weld penetration [2], [3], [4]. For welds
made in experiments the two welding regimes are
shown in Figure 1 by the weld cross section.
a) b)
Figure 1: Image of weld cross-section of the weld a)
weld in the conduction regime b)weld in keyhole
regime
The paper proposes the identification and use of
quantities that can characterize welds under
keyhole welding regime. It looks like those in
characterizing keyhole regime occurs sizes
associated with both weld cross section and weld
surface. Experimental factorial design 2
2
used
presents for higher levels of power keyhole
welding regime and for the lower power level
presents the conduction welding regime. It will
discuss variation of the response functions relative
to varied parameters power and welding speed.
70
2. Experiments
The experiment consisted in made lines of
fusion (welds), 110 mm long, on Dillimax 500
steel plates with thickness of 10 mm (carbon steel,
carbon content 16 . 0 %). Was used a Nd: YAG
Trumph Haas 3006D laser source with 3kW
maximum power on a continuous wave regime
CW. Laser beam was transmitted through a optical
fiber with core diameter of 0.6 mm
The focus system made a focal spot with 0.6
mm diameter. Lens focal length was 200 mm. As
protective gas argon was used with a flow rate of
20 l / min. Were used sheets of material with
dimensions of 10 130 100 mm for which were
made between 5 and 8 welds, with a distance of
over 10 mm between welds. Parameters varied in
the experiments are presented in Figure 2.
Figure 2: Scheme of keyhole welding pool with varied
parameters inwelding process
In experiments was varied the laser power,
welding speed and distance between focal plane
and piece surface (defocusing or defocusing depth)
figure 2. Welds were cut in the stable part of the
weld near the place where welding process was
stopped. Weld section was processed
metallographic. Weld width, near the piece
surface, and weld depth were examined using a
microscope with precision of 0.01 mm. Melted
area was measured directly by its footprint..
Defocusing values are considered negative if the
laser beam focus inside the piece.
In the experiments were varied power and
welding speed. To statistically analyze the effects
of parameters was necessary to introduce a
dimensionless parameter values.
Transformations between the two systems of
parameters values (actual values and coded
values) are based by following relationships:
2 P A [-] (1)
v B 22 . 2 33 . 2 [-]
(2)
The experimental plan is presented in Table 1 with
actual values that coded for power and cutting
speed.
Table 1: Varied parameters values in experiment
Analysis procedure consisted of presenting the
results of the mathematical model, ANOVA table
showing the correlation coefficients associated
with the mathematical model, Pareto chart showing
the hierarchy of effects and response surface is a
graphic representation of mathematical model. For
the mathematical model were presented two forms
for real values laser power and welding speed and
for coded system values. The first allows rapid
application of formulas and the second allows
direct analysis of the values of regression
coefficients. Based on these values were achieved
Pareto charts.Figure 3 shows the analysis weld
scheme and analyzed sizes that characterizing the
weld.
Figure 3: Scheme of weld with measured sizes
The paper analyzed variations the following
sizes: weld width, shape ratio of weld cross section
and the heat affected zone area on the weld cross
section.
Weld width L [mm] was obtained as a result of
three measurements on the weld surface at the
beginning, middle and end of the welding
process.The weld width characterized in general
71
the weld and it is independent of the area in this
cross section in weld was performed.
F ratio (w/h) is the ratio beetwen weld width
and weld depth on the weld cross section.
This ratiois associated with the welding regime
characterization.Values of F ratio below unity
shows keyhole welding regime.
A heat affected zone area S[mm
2
] was measured
on the weld cross section. It is given by the
isothermal line of transformation metal structure
which is well below the melting temperature. Area
heat affected zone containing molten zone area.
Heat affected area can be measured with greater
accuracy than the area of molten zone.
3.Weld width
The mathematical model for weld width at
defocusing 0 is given by relations (3) and (4).
Statistical analysis by ANOVA method is given in
Table 2. Figure 4 shows the Pareto chart for weld
width at defocusing 0 . It is noted that the weld
width increases with power and decreases with
welding speed.
The decrease effect with welding speed is
greater than the increase effect of power. The
interaction between power and welding speed
increases width of the weld width so the overall
effect of increasing with power exceeds the
decreasing effect with welding speed. The Pareto
diagram presented three effects have similar values
and have no statistical significance. It looks like
that along the weld there is a change for the first
effect between power and welding speed.
B A B A L 275 . 0 375 . 0 29 . 0 086 . 2
(3)
Pv
v P L
61050 . 0
0535 . 2 93075 . 0 64859 . 0
(4)
Table 2: ANOVA table for weld width L at 0
Figure 4: Pareto Chart for weld width at 0
Figure 5: Response surface for weld width at 0
Figure 5 shows the surface response for weld
width at defocusing 0 . It is noted that the
weld width has high values on the experimental
field. Lower values for the weld width are obtained
at low power and high welding speeds. There is a
sharp increase in weld width values on the
diagonal of experimental field from low power and
high welding speed on the situation in which
power is high and low welding speed. It is
recommended for high value of weld width. These
values can be associated with deep welds.
The mathematical model for weld width at
defocusing mm 2 is given by equations (5)
and (6). The correlation coefficients for the
mathematical model associated with the
mathematical and statistical method of ANOVA
analysis of variances are given by equations (7)
and (8).
72
Figure 6 shows the Pareto diagram for the weld
width at defocusing. It is noted that the weld width
increases with power and decreases with welding
speed. Effect of power have statistical significance.
The interaction between power and welding speed
decreases weld width. It is noted that the
cumulative effect of decreasing of the weld width
with welding speed is lower than the increase with
power. This shows the strong dependence of the
weld width with power.
B A
B A L
325 . 0
545 . 0 875 . 0 0966 . 2
(5)
Pv
v P L
72150 . 0
23310 . 0 63225 . 1 10195 . 0
(6)
95 . 0
2
R
(7)
87 . 0 ) . . (
2
f d for adj R
(8)
Figure 6: Pareto Chart for weld width at mm 2
Figure 7 shows the response surface for the
weld width at defocusing mm 2 . It is noted
that on the experimental field weld width increases
with power and decreases with welding speed.
Maximum values for weld width are obtained for
the maximum power and minimum welding speed.
The type of increase seen in Figure 5 exists in this
case for focusing within piece. The differencethat
arises is that this type of growth is valid throughout
the experimental field.
Focusing the laser beam inside the piece brings
the following changes relative to the weld width.
By focusing a laser beam inside the piece produce
a control of weld width through its dependence on
power and association of this dependence to
statistical significance. Focus inside piece changes
sign of interaction between power and speed.
To focus on the piece surface effect of the
interaction is associated with speed effect, for
focus within the piece the interaction effect is
associated with power effect. This shows the
increased contribution of dynamic phenomena that
occurring in the weld pool. It shows that the laser
beam focus inside the piece provides stability on
weld width and for overall the weld surface
characteristics of welds.
Figure 7: Response surface for weld width at
mm 2
4. Ratio F
The mathematical model for the ratio F at
defocusing 0 is given by equations (9) and
(10). Study of variation by the ANOVA method is
presented in Table 3.
AB B A F 425 . 0 375 . 0 025 . 1 535 . 1 (3)
Pv
v P F
9435 . 0
7195 . 2 01525 . 2 43925 . 6
(4)
Table 3: ANOVA table for Ratio F at 0
Figure 8 shows the Pareto chart for the ratio F
at defocusing 0 . Decreased of ratio F shows
getting keyhole welding regime and increase its
intensity. The figure notes that the experimental
field F Ratio strongly decreases with power. Effect
of interaction between power and welding speed
73
decreases the ratio F. Welding speed increases
Ratio F. The overall effect of power is much higher
than the welding speed effect. The three effects
have not shown statistically significance.
Figure 8: Pareto Chart for ratio F at 0
Response surface shown in Figure 9 indicates
that the F ratio decreases with the power and
welding speed on the experimental field. Lowest
values for the F ratio are obtained at high power
and low welding speeds. At high power welding
speed has no influence. In this situation gets
keyhole welding regime. Conduction welding
regime is carried out at low power and high speed
welding.
It notes that there is a sudden transition from
conduction welding regime to the regime keyhole
welding.
Figure 9: Response surface for Ratio F at 0
The mathematical model for the ratio F at
defocusing mm 2 is given by equations (11)
and (12). Study the variations by ANOVA method
is presented in Table 4.
AB B A F 2275 . 0 2425 . 0 0925 . 1 71 . 1 (11)
Pv
v P F
505050 . 0
548450 . 1 622575 . 1 520175 . 5
(12)
Table4:ANOVA table for Ratio F (w/h)at mm 2
Figure 10 shows the Pareto chart for the ratio F
at defocusing mm 2 . It is noticed that the
experimental field F ratio decreases with power
and also decreases with welding speed. The only
effect that increases the ratio F is given by the
interaction between power and welding speed. The
effects have not shown statistically significance. It
is shown that both increased with power and
welding speed leads to increased intensity of
keyhole welding regime. Focusing the radiation
inside the piece provides propagation and
absorption of radiation in Keyhole. High welding
speed leads to the production of two spots in which
the laser radiation absorption takes place. So this
may lead to loss of the radiation out of keyhole by
reflection, situation shown by the interaction
between power and welding speed.
Figure 10: Pareto Chart for Ratio F at mm 2
74
Figure11:Response surface for Ratio F at mm 2
Response surface in Figure 11 shows the
variation ratio F at defocusing mm 2 . It is
noticed that on experimental field the F ratio
decreases with power. Welding speed do not
produce significant changes. It looks as if the laser
beam focus within the piece setting of keyhole
welding regime depends solely on power.
Focusing the laser beam inside the piece,
compared with focusing on the piece surface,
increased the role of power in the setting of the
keyhole welding regime. It notes the changing
role of the interaction effect between power
effect and welding speed effect.
5. Heat affected zone area on the weld
cross section
The mathematical model for heat affected zone
area on weld cross section at defocusing is given
by equations (13) and (14). ANOVA statistical
analysis method is presented in Table 5.
B A B A S 125 . 3 875 . 3 875 . 6 25 . 9 (13)
Pv v P S 9375 . 6 2725 . 5 40625 . 0 03375 . 1 (14)
Table 5: ANOVA table for HAZ area S at 0
Figure 12: Pareto Chart for HAZ area at 0
Figure 12 shows the Pareto chart for the area of
heat affected zone to focus on the piece
surface. 0 . It is noted that the area of heat
affected zone increases with power and decreases
with welding speed. Power effect is statistically
significant.The interaction between power and
welding speed decreases the area of heat affected
zone. The overall decrease effect with welding
speed is stronger than the increase effect with
power. It is noted that all three effects are
important contributions for determining the heat
affected zone area.
Figure 13: Response surface for HAZ area at 0
Figure 13 shows the response surfacefor area of
the heat affected zone at defocusing 0 . It is
noted that on the experimental area of heat
affected zone increases with power and decreases
with welding speed on the field experiment. The
highest values for the area is obtained at high
power and low welding speed.
75
The mathematical model for heat affected zone
area on the weld cross section at defocusing
mm 2 is given by equations (15) and (16).
ANOVA statistical analysis method is presented in
Table 6.
B A B A S 625 . 2 875 . 3 125 . 8 4166 . 9 (15)
v P
v P S
82750 . 5
05250 . 3 24125 . 14 03715 . 10

(16)
Table 6: ANOVA table for HAZ area S at mm 2
Figure 14: Pareto Chart for HAZ area at mm 2
Figure 14 presents the Pareto chart for the area of
heat affected zone at defocusing mm 2 .It is
noted that the area of heat affected zone increases
with power and decreases with welding speed. The
interaction between power and welding speed
decreases the area of heat affected zone. Power is
the only effect that has statistical significance.
variation in heat affected zone area on weld cross
section at defocusing mm 2 .It is noted that in
the experimental field heat affected zone area
increases with power and decreases with welding
speed. Variations with welding speed are strongest
at high power than at low power.
Response surface deformation on experiment al
field is due by the interaction between power and
welding speed effect.
Figure15:Response surface for HAZ area at mm 2
For heat affected zone area laser beam focus
within the piece led to an increase in the role of
power. This increase is relatively small. At focus
within the piece the effect of the interaction
between power and welding speed decreases.
6. Discussion
Laser welding is widely applied in particular for
metals and steel. The laser beam is a concentrated
heat source.
In laser welding Keyhole phenomenon has two
important consequences. The first relates to
increasing the energy coupling between laser
radiation and material because the laser radiation
can propagate in the Keyhole and its absorption
occurs inside the material. This creates the
possibility of formation of heat sources inside the
material. It is considered that the presence of
Keyhole will increase the energy couplingbetween
laser radiation and material.
The second issue concerns the very strong growth
of Keyhole in the depth of material. The presence
of a strong heat source that produces sudden
changes in time and space heating favors strong
evaporation. Vaporization is followed by a rapid
increase in pressure followed by high in
vaporization temperature. Such gas-liquid interface
becomes overheated. Rapid disappearance of the
irradiated heat source will produce rapid
vaporization of superheated liquid by passing the
vapor.
In this paper we studied the variation of the
three sizes that characterize welds in keyhole
76
welding regime for two situations for the depth of
focus. Keyhole welding regime was obtained for
high values of power. The Ratio F shows subunit
values for keyhole welding regime. These F ratio
values are associated with high values of weld
width and with high values heat-affected zone area
on the weld cross section.
It looks like that all the three sizes analyzed in
the paper, weld width, Ratio F, heat affected zone
area can be used to separate the welds in the
conduction regime by the welds in keyhole regime.
Although the F ratio criterion defines the
separation between the two welding regimes, the
mathematical models for the weld width and area
of heat affected zone are useful in predicting
welding regime obtained. Focusing on the surface
of the piece introduces instabilities in the welds.
These are associated with the lack of statistical
significance for the mathematical models
associated with one dimensional size for welds.
Weld instabilities produced by focus on piece
surface are morepronounced for weld width, size
that characterizes the weld surface. Heat affected
zone area shows small with defocusing.
Mathematical models have shown high levels of
correlation coefficients.
7. Conclusions
The paper presents two important issues for the
study of welds made with laser beam the presence
of keyhole regime and the effect of defocus for
laser beam. For analysis of the welding process
was used a full factorial experimental design and
were used mathematical modeling and statistical
analysis of variations. The following were showed:
- The weld width and heat affected zone area on
the weld cross section can be used to characterize
the welding regime as well as the F ratio.
- For high power levels were obtained welds in
keyhole welding regime.
- Focusing the laser beam to the workpiece surface
produces instability in the welding process.
-Defocusing change did not affect the welding
regime obtained as evidenced by effects on the
heat affected zone area.
The aim of experimental study was the
assessing contribution of power and speed of
welding to weld characteristics. From the analysis
carried out were the following: power has main
effect on all response function and in conclusion
for laser melt capacity; power increase laser
melting capacity of steel and welding speed
decrease this capacity; mathematical models for
the weld width and weld depth provided a good
statistical correlation and statistical significance.
The experimental modeling achieved considered
only the main part of the laser processing. There
have been lines fusion lines and not welded joints.
Were varied only main factors of influence, power
and speed, while focal point position in relation
with the piece surface and assistant gas parameters
was maintaining constant. The plates used were
sufficient thick to not be penetrated during
welding. It is important that experimental research
should be applicable in slightly modified
experimental conditions. To fill the gap between
tow pieces using at laser welding using added
material in the form of wire. Mathematical models
performed make predictions for the laser welding
results for Dillimax 500 steel or other steels with
close thermal characteristics. Experimental
research methods used are generally applicable in
material processing.
References
[1] Alexander F. H. Kaplan, Masami Mizutani,
Seiji Katayama, Akira Matsunawa, Keyhole
Laser Spot Welding, International Conference
Applications of Lasers and Electro-Optics
2002
[2] Yukimichi Sasaki , Tadashi Misu , Shunro
Yoshioka , and Toshiyuki Miyazaki,
Monitoring of YAG Laser Spot Welding-
Detection of Porosity Defect by Acoustic
Signal International Conference Applications
of Lasers and Electro-Optics 2002.
[3] Inoue, Miamoto, Ono, Adachi, Matsumoto, In
process monitoring of penetration depth in 20
class CO
2
laser welding of thick sections,
International Conference Applications of
Lasers and Electro-Optics 1999.
[4] J ae Y. Lee, Sung H. K., Dave F. Farson,
Choong D. Yoo, Mechanism of keyhole
formation and stability in stationary laser
welding, J . Phys. D: Appl. Phys. 35 (2002)
15701576.
77
ASPECTS REGARDI NG THE QUALI TY OF CAST I RON
PRODUCED I N S.C.FONTUR S.A. SUCEAVA
Gramaticu Mihai
1
; Cosmaciuc Vasile
2
; Costiug Gina
3
; Hricu I onu Marius
1
1
tefan cel Mare University of Suceava, Faculty of Mechanical Engineering, Mechatronics and
Management, University Street, No. 13, 720225, Suceava, Romania,EU. Suceava, gram@fim.usv.ro
2
S.C. EXPROTERM S.A. Suceava, exproterm@yahoo.com
3
S.C, FONTUR S.A. Suceava,
contact@fontur.ro
Abstract: The authors analyze the quality of produced cast iron and molded forms from
SC FONTUR SA Suceava. For this purpose samples were cast from different batches of samples, from
which were made tensile specimens, samples for chemical and metallographic analysis and hardness
control. Based on analysis conducted authors show corrections that have to be made to
the chemical composition and of cast iron structure to obtain high mechanical strength. Also they give
recommendations on improving technologydevelopment.
Keywords: cast iron, chemical composition, microstructure, mechanical properties, moulding
1. General aspects
Increased production of iron castings is due
primarily to the fact that cast iron is the cheapest
casting metallic material with great technological
properties and tensile strength, that for nodular
ductile cast iron is simillar to those of steel. The
moulding cast iron technologies are more simple
and give better reproductibility of results.
Gray cast iron structure is made of: basic
metallic mass (mold) and graphite.
Basic metallic mass is simillar to hipo or
eutectic steel (C
mb
_0,77%). It can be:
- ferritic, carbon of metallic core mass is
- 0.1% C, polyhedral aspect, soft, HB =100 ...
120;
- ferito-perlite, C
s0,77
> C
mb
> C
sol
made
of white polyhedral ferrite grains (around
graphite) and lamellar pearlite, HB =140...180 ;
- perlite, C
mb
= C
s0,77
, with lamellar aspect,
hard and resistant, HB =250...340.
Where graphite is grossly spread , its influence
on cast iron strength is prevalent, the metallic
mass being insignificant, therefore primary coarse
graphite cast iron have reduced mechanical
strength [1].
From a structural point of view, graphite is
characterized by: shape, size, quantity and
distribution.
The graphite amount (occupied area) shows the
filling degree of graphite formations. It depends
on the total amount of carbon in cast iron
composition and especially the amount of spread
graphite in crystallization process. We need to
consider a minimal value of graphite quantity
restricted by moulding conditions. Therefore,
total carbon content of gray cast iron is taken
between 2.4 ... 3.8 [2],[3], [4].
With density four times lower than perlite, the
actual volume occupied by graphite is much
higher than the C
gr
value (free spread graphite
carbon form).
For exemple, for a cast iron with C
T
=3%, that
has basic perlitic mass (C
mb
=0,77%), C
gr
=2,23% it
coresponds a volume of 9% graphite.
Shape appears to be the most important
graphite classification criteria, characterized by
density degree of graphite spreads, respectively the
cutting degree of basic metallic mass and the
tensions focusing effect. The size of graphite
spreads also influences the mechanical proprieties
especially lamellar graphite.
The prevalent role of graphite concerning the
reduced level of cast iron mechanical
characteristics derive from its negative effects,
namely:
- cutting effect, respectively decreasing basic
mass usefull section in takeover mechanical
process of solicitations;
- isollating effect of basic metallic mass
portions or areas, favorizing a distribution and a
unequal takeover discontinuous of solicitations,
especially in graphite interdendritic distribution in
compact or discontinuous network form;
- focussing effect due to values of 10 to 100
times bigger regarding the medium value of
mechanical tensions, especially in lamellar spread
peaks area, causing premature rupture during
tablets.
78
The mechanical characteristic values indicate
the total unfavorable influence of lamellar graphite
with sharped peaks with maximal cutting effect
on rupture restistence (Rm=100...350MPa), and
also on specific elongation (A<0,5%), those cast
irons being considered fragile[4].
The experimental researches and current
practics showed that the most resitent cast iron is
the one with perlitic metallic mass, that contains
the most fine and uniform spread graphite.
The most important metallurgical processes
leading to increasement of mechanical resistence
and tenacity of cast iron are following, by primary
structure modication, the purposes given below:
a) graphite modification, by:
- obtaining the most compact form;
- the most uniform distribution into metallic
mould;
- dimensions decreasement of graphite
spreads;
- decreasement of graphite quantity.
b) metallic basis modification, by:
- crystalline refinement (grains dimension
decreasement and consequently grains number
increasement on volum unit);
- percentage decreasement of unmetallic
phases and segregations due to grains limit.
I n conclusion, for a better characterization of
cast iron mechanical behaviour is necessary an
interdependence analysis between structure and
properties.
In this paper authors have analyzed the quality
of cast irons produced by SC FONTUR SA
Suceava Company, following the purpose of
finding the technical solutions that would lead to
the increasement of its strength.
2. Experimental researches
S.C. FONTUR Suceava S.A. Company
regularly delivered banks of cast samples,
according to SR EN 561 from which have been
made:
- samples for chemical and microstructural
analysis, and hardness control;
- samples for tensile test.
2.1 Chemical analysis
Chemical analysis has been made by S.C.
EXPROTERM S.A. Suceava on a Quantovac 3460
machine.
The results of chemical analysis are given in
tab.1.
In obtaining gray cast irons with lamellar
graphite, with superior mechanical strength, basic
elements content (C, Si, Ph, Mn, S) are chosen in a
way that assures a minimal quantity of graphite in
cristalization, and on the other hand, when the
eutectic transformation is finished, we have to
achieve a complete perlitic structure.
In cast iron characterization from chemical
compositions point of view due to Fe-C binary
diagram are beeing used two notions: equivalent
carbon and eutectic degree of saturation. Those are
measures that allow position determination of a
cast iron given in relation to Fe-C eutectic.
The most common formula in equivalent
carbon determination is [5]:
C
ech
=C
t
+0,3 (Si +P)
where Ct is total carbon from cast iron
(chemical determinated). Cech of cast iron fixes its
position due to stabile eutectic (4,26%) or
metastabile (4,30%) of Fe-C alloys.
Eutectic saturation degree SC is another way to
put a given cast iron in relation to Fe-c eutectic in
silicon and phosphorus presence, the eutectic
saturation degree being equal to 1. Saturation
degree can be determined with one of the
following formulas:
Tab.1.- Chemical composition of analized banks
No.
crt.
Date
2010
Chemical composition, %
C Si Mn Ph S Cr Mo Ni A1 Co Pb Sn Ti V W
1 20.04 3.79 1.75 .436 .113 .153 .119 .048 .092 .005 .236 .015 .020 .026 .017 .007
2 25.06 4.01 1.95 .447 .085 .152 .155 .048 .104 .005 .341 .013 .027 .024 .027 .007
3 31.07 3.92 1.86 .414 .102 .148 .151 .049 .101 .005 .197 .011 .021 .023 .015 .007
4 25.08 4.00 1.88 .413 .085 .126 .161 .050 .101 .005 .319 .011 .028 .022 .025 .007
5 06.09 3.86 1.88 .407 .075 .125 .146 .041 .098 .004 .203 .009 .016 .022 .015 .007
6 07.10 3.95 2.05 .428 .059 .089 .132 .041 .106 .023 .141 .010 .017 .025 .015 .007
79
E C
E t
C
C C
C C
S
=
or
) ( 3 , 0 26 , 4 P Si
C
C
C
S
t
C
t
C
+
= =
where CC =4,26-0,3 (Si +P) is eutectic carbon
content.
The equivalent carbon, eutectic carbon and the
eutectoid one are calculated with relations given
below:
The equivalent carbon:
C
echiv.
=C +0,3( Si +P ) 0,03 Mn +0,4 S +
0,07Ni +0,05Cr +0,074Cu +0,25Al
The eutectic carbon:
Ceut. =4,26 - 0,3( Si +P ) +0,03 Mn - 0,4 S -
0,07Ni - 0,05Cr
C'eut. =4,30 - 0,3( Si +P ) +0,03 Mn - 0,4 S -
0,07Ni - 0,05Cr
The eutectoid carbon:
C
s
=0,68 0,15Si 0,05(Ni +Cr +Mn 1,7S)
C
s'
=0,80 0,15Si 0,05(Ni +Cr +Mn 1,7S)
Saturation degree in carbon :
S
c
=C / [ 4,26 - 0,3( Si +P )]
Cast iron structure moulded in samples with
different diameters is determined into the diagram
by knowing the limits of structural domaines due
to K characteristic and by sample diameter, shown
in tab.3.
This constant is given by formula:
|
.
|
\
|
+
=
Si C
Si K
3
5
1
4
3
Examining the determinated values for carbon
content results that those are bigger than
recomended to cast irons Fc200 si Fc250:
- sulfur and phosphorus contents are above
maximal limits normally recomended by standards,
-
-
most of it due to usement of waste from cast iron
radiators as raw material, in which content is
allowed more phosphorus to increase liquid cast
iron fluidity;
- equivalent cast iron carbon is above 4,26%,
which leads to primary and coarse graphite
appearance, and consequently to decresement of
mechanical strength and cast iron tenacity.
- calculated K constant is between limits
0,85...2,05, which coresponds to perlitic cast iron,
fact that results of metalographic analysis of
analyzed cast iron structure.
2.2 Metalographic analysis
From each bank have been made
metalographic samples with u20x25mm that had
been polished, glossed and chemical attacked with
nital.
2.2.1 Bank 20.04.2010
a no chimical attack
Tab.2.- Maximal saturation value and of K characteristic in carbon
No.crt. C
echiv..
C
eut.
C
eut.
C
S
C
S
K
1 4,428 3,641 3,681 0,398 0,516 0,812
2 4,707 4,035 4,075 0,366 0,486 0,940
3 4,583 3,920 3,960 0,379 0,499 0,883
4 4,665 3,334 3,374 0,375 0,495 0,902
5 4,513 3,206 3,243 0,376 0,496 0,886
6 4,628 3,271 3,311 0,351 0,471 0,985
80
b attack with nital 4%.
Fig.1.- Sample microstructure of bank 20.04 2010
Cast iron structure: P + F + G; where P perlite; F
ferrite; G graphite.
Microstructure shown in fig.1.a presents
lamellar graphite as forms of isolated separations
in metallic mold.
2.2.2 Bank 25.06.2010
a no chemical attack;
Cast iron structure: P + F + EP + G; where P
perlite; F ferrite; EP- phosphorous eutectic;
G graphite.
associated with punctiform graphite, in light
coloured metallic mold.
2.2.3 Bank 31.07.2010
Tab.3.- K for different structural domaines
Sample diameter,
mm
K
a
motley
K
perlitic
K
b
perlito-ferritic
30 0,65-0,85 0,85-2,05 2,05-3,10
20 0,75-1,10 1,10-2,25 2,25-3,40
10 1,05-1,50 1,50-2,35 2,35-3,50
81
Cast iron structure: P + F +EP +G
Where P perlite; F ferrite; EP- phosphorous
eutectic; G graphite.
associated with punctiform graphite, in light
coloured metallic mold. Graphite arched and
semiarched. Microstructure shown in fig.3.b
contains: perlite (dark colour), 94,26%; ferite
(light colour), under 2%; phosphorous eutectic
shiny white coloured grains, very well shaped;
lamellar and punctiform graphite.
2.2.4 Bank 25.08.2010
a no chemical attack
Cast iron structure: P +EP +G
prevalant lamellar graphite as forms of isolated
separations associated with punctiform graphite, in
light coloured metallic mold.
2.2.5 Bank 06.09.2010
Fig.5.- Sample microstructure of bank 06.09.2010
Cast iron structura: P +EP +G
Where P perlite; EP- phosphorous eutectic;
G graphite.
prevalant lamellar graphite as forms of isolated
separations associated with punctiform graphite, in
light coloured metallic mold.
2.2.6 Bank 07.10.2010
82
Fig.6.- Sample microstructure of bank 07.10.2010
2.3 Mecanical tests tensile and hardness
Of each given sample bank had been made:
- tensile samples with calibrated part diameter
of 20mm, calibrated part length of 100mm.
- samples for hardness determination, same
used in metallography.
Tensile test was made on universal machine of
FIMMM mechanical testing laboratory, University
Stefan cel Mare, din Suceava.
Number of samples : at least 3 of each bank .
Gross molded samples diameter: 30 mm.
Tensile samples diameter, after chip removing
process: 20 mm, as standard SR EN 561.
Hardness samples same prepared in
metalography.
Mechanical test results are given in tab.5.
Tensile tests were done as standards SR EN
561. on processed samples with 20 mm.
Hardness has been measured universal harness
testing machine CV-700 with ball penetrator of 2,5
mm, downforce 1839 N, F/D
2
=30, where D is ball
diameter.
Tensile and harness test results are given in
tab.5.
Based on numerous determinations it could be
established several correlations between saturation
degree, tensile strength and cast iron hardness:
2
9 , 36 139 122
30
+ =
C C r
S S o
C r
S 5 , 82 102
30
= o
or
C r
S 80 6 , 100
30
= o
HB
30
=538-355 S
C
or
HB
30
=465-270 S
C
where
30
r
o
si HB30 are respectively tensile test
strength and Brinell hardness on standard rods with
diameter of 30 mm.
Cast iron structure: P +F +G
Microstructure shown in fig.6.a presents prevalant lamellar graphite as forms of isolated separations
associated with punctiform graphite and annealing graphite, in light coloured metallic mold.
In quantity determination of graphite, perlite, ferrite and phosphorous eutectic, were used
microstructures captured with OPTIKA digital camera of Science and Material Engineering Laboratory, of
which overlapped measurement grids transposed by camera program. The quantity of each constituent has
been determined by points method:
Q
constituent
=
100
t
c
N
N
(%)
where: N
c
number of points (nodes or intersections) that fall into determined inner constituent; N
t

total network number of nodes or intersections.
Quantities of each constituent, for cast iron samples of each analized bank are given in tab. 4.
Tab.4.- Experimental determined values on constituents quantities of analyzed cast iron structure
No.
crt.
Date/2010 Structural constituents quantities, %
Q
F
Q
P
Q
G
Q
E
1 20.04 4,72 90,59 4,69 -
2 25.06 2,34 92,70 3,03 1,93
3 31.07 - 94,26 3,92 1,82
4 25.08 - 93,76 4,12 2,12
5 06.09 4,40 91,24 4,36 -
6 07.10 4,72 90,79 3,40 1,09
83
On this aspect, Namur [6] proposed as quality
apreciation cast iron criteria the mechanical
massiveness invariant:
]. / [
2
3
2
kg mm
HB
i
r
o
=
That is characteristic to a chosen cast iron,
because it does not depend on sample dimension
onto which is being measured of (on condition that
or si HB have to be measured on the same sample).
The following paragraphs presents tipical
values for i (in m2/kg, that eliminates 10-6 ):
- cast irons for stoves with 1-1,5% P
i=20-40;
- semiphosphorous cast irons (0,5-0,9% P)
i=35-50 ;
- unmodified hipoeutectic hematite cast iron
i=50-75;
- hematitical cast irons modified with 0,1-
0,2% SiCa i=65-90;
- perlitic mechanic cast iron i=80-110;
- cast irons with 0,25% Ti si 0,25% V
modified i=90-120;
- soft steel i=l100-1300.
At the same saturation degree, tensile test
strength can vary into a wide domain and it
allowed to Patterson [7] to propose a new
criteria of quality cast iron apreciation namely
relative strength RR or ripe degree (Reifegrad):
C
r
S
RR
5 , 82 102
=
o
or
C
r
S
RR
80 6 , 100
=
o
,
where counter (or) represents determined test
strength of chosen cast iron, and denominator
represents theoretical strength. Quality index
shows the increasement or decreasement of given
cast iron strength due to calculated medium value.
Variation limits of relative strength are between
0,6-1,3 and in vacuum overheated cast irons, this
may decrease up to 0,4. In the same way we can
define another index, called relative hardness,
RH:
C
S
HB
RH
355 538
=
or
C
S
HB
RH
270 465
=
,
that shows how much given cast iron hardness
varies due to medium value, theoretically
calculated as on formula .
Combining the ecuations we can determine
relative hardness due to strength:
Tab.5.- Tensile and hardness test results
No.
crt.
Tensile Brinell hardness, daN/mm
2
Cast iron brand as
standard
SR EN 561
Sample
diameter,
mm
Breaking
strength,
(N)
Rm,
(MPa)
Strengths
medium value,
(MPa)
Mesured
values
Samples
medium
value
Burden
yields
medium
value
1
1 20 68452 218
216,66
199 ;192 ;211 200,6
201,9 FGL200
2 20 70336 224 206 ;210 ;212 209,3
3 20 65312 208 194 ;198 ;196 196,0
2
1 20 76105 242.37
231,42
248; 244, 252 248
248 FGL250
2 20 68314 217,56 246; 244, 254 248
3 20 73650 234.35 240; 256, 252 249
3
1 20 60393 192.33
203,27
204 ;202 ;211 205
203 FGL200
2 20 50882 159.49 206 ;210 ;212 209
3 20 81015 258.00 194 ;198 ;196 196
4
1 20 76105 242.37
242.37
287 ; 308,297 297
254 FGL250
2 20 79542 253.31 248 ;234 ;241 241
3 20 84877 270,31 214 ;234 ;224 224
5
1 20 60882 193,89
206,80
300;356 ;328 328
256,6 FGL200
2 20 70704 225.17 246 ;249 ;247 247,5
3 20 63224 201,35 187 ;202 ;194 194,5
6
1 20 79542 253.31
237,10
287 ;308 ;297 297
254 FGL250
2 20 75400 240,12 248 ;234 ;241 241
3 20 68450 217,89 214 ;234;224 224
84
r
HB
RH
o 3 , 4 100+
=
or
r
HB
RH
o 4 , 3 125+
=
.
Value of RH can vary between 0,8-1,3 and
decreases by modification. Cast iron overheated
with lower RH, leads to its increasement.
Because lower hardness of a given strength is
an index of cast iron superior quality (greater
plasticity and tenacity, better processability),
general cast iron quality index (I C) is the
relation between relative values of cast iron
hardness and strength, for example:
2.4 Conclusions
From chemical composition analysis results
following conclusions:
- carbon content is higher than recomended
for cast irons Fc200 si Fc250;
- sulfur and phosphorus contents are above
maximal limits recomended by standard
procedures, most of it due to usement as raw
material in processing waste from cast iron
radiators, being admitted in their contents more
phosphorus to increase liquid cast iron fluidity;
- cast iron equivalent carbon is above
4,26%, that leads to primary and coarse graphite
appearance and consequently to a decreasement of
mechanical cast iron strength and tenacity.
- calculated K constant is between limits
o,85...2,05, that corresponds to a perlitic cast iron,
fact resulted from metalographic analysis of
analized cast iron structure.
- analyzed cast iron structure is mostly
perlitic, ferritic content being reduced. This
structure offers good mechanical strength to
processed cast irons.
- Graphite has lamellar shape and its
distribution as form of isolated separations; some
samples also have punctiform and annealing
graphite;
- Banks 2, 3, 4 si 5 have phosporous
eutectic in their structure, constituent that weaken
cast irons, reducing their mechanical strength to
shocks.
By determined strength, cast irons of
processed banks fit into standardized brands :
.
5 , 82 102
355 538
C
C r
S
S
HB RH
RR
IC
= =
o
Cast iron quality is improved as relative strength value increases and relative hardness decreases.
By knowing different technological parameters (chemical composition, eutectic cell number, relative
thickness of wall piece) we have to determine major mechanical characteristics based on different empiric
formulas established by processing numerous tests (tab.6).
Tab.6.- Mechanical characteristics calculation of cast iron with lamellar graphite (Fgl) based on
simplified relations[8]:
The characteristic Measuremen
t unities
Calculation relation Observations
Flowing limit, R
p0,2
MPa R
p0,2
=(0,75-0,85)R
m
R
m
breakage strength, Mpa
Tensile breakage
strength on rods of
30, R
m30
MPa R
m30
=122 139 S
c
+36,9 S
c
-
1
S
c
saturation degree in
carbon
) ( 3 , 0 26 , 4 P Si
C
S
c
+
=
Idem, R
m30
MPa R
m30
=102 82,5 S
c
Idem, R
m30
MPa R
m30
=100,6 80 S
c
Hardness HB
30
MPa HB
30
=538 355 S
c
Idem
Idem HB
30
MPa HB
30
=465 270 S
c
Tensile strength on
a sample or on a piece
with unknown diameter
(=X, R
mX
)
MPa R
mX
=R
m30
(HB
X
/ HB
30
)
3/2
R
m30
, HB
30
known for
samples 30; also, R
mX
known
for sample =X.
Tensile strength, R
m
MPa R
m
=7(13,5 2CE 2,3 log g) CE equivalent carbon, %;
G wall piece thickness, mm
85
- FGL 200 (Fc200) banks 1 ; 3 i 5 ;
- FGL 250 (Fc250) banks 4 i 6.
Recommendations:
- processing temperature increasement in
cupola ;
- use of air enriched with oxygen in
processing;
- supplementary desulphurisation and
dephosphorusation;
- changing processing aggregate, purchasing
an induction furnace for processing that will assure
superior qualities of processed cast irons and the
posibility of modifying those.
2.5. References:
1.- Piwowarski, E., Fonte de nalt calitate,
Editura tehnic, Bucuresti, 1967.
2.- Sofroni , L., Stefnescu, M.D., Fonte
modificate cu proprietti superioare, Editura
Tehnic, Bucuresti, 1971.
3.- Sofroni , L., Riposan, I., Chira, I., Fonte
albe rezistente la uzur, Editura Tehnic,
Bucuresti, 1987.
4.- Sofroni , L., Stefnescu, M.D., Fonte
speciale, Editura Tehnic, Bucuresti, 1974.
5.- Sengupta, P.K., J ordan, D.E., Die
Scheweibarkeit von austenitischen Gusseisen mit
Kugelgraphit. In> Giesserei, nr. 26, 1971,
p.801805.
6.- Volianik, N., Contribution a letude de la
graphitisationdes fontes malleables. n: Revue de
Metallurgie, Iunie 1967, p 567584.
7.- Iatan, N, s.a.- Consideratii privind influenta
principalelor elemente de aliere asupra durittii
fontelor destinate turnrii cilindrilor cu crust
extradur. n ;: Metalurgia, nr.9, 1971,
p.556559.
8.- Henke, F., Verschleissbestandige weisse
Gusseisen. n: Gusseiesen Praxis, nr.1, 1973,
p.121, nr.2,1973, p.3240, nr.3, 1973, p.123,
nr.4, 1973, p.6974.
9.-Farge, J .C.T., Chollet,P.,Yernaux, J . Effect
of composition, cooling rate and heat treatment on
properties of new-resistant White iron. n :
Foundrz Trade J ournal, nr. 2836,1971, p. 319-327
86
87
OPTI MI ZATI ON OF THE PL ASTI C I NJ ECTI ON PROCESS
THROUGH THE MODI FI CATI ON OF THE PROCESS FUNCTI ONAL
PARAMETERS
Teodor - Daniel Mndru, Ciprian Dumitru Ciofu &
,
Dumitru Nedelcu
Gheorghe Asachi Technical University of Iasi, Faculty Machine Manufacturing and Industrial
Management, e-mail: mindru.daniel@gmail.com
Abstract: The paper presents the optimization of the plastic injection process through the reduction of
the cooling and removing times. For the deployment of the experiment the PA 6 polyamide will be used.
The SolidWorks and Moldflow software packages will be used for CAD and process simulation, the
material flow into the mold, as well. The experiment aims to obtain the optimal variant of PA 6 polyamide
injection through the modification of the mold as well as the melt temperature in order for the mold
filling to be complete and the cooling time to be as short as possible, therefore generating an increase of
productivity.
Keywords: cooling channel, flow, melt, mold, temperature
1. Introduction
In order to plan a plastic injection process was
taken into consideration that the most important
technological parameter is the temperature of the
mold. We may consider that the process develops
under normal condition only the temperature of the
mold is stationary and may be controlled.
Depending of the plastic materials used, the
injection molds can be heated [1].
The plastic injection process is a cyclic one,
formed by several operations such as:
- The gauging [2];
- The heating and the melting of the material
in the machine cylinder;
- The mold clamping;
- The mold charge with under pressure
material;
- The solidification and the cooling of the
material;
- The opening of the mold;
- The removal of the injected piece from the
mold.
This paper will study the injection operation as
well as the charging and the cooling operations of
the material from a mold through the optimization
of the inlet temperatures of the injection process.
Cooling of the injected piece from the
maximum value of the temperature to the room
temperature value is an operation determined by
the thermal conductibility of the plastic material.
After the mold is opened, the cooling operation
continues. In plastic injection cycle, the cooling
times are the longest, getting up to 68% of the total
period of the process [2].
The injection technology aims to obtaining the
shortest cooling time possible that assures the
prescribed quality of the piece.
The figure 1 presents the temperature variation
inside a plastic component during the cooling
operation.
Figure 1. Temperature variation inside a plastic piece
during cooling operation; T
c
environmental
temperature, T
D
mold release temperature; T
r

cooling temperature.[2]
88
As a result of the analysis of figure 1, we may
express the calculus formula of the total time of the
injection process as:
d r u t
t t t t (1)
Take into consideration the relation (2):
d r u
t t t (2)
K t
d
(3)
and replacing relations (2) and (3) into relation
(1), this one may be written as:
K t t
r t
(4)
t
t - Cycle time
u
t - Injection time
r
t - Cooling time
d
t - Time Cleaners
K - Constant
2. Simulation process, results and discussions
The cooling system of the mold must be
designed according to the overall dimensions of
the injected parts or, as the current analyzed
situation, when the injection is made in cavity
distribution channels, thus obtaining six
components, the cooling channels must cover the
entire surface of the model.
The piece used for process optimizing is
presented in figures 2a and 2b. The figures are
realized using the SolidWorks software package.
a)
b)
Figure 2: The bushing type piece drawing, a) 2D
Model, b) 3D Model
The six cavities along the distribution channels
and their placement in the mold are represented in
Figure 3.
It was considered the cold type of sprue. The
initial parameters are as follow:
- Conical ledge: 23 mm;
- Inlet diameter: 5 mm;
- Outlet diameter: 9 mm;
- Semicircular runners type: 10 mm;
- Conicalgate: inlet diameter 5 mm. outlet
diameter 3mm;
- Dimensions of mold active parts:
165x105x50mm;
- Used material for mold: P20.
Figure 3: The 3D general view fo injection mold
The material used during the simulation is PA6
polyamide. In figure 4 it may be seen the variation
of the viscosity depending on the working
temperature. In figure 5 is presented the variation
of the specific volume by temperature.
Figure 4: The viscosity variation by temperature
Figure 5: The specific volume by temperature
In order to obtain accurate components, special
attention must be applied to the placement of the
cooling channels in relation to the component, to
89
the injection sites, the filling direction of the mold
etc. For a better cooling there were realized
several types of runners: with rectangular shaped
section, square shaped section and triangular
shaped section. As a result of the experiments, it
was noticed that the most efficient runners are
experimentally the square shaped ones and
theoretically the rectangular ones, see figure 6, a
and b. Constructively speaking, the easier to make
are the circular ones, because they are just drilled,
see figure 7 [2].
Figure 6: Runners. a) Rectangular shaped runner; b)
Square shaped runner.
Figure 7: Drilled runners
The most often used cooling circuit is made out
of straight runners. The connections between the
runners are made with the help of supplementary
runners, applied to the mold plates. The role of the
runners in the injection process is to maintain the
temperature of the mold plates at the established
value, without considerable variations.
The injection processs inlet parameters are:
mold temperature 70-110
o
C, melted material
temperature 230-300
o
C and maximum injection
pressure 180 MPa.
2.1 Process simulation
The analysis with Autodesk Moldflow of the
injection process offers the possibility of
visualizing the process evolution without being
necessary to build the mold and to solve the
eventual problems that can occur in the designing
process of new parts [3].
Due to the fact that the simulation is made for
an injection cycle, the experimental plan will not
take into consideration the runners but only the
temperature variation of the mold and of the
melted material.
The mechanical properties of the material used
for the simulation part are presented in table 1. In
table 2 are presented the process parameters
recommended by the manufacturers [4].
Table 1 The mechanical proprieties of the used plastic
material
Elastic modulus 2910 MPa
Poisson ratio 0.386
Shear modulus 1050
Table 2The process parameters
Moldtemperature range minimum 70
o
C
Moldtemperature range maximum 110
o
C
Melt temperature range minimum 230
o
C
Melt temperature range maximum 300
o
C
Absolute maximum melt temperature 340
o
C
Ejection temperature 133
o
C
Maximum shear stress 0.31 MPa
Maximum shear rate 100000 1/s
Table 3 highlights the parameters of the
simulation and the number of experience. As it
may be noticed, during the first nine experiments
the temperature of the melted material is constant
and we varied the temperature of the mold. During
the following 8 experiments the temperature of the
mold is constant and we varied the temperature of
the melted material 10
o
like step adh 5
o
for mold
heating temperature.
Table 3Parameters and experiments number
Exp.
Nr.
Mold
temp.
[
o
C]
Melt
temp.
[
o
C]
Exp.
Nr.
Mold
temp.
[
o
C]
Melt
temp.
[
o
C]
1 70 265 10 90 230
2 75 265 11 90 240
3 80 265 12 90 250
4 85 265 13 90 260
5 90 265 14 90 270
6 95 265 15 90 280
7 100 265 16 90 290
8 105 265 17 90 300
9 110 265
The paper presents the highlights experiments
from table 3
2.2 Results of the simulation
Following the simulation, we may analyze the
result presented in tables 4, 5, 6, 7 and 8 as well as
in the figures attached to each experiment.
90
Table 4The simulation results
Experiment 1 Figure 8,9,10
Actual filling time 0.32 (s)
Actual injection pressure 28.440 (MPa)
Clamp force area 24.8234 (cm^2
Max. clamp force during filling 5.216 (ton)
Velocity/pressure switch-over at
% volume
99.02 (%)
time
0.31 (s)
Estimated cycle time 7.38(s)
Total part weight 11.028 (g)
Shot volume 18.4877 (cm^3)
Cavity volume 11.8609 (cm^3)
Runner system volume 6.6268 (cm^3)
Figure 8: The fillingtimedistribution
Figure 9: The temperatureat flow front
Figure 10: The pressure at end of fill
Actual filling time 0.31(s)
Actual injection pressure 26.145(MPa)
Max. clamp force during filling 4.862(ton)
% volume
99.11 (%)
time
0.31 (s)
Total part weight 10.947(g)
Figure 12: The temperature at flow front
Figure 13: The pressureat end of fill
% volume
99.11 (%)
time
0.31 (s)
Estimated cycle time 7.36 (s)
91
Figure 16: The pressureat end of the
Actual filling time 0.70(s)
Max. clamp force during filling 25.222(ton)
% volume
99.00(%)
time
0.64(s)
Total part weight 11.724(g)
Figure 19: The pressureat end of thefill
% volume
99.09 (%)
Velocity/pressure switch-over at 0.16 (s)
time
Estimated cycle time 7.56 (s)
Runner systemvolume 6.6268 (cm^3)
Figure 22: The pressureat end of fill
2.3 Discussions
As a result of the experiments we can draw
graphics for the variation during the injection
process. In the figure 23 is presented a graphic for
the first 9 experiments, during which the
temperature of the melt temperature is constant.
The figure 24 presents the graphic display of the
variation during the last 8 experiments, for which
the temperature of the mold is constant.
Figure 23: The variation of the process time by the
mold temperature.
92
Figure 24: The variation of the process time by the melt
temperature.
Analyzing these graphics we may observe that
the optimal variant for designing the injection
process is by setting the technological parameters
to: melt material temperature, 265
o
C mold
temperature, 90
o
C. The optimal values were used
while performing five experiment. The values for
all the inlet parameters of the process may be
found in table 5. The simulation may be visualized
in figures 11, 12 and 13.
If go back to Eq. 1, and set the experimental
data into this formula, the cooling time for the
injection process using PA 6 polyamide may be
expressed as:
d r
t t 32 . 0 63 . 7 (5)
Applying Eq. 3, we obtain:
K t
r
31 . 7 (6)
Following there are several information
resulted from the simulation of number 5
experiment during the multi-cavity injection
process.
Figure 25 graphically presents the variation of
the temperature during the injection process. The
difference between the maximum value of the
temperature and the minimum is 3,735
o
C.
Figura 25: The temperature variation
Figure 26 presents the variation of the cooling
time. The difference between the maximum and
the minimum obtained values is 0,5337 s.
Figura 26: The cooling time distribution
During the simulation of the 10
th
and the 11
th
experiment appeared a production fault, named
incomplete filling, presented in figure 27. As a
result of these two experiments, we had wastage.
Figure 27: The incomplete filling
3 Conclusions
After the experimental research the main
conclusions are as follow:
- The time of deployment for an injection
process is increasing by maintaining constant the
temperature of the melting material and the
increase of the mold temperature;
- When the mold temperature is constant and
the melting material temperature is increasing, the
deployment time is decreasing;
- The injection time, as minimal value could be
obtain using the overlap of figure 23 and 24 and
thuts result the process parameters;
- The obtained results confirm that the injection
pressure decreases by the increase of the injection
time if the melted material temperature is constant.
The value of the pressure in this case decreases
from 28.440 MPa down to 26.145 MPa;
- If the mold temperature is constant, the
injection pressure decreases by the melted material
temperature from 111.486 MPa down to 23.683
MPa.
Acknowledgement
This paper was realised with the support of POSDRU
CUANTUMDOC DOCTORAL STUDIES FOR
EUROPEAN PERFORMANCES IN RESEARCH
AND INOVATION ID79407 project funded by the
European Social Found and Romanian Government.
4 References
[1] Fetecu, C., Injectarea materialelor plastice,
Bucharest, 2005, pp 127 - 156
[2] eres, I., 1996, Injectarea materialelor
termoplastice, Imprimeria de Vest Publisher, Oradea
(1996), pp 254 - 300
[3] Fetecau C., et all. Overmolding injection molding
simulation of tensile test specimen, International
J ournal of Modern Manufacturing Technologies,
Vol. II, No. 2/2010, Politehnium Publishing House,
ISSN 2067-3604.
[4] *** Plastics Data Charts - www.plastemart.com
Accesed: 23.12.2010
93
A PROBABI LI STI C DYNAMI C MODEL ASSOCI ATED TO A
J ACKSON NETWORK, WI TH SERI ES QUEUES;
THE SOLUTI ON OF THE ASYMPTOTI CALLY STABLE STEADY
STATE
POPA MARI N, DRGAN MI H, POPA MARI ANA
University of Bucharest, Technology Department, marpopa2002@yahoo.com
University of Bucharest, Technology Department, dragan_mihaita@yahoo.com
Abstract: This article studies in detail a subclass of Jackson networks class, namely the subclass of
computer networks with queues in series and comes to prove a theorem which is called the Jackson's
theorem, which provides a formula that gives an analytic expression of the probability distributions for
the asymptotically stable equilibrium state of this subclass.
The solution of this model in asymptotically stable equilibrium state will provide, at every moment,
the probability that in the network nodes to be a certain number of transitions, in the waiting tail of the
node or in processing by the server of the respectively network node.
Keywords: computer network, JACKSON network, network with series queues, server, sorting process.
1. Preliminary
The behavior of computer networks is
characterized by the presence of some congestion
points of transitions, called network nodes [1, 2].
In every network node forms a waiting queue
where the transitions arrived in the node wait to be
selected according to the discipline associated to
the waiting queue, in order to be processed.
So we consider a network node as an entity made
of a server or processing unit and a waiting queue,
depicted as in Fig. 1[1].
Collection of such network nodes and interaction
between them forms a computer network with
waiting queues.
Definition 1.1 Its called J ackson network a
computer network with waiting queues where
every server is preceded by a waiting string and
where every transition, after has been processed by
the server, is sent in the queue of another server.
The name of such a network class comes from
the name of the one whome, using stochastic
processes, like multi dimensional processes of
birth and death, discovered the solution of
mathematic model associated to the net in a
simplified shape of product.
J ackson network class contains subclasses of
network with series queues, with parallel queue,
acyclic, with feedback and with queues in local
balance.
Well build a mathematic model associated to a
J ackson type computer network, obtaining a
dynamic probabilistic model.
The solution of this model will give, in every
moment, the probability that in networks nodes to
arrivals
Waiting queue server
departures
Figure. 1 Network node
94
be a number of transitions in the waiting queue of
the node or in processing by the server of net node.
2. Theoretical results
The main result of the article requires the
following result due to Burke's demonstration,
whose demonstration is in [2].
Theorem 2.1. If in a net node the process of
transition arrivals is Poissonian with parameter and
the process of transition processing is Poissonian
with parameter, then the process of transition
departures of net node is Poisson with
parameter.
Theorem 2.2 We consider L a computer network
made of N net nodes in series, to which the influx
transitions is Poissonian with rate and the
processing flow in net node i is Poissonian with rate
i
, i= N , 1 . If utilization factors , ) 1 , 0 e =
i
i
for
every i= N , 1 , then the solution of model network in
series proper to the asymptotically stable steady state
is: p(n
1
,n
2
,.n
N
) = , )
[
=

N
i
n
i i
i
1
. 1
Proof
We note , ) t
i
Q the number of transitions in node i
at time N i , 1 , = t [1]
Sought to determine the probability that at time
t
to haven
i
transitions in node i, N i , 1 = , that is the
probability:
, ) , ) , ) , ) , )
N N N
n Q n Q n Q P n n n P = = = = t t t t ,..., , , ,..., ,
2 2 1 1 2 1
In the asymptotically stable state for any
N i n
i
, 1 , 0 = > there is , ) t
t
, ,..., , lim
2 1 N
n n n P

and
is finite, value noted , )
N
n n n p ,..., ,
2 1
.
The probability distribution , )
N
n n n p ,..., ,
2 1
,
proper to the asymptotically stable steady state,
shows a description of the computer net average
behavior on long-term.
We note
N
n n n
S
,..., ,
2 1
the network state proper to
the case in which, for N i , 1 = , in node i there aren
i
transitions.
Thus , )
N
n n n p ,..., ,
2 1
= , )
N
n n n
S P
,..., ,
2 1
that is the
probability distribution of the asymptotically stable
steady state represents the probability that the
network be in state
N
n n n
S
,..., ,
2 1
. Will obtain these
probabilities considering the mathematical model
associated to the network as a stochastic process type
birth and death multi-dimensional.
Next, we demonstrate the assertion from
enunciates through induction after N(N>2).
Case N=2.
Consider a J ACKSON network with two
series queues, as in Fig.2 in which the transactions
arrived at the end of queue 1,after a Poisson flow of
rate , waits to be processed by server of node 1.
After a transaction is processed by this server, will
go at the end of the queue of node 2, where waits to
be processed by this nodes server, while server 1
selects a new transaction from its queue, according
to the order chosen for the queue. As per BURKE
theorem, the arrivals to queue of node 2 are also
Poisson of parameter +
1
, where
1
is the rate o
servers 1 processing. Thereby the arrivals to queue
of node 2 are a Poisson flow of rate +
1
. Because
the rate of transitions processing by server 2 is
2
according to BURKE theorem, transitions that leave
this network are a Poisson flow of rate :
+
1
+
2
.
Figure. 2.
Follow to determine the probability that at time
to haven
1
transitions in node 1 andn
2
transitions in
node 2, that is:
P(n
1
,n
2
,) =P(Q
1
() =n
1
. Q
2
() =n
2
).
Therefore, we will calculate successively the next
probabilities:
P(0,0, + A) = probability of not having
transitions in any node at timeA.
P(0,n
2
, + A) =probability of not having only
in node 2, n
2
transitions at time
A(n
2
_1).
P(n
1
,0, + A) =probability of having only in
node 1, n
1
transitions at timeA(n
1
_1).
P(n
1
,n
2
, + A) = probability of having n
1
transitions in node 1 and n
2
transitions in node 2 at
timeA
1 (
1
> n
si
). 1
2
> n
queue
2
queue
1
Node
of net 1
node of
net 2
Arrivals
poissonn
depa
rture
s
server
2
server
1
95
Using the hypothesis of a stochastic process type
birth death, we obtain the following:
P(0,0, + A) =P(0,0,) (1- A) +P(0,1,)
2

A (1- At) +0(A)
P(0,0, + A) = P(0,0,) P(0,0,)A +
2
P(0,1,)A +0(A)
=
A
A +
t
t t t ) , 0 , 0 ( ) , 0 , 0 ( P P
.
) ( 0
) , 1 , 0 ( ) , 0 , 0 (
2
t
t
t t
A
A
+ + = P P
Passing to limit after A t 0 obtain:
) , 1 , 0 ( ) , 0 , 0 ( ) , 0 , 0 (
2
t t t
t
P P P
d
d
+ =
P( 0,n
2
, A) =
=P(0,n
2
,)[1-A)[1-
2
At]+P(1,n
2
1,)
1
At[1-
-A)[1-
2
A]+P(0,n
2
+1,)
2
A[1-A)+0(A) .
t
t t t
A
A + ) , , 0 ( ) , , 0 (
2 2
n P n P
=
=-(
2
+)P(0,n
2
,)+
1
P(1,n
2
1,) +
2
P(0,n
2
+
+1,)+
t
t
A
A ) ( 0
.
Passing to limit after At 0 obtain:
t d
d
P(0,n
2
,)=-(+
2
)P(0,n
2
,)+
1
P(1,n
2
-1,)+
+
2
P(0,n
2
+1,)
Analogue :
P(n
1
,0,A)=P(n
1
,0,)[1A)[1
1
A]+
+P(n
1
1,0,)A[1-
1
A]+P(n
1
,1,)
2
A[1- A)[1-
1
A] +0(A)
t
t t t
A
A +
) , 0 , ( ) , 0 , (
1 1
n P n P
=
=-(
1
+)P(n
1
,0,)+P(n
1
-1,0,)+
2
P(n
1
,1,)+
t
t
A
A ) ( 0
Thus:
t d
d
P(n
1
,0,)=-(+
1
)P(n
1
,0,)+P(n
1
-1,0,)+
+
2
P(n
1
,1,)
Finally:
P(n
1
,n
2
,A)=P(n
1
,n
2
,)[1A)[1
1
A][1
2
A]+
+P(n
1
1,n
2
,)A[1
1
A][1
2
A] +
+P(n
1
,n
2
+1,)[1 A)[1
1
A]
2
At +
+P(n
1
+1,n
2
1,)[1- A)
1
A(1
2
A) +0(A)
=
A
A +
t
t t t ) , , ( ) , , (
2 1 2 1
n n P n n P
=-(
1
+
2
+)P(n
1
,n
2
,)+P(n
1
1,n
2
,t)+
+
2
P(n
1
,n
2
+1,)+
1
P(n
1
+1,n
2
1,)+
t
t
A
A ) ( 0
Passing to limit after A t 0 obtain:
t d
d
P(n
1
,n
2
,) =-( +
1
+
2
)P(n
1
,n
2
,) +
+P(n
1
1,n
2
,) +
1
P(n
1
+1,n
2
2,) +
2
P(n
1
,n
2
+1,)
Conclusively, for the model of computer network
with two series nodes, is obtained the following
system of differential equation:
) , 1 , ( ) , 1 , 1 ( ) , , 1 (
) , , ( ) ( ) , , (
) , 1 , ( ) , 0 , 1 ( ) , 0 , ( ) (
) , 0 (
) , 1 , 0 ( ) , 1 , 1 ( ) , , 0 ( ) (
) , , 0 (
) , 1 , 0 ( ) , 0 , 0 ( ) , 0 , 0 (
2 1 2 2 1 1 2 1
2 1 2 1 2 1
1 2 1 1 1
1
2 2 2 1 2 2
2
2
t t t
t t
t
t t t
t
t
t t t
t
t
t t t
t
+ + + + +
+ + + =
+ + + =
=
+ + + + =
=
+ =
n n P n n P n n P
n n P n n P
d
d
n P n P n P
n P
d
d
n P n P n P
n P
d
d
P P P
d
d
0 , 1 ) , , (
2
0 ,
1
2 1
> =
>
t t n n P
n n
(complimentarily
condition)
In general case, this system is extremely difficult
to resolve. In the stabile asymptotically balance is
known that for any n
1
,n
2
> 0 there is, and id finite :
) , ( ) , , ( lim
2 1 2 1
n n p n n P =

t
t
The probability distribution p(n
1
,n
2
) of stabile
asymptotically balance delivers a description of the
average behavior of the computer network long-
term. In this condition, the system above becomes:
1
0 1 1 1
1
0 1 0 1 0
0 1 0 1 1 0
0 1 0 0 0
2
0
1
2 1 2 2 1 1
2 1 2 1 2 1
1 2 1 1 1
2 2 2 1 2 2
2
2 1
=
= + + + +
+ + + +
= + + +
= + + + +
= +
>
) n , n ( p
) n , n ( p ) n , n ( p
) n , n ( p ) n , n ( p ) (
) , n ( p ) , n ( p ) , n ( p ) (
) n , ( p ) n , ( p ) n , ( p ) (
) , ( p ) , ( p
n , n

We consider:
p(n
1
,n
2
) =0, n
1
<0 or n
2
<0
1
p(0,n
2
) =0
2
p(n
1
,0) =0.
relations that are natural if we think of their
interpretation.
Thus
1
p(0,n
2
) is the probability of processing a
transition in node 1, but in this node we have no
transition, and so
1
p(0,n
2
) =0.
Analogue
2
p(n
1
,0) =0.
In these hypothesis, the system of the first 4
equations proper to the state of stabile
96
asymptotically balance is reduced to a single
equation[1, 2], which is:
p(n
1
1,n
2
) +
1
p(n
1
+1,n
2
1) +
2
p(n
1
,n
2
+1)=
=( +
1
+
2
)p(n
1
,n
2
)
From this relation, deduce that in case of stabile
asymptotically balance, the flow of each state is
conserved, that is for any state, the input flow
coincides with output flow of this state.
To this relation adds the obvious
relation 1 ) , (
0 ,
2 1
2 1
=
> n n
n n p .
If note
2 1
,n n
S the network state appropriate to the
case in which in node 1 there aren
1
transitions and in
node 2 there aren
2
transitions, in the relation above
see that the states to which and from which have the
transitions with state
2 1
,n n
S are:
1 , , 1 1 , 1
2 1 2 1 2 1
, ,
+ + n n n n n n
S S S
as inputs of state
2 1
,n n
S
1 , 1 , 1 , 1
2 1 2 1 2 1
, ,
+ + n n n n n n
S S S as outputs for
2 1
,n n
S .
The interaction of these states is given by diagram
in Fig. 3.
Fig 3
Because in any system in stabile balance, the flow
rate of input in a state is the same with the flow rate
of output from that state, and that is for any state of
the system, results that in our case, by writing this
relation of flows preservation for state
2 1
,n n
S ,
obtain:
Flow rate of input for :
2 1
,n n
S =p(n
1
1,n
2
)+
1
p(n
1
+1,n
2
-)+
2
p(n
1
,n
2
+1).
Flow rate of output from :
2 1
,n n
S =( +
1
+
2
) p(n
1
,n
2
).
By condition of flows preservation in
2 1
,n n
S obtain
relation:
p(n
1
-1,n
2
) +
1
p(n
1
+1,n
2
-1) +
2
p(n
1
,n
2
+1)==
( +
1
+
2
) p(n
1,
n
2
)
which is actually the relation equivalent to the
probability system in case of stabile asymptotically
balance.
Notice that if we put into a rectangular network
n
1
On
2
, the neighbor states for
2 1
,n n
S ,then every flow
rate is well targeted, which is: from left to right,
2
downwards and1 direction SE-NV.
Since in the considered computer network with
series queues, the two net nodes are independent,
J ackson [1] comes with the idea of looking for
solutions to the system corresponding to stabile
asymptotic balance as a product, that is
p(n
1
,n
2
)=p
1
(n
1
)p
2
(n
2
), wherep
i
(n
i
) is the probability
of having in nodei, n
i
transitions in state of stabile
asymptotic balance and so:
, ) , ) = =
> > 0
2 2
0
1 1
2 1
1 1
n n
n p si n p (*)
With this choice, the equation above becomes:
1
p
1
(n
1
+ 1) p
2
(n
2
1) + p
1
(n
1
1) p
2
(n
2
) +
2
p
1
(n
1
) p
2
(n
2
+1)=(
1
+
2
) p
1
(n
1
)p
2
(n
2
) (1)
For n
1
=n
2
=0 and using the hypothesis above,
obtain the equation:
1
p
1
(1)p
2
(-1) +p
1
(-1)p
2
(0) +
2
p
1
(0)p
2
(1) =
=( +
1
+
2
) p
1
(0)p
2
(0)
2
p
1
(0)p
2
(1) =p
1
(0)p
2
(0)
p
1
(0)[
2
p
2
(1) p
2
(0)] =0.
We have the following situation possible:
1.If p
1
(0) =0 in relation (1) aboven
1
=0 obtain:
1
p
1
(1)p
2
(n
2
- 1)+p
1
(-1)p
2
(n
2
)+
2
p
1
(0)p
2
(n
2
+1) =
( +
1
+
2
) p
1
(0)p
2
(n
2
)
Using the hypothesis above and p
1
(0)=0 obtain:
1
p
1
(1) p
2
(n
2
-1) =0
If p
1
(1) =0 is demonstrated by induction that
p
1
(n
1
) =0 for anyn
1
_ 0.
If p
2
(n
2
1) =0 results that p
2
(n
2
) =0 for anyn
2
>
0.
Both situations are false, because they contradict
relations (*) above.
2.If
2
p
2
(1)p
2
(0)=0p
2
(1)=
2
p
2
(0) (2)
With this value go in relation (1) after n
2
=0.
Have:
1
p
1
(n
1
+ 1)p
2
(-1) + p
1
(n
1
1)p
2
(0) +
+
2
p
1
(n
1
)p
2
(1) =( +
1
+
2
) p
1
(n
1
)p
2
(0)
Results: p
1
(n
1
- 1)p
2
(0) +
2
p
1
(n
1
)p
2
(1) =
=( +
1
) p
1
(n
1
)p
2
(0)
Replacing p
2
(1) with its value from relation (2)
obtain:
S
n11,n2+1
S
n1,n2+1
S
n1,n
S
n1+1
,n
S
n1-,n2
S
n1+1,n2-1
S
n1,n2-1
97
p
2
(0)[p
1
(n
1
1)+
1
)p
1
(n
1
)]+
2
p
1
(n
1
)
2
p
2
(0)=0
p
2
(0)[p
1
(n
1
1) p
1
(n
1
)] =0.
If p
2
(0) =0 is shown that above are obtained null
values for p
1
andp
2
, which is false.
Thus, results that for p
1
(n
1
) =
1
p
1
(n
1
- 1) for any
n
1
_ 1.
Writing this relation as :
) 1 (
) (
1
1
k p
k p
=
1
, k _ 1
and doing the product:
[
=
1
1
1
1
) 1 (
) (
n
k k p
k p
=
1
1
n
|
|
.
|
\
|
obtain
1
1 1
1 1
) 0 (
) (
n
p
n p
|
|
.
|
\
|
=
and so ) 0 ( ) (
1
1
1 1
1
p n p
n
|
|
.
|
\
|
=
(**)
From relation: , ) =
>0 1
1
1 1
n
n p
, )
>
>
=
|
|
.
|
\
|
= =
|
|
.
|
\
|
0 1
0 1
1
1
, 1
1
1 1
1
1
1
) 0 ( 1 0
n
n
n
n
p p
,
if
1
<1, so that the series is convergence.

Note
1 =
1
and call
1
utilization factor in net
node 1.
So for
1
e (0,1) havep
1
(n) =(1
1
)
1
n
,
n _ 0,(3)
Use this result in relation (1) after n
1
=0.
Obtain:
1
p
1
(1)p
2
(n
2
1)+p
1
(-1)p
2
(n
2
)+
2
p
1
(0)p
2
(n
2
+ 1)=
=( +
1
+
2
) p
1
(0)p
2
(n
2
).
From hypothesis result:
1
p
1
(1)p
2
(n
2
1)+
2
p
1
(0)p
2
(n
2
+1)(+
2
)p
1
(0)p
2
(n
2
) =0
Replacing p
1
with value found in relation (**)
obtain:
1
1
p
1
(0)p
2
(n
2
1)+
2
p
1
(0)p
2
(n
2
+1)
-( +
2
) p
1
(0)p
2
(n
2
) =0
p
1
(0)[p
2
(n
2
1)+
2
p
2
(n
2
+1)(+
2
)] =0
Weve seen above that p
1
(0)
=
0 and so:
[p
2
(n
2
1) p
2
(n
2
)] =
2
[p
2
(n
2
) p
2
(n
2
+1)], n
2
_ 1.
Writing this relation as:
[p
2
(k 1) p
2
(k)] =
2
[p
2
(k) p
2
(k +1)], k _ 1
and making the sum of k from 1 ton
2
obtain:
[ , ) , ) [

= =
+ =
2 2
1 1
2 2 2 2 2
1 ) ( ) 1 (
n
k
n
k
k p k p k p k p ,
relation equivalent with:
[p
2
(0) p
2
(n
2
)] =
2
[p
2
(1) p
2
(n
2
+1)]
But from (2) p
2
(1) =
2
p
2
(0). Replacing this
above and reducing the termp
2
(0) obtain:
p
2
(n
2
+1) =
2
p
2
(n
2
)
In doing the above we obtain successively:
, )
, ) , ) 0
0
1
0
1
2
2
2 2
2 2
2 2
2
1
0
2
2
2 2
2
2 2
2
2
p n p
) ( p
) n ( p
) k ( p
) k ( p
k ,
) k ( p
k p
n n
n
n
k
|
|
.
|
\
|
=
|
|
.
|
\
|
=
|
|
.
|
\
|
=
+
> =
+
from
, ) , )

> >
=
|
|
.
|
\
|
=
0 2 0 2
2
1 0 1
2
2
2 2
n n
n
p n p
, 1
1
) 0 (
2
2
2
0 2
2

=
|
|
.
|
\
|
=
>
n
n
p
if
2
<1 so that the series to be convergence.

Note
2
2
= and call
2
factor of utilization
in node 2.
Thus for
2
e(0,1) obtain:
n
n p
2 2 2
) 1 ( ) ( = , . 0 > n (4)
Form relations (3), (4) and from :
p(n
1
, n
2
) =p
1
(n
1
) p
2
(n
2
) obtain the solution for
the state of stabile asymptotic balance of the model
as being:
0 , ), 1 )( 1 ( ) , (
2 1 2 1 2 1 2 1
2 1
> = n n n n p
n n
,
where , ) , ) 1 , 0 , 1 , 0
2
2
1
1
e = e =
.
Case N>2.
Assuming now that the relation form the
statements true for 1 , 2 = N k .
For k= N - 1 take the first N 1 series nodes
and form the net E
1
for which we know the
solution:
p(n
1
,n
2
,.,n
N 1
) = , ) , 1
1
1
[
=

N
i
n
i i
i
according the
hypothesis of induction.
98
Thus E is obtain connecting in series E
1
with
net nodeN, as in Fig. 4.
Note
N
n n
S
,
state corresponding to :
n = (n
1
,,n
N 1
) andn
N
and
p(n,n
N
)=p(
N
n n
S
,
) = ) (
, ,...,
1 1 N N
n n n
S p

=
=p((n
1
,,n
N 1
),n
N
) = p(n
1
,,n
N1,
n
N
).
As passing from E
1
in net node N is made
through net node N 1, it can be built the next
diagram of flow between states (similar to case of
two series net nodes)(see Fig.5) in which
) 1 , ,..., ( 1
1 2 1
=
N N
n n n n and
) 1 , ,..., ( 1
1 2 1
+ = +
N N
n n n n
Write relation of flow preservation in state
N
n n
S
,
and for resolving the obtained system we
look for solutions like:
p(n,n
N
) =p(n)p
N
(n
N
).
The preservation relation is:
N 1
p(n +1)p
N
(n
N
1) +p(n 1)p
N
(n
N
) +
N
p(n)p
N
(n
N
+1) =( +
N 1
+
N
) p(n)p
N
(n
N
) (5)
S
n-1,nN+1
S
n-1,nN
S
n,nN+1
S
n+1,nN
S
n+1,nN-1
S
n,nN
S
n,nN-1
N-1
N-1
N-1
N-1
N
N
N
N
Fig. 5
Arrivals
poisson
depart
nodN
nod N-1
nod
1
L
1
L
Figure. 4
99
Using that p(n) =
i
n
i i
N
i
) 1 (
1
1
[
=
(results
from induction hypothesis) it will be obtained a
relation of recurrence of order I for p
N
and
therefore:
N
n
N N N N
n p ) 1 ( ) ( = , where
) 1 , 0 ( e =
N
N
.
Thus : p(n
1
,,n
N 1
,n
N
) = p(n,n
N
)=p(n)p
N
(n
N
)=
=
i N i
n
i i
N
i
n
N N
n
i i
N
i
) 1 ( ) 1 ( ) 1 (
1
1
1
=
[ [
=
=
.
Indeed, replacingp(n) in relation (5) obtain:
[
[
=

+ +
+ +
2
1
1 1
2
1
1 1 1
) ( ) 1 ( ) 1 (
) 1 ( ) 1 ( ) 1 (
N
i
N N N N i
n
i
N
i
N N N N i
n
i N
n p n p
n p n p
i
i

[
[
=

+ + =
= + +
2
1
1 1 1
2
1
1 1
) ( ) ( ) 1 ( ) (
) 1 ( ) ( ) 1 (
N
i
N N N N i
n
i N N
N
i
N N N N i
n
i N
n p n p
n p n p
i
i

We divide this relation to [
=
2
1
) 1 (
N
i
i
n
i
i
and
) ( ) 1 ( ) (
) 1 ( ) 1 (
) ( ) 1 (
) 1 ( ) 1 (
1 1 1
1 1
1
1
1
1
1
1 1
1
1
1
1
N N N
n
N N N
N N N
n
N N
N N N
n
N
N N N
n
N N
n p
n p
n p
n p
N
N
N
N

+

+ + =
= + +
+ +
+

Divide through ) 1 (
1
1
1
1
N
n
N
N
and obtain:
= + +
+ +

) 1 (
) ( ) 1 (
1
2
1 1
N N N N
N N N N N N
n p
n p n p

) ( ) (
1 1 N N N N N
n p

+ + =
Reduce the term ) (
N N
n p with
) ( ) (
1
1 1 1 N N
N
N N N N N
n p n p
, divide
with
1
1
=
N
N
and obtain:
+ =
+ +

) ( ) (
) 1 ( ) 1 (
1 1
N N N
N N N N N N N
n p
n p n p

)] 1 ( ) ( [
)] ( ) 1 ( [
+ =
=
N N N N N
N N N N
n p n p
n p n p
a relation similar to the one form which we

obtainedp
2
. Forwards, proceed as for p
2
and finally
obtain p
N
as being
N
n
N N N N
n p ) 1 ( ) ( = where
) 1 , 0 ( e =
N
N
.
In order to obtain this result, it was also needed
the relation ) 0 ( ) 1 (
N
N
N
p p
= which is get if in
relation (5) put n =0 andn
N
=0.
Obtain:
+ + =
= + +
) 0 ( ) 0 ( ) (
) 1 ( ) 0 ( ) 0 ( ) 1 ( ) 1 ( ) 1 (
1
1
N N N
N N N N N
p p
p p p p p p

) 0 ( ) 1 (
0 )] 0 ( ) 1 ( )[ 0 (
N
N
N
N N N
p p
p p p

=
=
because p(0) = 0.
Observations[1].
1) J ackson has shown that forma the product of
the final relationship of a mathematical model is
available also in a more general case, in which are
allowed transitions between states form:
N j
n n n
S
,..., ,...,
1

N j
n n n
S
,..., 1 ,...,
1
+
, a transition enters
from the environment directly to the queue of
processor j.
N i
n n n
S
,..., ,...,
1

N i
n n n
S
,..., 1 ,...,
1

, a transition leaves
the system (computer net) through net nodei.
N j i
n n n n
S
,..., ,... ,...,
1

N j i
n n n n
S
,..., 1 ,... 1 ,...,
1
+
,
a transition passes from net node i straight to
nodej.
2) Also, J ackson has shown that the final
solution is obtained as a product for a more general
class of networks in which is allowed processing
the transitions several times in a certain net node,
that is allowing cycling the transitions between
entering and leaving a certain net node.
3. Conclusion
This article presents a concrete modality to
model systems for the particular case of computer
networks with series queues, which is the
mathematical method that leads to a equations
system, sometimes impossible to resolve without
imposing supplementary hypothesis which are, in
some cases, extremely restrictive, hypothesis that
ease up the model from the real modeled system.
In this case, working in the asymptotically
stable equilibrium state, the differential equations
system thats obtained, impossible to resolve in
general, is reduced to a resolvable equations
system, which will provide, through the probability
distribution p(n
1
, n
2
, ..., n
N
), a description of the
100
medium behaviour on long term of the computer
networkwith series queues.
In the same way the other subclasses of J ackson
network can be modeled, and also network type
BCMP, BUZEN, etc and the theoretical results can
be compared by mathematical models for
performance indicators, such as: using a node, use
two or more node in the same time, residence time
of transactions in network, medium length of
waiting tails, a nodes efficiency and many other,
with the results given by modeling through other
methods, such as modeling trough Coloured Petri
nets [3,4].
References
[1]. R. F. Garzia, M. R. Garzia, Network
modelling, simulation and Analysis, Ed.
Marcel Rekker, New York and Basel, 1990.
[2]. M. Popa, Modelarea matematica a retelelor
de calculatoare, Ed. Univ. of Bucharest, 2004
[3]. M. Popa, Mariana Popa., M. Dragan,
Modelling patient flow in a medical office by
stochastic timed Coloured Petri Nets, the 3rd
International Conference on
Telecomunications, Electronics and
Informatics, May 20-23, Ed. UTM, Chisinau
2010.
[4] M. Popa, M. Dragan, Mariana Popa,Study the
Markings Trap in the Coverage Graph of a
Colored Petri Nets(CPN), 16TH
International Conference on Soft Computing,
J une 23-25, Brno, Czech Republic, 2010.
101
TECHNI CAL SOLUTI ONS TO DESI GN AN EQUI PMENT FOR
I CE PARTS RAPI D PROTOTYPI NG
Nicolae IONESCU
1
, Aurelian VIAN
2
, Alexandru SAVI N
3
, Mihai TRIFNESCU
4
1
POLITEHNICA University of Bucharest, ionescu_upb@yahoo.com.
2
POLITEHNICA University of Bucharest, aurelianvisan2003@yahoo.com.
3
POLITEHNICA University of Bucharest, savinalexandru@yahoo.com.
4
POLITEHNICA University of Bucharest, trify_tl@yahoo.com.
Abstract: The paper presents some detailed technical solutions to design an equipment for ice parts
Rapid Prototyping. The authors designed two alternative solutions of the equipment applying a
methodology developed within the TCM Department of the POLITEHNICA University of Bucharest.
Both alternative solutions are based on the operation principle of selective freezing with liquid nitrogen of
water sprayed on the working surface according to section image conveyed by computer by slicing a
CAD model. The article focuses on the methods of solving several issues of conception, design and
achievement of the high complexity equipment.
Keywords: Rapid Prototyping, ice, equipment, design
1. Introduction
Rapid Prototyping procedures has gained wider
applicability recently, as a modern manufacturing
technique, due to short time machining of a large
type of parts by using this method. As it is well
known, there are currently applied with good
results several rapid prototyping techniques, such
as [7]: Stereolithography - SLA, Selective Laser
Sintering SLS, Solid Ground Curing SGC, 3D
Printing, Fused Deposition Modelling FDM,
Laminated Object Manufacturing LOM and Ink-
J et Printing.
Each of these methods had both advantages and
drawbacks, yet what all have in common is the fact
that the part is achieved by material deposition not
removal, as it is the case with classical
technologies.
Not withstanding all the latest improvements of
the recent years, all these rapid prototyping
methods have a major drawback in common that is
the very high cost of parts achieved, due mainly to
high cost of equipment and consumables.
One way to cut down the rapid prototyping
costs is to use less expensive raw materials.
Therefore, rapid achievement of ice prototyping is
based on using very cheap material, namely water.
At this moment, world wide achievements in this
domain are but at an early stage of experimental
research [2, 3, 4, 5, 6].
This article presents the principal aspects of the
extended research work carried out by the authors,
ending up with the achievement at final stage of an
equipment for ice parts rapid prototyping.
Considering the current research carried out to
the present moment, the authors are entitled to
state that the achievement of this equipment would
bring about special effects, both in scientific terms,
and practical applications. Among the most
important domains of application for the ice parts
rapid prototyping that are currently applied with
special outcome, it should be mentioned here of
the achievement of the fusible casting moulds [1,
8] and medical bio-models.
2. The principle of ice prototypes production
Most of the rapid prototyping techniques are
still very expensive; in addition, some of them
generate dust and smoke emissions, thus
endangering the operators health. One of them is
the method Rapid Freeze Prototyping RFP,
which is a remarkable technique [5, 6], featuring
high technical characteristics,
This method helps to obtain the intended
technical characteristics, as well as higher
operation speed, without generating noxious
agents, while the working material water, is
much less expensive than any material used for the
other rapid prototyping techniques [6].
102
Based on the RFP technology, a three-D model
can be obtained of ice by using the CAD model,
through selective deposition of water and water
freezing layer by layer, in conformity with the
principle technical diagram shown below in figure
1.
In this case, according to figure 1, water is
sprayed on a table and freezed by simultaneous
spraying of liquid nitrogen. The part formation
takes place in a tank with constant low temperature
maintained through thermostat control system.
After deposition of the first layer, the nozzle is
lifted, or the table is lowered with the thickness of
the deposition layer. At the end of the indicated
period when total water freezing takes place, the
process is resumed with the deposition of a new
water layer on top of the first.
Drop on demand nozzle
Feeding pipe
Water droplet
Ice part
Figure 1: Principle of Rapid freeze prototyping [2]
When the entire part is ready, the part thus
obtained can be used immediately, or it may be
kept in a refrigerator for a distant use.
3. Conceptual design of the ice parts
rapid prototyping equipment
The conceptual design of the RFP equipment
has been achieved based on a methodology that
had been developed during several years by the
TCM Department of POLITEHNICA University.
This methodology has been applied with special
results, both in didactic events, and in graduation
diploma studies of the master degree, as well as in
projects that are being in progress.
This methodology started to be applied when
the need has been identified and defined to rapidly
obtain cheap prototypes. Based on a questionnaire,
the clients demands regarding the rapid
prototyping equipment have been examined and
grouped according to ten primary requirements and
sequenced in hierarchy.
The next step was the analysis of competing
products, that is, in principal, the rapid prototyping
equipment corresponding to the most relevant
techniques, as mentioned in Introduction to this
article. The authors followed this procedure to
ensure competitive design of this type of
equipment, in comparison with the most advanced
existing international technologies; the equipment
technical specifications have been determined
accordingly.
The next step was to prepare the functional
design, by determining the general function of the
equipment that is ice prototyping in a wide range
of standard dimensions. The general function was
then submitted to a general examination procedure,
resulting in the first place the principal functions,
and then the complementary and supplementary
functions.
Considering the principal functions, a list was
set up of the critical functions, which determine the
commercial success of the product which
correspond to the size and demands with relatively
high importance: ensure the movement on the three
axes, conveying of information from computer to
equipment, water spraying on strictly controllable
sectors, simultaneous spraying of liquid nitrogen to
induce freezing and maintaining a negative
temperature in the part processing area.
To determine the partial solutions at conceptual
level, an external research was prepared mainly
consisting in the study of several patents,
consulting several experts in this domain, and
interviewing top users.
Following the sorting of concepts, two of them
were retained, which are shown in figures 2 and 3.
Figure 2: Concept no. 1
103
Figure 3: Concept no. 2
Applying the multi-criteria method of analysis,
Analytically Hierarchy Process (AHP) we
selected concept no. 1, to be further examined at
the detailed design phase.
4. Detailed design of the ice parts rapid
prototyping equipment
At this stage, the detailed design was prepared
of the product, considering both the total assembly
and its component parts, to provide the prototype
documentation.
In the framework of the constructive design of
the total assembly, the product architecture was
prepared, the configuration of parts and
subassemblies, as well as their preliminary
dimensioning, based on their functional role.
Design was carried out, while taking into
consideration a series of criteria in accordance with
the technical service life cycle of the product, such
as its functionality, usefulness, ergonomy, labour
protection, ecology, aesthetics, manufacturing,
product disposal and minimal cost.
For the selected product concept, the product
architecture was determined in several stages, as
follows.
To better observe the interfaces between the
product blocks, an approximate constructive
product plan was first created comprising five
large blocks the system of transmissions,
actuation, the item corpus, the water circulation
system and the liquid nitrogen circulation system.
Next, we developed the detailed constructive plan
of the part, as indicated in figure 4 and we
identified the fundamental and incidental
interactions.
Figure 4: Detailed constructive plan of the product
Based on this approach, the authors prepared
the final block diagram, stating the constructive
solutions for the principal interactions (figure 5).
Figure 5: Final block diagram
Based on the above mentioned, two alternative
solutions were prepared of the rapid prototyping
equipment.
In the first solution, the horizontal movement is
performed by the working head, while the vertical
displacement by the mass of the equipment. In the
second alternative solution, the displacements on
the three axes are provided by the working head.
Figures 6, 7, 8 and 9 show the constructive
solutions adopted under the first alternative
solution.
The shape of the working head is shown in
figure 6. The water is pumped by the cylinder and
piston unit upon disconnecting of power magnet,
then water is injected through nozzle, when the
power magnet is actuated through computer
control. Injected water comes in contact with the
cold environment produced by liquid nitrogen and
freezes, getting thus attached to the deposition ice
layers. This shape of the working head leads to the
significant decrease of the ice layer deposition
104
time, as compared to the working head with water
dropping on the working surface.
Water injection nozzle
Valves
Spring
Electromagnet
Piston
N2
N
2
injectionnozzle
Water
Figure 6: Construction of the working head
Figure 7 shows the working head movement
principle. Movement is conveyed through motor
V-belt gear assisted by planetary rotary speed
reducer. The movement of the working head is
achieved by linear transducer (with photodiodes)
and calibrated ruler. The intention was to improve
the positioning precision in the case of higher
working speed.
Motor
Column
directions
Dewar tank
Displacement
traducer
Working head
Trapezoidal toothed
belt drive
Scaleruleto measure
displacement
Figure 7: Working nozzle movement principle
The less sophisticated shape was selected for
the chamber of ice part prototyping and constant
temperature maintenance, to minimize the cost of
paying off the prototyping equipment and the ice
part prototypes, respectively (figure 8).
Working
Table
Ball screw nut
mechanism
Refrigeration
system
Wormand wheel
elevator
Figure 8: Ice part prototyping and temperature
maintenance chamber
Figure 9 shows a general view of the first
alternative solution of ice parts rapid prototyping
equipment.
Figure 9: General view of RFP equipment first
alternative solution
The constructive solution for the second
alternative was designed to provide the movement
of working nozzle along the three axes. Figure 10
shows the general view of the RFP equipment in
the second alternative solution.
105
Figure 10: General view of RFP equipment second
alternative solution
The principal constructive elements of the unit
have been selected, or, from case to case, designed,
as described below.
For the selection of motors, we examined
several alternatives from the products of
manufacturers available on the specialised market,
considering DC motors in miniature variant, with
or without reducer. The authors considered linear
motors, servomotors and step by step motors.
Linear motors have the advantage of high
precision and reliability, high feeding speed and
acceleration and relatively high cost with respect to
the cost objectives envisaged for the RFP
equipment. Other shortcoming of this item is that it
needs a source of compressed air to make it
possible to maintain the distance between rotor and
stator, which makes the equipment a rather
complicated construction.
Servo motors are direct motors which require
position actuators (encoders), which increase
significantly the cost of equipment. They are
however less expensive than linear motors, but
show lower repeatability performance.
Step by step motors rotate at a specified angular
step and require no position actuator. They are not
expensive and require no sophisticated electronic
control units. Other advantage is their rotation by
1/2 to 1/32 of the normal step, thus increasing the
positioning precision, yet at the expense of smaller
feed speed. The principal disadvantage of these
types of motors is the so-called phenomenon of
electro-magnetic skidding at high speed, that is the
motor skips over rotation steps. This would result
in low feeding speed, which is a minor
disadvantage, considering the smaller CNC
operation travel (150x200X350mm) in the case of
this construction of the authors. To prevent electro-
magnetic skidding of motors, the feed speed had
been limited to 500 mm/min. maximum, which
means smaller equipment output, still better quality
of the machined surface.
Considering the above-mentioned, the authors
selected step by step three-motors manufactured by
NEMA.
Also, due attention was paid to transmission
screws: we selected ball screws produced by ISEL
GmbH, made of special hardened steel CIF at
60HRC, which ensures high wear resistance and
core elasticity.
The screw thread flank is obtained through
rolling, which leads to high precision of
positioning of the nut, with respect to the leading
screw. The ball nut also from the producer ISEL
GmbH made of steel 16MnCr5 hardened and
nitrated in gas atmosphere, which provides low
wear percentage of the rolling flank.
The authors selected a free software equipment
that can be purchased via Internet, with the result
that the cost of this component part of the RFP
equipment is zero.
The performances of this software are
comparable with those of genuine software
equipment, its technical characteristics being
totally compatible with the design RFP equipment
of the authors.
This software makes the translation from
machine code G-code into one decodable by the
equipment controller, thus enabling the execution
of rapid working feed movements, rotary speed
and other auxiliary movements. This code
conversion is made by the so-called
Postprocessor specific to the RFP equipment
jointly working with the electronic controller.
Generally, this software is used with industrial
PCs integrated in the equipment construction, but
in the case of the equipment conceived by the
authors, this software will be installed on a laptop
controlling the equipment.
Two software items, Mach3 and K-CAM, have
been examined, the second being selected, since
Mach3 requires a stronger computer for its
optimum operation and in addition, it was noticed
that this software evinces a delay between the
106
computerised control and the real control response
from the equipment engines.
The electronic controller provides the
connection between software and hardware, being
compulsory for the control function of the RFP
equipment, which is feasible in two alternative
solutions:
1. The G code is hardware-integrated in
controller, which requires more complex
electronics and is more expensive.
(microprocessors and ROM memory).
2. Code G is software-generated and requires
lower costs.
The power supply for the switch unit is high
output, which is specific for these types of power
supply systems and is available for acquisition at
Conex Electronic Bucharest.
For water supply, the authors found the solution
of customizing a pump used for car wash, namely
front screen wash. This type of pump is computer
controlled in accordance with the slicing section of
the CAD model, water being pushed towards the
spraying system above the part.
The equipment is endowed with a liquid
nitrogen storage and circulation unit comprising
Dewar tank, insulated pipes, valves, electric fuse
for the tank pressurisation, etc. By the heating of
the electric fuse fixed on the pipe for nitrogen
extraction from the tank, part of the liquid nitrogen
gets evaporated and pressurises the tank. Thus, the
nitrogen is forced up the discharge pipe and
reaches the working zone through opening a valve
actuated in correlation with the way in which water
is sprayed.
The prototype of second alternative solution
was produced, tested and standardised in the Dr.
Kocher Center inside the POLITEHNICA
University of Bucharest.
5. Conclusions
Based on the above-mentioned descriptions and
definitions made by the authors, several
conclusions can be stated, as follows:
1. The authors proposed a solution of rapid
prototyping cost reduction consisting in the
utilization of less expensive raw material in the
process. Ice parts rapid prototyping involves the
utilisation of a very cost-efficient material, which
is water.
2. Conceptual design of RFP equipment is
based on the methodology developed during
several years of research by the authors within the
TCM Department of POLITEHNICA University of
Bucharest.
3. Two alternative solutions were prepared of
the RFP equipment. The first solution contains
horizontal movement by aid of working head, and
vertical movement by aid of the equipment table.
In the second alternative solutions, the movement
on the three axes are performed by the working
head.
4. Both solutions are based on the same
operation principle of selective freezing using
liquid nitrogen, of the sprayed water on the
working surface, in conformity with the slicing
section of the CAD model.
References
[1] Beeley P. R. and Smart R. F., Investment
Casting, 1st Edition, The University Press,
Cambridge, UK, 1995.
[2] Leu M. C., Zhang W., and Sui G., An
experimental and analytical study of ice part
fabrication with rapid freeze prototyping,
CIRP Annals, 49(1), pp.147-150, 2000.
[3] Liu Q., Sui G., and Leu M. C., Experimental
study on the ice pattern fabrication for the
investment casting by rapid freeze prototyping
(RFP), Computers in Industry, 48(3), August,
2002.
[4] Peters D. M. Patterns in ice, Foundry
Management & Technology, 123(8), 1995.
[5] Yodice A., Freeze cast process, US Patent
5,072,770, 1991.
[6] Yodice A., Freeze process cuts casting costs,
Advanced Materials and Processes, 155 (4)
(1999) 35-36.
[7] Wohlers Terry, Rapid Prototyping Moulds
the Future of Manufacturing Available from,
http://www.wohlersassociates.com.
[8] Wu X., Study on ceramic mould investment
casting based on ice patterns made by rapid
prototyping method, Bachelor Degree Thesis,
Tsinghua University, Beijing, China,1997.
107
COMPOSI TE SHAPE MEMORY ALLOYS USED I N ENERGY
DI SSI PATI ON APPLI CATI ONS
I. Cimpoeu*, S. Stanciu, A. Enache, D. Dan, V. Manole, P. Paraschiv
The Gh. Asachi Technical University of Iasi Faculty of Materials Science and Engineering, Prof.
dr. docent Dimitrie Mangeron Rd., 61 A, 700050, Iai, Romania
Abstract: Shape memory alloys exhibit through them martensitic transformation a high damping
capacity that can be used in many applications. Microstructural and chemical analyses were made
concerning a shape memory alloy based on copper that can be use as matrix material in a composite
material for mechanic energy dissipation. Scanning electrons microscope and X ray analysis equipment
were used to characterize some material properties.
Keywords: shape memory alloys, damping capacity, martensite transformation
1. Introduction
Shape memory alloys, because of their
unique thermo-mechanical behavior, have
developed considerable works towards the
performance of intelligent materials and
structures. Indeed, these alloys are particularly
useful when large deformation and recovery of
the shape is observed under a small rate of
stress or temperature. These properties are
attributed to the ability to undergo martensitic
phase transformation. This approach presents
the advantages to be easily incorporated in a
structural computation. Shape memory alloys
have ability to modify both their shape and
properties in response to the thermo-
mechanical environment. They may be
embedded into an inactive material to
transform it into an active composite: by a
suitable choice of their volume fraction, shape
and alignments, they accomplish a specific
function. Many applications can be profitably
explored in aerospace structures, vibration
control among others.
Actuation by shape memory alloys (SMAs)
has generated considerable interest and some
applications have been developed across a
wide range of industries. Successful
applications are still largely restricted to niche
markets, although the extent of their
applications is growing rapidly. Two factors
seem to have restricted growth into other
markets. These are the early dominance by the
medical market, where small sections of low
temperature materials have been developed.
This has partially led to the second limiting
factor, which is the lack of availability of
larger engineering sections and techniques.
This second limitation is partially due to
technical difficulties in scaling up of the
materials and techniques, plus a lack of
visibility of the benefits.
Nonetheless, impressive demonstrations of
larger scale advanced actuator applications
have been achieved, particularly in the
aerospace industry. Applications within a gas
turbine are an order of magnitude more
difficult due to the more robust requirements
and higher temperatures encountered [1].
Many research grants are in hand to investigate
the opportunities for smart technologies to gas
turbines, with SMAs attracting considerable
interest. The opportunities for enhanced
performance and cost reduction are generally
obvious within the industry, but this must be
achieved with realistic cost and reliability from
the technologies.
In general, a high integrity, robust and stiff
structure must be achieved. This has led to
several programme investigating larger
structures and supporting technologies such as
joining techniques. The control of noise
pollution has become increasingly important in
the aerospace industry. The generation and
radiation of noise from aero-engines poses a
problem of immense complexity. Over recent
years, noise reduction has become a prime
108
requirement in the specification for next-
generation propulsion systems.
Engine manufacturers are constantly
investigating ways that will reduce the noise
created by gas turbine engines. Commercial
SMAs are usually available off the shelf in
the form of wires and thin strips. The SMA
business is very much supply driven and
without major demands for large section
SMAs, suppliers are reluctant to invest and
develop the process technologies required to
manufacture them. This has been a major
factor hindering the advances in the alloy
development and process technology of SMAs
[2-4].
In this study a composite matrix material
made of shape memory alloy based on copper
is analyzed in martensitic state by
microstructure and chemical composition
points of view. Number of martensitic variants
and grains dimensions influence the materials
damping capacity through them boundaries
dimensions and number so is important to
analyze them characteristics [5-7].
2. Experimental details
To obtain the alloy was used a laboratory
furnace with graphite crucible using copper,
zinc and aluminum high purity materials with
reduce percentage of iron [8]. The heat
treatments were realized on a laboratory
Vulcan furnace with controlled temperature
variation [9].
Chemical composition was determined
through spark spectrometry analysis using
Foundry Master equipment (for matrix and
reinforcement elements chemical analysis) and
EDAX analysis as well for matrix study. In
this study different EDAX software
applications were used to determine the
chemical variation of the elements on Line
Point mode with automatic or element list
considerations. Microstructures of the
composite matrix were obtained with a
scanning electron microscope (SEM) LMH II
by Vega Tescan brand using a secondary
electrons (SE) detector to characterize the
matrix in martensitic state details.
Microstructural and chemical investigations
were carried out to analyze a shape memory
alloy based on copper used as matrix material
in a composite. SEM microstructure of the
composite surface is presented in figure 1 at a
500x image amplification. Composite material
is made of a shape memory alloy as matrix and
an arch material (steel) reinforced in metallic
matrix. Analyses concerning some interface
phenomena were done [10] presenting a nice
combination of the materials and a diffusion
made layer between SMA and steel materials.
Figure 1: SEM microscopy of a composite material with a shape memory alloy matrix
109
Microstructure of a shape memory alloy
used as matrix material in a metal-metal
composite material is analyzed in figure 2.
SEM microscopy presented in figure 2 were
realized with a SE (secondary electrons)
detector at a 30 kV alimentation tension lamp
and different image amplifications like 500x in
a), 1000x in b) and 10000x in c). Using the
Tescan Software 3D analyzes of the martensite
variants was done based on the electrons
analyzed from de sample.
a) b)
c) d)
Figure 2: Microstructural analyze of a shape memory alloy based on copper used as matrix in a composite
material in martensite state a) grains intersection b) detail of grains intersection froma) with martensite
variants type c) geometrical and dimensional characteristics of martensite variants d) 3D image of area selected
in c) for martensitic variants characterization
Martensite microstructure of the copper
based shape memory alloy propose for the
matrix material has different types of variants
like plates or V shape with different kind of
dimensions between 2 or 3 m to 50 to 100
nm. Secondary martensite variants appear as
well in the microstructure, marked with red
line in figure 2 c), having nanometric
dimensions and being formed between big
martensite variants. Different grains contain
different orientation variants of martensite,
presented in figure 2 c), parts of them
110
orientated at 90 degree. The 3 D representation
of figure 2 c) present three different types of
variants not only as shape and orientation but
as high as well.
Chemical investigations of the matrix shape
memory alloys were done on different areas of
4 mm
2
from the material surface and chemical
composition results (ten different areas) being
averaged in table 1.
Spectral energies characteristics to shape
memory alloy analyzed are presented in figure
3 with two energy connections for copper and
zinc and a single characteristic energy for
aluminum.
Figure 3: Spectral energy distributions of elements fromcomposite material matrix
In table 1 weight and atomic percentages of
the component elements are presented having a
72.84 % wt of copper, 20.42 % wt of zinc and
6.73 % wt aluminum. Atomic percentages
present a 67.11 contribution of copper, 18.29
% zinc and 14.597% aluminum that are
combined as CuZn, Cu
5
Al
9
and Cu
5
Zn
8
formations [11].
Table 1Chemical composition of matrix material consisting of a shape memory alloy
Element AN series Net [wt.%] [norm. wt.%] [norm. at.%] Error in %
Copper 29 K-series 77223 64,25326 72,84393 67,11193 1,617906
Zinc 30 K-series 18388 18,01927 20,42844 18,2902 0,487459
Aluminum 13 K-series 3272 5,934216 6,727622 14,59787 0,360829
Sum: 88,20674 100 100
In figure 4 distributions of chemical
elements copper, zinc and aluminum on a 100
m line are presented and can be considered
that the material homogeneity id very good
exhibiting nice properties on entire material
mass. Following the diagram from figure 4 can
be observe a smoother behavior distribution of
copper on the smaller martensite variants and
bigger variations of all elements in the other
parts of the material. Having a smoother
homogeneity shape memory alloys can be
better controlled in practical applications.
Shape memory alloys exhibit a nice damping
capacity due trough martensitic transformation
that can be used in dissipation applications.
Reinforcement of a shape memory alloy to
increase them strength and helping the shape
111
memory recovery represent a goal of the researchers in this field.
Figure 4: Chemical elements distribution on a 100 mline of a shape memory alloy based on copper
Energy dissipation in shape memory alloys
from mechanical to thermal is made especially on
boundaries matrix between grains and martensitic
variants. Modifying the grains number and
dimensions and the variants shape, dimension and
number can improve damping capacity of the
material.
4. Conclusions
A shape memory alloy based on copper was
obtained through classical melting methods and
analyze by Microstructural and chemical points of
view.
The material is proposed for a damping
composite having a high dissipation capacity in
martensitic transformation domain of mechanical
energy in thermal energy. Grain boundaries and
martensite variants support most of the energy
dissipation so is very important to control the
material microstructure.
Showing different martensite variants and
various dimensions the material present a good
potential for damping composite metallic
materials.
Acknowledgement
This paper was realised with the support of
POSDRU CUANTUMDOC DOCTORAL
STUDIES FOR EUROPEAN PERFORMANCES
IN RESEARCH AND INOVATION ID79407
project funded by the European Social Found and
Romanian Government.
References
[1] Webster, J ., Proceedings of the International
Society of Air Breathing Engines Conference,
Bangalore, India, September, 2001.
[2] Caicedo, M., J acobs, J.J., Reddy, A., Hallab,
N.J ., J. Biomed. Mater. Res. A 86 (2007) 905.
[3] Shabalovskaya, S., Anderegg, J ., Van
Humbeeck, J ., Acta Biomater. 4 (2008) 447.
[4] Sun, T., Wang, M., Surf. Coat. Technol. 205
(2010) 92.
[5] Cimpoeu, N., Stanciu, S., Meyer, M., Ioni, I.,
Cimpoeu Hanu, R., Effect of stress on damping
capacity of a shape memory alloy CuZnAl, ,
J ournal of Optoelectronics and Advanced
Materials, Vol. 12, Issue 2, pg. 386-391, IDS
570DE, ISSD 1454-4164, 2010.
[6] Paun, V.-P., Cimpoesu, N., Cimpoesu Hanu, R.,
Munceleanu, G. V., Forna, N., Agop M., On the
Energy Dissipation Capacity and the Shape
Memory. A Comparative Study between
Polymer Composites and Alloys, Materiale
Plastice, vol. 47, nr. 2 , pg. 158-163, 2010.
[7] Cimpoeu, N., Cimpoeu Hanu, R,, Vizureanu,
P., Ioni, I., Agop, M., Experimental and
theoretical results concerning internal friction
investigation of a shape memory alloy based on
copper, Metalurgia Internaional, vol. XV, nr.
12, pg. 48-59, ISSN 1582 2214, 2010.
112
[8] Mares, M A, Carcea, I, Chelariu, R, Roman, C,
Mechanical and tribological properties of
aluminium matrix composites after ageing
treatments, Euromat 97 - proceedings of the 5th
european conference on advanced materials and
processes and applications: materials,
functionality & design, vol. 1 - metals and
composites pg. 423-426, 1997.
[9] Cimpoeu, N., Enache, A., Stanciu, S.,
Nejneru, C., Achiei, D., Hopulele, I.,
Composite shape memory materials obtaining
methods, Simpozionul international Artcast
2008, Galati, pg. 293-296 ISBN 978-973-7845-
94-8, 2008.
[10] Enache, A., Cimpoesu, I., Stanciu, S.,
Hopulele, I., Axinte, M., Cimpoesu, N.,
Interface analyze of a composite material with
shape memory alloy matrix, Bul. Inst. Polit.
Iasi, t. LVI (LX), f. 1, 2011.
[11] Morton, A.J ., The -phase regions of the Cu-
Zn, Ni-Zn and Pd-Zn binary systems, Acta
metall., 27, , 863-8,1979.
113
I RESEARCH ON THE CORROSI ON OF CONSTRUCTI ON MATERI ALS
FI SSURE MACHI NERY HYDROCARBONS I N CONTAMI NATED
ENVI RONMENTS WI TH STAFF FROM CORROSI VE BY CRUDE OI L
DI STI LATI ON
Moroanu Marius
1
, Ovidiu Georgescu
2
1
marius.morosanu2003@gmail.com.
Abstract: In the atmospheric distillation of crude oil facilities and in vacuum of the fuel oil were
highlighted in uniform and uneven general corrosion, local corrosion in the form of points, spots, caves
and cracking, selective, crevice, etc. .. All these forms of corrosion are dangerons, corrosion cracking,
however, occupies a special place, because in a short time can switches off the machine or plant with hard
consequences sometimes to assess. If sulfur crude oil processing appeared to corrosion cracking of
austenitic stainless steel scales of DA-C1 column at the top, after 1.5 years of operation. Also, heat
exchangers at the top of the columns DA and DV, with monel or carbon steel pipe. Hydrocarbon
environments in the presence of additional factors, are potentially active for the emergence and
development of corrosion cracking. The paper presents some aspects of metal corrosion crack to metal
equipments in the crude oil distillation plants and fuel oil vacuum
Keywords: stress corrosion, load test, time fractional (max. 5)
1. Aims and background
Corrosion cracking under stress is one of the
most dangerous forms, given the destruction that it
causes to metal materials used in manufacturing
industrial machinery. This process is manifested by
intergranular cracking and transgranular,
accompanied by a reduction in strength properties
and finally, the fracture mechanics of materials in
the cracks.
Environments containing hydrochloric acid,
hydrogen sulfide, naphthenic acids, cyanides, etc..
are specific environments that cause cracking and
ultimately fracture of steels subjected to corrosion
cracking under stress in the normal operation of the
plant from atmospheric distillation of crude oil and
fuel oil vacuum. Sulfur compounds, chlorine,
nitrogen and oxygen are converted to hydrogen
sulphide, hydrochloric acid, ammonia and
respectively water [1-3].
The compounds of conversion in the absence of
corrosion inhibitors have a corrosive action on
metal equipment from carbon steel, stainless steel
or alloys CuZn, Cu Ni, itself and their interaction
products (ammonium chloride, ammonium sulfide,
ferric chloride, etc.. ). These building materials in
specific environment conditions, are susceptible to
corrosion cracking - [4-6].
2. Experimental
The paper presents experimental data obtained from
tests performed by the conventional method of
determining the time of failure and the propagation of
cracks in the corrosion of metal materials.
Materials used are carbon steel S235JR , stainless
ferrites steel 7AlCr130 and austenitic stainless steels
X6CrNiTi 18-10 and 17-12-2 X6CrNiMoTi of high
mechanical strength, whose composition is shown in
Table 2.1.
Tabel 2.1.- Chemical composition of studied steel
Also, test samples were stabilized brass (70%
Cu, 29% Zn, 1% Sn) and Monel W2.4230 (65.9%
Ni, 32.5% Cu, 0.03% Al, 1, 57% Fe).
Steel Components , %
C M
n
S
i
Cr Ni Ti M
o
P S
S235J R 0,18 0,
53
0,
3
0
- - - - 0,0
4
0,0
4
X6CrAl
13
0,08 0,
60
0,
6
0
13,
2
- 0,1 - 0,0
3
0,0
15
X6CrNi
Ti
18-10
0,06 1,
8
0,
7
18,
5
9,5 0,51 - 0,0
3
0,0
14
X6CrNi
MoTi
17-12-2
0,07 1,
7
0,
8
18,
3
11,
2
0,52 2,4 0,0
4
0,0
15
114
The specimens used in corrosion cracking tests
are of cylindrical shape with a diameter of 3 mm,
threaded at both ends to be caught in fast-
tensioning device to ECO Before each
determination, the metal specimens were machined
and polished with metallographic paper to 600 grit,
then degreased in acetone.
Tensioning specimens was performed under a
constant load voltage to a unidirectional, its value
to 50%, 70% and 85% of breaking strength of the
material tested.
Tests were conducted at temperature of 200C in
condensate aqueous collected from the top of the
atmospheric distillation column column AD-C1-C5
VD vacuum distillation, samples weher pH was
maintained in the 6.0 to 6.5 and the chloride
content was 30-60 ppm.
The samples tested were mounted in the
container unit composition aggressive
environment. Lever was loaded with weights until
the tensioning force of the specimen (Figure 2. 1).
Test time is measured by timer with digital
readout, and the cracking sound samples is
reported. The timer stops to the break, if the
sample and indicate the number of hours during the
sample remains under load.
1 cell solution;
2 heating jacket;
3 thermometer
4 corrosion
specimen;
5 refrigerant;
6 hydraulic
closure;
7 sealing;
8 lever;
9 greuti;
10 dispozitiv de semnalizare a ruperii.
Figure.2.1: Divice for determining a corrosion
crack with unidirectional voltage
Also, the tendency to corrosion cracking of
steels investigated was determined according to SR
SR I SO 7539-2:1994/A99:2002. The method
involves immersing for a specified duration of
epruvette, previously deformed in cold and tense
elastic device shown inFigure 2.2
Figure. 2.2 : Sample tensioning device
Test environments at the top of the columns
were maintained at 80 temperature and the
continuous refluxing and extensive examination of
the specimen surface was made after its removal
from solution. The samples were processed only to
the interior, which comes into contact with the
mandrel. The specimens were mounted in the
device so that the processed surface to be placed in
the compression and tension by tightening the
screws Co up to a distance of 2 mm between the
specimen cap and base plate of the device.
Specimens caught in the device tensions were
degreased with acetone and were placed in a
corrosive environment, which should completely
cover the specimens, the duration of the test.
The specimens were examined visually for
cracks control. The examination was conducted by
taking samples of the solution, washing with hot
distilled water and observing with the naked eye
and with a 10x zoom lens.
3 Results and discussion
Composition of condensate aqueous
collected from the top of the atmospheric crude
distillation column AD-C1 (Figure 3.1) and
vacuum distillation of fuel oil VD-C5 (Figure 3.2)
are presented in Table 3.1, and stress corrosion
cracking test results are presented intable 3.2.
Figure 3.1: The technological scheme for
atmospheric distillation plant (AD-C1)
115
Fig. 3.2: The technological scheme for vid
distillation plant (VD-C5)
Table 3.1Quality of aqueous condensate test
No Qualitative
Characteristics
UM Working place
Column
reflux
tank AD-
V1
Colum
reflux
recipient
Dv-V100
1 pH upH 6,4 6,7
2 Chlorides ppm 42 31
3 Sulfur ppm 52 45
4 Iron ppm 0,1 0,2
Table 3.2: Values for fracture time of the tested
materials at temperatures of 200C and 800C in the
alquons condensate fromthe top of the column AD-C1,
pH=6,0
It appears that after 1000 hours of testing in the
aqueous condensate from the top of the column
AD-C1 did not show cracks on the surface of
specimens under unidirectional tension of 50% of
the breaking strength of the material. At test loads
of 85% from the tensile strength, fracture time
decreased. This decrease becomes more
pronounced by increasing the operating
temperature of 200C to 800C.
Also tested were the same types of materials into
the tension that in Figure 1. Lack of large cracks
on the (upper) of the specimen showed resistance
to corrosion cracking under stress of the tested
materials (Table 3.3).
Specimens of carbon steel S235J R had
corrosion rates of 0.118 mm / year at the top of the
column DA-C1 and 0.130 mm / year at the top of
the column C5 DV-corrosion showed punctuate on
the whole surface. Austenitic stainless steels have
much lower corrosion rates and their surface is
uniform general corrosion. Furthermore, brass
CuZn 28 and Monel W24230 SN1 have corrosion
rates of 0,012 mm / year respectively 0.0103 mm /
year.
No
M
a
t
e
r
i
a
l
B
r
e
a
k
i
n
g
r
e
s
u
l
t
s

R
m
N
/
m
m
2
%

f
r
o
m

R
m
N
/
m
m
2
T
a
s
k
t
e
s
t
N
/
m
m
2
T
e
m
p
e
r
a
t
u
r
e
s
0
C
F
o
r
F
a
c
t
u
r
e

t
i
m
e
O
b
s
e
r
v
a
t
i
o
n
1
S
2
3
5
J
R
420
50
21
0
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
70
29
4
20 >1000
Nu s-a
rupt
80 750 S-a rupt
85
35
7
20 390 S-a rupt
80 105 S-a rupt
2
X
6
C
r
A
l
1
3
560 70
39
2
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
3
X
6
C
r
N
i
T
i

1
8
-
1
0
550
50
27
5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
70
38
5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
85
46
7.5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
4
X
6
C
r
N
i
M
o
T
i
1
7
-
1
2
-
2
540
50
27
5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
70
38
5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
85
46
7.5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
5
B
r
a
s
s
350 70
24
5
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
6
M
o
n
e
l
580 70
40
6
20 >1000
Nu s-a
rupt
80 >1000
Nu s-a
rupt
116
Tabel 3.3: Test results of the tested materials corrosion
cracking at 80 temperature for 96 hours
On visual inspection of samples there were not
found am cracks. The values of corrosion rates
ranged in the materials stability group metals. The
data presented show that the methods used the
studied materials fracturing time is high.
4 Conclusion
1. Corrosion cracking tests were performed
with samples under a unidirectional tensin at a
constant load and with samples deformed at cold
and elastic tensied.
2. After 1000 hours of testing in the aqueous
condensate from the top of the column DA-C1 did
not show cracks on the surface of specimens under
unidirectional tension of 50% and 70% of breaking
strength and break more quickly specimens at a
temperature of 800C.
3. By the cold deformation and elastic tension,
the samples exposed in the media at the top of the
column DA-C1-C5 DV, at boiling temperature,
were not affected by cracks.
4. The presence of chloride with hydrogen
sulfide accelerates corrosion phenomena,
compared with the corrosions processed only by
hydrogen sulfide. Hydrochloric acid aqueous
solutions under high unit action efforts, cause
corrosion cracking of austenitic stainless steels.
5. Low concentrations of chlorides, ppm order.
produce austenitic cracks when water with salts is
concentrated by boiling. The tendency to cracking
increases with the temperature increasing at a
constant voltage and with increasing salt content.
6. Variable-valence metal chlorides Fe2 +/ Fe3
+, Cu2 +/ Cu +give rise to pitting corrosion,
which may constitue germs of corrosion cracking.
References
[1] Shalaby H.M., Refining of Kuwait s heavy
crude oil: materials challenages, Workshop on
corrosion and protection of metals, Arab school
for science and tehnology, December 3-7
Kuwait
[2] Antonescu N.N., Moroanu M, Petrescu M.G.,
Niculae A.D., Georgescu L., Georgescu Ov.,
Issues on corrosion equipment in instalation by
distribution DAV naphthenic acids fromcrude oil,
Poster WP15- corrosion in Refinery Industry,
EUROCORR 2010, The European Corrosion
Congres, European Federation of Corrosion 13-
17 September 2010.
[3] Burlov V.V, Altsybeeva A.I., Kuzinova T.M.,
Sokolov V.L, Corrosion problems of modern
petroleumrefinery, Poster WP15- Corrosion in
refinery industry, EUROCORR 2010, The
European Corrosion Congres, European
Federation of Corrosion, 13-17 september 2010.
[4] Ropital F. Mecanismes et mecanique des
interactions plastic envirovmements. Cas
dindustrie petroliere, Plast Ox 2007 (2009),
p265-275., EDP Sciences 2009
[5] Lobley G.R., Stress corrosion cracking: case
Saudies in refinery equipment, The 6 th sandi
Engineering Conference, KFUPM, Dhahran,
December 2002, vol.5, 2002, p17-26.
[6] Feaugas X., Creus. i., Sabot R., El Alami H.,
Lange D., Sahal M., Savall C., Huvier C.,
Influence dun etat mecanique sur la reactivite
de surface en milieux aqueux des metaux c.f.c,
Plasto x 2007 (2009), p221-236, EDP Sciences
2009.
No Work
Environment
Nature
Materials
Time
to
break
after
140
hours
The corrosion
rate
Kg,
g/m2
h
P,
mm/an
1 Condensa
tion from the
top of
distillation
column AD-
C1
S235JR No
cracks
0,107 0,118
X6CrAl13 No
cracks
0,022 0,024
X6CrNiTi
18-10
No
cracks
0,014 0,016
X6CrNiMoTi
17-12-2
No
cracks
0,0044 0,0049
Brass No
cracks
0,011 0,012
Monel No
cracks
0,0093 0,0203
2 Condensa
tion from the
top of
distillation
column VD-
C5
S235JR No
cracks
0,12 0,13
X6CrAl13 No
cracks
0,024 0,027
X6CrNiTi
18-10
No
cracks
0,019 0,021
X6CrNiMoTi
17-12-2
No
cracks
0,0052 0,0058
117
SOURCES OF POLLUTI ON I N SI DERURGY AND TECHNI QUES FOR
REDUCI NG NOXI OUS EMI SSI ONS
Elisabeta Vasilescu , Ana Doniga , Marian Neacsu
1
Dunrea de J os University of Galati , elisabeta.vasilescu@yahoo.com
2,3
Dunrea de J os University of Galati , uscaeni@yahoo.com.
Abstract: The systematic identification of the various sources of pollution creates the necessary
background for organizing actions in order to prevent and fight against pollution. The diversity of the
sources and forms of pollution are characteristic of industrial activities, especially of those in the
chemical industry, metallurgical engineering, transportation industry, generally, industrial sectors dealing
with raw material manufacturing.
In the metallurgical industry, the air pollution takes place especially on the basis of the following
categories of polluting agents: solid particles resulted from powders, particles drawn by combustion
gases, particles of coal, ore, coke and fusing agents, slag, iron oxides, sulphur oxides (SO
2
and SO
3
),
nitrogen oxides (NO
2
and NO
3
) and, although in smaller quantities, by tars, hydrocarbons, soot, organic
acids and others.
Being familiar with the sources of pollution, their monitoring and control represent a first step
towards reducing the quantity and the toxicity of all emissions, focusing on applying a cleaner
production in the industry of elaborating and processing metallic materials, too.
This paper succinctly presents the main sources of pollution in the primary siderurgical sectors, as
well as a series of techniques to reduce pollution, especially in the process of steel elaboration.
Keywords: pollution, noxious emissions, siderurgy
1. Introduction
During the period between1950 and 1960, steel
industry represented a major source of pollution,
especially air pollution. The problem of reducing
air pollution was raised at once, which made the
specialists focus their attention, initially, on
modernizing the installations for gases and dust
collection, but this approach was replaced by the
introduction of new installations designed
according to the concept of cleaner production.
A cleaner technology is based on less polluting
production procedures or procedures which use
recycling technologies, reintroducing in the
manufacturing process the waste material resulted
from a certain technological phase, or procedures
which revaluate waste material turning it into raw
material for a secondary production. In a less
polluting procedure, investments become
productive and the expenses implied by
decontamination are an integral part of the
manufacturing process.
From the economical point of view, clean
technologies, by fighting against wasting, allow
achieving important energy and raw material
savings, which lead to cutting down the terms for
recovering the investment expenses.
2. Sources of pollution and techniques for
reducing pollution in the primary siderurgical
sectors.
In coking plants, agglomeration and
elaboration sectors (cast iron-steel), air pollution is
the most important environmental problem and
diminishing or completely eradicating noxious
emissions in the atmosphere is connected with the
preservation of energy and resources available
(fig.1).
Thus, residual gases represent a valuable
source of fuel, at the same time being an important
link in the energetic chain of a plant.
In the case of coking plants, the noxious
emissions in the air consist mainly in powders
resulted from coal supplying, coal processing and
118
coke sorting. Burnt gases from coking batteries,
coke quenching and cooling are also released in the
air. Generally, before being used, the resulted
gases are carried through different cleaning
systems, the residual water resulted containing
solid particles, heavy metals (in the case of furnace
gas and converter gas cleaning) or organic
polluting agents (cyanides, ammonia and others) in
the case of coking gas cleaning. Besides, powders
and silt containing iron can also be collected. Thus,
by reducing the emissions of secondary gases in
atmosphere, an ironworks can benefit from the
existence of a collected power source, reducing the
other expenditures, and as a result of recycling
materials containing iron, the consumption of the
available resources is diminished.
Figure1 Relative emissions in the air originating in the primary siderurgical sectors
Temporary gas emissions appear during the
operations of liquid pig iron transport, preliminary
treatment and operations of loading, charging and
evacuation of the LD converter. The secondary
stirring installation and the continuous casting one
are less important sources of noxious emissions in
the air.
An inert gas can be used to reduce the iron
oxides emissions during the operations of liquid
pig iron transport and of loading the converter, and
the emissions generated during loading can be
reduced by optimizing the loading technologies as
well as the characteristics of scrap iron (quality,
distribution in furnace).
In the case of steel works, there are emissions
resulted from secondary gas burning, temporary
emissions from the steel works, flashing emissions
and the cleaning of the secondary extraction gas.
The secondary emissions during charging can
be controlled by providing a tight adjusting
between the converter opening and the seal boot of
the primary residual gas system, but this can often
be prevented by the drops of slag and metal
produced during the charging operation.
The secondary gas from the LD converter can
be collected and recycled as fuel, although a part of
it has to be burnt every time at the beginning and
in the end of each charge when the CO
concentration is low. Thus, the efficiency of the
gas cleaning installation as well as of the gas
collecting will have an important effect on the total
quantity of the emissions in the air. Technologies
apply computerised control systems in order to
increase the efficiency of gas collecting.
The converter gas used as fuel has a caloric
power of about 9 MJ /m
3
, thus saving 0,7 GJ /t of
raw steel. In Germany, for instance, the recycling
of the converter gas saves the equivalent of 300
Nm
3
of natural gas.
119
The system works by carrying the hot gas
through a recuperator where the heat is used to
generate steam (fig.2). Then, the gas goes through
a purifying system or through a wet cleaning
system, and, when a certain value of the CO
concentration is reached, the gas is collected,
cooled and stored. The electric filter purifying
system has lower power consumption as compared
to the corresponding water systems and it is
installed in the place where the gas is collected
from the converter.
Figure 2The purifying and recycling systemfor LD converter gases
In electric steel works, a primary extraction
system can be used in order to directly collect the
emissions generated in the furnace during the
melting process and a secondary system can be
used to collect the temporary emissions in the steel
works during loading, melting and evacuation
(fig.3). The primary gas is mainly filtered in order
to remove the particles before their being released
in the atmosphere, but this system may not be
equally efficient in lowering the level of the
volatile compounds, of the smoke resulted and of
the traces of organic compounds, of volatile metals
(e.g. mercury) etc. The level of these compounds
can be controlled to a certain extent by sorting and
preparing the scrap iron. Another problem in the
case of electric furnaces is represented by the
dioxin type toxic gas emissions, for which a
series of researches have been made in order to
diminish or completely eliminate them.
Figure 3 The situation of the powders collected fromthe primary and secondary gases evacuated fromthe
electric arc furnaces (data processed from67 installations)
120
3. Modern technologies used to reduce the
dioxins emissions in the electric arc furnaces
Among the polluting substances emitted
during the metallurgical processes, dioxins are the
most toxic.
The problem of dioxins emissions is more
acutely raised in the case of electric arc elaboration
furnaces, where important quantities of scrap iron
(along with the organic and chloride constituents
which they inevitably contain) are melted in order
to obtain steel.
Dioxins define a group of compounds which
consists of about 75 polychlorodibenzodioxins
(PCDD) and 135 polychlorodibenzofurans
(PCDF).
At high temperatures, dioxins and furans
decompose.
In the gas cooling phase, they are remade by
the chemical reactions between the organic
compounds and the chlorides present in the gas and
powder emissions, at temperatures of 200 600C.
This novo synthesis reaches the maximum
intensity between 250 and 400C. In fig.4, it is
schematically presented the variation of the
reaction speed in the process of dioxins forming
and decomposing, function of temperature raise.
Dioxins remaking depends on the gases
kinetics, especially during the above mentioned
temperature interval. Thus, in order to reduce these
emissions, it is necessary a rapid cooling of the
gases resulted in the process, at temperatures
bellow 200C to prevent dioxins remaking.
The strategy for reducing the concentration
of dioxins in the electric furnaces emissions is
focused on three aspects:
- thermic decomposing as advanced as possible of
the dioxins contained in emissions by an extensive
exposure of these compounds at temperatures over
850C;
- prevention of their remaking (by novo
synthesis) during the gas cooling phase by
accelerating the process within the temperature
domain 600 - 200C:
- separating the residual dioxins from the emission
flow, by injecting an absorbent substance, followed
by a filtering operation.
Figure 4 Dioxins forming function of temperature
To reduce the dioxin emissions in atmosphere,
a modular system for treating emissions was
developed, which takes into account the specific
conditions of exploitation and the type of electric
furnace (classical or tub furnaces).
Each part of the system (equipment)
corresponding to a technological treatment stage
was optimised in industrial conditions (fig.5).
121
Figure 5 The efficiency of absorption depending on the filter temperature and the injecting of the absorbent
substance
In the case of classical electric furnaces, a
separator was installed, which captures the dust
particles and the drops of liquid slag. The hot blast
(which includes gases and solid particles) can
reach 1000 C and contains an important quantity
of gaseous CO which burns in contact with the
injected air.
Simultaneously, the volatile organic
compounds contained in the gaseous mixture,
together with the dioxins resulted from the organic
substances present in the scrap iron and the
recycled materials in the charge, are decomposed.
In the case of the shaft furnaces, a quantity
of air and liquid fuel are injected and burnt in a
combustion chamber located at the outlet port of
the charge preheating tub in order to achieve a
temperature high enough (over 850C) to assure
the decomposing of the organic materials and of
the dioxins contained in the mixture of gases and
powders evacuated.
The dense particles of powders and the slag
drops are evacuated quite efficiently through the
filter made by the charge column situated in the
preheating room.
Gas cooling
After combustion, hot gases are cooled to
approximately 600C in a pipe cooled by water,
and then they are passed either in a mist cooler or
in a gas-gas cooler with forced draught to reach a
temperature of approximately 200C. The
advantages of mist cooling are: the high cooling
speed and, due to a small pressure lowering, the
easier circulation through the system.
The forced draught cooling has the advantage
that it can function without water (the one which
makes the mist), which simplifies the exploitation.
Dioxins absorption
The residual dioxins contained in the gaseous
mixture are eliminated by condensation and / or
absorption followed by a filtering stage. The
efficiency of the process depends on a
homogeneous dispersion of the particles having a
certain size and shape (with large specific surface)
in the emissions mixture to obtain a maximum
degree of compression and absorption of dioxins.
For the purpose of assuring homogeneity, active
carbon (coke) or active carbon mixed with lime
were used. The quantity and the quality of the
absorption mixture are important factors which
assure the security of the installations within the
aggregate, with a view to reduce the danger of
taking fire and explosion inside the filtering
chamber.
The injection system of the absorbent
substance is made up of a bin (storage chamber)
for the absorbent substance, a (spider)
distribution aggregate and the proper infusion
system.
The particles enriched with dioxin are
separated from the emission flow in a filtering
chamber containing sack filters and unclogging
devices with counter-current air. Powders are
collected in a bin, and then they are agglomerated
to be recycled or evacuated in the end.
Industrial uses
1. System for a tub furnace.
122
The system of dioxin evacuation has a
combustion chamber equipped with CH
4
burners.
After leaving this chamber, the exhaust gas and
powder mixture goes through a mist cooler. An
absorbent infusion device is installed next to the
chamber where primary and secondary gases are
mixed. The purifying of the dioxin loaded gases is
achieved in an air filter.
2. System for classical electric furnaces.
The burnt gases go into a separator where thick
particles are eliminated either by the emission flux
or by a mist cooler or a forced draught air cooler.
Gases go then through a fine filter where the
smallest particles are eliminated.
To sum up, we can make the following
observations:
- It was noticed that the efficiency of the systems
for the removing of dioxins in the electric ovens
emissions has been substantially improved by
introducing absorbent substances infusion systems;
- The dioxins tendency of concentrating, by
absorption and condensation, on the particles of
very fine powders (which facilitates their
elimination in sack filters) becomes more
accentuated as the temperature of the gases at the
filter entrance goes lower. This temperature should
be under 80C to assure a removal of the dioxins
as advanced as possible.
In the case of the system treating the emissions
from an electric furnace, the efficiency of purifying
depends on the installation building as well as on
the exploitation technology of the equipment.
Conclusions
The dioxins concentration can be reduced at
much lower levels than those imposed by the
environmental legislation in force, due to the
combination between the thermic treatment of the
primary emissions (combustion gases), with rapid
cooling and gas filtering.
The solutions proposed and experimented
correspond to the most advanced techniques
concerning environmental protection the
introduction of some clean technologies in
metallurgical engineering.
References
[1] Poluanti organici persistenti n mediul nconjurtor.
Brosura informativa a Asociatiei expertilor de
mediu 2003
[2] Jiroveanu, M., Oprescu St. Captarea si epurarea
gazelor n industria metalurgic prelucratoare.
EDP, Bucuresti.
[3] Pumnea C.,Ionit N.- Tehnologii n industria
prelucrtoare. EDP Bucuresti
[4] Gheorghe N.,Gheorghe D. - Poluarea n industria
metalurgic si chimic Editura Performantica,
Iasi 1997.
[5] Voicu V. Combaterea noxelor n industrie ,
Editura Tehnic, Bucuresti 2002
[6] * * * Technical and Management Issues, pag.65
[7] Lehner J., Friedacher A., Gould L., Fingerhut W.
Des solutions economiques pour lelimination des
dioxines dans les fumees des fours electriques a arc
La Revue de Metallurgie nr.1-2004
123
THE TEORETI CAL CONSI DERATI ONS ABOUT THE PRECI SI ON
I NCREASE OF CONI CAL DRAWI NG PI ECES
Lucian V. Severin
1
, Traian Lucian Severin
2
1
University of Suceava, Romania, severin@fim.usv.ro
2
University of Suceava, severin.traian@fim.usv.ro
Abstract: The paper presents same theoretical considerations about the precision increase of conical
pieces drawing, among other things the angle calculus work zone of tools. This is influenced by
initial dimensions and the initial thickness of the work piece and the size of deformations.
Keywords: conical pieces, precision, drawing, deformations, angle.
1. Introduction.
The process of drawing conical pieces with
rigid tools is developing under more difficult
conditions compared with the cylindrical
parts. In this mode the deformation processing
is transmitted from the punch to semi-product
through a smaller area with the same degree
of deformation and therefore drawing stress
have greater values.
Also, in comparison with the cylindrical
drawing, the interstice between deformation tools,
at the beginning process has great relatively
values, what determined taking technological and
constructional solutions in technological
equipment conceiving, corresponding the relative
sheet metal thickness [1].
Comparative analysis of mechanical
deformation material scheme shows that these are
different in walls pieces (fig.1). Thus, in conical
pieces case the stress state is plane and the
deformation state is spatial.
The low and middle conical pieces have a little
deformation degree of the material and these
determined elastically arching when the pieces are
removed from the die, particularly at reduced
relative thickness of sheet metal. This difficulty is
eliminated if use at drawing processes the moulds
with special construction (moulds with strong
pressing of semi-product by blank-holder moulds
with redrawing rib and so on) [1, 2].
Low and middle conical pieces drawing
process, with a high precision realized in a single
operation, in many situation is needed to be
realized the drawing pieces with the dimensional
calibration. For that it must angle work zone
tools adequate determination. If these are realized
at adequate value from constructive reference
material, is not adequate realized the dimensional
calibration, or it must great deformation forces by
volumetrically calibration of conical piece wall
(fig.3).
o
1
o
3
o
2
o
3
o
2
o
1
c
1
c
3
c
1
c
3
c
3
c
1
c
2
R
M
R
p
Figure 1Mechanical scheme of
cylindrical pieces. drawing.
124
It is not produced a good calibration of piece,
because so that is showed by the deformation state
of semi-product in upper conical piece wall zone,
the thickness increase, from initial semi-product
thickness s
0
, tothickness s, at upper conical piece
wall.
The question is how to determine the
angles work areas of active elements because
at the end operation processing calibration
track its makeover on entire surface
2. Upper piece wall thickness calculus
It is considered a semi-product with the
thickness s
0
and radius R
0
(fig. 4). On this semi-
product is considered a point M placed on a circle
with radius o. After conical drawing of semi-
product, this point take positionM
1
on piece wall,
at distance r to piece axis, where wall piece
thickness iss
1
If it is considered a reference system in
cylindrical coordinate (o,u z) orientated on the
main directions of stress and deformations can be
writing:
3
0
1
2
c c c
c
u
= = = =
s
s
ln ;
r
ln
z
(1)
where c
o
i c
z
are the real main deformations in
point M
1
.
Writing Hooke relations between stress and
deformations for a plain stress state in this point,
and the constance volume relations in plastic
deformations, result:
0 = + +
=
z
z
z
; c c c
c c
c c
o
o
u
(2)
If it is note , a =
u
o
o
can be writing:
u
c c
a
a
z
+
=
2
1
(3)
Using relations (1), relation (3) become:
r
ln
a
a
s
s
ln
+
=
2
1
0
1
With this relation can be calculated the
thicknesss
1
by relation:
a
a
r
s s

+
|
|
.
|
\
|
=
2
1
0 1
(4)
In upper conical piece wall the thickness s is
determined with relation (4) where is substitutedo
with R
0
, r with R and a with zero value, because
the stresso
o
has zero value in point A. Therefore it
resulted:
2
1
0
0
|
|
.
|
\
|
=
R
R
s s (5)
If note the report R/R
0
with m, report what
defines piece drawing coefficient at the upper
wall, relation (5) can be write with relation:
m
s
s
0
= (6)
Suitable relations (4) along conical wall exist
a position of the point M
1
wherec
z
=0.
This position is obtained if a=-1 .Thus can be
writing [2]:
R
M
R
p
o
3
o
2
o
1
o
2
o
3
o
2
o
1
c
1
c
2
c
2
c
1
c
3
c
3
c
2
c
1
c
3
Figure 2Mechanical scheme of conical
pieces drawing
Figure 3. Drawing scheme with piece
calibration
R
M
R
p s
0
s
125
1
1
0
0
=
|
|
.
|
\
|

=
o
R
ln
R
ln
a
c
c
and by solving is
obtained:
0
0
607 0 R ,
e
R
= = (7)
This is the point M position on semi-product
what after drawing arrives on piece in thepoint
M
1
, without the change of material thickness.
3. Establish the angles of active zone tolls.
Because of the deformation state described
by the relation (4), the thickness conical wall of
drawing piece is variable.
The greatest thickness value is calculated by
relation (6) and this is in point A
1
, because the
material deformation degree is greatest in this
point.
The drawing process with calibration piece
dimensions must to be making on the mould with
corrected active zone tools with angle (fig.3).
If piece angle o is quoted on middle
thickness, then can be writing:
| o o =
p
and | o o + =
M
(8)
If piece angleo is quoted at piece inside, the
work zone tools will have angles:
o o =
p
and | o o 2 + =
M
(9)
and it is quoted at piece outside
| o o 2 =
p
and o o =
M
(10)
In relations (8), 9) i (10) it is noted:
piece angle-o; punch angle-o
p
; die angle-o
M
; angle
between inside generatrix of conical surface and
the outside generatrix surface -2.
Angle size can be established with the
position of point M defined by relation (7),
tacking account the semi-product thickness is not
modified in piece processed. Wall thickness in
point A
1
is determined by relation (5). If it is noted
generatrix length between point A
1
andM
1
withL,
can be writing:
L
s s
arctg
2
0
= | (11)
where is the angle between conical surface of
inside generatrix or conical surface of outside
generatrix and middle conical surface generatrix.
The sizeL can be determinate with relation
determined by material volume equality included
in plane circular crown with 0,393R
0
width with
the material volume included in conical piece wall
with middle generatrix by lengthL.
This volume can be calculated with the
second Pappus-Guldin theorem, determined with
relation [3]:
c p
x L
s s
V t 2
2
0
+
= (12)
where, figure 3 corespondly, x
c
has the value:
D=2R
s
L
s
0
0,607R
0
D
0
=2R
0
o
Rp
D=2R
L
y
c
x
c
s
s
0 2
Figure 4. Wall thickness change of drawing conical
piece
A
A
1
M
M
1
A
1
C
R
R
0
o
r
z
A
A
1
M
M
1
126
o sin y R x
c c
= (13)
where :
0
0
2
3 s s
s s L
y
c
+
+
= (14)
By y
c
substitution in relation (13) with the
value determined by relation it is obtained:
o sin
s s
s s L
R x
c
0
0
2
3 +
+
= (15)
Material volume of semi-product what is
processed in piece wall with length L is obtained
by relation:
2
0 0
632 0 R s , V
s
t = (16)
By equalization the relation (12) with (17)
and taking account by relation (6) and (15)
results:
, , ,
, o
o
sin m
sin m m , m m m m m R
L
2 1
2 842 0 1
2
3
2
2
0
+
+ + +
=
(17)
Substituting the length L determined with
expression (17) in relation (11)and taking account
by relation (6) it is obtained:
, ,
, , ,
(
(
(
(
+ + +
+
=
sino 2m m 0,842 m 1 m m m m
sino m 2 1 m 1
m 3R
s
arctg
2
2
0
0
(18)
With the angle size obtained with relation
(18) substituting in relations (8), (9) or (10) and it
are determined the work zone tools of the conical
piece with calibration drawing.
4. Conclusions
The thickening angle of the conical piece
drawing wall depend on initial semi-product
dimensions used, the initial thickness and
drawing coefficient in piece processing
determined on greatest diameter of this.
The angle size increases with growth of
initial semi-product thickness, with increasing
the degree deformation and decreases with
increasing semi-product deformed radius at
the same degree of deformation and the same
initial thickness.
4. References
[1] Romanovski, V., P., Stantarea i matritarea la
rece. Editua tehnic, Bucureti, 1970;
[2] Teodorescu, M., Al., Tehnologia presrii la
rece. Editura didactic i pedagogic, Bucureti,
1980;
[3] Bachman, K., H., Mic enciclopedie
matematic. Editura tehnic, Bucureti, 1980
127
METHOD FOR THE CONTROL OF THE GEOMETRI C AND
CONSTRUCTI VE ELEMENTS OF THE SPI RAL-FACE AND END
MI LLI NG CUTTER
Strjescu Eugen
1
, Pavlov Olimpia
2
, Dogariu Constantin
3
1,3
Politehnica" University of Bucharest-Romania,eugen_strajescu@yahoo.com
2,3
S.C. Munplast S.A. ostefu@yahoo.com.
Abstract: In the paper are presented the bases of a methodology for the determination of all the
geometric and constructive elements of the cutting tools for the milling machines, starting from the
getting of a solid model 3D of the measured tool. There are shown the possibilities to obtain a 3D model
and the graphic model of the geometric parameters control.
Keywords: Control, Cutting Tools, 3D Model, Geometry.
1. General I ntroduction
Unprecedented development of material
manufacture and technical orientation toward
higher accuracy of products required development
of the cutting tools' construction, made by using
new materials to increase the precision of their
execution. Increase bad-dimensional and geometric
precision of the tools required technological
processes and new machinery and apparatus at the
highest level.
The necessity to exactly endeavor the
dimensions and cutting tools' geometry, especially
complex, assembled or curved cutting edges and
ordered development of techniques and different
technologies, which would shorten the time needed
to control and would permit to obtain endeavor
parameters at all cutting edges, to verify if all the
geometric parameters are real and included
between acceptable limits, more or less close to the
optimum.
To control the cutting tools' geometry there are
imagined, developed and built a large variety of
devices and control instruments, from simple
templates, rulers, sublime and reporting workshop
microscopes to larger or smaller cars and the
coordinated measures [Minciu et al. 1995].
Currently, the explosive expansion of
computer's use and devices permitted a new
approach of the problems of the cutting tool's
control, based on other principles, that we did not
find in the autochthon literature or in the strange
literature [Strajescu 1989].
The original idea developed further make use
one of innumerable programs (software) made by
firms specialized in CAD, CAM (3D model-
horizon), in order to determine all the geometric
and constructive elements of cutting tool using 3D
solid model of the studied cutting tool obtained in
advance.
2. Obtaining 3D Solid Model Of Cutting
Tools
The methodology that we propose is based on
the getting of a 3D solid model of the controlled
cutting tool [Enache et al. 1988]. As is easily
intuited, if the project of the tool is made with a
computer using software CATIA, Solid Works,
Solid CONCEPT, the 3D model exists, because
that model is generated by the software. It is
known that at the design of the cutting tool, there
are some preconditions which allow the
construction of integral tool. Using 3D model
made on this route would allow the designer
visualize all geometrical elements at all edges of
the cutting edge which classical design
methodologies not allow. Other two possibilities to
obtain 3D model are taken into consideration
toward the authors of this article. This 3D model is
obtained by scanning the cutting tool (modern
scanners will produce geometric precision and high
dimensional, maximum deviations hundredth of a
millimeter is now toward normal), and by
128
obtaining the 3D model using photography, photos,
and treatment with specialized software.
2.1. Obtaining 3D model of the phase of
projects hard.
The trend of placing in the cutting tools' current
design of the technical computing, imposed and
developed by the personal computers requires a
complete description of the geometric shape and
dimensions of tools.
The assisted design of the cutting tools can be
optimized by programs.
For the cutting tools of any type, design
subsystems are built in the next order:
1. elaboration of the principles for the
development projects;
2. parameters' description of the processed
parts;
3. implementation of the model for calculating
the geometric and constructive elements of the
cutting tools;
4. elaboration of the block scheme of
calculation;
5. developing computer programs.
In order to realize an automatic design of the
cutting tool with efficiency, it would essentially
determine the correct informational structure of the
cutting tool. In this respect, it should be clarified
the following connections:
a) spatial, which determine place and
sequence of placement;
b) function, which determine the parameters'
size (e.g. parameters of the constructive elements
which determine strength and rigidity);
c) external, upon the interaction's character
and conditions with the manufactured piece.
The external connections determine the initial
data formation, the external factors being very
numerous.
The multiple integration, multidirectional,
programs from the computer-assisted sphere,
specialized programs on the problems of
generating geometry (3D drawing), and simulation
and calculation of static and dynamic behavior
presents a lot of advantages.
We gate away from the pre-project until to
reach the final product.
So, the successive stages are: specification of
the project's theme, the preliminary drafts,
realization of the pre-project, the model calculation
and analysis with logical for cinematic and
dynamic simulation, the results' getting, the
decision on the accuracy, the finite element
analysis, decision on the analysis' results, the
realization of the prototype or experimental
product, experimental research, simulation and
optimization through testing and confirmation of
the procedures.
There is currently a tendency to extend the use
of the solid model geometry, which is a
generalization of the geometry in space.
A solid model provided a complete and
consistent definition of the piece geometry.
Principles of the edition remain available in three-
dimensional space, with the remark that changes
affect mass properties of the ensemble (center of
gravity, moment of inertia) that can be calculated
automatically.
The parametric design of the simple parts or
assemblies is actually an effective concept and
represents the way to suit for the purpose of this
paper. This concept design integrates design and
analysis steps.
Using generalized cutting tools on different
types of tools are included in the concept of
parametric design.
In this case are necessary complex calculations
and the project no longer analyses directly the
piece geometry. Analysis is made on a generalized
or idealized model to be complete in a timely and
with an acceptable coast. Switching to an ideal
geometry ask for a fundamental change in the
information model. Idealized model may be
different depending on the type of analysis (e.g.
static or dynamic, structural or thermal, linear or
nonlinear). A simulated event is equivalent to a
single running of an experiment. Therefore it must
perform a statistical analysis of a set of
experimental results. The accuracy is higher and
the confidence in the results is greater when the set
of runs is greater. It is possible to run multiple tests
under different conditions. The results can be
statistically processed.
Following such a process to obtain high
accuracy with a 3D solid model of cutting tool that
can be cut with all the desired measuring planes,
the programs used for this purpose may put up all
searched angles, without calculations or other
operations.
2.2. Obtaining the 3D model by photography.
An absolute new path in the field, usable to
obtain a 3D model can be visualized today by the
existence of some computer programs that allow
reconstruction 3D model from photos made in
fixed positions that encircle the subject of the
129
wanted model. Initially such programs were
developed for the architecture, but appear variants
with net superior precision that can be applied to
the cutting tools. Software developments in this
area will bring increased safety with precision and
dimensional accuracy. We must say that these
programs function today, but the necessities for the
obtained images' processing are great time
consuming.
The D Sculptor computer program permits to
create computer photo-realistic models of a large
domain of objects, using common pictures of
objects in a relatively easy mode and quite fast. No
need for special hardware items, but only a
computer and a camera. It is recommended a
digital camera, but it is possible to use scanned
images.
Figure 1: The main screen of D Sculptor software.
The basic processes for the composition of the
base model are the next:
- it is placed the object for which we need the
model on the calibration plane (fig. 1);
- the object is photographed from multiple
angles;
- the pictures are imported in the D Sculptor
soft;
- D Sculptor automatically detects the model;
- the outside lines of the object in each
photography are marked, using the tools for
masking from the D Sculptor menu;
- D Sculptor computes the three-dimensional
model;
- the three-dimensional model is exported for
use with other software (as it is necessary shown).
While D Sculptor 2.0 brought improvements
both at the technical level and in terms of interface.
You can create models faster than before, and D
Sculptor 2.0 Professional last version has a high
accuracy.
2.2. Obtaining 3D model by scanning
A convenient way to achieve easier a precise
3D model of a cutting tool is by scanning and
reconstruction model. The method is really simple,
but its application is limited by the high coast of
the scanner device and, some time, by the necessity
of an ulterior addition of the model. This last
action can be complicate if the scanned object is
complex and have many concave surfaces or
masked by other parts surfaces.
Figure 2: The solid model for a lathe cutting tool. The
main screen of D Sculptor software.
For the present paper the solid model was
obtained using a scanner 3D (Model 2020i
Desktop 3D Scanner) that gave very good results.
In conclusion at the chapter 2, it is possible to
tell that the use of the 3D model obtained in the
project phase permits the integral determination of
the geometric and constructive elements in every
point of the active surfaces or cutting edges,
avoiding for the designer ulterior surprises.
The getting of the model by the two other ways
(photographing and scanning) permit the complete
and precise determination of the geometric and
constructive elements of the studied cutting tools.
3. The Use of the Solid 3D Model for the
Control of the Cutting Tools' Geometry
Starting from the solid model obtained in one of
the ways described above (Fig. 2) and using a
design mean as CATIA or Solid Works, by simple
commands consecutive to the points assessment in
which the geometry is measured it is possible to
draw planes considered as main planes: the
orthogonal plane P
O
, the tangent plane (actually the
plain of the edge projection on P
r
) P
T
,, the normal
P
n
(less important for practice, but extremely
useful in complex geometric calculations), P
f
frontal plane and the posterior plane Pp.
4. Application of the Use of the Solid 3D
Model for the Control of the spiral-face and end
milling cutter
130
This kind of tool work simultaneously two
perpendicular surfaces, having teeth on the
cylindrical and frontal part.
The exterior diameter is established in function
of the technological needs and of the tool type,
being content generally between 1 and 160 mm.
For the diameters between 1 and 75 mm there
are preferred the constructions with a cylindrical or
conic tail, and for the bigger ones there are used
constructions with circular holes.
Figure 3: Intersectii cu plane normale frontale.
Figure 4: Intersections with planes perpendicular on
the cylindrical helix
The used plans (perpendicular on the cutting
edges, named in ISO Standards orthogonal planes
P
O
) for the frontal or cylindrical teeth are shown in
the fig 3 and 4. Using the program possibilities, the
values of the cutting angles
n
and
n
are precise
for different points of the active edge.
0 0 0 0
- 3. 9 9 - 3 . 99 - 3 .9 9 - 3 . 99
- 5
- 4
- 3
- 2
- 1
0
n
n
n 0 0 0 0
n - 3 . 99 - 3 .9 9 - 3. 9 9 - 3 . 99
M 1 M 2 M 3 M 4
Figure 5: The graphic representation of the
variation of the angles value for 4 point.
Figure 6: Intersections with orthogonal planes.
69.38
67. 95
75. 33
73. 32
- 4 - 4 - 4 - 4
- 10
0
10
20
30
40
50
60
70
80
M1 M2 M3 M4
0
0
Figure 7: The representation of the section planes
in which are determined the
O
and
O
angles and the
graphic representation of the variation of the two
angles value for 4 points.
131
4.1. I ntersections with tangential planes
After the intersection of the tangential plane
with the reference plane, it was obtained the angle
K
r
, shown in fig. 8.
The method is especially useful for complex
tools, small, with the active surfaces and curved
cutting edges, to which access control with current
instruments is very slowness or impossible and
where the definition of theoretical planes control
angles is difficult to apply.
Figure 8: The tool sectioned with different planes in
which is shown the angle
r
.
66. 55 66. 33 66. 21 66. 09
0 0 0 0 0
10
20
30
40
50
60
70
k
k
k 66. 5 5 66. 3 3 66. 2 1 66. 0 9
k 0 0 0 0
M 1 M 2 M 3 M 4
Figure 9: The Graphic representation of the
r
angles variation..
10. Conclusions
A proposal for a complete control of cutting
tools made on the basis on the getting of a 3D solid
model is new and brings the first advantage that it
is possible to obtain all the angles searched at any
point of the cutting edges or active planes. In the
same time, the solid model permits at any time to
obtain again the lost values. Our research led to the
stroke for the control of a milling tool very low,
but getting a solid model consumed time and
resources, especially to correct the model obtained
by scanning, correcting imperfections due to their
software interpreting the results of the scan.
The other method of obtaining the 3D model
through photography has its limits, given by the
precision low shape recovery, but the rate of
increase in performance software creates good
premises for future use. An important observation
concerns the tools with edges and surfaces
curvilinear, at which we can determine all the
geometric and constructive parameters, achieving
with the help of additional software included in the
software graphics of the parameter's values of all
edge points. Further research will develop
methodologies for control of complex tools for
very large or very small, with curved surfaces,
methodologies able to change the angles of the
long edges or surfaces with steps as small, based
on programs developed independent and included
in the design software used for analyzing 3D
model.
11. References
[1] Minciu C., Strjescu E., .a., (1995) Scule
chietoare, ndrumar de proiectare. Editura
[2] * * * Standardul ISO nr. 6599/2/1982
chiere i scule achietoare /2. ISR
Bucureti.
[3] Strjescu, E., (1984) Proiectarea sculelor
aschietoare, Litografia I.P.Bucureti, l984.
[4] Enache, t., Strjescu, E., Minciu, C., (1988),
Metode i programe pentru proiectarea
asistat a sculelor achietoare. Litografia
I.P.Bucureti.
[5] * * * Standardul ISO nr. 6599/2/1982
chiere i scule achietoare /2. ISR
Bucureti.
[6] x x x Scule achietoare i portscule pentru
chierea metalelor. (1988 - 1989), Colecie
STAS, vol. 1 i 2, Editura Tehnic Bucuretti..
[7] Duca Z., (1967), Teoria sculelor
chietoare. Editura Tehnic, Bucureti.
132
133
ZI NC COATI NGS ON STEEL SUBSTRATE ATTAI NED BY DI FFERENT
ELEMENTS ADDED
Radu Tamara, Vlad Maria
University Dunarea de J os of Galati Romania
Abstract: The purpose of this work was to identify the influence of different process variables, such
as bath temperature, immersion time and bath alloy additions, on the morphology and coating thickness
attained using innovative alloys systems containing different percentages of Ni, Pb, Sn and Bi. Steel
samples were galvanized by the hot-dip method in micro alloyed zinc baths at different temps and
temperatures Experiments were aimed to obtain zinc coatings with increased resistance to corrosion,
which adheres very well to steel support, not conducive Fe-Zn reactions, not have negative environmental
impacts and presents uniformly dispersed phase structure in zinc matrix. Layers obtained in micro
alloyed zinc with nickel, bismuth, tin, lead are analyzed.
Keywords: galvanizing, Ni, Pb, Sn, Bi, thick layer, corrosion.
1. Introduction
Hot dip galvanised coatings are widely used in
industry for corrosion protection of steels. The
steel is protected by the Zn coating through a
barrier effect and a galvanic effect, in which Zn
acts as the sacrificial anode while steel acts as the
cathode. Although Zn has high corrosion resistance
in most non-aggressive atmospheric environments,
corrosion problems of Zn coatings have been often
found in aggressive atmospheric environments
where salt and sulphur dioxide are present. To
overcome this problem, new types of Zn coatings
with higher corrosion resistance are needed.
To increase the resistance to corrosion,
adherence, limiting pollution and zinc-iron
reaction, its use more and more alloyed zinc melts.
Alloyed zinc melts in galvanising is ollso
important in the case of Sandelin steel galvanising.
It is characterized by the formation of thick layers
and which can be hot - dip galvanised only in
micro alloyed zinc melts. There is a great interest
in carrying out research which explores the
influence of the alloying elements in zinc melts on
coating layers characteristics and the
characteristics of the melts [1, 2, 3, and 4].
The paper aims to analyse the effect of small
quantities of nickel, tin, bismuth and lead on the
zinc melts.
Lead has an important role in zinc baths [2] that
must be taken over by another micro-alloying
element and the research in this area suggest the use
of bismuth. Bismuth isnt toxic [3] and like lead it
has the same fluidization effect on the melt also
decreasing the superficial force [4; 5].
There is a great interest in carrying out research
[1], which explores the influence of the alloying
elements in zinc melts on coating layers
characteristics and the characteristics of the melts.
Thick layer of hot dip galvanizing process varies
with temperature and immersion time and depends
on the chemical composition of melt. Elements of
micro alloying change the fluidity and surface
tension of zinc melt and consequently drainage of
zinc when support is extract from the melt [2, 3].
Immersion temperature affects both quality and
quantity of zinc deposit on the surface of bands
and produce ash, slag and dross. The usual
working temperature in galvanizing process is
chosen between 450-460
o
C. At lower temperatures
is more difficult adhesion between zinc and steel
and at higher temperatures (over 470
o
C) when the
layer thickness starts to decrease, Fe-Zn alloys
begin to break passing interface in the melt and
form a large amount of dross. Time of immersion
is determined by the thickness of steel sheet and
desired thickness of the layer. For a given
temperature, a given composition to melt and the
same work speed, increasing the duration of
immersion leads to a corresponding increase in
deposit weight. The optimal duration is determined
by technology. Times too small leading to
134
defective adhesion and uniformity, and too large
leading a strong attack, large layer of alloy Zn-Fe
and the emergence of such compounds in the melt
(matt, ash), [5].
2. Experimental research
In the framework of the research were analyzed
four different zinc melts micro alloyedwith nickel,
tin, bismuth, lead whose chemical composition is
It has galvanized steel sheet to chemical
composition according to Table 2.
Table 1. The chemical composition of micro alloyed Zn
melts
Alloy
Ni
[%]
Bi
[%]
Sn
[%]
Pb
[%]
Zn
[%]
Zn-Ni-Bi-Sn
0.16 0.71 2.95 - 96.18
Zn-Ni-Pb-Sn
0.16 - 2.88 0.72 96.24
Zn-Ni-Pb-Bi-
Sn
0.16 0.41 3.49 0.43 95.51
Table 2. The chemical composition of steel support,
in %
C S Mn P S
0,030 0,030 0.300 0,015 0,010
Al Ti V Ni Cr
0.046 0.002 0.001 0.008 0.025
For the preparation of the zinc bath, pure zinc
(SHG) was used.
Micro alloying with nickel, bismuth, tin, and
lead was made directly, using metallic elements,
finely crushed, followed by mechanical mixing.
Laboratory experiments at the micro alloying zinc
melts were performed in the temperature range
typical galvanizing processes, working at 450-
460
0
C.
Experimental immersion times were 3, 5, 8, 15
seconds. In the experiments did not apply any
control and uniformity process, coating thickness
resulting in free flow of zinc from the sample.
According to Zn-Ni phase diagram (Fig.1),
nickel and zinc forms intermetallic compounds and
solid solutions based compounds. Nickel is
insoluble in zinc and at the 418
o
C and 94.8% Zn
forms the ( +Zn) eutectic.
Alloying with Ni was made directly through the
use of finely crushed Ni, (fine pieces < 1mm) and
mechanical stirring. Micro alloyed process is
longer, being at least one hour at the temperature
of 700
0
C.
Assimilation efficiency is low because nickel
was lost in slag and dross (Table 3 and Table 4).
Figure 1: The Zn - Ni phase diagram[6]
Table 3. The chemical composition of dross
seample 1 2 3 4 5
Ni, % 0,024 0,021 0,034 0,022 0,018
Fe, % 0,024 0,024 0,021 0,025 0,024
Zn,% diff. diff. diff. diff. diff.
Table 4. The chemical composition of slag
seample 1 2 3 4 5
Ni, % 0,019 0,021 0,024 0,016 0,018
Fe, % 0,016 0,020 0,021 0,020 0,020
Zn,% diff. diff. diff. diff. diff.
Assimilation efficiency of the direct alloying
experiments was found in 77%, measured three
hours after the introduction of nickel (to calculate a
concentration of 0.21% Ni and 0.16% Ni was
obtained). In the literature it is recommended both
directly alloying with metallic nickel and the use of
alloys with a maximum of 5% nickel [5].
In the equilibrium diagrams of Zn-Bi (Fig.2),
there is an insolubility of the two metals who
forms a eutectic at 97.3% Bi and 254.5
0
C. Given
the low melting temperature of bismuth (271
0
C)
micro alloying was made with metallic bismuth,
grinding and mixing in the melt by mechanical
stirring.
135
Assimilation process of bismuth in the melt was
stable, maximum efficiency is obtained. Bismuth is
a micro alloying element used to replace lead, the
same effect of melt fluidity and reduction of
surface tension without being toxic.
Micro alloying the melt with 0.1% Bi,
influences surface tension and fluidity similarly
using a Pb content of ~ 1% [5,7]. Although
bismuth is more expensive, the quantity needed for
alloying bath is much lower, costs are
compensated. Bismuth is also very stable in the
melt and requires replenishment just proportional
added zinc.
Figure 2: The Zn - Bi phase diagram[6]
In the experiments, micro alloying with tin was
made with metallic tin. Sn-Zn equilibrium diagram
(Fig.3) shows a total insolubility of this element in
zinc with the formation of a eutectic at 198.5
0
C
and 91.2% Sn. Tin, like most analyzed elements
form intermetallic compounds with nickel.
Figure 3: The Zn Sn phase diagram[6]
Microstructure of coatings obtained is presented
in Fig. 4. It shows a thin layer of intermetallic
compounds Zn-Fe and metallic layer consisting of
intermetallic compounds (Zn-Ni, Zn-Ni-Sn, Ni-Sn,
Ni-Bi) finely dispersed in zinc matrix.
a) Ni-Sn- Bi
b) Ni-Sn- Pb
c) Ni-Sn- Pb-Bi
Figure 4: Microstructure of coatings obtained in
zinc alloyed
Analysing changes in layer thickness of
coatings obtained from micro alloying with nickel
(and elements for improving fluidity and structure)
shown in Figs 5, 6 and 7. At 450
0
C, is observed
poor uniformity and high values of thickness,
compared with 460
0
C operating temperature of
melt. After the trials for these types of coatings is
90m
90m
90m
136
proposed technological temperature of 460
0
C and
maintenance period of 3-5 seconds. The elements
of micro alloying used can enhance the surface
quality of the zincked steel by: uniformity, texture,
and luminosity. This surface layer is dependent on
the melting fluidity, its superficial tension and the
solidification characteristics.
48
39
35.5
43.3
38.5
34 32.5
37
30
32
34
36
38
40
42
44
46
48
50
3 5 8 15
Imm er s i o n t i m e, [ s ec ]
L
a
y
e
r

t
h
i
c
k
n
e
s
s
,

[
m
]
T =450 0C
T =460 0C
Figure 5: Layer thickness variation depending on
temperature and duration of immersion, ( alloyed Zn
with Ni-Bi-Sn)
43.5
52
48
46
51.5
46
48
49
38
40
42
44
46
48
50
52
54
3 5 8 15
Im mer s i o n t i m e, [ s ec ]
L
a
y
e
r

t
h
i
c
k
n
e
s
s
,

[
m
]
T = 450 0C
T = 460 0C
with Ni-Pb-Sn)
63
39.5
61.5
57.5
55
38
36.5
33.5
30
35
40
45
50
55
60
65
3 5 8 15
Im m er s i o n t i me , [ s ec ]
L
a
y
e
r

t
h
i
c
k
n
e
s
s
,

[
m
]
T =450 0C
T =460 0C
with Ni-Pb-Bi-Sn)
According to the combination of the micro
alloying elements the surface can have different
types of metallic layer which can have an
important effect on the coating aspect.
One can notice the changing of the surfaces
morphology according to the micro alloying
elements as compared to surface morphology
resulted at the coating with pure zinc; thus, at
micro alloying with Ni-Sn-Pb, the surface
morphology reveals the forming of some big
crystals with a fan aspect (Fig. 8). If a part of Pb is
changed to Bi the crystals are significantly getting
finished (Fig. 9).
Figure 8: The coating morphology with Zn-Ni-Sn -Pb,
x50
Figure 9: The coating morphology with Zn-Ni-Sn-
-Pb-Bi, x50
For measuring the corrosion resistance, it was used
a potentiostate PGP type 201. A calomel saturated
electrode was used as reference electrode and a
platinum wire electrode as an auxiliary one. The
corrosive environment used for electrochemical
tests was a solution of 3% NaCl at room
temperature.
The samples were prepared for analysis by being
degreased with acetone, washed and dried [5].
137
From the experimental data obtained from
measurements we chose representations in the
form of Tafel, lg. I
cor
- f(E
cor
) polarization
curves. The analysis of the graphical
representations (Figs. 10-12) made it possible
to determine the characteristic quantities of
corrosion: corrosion current intensity I
cor
,
corrosion potential E
cor
, corrosion current density
i
cor
, corrosion speed v
cor
and penetration index p is
show in Table 5. Corrosion behavior in seawater
appreciated by electrochemical tests show a lower
resistance to corrosion coating with Zn-Ni-Pb-Bi-
Sn. Comparing the chemical composition of the
threetypes of coatings is a big difference in
content of tin (more than 3% in sample no.3).
Table 5. Values of the corrosion process
Tip of coating
i
cor
[A/m
2
]
v
cor
[g/m
2
h]
p
[mm/an]
1. Zn- Ni-Bi-Sn-Cd 8.6230 0.0253 0.0012
2. Zn- Ni-Pb-Sn 9.2242 0.0271 0.0013
3. Zn- Ni- Pb-Bi-Sn 13.8614 0.0405 0.0020
Figure 10: Tafel curve for Zn-Ni-Bi-Sn
Figure 11: Tafel curve for Zn-Ni-Pb-Sn
138
Figure 12: Tafel curve for Zn-Ni-Pb-Bi-Sn
Conclusions
- coating thickness is less at the same time adding
lead or lead and bismuth show a higher fluidity
for zinc melt.
- Microstructure shows a thin layer of
intermetallic compounds Zn-Fe and metallic layer
consisting of intermetallic compounds (Zn-Ni,
Zn-Ni-Sn, Ni-Sn, Ni-Bi) finely dispersed in zinc
matrix.
- At alloying with Ni-Sn-Pb surface morphology
reveals the forming of some big crystals with a
fan aspect. If a part of Pb is changed to Bi the
crystals are significantly getting finished.
- Corrosion behavior in seawater appreciated by
electrochemical tests show a lower resistance to
corrosion coating with Zn- Ni- Pb-Bi-Sn.
References
[1] J ohn Zervoudis and Graeme Anderson, p. 4, A
Review of Bath Alloy Additives and their
Impact on the Quality of the Galvanized
Coating, Teck Cominco Metals, Toronto,
Ontario, Canada 2007.
[2] Galvanizing Reactive Steels, a guide for
galvanizers and specifies, International
Lead Zinc Research Organization, 2009.
[3] Gagne, M., Zinc Bismuth Alloys for After
Fabrication Hot Dip Galvanizing, American
Galvanizers Association1997, Houston, TX.
[4] Krepski, R.P., the Influence of Lead in After-
Fabrication Hot Dip Galvanizing,
14thInternational Galvanizing Conference
(Intergalva85), Munich.
[5] Beguin, Ph., Bosschaerts, M., Dhaussey, D.,
Pankert, R., and Gilles, M., Galveco, A
Solution for Galvanizing Reactive Steel,
Intergalva 2000, Amsterdam.
[6] ASM Metals Handbook, Tenth Edition,
Volume 3, Alloy Phasse Diagrams, ASM
International, Metals Park OH.
[7] Tamara RADU, Florentina POTECASU,
Maria VLAD, Viorel DRAGAN , Research
on obtaining and characterization of zinc
micro-alloyed with bismuth coatings, Metal
2010.
139
ACCELERATI ON TEST MACHI NE
Marc J uwet
1
, Koert Bruggeman
2
, Filip De Bal
3
1
KAHO Sint-Lieven, Marc.J uwet@kahosl.be
2
KAHO Sint-Lieven, Koert.Bruggeman@kahosl.be
3
KAHO Sint-Lieven, Filip.DeBal@kahosl.be
Abstract: Road transport of cargo is increasing year by year. The European action program on
transport safety aims at an increased safety level. One important aspect of transport safety is load
securing: in normal road conditions - including sudden change of lanes and emergency braking - the
cargo should not move on the vehicle. In this text a very simple and efficient method for load securing is
described. The method is based on the idea of rigid load units. Also a test machine to certify and
optimize the rigidity of a load unit is described.
Keywords: road transport, cargo safety, acceleration test.
1 Cargo Securing Methods
The idea behind cargo securing is that the cargo
should not slide, tilt, wander nor deform on the
truck. Sliding and tilting can easily be understood.
Inertia forces that can cause sliding or tilting are
known very well: 0,8g in the forward direction,
0,5g in the sideward and backward directions.
Wandering is a phenomenon related to vibration of
the vehicle. At inspection it is found that the cargo
or part of the cargo has moved several cm or even
more in a random direction. Also deformation of
the load is unacceptable. Deforming load units can
cause unacceptable inertia forces and can endanger
the stability of the vehicle, e.g. a heavy piece of
metal sliding sideward in a rigid crate, can cause a
truck to turn over.
Cargo securing can be done based on 3
different principles, as outlined below.
1.1 Top Over Lashing Friction Lashing
Friction between the loading platform of the
vehicle and the bottom surface of the cargo should
prevent sliding. Friction is increased by top over
lashing: some lashings over the load pull down the
load and increase friction between the load
platform and the cargo. At the same time, friction
between several layers of the load is increased,
thus preventing sliding between layers. Top over
lashings at the same time prevent tilting and
wandering.
Top over lashing is very well known by almost
all truck drivers. On the other hand, it has three
major disadvantages:
1. In most cases the number of top over lashings
is very high. On average for a full truck of 25
tons up to 80 top over lashings are necessary
(depending on the type of lashings used,
depending on the materials and the position of
lashing points). Fixing and removing the
necessary lashings can take up to 2 hours,
causing an unacceptable workload for the
driver and an unacceptable cost for the
transport company.
2. The top over lashings have to be tensioned, in
most cases up to a tension force of 200 to
500daN depending on the design of the
tensioning device. A tensioned lashing can
damage the cargo unless some corner
protection is used. This will increase cost,
weight and material usage.
3. Very often the tension in a top over lashing
decreases during transport due to vibrations
and/or small movements of the cargo. In that
case, load securing is inadequate.
1.2 Direct Lashing
Load securing devices such as lashings, steel
wire ropes or chains are fixed between the load and
lashing points on the vehicle superstructure. For
un-deformable loads such as heavy construction
machinery, four different tensioning devices
are used to compensate for the inertia forces. For
palletized goods, one lashing device for each
direction is used: e.g. to compensate for the forces
during braking, a lashing is put horizontally in
140
front of the load and tensioned backwards at each
side of the load. Typically the actual tension force
in the tensioning device during transport is not
important. The tension will increase if the load
tends to move. This means that in most cases, the
load will not be damaged by the tensioning device
and that no repetitive tensioning during transport is
necessary. For palletized goods direct lashing can
be an economical solution if the load units do not
tend to tilt nor deform.
1.3 Blocking
The load is blocked against a certified part of
the vehicle or another part of the load. In this case
a so called XL-coded vehicle is very interesting:
the European standard EN12642:2006, code XL,
guarantees that the walls of a vehicle
superstructure will resist the forces from the load if
this load is un-deformable and put against the
wall without a gap. The walls of an XL coded
vehicle type are tested statically or dynamically.
The front wall resists a distributed load of 50% of
the loading capacity of the vehicle, sidewalls resist
40% and the back wall resists 30% of the loading
capacity. Such rigid walls and the friction between
load platform and load compensate for the inertia
forces. Gaps and deformation of load units are
unacceptable since these can cause impact forces
that are much higher than the inertia forces. Only a
small gap of some cm between the pallets and the
walls is generally accepted: 2 pallets of 120cm or 3
pallets of 80cm can be put between two sidewalls
of a standard XL coded vehicle without further
load securing. In many cases the loaded products
are somehow smaller than the pallet. Such a gap
between the products on neighbouring pallets is no
problem if the load units are really rigid.
Blocking is the most efficient load securing
method for palletized products. The load securing
efficiency does not depend on the efforts of the
driver. If the products are fixed rigidly on the pallet
and the XL coded vehicle is loaded without gaps
between pallets and vehicle walls, no additional
load securing action is required. A new XL coded
vehicle should not be more expensive than another
vehicle. No time is lost for load securing. The main
uncertainty is the rigidity of the load units.
2 Test Methods for Load Unit Rigidity
A load unit on a pallet that is blocked, has to
resist inertia forces of 0,5g or 0,8g depending on
the direction. The rigidity depends on several
parameters such as the product type (e.g. fertilizer
grains, liquid washing powder, concrete tiles, ),
the primary packaging (paper or PE bags, glass or
PE or PET bottles, carton boxes, ), the
secondary packaging (tray, American box, shrink
film, ) , the stacking pattern (interlocked, ),
the transport packaging (wrapped foil, stretch
head, tie sheets, corner protection, straps, ). It is
not possible to predict rigidity at a glance.
Therefore a test method has been developed.
2.1 I nclination Test
In the transport community, an inclination test
is well known. According to Newtons law an
inclination of 26,7 is supposed to be equivalent to
an acceleration of 0,5g. As Newtons law is not
valid for deformable goods, this test is not valid to
test the rigidity of palletized goods. Even for many
un-deformable goods this test is not applicable, e.g.
a concrete prefab wall of 10cm thickness and 2m
height will tilt if inclined under 26,7.
2.2 Vibration Test, I mpact Test
In the packaging community these two types of
tests are very well known. Vibration tests are being
used to simulate transport conditions in laboratory
circumstances. The deformation of the load unit
strongly depends on the frequency of the vibration.
Standardized frequency spectra or measured
frequency spectra can be used. In most cases a
vibration test will predict problems related to
primary and secondary packaging during road
transport, but it does not lead to an unambiguous
conclusion on transport packaging efficiency.
To test the transport packaging an impact test is
very well known. The load unit hits a wall with a
certain velocity. During such an impact, the
deceleration forces can rise up to 3 or 15g
depending on the damping material that is being
used. However the duration of these forces is very
short: 10 up to 80 ms. The effect of an external
force on a deformable product is a deformation
vague going through the product. The effect of the
external force will depend on the duration of that
force. During road transport, the inertia forces can
occur for seconds. Therefore in most cases an
impact test does not predict the effect of real
inertia forces in a truck.
141
2.3 Real Truck Test
Of course a real truck test is a relevant test to
certify the rigidity of a load unit. In practice a
certification test on a vehicle is very expensive and
difficult to do. The load unit is put on the loading
platform of a vehicle and sliding is prevented, e.g.
by putting a high friction material in between the
platform and the load unit. The vehicle is equipped
with some measuring system in order to measure
accelerations in forward and sideward directions.
A series of movements are performed with
increasing inertia forces. The load unit is inspected
permanently during this test.
2.4 Laboratory Acceleration Test
An acceleration test that can be performed in
laboratory circumstances appeared to be necessary.
Therefore an acceleration test machine has been
developed (Fig. 1). A table is put on sliding
bearing and can move along a horizontal trail. A
very high torque servo motor has been developed
to move this table. The acceleration of the table is
controlled based on a feedback loop in which the
actual position of the table is measured
permanently. The acceleration of the table can be
chosen between 1 and 10m/s. A load unit is put on
the table and is subjected to the chosen
acceleration in the same way as in a vehicle. The
deformation of the load unit during the
acceleration can be monitored by means of a high
speed video. The permanent deformation of the
load unit after the test can be measured in detail
after the test.
Figure 1: The acceleration test machine after
an acceleration test, with a failed load unit.
Up to 1200 tests with different load units have
been done until now. The influence of several
parameters of the test, has been examined.
2.4.1 Acceleration or Deceleration.
The acceleration test machine can be used in
two possible ways: the load unit can be accelerated
slowly (eg.0,2g) up to a certain speed and
decelerated with an accurately controlled
deceleration (eg.0,5g). Alternatively the load unit
can be accelerated in a controlled way (e.g. 0,5g)
in order to check the rigidity and decelerated
slowly (e.g. at 0,2g). The results of both alternative
tests should be the same for an un-deformable load
unit.
A real load unit of palletized goods is somehow
deformable and the results of both types of tests
are in most cases not identical: an acceleration will
deform the products in the backward direction.
During braking the deformed products are forced
to move back in the forward direction starting from
a deformed shape. In the second type of test, the
slow acceleration does not cause any deformation.
The deformation during braking starts from un-
deformed products. Therefore the acceleration -
deceleration sequence is the most demanding and
is chosen as the basis for a test standard.
2.4.2 Duration of the Acceleration.
The elastic and permanent deformation of a
load unit depends on the duration of the
acceleration. Several experiments have been done
with four types of load units: palletized bricks,
palletized bags with PE granulate, palletized
laminate planks, palletized carton boxes. About 10
identically packed pallets of each type have been
accelerated for different durations. Increased
duration of the acceleration results in increased
deformation up to a certain acceleration duration.
For palletized bricks a duration longer than 0,12s
does not result in more deformation. For palletized
bags with PE granulate, an acceleration over 0,25s
does not result in more deformation. Therefore an
acceleration duration of 0,275s is put forward as a
basis for a test standard.
2.4.3 Acceptable Deformation.
Almost all palletized load units will deform
under inertia forces, most often with some
permanent deformation. No scientific basis was
found to establish criteria for an acceptable
permanent deformation. Therefore several tests
142
have been done with different types of products on
different pallet types. Most industrially palletized
products are somehow understacked, meaning that
the product on the pallet is smaller than the pallet.
Around the pallet some cm of the pallet is not
used. On the other hand, some products are
overstacked, meaning that the product is larger
than the pallet.
Most often understacking is preferred since it
protects the products during manipulation.
Understacking and overstacking are very
unfavourable for the rigidity of a load unit, e.g. an
understacked load unit that is wrapped with foil
will tend to slide on the pallet if the acceleration
force is higher than the friction between products
and pallet. The wrapped foil will not prevent or
reduce sliding until the border of the product is
aligned with the border of the pallet. On the other
hand, sliding of the product on the pallet is not
harmful as long as the product does not continue to
slide over the border of the pallet. Therefore, a
criterion for an acceptable deformation must
distinguish between sliding on the pallet and
deformation of the palletized products. Several
tested products have been categorized by 20
transport or packaging experts in acceptable and
unacceptable permanent deformation.
Unacceptable deformation can be described as:
displacement before and after the test measured in
any horizontal plane is smaller than 6cm and
smaller than 4% of the height of the pallet.
3 Additional Advantages of Load Unit
Rigidity
The initial purpose of an acceleration test is the
verification of the rigidity of a load unit for reasons
of load securing. A rigid load unit in an XL
certified vehicle does not require further lashing if
gaps can be avoided. Some multinational
companies systematically implemented this
strategy for load securing. Unexpectedly this led to
two additional advantages.
3.1 High Speed Camera Recording
When doing an acceleration test with a
conventionally packed load unit, the deformation is
filmed with a high speed camera. In most cases the
deformation is unacceptable and in some cases the
load unit collapses. The deformed or collapsed
load units are very well known from pictures of
palletized goods arriving at their destination.
However, these pictures do not give a lead for
improvement of the packaging. This problem is
overcome by the videos. The cause of the failure
can be found, e.g. the damage starts with
inclination of the boxes on the second layer,
followed by tearing up the foil at the foot of the
pallet. The cause of the damage is not a lack of
rigidity of the foil at the foot of the pallet, but an
un-sufficient stiffness of the foil at the level of the
second layer. The transport packaging can be
improved systematically by correcting the primary
failure modes, in the above example by increasing
the pre-stretch of the wrapped foil at the level of
the second layer. At the same time packaging
material that does not contribute to the rigidity can
be removed. In many cases wrapped foil in the
upper half of a pallet can be decreased, the number
of tie sheets can be decreased,
The videos can provide information related to
the stacking pattern also. In most cases some
mixture of stacking strategy turns out to be most
efficient. Columnar stacking (one box on top of the
other with the same orientation) can be preferable
for the lower layers since this type of stacking can
carry higher loads. The stability of the load unit
can be increased by interlocked stacking (other
position or other orientation of the boxes in
different layers) in the top layers. The use of
interlayer sheets (paper, corrugated board, high
friction paper, ) strongly influences the optimal
stacking pattern. In case of a secondary packing
including corrugated board (trays, boxes, display
boxes, ), a global optimization of secondary
packaging, interlayer sheets, stacking pattern and
wrapped foil can give rise to considerable savings,
e.g. up to 20 per pallet.
3.2 Cost Reduction
Two independent large scale studies in Europe
conclude that approximately 4% of products are
damaged upon arrival at their destination. Probably
this percentage is even higher since most (small)
companies do not keep records of this type of
losses and they do not like to publish this kind of
figures. Insurance companies pretend that in some
industry sectors the percentage is about 6%. The
additional costs for the shipping company are even
more considerable (disturbed customer relation,
emergency delivery, claims, processing of
damaged products, ) The real cause of the
damage is in most cases not known: damage
related to manipulation, to forces occurring during
transport, to environmental conditions, . In the
143
past months the transport packaging of several load
units has been optimized and a rigidity up to 0,5g
has been certified. In at least 5 different cases the
figures of damaged products before and after the
optimization are available. On average for the 5
different cases, the percentage of damaged
products has decreased by 50%. Two types of
products are delivered as full truck loads directly to
the customer. At loading time in the production
plant they are most likely not damaged. They are
transported by road and inspected upon arrival.
Damaged products are in many cases not unloaded.
Damage is thus most likely related to forces
occurring during road transport. In both cases the
damage percentage was reduced by 80%.
4 Conclusion
Load unit rigidity is very important for efficient
load securing of goods during road transport. This
rigidity cannot be quantified with conventional
standard tests available in the packaging sector.
Therefore an acceleration test machine has been
developed and a new test standard is proposed.
The load unit is accelerated for at least 0,275s
and no significant permanent deformation is
allowed. This acceleration test does not only allow
for certification of load unit rigidity, it also allows
to optimize transport packaging in a systematic
way. In almost all cases this leads to a reduction in
packaging material and a reduction in damaged
products upon arrival. In many cases this even
leads to a reduction of the total packaging cost.
5 References
5.1 Standards Related to Packaging Tests
[1] DIN 55440-1, Packmittelprfung;
Stauchprfung; Prfung mit konstanter
Vorschubgeschwindigkeit
[2] ISO 12048, Packaging - Complete, filled
transport packages - Compression and
stacking tests using a compression tester
[3] ASTM D642-00, Standard Test Method for
Determining Compressive Resistance of
Shipping Containers, Components, and Unit
Loads
[4] ISO 527-3, Plastics - Determination of
tensile properties - Part 3: Test conditions
for films and sheets
[5] ISO 1924-2, Paper and board -
Determination of tensile properties- Part 2:
Constant rate of elongation method (20
mm/min)
[6] EN 60068-2-1, Environmental testing - Part
2-1 : Tests - Test A : Cold
[7] EN 60068-2-2, Environmental testing - Part
2-2 : Tests - Test B : Dry heat
[8] EN 60068-2-6, Environmental testing -Part
2 : Tests - Test Fc : Vibration (sinusoidal)
[9] EN 60068-2-27, Environmental testing -
Part 2-27 : Tests - Test Ea and guidance :
Shock
Part 2-64: Tests - Test Fh: Vibration,
broadband randomand guidance
[11] ASTM D4728, Standard Test Method for
Random Vibration Testing of Shipping
Containers
Part 2-30 : Tests - Test Db : Damp heat,
cyclic (12 h + 12 h cycle)
Part 2-80 : Tests - Test Fi : Vibration -
Mixed mode
Part 2-81 : Tests - Test Ei : Shock - Shock
response spectrumsynthesis
5.2 Books
[15] The ISTA 2011 resource book, International
Safe Transit Association, J anuary 2011
[16] M. J uwet, Basisregels ladingzekering voor
wegvervoer, 92p, Tandem, 2010
144
[17] M. J uwet, Rgles de base arrimage des
charges pour transport routier, 92p,
Tandem 2010
[18] M. J uwet, Basic principles of load securing
for road transport, 86p, Tandem 2011-03-28
5.3 Conference Papers
[19] M. J uwet, K. Bruggeman, Test method for
stability of load units, Dekra International
Conference on Load Securing,
Eurospeedway Lausitz, Duitsland, 8-9
October 09
[20] M. J uwet, K. Bruggeman, Opportunities for
transport packaging optimization, ISTA
Europe international conference, Valencia,
Spain, 26-28 October 2010
[21] M. J uwet, K. Bruggeman, Verpakken en
zekeren van lading bij wegvervoer , EVO
conference, Asten, Nederland, 18 november
2010
[22] M. J uwet, J . Dendauw, Methods to quantify
load unit rigidity, ITW seminar, Virton,
Belgium, 15 February 2011
[23] M. J uwet, Nieuwe Europese regelgeving
voor transportverpakking, Empack
conference, s Hertoghenbosch, Nederland,
15 maart 2011
5.4 Articles
[24] Zeng, F., Le Grognec, P., Lacrampe, M-F.,
Krawczak, P., A constitutive model for
semicrystalline polymers at high
temperature and finite plastic strain:
Application to PA6 and PE biaxial
stretching, Mechanics of Materials, 2010
[25] D. Briassoulis, A. Aristopoulou, Adaptation
and harmonisation of standard testing
methods for mechanical properties of low-
density polyethylene (LDPE) films,
Department of Agricultural Engineering,
Agricultural University of Athens, Athens,
Greece, November 2000
[26] Sangkeun Rheea, J ames L. Whiteb, Crystal
structure, morphology, orientation, and
mechanical properties of biaxially oriented
polyamide 6 films, Institute of Polymer
Engineering, The University of Akron,
Akron, USA, 8 J uly 2002
[27] Naresh Bhatnagara,_, Rakesh Bhardwaja,
Palani Selvakumara, Mathias Brieub,
Development of a biaxial tensile test fixture
for reinforced thermoplastic composites,
Department of Mechanical Engineering,
Indian Institute of Technology, New Delhi,
India, 6 September 2006
[28] Arthur Bobovitcha,b, Emmanuel M.
Gutmanb, Sven Henningc, Goerg H.
Michlerc, Morphology and stress-relaxation
of biaxially oriented cross-linked
polyethylene films, Department of
Engineering, Martin-Luther University,
Germany, 17 September 2002
5.5 Patents
[29] Test machine for palletized products,
application 2010/0155, March 9, 2010
145
DEVELOPMENT OF WASTE MANAGEMENT SYSTEMS I N AN
I NTEGRATED SHI PYARD
Daniela Buruiana
1
Dunrea de J os University of Galati, dnegoita@ugal.ro
Abstract. This paper aims to present a waste management system in the shipyard industrial area
and the solutions adopted to reduce the environmental impact caused by wastes from ship repair and
maintenance work. The waste impact (material resulting from the hull blasting , oil and paint wastes )
implies the occupation of the land in the vicinity of the shipyard studied, soil and water contamination
and landscape change. Even if some of them are inert, solid naval wastes cause storage problems ,
primarily because of the large quantities resulted . Identifying solutions for reuse of solid waste leads to
reduced costs and superior capitalization of those materials. The waste management policies to be
adopted by a shipyard must find solutions to their very source involving both reduction and recycling.
It is also necessary to limit the negative environmental impact by greening works in the already a
affected areas by the uncontrolled deposits
Keywords: management, wastes , recycling, environment, efficiency
1. General
An effective integrated waste management
must focus on the following purposes: identifying
the sources/activities generating emissions,
emissions monitoring and management, measures
to reduce emissions through the application of the
best practices in the sector concerned
Waste elimination and/or minimization should
be a priority in any integrated environmental
management plan. Full or partial recycling leads to
reducing the amount of waste being deposited, as
well as reducing the amount of raw materials used
in any technological processes
When referring to an efficient solid wastes
management in shipyards, it is necessary to first
conduct a thorough analysis of the processes of
construction, repair and maintenance of vessels
and the volume of materials used in shipbuilding
industry. The results of such an analysis would
provide for the knowledge, appreciation and
application of the best integrated waste
management solutions. For most shipyards, these
solutions would allow the recovery of waste in
other applications, cleaning the areas affected by
uncontrolled waste disposal .
The evolution of marine industrial systems in
the context of the modern concepts of total quality
and sustainable development promoted by any
environmental policy is based primarily on the
understanding of technological processes related to
marine environmental problems and restoring the
balance, particularly through recycling in the
relationship repair / maintenance- vessels-
environment recycling- material energy.
An issue of increased interest in the world
today, also supported by environmental policy
applied to a shipyard, is represented by wastes
from both an environmental and economic points
of view
Environmental concerns in the strategies pursued
and applied to shipyards around the world fall into
two directions:
- development of advanced technologies that
significantly reduce emissions;
- increasing waste recycling and recovery output
to levels close to 100%
2. Defining and identifying pollution sources
in a shipyard
Shipyards all over the world are faced with
the need to restructure their work so as to meet the
sustainable development requirement by
protecting the environment as efficiently as
possible since the activity in this sector requires a
major consumption of raw material and energy
while generating considerable amounts of polluting
emissions such as gas, solid waste and waste
waters. Compared with the practice and global
146
trends, Romanian shipbuilding industry features
deficiencies both in the collection, transport and
storage of all categories of waste and waste
recycling and /or reuse. The tendency to dismantle
hundreds of ships each year occurs continuously.
This is a consequence of rapid development in the
field of naval technology, the competition between
the carriers and the current crisis. We can give as
an example the situation in some Asian countries
where more than 80% of the ships are dismantled
in shipyards located on beaches in India,
Bangladesh, Pakistan and Turkey. Bangladesh is
the country where most ships are dismantled.
Most shipyards use the cheapest method, but
also the most harmful to the environment: "failed
on the beach" for dismantling ships. This method
has a high price in human lives and causes many
diseases due to exposure of workers to toxic
action. The European Economic and Social
Committee is aware that in the foreseeable future,
the failure of ships for dismantling on the beach
will remain the preferred method. Therefore,
current conditions in shipyards must be improved
so that they operate in a safe and ecological
environment
Figure 1: Technological processes specific to ship repair and maintenance sector and specific pollutant emissions
Blasting /
first treatment
Ship dismantling
Blasting and painting
on different portions
BLASTING AND
PAINTING HALL
Cleaning with
Solvents; chemical
Degreasing, removal
Assembling the
ship by welding
Maintenance
Painting
Sand, slag,
VOCs
Wastewater
Slag and smog
Sand, slag,
VOCs
Solvent
VOC emissions
Metal surface
cleaning, lubricating
products
Batteries, waste from
oil, VOC emissions
147
Conditions relating to environmental and social
protection are disastrous in South Asian shipyards.
They result in unfair competition against their
European counterparts. Moreover, high local
demand for recycled steel is an additional problem
for European competitiveness [2].
Worldwide, the shipbuilding technique in the
last decade has reached a high level of
performance, demonstrating a great capacity to
adapt to changes brought in by raw materials and
energy conditions, the need to increase
productivity, reduce consumption and comply with
the environment legislation that is becoming more
severe.
It can be considered the most representative
technological processes generating pollutant
emissions from a shipyard are those in the sectors
of dismantling, repair and maintenance of ships.
Pollutant emissions generated by the blast cleaning
of metal surfaces as well as the process of painting
or welding of metal components are the most
important Figure 1 illustrates the technological
processes in a sector of ship repair and
maintenance along with their pollutant emissions
and specific wastes.
Monitoring pollutant emissions in conjunction
with the process that generate them calls for, and
simultaneously allows for, the implementation of
an integrated environmental management in these
sectors.. Today it is accepted that, at the current
level of knowledge, is not yet possible to establish
a balance of environmental pollutants on global
development in the shipping industry. This the
main reason why, in their analysis on the
correlation energy-environment-waste or
recycling-environment, they still use energy
balances, material balances respectively, for the
process being analyzed
Material balances are made in companies in
order to be able to monitor the emissions based on
material inputs (Table 1). This model is promoted
by the shipyards in J apan, which managed to reach
what is found in the literature as " waste-free
industry" or "zero waste:
Table 1: Balance sheet materials emission and waste
Process Raw
materials/materials
Emissions
emissions
to air
waste
water
solid waste
Ship
dismantling
assembling by
welding
Electrodes Smoke
CO2
- Slag from
welding electrodes
Surface
Preparation
Sandblast Grit
Solvents
Degreasing Alkaline
and acid agents
Sandblast
Grit
Solvents
Dust
emissions
from the
cleaning
process
Solvents
and acid
water
Recyclable
waste grit
Slag Residues
from cleaning areas
Surface
Finishing
Sandblast Grit
Solvents Paint Primer
- - Waste Water
Containers of paint,
brushes
Reducing water pollution by increasing the
recovery of all categories of waste and their
controlled storage is a priority in environmental
policy analysis proposed in a yard analysis. As a
first step, it should be made rigorous management
of the waste generated at source with particular
focus on reducing the quantity and harmfulness of
waste, advanced recycling of waste by
reintroducing them into various stages of the same
technological flow
This ensures protection of natural resources,
increase reuse of waste materials by converting
them into raw materials for their own flows or of
148
other industries, controlled storage of all categories
of waste.
All these objectives are proposing an action of
inventory and management of sources generating
wastes and monitoring the pollution they cause.
Monitoring means the acquisition, storage and
processing input data to assess some processes
within the shipyard. This would allow making
correct decisions, argued on reconsideration of
technology and design in relation to each part of
the technological process
3. Peculiarities of waste management in a
shipyard
Waste management is a complex issue
consisting of a multitude of components. In
particular, waste management in a shipyard in the
European Union represents a particular challenge
in that it should not disrupt the European market of
ship construction and repair. Although there is no
single recipe that can be applied in all cases, the
European Union has firm principles underlying the
different levels of waste management systems [3]:
-principle of prevention - the production of waste
should be minimized and avoided if possible;
- principle of responsibility for polluting
generation and payment he who produces waste
or contaminates the environment must pay the cost
of his actions; precautionary principle
- potential problems must be anticipated proximity
principle - waste should be stored as close as
possible to the place where they were produced
The Romanian legislation stimulates the
polluter pays principle (Article 3 of Law
137/1997). The Ministry of Environment and
Water has transposed into national legislation on
the environment, which fully complies with this
principle. The basic principle of the Directive
96/61/EC on integrated pollution prevention and
control is best reflected in the concept of
sustainable development, which requires measures
to prevent. The concept of best available technique
involves taking all measures to prevent pollution
from the plant design phase and continue in the
construction, operation and decommissioning
stages. Also, the holder of activity is directly
responsible for any environmental damage he may
cause
In an environmental policy, waste management
is an essential component due to increased
production and subsequent increase in the wastes
with negative impact on the environment. The
negative effects of waste on the environment and
on human health are well known and publicized
throughout the world, and failure to adequately
approach this issue can have catastrophic
consequences
Therefore solutions must be found to reduce
the amount of waste generated and to limit their
negative effects, by reusing the recyclable wastes
in the production processes as secondary raw
materials (such as those resulting from the grit
blasting) or by neutralization and final controlled
disposal of those who cant be recovered. Waste
management is the model of industrialization
which allows increasing the contribution of
industry to achieve economic and social benefits
for the present and future generations resources
without damaging the environment.
In most cases, the dynamics of industrial waste
amount generated in a shipyard cant be precisely
determined due to lack of a consistent database and
change of the definitions. This problem is
particularly serious, because to find the total
volume of industrial waste generated, we should
know exactly correlated with the waste evolution
correlated with naval dynamics, then to develop a
coherent policy with objectives, criteria, measures
and investment costs.
The importance of cost / benefit analysis to
determine the overall economic efficiency of the
investment consider the direct economic efforts
(investment costs) and indirect economic efforts
(for example, pollution costs, costs allocated for
affected human health, environmental
rehabilitation costs , costs that are usually
overlooked or minimized).
4. Conclusions
A special chapter in the world today and
supported by the current environment policy of a
shipyard is represented by the wastes, both in
terms of environmental and economic points of
view. A successful management will lead to
finding the best solutions for the recovery of waste
generated and thus both costs and environmental
impact can be reduced.
The Concern for compliance with legislative
requirements on environmental protection and the
need for harmonization of the economic progress
with rational management of material and energy
resources should lead to the recovery of waste by
technologies that offer both economically and
ecologically optimal solution:
-rigorous management of waste and reduction of
waste quantity at source;
- controlled storage of all categories of waste
149
generated on the platform of a shipyard;
- advanced recycling of the waste produced by
their reintroduction into the various stages of the
process flow, thus ensuring protection of the
natural resources of raw materials;
- increased use of wastes by converting them into
raw materials for other industries.
We believe that a focus on a strong
management for the environmental protection as
part of the administration and operation of a
shipyard will enhance the environmental quality
performance. The evolution of marine industrial
systems in the context of modern concepts of total
quality and sustainable development promoted by
the proposed environmental policy is based
primarily on the understanding of specific
technological processes in conjunction with
environmental issues focused on restoring the
balance, particularly through recycling: process-
environment-recycling-energy. Environmental
excellence, incorporated into all processing
activities, can be promoted through the principles
supported by:
- continuous improvement of the
environmental performance through systematic
monitoring of the environment factors, minimizing
the environmental impact and using the principle
of pollution prevention;
- reducing waste at source, and where they
inevitably occur, they should be capitalized or if
this is not technically and economically feasible,
they must be controlled stored while avoiding or
reducing any impact on the environment;
- implementing clean technologies and best
practices in the field by investments leading to
environmental protection and savings of both
energy and raw materials.
References
[1] Negoi D, Contribuii la reducerea polurii
provocate prin funcionarea unui antier
naval, Teza doctorat 2007;
[2] Avizul Comitetului Economic i Social
European privind Comunicarea Comisiei ctre
Parlamentul European O strategie a UE
pentru ameliorarea practicilor de dezmembrare
a navelor, http://www.eurlex.europa.eu;
[3] Recycling Spent Sandblasting Grit and Similar
Wastes as Aggregate in Asphaltic Concrete
Naval Facilities Engineering Service Center
Port Hueneme, California 93043-4370.
[4] Rojanschi ,V, Bran, F., Politici i strategii de
mediu, Editura Economic, Bucureti, 2002.
[5] Negoi D., Dispersion of the atmospheric
pollutants on the platform of a shipyard,
published at the Sustainability for Humanity &
Environment in the Extended- Connection
Field Science-Economy Policy, Timioara, 24-
25 Febr. 2005, pp. 115-119.
150
151
METHOD FOR DETERMI NATI ON OF THE DI STANCE BETWEEN
ELECTRODES TO ELECTROHYDEAULI C UNDERWATER THROUGH
SCI NTI LLAS DI SCHARGE
Dumitru I ACOB
tefan cel Mare University of Suceava
Abstract: In this paper to present a method for fast-track determination of the distance between electrodes to
the discharge of the energy froma condensator batherry for a realize electrohysraulic effect, in function of the principal
work parameters of the installation, tension of loading U and capacity C, as well as maximal work tension of the
batherry
max
U .
Key words: electrical, energy, vapacitor, tension, capacity, electrohydraulic, effect.
1. I NTRODUCTI ON
It is knuw the fact that the
electrohydraylic effect to generate to the
sudden perforation of the electrolyte between
of the electrofs from the electrohydraulic
chamber, under electrical high tension, with
which is loaded yhe capacitor bathery, to the
sudden discharge of the electrical loaded
energy in the capacitor energy. To the sudden
discharge energy, to generate a pression
impuls which can be utilize in technological
purposes.
2. CONDI TI ONS FOR THE
REALI ZATI ON ELECTROHYDRAULI C
EFFECT
For the realization electrohydeaulic
underwater through scintillas discharge is
necessarytheelectrical perforation between
electrodes. There are following situations
respecting electrical perforation of the interval
from the two electrodes and namely>
- If rhe intensity of the electrical field
between electrods
m
KV
E 360 > to
realise sudden perforation, which is
favourable for execution of the
technological operations, because in
this case to generate the pression
impulse;
betwe enelectrods
m
KV
E
m
KV
360 30 ( ( to realise
thermal perforation, with heating and
place ionization of the liquide with to
realise the pression impulse;Clearly,
that this situation can not utilise in
teghnological purpose;
between electrods
m
KV
E 30 s the
perforation of the liquide is not
possible.
In other papers, to recommend that for the
determination of the maximum distance
between electrods which produce sudden
discharge, to utilise the equation:
j mm U K U C
E
2
1
2 3
max
06 , 0 = = o
(1)
in which:
U - the tension of loading pf the capacitor
batherry inKV;
C- the electrical capacity of the batherry
in F . From the first condition on the
electrical field from electrods, result
equations (2)and (3):
152
j j
j mm U U K
mm U m
U
C
= =
= =
=
777 , 2
360
1000
360
1
2
max
o
(2)
j j
j mm U U K
mm U m
U
= =
= =
=
333 , 33
30
1000
30
1
3
min
o
(3)
where:
naxC
o - the maximum distance for to
realise sudden discharge;
min
o - the minimum distance for which
the perforation of the liquide is not
possible.
The angles of inclination pf the rights,
in comparations wirh abscissa asis, are:
- For
C max
o :
0
2
20 , 70 777 , 2 = = = arctg arctgK o
(4)
- For :
min
o
0
3
25 , 88 333 , 33 = = = arctg arctgK |
(5)
From the equality equation (1) and (2) result:
2
1
2 3
2
06 , 0 U K U C U K = =
(6)
1
2
K
K
U
e
=
(7)
where
e
U is tension for which:
E C max max
o o =
(8)
The value 777 , 2
2
= K and the value
1
K for the capacity of the instalation from
laboratoy F C 70 = is:
247 , 0 70 06 , 0
3
1
= = K
(9)
For this case result:
KV
K
K
U
e
2 , 11
247 , 0
777 , 2
1
2
= = = (10)
The maximum distance between electrods in
this case is:
mm U K
e
2 , 31 24 , 11 777 , 2
2 max
= = = o
(11)
In figure 3 are represented the two curves:
j mm U U K
C
= = 777 , 2
2 max
o (12)
j mm U U K
E
2 2
1 max
247 , 0 = = o
(13)
To find that for
e
U U( , the condition
(15) is principalbecause if this is
accomplidhed, result that and the the
condition (14) is accomplidhed.
For
e
U U) the condition (14) is
principal, because if this is accomplidhed,
result that and the the condition (15) is
accomplidhed.
If the capacity of the bathery increase,
then the value of the tension
e
U desrease. For
example, if F C 350 = , the value
422 , 0 350 06 , 0
3
1
= = K and
KV U
e
6 , 6
422 , 0
777 , 2
1
= = .
153
So, with the increase of capacity, decrease
the interval for which the condition (15) is
principal. For the capacities very large, the
maximum distance between electrods, is
determined of the condition (14).
The optimal distance is:
2
max
o
o ~
o
(14)
because the variation of the electrical current
in the interval j
max
, 0o is about that in figure
4.
The value
E max max
o o = , if maximum work
tension of the insyalation
e
U U (
max
and
C max max
o o =
if maximum work tension of the insyalation
e
U U )
max
C
I - the intensity of the current determined in
very large part of electronic conducted
current: ;
T
I - the intensity of the current determined in
very large part of thermal ionization.:
For the capacity of the instalation
from laboratory F C 70 = , the tension
KV U
e
2 , 11 = and maximum work tension
KV U KV U
e
2 , 11 6
max
= ( = . So, in this case
mm mm
R
9 , 8 892 , 8
max
= = = o o . The optimaum
value result:
mm mm
E
45 . 4 446 . 4
2
892 , 8
2
max
0
~ = = =
o
o
(15)
vvalue with wich have work for
texperimentation.
The experimental determinations confirm
full previous assumpsions.
The technological work comain is under
curve OABC from the figure 3.
For the tensions KV U 6
max
s , proper for the
installation from plastic deformation
154
laboratoryof the mechanical engineering
faculty, maximum and the optimal values to
present in the table
Tabel 1.
3. CONCLUSI ONS
Virtually, for the determination of the
maximum and optimum distance from
electrodes, to underwater scintillas discharge,
to propose following methpdology:
- Function of the capacity of the
capacitor bathery, to determine the
value
2
K
- To determine the balue of the yension
e
U for which the two distances is
rqual:
j KV
K
K
U
e
1
2
= (16)
- To compare the maximum work
tension of the installation
max
U with
the value
e
U
- If
e
U U (
max
the maximum distance
between electrodes is
naxE
o ;
- If
e
U U )
max
the maximum distance
between electrodes is
naxC
o ;
- The optimal distance from electrodes
id:
2
max
o
o =
o
(17)
The value
max
o from the equation (19)
is equal with
naxE
o , or with
naxC
o ,function of
report
e
U
U
max
. Clear, that for
e
U U =
max
, is
whatever which from the two equation to
utilise, because to obtain same result.
REFERENCES
[1]. IACOB DUMITRU, Cercetri teoretice i
experimentale privind defprmarea plastic la rece a
semifabricatelor metalice utiliznd energia acumulat
n condensatoare, Tez de doctorat, Universitatea
Tehnic Gh. Asachi Iai, 1994.
[2]. IACOB DUMITRU, Braha V., Rusu B.,
Cercetri experimentale privind perforarea i
decuparea electro-hidraulic; Universitatea Maribor,
Slovenia, 27-29 octombrie 1994.
[3]. IACOB DUMITRU, Braha Vasile, Luca Liviu,
Cercetri experimentale privind razele de racordare n
cazul energiei minime obinute la perforarea i
decuparea electro-hidraulic; Revista Deformri
Plastice (J ournal of Plastic Deformation) Centrul de
Studii si Cercetri pentru Deformri Plastice,
Universitatea "Lucian Blaga" Sibiu, vol.1(1994)
Nr.2, pag.22-26, ISSN 1222-605X.
[4]. IACOB DUMITRU, Braha Vasile, La pression
maxime dans la chambre de travail a la deformation
par des chocs mecano-hydrauliqes; Buletinul tiinific
al I.P. Iai, TOM. XLII (46) Fascicola 1-2, Secia V,
Construcii de maini, 1996, pag. 136-140.
[5]. IACOB DUMITRU, Mironeasa Costel,
Deformarea electro-hidraulic, Editura Matrix,
Bucureti, 2008, ISBN 978-973-755-318-8.
[6]. IACOB DUMITRU, Mironeasa Costel, Camer
electrohidraulic cu electrod mobil n timpul
descrcrii, Brevet de invenie nr. 123073 din
30.09.2010.
TensionU, KV 1 2 3 4 5 6
Maximum distance
E max
o , mm
2 2
2 max
247 , 0 U U K
E
= = o
0,247 0,988 2,223 3,952 6,175 8,892
Optimal distance
0
o , mm
2
max
0
E
o
o =
0,123
0,1
0,494
0,5
1,111
1,1
1,976
2,0
3,087
3,1
4,446
4,5
155
SUPERI OR WASTE RECOVERY I N THE METALLURGI CAL
I NDUSTRY
Elisabeta Vasilescu , Ana Doniga, Alexandru Chiriac
1
Dunrea de J os University of Galati , elisabeta.vasilescu@yahoo.com
2,3
Dunrea de J os University of Galati , sandu.chiriaci@yahoo.com.
Abstract: On the global scale the generating rhythm of the scraps exceeds the natural assimilating
capacity of the most regional ecosystems.
The numerous studies mode about the good practices spotlighted the succe of the ecological
technologie, but expoed the exitence of many indutrial ineficient, polluant and dangerous installation too,
wich are, still, into operation. To the last evaluation, the industrialized countries should cut down to
minimum four times the row material consumation, so that the considerable effort are necessaryto
decrease the quantities of scraps resulted in the industrial processes and increase their recirculation.
A clean technology could rely on a new, less polluted production procedure, that recirculates the
resulted scraps in a certain phase of fabrication or, on the utilization procedure of the scraps becoming
raw material for a secondary production.
The conception of the scraps recovery and recirculation should be brought up to date emphasizing the
management and the marketing activities to find the utilization fields of the accumulated scraps in the
industrial dump yards.
Keywords: scraps, metallurgical sources, metallic slags, metallic powders
1. Scraps Sources from the Metallurgical
Plants
In the frame work of the technological flours,
typical to the iron and steel industries, besides
main products the important quantities of the by
products and scraps outcome. In generally, the by
products are re introduced in the technological
flow, boving high degree of utilization.
The scraps are stored on the particularly fited
out surfaces but, at present, the recirculating of
such materials is pursued by the nonpolluted
technologies as possible.
A classification of such scraps has in view the
following criteria:
-the content of the certain chemical elements,
depending on the technological process from
which they proceeded;
-risk degree related to the environment and
health impact;
-aggregation state at generation moment ; the
intinsic value of the scrap and its utilization
possibility;
-generation sources (technological flow and
operation which generated the scraps).The main
scrap generating sources from metallurgical fields
are: coke chemical plants, sintering, Blast
furnaces plants, Steelworks, Rolling Mils, Forging
plants, Thermal treatments shops, Mechanical
processing works and their types of scraps are:
powders, sludge (slurry), slags, refractory debris
chips (splinters), scales, etc. For example, from
rolling mills the scraps which the following
characteristics resulted (tables 1, 2):
156
Table 1Average grain size of the scale, slag and metallic melting.
Table 2Chemical composition of the metallic scraps
Scrap type Chemical composition (%)
Fe SiO
2
CaO Mn
Scale (%) 66.97 71.23 0.70 1.97 0.36 1.47 -
Metallic slag (%) 63.66 70.16 0.11 1.46 1.14 1.52 0.42 0.8
Metallic melting
(%)
74.48 78.37 0.30 0.45 1.08 1.35 0.43 1.13
Having in view the high iron quantity of such
scraps, their higher valorification was layed on the
table, aiming to the iron extraction and its use in
other fields.
One of the up-to-date valorification method of
ferrous scraps is the powder getting used for small
parts manufacturing, with complex geometry, of
which fabrication, by casting, isnt profitable from
technical and economical point of view, or parts
manufacturing which, by their operation nature,
should have a porous structure (filters, self-
lubrication bearings, porous plates for
accumulators etc.).
At present, the utilization methods of the
metallic powder-scraps should answer to the
following questions:
-getting of the acceptable purity powders for
those fields where they will be used;
-achievement of the wished phisical (shape,
grain-size etc) and technological characteristics
(compression, sintering capacity)
-acceptable cast of the row materials; resources
availability; value of the processing investments.
A large range of the products
manufactured by the typical methods of the
powders metallurgy is that of the iron and steel
products with higher characteristics, used for
mechanical engineering, welding rods (electrodes),
electromagnetical couplings, sintered bearings, et.
Generally, the powders, necessary to these
items manufacturing, are gotten from the oxides
resulted from metallurgical and chemical process,
and they are pure products; their purity degree and
shape and grain-size depending on the fabrication
method.
The scraps resulted from the parts and tools
manufacturing as well as scraps of the plastic
deformation are recovered by the alloyed melting
spraying (pulverization) or by the mixing of the
pure metal powders in the given ratio.
First procedure shows the draw hack of the
spherical grain-size powder getting good for the
cases where the required porosity parts are
necessary only and that of the contamination by the
melt contact to the furnace refractories.
The second procedure supposes the pure
metallic powder manufacturing, powders, that, not
in all cases, could be gotten with the necessary
characteristics and either in convenient economical
conditions.
Manufacturing of the alloyed powders, in the
advanced homogeneity and convenient economical
conditions, on the electrochemical way, removes
the above mentioned, drawbacks.
Complying with this technology, the anodic
dissolution of the respective alloy takes place from
semi-finished or worn out parts, in an electrolyte of
sodium or amonium chloride aqueous solution,
having concentration of 20%, voltage 12-24V,
current density 0.01A/dm
2
Formation of the metallic hydroxides takes
place, which are washed, filtered and dewatered in
air or nitrogen.
The gotten metallic oxides are reduced by
hydrogen at temperature between 500-1000C
with or without final crushing.
To get powders from high alloyed steels, the
anode is made of the worn out tools or scraps
resulted from the respective steel-grade tools
fabrication. The cathode is made of copper, nickel
or graphite. The gotten powder is screened, graded
Scrap type Grain size (mm)
01 15 510 115 1520 >20
Scale (%) 8.24 25.06 28.72 32.78 4.80 -
Metallic slag (%) 7.14 9.22 7.15 9.50 31.85 33.30
Metallic melting (%) - 3.6 4.72 11.78 10.62 68.78
Sludge scale (%) 71.4 28.50 - - - -
157
and mixed up with graphite powder additions to
achieve, finally, the necessary carbon content.
The powder is pressed in plates and sintered,
afterwards.
By same technology the non-alloyed metallic
powders could be gotten using pure metallic
anodes and electrolyte consisting of the aqueous
solution of pure amonium chloride, the powder
purity being some to that of the anode (purity).
The cathode could be same metal (alloy) as the
anode: copper or graphite.
Technological steps :
The chemical composition of the gotten powder
is same as that of the anode and the grain size is
polyhedronical shape. Grain-size is direct variable
to the current density.
By these technologies the followings are gotten
1) The alloyed powders, for every part of
the mechanical engineering, made in the
technological flaw: pressing, sintering; as well as
every tool, using the worn-out parts and tools and,
also, by the valorification of the scraps resulted
from the fabrication process of the parts, tools,
rolled, forged, extruded and cast semi-finished
products.
2) The non-alloyed powders of high purity
metals, the powder purity being same as to that of
the anode.
3) The scraps, as stainless steel chips,
could be changed in powders using the
intercrystalline corrosion method. For that it
should be started from the material with higher
carbides precipitated on the crystalline grain-size
borders, in order that the intercrystalline corrosion
process to be developed quickly. This method uses
the austenitic stainless steel chips proceeded from
the mechanical processing.
Several thermal treatments are used to
precipitate the carbides of the grain size borders
and to sensitive the steel at the intercrystalline
corrosion:
- first treatment is quenching in solution
consisting in steel heating at temperatures ranged
in the stability zone of the solid solution, enough
maintained to dissolve, partially or totally, the
secondary phase grain size (carbides) in matrix
(austenite) followed by the high speed cooling, to
prevent the precipitation of the secondary phase
and to maintain the solid solution in solved state,
thus under-cooled and supersaturated.
After treatment a homogeneous austenite was
gotten where the carbides of steel were dissolved.
- the second treatment is named sensitiveness
and it consist of the heating between 650-
680C, with maintain at heat 1 2 hours for a
massive precipitation of the carbides at grain size
borders.
- thus treated the steel is corroded by boiling in
an aqueous solution of 10% H
2
SO
4
+11% CuSO
4
for 48 hours. Steel becomes breakable, so that,
after this operation, by mortaring, stainless steel
fine powder is gotten.
In the engineering industry, in the hot-rolling
departments, the surface laminates is covered with
a layer of dross (oxide) to be removed by different
methods work because the damage and harm the
quality of equipment processed material. Skims
layer consists of three types of oxides: FeO
(wstite), Fe
3
O
4
(magnetite), Fe
2
O
3
(hematite).
In cold rolling departments, the hot rolled band
is cleansed of dross layer by chemical etching with
acid solutions (HCl or H
2
SO
4
).
In the technology of the cold strip rolling, an
important operation is the pickling of the semi-
finish (hot rolled strip) to remove the oxides and to
clean the strip surface.
Electrolyses
Filtering
Metallic hydroxides
Drying
Calcination
Metallic oxides
Processing with H
2
Screening
Grading
Metallic powders
158
A solution of H
2
SO
4
is used, having the
concentration of 15 20% in water at temperature
of 80 - 90C. The following chemical reaction take
place:
Fe
2
O
3
+3H
2
SO
4
=Fe
2
(SO
4
)
3
+3H
2
O
Fe
3
O
4
+4H
2
SO
4
=Fe
2
(SO
4
)
3
+FeSO
4
+4H
2
O
FeO +4H
2
SO
4
=FeSO
4
+H
2
O
Fe +H
2
SO
4
=FeSO
4
+H
2
Also, in the bath, the reduction of the ferric
salts to the ferrous salts takes place in the base of
reaction:
Fe
2
(SO
4
)
3
+H
2
=2FeSO
4
+H
2
SO
4
After pickling, an important quantity of the
crystalline light-green ferrous sulphate FeSO
4

7H
2
O (vitriol iron sulphate) is gotten and, which is
easy oxidated in contact with the oxygen of the air.
Ferrous sulfate is practically a waste resulting
from the etching. It can be sold or used for
industrial water treatment industry dyes,
medicines, agriculture.
The crystallized ferrous sulphate FeSO
4
7H
2
O,
and wastewater of the sulphuric acid pickling
installation could be used, with good results, to get
some aqueous or non-aqueous systems, which
contains the iron oxides or hydrated oxides, in
colloidal state or in suspension.
Finally it seeks magnetite (Fe3O4), a powder
with good magnetic properties and different uses.
A simple method for obtaining Elmore Fe3O4 was
experienced since 1938 and consists of magnetite
precipitation from solutions of salts of bi-and
trivalent iron, by the action of excess sodium
hydroxide.
This method was applied based on the solution
of FeSO4 7H2O Etching bands resulting from the
Cold strip mill at Mittal Steel Galati. It is usually
added and a solution of FeCl3 6H2O.
This method used the chemical reaction:
Fe
(aq)
+2
+2Fe
(aq)
+3
+4HO
-
=Fe
3
O
4
+4H
+
Fe
3
O
4
is gotten in suspension or colloidal
solution in various aqueous or non aqueous
mediums.
In order not to cause a rapid decrease in pH -
his alkaline solution is added in excess (10%
NaOH in water).
After precipitation of Fe
3
O
4
oxidation process
can proceed both in solution and in the suspension
during filtration, washing and milling.
Technology of getting Fe
3
O
4
from solution,
which contains Fe II and Fe III atoms from the
system, the following operations are presented in
figure 1.
Figure 1The main operations of getting Fe
3
O
4
fromsolution
Dissolution is made in distilled water to prevent
further oxidation and eliminate some of the
physically dissolved oxygen.
The mixing done so that the oxygen of
atmosphere is not driven in solution rather than in
small extent.
The mixed oxides of Fe II and Fe III is known
as magnetite (ferroferric oxides) in technic; it is
spinel type (M
II
OH
2
II
O
3
) with magnetical
characteristics. The numerous utilizations it has,
both as the colloidal state (magnetical fluid) and in
the shape of the dried powder in electronic and
electrical engineering, radio-broadcasting, non-
distructive control and electro-domestic
equipments. Filtering is made by the under-
Dissolution
FeCl
3
Mixing Precipitation Filtering
Dissolution
FeSO
4
+7H
2
O
Washing Drying Crushing Grading
Fe
3
O
4
grade
159
pressure filters, maintaining a liquid cushion, as
long as possible, on the precipitation surface. In the
same way running and washing.
The drying is made in vacuum drying stove at
105C to avoid oxygen action on the precipitate.
Depending on how to exploit Fe
3
O
4
obtained
through this process, choose the appropriate size
range. Sorting can be done on an assortment of
vibrant site. If they need a very fine grained (<1
m) provides a wet or dry milling. An optimum
grain can be obtained by passing an appropriate
aqueous phase precipitated in the organic phase.
Keeping quasicoloidale very fine particles in
solutions can be achieved with surface-active
systems. Oleic acid precipitate moisten and mix
one more time, so that each particle to be covered
with a thin film of oleic acid, which will prevent
congestion. For an advanced mixing colloid mill
can be used according to the method Bibik [..]
Finally you can add a quantity of toluene for a
variable density, depending on the purpose. You
get a ferrofluid with magnetic properties with
multiple uses:
- ferrofluid seal;
- ferrofluid bearings;
- the effects of levitation applications;
- ferrofluid systems of writing and posting.
The powder obtained by precipitation method is
black, very fine, the reason is not flowing, form
small clusters and is magnetic. Density is about.
0.33 g/cm3. Particle size, d = max.3, 7 m
Diffractometer analysis showed a high purity of
the powders (Fig. 2)
Figure 2Fe3O4 powder diffractometry
2. Conclusions
To get high purity metallic powders from the
metallic or non-metallic scraps is one of the up to
date method of higher valorification of the scraps,
from the metallurgical plants.
Non-polluted clean-technologies are used,
generally, with low consumption of energy and
materials.
Technologies require simple installation with
high performances
By a good management of the technological
operation the wished purity of powder could be
gotten, into a large range of grain size, grain
shapes and with multiple utilisations.
References
[1] Negulescu, M; .a. Protecia mediului
nconjurtor, Editura Tehnic, Bucureti,
1995
[2] Rojanschi,V; .a. Protecia i ingineria
mediului, Editura Economic , Bucureti 2001
[3]*** Gospodrirea deeurilor-Buletin de
informare INFOTERA, Bucureti,1998.
160
[4] *** Unser Metall Recycling zu 100%,
nr.6/1997, pag.398.
[5] *** Deeurile-restricie sau stimulent al
dezvoltrii- Simpozionul CMRM;
Bucureti,1993
[6] E. Vasilescu; A. Doniga Valorificarea
superioara a deseurilor la pulberi micronice
cu caracteristici fizico-chimice impuse,
Contract de cercetare nr.262/ 2000
[7] M.Vlad, E. Vasilescu, A. Doniga Cercetari
privind valorificarea deseurilor identificate
in actiunea de implementare a sistemului de
management de mediu, Contract cercetare
nr.345/2001
[8] V.Popa, C.Caraci Valorificarea deseurilor
metalice Editura Tehnica Bucuresti1973
[9] V. Constantinescu, R.L Orban s.a. Aspecte
tehnico-economice ale elaborarii pulberilor
prin valorificarea superioara a materialelor
metalice scoase din uz, Conferinta de
metalurgia pulberilor, 1983, vol. 12, p.119.
[10] Gh.Facsko, C.Radovan, C Daminescu
Obtinerea pulberilor de nichel pe cale
electrolitica, Conferinta de metalurgia
pulberilor, 1975, vol. 1, p.19.
[11] F.Oprea , I. Constantin, P. Nicolae, C.Rohr
Procedeu semicontinuu de valorificare
complexa a materialelor refolosibile din
carburi metalice sinterizate, Conferinta de
metalurgia pulberilor, Cluj-Napoca 1983, vol.
2, p.125
161
I MPACT OF STREETS LANDSCAPE ON PEOPLE ORI ENTATI ON
AND ROAD SAFETY I N THE CI TY
A. Polyakov, V. Shvets, O. Veremiy, M.Grabenko
State Technical University Vinnitsa, Ukraine, elena.veremiy@mail.ru
Abstract: in this work is shown a sight of the effect of roads gardening on its aesthetic perception
and on road safety in cities. Planting of trees are described and the principles of their use as well as the
classification of trees by psychological influence on people.
Keywords: gardening, orientation, street, traffic safety.
Introduction
The most progressive type of human
settlements development is the creation of urban
agglomerations, which allows people living in
large areas to get to administrative and industrial
centers of cities and consequently get higher
paying jobs and improve their standard of living.
With the growth of cities, its industrial
development is becoming a more difficult
problem for the environment and the creation of
good conditions for human recreation. Intensive
development of the industrial economy is
accompanied by significant violations of the
surrounding environment.
Gardening of the cities of Ukraine, during the
years of economic crisis, was increasingly
abandoned, there is a lack of recreational, arts and
artistic features, the existing do not carry any
information and meaning.
With the building of urban agglomerations
increases the number of transports means,
especially individual, the number and length of
highway between the cities of
agglomeration. There is therefore a significant
complication in road markings, making it difficult
to orientate drivers on the road and usually leads
to accidental road impacts.
To solve the above problems, we propose to
approach the landscaping of the streets not only
by improving the ecological condition of
decorative and aesthetic appearance, but also by
paying attention to the informative planting of
trees and their psychological impact on drivers.
Main part
Road-maintenance organizations involved in
gardening of roads mostly carry out monotonous
ordinary planting that did not change during an
extended period of time. This is not the best
solution. There is a clear relationship between the
aesthetic qualities of highway and movement
safety (in primitive solution, monotonous roads
planting, traffic accidents are much higher). That
is why it is often said: "The issue is not about how
much would cost the construction a citys
beautiful highway, but about what will cost its
imperfection.
Visually smoother road is essential for safe
and confident driving. This option does not only
requires the usual driver's visual reference points
(such as the edge of the carriageway, brow sub
grade, concrete pavement joint axis), but also
additional information about the direction of the
movement (such as the use of contrasting lines
marking, boundary lines, columns and bar guides
'interiors, vegetable crops).
It should be noted the ability of shrub planting
to soften the consequences of a car coming of the
road for any reason. For example, a shrub of 30-
162
40% provides the deceleration of a car (which
came at a speed of 90 km/h on the road and at an
angle of 30 to the axis of the road) on only 3-4
m.
In lowland areas it is necessary to avoid
monotonous planting. Longitudinal planting on
urban roads was formerly widespread in Western
Europe, directly and not on too long plots. They
are good in terms of traffic safety in fog or snow,
but under sunlight they give some alternation of
shadow bands, that tend to tire the drivers (the
most dangerous frequency of alternating
illumination and shaded areas are 10-15 per
second, which corresponds with a speed of 80-100
km/h and the distance between trees of 2-3 m). In
addition, longitudinal plantations create a corridor
and close views of the surrounding landscape and
building. Application of longitudinal planting in
reconstruction and construction of new urban
roads is appropriate only for high slopes and
embankments along the banks of rivers.
The practice of decorative landscaping
orientation of roads provides a visual planting
guidance, which can be divided into 4 groups:
1) Direct planting (linear reception)
indicates a change in the movement, from far
prompts the driver about the degree of a
turn. They can only be linear, parallel to the axis
of driving, beyond the roadbed. Their length
depends mainly on the radius of the turn, and their
line should visually cover the entire width of
lanes, a look at the turns at their the approaches
(Fig. 1);
Figure 1: Direct planting on a turn shown in a
planar view:
a) Turn with a small angle and a large radius;
b) For small radius turn
2) Barrier planting - prompt the driver about
the inability to continue moving in its direction,
and at the same time is the visual "reflector"
view that forces to move to the right
direction. They are planted on the same principle
as direct planting: they are needed mostly at
intersections, bus stops, traffic intersection, but
can be used at venues and leisure complexes and
maintenance places (Fig. 2);
Figure 2: Examples of barrier planting on
adjacent roads:
a) Opposite the driving direction b) at the end of a
transitional speed zone;
c) At a convex vertical turn
3) Emphatic decorating planting are
intended to prevent driver distraction from the
most important or potentially dangerous parts of
the road (decoratives) or, conversely, to draw its
attention or its focus on points of importance or of
safety, or for architectural reason of the road (e.g.,
separation of pools). An example of new
approaches and planting can be a gate made by a
fractured convex longitudinal profile (Fig. 3);
Figure 3: Paired
planting trees on a
convex fracture of a longitudinal profile:
a) In an enclosed area, b) in open area
4) Safety (ordinary) planting - the protection of
drivers from blinding headlights of oncoming
cars; protection from side wind by ordinary trees,
shrub planting arrangement that delay car that are
leaving the road (Fig. 4);
163
Figure 4: Example of mixed mode of decorative landscaping of the road:
1 snow protection band, 2 regular tree planting, 3 landscape planting
To ensure visibility at intersections and
highways adjacent to a level, planting strips are
arranged according to figure 5. Estimated distance
visibility road surface (La, LB) should be
equivalent to the average speeds on the roads and
taken in table. 1, A width of the strip adjacent to
the road, provides side visibility (LF) should be
25 m (from edge of carriageway for) roads of I-III
rd categories and 15 m for roads of IV and V
categories.
Table 1: Estimated distance for visibility of
roads surface (La, LB), m
D
e
s
i
g
n

s
p
e
e
d
,
k
m

/

h
150
E
s
t
i
m
a
t
e
d
d
i
s
t
a
n
c
e

f
o
r
v
i
s
i
b
i
l
i
t
y
,

m
250
120 175
100 140
80 100
60 75
50 60
40 50
30 40
It is also important when selecting trees to
take into account the nature of their psychological
impact on humans (Table 2).
Figure 5: Scheme of planting strips for visibility
at intersections of roads
Table 2: Classification of trees by psychological impact on humans
Form type sketch Scope of use
1. Annoying Used for drawing the attention of drivers
1.1Widely sprawling Create snow protection strips; design border of country,
region, district; fulfill the conditions of the visual orientation
at crossroads, at plans with turn and areas with limited
visibility
1.2
Narrowly
sprawling
1.3Conical Made for the visitsof city, entrances of memorial space,
creating dominant zones
164
1.4Column form
2. Decelerating Used to create places of rest and quiet and slow driving rate
2.1Oval form plantingfor recreation, special areas, decorative plantings on
Historic streets
2.2Spherical form
2.3Umbrella form Planting for recreation, create formsonwhich are placed
information signs
2.4Tear-drop form
Conclusions
We have shown the clear relationship between
the aesthetic qualities of highway and traffic
safety. To ensure the visual smoothness of the
road it was proposed to use decorative
landscaping orientating roads, namely: direct,
decorative, emphatic and protective planting.
To ensure traffic safety at intersections and
highways adjacent to a level, planting strips are
placed according to the dependence between the
estimated speed and estimated distance for
visibility of road surface.
When choosing trees, it is necessary to consider
the nature of their psychological impact on
people. Thus, we distinguished class of annoying
and decelerating plantings. We should also keep
in mind that the form of an object is perceives
well only under the comparison with other similar
objects, that is why it is necessary to consider the
nature of the combination of two volumes of
crowns of different architectural structure.
References
[1] Hanma]rne ra ]ironenornune ]opmynannx
nicosaxncnnx cmyr anromnxxin i sanisnnnt:
nayx. nicnnx HHTV Vxpann, 2008 p. / acnp.
T.I. Bnxo. Htnin: HHTV Vxpann, 2008
C. 195-199.
[2] Hanma]rnoe npo+xrnponanne [Enexrponn
pecypc]: xnaccn]nxannx perynxpntx ]opm
penecntx pacrenn / uepxacony M.H. 1960
r. Pexnm ocryny o crarri: http://landshaft-
m.at.ua/publ/
klassifikacija_reguljarnykh_form_drevesnykh_
rastenij/1-1-0-12.
[3] Vxasannx no apxnrexrypno-nanma]rnomy
npoexrnponannm anromonntntx opor:
BCH 18-84 [Hauano ecrnnx or
01.01.1986]. M.: Mnnanroopom PCuCP,
1984 r. 74 c.
165
I MPLEMENTI NG A SPC I NTEGRATED SYSTEM TO I MPROVE
MANUFACTURI NG PROCESSES I N AUTOMOTI VE I NDUSTRY
L., Lobon
1
, C. V., Kifor
2
, C., Oprean
3
and O., Suciu
4
1
Lucian Blaga University of Sibiu, lucian.lobont@ulbsibiu.ro
2
Lucian Blaga University of Sibiu, claudiu.kifor@ulbsibiu.ro
3
Lucian Blaga University of Sibiu, rector@ulbsibiu.ro
4
SC Compa SA Sibiu, octavian.suciu@compa.ro
Abstract: Nowadays challenges emphasize greatly on the accomplishment of a high level of quality
in manufacturing processes, especially in automotive industry. Various methods have been used in order
to find ways to organize monitor and control the manufacturing processes. A method used widely to
achieve these requirements is statistical process control (SPC). The objective of this study is to present a
system which integrates SPC with the results from 3D measuring systems in order to calibrate and
monitor manufacturing processes from an automotive industry organization. The implementation of the
integrated system has resulted in improvement of the manufacturing processes and, consequently, to a
better quality of the final products.
Keywords: measuring systems, 3D, SPC
1. Introduction
The challenges in nowadays industry are very
strong, the emphasis stressing the need for
accomplishing a high level of quality during
manufacturing processes. Various methods have
been used, especially to find ways to control the
manufacturing processes. A method used widely is
statistical process control (SPC). Statistical process
control is recognized as a technique to achieve
cost-effective quality control through continuous
manufacturing process improvement [1]. Statistical
Process Control (SPC) is a statistical based
approach able to determine whether a process is
stable or not by discriminating between the
presence of common cause variation and
assignable cause variation. It is a well-established
technique, which has shown to be effective in
manufacturing processes.[2]
It is impractical to inspect quality into a
product; the product must be built right the first
time. The manufacturing process must therefore be
stable or repeatable and capable of operating with
little variability around the target or nominal
dimension. Online statistical process control is a
powerful tool for achieving process stability and
improving capability through the reduction of
variability.[3]
The statistical process control use different
devices and instruments for initial acquisition of
data. The evolution of SPC can be seen by the
large number of specialised hardware and
software. In its earlier time SPC means only data
analysis from manufacturing processes, today SPC
means informatic systems with complex analysis
functions, decisions and management. [4].
A widely used SPC system configuration is
shown in figure 1.
SPC systems have been the central point of
interest in various researches and have known
different approaches according to the specific
objectives of each study. Following we present
some of these approaches.
Systems that use previous collected data to
improve the quality of the manufacturing
systems. In one of the studies adopting this
approach [5] the authors developed a
framework that makes use of self-learning
algorithms that enable the manufacturing
system to learn from previous data and
results in eliminating the errors and
consistently producing quality products.
The framework relies on knowledge
discovery methods such as data mining
encapsulated in a process analyzer to
derive rules for corrective measures to
control the manufacturing process.
166
Analysis of statistical instruments used in
SPC, such as control charts with the
objective of identifying the real time of the
process change. [6]. The study presented a
method based on the concepts of fuzzy
clustering and statistical methods and
developed a novel hybrid approach which
is able to effectively estimate change-
points in processes with either fixed or
variable sample size. The proposed
approach can be employed for processes
with either normal or non-normal
distributions. Also, it can estimate the true
values of both in- and out-of-control
states parameters.
Decision support system used to make
decisions after analyzing statistical data
from SPC [7]. There is presented an
advisory decision support system. This
system helps in collecting statistical data
and thereafter analyzes the enormous
volume of data and aids in making quality
related decisions. Unlike the conventional
SPC applications where the analyzed
results have to be interpreted by quality
control specialists, Manufacturing
execution system (MES) based unmanned
manufacturing environments require
automation of the interpretation process.
The developed advisory system helps in
selecting and designing control charts
based on various cost, rule or heuristics
models. The system also provides
interpretation expertise by configuring and
applying various rule sets. On violation of
these rules, signals are generated by the
system and the expert system advices for
appropriate remedial actions. Thus the
system acts as an advisory support system.
Analysis of production system architecture
which use SPC systems and techniques.
The authors [8] presented a new analytical
method for evaluating the performance of
production systems in which statistical
process control (SPC) techniques are
implemented. Machines behaviour is
monitored by measuring quality
characteristics of the produced parts
through off-line inspection devices and
sampling inspections. The numerical
results show the good accuracy of the
proposed method, provide new insight in
the relations among the two areas and pave
the way to the joint design of production
logistics and quality control systems.
In concordance with the recent researches
regarding SPC used in manufacturing, the
objective of this study is to present a system which
integrates SPC with the results from 3D measuring
systems in order to calibrate and monitor
manufacturing processes from an automotive
industry organization.
Figure 1: Generic SPC system.
2. System development 2.1. Requirements
167
The development of the system started from a
set of specific requirements:
- monitoring workstations in order to rapidly
take action when the process is out of
limits;
- validation of settings following the change
of part, tools or if the process is out of
control limits;
- statistical sampling of products key features
which cannot be directly monitored at the
workstation because of the complexity of
the measuring device;
- setting and monitoring of a large number of
workstations that otherwise would have
needed a numerous and sophisticated
measurement devices;
- reduction of cost related to calibration and
maintenance of a large number of
measurement devices.
2.2. Procedures
The design and development of the system is
based on a series of quality procedures developed
within the organization. Such procedures are: the
procedure for process statistical analysis (fig. 2),
process monitoring, process validation, process
control through SPC, etc.
Figure 2: SPC procedure.
2.3. SPC conceptual framework system
Confirmed
capability
?
Thereareprevious
data?
Control plan
I-QS-.
CPPM
SPC process plan
Quality engineer
SPC Team
SPC characteristics analysis
Previous data
Technological documents
CPPM
SPC Team
Acquiring process information
I-QS-
SPC Team
SPC Team
Establishing sample size and sample collection
frequency
SPC process plan
Control chart
Quality engineer
Control chart
Quality engineer
Operator
Measuring and data collection
Control chart
Quality engineer
Control chart
Quality engineer
SPC Team
Perform calculations and analysis results
Plan process SPC
Quality engineer
Control chart
Quality engineer
SPC Team
Variation causes analysis and their elimination
Action plan
Quality engineer
YES
YES
168
The SPC system framework (fig. 3) was
developed after analyzing the requirements and the
procedures. The system is composed of four main
elements: 1. the workstations where the
information comes from; 2. coordinate 3D
measuring system; 3. the monitoring room where
the process parameters are assessed and visualized;
4. SPC system which integrates the information
and transmits the results to available displays for
validation and monitoring in the workstations area
and the monitoring room.
Figure 3: Conceptual structure of integrated SPC system.
2.4. System realization
The real system has two main components: the
hardware all the physical equipments that form
the SPC system and the software which
comprises the SPC analysis and monitoring
application and ensures the information transfer
within the system. A detailed presentation of these
components follows:
The hardware
Coordinate 3D measuring system (fig. 4). It is used
for complex and precise measurements of the
collected samples from the workstations.
Figure 4: 3D coordinate measuring system.
Computers, displays, transferring data systems etc.
(fig. 5). All these components are used for data
collection, analysis, transfer and visualisation.
They assure the correct flow of information and are
positioned near the workstations, the measuring
systems and the monitoring room.
information flow
information flow
information flow
SPC SYSTEM
hardware+software
3D coordinate
measuring system
workstation 1 workstation 2 workstation 3
display display display
Global view display
Monitoring room
169
Figure 5: Components of the SPC system.
The software
1. The SPC analysis and monitoring application
is developed and implemented within the
organization allowing:
a) for SPC system:
-data collection from measuring devices;
-data processing and calculation;
-generation of specific SPC graphical and
numerical reports (fig. 6);
-process capability study;
-export of reports data;
b) for monitoring (fig. 7):
-support for generating organization layout;
-establishing the connections between
system elements;
-real time graphical representation of
problems encountered in fabrication
process.
Figure 6: SPC graphical report.
170
Figure 7: SPC monitoring layout.
3. Results following the systems
implementation
The system was implemented in an automotive
industry organization and has been functional for
two years. The results obtained in this period of
time are reflected in several areas:
-the reduction of time with 5%-10% for
preliminary and ongoing settings
required for the manufacturing
processes;
-ppm reduction with 5%-20%;
-other costs reduction 1,5%-3% (example:
costs related to calibration and
maintenance of a large number of
measurement devices);
-raising the awareness of the workers about
the functioning of the entire
manufacturing system and their role as
a part of it;
4. Conclusions
This study has presented a system which
integrates SPC with the results from 3D measuring
systems in order to calibrate and monitor
manufacturing processes from an automotive
industry organization. The development of the
system had followed certain steps. First there were
identified the specific requirements in the
organization where the system was created. Then,
the design and development of the system followed
a series of quality procedures developed within the
organization. The next step was to conceptualize a
framework for integrating the main components of
the system. Based on this framework the physical
system was developed combining hardware and
software elements. All the software components
were realized within the organization so as to fit
best for the purpose intended.
Its important to notice that, although the
system was created for an organization from
automotive industry, it can also be adjusted with
minimum changes for manufacturing processes
from other industries.
5. References
1. Zhu, Y. D., Wong, Y. S. and Lee, K. S.
Framework of a computer-aided short-run SPC
planning system, The International J ournal of
Advanced Manufacturing Technology,
Springer, London, 2006.
2. Baldassarre, T., Boffoli, N., Caivano, D. and
Visaggio, G., Managing Software Process
Improvement (SPI) through Statistical Process
171
Control (SPC), Springer, Berlin Heidelberg,
2004.
3. Montgomery, D. C., and Runger, G. C., Applied
Statistics and Probability for Engineers - Third
Edition, J ohn Wiley & Sons, Inc., Arizona
State, 2003.
4. Oprean, C., Lobon, L. and Kifor, C. V.,
Integrated statistical process control system for
manufacturing processes, Academic J ournal of
Manufacturing Engineering, Vol. 5, No. 2,
2007.
5. Kumara, S., Nassehia, A., Newmana, S. T.,
Process control in CNC manufacturing for
discrete components: A STEP-NC compliant
framework, Robotics and Computer-Integrated
Manufacturing23, pp. 667676, 2007.
6. Alaeddini, A., Ghazanfari, M., Nayeri, M. A., A
hybrid fuzzy-statistical clustering approach for
estimating the time of changes in fixed and
variable sampling control charts, Information
Sciences 179, pp. 17691784, 2009.
7. Chakraborty, S., and Tah, D., Real time
statistical process advisor for effective quality
control, Decision Support Systems 42, pp. 700
711, 2006.
8. Colledani, M., and Tolio, T., Performance
evaluation of production systems monitored by
statistical process control and off-line
inspections, Int. J .Production Economics 120,
pp. 348367, 2009.
172
173
A NEWAPPROACH OF THE
MAI N LANDI NG GEAR EQUATI ONS
Daniel BOSNI CEANU
1
1
Military Technical Academy, Bucharest, Romania bosniceanud@yahoo.com
Abstract: In this paper an approach for modeling landing gear systems is presented. Specifically, a
nonlinear model of an main landing gear is developed. This model includes nonlinear effects such as a
polytropic gas law, velocity squared damping, a geometry governed model for the discharge coefficients,
stick-slip friction effects and a nonlinear tire spring and damping model. An initial model was developed that
only included the air-spring above the fluid, fluid dynamics through a fixed orifice, and a linear tire spring
term.
Key words: main landing gear, hydraulic spring, shock damper, linear model of equations
1. I nitial landing gear investigation
This chapter is intended to familiarize the reader
with landing gear terminology and to demonstrate
a mathematical development of the equations of
motion for a telescoping landing gear. Figure 2-1 is
intended to acquaint the reader with basic landing
gear components. It shows the simplified
components of a telescoping, main landing gear
(as opposed to a nose gear).
Point 1 on the figure is a rigid body
representation of the aircraft fuselage. Point 2 is a
chamber containing compressed nitrogen which
serves as a spring that carries the weight of the
plane in ground operations. Point 3 refers to the
main, upper cylinder which houses the
compressed gas, hydraulic fluid, and within which
the piston slides. Point 5 is the orifice plate. It is
essentially a circular plate with a hole in the center
through which the hydraulic fluid flows when the
strut is stroking. It, along with the metering pin,
point 6, controls the damping characteristics of
the gear. Point 7 locates one of many rebound or
snubber orifices. These holes lead into a small
volume on the backside of the piston head (point
8) called the rebound or snubber chamber. The
purpose of the snubber is to provide damping
when the strut extends. The Point 9 is the piston.
Figure 1. Schematic of typical telescoping main
landing gear studied
It houses the metering pin and is also the rigid
connection of the wheel axle. Finally, point 10 is
the tire. This element of the gear adds both spring
and damping characteristics to the overall
performance of the gear, and is selected carefully
for various applications.
174
2. Nonlinear model development
This research discusses an independent
development of a mathematical model of a main
landing gear with all the relevant physical
parameters included. The nonlinear equations of
motion are developed for a telescoping main gear.
An initial model was developed that only
included the air-spring above the fluid, fluid
dynamics through a fixed orifice, and a linear tire
spring term. This simple model allowed some
trend comparison between the results of this model
and the early results of the linearized gear. A
metering pin was then added to change the main
orifice effective diameter as a function of stroke.
Another variation was the addition of a snubber,
or rebound chamber. This feature provides
damping while the gear is extending. The model
includes constant seal friction as well as a variable
friction that is a function of stroke. In a further
effort to be realistic, a nonlinear tire model was
added. This tire model has a spring rate that is a
function of tire deflection and damping
proportional to compression rate. In the equations
developed below, the spring and damping
coefficient are used as if they were constant. The
nonlinear characteristics of each of these terms is
included in the equations of motion that are
actually integrated.
Figure 2 is a schematic of the gear used in the
development of the equations of motion. This
schematic is representative of a general
telescoping-type main landing gear. It includes the
aerodynamic lift on the plane, Lift, the upper mass
(of the plane's fuselage) and the mass of the main
cylinder lumped together as a rigid mass, M,, and
the mass of the piston and the mass of the tire, also
lumped together as M
L
. The inertial coordinate of
the upper mass is X
wg
. The zero value for X
wg
is when the
gear is fully extended with the tire just touching the
ground. Fromthis same gear configuration, X
a
, the
coordinate of the lower mass, is taken as zero at the
axle of the tire. Therefore, when the gear is in some
compressed state, X
a
measures the deflection of the tire
when theground input, U(t), is zero.
In the compressed nitrogen chamber (upper
cylinder) with cross sectional area of A
u
thepressureis
P
u
. Likewise, in the lower chamber with cross sectional
area of A
L
, there is a pressure of P
L
. In the snubber
chamber, with annulus area of A
R
, the pressure is
defined to be P
s
. The orifice plate has a hole of
diameter D
op
through which the metering pin, with
variablediameter D
pin
moves. Fluid reaches the snubber
chamber through the orifices d
s
c
and d
s
E
, where the
superscripts represent either the compression mode or
extension mode respectively. The diameter of the piston,
D
pi
, is used to calculate A
R
. Simply subtract the area of
the piston shaft fromthat of the lower cylinder to get
A
R
. The tire is also shown in Figure 2 with a distinction
of pointing out that the tire spring and damping
coefficients, K
t
and C
t
are nonlinear and contribute to
the calculation of the tire force F
t
.
Figure 3 shows the forces acting on the upper
mass. Balancing the forces on the upper mass
gives the following equation:
f A P A A P A P L g M X M
R S O L L U U U
wg
U
+ = ) (
..
(01)
The term on the left hand side of Eq. (01) is the
inertial motion term, g is the gravitational
acceleration, f is the friction present in the gear,
and all other terms are as described previously.
This equation assumes that the fluid pressure in the
upper cylinder is identical to the pneumatic
pressure.
Figure 2 - Schematic of telescopinc landing gear
175
Figure 3 - Schematic of upper mass and main cylinder
In this area, reflects the fact that the metering
pin is included, i.e. it is a variable cross-sectional
area depending on stroke.
Figure 4 - Schematic of lower mass
Figure 4 shows the forces acting on the piston.
Summing the forces on the lower mass (piston)
the force balance equation is:
f F A A P A A P g M X M
t S R S S L L L
a
L
+ = ) ( ) (
..
(02)
Where the left hand side of Eq. (02) is the
inertial motion of the lower mass and A
s
is the area
of the snubber orifice. F
t
is the force that is
transmitted through the tire from the ground and
has the form:
) ( ) (
. .
U X C U X K F U
t a t t
+ + + = (03)
where the tire force is a function of a nonlinear tire
stiffness and a damping force that is composed of a
damping coefficient that is proportional to the tire
stiffness and the time rate of change of the tire
deflection.
3. Relation of pressures to stroke position and
stroke rate
The pressure terms in Eqs. (01) and (02)
are as yet unknown and need to be related to the
positional variables X
wg
and X
a
or their derivatives.
The pressure of the compressed nitrogen in the
upper cylinder can be described by the polytropic
gas law for a closed system as:
(04)
where X
s
is the strokeavailable, given by:
a Wg S
X X X = (05)
with X
SI
as some initial length, P
SI
, the
charge pressure at X
max S
, and , the polytropic
gas constant. X
S
max
is the maximum value to
which the gear can be extended. This form of
representation of the pressure change is assumed
to happen as a quasi-equilibrium process. The
significance of the polytropic gas constant is that it
describes the type of process that occurs. An
average value is usually sufficient in application
Equation (04) was defined in such a manner
that P
u
will become very large when X
s
is near
X
S
max
, i.e. the gear is nearly completely
collapsed. This is a suitable representation of the
process, with only the polytropic gas constant y
as an unknown.
The pressures (P
L
and P
s
) of the fluid in the
lower cylinder and in the snubber are related to the
flow rates of the fluid into and out of those
regions. The volumetric flow rates through the
orifice plate hole, Q
c
, and the snubber orifices, Q
s
,
can be determined by combining the continuity
) (
max S S
SI
SI U
X X
X
P P
=
176
equation and Bernoulli's equation for fluids. Flow
is always from the higher pressure to the lower
pressure. Bernoulli's equation for an
incompressible fluid states that along a
streamline,
(06)
where P is the pressure at some point, g is the
gravitational acceleration, V is the velocity of the
flow, vis the specific weight of the fluid which is
equal to the fluid density (p) multiplied by the
gravitational acceleration (g), and Z is the height
difference from some zero reference. This
equation assumes that the viscous effects within
the fluid are negligible, the flow to be steady and
incompressible, and that the equation is applicable
along a streamline.Equating Bernoulli's equation
(Eq. (06)) at two points in the flow along the same
streamline yields:
(07)
In the case of a landing gear, the potential
distance between Z
1
and Z
2
can be neglected as
the distances involved are very small compared to
the other terms. Equation (07) with the continuity
equation for incompressible fluids which states Q =
A
1
V
1
= A
2
V
2
allows for the solution of this
equation in terms of one of the velocities.
Assuming that P
1
>P
2
, i.e. the flow is from P
1
to
P
2
, then solve for V
1
from the continuity equation
as:
(08)
When the flow reverses, i.e. P
1
< P
2
, then the
velocity at point 2 is described by the above
equation with the pressure terms switched and a
negative sign on the square root. The ideal
volumetric flowrate (Q
ideal
) for an incompressible
fluid can be expressed as Q
ideal
=A*V.
Now we have:
V AC CdQ Q
d ideal real
= = (09)
Substituting Eq. (08) into Eq. (09) for velocity:
2 1 2 1
4
2
1
) ) ( 1 (
2
P P P P
D
D
AC Q
d real

(10)
For our landing gear, there are two flows that
are of concern, the flow through the orifice plate
and the flow into and out of the snubber chamber.
Define Qs
C
as the flow rate into the snubber
chamber in the compression mode, where the
snubber orifice area (A
s
) becomes A
s
c
, which
allows larger flow. The flow rate through the
snubber orifice during the extension mode is
defined as Q
S
E
, and the area A
s
becomes A
S
E
, which
only allows small, restricted flow. In both cases,
the flow through the main orifice plate is Q
O
.
Figure 5 - Control volume between piston and orifice
plate
Figure 5 shows the direction of fluid flow into and
out of a control volume in the lower chamber as a
function of stroke mode (extension or
compression). In relating the flow rates to the
pressures, defining a control volume as shown by
the dashed line in Figure 5 is necessary. The
stroke rate is defined as
a Wg S X X X
. . .
= (11)
where the compression mode is given by X
s
>
0.0, and the extension mode byX
s
<0.0. The flow
is assumed to be negative leaving the control
volume, and is positive entering it. For an
incompressible fluid, the volumetric flow rates for
compression and extension can be written as:
) 1 (
) ( 2
4
1
4
2
2 1
2 2 2
1
2
2
1
D
D
P P
V V
D
D
V
= =
. )
2
1
(
2
const Z V
g
P
= + +
u
2
2
2
2
1
2
1
1
)
2
1
( )
2
1
( Z V
g
P
Z V
g
P
+ + = + +
u u
177
0 . 0
.
= + + S
L
C
S a
X A Q Q (12)
during the compression mode and
0 . 0
.
= + + S
L
E
S a
X A Q Q (13)
during the extension mode. Equation (10) defined
the general form of the equation for a flow rateand
can be written as:
U L U L
L
o
d o o
P P P P
D
d
C A Q
=
) ) ( 1 (
2
4
(14)
where d
0
is the effective diameter of the main
orifice, D
L
is the diameter of the lower chamber,
and C
d
is the discharge coefficient of the main
orifice. The flow through the snubber orifices
during this mode is described by:
with d
s
c
as the diameter of a snubber orifice, D
L
as
described above, C
dS
C
is the discharge coefficient
of the snubber orifice and A
s
c
is the effective area
of the snubber orifice. Similarly,
for the extension mode , where flow is into the
control volume (P <
L
P
U
and P
S
) .
L U L U
L
o
d o o
P P P P
D
d
C A Q
=
) ) ( 1 (
2
4
(16)
where the difference between this equation and Eq.
(14) is that the pressure terms have exchanged
positions and the whole term is now positive. The
flow rate through the snubber orifices during the
extension mode is given by
where D
R
is the effective diameter of the annulus
snubber chamber, d
s
E
is the diameter of a snubber
orifice, A
S
E
is the effective area of the snubber
orifices and C
dS
C
is the discharge coefficient of
the snubber orifices in the extension mode. To
simplify Eqs. (14), (15), (16), and (17), let the
non-pressure terms be redefined as:
) ) ( 1 (
2
4
1
L
o
d O
D
d
C A E
,
1 3
E E =
) ) ( 1 (
2
4
1
L
o
d O
D
d
C A E
,
) ) ( 1 (
2
4
4
R
E
E
dS
E
S
D
d
C A E
E
, respectively.
Substituting Eqs. (14) and (15) into Eq. (12) and
Eqs. (14) and (15) into Eq. (13) using this new
notation, rewrite Eqs. (12) and (13) as
0 . 0
.
2 1
= + S
L S L U L
X A P P E P P E for
S X
.
>0 (12.a)
0 . 0
.
4 3
= + S
L L S L u
X A P P E P P E for
S X
.
<0 (13.a)
Figure 6 - Control volume for the snubber chamber
) 15 (
) ) ( 1 (
2
4
S L S L
L
C
s
C
dS
C
S
C
S
P P P P
D
d
C A Q
) 17 (
) ) ( 1 (
2
4
L S L S
R
E
s
C
dS
C
S
E
S
P P P P
D
d
C A Q

178
Additional information about the flow rate-
pressure relationship can be gained by studying a
control volume in the snubber chamber as shown
by the dashed line in Fig.6 .
The variables A
R
and D
R
in Fig.6 are the
rebound chamber annulus area and effective
diameter respectively. P
s
is the pressure in the
rebound chamber and d
s
c
and d
s
E
are the diameters
of the snubber orifices in the compression mode
and extension mode respectively. In the case of
compression, whereX
s
>0.0 and P
L
>P
s
,
0 . 0
.
= + S
R
C
S
X A Q (18)
Substituting the flow rate
C
S
Q of Eq. (15) into Eq.
(18) yields: (19)
From previous notation of E
i
this expression
becomes: 0 . 0
.
2
= + S
R S L
X A P P E (20)
Rearrange Eq. (20) to get an expression for the
pressures in terms of the stroke rate as:
S
R
S L
X
E
A
P P
.
2
= (21)
2
.
2
1
) ( S
R L
U L
X
E
A A
P P

+ = (22)
where P
u
is given in Eq. (04). Square both sides of
Eq. (21) and solve for P
s
as:
P
S
=P
L
- (
2
E
A
R
)
2
2
.
S
X (23)
Similarly, for the extension case withX
s
<0.0:
2
.
2
3
) ( S
R L
U L
X
E
A A
P P

= (24)
P
S
=P
L
- (
4
E
A
R
)
2
2
.
S
X (25)
These known pressures [Eqs. (04), (22), (23),
(24), (25)] can now be substituted into Eqs. (23)
and (24). Algebraic simplification of these
equations leads to the compression and extension
cases in terms of readily measurable quantities
as:(01a)(02a)(01b)
(02b)
Introduce a new notation using subscripts to
simplify the above equations: "1" and "2" will be
associated with compression (equation set (a)),
and "3" and "4" with extension (set (b)). With this
change, the equations can be written
in the form:
(01c)
(02c)
where the coefficients of the stroke rate squared
term are assigned the C
/
i
s, and the coefficients of
the stroke position term are the K
i
's.
f X A A
E
A A
A
E
A
E
A A
X
X
P A A L g M X M
S
L
R L
R
R R L
S
SI
SI L R U wg U
+
)
`
+ + =
2
. .
0
2
1
2
2
2
1
..
) ( ) ( ] ) ( ) [(
) ( ) (

f F X
A A
E
A A
A A
E
A A
E
A
X
X
P A A g M X M
t
S
C
S L
R L C
S R
R L R
S
SI
SI R L L a L

)
`
+ =
2
.
2
1
2
1
2
2
..
) ( ) ( ) ]( ) ( ) [(
) ( ) (

f X A
E
A
E
A A
A A
E
A A
X
X
P A A L g M X M
S
R
R R L
L
R L
S
SI
SI L R U wg U
)
`
+ + =
2
.
2
4
2
3
.
0
2
3
..
] ) ( ) [( ) ( ) (
) ( ) (

f X K X C L g M X M
S 3 / 1
2
S
.
3 / 1 U
wg
..
U
+ + =

f F X K X C g M X M
t S
S
L
a
L
+ + =

4 / 2
2
.
4 / 2
..
0 . 0
) ) ( 1 (
2
.
4
= +
S
R S L
L
C
s
C
dS
C
S
X A P P
D
d
C A
f F X
A A
E
A A
A A
E
A
E
A A
X
X
P A A g M X M
t
S
E
S L
R L E
S R
R R L
S
SI
SI R L L a L
+
)
`
+ =
2
.
2
3
2
4
2
3
..
) ( ) ( ) ]( ) ( ) [(
) ( ) (

179
The only unknown term left in these equations
is friction. As mentioned previously, friction in
this gear comes mainly from two sources, friction
due to tightness of the seal and friction due to the
offset wheel (moment). The seal friction is
assumed to be a maximum value statically and
some function of velocity in the dynamic state.
The functional relationship between frictional force
level and velocity could be determined through
testing. The friction due to the offset wheel is the
result of the moment produced by the nonaxially
loaded piston within the cylinder.
It can be seen from Fig.7 that the force between
the piston head and the cylinder, N, is a result of
the tire force, F
t
, applied at moment arm, ma, from
the centerline of the piston. The frictional force
due to the offset wheel (F
ow
is assumed to be of
the form (refer to Fig.7):
Figure 7 - Schematic of gear for friction model
development
N F
OW
= (26)
Where N is the normal force of the cylinder
wall resisting the side of the piston head, and is
the coefficient of friction between the two parts.
To find the unknown force N, sum the moments
about point O to zero to get:
0 ) ( :
0
= +
stp X N ma F M
S t
(27)
Where stp is the minimum distance between
the piston head and the lower seal when the gear is
fully extended. Rearrange Eq. (27) by isolating N,
and then substitute N into Eq. (26) to get an
explicit form of F
ow
:
(28)
(29)
The total friction in the landing gear, f, in
equations (01c) and (02c) is now assumed to be:
This paper assumes that a proportionate part of
the fuselage (half of the 80% of the total weight
that rests upon the main gear) is treated as a lump
mass centered at the centerline of the main upper
cylinder. Also, this model takes into account only
vertical loads on the strut. The tire is modeled as
a nonlinear spring and damper. This tire model
does not take into account spinning stiffness
(because the test tire does not spin) or spin-up drag.
The fluid is assumed to be incompressible and all
structural members are assumed to be rigid, with
each having only a vertical degree of freedom.
These assumptions are good only for straight-line
taxiing over runway profiles and landing impact
(spin-up drag on the tire does not significantly
effect the vertical loads on the strut). Any braking
or turning maneuvers are not covered in the
development. The equations developed here are
the basis for a "rollout" simulation.
4. Conclusions
In this, the nonlinear equations of motion were
developed for a general, telescoping main landing
gear.
These equations contain a pneumatic spring
that is determined based on the polytropic gas
compression law, a hydraulic damping that is
proportional to the stroke rate squared,
gravitational forces, lift, inputs from a runway, and
finally friction, which is composed of both a
constant seal friction and a variable bearing
friction. These equations explicitly contain the
empirical parameters of polytropic gas constant,
discharge coefficients for both the main orifice and
the snubber orifices, and the friction levels in the
gear. These parameters are the only variables that
appear in equations (01) and (02) that cannot be
directly measured.
Equations (01) and (02) are highly nonlinear
and are discontinuous due to the differing values
of friction and discharge coefficient as a function
of extension and compression. Future work will
discuss more about the nature of these equations
and present a method of solving these equations
for gear displacements and velocities.
stp X X
F ma
N
a wg
t
+
-
=
) (
stp X X
F ma
F
a wg
t
t
+
-
=
180
References
[1]Greenbank,S.J ., Landing Gear-The Aircraft
Requirement, Proceedings of Institute of
Mechanical Engineers, Vol. 205, 1991;
[2] Norman, S. - Aircraft Landing Gear Design:
Principles and Practices, Lockheed
Aeronautical Systems Company AAIA
Education series;
[3] O'Massey, R. C., Introduction to Landing
Gear Design, ASM Paper No. W70-18.1,
March 1970;
[4]Rashant, K. Simulation of Landing Gear
Dynamics and Brake-Gear Interaction,
Germany, 2008
[5] Sonny,T.,Mason,H., Landing Gear Integration
in Aircraft Conceptual Design,
Multidisciplinary Analysis and Design Center
for Advanced Vehicles, Virginia Polytechnic
Institute and State University Blacksburg,
Virginia,September 1996;
181
RESEARCH ON HOLLOW CATHODE EFFECT AND EDGE EFFECT
AVOI DANCE I N PLASMA NI TRI DI NG TREATMENT
AXI NTE Mihai
1
, NEJ NERU Carmen
1
, PERJ U Manuela Cristina
1
, CIMPOEU Nicanor
1
,
HOPULELE Ion
1
1
Technical University Gheorghe Asachi of Iai-Romnia, E-mail address:
mihai.axinte@gmail.com
Abstract: This paper contains a description of the nitriding installation with unpolarized grid in order to
avoid hollow cathode effect and edge effect. For the experimental study we have made wedge samples cut, and
triangular prism-shaped samples. We also made a theoretical study on these unwanted phenomena.
Keywords: hollow cathode effect, edge effect, grid
1. Introduction
Plasma technologies have been long
studied in surface engineering because they
provide are important regarding technical and
environmental benets over salt bath or gas
treatments. Among them, the active screen
technology for plasma surface engineering offers
multiple advantages over conventional direct
current plasma treatments, like: improved surface
quality and more uniform material properties [1].
In general, for the plasma nitriding
process, the components to be treated are subjected
to a high cathode potential and the grounded wall
while the furnace wall forms the anode. The parts
are directly involved in the discharge process [2].
Conventional direct discharge plasma treatments
apply considerable cathodic potentials to the
components in order to reach the treatment
temperature by means of bombardment with
energetic ions [1].The electric voltage applied
between anode and cathode during plasma
nitriding is 400700 V [2]. This makes the
discharge processes sensitive to non-conductive
substances, and all parts must be thoroughly
cleaned to avoid severe arcing during the
treatment.
The positive ions generated by glow
discharge are accelerated in the cathode fall region
near the cathode surface and bombard the surface
of the specimen. The ion bombardment causes
sputtering, transfers kinetic energy to the
component and raises its temperature. Many
studies have indicated that the geometry, size and
ratio of the surface to mass of the component have
substantial effects on the temperature distribution
within the component and lead to inhomogeneous
response to nitriding.[2]
Plasma characteristic, correlated with the
specific geometry and shape are responsible for a
number of problems:
1) Inhomogeneous batch heats not
uniformly. Problems arise when the treated batch
has a significantly varying surface mass ratio. If
this ratio differs much from the same characteristic
of a standard specimen with thermocouple, some
of the parts in the batch can be overheated or
unheated [3]
2) The hollow cathode effect is a special
situation for the glow discharge between two
closely separated cathode surfaces. This effect
occurs when the dimension of the cathode fall
region becomes as large as the separation distance.
The loss of electrons is low due to the special
geometry and because they are repelled by the
negative walls of the cathode, and in fact, they
oscillate between the sample and the wall (figure
1). The plasma density (that is the electron
concentration) increases and reaches values as high
as 1012 cm
-3
[29]. As a consequence, the
production of ions rises too, and the ion flux
density on the substrate surface increases. Due to
these factors, such surfaces can be heated to
extremely high temperatures even at relatively low
bulk substrate temperatures of 200500 8C and the
local heating can produce melting surface, which
means destroying the part [4]. The effect manifests
itself at a specific composition of the gas, which is
182
characterized by pressure p[Torr], and specific
distance D [mm] between the face to face surfaces.
Figure. 1Highlighting the hollow cathode effect. Cross
section through a narrow channel V shape
3. Edge effect is the growth rate of the
layer different on the edges and flat regions of the
surface of a part. The elementary volume on the
edges receives more energy than the elementary
volume on the flat surface in the same time period
[3] (Fig. 2). Due to distortions of the electric field
around the corners and edges, the shape of plasma
sheath, which is connected to the shape of samples,
determines the ion flux distribution, which, in turn,
affects the uniformity, hardness and surface phases
of coating, erosion rings occur, characterized by
the reduction of hardness [5].
Figure 2. Quantitative energy difference between flat,
edge and corner areas, at edge effect.
To avoid this disadvantage an
experimental installation was build. This
installation is equipped with an polarized screen. In
such a case a glow discharge is largely
transferred from the surface of the parts to the
screen surface, and the ion bombardment of the
parts becomes less intense [3]. This grid is made of
stainless steel, and has the role to modify the
electrical discharge field between anode and
cathode, on the ionic triode principle. According to
[6], the use of active screen virtually does not
reduce the growth rate of the layer. Such a system
is a triode.
The triode ionic screen interposed between
cathode and anode is polarized, so it can change
the electrical field configuration. Modifying the
electrical field between anode and cathode,
changes the discharge conditions and also the
hollow cathode effect occurrence, by adjusting the
grid electric potential [3].
Constructive principle of the installation is
presented in figure 3.
Figure 3. Constructive principle for installation,
Electrical block diagram
In the figure 4 is presented the detailed
electrical diagram. In ionic triode case the
interposed grid between anode and cathode is
electrically polarized, so it changes the
configuration of the anode and cathode electric
field.
Changing the electric field between anode
and cathode changes the degeneration conditions
183
for arc discharge, and double-cathode formation by
adjusting the electrical potential of the grid.
Figure 4 Detailed electrical diagram
The shape for the grid, or the active screen
varies depending the dimensions and the part
configuration. It can have the following shapes:
planar grid, circular grid, cage grid. Materials for
the screen: stainless steel, copper, steel sheet with
different diameter holes in it [7].
The installation function with active screen
represents the grid activation in a certain stage for
plasma nitriding, treatment technology
After cleaning the vacuum chamber and
creating the 10
-1
vacuum value, the anodic current
is connected to the installation. Then the discharge
current is adjusted and proceeds to reducing
vacuum level at the specific value by introducing
gas. Anode current is adjusted to the desired value.
Due to the discharge changing conditions the grid
polarization is modified, conform to the necessary
circumstances.
When the temperature is near the
prescribed one automatically the anodic discharge
electrical current is reduced.
a) b)
Figure 5. Samples different shapes: a) holes different
diameter, b) prismatic shape with sharp edges.
In figure 6 is presented the plasma
nitriding process and the of hollow cathode effect
occurrence due to the configuration of the sample.
The sample has a V shape. This shape allows us to
highlight the hollow effect appearance conditions.
Figure 6. V shape cut sample dimensions
The sample has a cylindrical shape. The
height is 50mm, exterior diameter is 10 mm,
interior hole diameter is 8 mm, and the V shape cut
has the height 25 mm, figure 6. As it is graphically
184
explained in fig. 1 at the cathode fall intersection
the electrons are caught between the two surfaces
and the ionisation increases very much, also the
temperature. In figure 7 is presented the hollow
cathode effect during functioning.
Figure 7 Hollow cathode effect on a V shape sample
during plasma nitriding process, with no active screen
between anode and cathode.
2. Conclusions
1. The installation will be used to avoid classical
plasma nitriding defects by triode effect from the
polarized grid.
2. The installation setings permits us to highlight
the cathode effect apearance conditions.
3. Following researches on avoiding the
disadvantages attached to general plasma nitriding:
edge effect, hollow cathode effect, arc discharge,
not uniformly deposition layers.
4. The ionic triode reduce the sputtering effect ,
positively influencing the surface quality,
decreasing the roghness value.
References:
[1]S. Corujeira Gallo, H. Dong, Study of active
screen plasma processing conditions for
carburising and nitriding austenitic stainless steel,
Surface & Coatings Technology 203 (2009) 3669
3675
[2]Y. Li et al., Plasma nitriding of 42CrMo low
alloy steels at anodic or cathodic potentials,
Surface & Coatings Technology 204 (2010) 2337
2342
[3] S. J anosi et al., Controlled hollow cathode
effect: new possibilities for heating low-pressure
furnaces, Metal Science and Heat Treatment, Vol.
46, Nos.78 (2004) 310-316
[4] C. Alves J r. et al., Nitriding of titanium
disks and industrial dental implants using hollow
cathode discharge, Surface & Coatings Technology
194 (2005) 196202
[5] C. Alves J r. et al., Use of cathodic cage in
plasma nitriding, Surface & Coatings Technology
201 (2006) 24502454
[6]. C. X. Li et al , Active screen plasma
nitriding of austenitic stainless steel, Proc. 4th
European Stainless Steel Science and Market
Congress, Paris, France, Vol. 2 (2002), pp. 297
303.
[7]. M. Axinte et. al., Facility for study heating
and diffusion process, using a ionic triode in a
plasma nitriding installation, Tehnomus XV,
Suceava (2009), ISSN 1224-029X
[8] G. Vermesan, V. Deac, Bazele tehnologice
ale nitrurrii ionice, Editura Universitii din Sibiu,
Sibiu, 1992
185
AN ALGORI THM FOR SOL VI NG QUALI TY PROBLEMS,
USI NG QUALI TY TOOL S
Marius BU
1
, Mircea CI OBANU
2
1
Stefan cel Mare University of Suceava, Romania, marius.baesu@fim.usv.ro
2
Stefan cel Mare University of Suceava, Romania, mircea.ciobanu@fim.usv.ro
Abstract: In this work is has been developed an algorithm, to address shortcomings in
communicating with customers and lack of coordination. As a support for this algorithm, there was
created an Excel program which combines two of the tools proposed to solve the problem, respectively
Ishikawa diagram and Pareto chart.
Keywords: Quality tools, Ishikawa, Pareto
1. I ntroduction.
The essence of solving quality problems is
establishing, elaborating, identifying,
implementing and monitoring the corrective /
preventive actions for quality improvement. In
the current standards (e.g. EN ISO 9004: 2010
and other specialty papers [1], [11], [13], [14],
[15]) there are presented various models and
methodologies to solve quality problems, to
improve quality, respectively to establish,
develop, identify, implement and monitor
corrective / preventive actions. They vary in
number, form, sequence, level of detail and
content of stages, phases, tasks, steps, specific
sequences, and by grouping them in stages. All
of them are using a mixture of statistic and
non-statistical quality tools, known also as The
Seven Basic Tools of Quality and respectively
as The Seven Management and Planning
Tools.
In table 1 there is a briefing of these tools.
There are presented the objectives of each one
of these tolls, and some hints when to use
them.
Table 1: The quality tools
2. The methodology of elaborating algorithms
in using the tools to solve quality matters.
Based on the theme or problem to be solved (in
planning, assurance, and quality control), the
development of algorithms (strategies) for the use
of methodological instruments in solving the
quality problems (planning, assurance, control and
quality improvement) suppose the browsing of a
proposed synthetic methodology, depicted in
Figure 1.
186
Figure 1: The methodology of elaborating algorithms in
using the tools to solve quality matters.
The methodologies specific to the theme or
problem to be solved, could be graphically
represented in the following forms:
-ordered list of stages, phases, tasks,
sequences and methodological tools;
-organization chart, flow chart, block
diagram with rectangles and words;
diagram, with words and symbols or drawings
of methodological tools;
diagram, with symbols or drawings of
methodological tools;
-combinations of the previous forms.
To indicate logical relationships between
the stages, phases, activities, sequencing
methodology, respectively the methodological
tools used to address each stage, arrows should
be used.
The organizational charts, logic diagrams
and block diagrams, as those presented below
(figure 2), could be considered as a reference,
as an example of a specific methodology to
solve a quality problem [4], [17].
Figure 2: Example of a specific methodology to
solve quality problems
3. Case study
For the case study considered in this paper, data
regarding customers complaints about a service
department were collected. The data were collected
during September 2008 - March 2009.
The proposed methodology for identifying,
detailing and improving the causes of those
complaints is as follows:
- collecting data (tables);
- sorting data in descending order, according to
Pareto chart;
- identifying the causes that led to an increased
number of complaints, with negative effects on the
company image (using Ishikawa diagram)
- implementing the corrective and preventive
actions, resulting from brainstorming activities;
- collecting a new set of data, in the same
period of time as that when the initial data were
collected, the difference between being a calendar
year;
- building a new Pareto chart.
- comparing data and drawing the conclusions.
The algorithm is shown in Figure 3.
187
Figure 3: The algorithmproposed
As said before, for the case study considered in
this paper, data regarding customers complaints
about a service department were collected. The
data were collected during September 2008 -
March 2009.
It is worth mentioning that during the data
collection period, the register of complaints was
kept by "the workers on the shelf" and the vendors.
As noted, most complaints have occurred
during December-J anuary, when the workload for
commercial workers is significantly higher
(inventories, closing the year, etc..,)
The main problems, as is about to see, are
caused by the fact that there was not, at that time, a
central dispatch for service problems, lack of
competent people to solve problems or even to
provide some relevant solutions. By competent
people means specialists in this field, because, as
already indicated, both the role of "dispatcher" and
the "primary dealer" was performed by commercial
workers, most of them mainly with basic
knowledge of sales and technical matters. In
addition, some of the claims submitted by service
aimed at a new batch of heating products, just put
on the market (it is late summer, early autumn
2008), which, as noted later, had not undergone
enough testing.
3.1 Data collection
Data were taken from the book of suggestions
and complaints and tabled.
Table 2: Data collected between September 2008-
March 2009
According to the methodology, the next step
being the establishment of a Pareto chart, in order
to identify the most common problems, the data
are arranged in descending order.
In Table 3 these data are arranged downward.
Table 3: Arranged data
According to the typology of drawing a Pareto
chart, the cumulative frequency of the recorded
data must be calculated.
On the Internet there are a lot of
Excel/Spreadsheet applications, freeware or
shareware, that help in preparation of such
diagrams. The only problem is that there is not an
application that includes the same excel
spreadsheet and a chart of data descending
arranged, and the cumulative frequency table and
the resulting Pareto diagram.
Figures 4 and Table 4 summarize the data
collected and the cumulative frequencies and the
Pareto diagram that shows further processing.
188
Table 4: The data frequencies calculation
Figure 4: The Pareto Chart
Figure 5 shows the excel application developed
by the authors of this paper to create Pareto charts,
application used by students in the Quality
Management applications.
Figure 5: Excel application
When finished, the chart reveals that the
most frequent complaints are of the type: time
problem solving, information provided,
treatment, working program.
So, looking at the graph, it appears that we
first must operate on the four cases presented
in the previous paragraph.
Therefore, after a brainstorming session, it
was compiled a Ishikawa diagram, where we
tried to find the factors (the causes)
influencing the increasing number of
complaints over the service department.
Figure 6: The resultedIshikawa diagram
Therefore, following the chart analysis and
discussions it had been established a set of
measures:
the founding of a central dispatch service,
which:
- to receive complaints directly to a toll free
telephone number,
- to allocate a time resolution of each complaint
and at the end of this period to contact the
complainant and find out the steps taken and the
state of play in the case of larger problems;
providing in each subsidiary stocks of spare
parts, and in case of unique or the imported ones,
ensuring the lowest possible delivery dates;
allocation of the servicing to the collaborating
firms;
giving up some manufacturing companies
whose products brings the number of complaints
and guidance to other international companies;
staff training in marketing;
overtime (extra hours) payment.
So, being set all that, a new series of data was
collected, over the same period of time, (ie
September 2009 - March 2010), and after drawing
a new Pareto chart, it was noted that these
measures had the desired results, ie fewer
complaints and complaints.
189
Figure 7:Data collected between September 2009 -
March 2010
After the data collecting, we proceeded to
arrange them in descending order, as we did in the
first step, according to Table 5 and, then we
calculated the relative and cumulative frequency
(Table 6).
Table 5:Data rearranged
Table 6: The data frequencies calculation
The application of the measures outlined above
has led to a drastic decrease in complaints to the
service department, as shown in Figure 8.
Figure 8: Pareto chart for the data collected in the
second phase
The Excel application adds a plus, meaning that
on a separate sheet there can be seen the two
Pareto charts, the initial one and the final one, after
applying the measures that have resulted from the
first interpretation of Pareto charts.
As shown, the amount of complaints has
dropped by almost 70%.
The application brings a benefit, meaning that,
after the drawing of the first chart and after
applying the necessary measures (as a result of
brainstorming sessions and use of Ishikawa
diagram), starting with the recording of a new data
series for the same amount of time, introducing a
new data series (e.g. at the end of each month), it
can be observed, in real time, the effect of the
implemented measures by comparing the two
charts, so it could be applied in due time the
inherent corrective or preventive actions.
Figure 9: Comparison of the two Pareto charts (the
initial and the final ones)
4. Conclusions
The elaboration of an algorithm for the use of
methodological tools is helpful in solving quality
problems, especially if supported by a program or
software.
In this work, it has been developed such an
algorithm, meant to solve the deficiencies in
communications with clients, communication
between subsidiaries and the lack of coordination.
190
In support of this algorithm an Excel application
was created, which combines some of the tools
proposed to solve these problems, in this case
study as these two statistical tools, respectively the
Ishikawa diagram and Pareto diagram.
The application brings a benefit, meaning that,
after the drawing of the first chart and after
applying the necessary measures (as a result of
brainstorming sessions and use of Ishikawa
diagram), starting with the recording of a new data
series for the same amount of time, introducing a
new data series (e.g. at the end of each month), it
can be observed, in real time, the effect of the
implemented measures by comparing the two
charts, so it could be applied in due time the
inherent corrective or preventive actions.
References
[1] u, N., Dima, O., Guru, Gh., Gonzales
Barajas A., - Sisteme de asigurare a calitii,
Editura J unimea, Iai, 1998.
[2] Ciobanu, M., Iacob, D., Mironeasa, C., -
Ingineria calitii, Editura Printech, Bucureti,
1999
[3] Entescu, A.-M., Entescu, M.-Al., - Calitate.
Terminologie comentat, Editura Tehnic,
Bucureti, 2000.
[4] Gherghel, N., - Ingineria calitii. Aplicaii de
sintezi teste, Editura Cermi, Iai, 2006
[5] Gherghel, N., - Instrumentele controlului
calitii, Evaluarea i controlul calitii, p. 117-
254, Editura J unimea, Iai,1998.
[6] Gherghel N., Seghedin, N., - Consideraii
asupra diagramelor cauz efect utilizate n
controlul calitii, Tehnomus X, Tehnologii
i produse noi n construcia de maini, vol. V,
seciunea Calitate i fiabilitate, p. 122 129,
Suceava, 28-29 mai 1999.
[7] Gherghel N., - Noi tipuri i variante de
diagrame cauz - efect pentru controlul
calitii, Tehnomus X, Tehnologii i produse
noi n construcia de maini, vol. V, seciunea
Calitate i fiabilitate, p. 130- 137, Suceava, 28-
29 mai 1999.
[8] Gherghel N., - Managementul calitii n
sisteme de fabricaie, Note de curs masterat,
Univ. Tehn. Gh. Asachi Iai, Facult. Constr.
de Ma., 2003 - 2005.
[9] J uran, J . M., - Planificarea calitii, Editura
Teora, Bucureti, 2000.
[10]Mittoneau, H., - O nou orientare in
managementul calitii: apte instrumente noi,
Editura Tehnic, Bucureti, 1998.
[11]Noye, D., - Ghid practic pentru controlul
calitii. Principii, metode, mijloace, Editura
[12]Perigord, M., - Etapele calitii. Demersuri i
instrumente, Editura Tehnic, Bucureti, 1997.
[13]Poti, Ch., - Diagnosticul calitii. Metode de
expertiz i investigaii, Editura Tehnic,
Bucureti, 2001.
[14]Trandafir, M., Antonescu, V., - Calitatea.
Metode i tehnici de lucru. Analiz, evaluare,
control, O.I.D. I.C.M., Bucureti, 1994.
[15]Zetu, D., Carata, E., ura, L., - Ingineria
calitii n sisteme de fabricaie, Editura
J unimea, 2000.
[16]Bass, I. - Six Sigma Statistics with EXCEL and
MINITAB, McGraw-Hill, NY, 2007
[17]Durkee, J ., - Tools for the management of
qualitypart II, Metal Finishing, Volume
105, Issue 3, pp70-71, March 2007.
[18]Gryna, F., Chua, R., Defeo , J . - J uran's
Quality Planning and Analysis for Enterprise
Quality, McGraw-Hill Series in Industrial
Engineering and Management, NY, 2005
[19]Luo, X. G., Kwon, C. K., Tang, J . F. -
Determining optimal levels of engineering
characteristics in quality function deployment
under multi-segment market, Computers &
Industrial Engineering, Volume 59, Issue 1,
pp. 126-135, August 2010.
[20]Pyzdek, T., Keller, P. A. - Quality
Engineering book, Second edition, M. Dekker,
N.Y 2005
[21]*** - Tools for Developing a Quality
Management Program: Proactive Tools
(Process Mapping, Value Stream Mapping,
Fault Tree Analysis, and Failure Mode and
Effects Analysis), International J ournal of R. O.
B. P., Volume 71, Issue 1, Supplement 1, pp.
S187-S190, 2008
[22]***www.bain.com/management_tools, 2010
[23]***www.borland.com/us/rc/lifecycle-quality-
management, 2010
[24]*** www.epa.gov/quality/qmps, 2010
[25]***www.mastercontrol.com/quality-
management-software/total-tqm/tools, 2010
[26]***www.sixsigmaiq.com, 2010
[27]***www.skymark.com/resources/tools, 2010
191
ABOUT NATI ONAL STRATEGY OF ROAD SAFETY
I N PRESENT AND FUTURE
CI UBOTARI U DANUT
1
, NECULAI ASA VASI LE
2
1
Ministry of Domestic Affairs and Administration, SPCRPCIV, Botosani, Romania, email
danutciubotaru@yahoo.
2
Gheorghe Asachi Technical University, Iasi, Romania
Abstract: After the year 1990, the car park and also the number of drivers have almost tripled,
being in a complete contradiction with the slow development of the infrastructure, leading to an
increase in the percent of material and human damages as a result of traffic accidents. As far as
Romania is concerned at the level of the European Union, during 2001-2010 recorded a 25%
increase of the number of deceased persons, which places our country on the last but one place in
the European top.
Keywords: deceased, seriously injured, traffic accidents.
1. INTRODUCTION
The continuous increase of the motorization indices
together with the increase of the number of drivers as
well as the various infrastructure rehabilitation sites for
national and European roads led during the last years to
the blocking traffic, therefore influencing the behaviour
of certain traffic participants and implicitly the
accidents dynamics evolution in Romania.
The present project tries to draw attention over the
increasing number of accidents with several victims
and also their severity.
2. THE TRAFFIC ACCIDENTS REPORT
RECORDED AT THE WORLD LEVEL,
EUROPEAN UNION AND ROMANIA IN 2009
a) At the world level, the traffic accident holds the first
position among the causes of violent deaths [3,4]:
- over 1,26 million deceased people 3200 every day;
- over 35 million injured people 50.000 every day;
- over 10 million handicapped people as a result of the
accident;
The most traffic events are recorded in developing
countries and in transition countries, although they
have only 32% of the total number of vehicles.
b) At the European Union level, the situation of the
traffic events is as follows:
- over 50.000 deceased people;
- 1,5 million injured people;
- over 160 billion Euro for social costs;
Most of the experts consider that roads can be
described as being far the most dangerous and exposed
to accidents of all means of transport; [5].
The absence of a strong and well-defined strategy of
supervision and control encumbers the development of
a safety culture of the traffic leading to an idle violent
attitude among the traffic participants.
c) In Romania the number of the traffic accidents
reported to the number of inhabitants, vehicles, holders
of driver licenses is well above the EU average as
follows:
- over 2400 deceased people annually;
- over 6000 people seriously injured every year;
- road traffic acquired a special significance in the last
20 years as follows:
- the car park developed from 2,2 million vehicles in
1990 to 5.323.960 in 2009
- the number of drivers increased from 3,3 million in
1990 to 6.063.410 in 2009
- the road infrastructure has remained at the same
parameters as in the 90s.
The contradiction between the dynamics of the
above mentioned components and the slow
development of road infrastructure, unable to ensure
the fluency and safety conditions of the crowded
traffic, increased the threat of traffic events.
192
Figure 1: The evolution of serious traffic accidents during 1990 2008
Table 1: The evolution of serious traffic accidents during 1990 2008
Year No. of accidents No. of deceased Severely injured
1990 9708 3782 6137
1991 8948 3078 7789
1992 8781 2816 6960
1993 8791 2826 8302
1994 9381 2877 8198
1995 9119 2863 7698
1996 8931 2845 7504
1997 8801 2863 7451
1998 8457 2778 7221
1999 7846 2505 6594
2000 7555 2499 6315
2001 7244 2461 5963
2002 7236 2411 5976
2003 6689 2229 5585
2004 7068 2449 5775
2005 7211 2629 5885
2006 7164 2587 5780
2007 8503 2800 7089
2008 10592 3053 9350
Total:
Accidents: - 158.025
Deaths: - 52.351
Severely injured: - 131.572
The outburst of road traffic after 1990 led to the
doubling of the number and consequences of the
accidents with victims compared to those before 1989,
offering another dimension to this phenomenon.
3. THE CATEGORIES OF COSTS RELATED TO
TRAFFIC ACCIDENTS [1]
- medical costs, material damages and losses for the
society;
- administrative costs (insurance policies, police, etc);
- the evaluation of personal sufferings;
- damages as a result of some slight accidents, with
small material losses, which are not recorded in the
police
statistical reports;
- the life value reported to the value of the average life
time;
In most countries the first two categories are taken into
account.
There are two evaluation modalities:
Approximate evaluation and estimated evaluation.
3.1. The approximate evaluation
In the formula used at the European level, the
following definitions have been taken into account:
dead =when the decease occurred during the first
30 days after the accident;
seriously injured =the injure which requires an
immediate hospitalization;
slightly injured =the other categories that are not
included in the above-mentioned categories;
According to these conditions, we have the formula:
C =PNB (25,1D +1,64R +0,25r) (1)
Trend of severe acci dents 1990-2008
0
2000
4000
6000
8000
10000
12000
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Accidents
Killed
Severly injured
193
PNB the gross domestic product for each
inhabitant;
D the number of people deceased in accident;
R the number of people seriously injured in
accident;
r =the number of people slightly injured in accident;
3.2. The estimated evaluation
Although it is not so accurate, it shows a larger degree
of precision than the approximate evaluation due to the
additional information introduced in the formula.
Epd =PNB (Vv - X) +12 Cu (Vm - Va) +CMS (2)
Where:
PNB =the gross national product;
Vv years length of life expectancy;
Cu the monthly pension rate due to the heirs;
Vm age at which the pension is paid to those
entitled.
Va the age of the entitled people when they
received the first pension;
CMS medical costs as a result of the accident.
Romania wastes every year a billion Euro for the traffic
accidents, direct social costs calculated by the World
Bank.
4. THE NATIONAL STRATEGY FOR ROAD
SAFETY 2011-2020 [2]
The road traffic national strategy for 2011-2020
represents a public policies document on a long term
which proposes the creation of a new public policy
related to road safety, taking into account the poor
results of the actual policies. During 2001-2008
Romania recorded a 25% increase of the number of
deceased people, which ranks our country on the last
but one place in the European chart, according to figure
2.
5. CONCLUSIONS
After examining the statistics we concluded the
following:
-Most of the drivers do not acknowledge the
relationship between the traffic rules and the safety of
the road traffic, being willing to break the rules if they
consider that this does not endanger their safety at that
moment.
-The slow development of road infrastructure in
contradiction with the number of vehicles and holders
of driver licenses which almost tripled after 1990.
-In this situation and in the context of adopting a new
Plan of road safety actions for 2011 2020 by the
European Commission, plan which has as objective the
50% number decrease of injured people and victims
from traffic accidents at the level of the European
Union, Romania has to adopt a long term strategy
which should aim to stop the increase of the number of
victims and attaining the European objective.
REFERENCES
[1] Asandei Cezar Doctors Thesis: Researches over
the dynamics of traffic events pedestrian vehicle
Transilvania University, Brasov 2001
[2] The European Council of Road Safety ETSC,
J une 2010: PIN the index of performance in the
field of road safety.
[3] Avramescu N, etc The Dynamics of serious
traffic accidents 2002
[4] Statistics: the Traffic Police Department of
Bucharest and the Community Public Department
of Driving Licences and Vehicles Registration of
Botosani.
[5] Curt Rich, Drive to survive, USA, 1999.
Figure 2: Increase the number of decease
-55
-50
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
L
V
E
S
*
P
T
E
E
F
R
*
L
T
I
T
*
I
L
I
E
*
D
E
*
S
I
S
K
C
H
B
E
*
F
I
*
S
E
A
T
H
U
N
L
C
Z
L
U
*
U
K
*
*
D
K
C
Y
N
O
E
L
*
P
L
B
G
R
O
M
T
-36%(EU27)
194
195
RESEARCHES CONCERNI NG THE CORRELATI ON STRUCTURE OF
PROPERTI ES OF STEELS FOR METAL CONSTRUCTI ONS
STRONGLY REQUESTED
Stela Constantinescu
1
, Maria Vlad
1
1
University Dunarea de J os of Galati, Faculty of Metallurgy , Materials Sciences and Environtment,
stela.constantinescu@email.ro
Abstract: The RV52 steel plates are used to manufacture liquefied and compressed gas reservoirs,
recipients and pressure vessels which operate at low temperatures, big pressures and highly loaded
structures . The making technology and thermique treatment determine high properties such as : strength,
wear resistance and tenacity . The normalized plate sample were studied to determine the effect of the
tension release heat treatment on the properties of steel plates used for metallic welded construction .
From the relations expressing the link between the strength and tenacity relation o none hand and the
structural characteristics, on the other it is obvious that the only factor which leads both to increasing
strength properties and decreasing transition temperature is the finishing of the ferrite grains .
Key words: thermique treatment , strength, tenacity, ferrite grains
1.Introduction
The plates of RV52 steel are used to
manufacture liquefied and compressed gas
reservoires , pressure vessels and devices
operating at low temperatures and within heavily
loaded metal structures.
During the processing and operation in the
basic steel mass a number of phenomena can
ocurr such as:lamelar spreading, fissures in the
welded areas and easy breaking. Occurrence and
development of these phenomena depend on the
steel chemical composition and the semiblanks
elaboration and processing conditions.The
normalised plates of RV52 steel feature a ferite-
perlite structure, the obvious tendency being that
of reaching an as low as possible content of
perlite with the corresponding decrease in the
carbon content.The heat normalising treatment of
the thick plates took place in the roller continuing
furnaces from the rolling mill of SIDEX Galati
observing the follwoing parameters : heating up to
temperatures A
C3
+20
o
40
o
C, air cooling.
The analysis of the relation ship between
strength and tenacity, on the one hand , and
structure characteristics, on the other hand, shows
that the heat tension-relieving treatment and
especially, the normalising treatment considerably
increase the strength /resistance and decrease the
transition temperature due to the ferrite grain
shrinking process [1].
Manufacturing plates of high chemical and
structure homogenity leads to an isotrophy
corresponding to these properties, which attracted
the researchers interest .
2. Experimental researches and results
In order to carry out the researches on the
correlation of microstructure with the properties
of steel RV52 , normalized and detensioned, the
following working variants have been established:
elaboration in a 50t electrical furnace of 15 t
thick plates by ingot casting, normalising
treatment , and sampling for heat treatment
detension in laboratory;
elaboration in a 150t converter , by 25 t ingot
casting and continuing casting in slab , of thick
rolled plates , normalizing treatment , and
sampling for heat treatment detension in
laboratory [2,3]. The chemical composition of the
steel investigated is given in table 1.
196
Table 1. The chemical composition of material.
Chemical composition, [%]
C Mn Si P S V Al Ni
0,14 1,4 0,26 0,016 0,010 0,050 0,024 0,28
Sampling the steel RV52 plates before and after
normalising has been carried out acc to the
following scheme :
- the samples of the plates made by ingot roling
have been taken from the edges and the axis
corresponding to the upper, middle and bottom
parts;
- the samples of the plates made by cast brame
rolling have been taken from the edge and axis.
The metallographic structures of the samples
without heat treatment, from the 15t steel ingots
made in the electrical converter , for upper,
middle and bottom sides are illustrated in figures
1a,1b,1c .
Figure 1. Metallographic structures of the samples without heat treatment fromthe 15t steel ingots made in the
electrical converter : a) upper; b) middle; c)bottom
The samples have been taken from the central
axis of the plate corresponding to : a) upper
side/head, b) middle part ; and c) bottom side/foot
of the ingot [4].
The microscopic structures of the plates made
from steel elaborated in the LD converter and
continuously cast, positions edge and axis, are
illustrated in figures 2a, 2b.
197
Figure 2. Metallographic structures of the plates made fromsteel elaborated in LD converter : a) edge; b) axis
Metallographic slifs have been sampled from the
normalized plate samples. The slifs have been
prepared in longirudinal cross section acc to the
classical procedure [5,6]. Determinations have
been performed acc to STAS 5949-95, and the
results are given in table 2 .
Table 2 . Non- metalic inclusions score
Score
Types of non- metalic inclusions
Total score over
the same field Sulphides
S)
Oxides
OL+OP
Silicai
SF+SP(SN)
Nitrides
NT+NA
Non- metalic inclusions score
1 2 3 4 5 6
a
1
0,5
1
1,5
1,5
1
1
1
1,5
0,5
0,5
0,5
4
3,5
4
Max. score
from types of
inclusions
1 1,5 1,5 0,5
b
0,5
0,5
1
1,5
1,5
0,5
1,5
1,5
1,5
0,5
0
0
4
3,5
3
Max. score
from types of
inclusions
1 1,5 1,5 0,5
c 1
1
1
0,5
1,5
1
1
1
1,5
0,5
0
0
3
3,5
3,5
Max. score
from types of
inclusions
1 1,5 1,5 0,5
198
The elaboration technique has also been
assessed by determining the non- metalic
inclusions score as shown in table 2.
The impurest zones, irrespective of the sample
position, are those corresponding to the upper and
middle side of the ingot . In the axis purity is
lower as compared with the sample edge [7,8].
When elaborated in the electrical furnace, a higher
purity is reached than that in a converter. In the
case of continuing casting, the differences reported
between edge and axis are considered higher that
with ingot casting. The purity results show that the
C, Mn and S segregation zones are also zones
which include the non-metallic inclusions [9,10].
A typical image in terms of purity is given in
figure 3 .
Figure 3. Typical aspect of purity of RV52 steel plates,
without attack.
Magnified x100
Micro-structural measurements have confirmed
the presence , in the max C and Mn segregation
zones , of higher perlite proportions of these
substances as shown in figures 1a,1b,1c and 2a,2b.
The metallographic structure of the heat treated
plates is given in figure 4 which highlights a fine-
granulation ferrite-perlite structure [11,12].
Figure 4. Micro-structure of normalized RV52 steel
plates Magnified x 100
Variations and distributions of the mechanical
properties : ultimate strength (R
m
), yielding point
(Rp
0,2
), breaking elongation (A
5
%), resilience on-
v-grooved longitudinal samples, tested to -20
o
C
and 50
o
C, from upper, middles and bottom sides
are graphically illustrated in figures 5,6,7,8.
Sample position in comparison with the ingot Sample position in comparison with the ingot
Figure 5. Tensil strength (Rm) values distribution Figure 6. Yield strength (Rp
0.2
) values distribution
of samples in various zone of the sheet of samples in various zone of the sheet
199
Figure 7. Elongation A
5
values distribution of Figure 8. Break energy at shock (KV
L
) values
of samples in various zone of the sheet distribution on the longitudinal piece test at --20
0
C.
Figure 9. Break energy at shock (KV
T
) values distribution
on the transverse piece test at -20
0
C.
The resilience of the cold- deformed and heat-
treated samples under the above mentioned
conditions has been determined at temperatures of
- 20
o
C i 50
o
C, acc to figures 9.
In order to establish the influence of the heating
temperature subsequent to cold deformation of the
samples subject to 4%,8% and 12% degree of
deformation , these have been treated acc to the
treatment cycle described below :
- heating: 250
0
C; 500
0
C; 650
0
C;
- exposure: 160 min. As a result of 2 min/ mm
exposure;
- air cooling. [%]
From the analysis of the tension-relieving
graphics, it has been found that :
- as a result of the tension-relieving heat
treatment at 250
o
C, the shock ultimate strength
at -20
o
C and 50
o
C decreases with respect to the
values of of the samples from plates of 4%,8%
and 12% degree of deformations , and the values
of the normalized samples figures 10, 11.
- determining the shock behavior is more obvious
when the degree of deformation is higher.
- with degrees of deformation of 12%, the shock
ultimate strength at -20
o
C and 50
o
C takes lower
values than those min admissible as indicated by
STAS 11502-90.
200
- heating to 650
o
C results in recovery of the
tenacity properties of RV52 steel , without having
reached the level of normalization.
- heating at 250
o
C has disastrous consequences,
because tenacity is completely damaged both with
respect to normalization and deformation states;
- by heating at 500
o
C, the values of the shock
ultimate strength increase as compared with
those rached by cold deformation, while keeping
however the influence of the cold deformation, i.
e. lower values for higher degrees of deformation
[13,14].
Figure10.The influence degrees of deformation and Figure11.The influence degrees of deformation and
heting for values break energy at shock, KV
L
(-20
o
C) heting for values break energy at shock, KV
L
(- 50
o
C)
3. Conclusions
The researches were focused on determining the
factors that cause properties variation in different
zones of the RV52 steel plates manufactured by
Arcelor Mittal Galati .
In order to obtain highly improved properties a
good correlation should be achieved between the
conditions of elaboration, deformation and heat
treatments highlighted by the chemical
composition and structure .
The cold plastic deformation of low degrees of
deformation: 4, 8, 12% results in poorer tenacity
properties as compared with the normalised state
values , which is more sensitive with higher
degrees of deformation.
The best values of shock ultimate strength both at
-20
o
C and 50
o
C have been obtained after a
tension relieving treatment at 650
o
C for 6 %
degree of deformation
The measurements of the sulphide , carbon and
manganese segregation , purity, microscopic
structures and the mechanical properties in these
zones have all shown that :
The C, Mn, and S segregation phenomena and
therefore the variation of the properties acc to
different directions are much more obvious in a
less careful elaboration in the 150t converter, as
compared with the elaboration in the vacuum-
degasing electrical furnace.
The chemical, structural and properties
anisotropy is much more obvious in case of 25 t
ingot casting as compared with 15t ingot
casting.
In case of ingot plates , the segregation and
anisotropy phenomena are more pregnant in the
plate axis than in its edges, i. e. In the upper and
middle sides than in the bottom side.
Taking into account the results obtained , in order
to diminish the chemical segregation which
causes the structure modification and properties
anisotropy , it is necesary to take technological
steps able to eliminate this phenomenon. Such
steps should include both the elaboration
201
technological parameters ( casting temperature,
vaccum technology) and the casting parameters (
ingot size) along with the use of some raw
materials ( liquid cast iron, refractory staff) which
should cause a min. impurification of the steel.
4 References
[1] Constantinescu, S., Metals properties and
physical control methods, Didact i Pedag.
Publishing House , Bucureti , 2004.
[2] Constantinescu S. , Drgan V., Destructive
and nondestructive testing of metals , Ed.
Evrika Brila , 2000.
[3] Walanbe T., Mechanical proprieties of Cr-
Mo steels after eleveted temperature service,
Partea I: Document II S-IX-116-79; Partea II:
Document II S-IX-1167, 1990.
[4] Steven N., Jise, Steel , 1989, Vol.193, p.141
[5] Sugiyama T., Kobe Seel Engineering
Reports,1995, Vol.25, nr.4, p.4.
[6] Constantinescu S., Drugescu E., Studies on
Mechanism and Kinetics of Phase
Transformation in Superficial Layers Using
Unconventional Procedures. Research
Contract no. 5005, Galati 1995, p. 53- 88.
[7] Constantinescu S., Drugescu E.,Radu T.,
Practical application of AE of the different
grades of steel, in Proceedings of the
25
th
European Conference on Acoustion
Emission Testing EWGAE 2002 Prague
Czech Republic , 11 13 september, 2002 , p.
135-143.
[8]Constantinescu S., Influence of
manufacturing process on chemical and
structural homogeneity of welded pipe sheets
for tanks and vessels working under pressure,
in Proceeding of the International Conference
on Advances in Materials and Processing
Technologies, september 18 21 , 2001,
Leganes, Madrid, Spain, p.57 .
[9] Constantinescu S., Researches on the optimum
heat treatment for the low alloyed steels used
at high temperatures, Metalurgia Internaional
, nr. 6 /2005 , ISSN 1582- 2214 , p.3-7 .
[10] S. Constantinescu, L. Orac ,Characterization
of steels in terms of their physical and
mechanical properties, Buletinul
Institutului Politehnic din Iasi, Tomul LI II
( LV II ) , fasc. 3, ISSN 1453 1690, p.51-
58, 2007
[11]Vlad M., Constantinescu S, Stratulat, E.
Improvent of steel sheet corrosion
resistance by thermic coverage method with
zinc-alluminiumalloys, Annals of Dunarea
de J os University of Galati, Fascicle IX
Metallurgy and Material Science, ISSN
1453 083X, 2007,Vol. 1, p. 32-38.
[12] Constantinescu S., Non-distructive control of
hot rolled products, BRAMAT 2007,
International Conference on Materials
Science and Engineering, February 22-24,
2007, Braov , ISSN 1223 9631, p.241-
246.
[13] Constantinescu, S., Orac L., Mechanical
Properties of thin films investigated using
micromachining techniques, METAL 2010,
Ostrava, Czech Republic ISBN: 978-80-
87294-17-8, p.926-931,2010 TANGER,
spol.s.r.o., www.METAL2010.com
[14] Radu, T. ,Constantinescu, S., Vlad, M., .
Morphologies of Widmanstatten structures
and mechanismformation in steels, J ournal
of Materials, Science Forum Vol. 636-637,
pp 550-555, Trans Tech Publications,
Switzerland,2010, ISSN 1662-9752,
www.scientific.net / M6F. 636- 637.550
202
203
DETERMI NATI ON OF THE EFFORTS THAT OCCUR I N THE CASE OF
PERFORMI NG I NNER THREADS WI TH FORMI NG ROLLERS
ION CRISTEA
1
, CRI NA RADU
1
1
Department of Engineering Vasile Alecsandri University of Bacau
Calea Marasesti 157, 600115, Bacau COUNTRY ROMANIA icristea@ub.ro
Abstract: - The aim of this paper is to present a simplified model for the theoretical calculation of the
deformation efforts that occur in the case of performing inner threads by forming rollers. Firstly,
determination of the normal stress distribution along the roll-blank separation line is presented and then
the calculus of the resulting force of deformation.
Key-Words: -Inner threads, forming rollers
1. Introduction
Although known since the last century, internal
threading by plastic deformation is not widely used
because of a poor number of theoretical and
practical researches made on this area;
consequently, it would be necessary to continue
and deepen the researches in order to setup the
theoretical and practical basis concerning the tool
geometry and its influence on the material
behavior.
The execution of inner threads by rolling is
made with thread rolling heads and it is specific to
the threads with internal diameter bigger than 30
mm. Their breeding range is restricted because of
their very high complexity.
Achieving inner threads by rolling differs from
tapping; in the first case, rollers have a rolling
motion over the interior surface of the workpiece,
combined with a permanent plastic deformation
made in the part material. The operation is similar
to the exterior rolling but in the last case the rollers
have a larger diameter. In the case of rolling with
heads, the material flow will appear in a
perpendicular plan to the diametric plan of the
roller [1].
The methodology of examination the
technological deformability of metals and alloys is
highly complex, mainly because it is very difficult
to objectively classify the metals plasticity under
different strain conditions and in function of
different technological parameters (temperature,
degree of deformation, deformation speed, etc).
If in the case of tap threading the threading
moments increase up to 100%, in the case of pipe
threading with vortex roll-process (fig. 1), the
loadings are 10 - 15 times smaller compared to the
classical processing by cutting.
The forming roller 1 rotates around the axis of
the device inserted inside the workpiece, eccentric
to its axis. Every rotation will bring the forming
roller in contact with the workpiece, forming the
thread profile. Meanwhile, the workpiece rotates
around its axis and moves on axial direction with a
rotation step. From theoretical point of view this is
the theory of final plastic deformation with a
complex kinetics of the deformation. To avoid the
great difficulties in solving these problems [2], a
simplified model for the theoretical calculation of
the deformation efforts, based on the use of slip
lines properties, is proposed[3].
Figure 1 Roller threading scheme
204
2 Problem Formulation
Assuming that the material has rigido-plastic
properties and taking into account the marginal
efforts for stress and speeds, it is necessary to
calculate the efforts acting on the spindle of rolling
head.
Considering the state of plane deformation, we
determine the normal stress distribution along the
roller-blank separation line and then the resulting
force of deformation. In this case the contact arc
may be replaced by a broken line [1], each segment
representing the edge of penetrating wedge. Thus,
the simplified scheme reduces to solving the
problem of plasticity of plane deformation state.
The differential equilibrium equations will be
reduced to a system of two equations. [2]
0 = +
y
xy
x
x o
ot
o
oo
(1)
0 = +
y
y
x
xy
o
oo
o
ot
The plasticity criteria of Mises and Treska-
Saint Venant are analogous and differ only by the
size of the constant k:
0,25 ( o
x
- o
y
)
2
+t
2
xy
=k
2
(2)
Further it will be analyzed the penetration of
the symmetrical sectioned wedge into the rigid-
plastic semiplane, by assuming that the plastic
mass are extruded on both sides. The boundary
conditions of the problem for the right side of the
semiplan are set up, by expressing the friction on
the straight lateral contour lines through the angle
:
t =k sin 2o, 0 _ o _ a/4 (3)
It is made the assumption that the roller acts
with the speed v. The sliding field in the xy flow
plane is shown in fig. 2. In our case the contour of
roller consists of two sectors. The first sector, BP,
is parallel to the x axis while the second one, PA,
is tilted with the angle a/3 towards the first sector.
Along the contact sector BP, the following
relationships exist:
o
y
=- P, t
xy
=0, v
n
=v
0
(4)
Along the sector PA:
o
n
=- n , t
nt
=- t , v
n
=- v
0
cos(a/3) (5)
Along the free surface AG:
o
n
=t
nt
=0 (6)
According to fig. 2, the angles , ,
0
and
are linked by relationships:
0
0
= 0 - o = - (7)
Figure 2The field of plane stresses
By pressing depth it can be expressed the length
of the contact sector c, with the following
relationship:
, ) j , ) h a c =
0 0
sin . sin cos u u (8)
The uncompressed condition of the material
leads to the equality between the surface under the
x-axis, pressed by the roller, and the area of
triangle situated above the x axis, represented by
the extruded material. Hence, the equation (9) is
obtained:
2ah +h
2
tg =(c +a) (c cos h) cos ( - 0
0
)
+
+(c cos - h) tg (9)
The equations (8) and (9) allow us to establish a
link between the angle and the depthh, as well as
between andc.Further, the equations (1) and (2)
need to be solved such that to satisfy the boundary
conditions and then the inverse problem must be
solved. Assuming that the material is in plastic
condition, the stresses distribution along the border
of roller must be determined. From the equilibrium
205
condition we can determine the pressing force of
the roller in semiplane.
By analogy with the problem of stabilized flow
when introducing a sharp wedge, the case of the
sectioned wedge was examined. In this case the
geometric similarity was violated, thus difficulties
were encountered in the determination of the
particle trajectory in the speed field. But there were
no difficulties to establish the stresses field at a
certain deformation step.
In this case, for each step the deformation is
considered small. The stresses field is determined
based on the integrals of the plasticity equations
for the OXY coordinates. The necessary
constructions in the OXY work plan are presented
in Fig 2. In the areas BMP, PDEA and AFG,
joined by central fields of slip lines, uniform
stresses states occur.
In the AFG area the stresses field is determined
from the conditionat the boundary AG: =t
nt
=
0. Hence:
o =- k, f =a /2 - (10)
where: o - medium stress; f - angle between
the x axis and the highest component of the main
stress, o1.
As simple stresses states are analyzed, along the
characteristics are valid the integrals of the
plasticity equations:
o =2 kf +c
1
(11)
where c
1
is a constant determined in the AGF
field:
(12)
In the area PAED, the stresses field has the
following expresion:
o =k cos 2o n; f =a / 2 - o (13)
while in the area BMP:
o =k - p; f =0 (14)
Knowing the c
1
constant for the normal stress
components, it results:
n =2 k ( - +o +cos
2
o) (15)
p =k (a - 2 + 2)
The punch pressure force, P, must be balanced
by the normal and tangential stresses on the right:
P =2aP +2c(nsin - tcos) (16)
Finally we will obtain:
P =2ka [a +2 - 2( - 0)] +4kc[0
0
+o +
+cos o sin
-1
sin ( +o)]sin (17)
Establishing the pressing depth, h, the angle
and the size of the coil facet, a, from the equations
(8) and (9) it can be obtained the length of the
contact sector, c, as a function of
0
:
, )
, )
0
0
sin cos
sin
u
u

+
=
h a
c (18)
By replacing eq. (18) into eq. (17), the
following relation results:
, )
, )
, )
, )
, )
, )
, )

|
|
.
|
\
|

u
+ u
+
+ u
|
|
.
|
\
|

u
+ u
|
|
.
|
\
|
+
u
+ u
= +
tg h
h a
h
h a
a
h a
tg h ah
. cos .
sin cos
sin
cos . cos .
sin cos
sin
.
.
sin cos
sin
2
0
0
0
0
0
0
0 2
(19)
which is a transcendent equation in 0
0
. It can be
solved by halving the interval [0; a/3].
4 Conclusion
An easier way to solve the transcendent
equation consists in expressing the three variables
a, h and c, as a function of the thread pitch. This
can be done for metrics, trapezoidal and Whitworth
threads by using the following relationships:
h =b
1
p; a =b
2
p; c =b
3
p (20)
The standardized values of the three
coefficients are given in table 1 [4]
Table. 1
206
Coefficients Metric Whitworth Trapezoidal
b
1
5/16 tg
u/2
0,64 0,5
b
2
0,25 0,137 0,36
b
3
5/16 sin
u/2
0,72 0,517
By replacing eq. (20) into eq. (3) results:
b
3
p[cos - sin( - 0
0
)] - b
2
psin( - 0
0
) =b
1
p (21)
By dividing withp we obtain:
b
3
cos - ( b
3
+b
2
) sin ( - 0
0
) = b
1
(22)
and hence:
, )
2 3
1 3
0
cos
sin
b b
b b
+

=

u (21)
From the eq. (21) we can determine the angle
0
as:
2 3
1 3
0 0
cos
arcsin
b b
b b
+

=

u (22)
From the eq. (4) we can determinethe thread
pitch, p:
2b
1
b
2
p
2
+b
1
2
p
2
tg =(b
2
+b
3
)p
2
(b
3
cos - b
1
)cos(-
0
0
) +
+b
3
p(b
3
cos - b
1
p)tg (23)
and hence:
, ), )
1 3 3 2
2
1 2 1
3
cos 2
cos
b b b b tg b b b
b
p
+ +
=

.
, ) u tg b b
1 3 0
cos
1
+
(24)
where the value of 0
0
is obtained from eq. (8).
References:
[1] G.A. Toropov, Dispozitiv pentru filetat interior.
Brevet S.U. 1551458-1990.
[2] V.V. Sokolovskii, Construction des champs de
contraintes et de viteses dan les problems de
lecoulement plastique, Inj. J ournal T., 1963
[3] R. Lee, S. Tupper, Analysis of Plastic
Deformation in a Steel cylinder Striking a Rigid
Target, J ournal of Aplied Mechanics , 21 (1),
63-71, 1954
[4] I. Cristea, Tehnologii de executie si control a
filetelor interioare deformate plastic, Ed.
J unimea 1999
207
CONSI DERATI ONS REGARDI NG THE USE OF CHUCK COLLETS I N
MECHANI CAL SYSTEMS
Traian Lucian SEVERIN
1
, Vasile RATA
1
1
Management, University Street, No. 13, 720225, Suceava, Romania,EU. severin.traian@fim.usv.ro,
rata.v@fim.usv.ro
Abstract: In this paper the authors aim to present the theoretical study of of constructions such as clenching
mechanisms with chuck collets. For this to has been modernized the test stand clenching mechanisms with chuck
collets existing in the laboratory. Also in the paper were aimed experimental tests of clenching characteristics:
clenching stress, dependence of clenching stress upon the main influencing factors
Keywords: clench, chuck collets, acquisition board, force.
1. General characterization of mechanical
systems with chuck collets
The main use of chuck collets:
- in the construction of grabbing systems of
robots, of fixing mechanisms from working
technical structures, self-feeding systems
within working centres, etc.
- technical helping systems to fix working
tools: drills, cutters, reamers, etc.
- control systems, positioning systems and / or
measurement systems, etc.
Chuck collets have one or two conical inside or
outside surfaces, a series of longitudinal nicks
delimiting a series of elastic areas, called jaws.
Besides chuck collets, the mechanisms of type
MA comprise rigid collets which act upon chuck
collet which is deformed until the active surfaces
of the latter one get in contact with the piece
centring surface, making its centring and
clenching.
Chuck collets-mechanisms have some
advantages, some of which are mentioned below:
- provide high centring accuracies (radius range
of centred surfaces being under 02...0.05mm);
- allow clenching of thin walls, easily deforming
and law gauge pieces;
- due to the relatively uniform repartition of
clenching stress , they are easily to be made and
so low cost.
Their main disadvantage consists in the
working field limited by values of 0.05 D,
where D is the diameter of centring surfaces. To
cover a wide range of D dimensions, sets of
different dimension chucks are required. If the
centring dimension tolerance does not fit within
the working region, then the mechanism does not
ensure the clenching stress required by minimum
dimensions or they are overdemanded, up to
breakage in the case of high dimensions when
being centred on outside surfaces. Also, under such
conditions, they lead to high accuracy and low
rigidity as a result of some (imperfect) line
contacts between the chuck collet and piece,
between chuck collet and rigid mandrel or collet
respectively. In some constructions there may
occur errors of axial orientation as well, especially
when the chuck collet has axial mobility.
2. Mathematical modeling of calculation
2.1 Mathematical model for stress
calculation
The drive force calculation Q of chuck
collet and mechanisms is made generally for
each type of stress, depending on constructive
peculiarities.
The physics model for stress calculation
208
Figure 1. Physics model of calculation
The calculation relations of drive forces Q
depending on total clenching forces S are:
- for mechanisms in fig. 1a and 1b:
Q =T1 +F
2
=S tg( +
1
) +
2
S (1)
- for mechanism in fig. 1c:
S =Qctg( +
1
) - Q tg
3
. (2)
Calculation relations are valid whether chuck
collets 1 or rigid mandrels 2 are mobile.
Frontal anterior 4 (fig. 1a), or posterior (fig. 1b)
bearings lead to appearance of axial friction forces
F
2
=
2
S between piece 3 and chuck collet 1
whereas the frontal bearing 5 of chuck collet (fig.
1c) leads to the appearance of radius friction force
F3 =
3
Q.
The significance of noted items of relations is:
-1and 3 are friction coefficients between
the chuck collet 1 and rigid collet 2, frontal
bearing 5 respectively;
-2 friction coefficient between chuck
collet and piece, having values between
0.2... 0.7, depending on the shape of centring
surfaces of chuck collets (fig.3).
Chuck collets- mechanisms have become
widely spread, especially in the construction of
devices such as self-centred mandrels and chucks.
They are frequently used in works from pig and
pipe on turret lathes, semi-automatic or automatic
lathes for large-scale series production and for
machine equipment as well universal tools meant
especially to fine mechanics. In this case the
devices are fitted with sets of chuck collets to
satisfy the current centring demands of pieces even
under conditions of small series production or
unique. The fixing of straight-shank tools on
cutting machines is also made by the help of
devices containing such mechanisms. Under these
working conditions, constructive and functional
restrictions are called for regarding the fixing
accuracy and axial shares carrying out.
Between the axial run of a chuck collet and the
run c
e
of a jaw, the following relation may be
written:
c
e
=c
i
tg. (3)
3 Experimental stand
Figure 2. Structure of experiment stand, 1-computer labviewsoft, 2-dynamometer, 3-chuck collet, 4-dynamometer
key, 5- tensiometer bridge, 6-Labjack acquisition board
209
Figure4Login scheme of data acquisition soft.
On the above presented stand, the value of
effective clenching forces F
ef
at the level of chuck
collet jaws was taken and compared to the
calculation force F
Sc
approximated by model F
Sc
(1), (2).
At the same time, the precison degree of
clenching forces, depending on the level of
working/stress force could be determined.
The research stand of functional characteristics
of fixing chuck collets mechanisms uses a data
acquisition soft of a highly appreciable originality.
The experimental results obtained are shown in
table 1
Tabel.1 Experimental results of chuck collets with experimental data processing
Nr.crt. Nr.div.(V) K
f
(N/V) F
i
(N) F
Sc
(N) F
ef
(N)
Introduced
Moment
M
i
(N/m)
Measured
Moment
M
f
(N/m)
Moment
measured
by S
c
M
f
(N/m)
i=
F
ef
/
F
Sc
1. 1.74
1385
2409.9 1025.05 388.88 35 42 46.127 0.32
2. 2.481 3436.18 1401.58 435,.8 40 47 63.071 0.31
3. 2.743 3799.05 1615.92 555.55 45 60 72.716 0.34
4. 3.014 4174.39 1775.58 657.4 50 71 79.901 0.37
5. 3.218 4456.93 1895.75 712.96 55 77 85.309 0.376
6. 3.345 4632.82 1970.57 842.59 60 91 88.676 0.427
7. 3.480 4819.8 2050.1 916.6 65 99 92.254 0.447
8. 3.605 4992.92 2123.74 981.48 70 106 95.566 0.464
9. 3.724 5157.74 2193.84 1018.51 75 110 98.723 0.465
10. 3.862 5348.87 2275.14 1250 80 135 102.381 0.54
11. 3.951 5472.13 2327.57 1268.51 85 137 104.74 0.544
12. 4.105 5685.42 2418.23 1314.81 90 142 108.82 0.546
210
Fig.5. Variation of calculation forces and experimental forces
4. Conclusions
From the analysis of experimentally obtained
graphs, the following conclusions may be drawn:
- experimental forces are smaller than those
obtained by calculation, as friction forces
occur between collet and conical bearing,
between collet and semiproduct;
- an efficiency field in using the mechanisms
in technical systems may be defined;
- for mechanical systems of automatized
structures, corrections are necessary to be
taken to have in view determination errors of
these characteristics.
References
[1.] Rata V, Severin T. Managementul
proiectrii dispozitivelor mecanice.
Tipografia MATRIX Bucureti, 2008.
[2.] Rata V, Maiorescu A Teoria proiectrii
dispozitivelor. Universitatea tefan cel Mare
Suceava, 1992.
[3.] Rocule S. s.a - Proiectarea Dispozitivelor.
Editura Didactic i Pedagogic Bucureti,
1982;
[4.] Tache V, Brgaru A. - Dispozitive pentru
maini unelte. Proiectarea schemelor de
orientare i fixare. Editura Tehnic Bucureti,
1977;
211
THE CORROSI ON OF TUBULAR FURNACES I N PLANTS.
ATMOSPHERI C DI STI LLATI ON OF CRUDE OI L AND I N VACUUM FOR
FUEL OI L
Moroanu Marius
1
, Ovidiu Georgescu
2,
1
marius.morosanu2003@gmail.com
Abstract: The paper aims to investigate the corrosion of the furnace tube elements of intalatiile
Atmospheric crude oil distillation and vacuum of oil. Such phenomena occur in the ceiling and
accentuates intalatie distillation as we approach the staging area. Although the corrosion phenomena is
not frequently met for the tubes from the furnace, it produces serious damage because when a tube
reached limit corrosion in one spot it must be immediately replaced taking into accomrnt the fact that the
remaining tube is in perfect condition.
Keywords: corrosion, erosion, Sulfur compounds
1. I ntroduction
For unsulphurous fuel oil, the pipes from the
formance are made from steel P235GH, and for the
sulphurous fuel oil, the pipes are made from steel
with 5-9%Cr. The pipes are affected not only by
the corrosion phenomena, but also by the erosion
phenomenon at the end of the cycle in areas where
the speed of the fluid increases. The same
phenomena appears in the ceiling and increases as
it get closer to the transfer area. Solid salt particles,
sulphates and fuel, resulted from the water
evaporation, are strongly hit by the meandres of
introduction and by the ends of the tubes where the
fuid changes its course. At high speeds remarcable
erosions are generated. At the first tubes entering
the ceiling, where the speed of the fuid is 25m/s,
the speed of corrosion is 0.2 mm/year and 1,5-2,5
mm/year for the tubes before leaving the furnace,
where the amount of steem is bigger with a flow
rate of cc78 m/s.
Although the corrosion phenomena is not
frequently met for the tubes from the furnace, it
produces serious damage because when a tube
reached limit corrosion in one spot it must be
immediately replaced taking into accomrnt the
fact that the remaining tube is in perfect condition.
In the furnace, the fuel oil is heated up at 300-
340
0
C,the temperature at which magnesium
chloride hydrolyzes in proportion of 60-90%, and
the calcium chloride in proportion of 10%,
hydrochloric acid being released.
MgCl
2
+H
2
O Mg(OH)Cl+HCl
CaCl+H
2
O Ca(OH)Cl+HCl
Basic magnesium and calcium salts decompose
at temperatures exceeding 3200C.
Mg(OH)Cl + H
2
O Mg(OH)
2
+ HCl
Ca(OH)Cl+H
2
O Ca(OH)2+HCl
Hydrochloric acid in gas form is corrosive for
carbon steel, meanwhile the sulfide hydrogen
reacts even in the absence of water.
When processing the naphthenic oil with high
chloride content, the transfer lines are sublimated
to a combined effect of erosion-corrosion because
of the naphthenic acids and crystallized salts.
Naphthenic acids dissolve the protective layer of
FeS and with the sodium hydroxide, introduced
after salt removal, form sodium naphthenate, with
detergent role which cleans the surface of the
metal.
So the clean metal is subjected to the erosive
action of the crystals ( NaCl, CaCl, MgCl
2
,
Na
2
SO
4
).
212
2. Work tehnique
The experimental determinations regarding the
corrosion behavior of ferrous metals as a result of
the aggressive action of circulating oil in the
furnace tubes were made with test-tubes of steel
with chemical composition presented inTable 2.1.
The samples were mechanically prepared then
they were chemically etched with 5% HCl
inhibited solution, washed with demineralized
weak alkaline water and finally dried with acetone
in a desiccator. After testing, individual
observations were made on the condition of the
surface and on the quantitative determination of
corrosion.
From the pipe affected by corrosion, samples
were taken and were chemically and
microstructurally analyzed.
2.1 Experimental data anaysis
Figure 1 shows a schematic diagram of the AD
furnace with air heaters at the top, gas output.
Figure 1: Schematic diagram of the AD
furnace with air heaters
Tubular heat exchange elements are
simultaneously sublimated to both external
corrosion due to combustion gases and internal
corrosion due to technologic fluid and defense
requirements, which vary from one facility to
another.
Resulting corrosion products of the outer
surfaces due to combustion gases and on surfaces
within the same tube determined by the action of
the technologic fluid, are acting simultaneously on
the metall material. This simultaneously action not
only determines modifications of the material
structure, but also can have a negative influence on
the main mechanical properties. Accurately assess
of the changes occurring in the morphology of
steel constituents and the nature of the subsided
and differently dispersed phases allow more real
appreciation of the possible use of such tubular
elements.
Intemal corrosion of the tubular elements of the
furnaces from AD and VD plants is determined on
one hand by the aggressiveness of the technologic
fluid, and on the other hand by operating
conditions. Processed raw material is made from
paraffinic, naphthenic and aromatic hydrocarbons,
sulphurous hydrogen, mercaptans, various
chlorides, etc. For example, sulfur is dissolved in
fuel oil at a rate of 0.20% to 0.24% and chlorides
from 20 to 250 ppm. With technology operating at
temperatures above 4000C and pressure of 20bar,
fuel oil undergoes thermochemical decomposition
reactions, resulting a surplus of unwanted
corrosive agents, in addition to those existing in
the crude oil, or those witch are added as needed
for processing it. The intensity of metal tubular
elements degradation processes by corrosion is
additionally influenced by the speed of the
tehnologic fluid circulation and by their
turbulence within the pipes in the oven and transfer
lines. The fluid flow is biphasic, corrosion
processes, as it consisted, depend not so much on
the total acidity of crude oil, but on the total acidity
of the non-vaporised residue from the system
(Table 2). In order here to calculate the total
acidity coefficient of the non-vaporised was
determined residue efficiency in distilled fractions
from native (TAC) of crude oil remaining after
each fraction was removed.
The data presented in Table 2.2 show that the
total acidity coefficient of the non-vaporised
residue reaches maximum of 2.56 mg KOH / g
corresponding to the fuel oil exiting the furnace.
The flow rate gradually increasing with the
vaporization oil of a new quantity of hydrocarbs
No Types
of
steel
The concentration of elements, %
C Mn Si Cr Ni Ti P S
1 S235j
R
0,1
9
0,63 - - - - 0,05
0
0,07
0
2 P235
GH
0,1
6
0,52 0,22 - - - 0,03
0
0,02
9
3 X6Cr
Al13
0,0
7
0,08
5
0,82 13,2 - - 0,03
5
0,02
5
4 X6Cr
NiTi1
8-10
0,0
6
1,25 0,75 18,5 11,2 0,45 0,04
0
0,01
5
5 Sectio
n with
breaki
ng
area
0,1
1
0,42 0,29 - - - 0,02
1
0,02
4
213
determines a maximum of the total acidity
coefficient, of flow rate and temperature. All these
values are reached simultaneously at about exiting
of the fourmance. Given the severity of this
condition the speed of the corrosion should be very
high compared to the distillated fractions from the
same crude oil. However, this fact was not
determined in reality. This fact can be explained by
assuming that in this case crude oil contains
substances that ensures a protective layer of coke,
as not in the case of distillates. This layer can be
partially removed if the fluids turbulence is high,
fact that allows the installation of a localized
corrosion attack.
Table 2.2: TAC Total acidity coefficient of distillated
fractions and non-vaporised residue
Distillated
fraction
Efficiency
(%)
TAV
fraction
(mg
KOH/g)
TAV
calculated
for non-
vaporised
residue (mg
KOH/G)
Gasoline
70100
0
C
12 0,03 2,21
Petroleum I
170230
7 1,21 2,29
Petroleum II
230..280
0
C
6 1,25 2,38
Gas oil I
270310
0
C
16 1,82 2,53
Gas oil II 310-
350
0
C
3 1,99 2,56
Fuel oil >
350
0
C
56 3 0,55
Corrosion products resulting from the attack of
naphthenic acids are soluble iron naphtens which
can not provide protection as that of iron sulphide
resulting from the interaction with hydrogen
sulfide.
Along with naphthenic acid crude oil contains
sulfur compounds, hydrogen sulphide,
mercaptans, etc., whose attack on steel plant
enhances the corrosion processes specially at high
temperatures
Following this attack a layer of pores reducing
the speed of corrosion of naphthenic acids is made.
Test-tubes from carbon steel S235J R and
stainless steel X6CrAl13 and X5CrNi18-10 with
and without sulphide layer were tested in solutions
of naphthenic acids with TAC 145 mg KOH / g
during 20 days at a temperature of 100
0
C (Table
3).
It can be seen from the data presented in Table
2.3 that the existence of a compact layer of iron
sulfide reduces significantly the corrosion rate due
to naphthenic acids.
Material Protected
with FeS
layer
Corrosion rate
K
g
-
(g/m
2
/h) P
(mm/year)
S235jR No 0,092 0,102
Yes 0,071 0,079
X6CrAl13 No 0,074 0,082
Da 0,0092 0,0010
X6CrNi
18-10
No 0,014 0,016
Yes 0,0053 0,006
But in industrial conditions, it is possible that in
the presence of large amounts of naphthenic acids
in crude oil, with a high-speed of technologic fluid,
the FeS film can be damaged (broken) which
allows the initiation and localization of new firms
of localized corrosion. In these circumstances,
inner and outer surfaces of the tubular element are
highly stressed by the corrosion products.
The tube element from the AD furnace, which
carries crude oil out of the furmace at a
temperature of 340
0
C and pressure of 80 bar of
P235 GH steel, after 60,000 hours of operation,
had shown corroded areas with cracks.
The extermal diameter of the pipe, outside the
cracked area is 10% higher than the minimum
diameter, and in thefractured zone the remaining
elongation is 12.4% with corresponding reduction
of wall thickness to 0.9 mm and a reduction of
91% compared to nominal thickness of the tubular
element.
In the breaking area, the tubular element shows
an excessive oxidation due to local thermal
overstressing. This is due to the deposit of
corrosion products that reducing heat transfer,
determined the appearance of cracks with a depth
of 1.2-2.0 mm. Te cracking and further material
breaking were favorites by the greater thickness of
the decarbonized layer, caused by the mentioned
thermal overstressing, respectively by exceeding
its resistance capacity under the action of internal
pressure (Figure 2.2)
214
Figure 2.2. Cracks in the tube of the AD furnace
Metallographicall research conducted on
processed cuts from both corroded and unaffected
areas (2% solution of metal activation) revealed
microstructural changes resulting from corrosion.
In the unaffected area no changes were found in
the feritico-pearlitic layers, specific for P235GH
steel (Figure 2.3) and the corroded area presents a
modified feritico-pearlitic structure with deformed
grains.
Figure 2.3. The
materials
microstructure from
the unaffected area
(x500)
Figure 2.4. The
materials
microstructure from the
breaking area (x 500)
Also from the analysis of metallographic cuts, it
is revealed a spheroidizing of the carbides from the
pearlite and a precipitation at their ferrite grains
limits.
These specific types of precipitation are
specific for the steel which is maintained at higher
temperatures, copmared to the recommended
temperature range.
3. Conclusions
In the AVD plant which processes non-
sulphurous oils, corrosion processes are
determined primarily by naphthenic acidity and
distilled from crude oil, expressed by total acidity
coefficient (TAC), by the sulfur and chlorine
compounds, and operating conditions.
Sulfur compounds present in crude oil have
naphthenic corrosion braking action, as iron
sulphide skin, as a product of the interaction
between steel and sulfur compounds, is more
compact and more adherent than iron naphthenate
skin.
The flow rate of technologic fluid and its biphasic
nature, corrode with the vaporization phenomenon
to in balance and the start of thevaporos, liquid
stages distress determines the intensification of the
transfer lines corrosion and of the furnace pipes,
because in these parts of the installation there
combines cavity, corrosion and erosion phenomena
The breaking of the AD furnace tube was due to
exceeding the endurance limit of the steel as a
result of local overheating due to corrosion
products deposited layers, up to 8.2 mm thick.
Overheating caused a pronounced decarburization
and the emergence of cracks up to 2 mm grooves
on the esternal surface of the pipe.
References
[1] Shalaby H.M., Refining of Kuwait s heavy
crude oil: materials challenages, Workshop on
corrosion and protection of metals, Arab school
for science and tehnology, December 3-7
Kuwait
[2] Antonescu N.N., Moroanu M, Petrescu M.G.,
Niculae A.D., Georgescu L., Georgescu Ov.,
Issues on corrosion equipment in instalation by
distribution DAV naphthenic acids fromcrude oil,
Poster WP15- corrosion in Refinery Industry,
EUROCORR 2010, The European Corrosion
Congres, European Federation of Corrosion 13-
17 September 2010.
[3]Ropital F. Mecanismes et mecanique des
interactions plastic envirovmements. Cas
dindustrie petroliere, Plast Ox 2007 (2009),
p265-275., EDP Sciences 2009
[5] Lobley G.R., Stress corrosion cracking: case
Saudies in refinery equipment, The 6 th sandi
Engineering Conference, KFUPM, Dhahran,
December 2002, vol.5, 2002, p17-26.
215
APPLI CATI ONS OF FULL FACTORI AL DESI GN EXPERI MENTS
FOR L ASER WEL DI NG
Remus Boboescu
1
1
Politehnica University Timioara, remus_boboescu@yahoo.com.
Abstract:. Laser welding using Nd: YAG laser with continuous emission is applied for a low alloyed
steel. The study pursued molten areas characteristics in the material. On the weld cross section was
measured weld width, weld depth and the weld molten zone. Their variation was analyzed with power
and welding speed. A full factorial experimental design was applied for two particular values of the
distance between focal plane and the workpiece surface (defocusing depth). It presents mathematical
models, the ranking effects by Pareto charts, response surface method and the multiple ANOVA analysis
of variance. It showed the main effect of laser power in determining the weld characteristics.
Keywords: laser welding, Nd:YAG laser, full factorial design, laser beamdefocusing, mathematical model.
1. Introduction
Laser welding is widely applied machining of
metals and in particular for steels. The laser beam
is a concentrated heat source that allows for the
intensity of the laser beam 10
4
-10
7
W/cm
2
and for
interaction time between laser radiation and the
material 10
-3
-10
-2
s obtaining molten zones. Melting
material is used for welding. Several papers have
presented the features of the steel melted zone for
laser welding [1].
Main information about the weld characteristics
are given of the weld cross section. On this are
measured the weld width, weld depth and the weld
molten zone area. Sectioning the welds in a stable
area provides information about the ability to
perform laser beam melting material. Applied the
mathematical modeling for the molten zone of
welds is useful for estimating the parameter values
used in making welded joints. Type the full
factorial experiment with the statistical method of
ANOVA analysis of the multiple underlying
several methods of statistical analysis in
experimental research [2] [3].
This paper proposes a study on the laser welds
made on the low alloy steel plates.
Apply a complete factorial experimental design
type2
2
for two different experimental situations
given by the focal plane position relative to the
workpiece surface (Defocusing) for welds made
using laser irradiation Nd: YAG under continuous
regime. In experiments were varied laser power
and welding speed.
2. Experiments
The experiment consisted in made lines of
fusion (welds) ,110mm long, on Dillimax 500 steel
plates with thickness of 10 mm (carbon steel,
carbon content 16 . 0 %), figure 1.
Figure 1: Plate with welds a) Surface plate b) cross-
sections through the plate
216
Was used a Nd: YAG Trumph Haas 3006D
laser source with 3kW maximum power on a
continuous wave regime CW. Laser beam was
transmitted through a optical fiber with core
diameter of 0.6 mm
The focus system made a focal spot with 0.6
mm diameter. Lens focal length was 200 mm. As
protective gas argon was used with a flow rate of
20 l / min. Were used sheets of material with
dimensions of 10 130 100 mm for which were
made between 5 and 8 welds, with a distance of
over 10 mm between welds,.
In experiments was varied the laser power,
welding speed and distance between focal plane
and piece surface (defocusing or defocusing depth)
figure 2. Welds were cut in the stable part of the
weld near the place where welding process was
stopped. Weld section was processed
metallographic. Weld width, near the piece
surface, and weld depth were examined using a
microscope with precision of 0.01 mm. Melted
area was measured directly by its footprint.
Parameters varied in the experiments are
presentations in Figure 2. To focus within the piece
defocusing values are considered negative.
Figure 2: Parameters varied in the experiments
In the experiments were varied power and welding
speed. To statistically analyze the effects of
parameters was necessary to introduce a
dimensionless parameter values. Transformations
between the two systems are based on real-coded
relationships following:
2 P A [-] (1)
v B 22 . 2 33 . 2 [-]
(2)
The experimental plan is presented in Table 1 with
actual values that coded for laser power and
welding speed.
Analysis procedure consisted of presenting the
results of the mathematical model, ANOVA table
showing the correlation coefficients associated
with the mathematical model, Pareto chart showing
the hierarchy of effects and response surface is a
graphic representation of mathematical model. For
the mathematical model were presented two forms
for real values laser power and welding speed and
for coded system values. The first allows rapid
application of formulas and the second allows
direct analysis of the values of regression
coefficients.
Table 1: Varied parameters in the experiment
Based on these values were achieved Pareto charts.
On the weld cross section was measured near
the surface of the weld the width w[mm], at the
center of the weld the depth h [mm] and melted
area MA [mm
2
]. These measurements are shown
in Figure 3.
Figure3: Weld cross-section with the considered sizes
3. Effects of welding process parameters
The varied parameters laser power, welding
speeds and defocusing have the following effects
on molten zone dimensions:
- Laser power. Increasing the laser power produce
increase the intensity on piece surface therefore
melted material amount increases. From a certain
value, intensity not too high melting material
amount, favours material vaporization.
- Welding speed. Increasing welding speed
decrease the interaction (time) between laser
radiation and material. If the interaction time is
less then the molten zone dimensions are smaller.
- Defocusing. Defocusing by lowering the focal
217
plane within piece produce lower intensity at piece
surface by increasing laser spot area at piece
surface and from same issue will increase
interaction time between laser and material.
Focus within piece associated with the presence of
keyhole in welding bath will increase the spread of
radiation in keyhole and coupling of laser radiation
and material. Defocusing may thus have different
effects on melting material. You can not
predetermine a clear trend of increasing or
decreasing the molten zone. Defocusing effects
will be analyzed based on experimental results.
4. Weld width
For the weld width at defocusing 0
polynomial mathematic model is given by relations
(3) (4). Statistical analysis of variation is given in
Table 2.
Pv v P w 111 . 0 999 . 0 4335 . 0 7348 . 0 (3)
AB B A w 05 . 0 55 . 0 55 . 0 8833 . 2
(4)
Table 2: ANOVA table for weld width w at 0
Pareto diagram on Figure 4 shows that the weld
width increases with power and decreases with
welding speed. The two parameters are equal but
opposite sign effects on weld width. The
interaction between power and welding speed has
an effect on the decrease weld width. The
mathematical model presented is statistically
significant data for the effects of speed and power.
variation of weld width with power and welding
speed at defocusing 0 .It is noted that on the
experimental field weld width increases with
power and decreases with speed. Maximum values
for the weld width are obtained at the highest
power and lowest speed. Lower values of weld
width are not recommended because they are
associated with low weld depth.
Figure 4: Pareto Chart for weld width at 0
Figure 5: Response surface for weld width at 0
For the weld width at defocusing mm 2
mathematic model is given by polynomial
equations (5) (6).Statistical analysis of variation is
given in Table 3. Pareto chart in Figure 6 shows
that the weld width increase with power and
decreases with welding speed. Both effects are
statistically significant. Power effect is greater than
the welding speed. Interaction between the welding
speed and power decreases the width of the weld.
Pv v P w 666 . 0 444 . 0 999 . 1 5826 . 0 (5)
AB B A w 3 . 0 8 . 0 3 . 1 7166 . 2
(6)
Table 3: ANOVA table for weld width w at mm 2
218
Figure 6: Pareto Chart for weld width at mm 2
variation of weld width with power and speed at
defocusing mm 2 . It is noted that on the
experimental field weld width increase with power
and decreases with speed. The variations of weld
width are low at low power and high welding
speeds. It looks like that under weak irradiation
weld width may be associated with the laser beam
spot on the workpiece surface.
Thus the focus within the piece can produce
acceptable weld width even if using low power or
high values of welding speed.
Figure 7:Response surface for weld width at mm 2
The variation observed previously shown that
the weld width becomes more dependent on power
for the laser beam focus within the piece.
Interaction effect between laser power and
welding speed increases weld width. It is
associated with dynamic phenomena occurring in
the weld pool that influence laser radiation
absorption.
5. Weld depth
For the weld depth at defocusing 0
polynomial mathematical model is given by
equations (7) (8).Statistical analysis of variation is
given in Table 4. Pareto chart in Figure 8 indicates
that the weld depth increases with power and
decreases with welding speed. Effect of power is
much greater than the effect of welding speed. The
interaction between power and speed has a small
role and does not present statistical significance.
Pv v
P h
06105 . 0 60495 . 0
903425 . 1 566625 . 1

(7)
B A B A h 0275 . 0 3275 . 0 9675 . 1 3 (8)
Table 4: ANOVA table for weld depth h at 0
Figure 8: Pareto Chart for weld depth at 0
variation of weld depth with power and speed at
defocusing 0 . It is noted that on the
experimental field weld depth increases sharply
with the power and almost no variation with
welding speed. Welds with high depth presents
keyhole welding regime.
219
Figure 9:Response surface for weld depth at 0
For weld depth at defocusing mm 2
polynomial mathematical model is given by
equations (9) (10). Statistical analysis of variation
is given in Table 5. Pareto chart in Figure 10
shows that the weld depth increases sharply with
power. Power effect is statistically significant.
Weld depth and decreases the interaction between
welding speed and power. The two effects
mentioned above are close.
They are not given by the mathematical model
with statistical significance. It looks like that for
the focus inside piece the effect of power becomes
important. Dynamic phenomena occurring in the
weld pool becomes more intense, which is
reflected in the increasing role of the interaction
between power and welding speed.
v P v P h 8436 . 0 6216 . 0 7554 . 2 8541 . 1 (9)
B A B A h 38 . 0 48 . 0 87 . 1 5383 . 2 (10)
Table 5:ANOVA table for weld depth h at mm 2
variation of weld depth with power and speed at
defocusing mm 2 . It is noted that on the
experimental field there is a weld depth increase
with power. At high power has been a slight
decrease in the weld depth with welding speed.
Effect of welding speed is obvious by reducing the
increase with power at high welding speeds.
Figure 10:Pareto Chart for weld depth at mm 2
Figure 11:Response surface for weld depth at mm 2
Effects on weld depth analysis show that it
depends strongly on the power in both cases of
defocusing depth. Focusing inside the piece
increases the welding speed effects due to
increased contribution of dynamic phenomena in
the weld pool. They are very sensitive to variations
in time of interaction between laser radiation and
material controlled by the welding speed.
6. Melted area at the weld cross section
For melted area mathematic polynomial model
at defocusing 0 is given by equations (11)
(12). Statistical analysis is presented in Table 6.
Pareto chart in Figure 12 shows that the melted
area increases with power and decreases with
welding speed. The interaction between power and
speed decreases melted area. Power is the only
effect that has statistical significance. Contribution
of welding speed by its effect and by its interaction
with power is lower than power effect.. But the
effect of welding speed is close to power effect. It
looks like that for the melted area on weld cross
220
section are important both laser beam intensity at
workpiece surface and interaction time between
laser radiation and material.
v P v P MA 776 . 1 222 . 0 836 . 0 7664 . 1 (11)
B A B A MA 8 . 0 7 . 1 7 . 2 866 . 3 (12)
Table 6: ANOVA table for melted area MA at 0
Figure 12: Pareto Chart for melted area at 0
variation of cross section area of the molten weld
welding power and speed at defocusing 0 .
High values for melted area are recommended for
joint the two pieces in laser welding. It is noted
that the melted area increases with the power and
decreases with welding speed. The decrease with
speed is lower at the low power than the high
power. This shows the role of interaction between
power and speed. Highest values of the melted area
are obtained at maximum power and welding
speed on the experimental field.
For melted area mathematic model polynomial
at defocusing mm 2 is given by equations
(13) (14). Statistical analysis is presented in Table
7.Pareto chart in Figure 14 show that the melted
area on the weld cross section increases with
power. Power is the only effect that is statistically
significant.
Figure13:Response surface for weld melted area 0
Welding speed and the interaction between
power and welding speed lower melting area. The
two effects are equal. It is noted that the overall
cumulative effect of decreased melted area with
speed and with interaction between speed and
power is greater than the effect of increasing the
power given. It looks like the melted area is
heavily dependent on the time of interaction
between laser radiation and material.
v P v
P MA
1625 . 4 16250 . 4
39375 . 7 93545 . 6

(13)
B A B A MA 875 . 1 875 . 1 025 . 3 4843 . 3 (14)
Table 7:ANOVA table for melted area MA at
mm 2
variation in area melted on weld cross-section at
defocusing mm 2 .It is noted that on the
experimental field melted area increases with the
power and decreases with welding speed. The
decrease with welding speed is stronger at high
power. Increasing with power is greater at low
welding speeds. It looks like the importance of
interaction between power and speed. Maximum
values for the melted area are obtained at
maximum power and minimum welding speed. At
low power and high welding speeds there is a
domain withsmall variations for melted area.
221
Figure15:Pareto Chart for melted area at mm 2
Figure16:Response surface for weld melted area
mm 2
Focusing within piece shows as additional
elements for the weld cross section melted area
increased the welding speed contribution. This is
accomplished through both its effect and its
interaction with power. It is noted that the effect of
welding speed on melted area is stronger than the
effects of welding speed on weld width and weld
depth.
7. Discussion
Mathematical modeling using full factorial
design showed high values of correlation
coefficients and statistical significance was
obtained for sizes analyzed with power effect.
For weld width and weld depth the focus within
the piece led to an increase in power effect.
For melted area on weld cross-section variation
was contrary effect, namely increased speed effect.
It looks like that lower intensity laser beam to the
workpiece surface by focusing the laser beam
inside piece lowers laser beam intensity to the
workpiece surface. The intensity is mainly
controlled by power. So considered sizes are more
sensitive to changes in power. Fused area is related
to the total amount of melt obtained after
irradiation. At defocusing mm 2 the laser
beam heat source is inside the Keyhole at its front
wall. Keyhole front wall angle depends on the
welding speed.
Strong inclination of the keyhole front wall of
the laser radiation encourage reflection outside
keyhole. This reduces the coupling between laser
radiation and the material thus obtained decreases
the amount of melt. The mathematical models
presented are useful in determining the size of area
that can be melted by laser beam. The deep
welds are in keyhole regime. Molten zone should
be sufficient to melt the pieces and filler material
used in welding.
Mathematical modeling presented in this paper
can be used to determine the characteristics of
welds in concrete terms the imposition of
restrictions related to carrying out the welding.
For deep weld the weld width must be high.
Focusing the laser beam inside the piece from the
start ensures high values for weld width due to the
increased the size of laser beam spot on the
workpiece surface. Experiments have shown that
inside piece weld width decreases in comparison
with the weld at the surface but a certain amount is
kept relatively constant with depth in the material.
The high level of power ensures an appropriate
weld width both of the weld surface and inside the
piece.
Increasing the depth of the weld is given by
power level. The high level of power produce
welds in keyhole regime. Getting the welds under
keyhole regime is strongly associated with the
formation of pores in the material. Increasing
power and decreasing the welding speed leads to a
situation in which many pores are produced in
welds. Low welding speed produce significant heat
affected zone. Appears important to control the
laser beam intensity at the workpiece surface by
changing the defocus. Decrease laser beam
intensity independent of power provides a
moderate keyhole welding regime is useful for
reducing porosity in the weld.
For melted area is recommended the maximal
values in all cases. This can be achieved by
increasing power and decreasing welding speed.
222
Be sure that it is not excessive variations do not
produce enlarged pores and heat affected zones.
8. Conclusions
This paper has analyzed the characteristics of a
cross section through the weld associated with a
period in which the welding process showed
maximum stability. The effects of power and
welding speed are considered for two situations for
the laser beam focus relative to the workpiece
surface.To control variation in the welding process
by power and welding speed mathematical
modeling associated with experimental research
has shown the following:
-Power is the main effect on the characteristics of
the weld. We recommend a high level of power.
-In relation to central point of a factorial
experiment is recommended to increase power and
lower welding speed without reaching the extreme
values for both parameters.
- It is recommended to focus the laser beam inside
the piece to achieve a moderate keyhole welding
regime that is associated with reduced porosity.
- Modeling of the melted zone of the weld cross
section will be useful for the design of welded
joints.
The paper presents several ways of presenting
the measured sizes. The mathematical model,
Pareto diagram and response surface shows the
same variation. Each of them has a different utility.
From the technological point of view these
types of representations can be used to determine
the objective function values that characterize the
welds for data values of welding parameters
power, welding speed and defocusing. It can
address the inverse problem required to determine
the values of parameters power, welding speed ,
defocusing to obtain employment objective
functions, melted area, weld width, weld depth
within certain limits. For this problem is to use
utility of response surfaces. Response surface is a
function of grade 2.Applications of response
surface in optimization of this type, refers to the
determine of maximum and minimum extremes on
the experimental field from response surface
figure. Thus the presentation of multiple
mathematical models for the same experimental
situation is useful.
For laser machining processes defocusing is an
important parameter. Defocusing can be varied
continuously but that it may be the range is narrow
because much lower intensity of laser beam to the
workpiece surface compromise the welding
process. Small variations in defocus make that its
effect can not be seized. Experimental situation
presented in the paper is a case in which
defocusing variations are obtained without
excessive lowering laser beam intensity to the
workpiece surface.
Using mathematical modeling and statistical
analysis of the changes is a powerful tool for
understanding the causes of variations for analyzed
sizes and control the welding process.
References
[1] Kazuhiko Ono, Karoru Adachi, Yasuichi
Matsumoto, Isamu Miyamoto, Takashi Inoue,
Ryuichi Narita, Influence of oxide filmon weld
characteristic of mild steel in CO
2
laser
welding, ICALEO 1999
[2] Benyounis, A.G. Olabi, M.S.J. Hashmi, Multi-
response optimization of CO
2
laser-welding
process of austenitic stainless steel, Optics &
Laser Technology 40 (2008) p:7687.
[3] Lung Kwang Pan, Che Chung Wang, Shien
Long Wei , Hai Feng Sher, Optimizing
multiple quality characteristics via Taguchi
method-based Grey analysis, J ournal of
Materials Processing Technology 182 (2007)
p:107116.
.
223
U.S. QUENCHI NG AND DI MENSI ONAL STABI LI TY I N TI ME
OF 100Cr6 STEEL
Nicolai BANCESCU
1
, Constantin DULUCHEANU
1
, Traian Lucian SEVERIN
1
1
Management, University Street, No. 13, 720225, Suceava, Romania,EU. bancescu@fim.usv.ro,
dulu@fim.usv.ro, severin.traian@fim.usv.ro
Abstract: Due to the relatively high susceptibility to deformation and cracking, the bearings steel
creates major problems to quenching, especially in the case of complex geometric configurations. On the
other hand, the dimensional stability of bearing elements influences significantly bearing durability. In
this paper, the authors present a series of experimental results on the dimensional time behavior of
100Cr6 steel for several martensitic volume quenching.
Keywords: quenching, bearing steel, dimension stability.
1. General consideration
Sustainability of precision bearings largely
depends on the dimensional time stability provided
by the structure elements obtained from the final
heat treatment. In this case the notion of structure
defines the nature, form, size, distribution of
phases and structural constituents also the size and
sign of residual stresses induced in product
following due to the heat treatment applied.
Martensitic quenching is a heat treatment of the
technologies that induces stress and strain in
dimensional stability with implications for long
periods of time. The transformation of austenite
into martensite in quenching induces large amounts
of residual material stress, due to the difference in
specific volume of the constituents (martensite has
higher specific volume than austenite from which
it came). Finally, on the quenching product surface
tensile stress is formed which is summed
algebraically with the tensions caused by thermal
shock. Parts of the residual stresses are
"downloaded" during its return through specific
mechanisms. On the other hand in the structure of
quenching steel the residual austenite and
martensite phases are out of balance and so
unstable over time. These phases tend to evolve
over time, even at ambient temperature, the phase
constituents close to balance or causing
dimensional changes in some operational situations
affecting the product behavior in exploitation.
Collective concerns of the thermal treatments
laboratory FIMMM Suceava on optimization of
heat treatment of bearing steel dates back to 1985.
In the article are presented results of research
regarding the influence of quenching regime for
dimensional stability in time of 100Cr6 steel
(symbolizing the ancient RUL1).
2. Research methodology
Typically, dimensional stability in time is
expressed as relative linear strain variation over
time; the measurements are made until the full
stabilization of the dimensions. To study the
dimensional stability in time was designed and
implemented a device like in figure 1. With five
workstations, the device allows embedding of
cylindrical specimens with 10 x 150 mm,
between a high rigidity wall and an elastic wall on
that are bonded strain gauge.
Figure 1. Device for testing dimensional stability
224
To limit thermal deformation caused by
changes of ambient temperature environment after
mounting the samples device are placed in a
thermostatic chamber in which temperature is
maintained at 20
C with an accuracy of 0.2
C.
Strain gauges in full bridge are connected to five
bridges N2305, selected in advance as stability
while balancing adjustment. In order to ensure
stable mechanical properties over time after
implementation, the device was subjected to
thermal stress relief treatment of 100 hours at 200
C. Immediately after heat treatment, samples were

mounted in the device, gathering performing with
torque indicator handle wrench to maintain the
same measurement conditions. After placing the
device in the thermostat room and after
temperature stabilization was done to balance each
strain gauge bridge. Calibration of each station was
done with a orthotest. We worked with lots of five
specimens for each group thermal treatments are
presented in Table 1. Austenitizarea for quenching
was performed in a CARBOLYTE electric furnace
heated to 850
C, keeping time was 40 minutes. As

cooling environment mineral oil Lubrifin MET 1 R
type II was used, recommended for quenching
bearing elements.
Tabel 1. Quench cooling variants of the samples group
Group
code
Getting austenite
Mineral oil cooling(40
C)
LUBRIFIN MET 1 R Tip II
Draw back
Running
maintenance
Cl 850
C/40 minutes Mechanical agitation - -

Cl +R 850
C/40 minutes Mechanical agitation 150
C 2 h
Mg 850
C/40 minutes Continuous magneticfield300 G - -

Mg +R 850
C/40 minutes Continuous magneticfield300 G 150
C 2 h
US 850
C/40 minutes Ultrasonicfield40,4 kHz and 4W/dm

2
- -
US+R 850
C/40 minutes Ultrasonicfield40,4 kHz and 4W/dm

2
150
C 2 h
For magnetic field quenching was used an oil
tank with 40 dm
3
capacity, fiberglass located in the
center of a DC powered coils. In the center of the
basin is done a magnetic field strength of 300
Gauss, [1].
Ultrasonic field quenching was made in a tank
with a capacity of 40 dm
3
, stainless steel sheet, in
which at the bottom was placed a piezoelectric
transducer with U.S. power of 200 W and the
resonant frequency of 40.4 kHz. Power transducer
was made from an electronic generator, developed
in the laboratory, capable of ensuring a U.S. field
strength of 4 W/dm
2
, [1].
Quality assessment was done by measuring the
heat treatment hardness.
Figures 2.a. and 2.b. are the results of hardness
measurements and their statistics. From their
analysis results the superiority of U.S. quench field
over other hardening technologies experienced.
Measurement of dimensional stability in time
was made in a period of 204 hours. After this time
installation was found to stabilize the specimens
dimensional. Measurement results are presented in
Table 2 and chart in Figure 3.
a
b
Figure 2. Results of hardness measurements for heat-
treated groups: a) hardness values chart; b) statistical
measurements.
225
Tabel 2. Dimensional sability over time for different variants of heat treatment [m]
Time[h] Cl Cl.+R. US US+R. Mg Mg+R
0 1,33 1,33 0,66 0,66 1,33 1,33
12 2,66 2,66 0,66 2,66 2,66 4,00
24 4,00 2,66 0,66 2,66 2,66 4,00
36 4,00 4,00 1,33 3,33 4,00 5,33
48 5,33 5,33 1,33 4,00 4,00 6,66
60 5,33 6,66 2,66 4,00 4,00 8,00
72 6,66 6,66 4,00 5,33 5,33 8,00
84 6,66 6,66 5,33 5,33 5,33 8,00
96 6,66 6,66 5,33 6,66 6,66 8,00
108 8,00 8,00 6,66 6,66 6,66 8,00
120 9,33 9,33 6,66 6,66 6,66 9,33
132 10,66 9,33 8,00 8,00 8,00 9,33
144 10,66 10,66 9,33 8,00 9,33 9,33
156 12,00 10,66 10,66 8,00 9,33 9,33
168 13,33 12,00 11,33 8,00 10,66 10,66
180 13,33 13,33 12,00 8,00 10,66 10,66
192 13,33 13,33 12,00 8,00 10,66 10,66
204 13,33 13,33 12,00 8,00 10,66 10,66
Figure 3. Variation of dimensional changes in time of the specimens heat treated [m].
4. Conclusions
1. Quenchig technologies significantly
influence the structure and properties of bearing
steel. The amount of residual austenite in the
structure, characteristics and size of residual stress
of martensite is directly influenced by the
parameters of the cooling operation.
Metallographic structural analysis for various
quenching regimes tested (Figure 4) is clearly
observed differences between the quantity and
distribution phases.
226
a
b
c
Figure 4. Metallographic structure of quenched
samples: a) classic quenching b) magnetic field
quenching, c) U.S. quenching field. MO x 400
2. Activating U.S. cooling environmental
drastically reduce calefaction on cooling and
uniform heat exchange intensity in the first part of
cooling, as confirmed by less dispersion of
hardness values, as apparent from Figure 2.b.
Furthermore, the U.S. vibration frequency of the
fluid cooling determines the cooling capacity
growth, and hence the cooling rate directly affects
the amount of residual austenite in the structure.
Compared with other types of heat treatment, U.S.
quench ensure sensitive reduce of the residual
austenite amount. (Figure 4.c.).
3. Mineral oil used as a cooling environment
can be considered incompressible and ensure
transfer of mechanical energy from the transducer
to the product. Additional energy intake
determines the transformation of additional
austenite quantities into martensite at quench with
direct implications on the content of residual
austenite, [2, 3].
4. Following the additional energy intake, even
during the quench, product stress relief is realized,
phenomenon confirmed by experimental research,
[1].
5. Dimensional stability over time is directly
influenced by the proportion of phases and
constituents in the metallographic structure. In
bearing steel case is mainly on the amount of
residual austenite in the structure, [1]. Among
variants examined by heat treatment, the U.S.
quenching ensures the most dimensional stability
in time by dramatically reducing the amount of
residual austenite in the structure in the first place.
6. As an alternative technology at industrial
scale, the U.S. quenching has the advantage that it
not requires major investment or significant
changes in heat treatment lines or excessive energy
consumption.
7. U.S. heat treatment oil activation reduces the
negative effect of waste water and increases its
use.
References
[1] Bancescu, N., Studii si cercetri privind
cresterea durabilittii contactului cu rostogolire
prin tratamente termice neconventionale
aplicate tribostratului, Tez de doctorat,
Universitatea Dunrea de jos Galai, 1995.
[2] Saban, R., Dumitrescu, C., Petrescu, M., Tratat
de Stiinta si ingineria materialelor, Academia de
Stiinte Tehnice din Romnia, Editura AGIR,
Bucuresti 2006.
[3] Amza, Gh., Ultrasunetele. Aplicatii active,
Academia de Stiinte Tehnice din Romnia,
Editura AGIR, Bucuresti 2006.
227
I NFLUENCE OF POLYMER CONCENTRATI ON ON THE
PERMEATI ON PROPERTI ES OF NANOFI LTRATI ON MEMBRANES
Balta Stefan
1,2
, Bodor Marius
1
, Benea Lidia
1,2
1
Universitatea Dunrea de J os din Galai, Facultatea de Metalurgie i tiina Materialelor,
2
Centrul de Cercetare Interfee Tribocoroziune i Sisteme Electrochimice, Romania.
E-mail: Stefan.Balta@ugal.ro
Abstract: Driving force membrane processes seem to be most useful for water treatment. Membranes are
very effective in removing a wide variety of water contaminants. Therefore, the use of these processes in water
purification to replace or to improve conventional treatment has increased. An inherent problem of membranes is
fouling, the accumulation of materials (foulants) near, on, or within the membrane that causes a reduction in the
amount of product water over time. As a result of fouling, capital and operating costs of membrane systems are
higher, making them less attractive. A membrane is an interphase between two adjacent phases acting as a
selective barrier, regulating the transport of substances between the two compartments. The main advantages of
membrane technology as compared with other unit operations in (bio) chemical engineering are related to this
unique separation principle, i.e. the transport selectivity of the membrane. Separations with membranes do not
require additives, and they can be performed isothermally at low temperatues and - compared to other thermal
separation processes - at low energy consumption. Nanofiltration (NF) separate or remove small molecules or
ions from a solvent (most often water) by means of pressure driven filtration through a dense polymeric
membrane. Combination of selectivity with a high permeability to water and mechanical strength sufficient to
withstand high pressures is achieved by using thin film composite membranes comprising a dense film of 10-200
nm (active layer) supported by a thick asymmetric porous film. In this paper is describing the manufacturing
processes of Polyethersulfone membranes (PES). A Polyethersulfone membranes was made with different
concentration of polymer in N-Methyl-pyrrolidone (NMP) solvents. The influences of the polymer concentration
on the membranes permeation properties were studied. After the preparations all membranes were studied for a
comparison with cross flow and dead-end equipments to see the flux and permeability of pure water. The
permeation results and the SEM photography show the influence of the polymer concentration, increasing
concentration permeation properties are decreasing.
Keywords: membrane, nanofiltration, PES
1. Introduction
Nanofiltration (NF) is widely applied in the
treatment of waste water and in the production of
drinking water [1,2] because provide a feasible
process allowing a high retention of multivalent
ions as well as organic molecules. However, one of
the main drawbacks of the nanofiltration
performance is the fouling phenomenon, usually
attributed to adsorption of organic substances on
the membrane surface. Membrane fouling lead in
diminished the membrane performance, serious
deficient production, and excessive operating costs
[6,7]. Because of the fouling, the dyes rejection [8-
10] and the permeation properties [11-13] of the
membranes decrease due to a higher
hidrophobicity of the membrane surface.
Membrane fouling depends by the membrane
characteristics [14-17] and by the filtration mode
(cross-flow or dead-end filtration) [18].
In order to increase the effectiveness of
nanofiltration membranes, some properties such as
hydrophilicity and fouling resistant should be
improved. The membranes were synthesized at
four different polymer concentrations 25, 27, 30
and 32 wt.% by the phase inversion method. This
method means that after dissolution of the polymer
in a solvent, the polymer solution has to be cast to
a thin film with different thickness on a support
layer. In our case the thickness of the polymer
layer was 250 m. The support layer, with the thin
polymer film on it, is then immersed in a non-
solvent bath, deionized water. Due to the diffusion
of the non-solvent in the polymer film, the polymer
solution becomes thermodynamically unstable,
resulting in two phases: a polymer-poor phase (the
228
pores of the membrane) and a polymer-rich phase
(the matrix).To make a good comparison of
membrane characteristics some properties like
hydrophobicity, permeability and morphology
were studied.
2. Experimental
2.1. Materials
The solvent used was 1-Methyl-2-pyrrolidone
and he support layer (type Viledon FO2471) used
for the membrane manufacturing was obtained
from Freudenberg (Weinheim, Germany). The
polymer, Polyethersulfone type Radel, was
supplied by Solvay (Belgium) and was used as the
base polymer. To determine the membranes flux
and the permeability was used distillated water.
2.2. Membrane preparation
NET polymeric membranes were manufactured at
four different concentration of polymer (25, 27, 30
and 32 wt.%) in NMP, using the phase inversion
induced by immersion precipitation method.
Preliminary experiments made by others
researchers showed that the membrane with 30 and
32wt.% of PES are the most suitable
concentrations to obtain NF membranes.
The casting solution was obtaining adding
polymer in the solvent solution and mixed at 40
0
C
on the mechanical stirring at 200 rpm for 24 hours.
On a polyester support a thin film of the polymer
solution with a thickness of 250 m was cast with
a filmograph (K4340 Automatic Film Applicator,
Elcometer). Membrane was immersed in distillate
water for precipitation and after 15 minutes was
washed to remove the excess of solvent. For every
type of membrane, four different solutions were
made and from every solution three membranes
was manufactured and tested to obtain the true
values of the membranes properties.
2.3. Filtration experiment
To study the performance of the membranes,
permeability and flux, were used a dead-end (Fig.
1a.) and cross flow filtration (Fig. 1b.)
installations.
Figure 1. Filtration equipment: a) dead-end and b) cross flow
The pure water flux experiments were carried
out with a commercial cross flow unit on
laboratory scale. The permeability of the prepared
membranes were studied using two dead end
modules Sterlitech HP4750 at the room
temperature and desirable pressure. The pressure
was realized with a nitrogen cylinder and a
pressure regulator, connected to the dead-end cell.
The solution volume used for every experiment
was 250 ml and the permeate was collected in a
graduate cylinder. The pure water flux was
determined at 10 bar pressure and the time was
measured at every 5 ml of permeate.
To determine the pure water permeability
(PWP) was measured the water flux (J
w
) at six
different pressure (P) from 5 to 20 bar. ThePWP
was calculated by the following equation:
P
J
PWP
w
1)
To determine the pure water flux and
membranes behavior for a long time all the
membranes was tested in a cross flow installation.
All experiments were realized at 24
o
C and the
applied pressure was 8 bar. The membranes
229
surface area was 0.0059m
2
and the time for every
experiment was 24 hours.
2.4. Characterization of the membrane
surface and morphology
To study the hydrophilicity/hydrophobicity of
the membranes was used a Drop Shape Analysis
System DSA 10 Mk2 (fig. 2b.). On the cleaned and
dry membrane surface was placed a distillate water
droplet of 2 l and the contact angle between the
membrane surface and the droplet was calculated
(fig. 2a). The final value o the contact angle for
every type of membranes was the average of 21
measurements, seven determinations for three
different membranes.
Figure 2. Contact angle measurements: (a) the
principle and (b) the setup
Scanning electron microscopy (SEM)
measurements were performed for characterization
of the surface and cross-section of the membranes.
For the cross-section analysis the samples was
prepared by fracturing the membranes in liquid
nitrogen and sputtered with gold. The images were
made with a Philips XL30 FEG and Phylips FEI,
QUANTA 200 instruments. . Surface SEM images
were made with a Phylips FEI, QUANTA 200
instrument with an accelerating voltage of 20 KeV
3. Results
3.1. Pure water flux and permeability
For the determination of pure water flux the
filtration experiments were carried out with a
commercial nanofiltration unit on laboratory scale.
In all experiments the applied pressure was 8 bar
and the temperature was 24
0
C. To minimize
concentration polarization a feed velocity of 4.0
m/s was used. The membrane surface area was
0.0059m
2
. The evolution of flux was followed in
time during 24 h. For analysis and comparison the
values after 24 h of filtration were used.
Figure 3 shows that the permeation flux of PES
membranes. The polymer concentration has an
important influence on the pure water flux. When
the polymer concentration decrease the pure water
flux increase. The best membranes in term of water
flux are at 25 wt% of PES.
0
200
400
600
25 27 30 32
F
l
u
x

[
J
(
L
/
m
2
h
)
]
PES concentration (wt.%)
Figure 3. Pure water flux at different concentration of
PES
Membranes with 25% of PES have a good
permeability but because of the weaker mechanical
resistance have an important instability of flux in
time (figure 4). Hence, the increase of polymeric
material enhances the membrane mechanical
resistance.
Figura 4. Pure water flux for different PES
concentration
0
5
10
15
20
25
30
35
25 27 30 32
P
e
r
m
e
a
b
i
l
i
t
y
(
L
m
-
2
H
-
1
b
a
r
1
)
PES concentration (wt.%)
Figure 5. Permeability at different PES concentration.
Increasing the concentration of PES the
membranes pore size decrease and in consequence
permeability decrease. Figure 5 confirm the same
influence of the polymer concentration on the
permeability like in the case of pure water flux.
0.00
100.00
200.00
300.00
400.00
500.00
600.00
700.00
5
1
0
1
5
2
0
2
5
3
0
3
5
4
0
4
5
5
0
7
0
7
5
8
0
8
5
9
0
9
5
1
0
0
Volume [ ml]
F
l
u
x

[
J
(
L
m
2
h
)
]
32NMP
30NMP
27NMP
25NMP
230
Membrane at 25 wt.% of PES have the
higher permeability but because of them instability
in time are not the ideal membranes to be selected
for future experiments. Membranes a 27wt.% of
PES appear to be the best membranes for
nanofiltration experiment.
3.2. Contact angle
Contact angle determination is a well-known
method to study the membrane surface
hydrophobicity, since a hydrophilic membrane
surface gives rise to a low contact angle [16].
Figure 6 shows the measured contact angles for
neat PES membranes for four different polymer
concentrations, indicating that membrane
hydrophilicity increases as decreases the polymer
concentration. The effect of polymer concentration
on membrane hydrophilicity should be explained
in terms of pore size and porosity considering.
Results are in concordance with permeation
properties. Membrane with 25 wt.% of PES have
the most hydrophilic surface.
Figure 6. Contact angle at different PES concentration
3.3. Membrane morphology characterization
The permeation properties of neat membranes
can be better explained by the Scanning Electronic
Microscopy (SEM) analyses. Figure 7 presents
SEM images of the cross-sections of 25, 27, 30 and
32 wt.% PES. From the SEM images is observed
that polymer concentration has a clear effect on the
membrane structure, which can be described in
terms of membrane pore size and porosity
variations
Figure 7. Cross section SEM photography of membranes at different PES concentration: a) 25%PES, b)27%PES,
c)30%PES, d)32PES
Membranes with 25% of PES (figure 7 a) have
macrovoids in the structure who lead to a better
permeability but, how was observed at the pure
water flux, with an instability of flux in time. The
pore size and geometry change as the polymer
concentration increase, figure 7 a,b,c,d for 25, 27,
30 and respectively 32 wt.% of PES, in the same
time suppress the macrovoid and increase the
thickness of the top layer [17]. For 30 and 32 wt.%
of PES the permeability and pure water flux is
smaller because the bottom with the top layer are
not connected by macrovoids and a sponge
structure is formed. For membrane with 27 wt.% of
PES the bottom is connected with the top layer,
porosity is uniform distributed and the pores
structure are like fingers.
231
Figure 8. Surface SEM photography at different concentration of PES: a) 25, b) 27, c) 30 and d) 32wt.%.
Figure 8 show surface SEM photography of
membranes at different concentration of PES. By
increasing the concentration of PES it was
observed that the porosity of the membranes
decreases. Membranes with 25 wt.% of PES are
the most porous in comparison with membranes at
higher concentration.
4. Conclusions
A systematic study of influence of polymer
concentration was carried out, testing a large
number of membrane samples.
The polymer concentrations have a high
influence on the membranes properties and have a
negative effect on the water permeation and
hydrophobicity. The results from permeation
experiment and from analysis of membrane
morphology show the negative influence of
polymer concentration. At 25 wt.% of PES,
membranes have a good permeability but flux
show instability in time. Because of this behaviour
for industrial application membrane with higher
concentration of PES need to be selected.
Acknowledgements
Stefan Balta would like to acknowledge the
support provided by the European Union,
Romanian Government and Dunarea de J os
University of Galati, through the project POSDRU
6/1.5/S/15.
Bodor Marius would like to acknowledge
the support provided by the European Union,
Romanian Government and Dunrea de Jos
University of Galai, through the project
POSDRU 107/1.5/S/76822.
References
[1]. IC. ESCOBAR, AA. RANDALL, SK.
HONG, Removal by Nanofiltration: Full and
Bench-Scale Evaluation, J Water Sup Res
Technol Aqua 51 (2002) 6776.
[2]. A. MATILAINEN, R. LIIKANEN, M.
NYSTROM, Enhancement of the natural
organic matter removal from drinking water
by nanofiltration, Environ Technol 25 (2004)
28391.
[3]. E. Matthiasson, B. Sivik, Concentration
polarization and fouling, Desalination 35
(1980) 59.
232
[4]. D.E. Potts, R.C. Ahlert, S.S.Wang, A critical
review of fouling of reverse osmosis
membranes, Desalination 36 (1981) 235.
[5]. B. Van der Bruggen, M. Manttari, M.
Nystrom, Drawbacks of applying
nanofiltration and how to avoid them: A
review, Sep Pur Tech 63 (2008) 251263
[6]. B. Van der Bruggen, L. Braeken, C.
Vandecasteele, Flux decline in nanofiltration
due to adsorption of organic compounds,
Sep Pur Tech 29 (2002) 2331
[7]. L D. Nghiem, PJ . Coleman, C. Espendiller,
Mechanisms underlying the effects of
membrane fouling on the nanofiltration of
trace organic contaminants, Desalination 250
(2010) 682-687.
[8]. L. Braeken, R. Ramaekers, Y. Zhang, G.
Maes, B. Van der Bruggen, C.
Vandecasteele Influence of hydrophobicity
on retention in nanofiltration of aqueous
solutions containing organic compounds . J .
Membr. Sci. 252 (2005) 195-203.
[9]. J M. Arsuaga, MJ . Lpez-Muoz, A. Sotto,
Correlation between retention and adsorption
of phenolic compounds in nanofiltration
membranes, Desalination 250 (2010) 829-
832
[10]. L. Braeken L, K. Boussu K, B. Van der
Bruggen B, C. Vandecasteele, Modeling of
the adsorption of organic compounds on
polymeric nanofiltration membranes in
solutions containing two compounds,
Chemphyschem 6 (2005) 1606-1612.
[11]. J .M. Laine, J .P. Hagstrom, M. Clark, M.
Mallevialle, Effects of ultrafiltration
membrane composition, J . Am. Water
Works Assoc. 81 (1989) 6167.
[12]. M. Elimelech, X. Zhu, A.E. Childress, S.
Hong, Role of membrane surface
morphology in colloidal fouling of cellulose
acetate and composite aromatic polyamide
reverse osmosis membranes, J . Membr. Sci.
127 (1997) 101109.
[13]. C. J ucker, M.M. Clark, Adsorption of
aquatic humic substances on hydrophobic
ultrafiltration membranes, J . Membr. Sci. 97
(1994) 3752.
[14]. J . Cho, G. Amy, Interactions between natural
organic matter and membranes: rejection and
fouling, Water Sci. Technol. 40 (9) (1999)
131139.
[15]. D.B. Mosqueda-J imenez, R.M. Narbaitz, T.
Matsuura, Membrane fouling test: apparatus
evaluation, J . Environ. Eng. ASCE 130 (1)
(2004) 9098.
[16]. C.R. DIAS AND M.N. DE PINHO, Water
structure and selective permeation of
cellulose-based membranes, J . Mol. Liquids,
80 (1999) 117132.
[17]. H. Susanto and M. Ulbricht. Polymeric
Membranes for Molecular Separations.
Membrane Operations. Innovative
Separations and Transformations. WILEY-
VCH, Weinheim. 2009.
TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies"
233
STATI C DEFORMATI ON OF A WORKPI ECE FI XED I N
UNI VERSAL CHUCK AND LI FE CENTRE
Laureniu Sltineanu
1
, Lucian Tcaru
2
, Margareta Cotea
3
, Teodor Tilic
4
, Mihai
Boca
5
, and Irina Grigora (Beliu)
6
1, 2, 3, 4, 5
Gheorghe Asachi Technical University of Iai - Romania, slati@tcm.tuiasi.ro,
ltabacaru@tcm.tuiasi.ro, mcoteata@tcm.tuiasi.ro, teodortilica@gmail.com, mihaitzaboca@yahoo.com,
6
University tefan cel Mare of Suceava- Romania, grigorasirina_sv@yahoo.com.
Abstract: The rigidity could be defined as the capacity of a system or subsystem to oppose to the
elastic deformation generated by the action of external forces. In this paper, the authors highlight some
possibilities to evaluate the static rigidity of the technological system corresponding to a turning process.
Some remarks were elaborated by considering the workpiece as a bar fixed at one end and simply
supported at the other end. Experimental tests were developed to measure the static elastic deformation of
the workpiece under the action of a variable loading force.
Keywords: elastic deformation, rigidity, turning process, deflection, applied force
1. Introduction
The rigidity of a solid body is the capacity of
this body to oppose to the deformation generated
by the action of external forces [2, 3, 4]. Of course,
there is the possibility to evaluate the rigidity of an
individual part, but there is also the possibility to
take into consideration the rigidity of a system or
subsystem. In the case of the so-called classical
technological system constituted of the machine
tool tool device workpiece, the rigidity will
refers to its capacity to oppose to the deformation
generated by the cutting process.
There are various methods of evaluating the
rigidity of the technological systems; thus, there
arestatic anddynamic methods. In the case of the
static methods, a way to generate a force similar to
that appearing during the cutting process is used.
As known, in the case of turning process, the main
cutting force can be decomposed in components
along the main axes belonging to a coordinate axis
system attached to the technological system. The
deformations appear for each axis along which a
component of the cutting force acts. These
deformations can be measured and by determining
the ratio between the forceF and the deformationx
generated along the direction of the force, an
image about the static system rigidity along the
axis Ox is obtained. In some countries, there are
standardized methods for the static evaluation of
the technological system; some recommendations
concerning the static evaluation of the lathe are
included in an old Romanian standard.
On the other hand, there are methods in which
the deformation generated by a real cutting process
are determined; because there are some difficulties
in measuring the deformation during the cutting
process, the experimental researches were directed
to the measuring the dimensions of the machined
surfaces. By taking into consideration the results of
the measuring the dimensions of the machined
surfaces, a certain mathematical processing of
these results allows the evaluation of the
technological system rigidity in dynamic
conditions.
Information concerning the rigidity of the
technological system are important when some
precise machining processes must to be applied;
this means that during the design of the cutting
process, the establishing of the depth of cut a
p
has
to take into consideration the deformation of the
workpiece, so that finally the dimensions of the
machined surfaces be inscribed in a narrow
tolerances field.
Some decades ago, Korsakov developed a
thorough theoretical study concerning the elastic
deflection of the technological system in the case
of the turning process; various possibilities of
fixing the workpiece were taken into consideration
[2]. J ianliang and Rongdi [1] proposed a united
model for predicting diametral error of slender bar,
by applying a finite analysis to highlight the
workpiece deflection, when a follower rest is used
234
during a turning process. They concluded that the
diametral error mainly depends on the location of
follower rest, depth of cut and feed rate.
Having the intention to more detailed study the
influence exerted by the technological system
rigidity on the machining accuracy, this paper
presents some considerations concerning the static
evaluation of the technological system in the case
of a turning process.
2. I nitial hypothesizes
One of the machine tools existing in the
majority of the mechanical workshops is the lathe;
essentially, the turning supposes a main rotation
motion of the workpiece, while the tool
materializes a feed motion in the axial plane of the
workpiece. In such a situation, the turning
accuracy depends on the accuracy of the motions
performed by the workpiece and the tool in relation
with a fix coordinates system; in our case, such a
coordinates system could be defined by
considering the rotation axis of the workpiece as
Oz axis (fig. 1, a).
There are three ways to clamp the workpiece.
Thus, although it does not ensure a high alignment
a
b
c
d
Figure 1: Workpiece deformation as a bar rigid fixed at one end and simply supported at the other end
235
of the machined surfaces when the workpiece must
be twice clamped, the positioning of the workpiece
in the universal chuck and the life centre is the
most frequently used in the case of the long
workpieces. A better accuracy from the point of
view of alignment when two positions of the
workpiece are necessary and the workpiece is long
enough corresponds to the positioning between life
centres; for such reasons, the workpiece clamping
between the life centres is preferred within the
finishing cutting operations (turning, grinding).
The both above mentioned ways of the workpiece
positioning suppose the existence of the centre
holes on the frontal surfaces of the workpieces.
The third way to clamp the workpiece, applicable
in the case of short workpieces is by means only of
the universal chuck; obviously, if two clampings
are necessary, there is a high probability that the
alignment of the machined surfaces at the two
clampings not be high enough.
Within this paper, only the case of the
workpiece clamping in the universal chuck and life
centre will be considered, this being widely used
way in industrial practice, although, as above
mentioned, this way does not ensure a high
accuracy from the point of view of the alignment
of the surfaces obtained by two clampings.
During the turning process, the workpiece and
the other components of the technological system
(machine tool, devices, tools) are affected by
deformation phenomena; the presence of the
cutting forces determines the deformation of the
above mentioned mechanical components. Of
course, only elastic deformation could be accepted;
the operation conditions must be established so
that the internal stresses to not exceed the limit of
elastic deformation of the material from which the
components of the technological system are made.
Due to this elastic deformation of the
technological system components, the size of the
actual depth of cut is not the same with the desired
one; under the action of the cutting forces, the
components are affected by elastic deformations
and the current depth of cut a
pc
is lower that the
designed depth of cut a
pd
(fig. 2):
pd pc
a a
(1)
This relation could suggest a way of
increasing the machining accuracy; if the
differences between the current depth of cut a
pc
and the desired depth of cut a
p
are known, the
designed depth of cut could be so established that
even by taking into consideration the elastic
deformation of the technological system
components, an increased machining accuracy be
obtained. A way used to obtain information
concerning the elastic deformation of the
technological system components is based on the
study of the static rigidity which characterize the
technological system. It is expected that the bigger
the technological system rigidity is, the higher the
machining accuracy is.
The situation of the workpiece fixed in the
universal chuck and the life centre (fig. 1, a) could
be considered as similar to a bar rigid fixed at one
end and simply supported at the other end (fig. 1,
b). A coordinate systemxOyz could be taken into
consideration; the origin of this coordinate system
could be placed at the intersection of the rotation
axis with a frontal plane corresponding to the free
surface of the jaws of the universal chuck; this is
the case of the turning an external cylindrical
surface on a lathe, by using the mechanical
longitudinal feed.
The cutting force generated by the turning
process can be decomposed in components
directed along the axes corresponding to the
coordinate systemxOyz (fig. 3).
Figure 3: Components of the cutting force which acts
on the lathe tool during the turning process.
Figure 2: Decreasing of the depth of cut during the
elastic deformation during the turning process
.
236
The position of the point where the cutting
force is applied changes during the turning process
simultaneously with the position of the lathe tool
tip; this means that just the deformation of the
technological system changes, by taking into
consideration the position of the lathe tool tip.
As consequence of the turning process, the
cross section of the workpiece is not the same
along the workpiece (the machined surface has a
low diameter in comparison with the not machined
surface), but in order to simplify the theoretical
modeling, one can consider a workpiece having the
same diameter along all its length. In fact, in the
case of the workpieces having big diameters, the
decrease of the diameter as consequence of the
turning process is not significant and the
considering the same diameter along the entire
workpiece does not generate significant errors.
Also in order to simplify the mathematical
model, only the action of the cutting force
component F
x
will be considered; in real turning
conditions, elastic deformations are generated yet
along the axis Oy and Oz, but one can appreciate
that from the point of view of the diametral
accuracy, they exert a less significant influence.
For a bar rigid fixed at one end and simply
supported at the other end, the relation valid for the
elastic radial deflection is given [4] by the relation:

3
2 3
12
4
l EI
z l z l z F
f
z
x

(2)
where F
x
is the size of the cutting force component
developed along a direction parallel to theOx axis,
z is the distance along the Oz axis from the
coordinate system origin to the tool lathe tip, l - the
total length of the bar, E - the bar material
elasticity modulus andI
z
the second moment area
of the cross section corresponding to the bar
workpiece.
The elastic deformation is not the same along
the workpiece; just the different fixing of the
workpiece at the two end (considered as a bar rigid
fixed at one end and simply supported at the other
end) determine the changing of the elastic
deflection along the workpiece. On the other hand,
just the universal together with the main shaft of
the lathe are affected by an elastic deformation;
this means that the points where the workpiece is
fixed at the two ends have not a fix position. This
means that if one tries to measure the elastic
deformation of the workpiece near the universal
chuck, this deformation will be different in
comparison with the elastic deformation generated
near the life centre (fig. 1, c andd).
3 Experimental testing
In order to verify the validity of some of the
previous mentioned considerations, an
experimental testing was developed [5]. A
universal lathe type SNA 500x1000 was used
(workpieces having diameter up to 500 mm and a
length up to 1000 mm can be machined on such a
lathe); a workpiece having the diameter of 45 mm
and a length of 817 mm was clamped in the
universal chuck and the life centre (fig. 4).
Several dial gauges were placed in order to
obtain an image concerning the elastic deformation
of the workpiece (fig. 5), of the lathe main shaft in
the position corresponding to the universal chuck
and of the life centre. We tried to place the holder
of the dial gauges on the lathe bed. An elastic
deflection could also affect the lathe bed, when a
force is exerted on the technological system, but
due to the high rigidity of the lathe bed, this elastic
deformation is very low and one can consider that
this deformation can be neglected.
Figure 4: Measuring the elastic deformation of the technological systemcomponents
237
The experiments were developed by applying
an increasing force F
x
in various points along the
workpiece fixed in the universal chuck and live
centre; some of the experimental results were
included in the table 1. As one can see, in the
columns no. 3, 4, 5, 6, 7, 8 there are the sizes of the
deformations corresponding to the universal chuck,
to the life centre and to the tool holder; the
deformation registered by the dial gauges used to
measure the deformations of the test piece in four
established places were inscribed in the columns
no. 9, 10, 11, 12, 13, 14, 15 and 16. For each case,
there were measured the deformations both during
the system loading (when the force increases) and
unloading (during the decrease of the force size).
In order to obtain an image about the rigidity of
the technological system at the level of the
universal chuck and the live centre, the force
exerted on the test piece can be devised in two
components which act on the universal chuck and
on the live centre, by taking into consideration the
distances among the point of applying the force
and the points where the test piece is supported.
The moment equations allows the determining the
component F
x1
=102 daN which acts on the
universal chuck and F
x2
=138 daN (this component
acts on the live centre) when the force applied to
the workpiece has the sizeF
x
=240 daN.
Knowing the maximum sizes of the force F
x
components and of the elastic displacements x
(from table 1) affecting the universal chuck
(x
1max
=0.0242 mm) and the live centre (x
2mx
=0.065
mm), the system rigidity at the level of these
subassemblies could be estimated:
x
F
R
x
x

(3)
By means of the relation (3), one obtain
R
x1
=4214 daN/mm for the universal chuck and
R
x2
=2123 daN/mm for the life centre. Taking into
consideration the maximum elastic displacement of
Table 1: Elastic deformations during the increasing and decreasing of the radial force
applied at the distancez=470 mm
Exp.
No.
Force
F
x
,
daN
Elastic deformations
Universal
chuck
Life centre Tool holder z=70 mm z=240 mm z=470 mm z=770 mm
Loa-
ding
Un-
loa-
ding
Loa-
ding
Un-
loa-
ding
Loa-
ding
Unloa-
ding
Loa-
ding
Unloa-
ding
Loa-
ding
Unloa-
ding
Loa-
ding
Unloa-
ding
Loa-ding Unloa-
ding
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1 0 0 0 0 0.0195 0.00 0.000 0 0.003 0 0.00 0.000 0.000
0.0000 0.0076
2 20 0.0005 0.003 0.002 0.026 0.00 0.033 0.010 0.0135 0.025 0.030 0.025 0.040
0.0203 0.0267
3 40 0.002 0.0045 0.006 0.033 0.025 0.050 0.020 0.0240 0.050 0.059 0.060 0.075
0.0356 0.0457
4 60 0.004 0.0065 0.010 0.037 0.035 0.058 0.029 0.033 0.075 0.075 0.093 0.110
0.0533 0.0635
5 80 0.006 0.009 0.018 0.040 0.040 0.065 0.039 0.042 0.100 0.110 0.130 0.150
0.0711 0.0826
6 100 0.0082 0.011 0.020 0.045 0.050 0.070 0.039 0.044 0.125 0.140 0.170 0.190
0.0876 0.0991
7 120 0.0105 0.013 0.028 0.050 0.055 0.076 0.040 0.044 0.155 0.165 0.210 0.220
0.1087 0.1168
9 160 0.01550.0175 0.036 0.060 0.065 0.083 0.040 0.044 0.210 0.220 0.290 0.300
0.1448 0.1562
10 180 0.01750.0195 0.044 0.064 0.070 0.087 0.040 0.044 0.235 0.245 0.325 0.340
0.1638 0.1753
11 200 0.0195 0.022 0.051 0.065 0.075 0.089 0.044 0.044 0.250 0.270 0.365 0.380
0.1829 0.1930
12 220 0.022 0.024 0.056 0.065 0.080 0.090 0.044 0.044 0.250 0.295 0.405 0.415
0.2007 0.2083
13 240 0.02420.0242 0.065 0.065 0.090 0.090 0.044 0.044 0.295 0.295 0.440 0.440
0.2172 0.2172
0
0.1
0.2
0.3
0.4
0.5
0 100 200 300 400 500 600 700 800 900
Co o r d i n at e z , m m
D
e
f
r
o
m
a
t
i
o
n
,

m
m
F=100 d aN
F=200 d aN
F=240 d aN
Figure 5: Workpiece deformation under the action of the force
238
the tool holder (x
3max
=0.090 mm) and the
maximumsize of the F
x
force (F
x
=240 daN), the
rigidity of this subsystem could be considered as
being R
x3
=2666 daN/mm. These values highlights
a relatively low rigidity of the considered
subassemblies (it is known that for a machining of
high accuracy, the rigidity must have values close
to 20000 daN/mm) and this fact could be generated
both by the clearances between various
components of the subassemblies and by the wear
of the same subassemblies. An image concerning
the elastic displacements of the universal chuck,
life centre and tool holder is presented in figure 6,
where the mechanical hysteresis could be also
remarked.
4 Conclusions
The study of the technological system rigidity is
important due to its influence exerted on the
machining accuracy; it is expected that a bigger
rigidity of the technological system allows
obtaining a higher machining accuracy. A
workpiece fixed in the universal chuck and the life
centre could be considered as a bar fixed at one
end and simply supported at the other end; for such
a case, the elastic deflection of the workpiece
could be estimated by means of mathematical
relations used in the strength of materials. In such a
case (workpiece fixed in the universal chuck and
the life centre), the rigidity is not the same along
the workpiece. Theoretical considerations and
experimental results showed that the rigidity is
higher near the universal chuck and lower near the
life centre. In order to measure the elastic
displacements of the universal chuck and life
centre and the elastic deformation of the test piece,
a set of dial gauges were used. The experimental
research proved also the presence of the
mechanical hysteresis phenomenon.
References
[1] J ianliang, G. and Rongdi, H., A united model
of diametral error in slender bar turning with
follower rest, International J ournal of
Machine Tools & Manufacture, 46, 2006,
1002-1012.
[2] Korsakov, V. S., Tochinosti mekhanicheskogo
obrabotka, Mashinostroenie, Moskva, 1961.
[3] Militaru, C., Fiabilitatea i precizia n
tehnologia construciilor de maini, Editura
Tehnic, Bucurei, 1987.
[4] Pico, C., Coman, G.. and Pruteanu, O.,
Tehnologia construciei mainilor-unelte.
Volumul I, Institutul Politehnic Iai, 1970.
[5] Sltineanu, L., Cotea, M., Beliu, I., Boca,
M. and Pop, N., Consideration concerning the
elastic deformation of the workpiece at
turning, Proceedings of the 14th International
Conference Modern Technologies, Quality and
Innovation, Iai - Chiinu - Belgrad,
ModTech 2010, New Face of T.M.C.R., 20-22
May 2010, Slnic Moldova, Romnia, 539-
542.
-0,02
0
0,02
0,04
0,06
0,08
0,1
0 100 200 300
D
e
f
o
r
m
a
t
i
o
n
,

m
m
Fx
universal chuck loading
universal chuck unloading
life centre loading
life centre unloading
tool holder loading
tool holder unloading
Figure 6: Elastic displacement of some subassemblies of the technological system
239
RESEARCH REGARDI NG THE USE OF QUALI TY TECHNI QUES AND
I NSTRUMENTS I N VI EW OF MAI NTAI NI NG AND IMPROVI NG
QUALI TY MANAGEMENT SYSTEMS
Liliana Georgeta Popescu
1
, Mihai Victor Zerbes
1
, Marilena Blaj
1
, Radu Vasile Pascu
1
,
Roland Blaj
2
1
Lucian Blaga University of Sibiu, lilianaescu@ulbsibiu.ro
2
University of Agricultural Science and Veterinary Medicine of Cluj-Napoca,blajroland@yahoo.com
Abstract: In Romanian industry there is still a lot of think to be done in the field of maintaining and
improving quality management systems (QMS). It is imperative to design a strategy for relating with the
employees as well as the managers of the organizations in this field, aimed to make them better and easier
understand the concept of QMS. The present paper sets forth as a starting point for this strategy the use of
quality techniques and instruments (QTI), especially the 5S technique which is a good premise for the
maintenance and improvement of QMS. The 5S technique is a structured method for the maintenance and
improvement of working organization and standardization.
Keywords: inspection, cleaning, progress, 5s technique
1. 5s Technique
The ISO 9000 set of standards proves helpful in
the creation and implementation of a reliable
quality management system, as it sets forth the
fundamental principles requirements, glossary that
make up the system as well as the guidelines for
quality improvement. Mention should be made that
these standards tell only what to do not how to
do which shall be decided by each organization.
Customer requirements concerning the quality
of products and services entail the use of high-
performing numerical command machine-tools etc.
as well as highly qualified personnel. Thus a
quality management system is essential for
industry organizations to understand, coordinate
and correlate various processes necessary for
accomplishing high quality products and services
that gain a competition edge on the market.
Furthermore, an adequately implemented,
maintained and improved system shall unfailingly
lead to efficiency and efficacy, thus to
organizational progress.
The use of the 5S technique represents a step
forward to the improvement of such a system. 5S
stands for the initial letters of five Japanese words
that define an effective organization of the working
place and standardized working procedures. 5S is
more than mere administration, it differentiates
between a common and an outstanding
organization. The 5S technique represents a
structured method for the improvement of work
place organization and standardization. A well-
organized work place can motivate all employees.
5S highlights security, output, efficiency and it
also creates a feeling of propriety belonging.
The golden ruleof the 5S technique is that the
employee should be able to identify anything
within his work place in less than 30 seconds and,
respectively, in less than 5 minutes for anything
that transgresses the boundaries of his work place,
without any prior talk to a colleague, opening a
book or computer searching.
5S is a method whose results can be noticed in
a well organized and ordered work place, an
environment providing a place for anything is in
its place, whenever you need something. The
work place shall be clean, ordered and safe, thus
the employees become motivated and committed.
This is a simple concept. However it requires long-
standing commitment and hard work to attain it.
The 5S technique requires the coordinated
unfolding of the following five categories of
activities:
Sorting: removing everything that is not necessary
at the work place: waste material, semi-products,
unnecessary tools and equipment, scraps or
obsolete documents;
240
Figure 1: Activity 1 - Sorting
Setting in order: arranging the useful objects
subsequent to the previous operation and preparing
them to be used at any further moment;
Figure 2: Activity 2 Setting in order
Cleanliness: cleaning the entire work place
including the objects within the working area
Figure 3: Activity 3 Cleanliness
Standardizing: setting clear rules for maintaining
a perfect hygiene and pleasant atmosphere at the
work place;
Figure 4: Activity 4 Standardizing
Self-discipline: accurate observance of the
established working procedures and assimilation
by all employees of the proper method of
performing previous operations of sorting,
arranging, cleaning and standardizing and thus be
able to apply them from their first working day;
Figure 5: Activity 5 Self-discipline
2. Steps to Successful I mplementation of the
5S Technique
a. Everyone has to be involved.
It is essential to understand that the
implementation of the 5S technique can only be
achieved by a team. Everyone is accountable for
the 5S implementation not only a few individuals,
it is a process involving all individuals. Al
241
managers should take part in the decision-making
process in view of a better functioning of the 5S.
b. Managerial approval
The 5S technique shall not be performed
secretly or surreptitiously as overtime activity.
Instead, the 5S technique has to be approved by the
organization`s administration. Monthly meetings
may be organized where the company
representatives and managers may suggest more
topics.
c. The general manager has the final responsibility
The 5S technique shall be taken seriously and
completely assumed only when the managers and
general manager himself are responsible and fully
committed. There is nothing worse for the 5S
technique than the managers transferring
responsibility of project implementation to the
employees.
d. The 5S technique should be properly understood
and assimilated
People should no longer ask questions such as:
Why do we have to stick red labels to objects? or
Is the 5S implementation really necessary before
any improvement? Therefore 5S meeting shall be
organized in order to explain the concept and
answer the participants questions.
e. The manager should inspect the work place
The manager should personally inspect work
places and desks in order to clarify positive and
negative aspects. The manager should point out
elements subject to improvement and bring them
up at 5S meetings.
f. Do not leave things half-done!
Once initiated, the 5S process has to be
continued until completion. All participants to this
process have to be involved. As soon as the
principles have been established and atmosphere of
discipline and order has to be maintained.
The implementation of 5S shall probably
increase productivity, efficiency and thus an
atmosphere of professionalism shall be maintained
and encouraged. Developing employee attitudes
and discipline is more important rather than the
physical reorganization of the work place.
3. Stages in promoting the 5S
3.1 Mandatory activities:
- setting up the work team;
- training the work team;
-identifying work areas subject to 5S
implementation;
- organizing the four stages of 5S;
- training all employees to observe the 5S;
- 5S auditing;
Internal audits are necessary for the
improvement and regular reporting on the status of
5S implementation and maintenance
2.2 Stages in the 5S audit accomplishment
- selecting the work area;
- setting up the work team - in view of
implementing the 5S method, the pilot project
team members shall coopt individuals who perform
their professional activity within the work area (i.e.
foreman, adjuster, worker, etc.);
- training the work team members;
- assessment of the work area;
i). accurate study of the area and collecting data in
keeping with the Quotation column (table 1);
Table 1: Sample of audit form
ii). each member of the work team shall assign a
qualification (grade) for each assessment criterion
in the audit form;
iii). Summary of data collected from the quotation
forms and establishing the average quotation level;
iv) Identifying the objective (quotation level to be
reached for the ongoing year), filling in the 5S
indicator graph during the brainstorming session
and posting the results in the respective area
(figure. 6);
Figure 6: Sample formfor filling in graph
242
v) Drawing up the improvement plan and
thoroughly observing those criteria with poor
grading (table 2);
Table 2: Sample for the action plan
vi) Tracking the status of improvement plan by the
steering team;
vii) Resuming application of the method until
reaching the established level;
viii) Establishing a higher level;
ix) Permanent application of the method.
3 Conclusions
Thorough and continuous implementation and
application of the 5S technique may enable you
to design the following objectives:
- zero preparation time;
- zero loss;
- zero defects;
- zero delays;
- zero labour accidents;
Here are the advantages of 5S application:
a. Diminishing costs by means of:
- store decreasing
- preventing object misplacement;
- Preventing oil/water/air/energy/material waste
b. Efficiency accomplished by means of:
- effective use of the area
- doing away with useless search and waste
cleaning and inspecting critical points of the
equipment
c. Maintenance of auxiliary equipment and
capacity
d. Quality
e. Occupational health and safety accomplished by
means of:
- removal of dangerous areas;
- improvement of work conditions
- doing away with the causes of labour accidents
and occupational harm;
- Decreasing errors entailed by carelessness
f. Personnel motivation
However, let us not forget that it is imperative
to:
- Standardize habits in view of obtaining better
results;
- Communicate and train in order to attain quality;
- Communication is a two-way process
- Proceed in such a way that each individual should
feel responsible and allow everyone to express
opinions on personal responsibility, as well as to
be able to respond whenever there is any anomaly
related to ones responsibilities. Implementation of
the 5S technique shall lead to process
improvement, creating a positive image to
customers, and increasing organizational
efficiency. It will also enable the employees to feel
better and the organization thrive and become
more competitive on the market.
The method can be used in different domains.
For example we apply successfully 5S technique in
the factory (figure 7 a, b)
Figure7a: Before to implement 5S
into a productive sector
Figure 7b: After implementation 5Sinto a
productive sector
243
Than we apply 5S technique in the office (see
figure 8 a, b).
Figure 8a: Before to implement 5S in office
Figure 8b: After implementation 5S in the office
5S technique could be implemented even on the
drawer of the desk (figure 9 a and 9b),
Figure 9a: Before to implement 5S indrawer
Figure 9b: After implementation 5S in the drawer
and on the computer's desktop (figure 10 a, b)
Figure 10a: Before to implement 5S on desktop
Figure 10b: After implementation 5S on desktop
244
To conclude, implementation and maintenance
of the 5S technique may lead to progress.
Figure 11: The 5 steps
Success may be attained not only by means of
certain applicable standards, instead it is advisable
to study a number of guidelines, brochures, codes,
sciences, standards, and especially methods and
practices successfully employed by others. The
next step is to identify and select the optimum and
useful elements for the particular businesses and
organizations, which should be further promoted
and implemented in an integrated manner as a
management system which, in time, shall lead to
outstanding quality and successful management
and business.
References
[1] Bosanceanu, M., Interdependenta calitatea
productiei-principiul 5S, mmq.ase.ro,
Universitatea Stefan cel Mare Suceava.
[2] Budau, G., Zerbes, M.V., Ispas, M., Cercetari
privind necesitatea i oportunitatea
introducerii sistemului de management al
calitii n organizaiile economice din
domeniul industriei lemnului, Proceedings
ICWSE, Brasov, noiembrie 2003.
[3] Narayanan, L., The 5S Philosophy. A better
Work Environment for Everyone, Available
from, http://www. tpmonline.com.
[4] Kifor, C.V., Oprean, C., Ingineria calitatii,
Editura Universitatii Lucian Blaga Sibiu,
2002.
[5] Imbunatatirea competitivitatii prin
cresterea productivitatii in intreprinderile
din Romania, Bucuresti, 31 mai 2005;
Available from, http://www.lean.ro/LEAN
_imbunatatirea_competitivitatii.pdf.
245
PROPERTI ES OF HARD ALLOYS SI NTERED FROM
METALLI C CARBI DES
Constantinescu Stela
1
1
University Dunarea de J os of Galati, Faculty of Metallurgy , Materials Sciences and Environtment,
stela.constantinescu@email.ro
Abstract: The TiC or TaC alloys, the simple WC-C0 alloys with the same cobalt content are much more resistant
to bending and breaking and feature improved electrical and heat conductivity. The oxidation resistance of the simple
WC-C0 alloys is considerably lower which results in a pregnant tendency for chip welding and low resistance.
Hardness is affected by micro- porosity, granulation of WC phase, purity and composition, extent of homogenization of
the liquid and carbons. The magnetic saturation increases with higher Co contents; it is worth mentioning that bigger
grain alloys have a force considerably lower than that of fine grain ones.
Key words: heat conductivity, hardness, density, porosity, hard alloys
1. I ntoduction
The excessively high sintering temperature
results in a lower density which further negatively
affects the mechanical strength. Due to over-
heating and granulation growth by re-
crystallization an acute decrease in the bending
ultimate strength occurs.
Under or de- carburated materials also
containing fragile phases , feature poor ultimate
strength. The max values taken by the bending
ultimate strength are reached at Co 20 - 25 % and
suddenly decrease with high Co contents. At
theses compositions there are no contact bonds
among the carbons, the carbon crystals being
individual and surrounded by the Co metallic mass.
The alloys mainly used for short chip material
cutting (cast iron, porcelain, etc) are alloys of WC-
Co type, in some cases with small additions of
other carbons. The same compositions but of
different granulometric classes are also used for
pieces exposed to wear ( wire drawing, moulds,
etc).
The higher performance of the WC-Co alloys
in short chip machining applications is accounted
for by its very good heat conductivity which is 2-3
times higher than that of fast steels
The alloys WC-Co, WC-TiC-Co , are mainly
used for short chip material cutting and wear -
resistant piece applications [1]. The hard- alloys
producers can resort to a wide range of possibilities
to achieve variation of the properties of a WC-Co
composition and thus they can adapt their products
to the particular types of tools they may choose to
manufacture [2] .
The present metallography technique allows for a
correct identification and evaluation of hard
sintered carbons structures. The metallography
approach is an indispensable method of
investigation and control in industry.
2. Researches and experimental results
The experimental tests have shown certain
features previously identified as requirements for a
good quality deposit layer, which are: high
compression strength, good impact resistance,
good resistance to high temperatures and thermal
shock, fine surface roughness, pure layer structure
combined with equal grain size, deposition of
materials with continuous change of composition,
regarding the increase of deposit layer (sandwich
layers), deposition of materials impossible to
obtain through other procedures, almost unlimited
opportunity to choose deposition materials, good
adherence between basis material and deposit
layers.
The reaction will be possible
thermodynamically if the calculated concentrations
(partial pressures) of the reactants , under
equilibrium conditions, are less than their original
concentrations
The calculation of the equilibrium
concentrations from the equilibrium constant
involves a good choice of the number of gas spaces
246
which can be higher than two and the number of
independent relations A relation implies the
equilibrium expression depending on the free
standard reaction energy and temperature. The
other relation consists in that the systempressure
is the sum of the partial pressures . If some
reactants possess more than one valence state, the
reaction should contain the reactant under its
most stable valence state.
After making all the calculations the entire
range of deposition parameters (temperature,
pressure, gas initial composition) is obtained.
The phases coming from the vapor which
contain the diffusion element and the carrying
gas pass through three main stages; vapor
formation, transportation and deposition. These
stages differ in terms of chronology and make up a
whole process. According to its essence, such a
process is a chemical transport reaction expressed
by the solid substance interaction (A) with the gas
or vapors (B).
Upon conveying the gas, two processes are
possible: isotherm , without forced flow;
anisotherm, with forced flow
With the latter process, saturation with gas
diffusion by contact-free procedure is reported.
Figure 1: Metallographic appearance of free
carbon in fine-grained phase
When evaluate the structure of the hard
sintered carbons structures by metallography
method, the distinction is made between the
metallographic aspect of the WC-Co, WC-TiC-Co
by WC-TiC-TaC-(Nb)C-Co alloys. Particular
attention is paid to the distribution of the sintering
binder, namely the cobalt, to see the correlation
between the alloy properties and cobalt content
In the case of two alloys having the same
cobalt content but different WC granulations
.it is found that a harder alloys involved finer
granulations while milder alloys implies rough
granulations.
The cobalt, which takes the form of very fine
inclusions with alloys 5% Co and 95% WC, is
much more agglomerated between the WC crystals
along with a slight porosity .
In figure 1 and 2, is illustrated structurally free
carbon metallographic appearance in two situations
compared with the standard scale.
Thus one can see that it always appears at grain
boundaries of carbide WC - Co and that depending
on the carbon surplus in the balance of the alloy
material, the amount of free carbon as a separate
independent phase may be higher or lower.
Figure 2: Metallographic appearance of free
carbonphaseat thegrainaverage
Macro aspect of the fracture allows accurate
observation of the sintered alloy grain size, its
color and luster, the presence or absence of free
carbon and creating an impression that the exact
properties to be expected. Currently simple
macrostructure in fracture assessment allows a
faster control routine in the production of sintered
carbide parts.
An accurate assessment of the microstructure of
the sintered alloy requires an appropriate
metallographic attack. Preparation of samples for
metallographic study is relatively difficult due to
the hardness of the material.
247
Figure 3 illustrates the metallographic structure
of the alloy containing 6% Co and 94 % WC and
average granulation. The basic constituent, WC, is
under recrystallization form, called
2
which
structurally stable crystals under triangle prismatic
shapes of rectangular bases.
Figure 3: The metalographic aspect of the alloy
containing 6% Co and 94 % WC,sol.KOH-K
3
(Fe(CN)
6
)
As to hardness, there is a tight correlation
between the Co content and the WC-Co alloys
produced under identical conditions. As shown in
figure 4, hardness decreases with the increase in
the cobalt content. The sintering temperature and
the exposure time along with the type of grinding
and mixture homogenizing can decisively affect
the sintered metallic carbons alloys[3,4].
Figure 4: Variation of hardness, breaking and
bending resistance and compression resistance of the
alloys WC-Co , influenced by the Co content.
Hardness reaches max values with the optimum
sintering temperature and then it decreases as a
result of carbon recrystallization and alloy super-
sintering. An excessive sintering time, even if the
temperature is optimum, has the same effect ie
lower hardness. The WC-Co alloy density depends
on the Co content and the extent of sintering.
Figure 5 shows the density variation depending
on Co content in the alloy; it implies that the real
measured density takes values within 0, 5 - 3%,
under the theoretical calculated values. This may
be accounted for by the residual porosity which is
the result of a normal sintering process because of
insufficient mixture homogenizing or slight low-
or high carburing which may occur with sintering
[5,6] .
Figure 5: Alloy density vs Co content
As to the density variation, the specific
pressure for pressing the compressed pieces has a
lower influence while the sintering conditions,
such as the sintering temperature and its being kept
at lower values, play a decisive role in determining
density (figure 6). Modern methods such as hot
pressing or hot iso-static pressing, allow for
reaching densities identical to the calculated ones
[7].
The breaking/bending strength of the WC-Co
alloys increases with the cobalt content but the
relation is not lineal [10]. The increase in the
breaking/bending strength takes place also in the
alloys having more than 20% Co provided these
are sintered under special conditions (protected
against carburing) With alloys of 10% Co no
permanent deformations of the material can be
248
noticed during the breaking strength test, while
they are quite obvious with Co contents higher
than 20%.
Figure 6: The effect of the sintering temperature and
time on the WC-Co alloy containing 6% Co density and
hardness.
The compression resistance of the WC-Co
alloys, as shown in figure 4 increases with the Co
content and then significantly decreases after 4%
limit is exceeded .[8,9].
The resilience or impact resistance of the
metallic sintered carbon alloy is a measure of the
mechanical chock resistance. Chock resistance
variation vs. temperature is illustrated in figure 7.
The values of this curve have been determined
on groove-free samples, 6 x 6 mm cross section,
leaning against seats 40 mm distance apart. The
curve shape reaches a peak at the optimum alloy
sintering temperature [11]. The elongation
strength is quite difficult to determine for such a
fragile material like the hard sintered carbon alloy
All the composition WC-Co containing up to 10%
Co show no permanent deformation of the
material.
It can however be noticed elastic deformations,
without plastic deformations, immediately after
breaking alloys containing more than 25 % can
feature a measurable elongation [12,13].
Figure 7: Chock resistance variation vs temperature
on the WC-Co alloy containing 6% Co
The WC -Co alloy elasticity module is
interesting for applications involving elastic strain.
As shown in figure 8, the elasticity module
decreases with the Co content.
Figure 8: Elasticity module variation with Co
content for WC-Co alloy.
As regards the magnetic properties of the WC-
Co alloys, magnetization upon saturation is related
to the Co content (the phase - double Co and W
carbon) and the force acting upon it strongly
depends on sintering extent and the grain size.
Thus the magnetic measurements can be of use in
quality control [14].
The scope of application of WC-Co alloys is a
consequence of their properties. The applications
of the metallic alloys depends on their composition
according to which there are 4 groups as shown
below:
Group I- 97% WC- 3% Co and 95,5% WC
4,5% Co has the following scope of application:
cutting of graphite, ceramics and other metallic
249
materials, grinding, accuracy drilling of cast iron,
non ferrous material machining, drawing
machines .
Group II- 94,5% WC 5,5% Co and 93,5%
WC 6,5% Co is used for:
a) global granulation sorts ( cast iron processing,
non- ferrous materials and alloys, synthetic and
plastic materials, sensors for wear resistance
tools and pieces requiring no high tenacity ,
drawing machines .
b) fine granulation (machining of gray cast iron,
mild cast iron, steels of ultimate strength
higher than 1750 MPa , bronze, Si alloys,
drawing machines .
Group III 91% WC 9% Co , 89% WC
11% Co , 87% WC 13% Co used for machining
wood , synthetic resin , easy machining of steels,
brass and bronze grinding ,plates for agriculture
tools, wear pieces calling for high tenacity.
Group IV 85% WC 15% Co, 80% WC
20% Co , 75% WC 25% Co , 70% WC 30%
Co used for: wear resistant parts calling for high
tenacity ( cutting tools etc.) .
3. Conclusions
Hardness is affected by micro- porosity,
granulation of WC phase , purity and composition ,
extent of homogenization of the liquid and carbons
The excessively high sintering temperature results
in a lower density which further negatively affects
the mechanical strength.
The max values taken by the bending ultimate
strength are reached at Co 20 - 25 % and suddenly
decrease with high Co contents. At theses
compositions there are no contact bonds among the
carbons, the carbon crystals being individual and
surrounded by the Co metallic mass.
Due to overheating and granulation growth by
re-crystallization an acute decrease in the bending
ultimate strength occurs. Under or de-carburated
materials also containing fragile phases , feature
poor ultimate strength .
The compression resistance first increases with
the Co content then considerably decreases after
the Co 4% is exceeded.
The resilience of a Co 6% alloy increases with
temperature and reaches a max value at 1600
o
C.
The elongation vs compression resistance ratio
is estimated as 1: 3 while the elongation
resistance vs ultimate strength ratio is three times
higher that that of steel
The magnetic saturation increases with higher
Co contents; it is worth mentioning that bigger
grain alloys have a force considerably lower than
that of fine grain ones.
4. REFERENCES
[1] Mitoeriu, O., Constantinescu, S., Radu, T.,.a.,
Modern methods to performthe properties of
metal materials, University Dunrea de J os
of Galai , 1998.
[2] Constantinescu S., Nitride coatings on widia
substrate for mechanical applications ,
J ournal of Surface Engineering vol.52,
nr1,p.77-81, ISSN 0267-0844, 2009, USA.
[3] Constantinescu S., Radu T., Modern methods to
performthin layers, Romanian Metallurgical
Foundation Scientific, Publishing House ,
Bucureti , 2003.
[4] Constantinescu, S., Metals properties and
physical control methods, Didact i Pedag.
Publishing House , Bucureti , 2004.
[5] Radu, T., Constantinescu, S., Vlad, M.,
Morphologies of Widmanstatten structures
and mechanismformation in steels , J ournal
of Materials, Science ForumVol. 636-637, pp
550-555, Trans Tech Publications,
Switzerland, ISSN 1662-9752
2010,www.scientific.net
[6] Constantinescu, S., Practical experience of
using chemical vapour deposition coatings to
resist wear, TEHNOMUS New technologies
and products in machine manufacturing
technologies, J ournal no.17, 2010,Ed.
Universitatii Stefan cel Mare din Suceava, p.
84-88, ISSN:1224- 029X, 2010
.www.tehnomusjournal.fim.usv.ro.
[7] Constantinescu, S., Orac, L., Properties and
application of B
4
C coats within the
chemical vapour deposition, METAL 2009, ,
p.96-101, ISBN 978- 80 87294- 03- 1, 19.
21. 5. Hradec nad Moravici , Cehia ,
2009,www.metal2009.com
[8] Constantinescu, S., Orac, L., Characterisation
of niobium carbide coating obtained by
chemical vapour deposition, The Annals of
Dunarea de J os of Galati, Fascicle IX
Metallurgy and Materials Science,No. 1, 2010,
ISSN 1453-083X .
[9] Constantinescu, S.,Orac L.,Wear resistance of
chemical vapour deposition coatings,
4
th
International Conference Welding in
Maritime Engineering , BRAC - SPLIT-
Croatia 13 -16. 05. 2009,
branko.bauer@fsb.hr
250
[10] Constantinescu, S., Nitride coatings on widia
substrate for mechanical applications, J ournal
of SurfaceEngineeringvol.52,nr1,p.77-
81,2009,ISSN0267-0844,
www.ingentaconnect.com
[11] Constantinescu, S., Orac, L., Mechanical
Properties of thin films investigated using
micromachining techniques, Proceedings of
the 18-th International Metallurgical &
Materials,METAL 2010, p.926-931, ISBN:
978-80-87294-17-8, 2010, Ostrava, Czech
Republic, www.metal2010.com
[12] Constantinescu, S., Researches on the effect
of the tempering temperature on both
structure and properties of thick plates made
from steel of 700 N/mm
2
yielding point,
Revista METALURGIA INTERNAIONAL
, nr.6 / 2004 , ISSN 1582- 2214 , p.12-16
,www.metalurgia.ro.
[13] ConstantinescuS., Non-distructive control of
hot rolled products, Intenational
Conference on Materials Science and
Engineering, BRAMAT 2007. Organized by
Transilvania University of Braov and
Romanian Academy of Technical Sciences,
ISSN 1223 9631, 2007,
[14] Constantinescu ,S., Vlad, M., Mitoseriu, O.,
Orac , L., Making thin coatings of bn by
chemical deposition from vapour phase at
normal pressure, al-7-lea Congres
International I SSI M - Iasi, Buletinul
Institutului Politehnic Iasi, Tomul LV(LIX),
fasc.2, 2009, ISSN 1453 1690, p. 151-156.
251
RESEARCH ON THE CORROSI ON BEHAVI OR OF ZI NC COATI NG BY
MEASUREMENT OF POLARI ZATI ON RESI STANCE
Radu Tamara, I strate Gina Genoveva,
University Dunarea de J os of Galati Romania
Abstract: Measurement of polarization resistance of micro-alloyed zinc layers followed corrosion
products characterization and corrosion behavior of the default layers analyzed. We have studied micro-
alloyed zinc coatings with different percentages of nickel, tin, bismuth, lead, cadmium, exposed to a
corrosive environment with 3% NaCl at room temperature. For experimental determinations was applied
electrochemical method. A calomel saturated electrode was used as reference electrode and a platinum
wire electrode as an auxiliary one. Analyzing the variation of polarization resistance for each sample, we
can observe how initially the sample surface becomes active, which causes a decrease in polarization
resistance followed in time by the formation of a layer of corrosion products which will cause a variation
of polarization resistance.
Keywords: polarization resistance, zinc alloy, coatings.
1. Working method and apparatus used
In most atmospheric environments, Zn
corrodes much less than steel, due to the
formation of a protective layer consisting of a
mixture of Zn oxide, Zn hydroxide and various
basic Zn salts depending on the nature of the
environment. Thus the protection of steel by a Zn
coating is mainly through the barrier effect.
However, at the places where the Zn coating is
damaged and the steel underneath is exposed,
such as at cuts or at scratches, the galvanic action
between steel and Zn can protect the exposed steel
from corrosion.
The measured values of polarization potential
include the voltage drop (ohmic loss) due to the
electrolyte and reaction products film on the metal
surface. The value of this voltage drop is the
polarization resistance [1; 2]. Ohmic resistance
drops sharply with loss of power, unlike
electrochemical polarization which decreases with
a slower rate [3]. By measuring the polarization
resistance, for alloyed zinc coatings, it is intended
to assess the corrosion products film
characteristics and corrosion behaviour of the
default layers.
Measurements were performed on samples of
steel sheet coated with zinc compositions
corresponding Table 1, using as a saline
environment corrosion 3% Na Cl at room
temperature [4]. Layers of protection were
obtained by hot dip galvanized, with the classical
preparation of the sample surface.
Since nickel has a much higher melting
temperature than zinc micro alloyed a zinc-nickel
alloy, with 2% nickel, was made. Bismuth,
cadmium, tin and lead are easily assimilated into
molten zinc and micro alloying was achieved by
introducing these elements, finely chopped
directly into zinc bath. Bismuth is a novelty
element for micro alloying zinc melts, used to
replace lead, having the same effect on melt
fluidity and reduction of surface tension without
being toxic.
In all cases it was applied to the melt
homogenize the mechanical mixing.
For measuring the polarization resistance, it
was used a potentiostatePGP type 201. A calomel
saturated electrode was used as reference
electrode and a platinum wire electrode as an
auxiliary one. The samples were prepared for
analysis by being degreased with acetone, washed
and dried [5]. After connecting the cable
conductor, the delimitation of the working surface
(equal for all samples) was made by insulating
with resin. Measurement of polarization resistance
at micro-alloyed zinc layers was made for
corrosion products and corrosion behavior of the
default layers analyzed. The variation curves of the
polarization resistance (R
p
) were experimentally
determined as a function of time for galvanized
samples and samples coated with alloy.
252
Results of the electrochemical analysis consist
in the curve of the polarization resistance function
of time. Analyzing the variation of polarization
resistance for each sample( table 1), we can
observe how initially the sample surface becomes
active, which causes a decrease in polarization
resistance followed in time by the formation of a
layer of corrosion products which will cause a
variation of polarization resistance.
Knowing the R
p
, the value of I
cor
per surface unit
can be calculated using the Stern and Geary formula
(1)
, )
, ) , )
2 2
.
cm S k Rp
mV B
cm
A
I
cor
O
=
|
.
|
\
|
(1)
The value used for the B coefficient was of
26mV.
Surface, S, for each sample was 1 cm
2
.
Table 1. Chemical composition of coatings tested
Code Type of coating
Chemical composition, [wt%]
Ni Bi Sn Cd Pb Al Zn
0 Zn 0 0 0,0005 0,0004 0,0014 0,0005 diff.
I Zn- Ni-Bi-Sn 0,16 0,71 2,95 0,26 0 0 diff.
II Zn- Ni-Pb-Sn 0,16 0 2,88 0 0,72 0 diff.
III Zn- Ni- Pb-Bi-Sn 0,16 0,41 3,49 0 0,43 0 diff.
IV Zn-Bi I 0 0,27 0 0 0 0 diff.
V Zn-Bi II 0 0,36 0 0 0 0 diff.
VI Zn-Bi III 0 0,52 0 0 0 0 diff.
In the case of hot dip galvanized samples,
without addition of other metals corrosion product
formation leads to an increased polarization
resistance to a 1 kcm
2
and because it did not
reached zero polarization resistance during the
analysis result shows that the corrosion products
layer has a slow process of compaction (Fig. 1).
Figure 1: Polarization resistance function of time for the sample coated with zinc
253
Sample I, zinc-coated complex micro alloyed
with nickel, bismuth, tin, cadmium, forming a
film of corrosion products with a polarization
resistance of 90 kcm
2
very stable and dense.
Value of 0 /cm
2
(circuit interruption) was
reached after about five minutes (Fig. 2)
Sample II, coated with zinc and nickel, tin, lead,
forming a film of corrosion products with a 45
cm
2
polarization resistance and respectively
27.5 kcm
2
, has a similar evolution in time as the
galvanized sheet (figure 3)
The sample III, coated with zinc and
nickel, tin, lead, bismuth, forms a film of
corrosion products with a very low
polarization resistance of 0.38 kcm
2
which
did not provide surface passivity during the
analysis (Fig.. 4.).
Figure 2: Polarization resistance function of time for the sample coated with Zn- Ni-Bi-Sn
Figure 3: Polarization resistance function of time for the sample coated with Zn- Ni-Pb-Sn
254
Figure 4: Polarization resistance function of time for the sample coated with Zn- Ni- Pb-Bi-Sn
Sample IV, coated with zinc and 0.27%
bismuth, shows a variation of polarization
resistance, very similar to the pure zinc coated
sample , polarization resistance is 1.23 kcm
2
(Fig. 5)
Sample V, coated with zinc and 0.36%
bismuth shows low polarization resistance (0.4
cm
2
) for corrosion products formed (Fig. 6).
Figure 5: Polarization resistance function of time for the sample coated with Zn-Bi I
255
Figure 6: Polarization resistance function of time for the sample coated with Zn-Bi II
Sample VI, coated with zinc and 0.52%
bismuth shows a variation of polarization
resistance similar to the previous zinc with 0, 36%
bismuth coated samples (Fig.7).
Figure 7: Polarization resistance function of time for the sample coated with zinc Zn-Bi III
Comparing the polarization resistance of
samples coated, with different layers of zinc
micro alloyed, with the polarization resistance of
the zinc coated sample, we can observe:
- Rp
I
Rp
II
Rp
IV
Rp
Zn
Rp
VI
Rp
V
and
I
corI
I
corII
I
corIV
I
corZn
, result a higher
corrosion resistance of samples coated with zinc-
nickel-tin-bismuth-cadmium (code I);
256
Coatings with zinc-tin-bismuth have a
corrosion resistance similar to zinc coated or
lower:
I
corZn
I
corIV
I
corVI
I
corV.
Conclusions
- Analyzing the variation of polarization
resistance for each sample, we can observe how
initially the sample surface becomes active, which
causes a decrease in polarization resistance
followed in time by the formation of a layer of
corrosion products which will cause a variation of
polarization resistance.
- The sample coated with zinc-nickel-bismuth-
tin, has form a layer of corrosion products with a
polarization resistance of 90 kcm
2
very stable
and dense, presenting a very good corrosion
behavior compared with unalloyed zinc coating.
-Samples with layers alloyed with nickel
shows a bigger corrosion resistance compared
with unalloyed zinc coating: I
corI
I
corII
I
corIV
I
corZn.
- Coatings with zinc-tin-bismuth have a
corrosion resistance similar to zinc coated or
lower.
References
[1] Bradea,T.,Popa, M.V., Nicola, M., Stiinta si
ingineria coroziunii, Ed. Academiei
Romane, Bucuresti, 2004.
[2] Bard, A.J ., Faulkner, in Electrochemical
Meihods: Fundamentals and Applications,
New York, 1980.
[3] Vermesan, H., Coroziune, Ed.Risoprint Cluj
Napoca, 2005, p.5-221.
[4] Constantinescu, A., Detectarea si masurarea
a coroziunii, Ed.Tehnica, Bucuresti,1976,
p.14-22.
[5] Balint, L., Ciocan, A., Balint, S., Potecau,
Fl., Interaction of the zinc protective layer
with an aggressive atmosphere, Analele
Universitii Dunrea de J os din Galati,
2006, Fascicula IV, ISSN 1221-4558, p.345-
350.
[6] S. Baicean, L., Palaghian, F. Potecasu,
A.Ciocan, C.Gheorghies, Macro and
microstructural changes in naval steel under
fatigue corrosion, Metallurgy International
Review ISSN 1582-2214, No.10 special
issue, 2009, p.34-41.
257
MATHEMATI C MODEL FOR OPTI MI ZATI ON OF ZI NC-NI CKEL
ALLOY CO-DEPOSI TI ON PROCESS
1
Violeta VASI LACHE,
2
Marius BENA
1
Universitatea tefan cel Mare, Suceava, Str. Universitii, Nr.13, 720229, Suceava, violetav@fia.usv.ro
2
Universitatea Transilvania Brasov, Bulevardul Eroilor, No.29, 500036, Brasov,
Abstract: Any optimization method implies a mathematical model which should resolve the
quantitative requires of the problems. This model is based on the substrate effect and has to calculate the
partial current densities and so has to give prediction as far as concern the quantities of metal
electrodeposited and the energy involved.
Keywords: mathematical model, anomalous co-deposition, zinc-nickel alloys, mass transfer,
electrochemical kinetic
1. I NTRODUCTI ON
Electrodeposition of a simple metal has also a
simplier mathematical model, but situation is
different for alloys deposition. In our proposed
model kinetic parameters are analyzed and it was
made the supposition that current distribution and
mass transport are homogeneous on working
electrode. Also we tried to determinate the reaction
mechanism.
Figure 1. The scheme which shows partial current
densities for A and B components. (a) both components
under activation kinetic control, present identic Tafel
slopes. (b) both components present limitation of the
current. (c) both components under activation control,
but with different Tafel slopes. (d) component A
presents current limitation, component B under
activation control.
2. EXPERIMENTAL CONSIDERATI ONS
2.1. Electrochemical processe at cathode
In our experiments it was deposited zinc-nickel
alloy on gold substrate, previous deposited by
sputtering on glass plates. This methode was
chosen because it permits to analyze the layers
with XRD and SEM-EDX techniques.
The mechanism of electrochemical
reactions which occur on cathode surface has two
steps, as Matlosz described Matlosz [4], [5]. Zinc
ions are deposits on their own substrate, on gold
substrate and on nickel substrate. Also nickel ions
are deposited on their own substrate, on gold
substrate and on zinc substrate. More, there are
secondary reactions, Zn
2+
ions are combining with
hydrogen to formZnH
+
, and similar Ni
2+
ions are
combining to hydrogen to form NiH
+
. These
intermediate species, formed in adsorption process,
finally will be decomposed to metallic zinc and
nickel respectively.
The mechanism of electrochemical
reactions could be written as follow:
Ni
2+
+ e
-
Ni
+
ads
(1)
Ni
+
ads
+ e
-
Ni (2)
Ni + H
+
+ NiH
+
ads
(3)
NiH
+
ads
+ H
+
+ 2e
-
Ni + H
2
(4)
Zn
2+
+ e
-
Zn
+
ads
(5)
Zn
+
+ e
-
Zn (6)
258
Zn + H
+
ZnH
+
ads
(7)
ZnH
+
ads
+ H
+
+ 2e
-
Zn + H
2
(8)
Ni
2+
and Zn
2+
are disolved as metallic ions,
hydrolized or not. Ni
+
ads
and Zn
+
ads
which could
contain or not the group hydroxyl are adsorbed in
intermediate reactions. Ni and Zn are metallic
deposits of nickel and zinc respectively [6,7]. The
kinetic of mass transfer is supposed to respect
Butler-Volmer equation. In far from equilibrium
states anodic reactions could be neglected.
2.2. Determination of partial current
densities
For a binary alloy AB and a thicknes of deposit
Ad, partial current density of B element is,
B
B
B
B
m
t m
F n
i
A
= (9)
Here m
B
is mass of element B deposited in
alloy, M
B
is atomic mass of element B, At is
deposition time and n
B
is number of electrons
implied in reaction of element B.
2.3. Normal and anomalous co-deposition of
zinc-nickel alloys
Electrodeposition of zinc-nickel alloys is
generally an anomalous co-deposition, after
Brenners deffinition, because the metal less noble,
zinc, is deposited preferential and its percent in
deposit is higher than in electrolyte. Anyway,
normal co-deposition of zinc-nickel alloys is
possible only in particulary experimental
conditions. Co-deposition of zinc-nickel alloys
from different electrolytic bath was studied
potentiostatic and galvanostatic, function of
different variable parameters during the
electrodeposition [1,2].
2.4. Mathematical modeling of zinc-nickel
alloy co-deposition. The model of substrate
effect
The initial nucleation of adsorbed nickel
on electrode surface acts as a catalyser for zinc
deposition, resulting an inhibition of nickel
deposition. Also it was shown that pure zinc
cannot be deposited from aqueous electrolyt
solutions at UPD (underpotential deposition), but
it could be co-deposited with nickel. These
phenomena can be explaine by the fact that nickel
nucleation catalize zinc deposition. At potential
more negative than zinc equilibrium potential, zinc
deposition rate is enough higher and inhibits nickel
deposition resulting an anomalous co-deposition.
The alloy deposition performs uner
substrate effect. Not only nickel affects zinc
deposition, but zinc too affects nickel deposition.
Figure2. The diagramof zinc-nickel alloy
co-deposition
In figure 2 is represented the diagram of the
effects of different substrates during electroplating
with zinc-nickel alloys. The initial electrode
surface is divided in two parts. The first is
corresponding to u
Ni
which is the surface covered
by nickel and the second is the surface covered by
zinc, u
Zn
. Every surface is then divided in four
parts. So, for nickel deposition, u
Ni
u
1
corresponds
to the area of Ni substrate surface covered with
Ni(I)
ads
. u
Ni
u
2
corresponds to the area of Ni
substrate surface covered with NiH
+
ads
, u
Ni
u
3
corresponds to the area of Ni substrate surface
covered with Zn(I)
ads
. Free surface u
Ni
(1-u
1
-u
2
-u
3
)
corresponds to the area of Ni substrate surface non-
covered.
For zinc deposition, u
Zn
u
6
corresponds to
the area of Zn substrate surface covered with
Zn(I)
ads
. u
Zn
u
5
corresponds to the area of Zn
substrate surface covered with ZnH
+
ads
. u
Zn
u
4
corresponds to the area of Zn substrate surface
covered with Ni(I)
ads
. Free surface u
Zn
(1-u
4
-u
5
-u
6
)
corresponds to the area of Zn substrate surface
non-covered.
2.5. Theoretic model. General mechanism of
electrode reactions
A mechanism of reactions was developed as
effect of substrate. his model is based on the
supposition that every individual component is
deposited after a two steps-reaction, as Matlosz [4]
described. The nickel ions are deposited on their
own substrate and on zinc substrate. Also zinc ions
are deposited on their own substrate and on nickel
substrate. More, hydrogenated species ZnH
+
and
259
NiH
+
are strongly bonded at electrode surface.
Ni(II) will react givingNiH
+
ads
and these adsorbed
species will react after with deposited nickel. Also
Zn(II) will react givingZnH
+
ads
and these adsorbed
species will react after with deposited zinc.
2.6. The mass transfer effect
The material balance in equilibrium state
through diffusion layer for species Ni(II), Zn(II)
andH
+
, 0<x<d, can be written
0
) (
= V
II Ni
N (10)
0
) (
= V
II Zn
N (11)
0 = V
+
H
N (12)
+
=
H OH
W
C C K (13)
Supposing a constant diffusion coefficient,
D, the flow of every species i, into diffusion layer
is dx DdC N
i i
/ = . Accepted values for
difuusion coefficients are 410
-10
m
2
s
-1
for Ni(II),
5,0910
-10
m
2
s
-1
for Zn(II), and 9,310
-9
m
2
s
-1
for
solvated protons, and 5,510
-9
m
2
s
-1
for hydroxid
ions [3]. Intermediate species, Ni(I)
ads
, NiH
+
ads
,
Zn(I)
ads
and ZnH
+
ads
existe only on the electrode
surface so one their concentration is equal zero in
solution [8], [9].
2.7. Electrochemical kinetic
The charge transfer kinetic is supposed that
respects Butler - Volmer equation. Far from
equilibrium, anodic reactions could be neglected.
A Tafel modified expression describes the
electrochemical reactions rate on surface and is
adapted to calculate the partial current. For
example, on the first step of deposition reaction on
nickel substrate, partial current density, i
11
, could
be wrote as:
i
11
=-Fk
0
11
C
Ni
2+
u
Ni
(1-u
1
-u
2
-u
3
)exp(-b
11
q
11
)
2.8. Software for predictive calculus of zinc-
nickel alloy composition
To simulate zinc-nickel alloy deposition, it was
elaborated a soft-ware, which uses the calculus
relationship fromdescribed mathematic model. The
calculus stops when difference between a calculated
value and a previos one is smaller than 10
-5
.
The program lines are written as follows:
using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Text;
using System.Windows.Forms;
namespace CalculTETAi
{
public partial class frmMain : Form
{
public frmMain()
{
InitializeComponent();
}
double F, Ki0, Bi, EtaI, C0, ITotal, Ai, t,
TetaS, CH, CNi, CZn, TetaIInitial, TetaZn, TetaNi;
double TetaICalculat;
double Ci, Ii;
private void m_btnCalculeaza_Click(
object sender, EventArgs e )
{
m_tbTetaICalculat.Text ="";
try
{
F =double.Parse( m_tbF.Text );
Ki0 =double.Parse( m_tbKi0.Text );
Bi =double.Parse( m_tbBi.Text );
EtaI =double.Parse( m_tbEtaI.Text );
C0 =double.Parse( m_tbC0.Text );
ITotal = double.Parse(
m_tbITotal.Text );
Ai =double.Parse( m_tbAi.Text );
t =double.Parse( m_tbT.Text );
TetaS = double.Parse(
m_tbTetaS.Text );
CH =double.Parse( m_tbCH.Text );
CNi =double.Parse( m_tbCNi.Text );
CZn = double.Parse( m_tbCZn.Text
);
TetaIInitial = double.Parse(
m_tbTetaIInitial.Text );
TetaZn = double.Parse(
m_tbTetaZn.Text );
TetaNi = double.Parse(
m_tbTetaNi.Text );
}
catch
{
MessageBox.Show( "Introduceti valori
corecte", "Atentie", MessageBoxButtons.OK,
MessageBoxIcon.Warning );
260
return;
}
int Contor =0;
int NumarBucle =1000;
TetaICalculat =TetaIInitial;
do
{
TetaIInitial =TetaICalculat;
Ci =C0 * Math.Exp( -Ki0 * Ai * t );
Ii =-F * Ki0 * CNi * CZn * CH *
TetaZn * TetaNi * ( 1 - TetaIInitial - TetaS ) *
Math.Exp( -Bi * EtaI );
TetaICalculat =Ii / ITotal;
Contor++;
if( Contor >NumarBucle )
break;
} while( Math.Abs( TetaIInitial -
TetaICalculat ) >1 * Math.Pow( 10, -5 ) );
if( Contor >NumarBucle )
MessageBox.Show( "S-a depasit
numarul de bucle!", "Atentie",
MessageBoxButtons.OK,
MessageBoxIcon.Warning );
else
{
m_tbTetaICalculat.Text =
TetaICalculat.ToString();
}
}
private void m_btnIesire_Click( object
sender, EventArgs e )
{
this.Close();
Figure 3. User interface with the software for
calculus of alloy composition
3. CONCLUSI ONS
This model establishes a mathematic apparatus
to describe zinc-nickel alloy co-deposition
processes, using the substrate effect model for
different concentration of electrolyte and for
different applied potentials. There was a good
correlation between experimental data and the
prediction of this model.
4. ACKNOWLEDGMENTS
This paper was supported by the project
"Progress and development through post-doctoral
research and innovation in engineering and applied
sciences PRiDE - Contract no.
POSDRU/89/1.5/S/57083", project co-funded from
European Social Fund through Sectorial
Operational Program Human Resources 2007-
2013.
5. REFERENCES
[1]. Bard, A.J., Electrochemical Methods.
Fundamentals and Applications, J ohn Wiley and
Sons, New-York, 2001
[2]. Di Bari,G.A., Modern Electroplating, Fourth
Edition, Edited by Mordechay Schlesinger and
Milan Paunovic, John Wiley & Song, Inc., 2000
[3]. Teeratananon, M., Saidi, K., Fenouillet, B.,
Vergnes, H., J ournnes delectrochimic, Poiters,
France, june 2008
[4]. Matlosz, M., J ournal Electrochemistry Soc.,
140(1993)2272
[5]. Soares, M.E., Souza, C.A.C., Kuri, S.E., Corrosion
resitence of Zn-Ni electrodeposited alloy obtained
with a controlled electrolyte flow and gelatin
additive, Science Direct, vol.201, Issue 6, dec.2006,
p.2953-2959
[6]. Schlesinger, M., Electrodeposition of Alloys,
Modern Electroplating, Fourth Edition, J ohn Wiley
and Sons, Inc. New-York, 2000
[7] Brenner A, Electrodeposition of Alloys, Vol.I,
Academic Press, New York, 1963
[8]. Vasilache,V., Gutt,Gh., Vasilache, T., Rev. Chim.
(Bucuresti), 59, nr.8, 2008, p. 915
[9]. Vasilache,V., Ph. D. Thesis, University Stefan cel
Mare of Suceava , 2008
[10]. Vasilache V., Gutt Gh. Vasilache T., Studies about
electrochemical plating with zinc-nickel alloys. The
influence of potential through stoichiometric
composition, Revista de Chimie, Bucureti,
vol.59,nr.9 (2008)
[11] Vasilache V, Gutt S, Gutt Gh, Vasilache Tr, Sandu
I, Sandu IG. Determination of the Dimension of
Crystalline Grains of Thin Layers of Zinc-Nickel
Alloys Electrochemically Deposited. Metalurgia
International, 2009, 14: 49-53;
261
MODUL AR DESI GN FOR A FAMI LY OF MECHANI CAL
ANTHROPOMORPHI C POL Y-MOBI L E GRI PPERS WI TH 4
FI NGERS FOR ROBOTS
Ionel Staretu
Transilvania University of Brasov, Product Design and Robotics Department, Brasov, Romania,
e-mail:istaretu@unitbv.ro.
Abstract: In this paper one anthropomorphic modular gripper for robots are described. The stages of synthesis,
analysis design and functional simulation are presented. The structural synthesis of the anthropomorphic grippers for
robots can be made regarding the following main criteria: the number of fingers, the number of phalanxes, the relative
dimensions of the phalanxes, the relative position of the fingers, the degree of freedom of the gripping mechanism
and the characteristic constructive elements used. We choice a version with four identical fingers with three
phalanxes on finger. The kinematic synthesis is used to obtain a correct closing of the finger and of the gripping
mechanism. The function of position, the function of speeds and the function of acceleration for characteristic points
are obtained from the kinematic analysis. The static synthesis solves the problem to obtaining the necessary gripping
force on each finger and the total gripping force. The calculation of strength was made in function of the internal
forces which act between elements. With the constructive dimensions a 3D model can be obtained using CATIA soft.
Some aspects regarding functional CAD and virtual simulations are shown too. For one variant of this type of gripper,
with four fingers, the technical documentation is completed and the technical project has all the conditions for
practical achievement. There are two main constructive modules: the support the palm and the finger.
Keywords: Robotics, Anthropomorphic grippers, Mechanism, Design, Functional simulation.
1. Introduction
The mechanical anthropomorphic grippers for
robots have as main mechanical component a
similar mechanism with the biomechanism of the
human hand. This mechanism has only pivot joints
and two or more fingers with two or three
phalanxes.
These grippers for robots comparatively with
others mechanical grippers (mechanical grippers
with jaws, mechanical tentacular grippers) have
more advantages like: a bigger degree of dexterity
(99% for four fingers, 90% for three fingers and
40% for two fingers comparatively with human
hand), a larger domain of utility (many types of
objects can be grasped) and that the grippers can
do micro-movements with the grasped object
between the fingers (if the degree of freedom is
equal or bigger like the number of fingers).
In the paper are shown the stages of synthesis,
analysis, design and simulation for a modulated
family of anthropomorphic grippers. There are
shown the anthropomorphic poly-mobile grippers
with four fingers from this family.
2. Structural Synthesis and Analysis
2.1 Structural synthesis
The structural synthesis can be made regarding
the following main criteria: the number of fingers,
the number of phalanxes, the relative dimensions
of the phalanxes, the relative position of the
fingers, the degree of freedom of the gripping
mechanism and the characteristic constructive
elements used [1,2].
For our family of grippers these criteria were
adapted in order to obtain a good performance:
four identical fingers, three phalanxes on each
fingers, relative position of the fingers like in
Fig.1( for 4 fingers ), the degree of freedom M=n
Figure 1: The relative position of the fingers
262
(n the number of fingers), linkage mechanisms.
The main structural module is accordingly with a
finger and it is shown in Fig.2.
Figure 2: The structural scheme of the finger
2.2 Structural analysis
The mechanism of the finger (Fig.2) is a
polycontour mechanism with two outside
connection L=2 (v
1
, F
1
; v
P1
, F
P1
Fig.3,a) and
degree of freedom M=1.
The degree of freedom is obtained with

c i
f M M
, where M
i
is the degree of
freedom for monocontour i mechanism and
c
f
is the degree of freedom for common joints
(Fig. 3,b).
For each monocontour mechanism the degree
of freedom is obtained with
K i
f M
(where
i
f
is the degree of freedom of the joints
and
K
is kinematic degree of the monocontour k

mechanism [1]).
Figure 3: The block scheme and the graph of the
mechanism
For the mechanism shown in Fig.2, in
accordingly with the graph of Fig.3,b, the
following relations are obtained:
M
I
=f
A
+f
B
+f
C
+f
D
I
=1+1+1+1-3=1
M
II
=f
D
+f
E
+f
F
+f
G
II
=1+1+1+1-3=1 (1)
M
III
=f
L
+f
M
+f
N
+f
E
III
=1+1+1+1-3=1,
and f
C
=f
D
+f
E
=1+1=2.
The degree of freedom will be:
M=M
I
+M
II
+M
III
f
C
=1+1+1-2=1 (2)
M=1 has the following significance: one
independent movement (speed):

1 1
s v
and one
function of the external forces: F
1
=F
1
(F
P1
). L-M=1
represents one function of movement: v
P1
= v
P1
(v
1
)
and one independent force: F
P1
the contact force
between finger and grasped object.
3. Kinematic Synthesis and Analysis
3.1 Kinematic synthesis
The kinematic synthesis is used to obtain a
correct closing of the finger and of the gripping
mechanism. This situation is obtained with a good
correlation between the dimensions of the
phalanxes and a good relative position of the
fingers. The first and one intermediary position of
the finger are shown in Fig.4.
Figure 4: Two configuration of the finger
3.2 Kinematic analysis
The function of position, the function of speeds
and the function of acceleration for characteristic
P
i
points are obtained from the kinematic analysis.
The vectorial close chain method is use
successively for each monocontour mechanism.
The vectorial equations are:
0 ' ' A D DD CD AC (Fig.5,a),
0 GD FG EF DE (Fig.5,b) and
0 LE ML NM EN (Fig.5,c) [3].
a
263
b
c
Figure 5: The kinematic schemes
The implicit form for the equation of positions
is:
72i
=
72i
(s
1
), i- the member of the fingers:
i=1,2,3,4.
The functions for speeds are the derivative
function of time of the functions for positions and
the functions for accelerations are the derivative of
the functions for speed:
Pi Pi
i Pi
v a
v
72
(3)
4. Static Synthesis and Analysis
4.1 Static synthesis
The static synthesis solves the problem to
obtaining the necessary gripping force on each
finger and the total gripping force.
4.2 Static analysis
The function of the external forces is obtained
from the theorem of balance between the powers of
entrance and emergence of mechanism:
v
i
F
i
+v
Pi
F
Pi
=0, and
i
Pi Pi
i
v
F v
F

(4)
The internal forces are calculated using the
theorem of the joints and, after-words, with the
balance static equations of the mobile elements [1].
5. Constructive Design and 3D Model
The calculation of strength was made in
function of the internal forces which act between
elements. With the constructive dimensions a 3D
model can be obtained using CATIA soft[6,7].
There are two main constructive modules: the
support the palm (Fig.6,a) and the finger
(Fig.6,b)[4,5].
a
b
Figure 6: The main constructive modules
The family of anthropomorphic grippers is
obtained using fingers in 5 relative positions (see
Fig.1).
For instance the possible variants with four
fingers are shown in Fig.7: one variant with fingers
with parallel axes (Fig.7,a); one variant with
fingers with parallel axes but with a interval (Fig.
7,b) ; crossing axis (Fig.7,c); three fingers with one
a
264
b
c
d
e
Figure 7: The family of the anthropomorphic
grippers with four fingers
in opposite and central position (Fig. 7,d) and
three fingers with one in opposite and lateral
position (Fig. 7,e).
For these five variants the technical
documentation is completed and the technical
project has all the conditions for practical
achievement.
6. Functional Simulation
A functional simulation (Fig.8,a,b,c) was made
to check the correct work and to identify the
solutions to obtain the optimum variant for this
grippers.
a
b
c
Figure 8: The functional simulation
Other functional simulation is made with a
piece(Fig.9). These gripper has four degree of
freedom and its can grasp objects with regular or
irregular forms.
a b
265
c d
e
Figure 9: The functional simulation with piece
This gripper, with one specific intermediary
piece, can be mounted on any industrial
commercial robot(see Fig. 10) . One of its
configuration can be obtained, during the gripper is
mounted on robot, with change the relative
position of the fingers, regarding the form of the
grasped object.
Figure 10: Example with the gripper mounted on
ABB robot
For functional simulation of the grasped
operations, the robot with the gripper were
transferred in virtual reality VRML soft (Fig.11).
Figure 11: Transfer robot with gripper in VRML soft
Here we can test different grasping operation
for different objects. Then, the results, for one
correct grasp, can be used for programming the
real gripper
7. Action and Command Scheme
7.1 Action scheme
The gripper is acted with four pneumatic
motors. The dimension of the piston of the
pneumatic motor will be : s = F
m
/p( where p is the
pressure). For concrete adopted values is selected
the motor: DSNU 25-25-P-A-MA-S2. The
pneumatic scheme is shown in Fig. 12.
7.2 Command scheme
For command are used the following devices:
drossels (LRMA-1/8-QS-8), adapter ( SGS-
M10x1,5), end component (CPE 14 PRS EP),
expanding bloc ( CPE14 PRSEO-2), end element
( CPE14 PRSGO 2), blocked element ( CPE14-
PRSB).
Figure 12: Pneumatic scheme
The control subsystem is make of eight sensors
CZN-CP15 type with the following characteristics:
- 40 C degrees until +85 C degrees; 0,2 until 100
N grasp force; intensity: 1 Ma; period of life at 35
N:10 million of operations and a signal
convector(1 M 36-22 Ex-U).
The general scheme for motor, command and
control subsystem for one finger is shown in Fig.
13.The grasp process has the main following
stages: start signal for closing the gripper (electro-
mechanical , electrical or voice); sensing or not
sensing of the object by the tactile sensors;
obtaining the grasping force; transfer of the
object; open the gripper ( similar as the closing
stage of the gripper).
266
Figure 13: Command and control scheme
8. CONCLUSI ONS
The next conclusion can be formulated in
according to the considerations presented:
a) The main stages for to design the
anthropomorphic mechanical grippers are:
structural synthesis and analysis, kinematics
synthesis and analysis, static synthesis and
analysis, constructive design and 3D model and
functional simulation.
b) These grippers can be obtained using two
main modules: the support the palm and the
finger.
c) The family of the mechanical
anthropomorphic grippers for robots with two,
three, four and five identically fingers has more
variants, what can be obtained in accordance with
the number and the relative position of the fingers.
d) The aspects shown in this paper can be used
at families of the anthropomorphic grippers with
more than four fingers, with five or six identical
fingers.
e) Each finger can be acted with one
pneumatic motors and for command and
control can be used one classical command
scheme.
References
[1] Staretu, I., Daj, I. Mechanisms and Machine
Elements (in Romanian), Ed. Lux Libris,
Brasov, Romania, 2000.
[2] Staretu, I., Gripping systems (in Romanian),
Ed. Lux Libris, Brasov, Romania, 1996.
[3] Staretu, I., a.a. Mechanical Hand.
Anthropomorphic Gripping Mechanism for
prostheses and robots (in Romanian), Ed. Lux
Libris, Brasov, Romania, 2001.
[4] STARETU, I., BOLBOE, M., Synthesis,
Analysis, Design and Functional Simulation
for a Family of Anthropomorphic Grippers for
Robots, Proceedings of the SISOM 2007, pp.
131-134, Bucharest, 29-31 May, Romania.
[5] STAREU, I., Anthropomorphic Gripping
Systems with J ointed Bars or Wheel and wires
for Industrial Robots Constructive Synthesis,
Analysis and Design, New Trends in
Mechanisms, Editors S.M. CREU and
DUMITRU N. Academica-Greifswald
Publishing House, 2008, pp. 133-144.
[6] STAREU,I., Sisteme de prehensiune
(Gripping Systems)-editia a II-a, Ed. Lux
Libris, Braov, 2010.
[7] STAREU,I., Gripping Systems, Derc
Publishing House, Tewksbury,
Massachusetts,U.S.A., 2011.
.
267
THE I NFLUENCE OF ENVI RONMENT TEMPERATURE VARI ATI ON
ON THE STRENGTH CHARACTERI STI CS OF COMPOSI TE MATERI ALS
TYPE ALUCOBOND"
Constantin DULUCHEANU
1
, Traian Lucian SEVERIN
1
, Nicolai BANCESCU
1
,
1
Management, University Street, No. 13, 720225, Suceava, Romania, EU. dulu@fim.usv.ro,
severin.traian@fim.usv.ro, bancescu@fim.usv.ro
Abstract: In this paper the authors present an analysis of the influence of temperature variation on
the strength characteristics of the composite material called ALUCOBOND;, a material used in
construction on plating exterior wall of buildings and the execution of ornamental ensembles for
advertising purposes.
Keywords: composite material, environment, temperature, resistance
1. General consideration
The composite material called ALUCOBOND
is made of aluminum plate of 0.5 mm, joined by a
core of permanent polyethylene; blend is achieved
at constant chemical and mechanical (hot-rolled
and pressure) and the grip is perfect on all the
material. This structure provides a surface rigidity
of the composite material, resistance to bending,
perfect flatness and ease of processing.
Even if they have a lightweight compared to
other materials, ALUCOBOND boards have a very
good resistance to breaking or bending, while the
plates have exceptional qualities in terms of
resistance at high pressures and buckling.
Polyethylene core which is non-toxic have a good
density and improves resistance to fire.
ALUCOBOND dampens vibrations of 9 - 10
times more efficiently than aluminum sheet; he
behaved as an electromagnetic coating and for this
reason is a very good solution for plating hospitals,
airports, military bases, buildings in city centers
etc. Also the manufacturers of these composite
materials say that ALUCOBOND does not change
the mechanical properties between -50
C and +80
C, [3]
Because this composite material is used for
external plating of the buildings walls and the
execution of ornamental ensembles for advertising
purposes which is used as a support for various
objects, the authors proposed to determine whether
the environmental temperature changes affect its
resistance properties.
To do this, lots of 10 pieces of 4 mm thick by
ALUCOBOND were subjected to following
laboratory thermal stresses:
I. without thermal stresses (samples in delivery
state of material);
II. long cycle of heating-cooling;
III. two long cycles of heating-cooling;
IV. three long cycles of heating-cooling;
V. six short cycles of heating-cooling.
A "long cycle" consisted of heating samples in
an oven at 37
C for one hour followed

immediately by cooling to -30
C for one hour,

cooling done in a laboratory refrigerator existing in
thermal treatments at "Stefan cel Mare" University
Suceava. The "short cycle", heating and cooling
were performed at the same temperature (37
C and
-30
C), but times were held at that temperature

were only 15 minutes. The samples were made in
accordance with applicable standards (SR EN
10002-1:2002) [2].
After the planned thermal stresses, the samples
were traction tested. These tests were conducted in
Technology of Cold Pressing laboratory on a
traction test machine of steel specimens (Figure 1),
it is equipped with a linear displacement transducer
which is connected with a strain gauge deck plate
268
by acquisition board LabJ ack U12 and makes
possible the transmission of values to a computer
[4]. Using special software, the data transmitted
can be automatically calculated and displayed
Figure 1. Traction test machine of steel samples
Acquisition and evaluation size program is
made in LabView programming environment; it
takes the sensor information in a chart
representation as presented in Figure 2. The
application contains a button to save data to a file
type. ".dat" The charts of the two sizes
are represented in different colors: red - "force"
and green - "elongation".
Figure 2. Chart representation of the sizes results in the
traction test
The values recorded in the file. ".dat" are open
with Microsoft Office Excel program (Figure 3)
are taken only force and elongation recorded
values and placed in another Excel program that
can generate the following representations charts:
conventionally (); conventional (), real
experimental (), real experimental (), real
theoretically (), real theoretically (), [1].
Figure 3. Values recorded in the file .dat
269
After thermal stresses according to schedule,
the samples were subjected to traction tests, results
are presented in Table 1 and Figures 4 6.
Tabel 1. Experimental results
Sample
lots

max

r
[MPa] [%] [MPa] [%]
I 27,74 100 26,76 100
II 27,66 99,72 26,68 99,70
III 26,78 96,54 25,97 97,05
IV 26,14 94,23 25,17 94,06
V 26,40 95,17 25,53 95,40
Figure 4. Graphical representation of experimental results for sample 3 in group IV
Figure 5. The influence of thermal stresses on
max
Figure 6. The influence of thermal stresses on
r
.
270
Analyzing the results is found a reduction of the
resistance characteristics from 27.74 MPa for
max
,
on samples in delivery state (group I), to 26.14
MPa for samples in group IV (samples subjected to
three long cycles of heating-cooling), that
corresponds to a reduction of 5.77%, in the case of
r
values, the decline was from 26.76 MPa to
samples from group I to 25.17 MPa for those in
group IV, ie 5.95% reduction. Similar results were
recorded for specimens in group V were subjected
to six short cycles of heating and cooling:
max
=
26.40 MPa (4.83% reduction) and
r
=25.53 MPa
(4.60% reduction).
4. Conclusions
Composite material type ALUCOBONT is used
in plating exterior walls of the buildings and the
execution of ornamental ensembles for advertising
purposes which is used as a support for various
objects.
To check reaction of environmental temperature
changes the samples for traction test were
subjected to thermal cycles involved heating at 37
C and cooling to -30
C, maintaining these
temperatures for 60 minutes (cycle "long" ) and 15
minutes (cycle "short").
The analysis of the results was a decrease in
resistance characteristics (
max
and
r
) with values
contained between 0.3% and 5.77%, depending on
the number of heating and cooling cycles and
applied times to maintain use.
References
[1] Severin, L., Iacob, D., Severin, T.L.,
Tehnologia presrii la rece, Editura
Universitii Suceava, ISBN 973-666-149-0,
Suceava, 2005.
[2] SR EN 10002-1:2002.
[3] * * *, www.geplast.ro.
[4] * * *, www.labjack.ro.
271
MATHEMATI CAL MODEL S FOR SMOOTHI NG PROCESS BY
MECHANI CAL SHOCK S OF CYL I NDRI CAL SURFACES
Raluca Tanasa
1
, Traian Gramescu
1
Technical University Gheorghe Asachi of Iasi-Romania
Department of Machine Manufacturing Technology
raluca.tanasa@gmail.com, grat@tcm.tuiasi.ro
Abstract: Smoothing by mechanical shocks is a processing method based on the impact effect of the
balls with the work piece material. Specific phenomena such processing methods result in changes of
mechanical properties of the surface layer of the piece; also change the appearance of the original surface
layer. The surface obtained after processing the outer cylindrical surface by mechanical shock is
considered a chain of small cavities generated by the impact of the balls with the workpiece material. The
outer cylindrical surfaces roughness obtained after the smoothing using centrifuged balls, is an important
parameters of this process. The aim of this paper is to present ways to develop mathematical models of
surfaces roughness. Also are presented experimental results that confirm the validity of mathematical
models and their level of approximation of the real situation.
Keywords: Smoothing, Roughness, Plastic deformation, Hardening
1. Introduction
Industrial practice highlights a number of
technological possibilities for surface finishing of
pieces belonging to metal parts. The specialists
consider that the cold plastic deformation of the
metallic workpieces could be included into the
group of cold manufacturing processes, when the
temperature during the process does not exceed 30
% of the melting temperature of the respective
material. There are various cold manufacturing
processes; even the machining processes (cutting,
drilling, turning, milling, grinding etc.) have as
initial basic phenomenon the plastic deformation of
the workpiece material. Along with cutting or
nonconventional processes, cold plastic
deformation is used extensively in surface
finishing.
Extensive research conducted over the last
categories of interest to show their development
processes [6]. The advantages of these processes,
compared to the other are obvious: can be applied
to a variety surfaces, high mechanical
characteristics of the surface layer, surface
roughness comparable to that obtained from
cutting, much lower costs related to processing. A
manufacturing techniques based on the cold plastic
deformation is caused by the impact effect of the
hard metallic balls with the workpiece material.
From finishing processes of surfaces by plastic
cold deformation, cold centrifuged ball striking
process, so-called ball peening [4] appears more
economically due to its simplicity and because it
can be applied to pieces with lower stiffness.
The method involves the hitting of the
workpiece material by balls found in motion, so
that each impact contributes to the generation of
small cavity in the workpiece surface layer.
The actions of hard balls sent with specific force
and frequency over metallic surface of the
workpiece have significant effects on the structure
of the superficial surface layer but also on the size
of micro-irregularities surface.
The balls can be directed to the workpiece
surface in different ways; they could be transported
by a gas jet [4], but there are also equipments
which direct the balls just to the workpiece surface
to be processed.
During the process there is a clash of distorting
element (hard ball) to piece surface. The ball
hardness is much higher than the one of the piece
to be processed. The shape of piece after crash will
result in form of a spherical cap with size
dependent of working conditions.
Observing the condition of generating
cylindrical surface, strikes (collisions) will be
carried throughout the circumference of the piece
by turning it with speed n
p
. It can be concluded
272
that collisions constantly producing at part rotation
angle at the center
p
(Fig. 1) there will always be
an oscillation of level with a certain height R
max
surface, which is micro-irregularities of the process
with the same size [3].
Figure 1: Schematic diagramof the smoothing process
with centrifuged balls.
Methodology presented in this paper is to
determine theoretical mathematical expressions
(models) of the roughness R
max
. Also are presented
some experimental results obtained after applying
the smoothing process by mechanical shocks to
determine the reliability of the models obtained,
compared with experimental results.
2. Developing mathematical models
From Fig. 1 can be observed that the surface
deformation is in the shape of a spherical cap, with
a radius equal with balls radius, r
b
. This highlights
the possibility that a complete rotation of the part
obtain a number of spherical caps N
c
, that can be
computed with Eq.(1).
2
'
p
c
c
r
N
l
t
= . (1)
The spherical cap length l
c
(
ADB arc) can be

approximated with length rope ( AEB), the value
of segment ED is insignificant in relation to the
size of rope. Using geometrical relations existing
in the triangle
b
O AE can be expressed the Eq. (2).
2
max max
2 2
c b
l r R R = . (2)
By entering into Eq. (1) the obtained value of l
c
,
the resulted relationship it will have the form of
Eq. (3).
2
max max
2
p
c
b
R
N
r R R
t
=
.
(3)
On the disk tool are mounted a number of balls,
N
b
. After processing will be performed a number of
spherical shape on the parts surface, expressed by
the Eq. (4):
d
c b
p
n
N N
n
= .
(4)
In Eq. (4), with then
d
is expressed the disk tool
rotation, andn
p
represent part rotation. Combining
Eqs. (3, 4) results equation (5):
2
2
max max
2 0
p p
b
d b
r n
R r R
n N
t | |
+ =
|
\ .
. (5)
The solutions of Eq. (5) are:
2
2
max
p p
b b
d b
r n
R r r
n N
t | |
=
|
\ .
. (6)
It will be considered for the value of R
max
only
2
2
max
p p
b
d b
r n
R r
n N
t | |

|
\ .
.
Because for the higher values of R
max
than the ball
radius r
b
, the circle of maximum section has the
radius / 2
b c
r l = , and the approximation is
inconclusive.
The analysis of Fig. 1 shows the fact that to each
spherical cap from the surfaces part corresponding
a center angle:
2
p
c
N
t
= .
(7)
The N
c
parameter from Eq. (7) represent the
number of spherical cap resulted on the part
surfaces in one pass of disk tool over the piece.
Accepting the idea that the geometrical figure
ADBE represent a spherical cap, can be written the
following mathematical relations:
2
2 sin ;
4
2 sin
2
p
p
p
c p
ED r
l AEB r
=
= =
. (8)
Replacing theN
c
parameter from Eq. (4) in Eq.
(7) we get next relation:
273
2
p
p
d b
n
l
n N
t
=
. (9)
The AEBrope is common to both circles with
radius r
p
(piece) and r
b
(deformation ball).
Relations based Eqs. (8, 9) can be written:
2
2 sin 2 sin
2
p
b
b p
d b
n
r r
n N
t
. (10)
Applying mathematical rules can be obtained
the expression for calculating the angle
corresponding to a center ball of a spherical cap:
2
2arcsin sin
p p
b
b d b
r n
r n N
t
| |
=
|
\ .
. (11)
b
angle allows finding the height of spherical
cap (CE ) in the relationship
2
2
1 cos 2 sin
2 4
2
1
2 sin arcsin sin
2
b b
b b
p p
b
b d b
CE r r
r n
r
r n N

t
| |
= = =
|
\ .
( | |
=
( |
( \ .
(12)
To determine the expression for calculating the
roughness surface R
max
we will consider the Eqs.
(8, 9, 12).
max
2
2
2
1
2 sin arcsin sin
2
2
2 sin
p p
p
b d b
p
p
d b
R CE ED
r n
r
r n N
n
r
n N
t
t
= + =
( | |
= +
( |
( \ .
.
(13)
From the analysis we see that for the theoretical
calculation of height, R
max
, of micro irregularities
can use different expressions (see Eqs. (6, 13).
3. Experimental research
To see the approximation of the theoretical
results by applying the expressions referred to
those achieved in practice, were performed
experimental tests. To convert the surface
roughness R
max
in R
a
was used the approximate
relations [2]:
max
max
10
5
1
10
a a
a a
R
R for R m
R
R for R m
= >
= <
.
(14)
2.1 Equipment and work piece material
Research, [8] has been conducted with a special
device mounted on a universal lathe (Fig. 2). The
device features a disk tool with balls. It used a
special tool to direct balls to the surface of the
workpiece. It consists of a cylindrical disk with
holes in witch are placed in predetermined
diameter balls.
Figure 2: Device for smoothening by plastic
deformation with centrifuged balls.
At the periphery of this disc, there is a ring
which is made a number of holes equal to the
number of holes; diameter of these holes is smaller
than the diameter of balls inserted. In this way, the
balls move short distances in cylindrical holes;
when the disc is rotated, (Fig. 3) the balls are rolled
on the ring under the action of centrifugal force
and parts thereof beyond the outer cylindrical
surfaces of the ring.
To be affected by the ball peening process,
the hardness of the ball material has to exceed
the hardness workpiece material with at least
several HRC unities. The process of the ball
peening presented in this paper is applied in
the case of the workpiece materials able to be
affected by a certain plastic deformation and
hardening; this means that usually, the
workpiece material must have a phase
structure corresponding to that obtained by an
annealing heat treatment.
disk tool
device
work piece
274
In the considered case, the axis of the
workpiece and the axis of the disk tool are parallel
and placed in the same horizontal plan.
When the disk tool does not perform the
rotation motion, the balls existing in the holes from
the disk tool body has a position essentially
determined by the presence of their weight force.
Figure 3: Disks with deforming balls (different
diameters of balls).
Striking force was achieved by centrifugal force
resulted at disk tool turning withn
d
speed.
The workpiece and the disk tool perform
rotation motions in the same direction; due to this
fact, the ball takes contact with the external
cylindrical surface of the workpiece in a position
situated over the horizontal plan determined by the
workpiece and the disk tool axes.
As consequence of the impact, if the pressure
exerted by the ball under the action of the
centrifugal force on the workpiece surface layer is
higher than the compression resistance of the
workpiece material, a small cavity is generated on
the workpiece surface. At the same time, the ball is
pushed on a low distance in the conical hole
existing in the disc tool body, but it is obliged by
the ring to continue the motion along a trajectory
which is partially circular. Because in the contact
zone the workpiece surface and the ball perform
motions in opposite directions, the shape of the
cavity generated in the workpiece surface layer is
modified; of course, really the initial theoretical
spherical shape of this cavity is also affected by the
elastic recovery of the workpiece material.
Experimental researches were complying with
the scheme from Fig. 1.
The experimental test samples for smoothing by
mechanical shocks were selected from commonly
used materials in tree construction. Was considered
that smoothing process by plastic deformation is
applied mainly journal shaft couplings that are
used in friction [1]. The test pieces were made of
rolled steel containing 0.37% carbon, quality steel
containing 0.45 % carbon and an alloyed steel
containing 0.17 % carbon, 0.22 % molybdenum,
0.95 % chromium, 1.3 % nickel.
Experiments were performed on cylindrical
samples of diameter = 45 mm and length L =
300 mm. To ensure good crosses were made
during experiments centering holes. Cylindrical
surface was divided into portions with a length of
30 mm, Fig. 4; with "S is noted the safety section
left to avoid possible influence of the chuck over
the finishing operation results.
5
F
4
5
30
L=300
S
Figure 4: Dimensions and geometry of workpiece
Before being subjected to experimental
research, all samples were processed by finishing
turning.
2.2 Smoothing conditions
Smoothing by mechanical shock occurs at an
ambient temperature, the temperature well below
the recrystallization, which draws after itself that
all changes in the metal structure can be easily
recognized, as a result of plastic deformation
occurred [5]. The considered parameters in order to
express the theoretical point of view (Eqs. 6, 13),
of a cylindrical outer surface roughness results
applying the finishing process by mechanical
shocks, were: parts radius (r
p
), part speed (n
p
), disk
tool speed (n
d
), number of balls on disk tool (N
b
)
and ball radius (r
b
).
The experimental research carried out will take
into account the initial roughness (R
ai
) of the
workpiece undergoing processing by mechanical
shock [7]. The parameter constant during the
experiment was piece diameter; the used diameter
disk tool
ring
ball
275
was =45 mm. Smoothing process conditions are
given in Table 1.
Table 1: Synthesis of smoothing conditions
T
h
e
o
r
e
t
i
c
a
l

p
a
r
a
m
e
t
e
r
s
E
x
p
e
r
i
m
e
n
t
a
l

p
a
r
a
m
e
t
e
r
s
Part radius
(mm)
5 30
17,5
Part speed
(rot/min)
16
Disk tool speed
(rot/min)
2890
Number of balls on disk tool
(mm)
16
Ball radius (mm) 4
Initial roughness
m)
2 4
The experimental study of the ball meaning
process was made by using the device
schematically presented in Fig. 2; the device was
mounted on a universal lathe, instead of the tool
holder. The disk tool had a diameter D
d
of 200
mm; it presented a single balls row, the ball radius
beingr
b
=8 mm.
In order to study and discuss the effect of the
input parameters and to determine the theoretical
models that express surface roughness near reality,
were realized the construction from Figs. 5 8.
These graphs show the effect of smoothing
conditions in four different conditions in the cases
of expression (6) and expression (13) considered
over the workpiece surface. In this condition, four
of the input parameters were constant, and varied
only one.
The effect of deforming ball diameter on the
variation of surface roughness, using theoretical
models, can be appreciated from figure 5. In the
case of first mathematical model (Eq. 6), as the
deforming ball diameter increases, the surface
roughness decreases.
0
10
20
30
40
50
60
70
80
0.50 1.00 1.50 2.00 2.50
rb [mm]
R
m
a
x

[
m
]
Rmax th. (Eq.6)
Rmax th. (Eq.13)
Figure 5: Theoretical results of roughness surface
expressed with theoretical models at different deforming
ball diameter
Figure 6 constitute the effect of the part speed
on the theoretical surface roughness, in case of the
mathematical models. With increases of part speed
a good value for roughness surface is indicated by
the Eq. 6. From the graph it is preferable to avoid
smoothing at high part speed, because surface
roughness increases.
0
2
4
6
8
10
12
14
16
18
16 20 25 32 40
np[rot/min]
R
m
a
x

[
m
]
Rmax th. (Eq. 6)
Rmax th. (Eq. 13)
expressed with theoretical models at different part
speed
Number of balls on disk tool represents an
important smoothing parameter that affects the
results of process. The increase of number of balls
on disk tool will improve (decreases) the roughness
surface. The shape of graph recommends using Eq.
13 from the theoretical models.
0
2
4
6
8
10
12
14
16
18
1 4 8 12 16 Nb
R
m
a
x

[
m
]
Rmax th. (Eq. 6)
Rmax th. (Eq. 13)
expressed with theoretical models at different number of
balls on disk tool
Effect of workpiece diameter on the theoretical
roughness is shown in Fig. 8. The graphics
rendering is seen as a small roughness is obtained
when is used the mathematical model 2 (Eq. 13).
Also it is observed that lower values of surface
roughness parameter, are obtained at different
diameters of the part but maintaining a
constant optimal value of the other input
parameters in the process.
Mathematical models and experimental results
(Fig. 8) were verified. Values obtained for a given
situation confirms the practical approximation
model 2 (Eq. 13).
276
0
0.5
1
1.5
2
2.5
3
3.5
4
0 5 10 15 20 25
rp[mm]
R
m
a
x

[
m
]
Rmax th. (Eq.6)
Rmax th. (Eq.13)
OLC45 (Rai1)
17MoCrNi14 (Rai1)
OL37 (Rai1)
OLC45 (Rai2)
17MoCrNi14 (Rai2)
OL37 (Rai2)
expressed with theoretical models at different part
radius; experimental points, resulted in same working
conditions, on different materials.
Deviations from theoretical values of the
experimental results may have several causes:
- processed material behaves elastic plastic due to
the phenomenon of hardening;
- as a result of the initial hypothesis, according to
which was considered the circular contour
footprint, occur deviations;
- due to superficial hardening phenomena, every
shot is applied near the previous. That why the
processed material by mechanical shocks it is
partial hardened, fact that gives elastic plastic
properties, which change in practice the initial
assumptions made.
- practice values deviation in comparison with the
theoretical values can be explained due to the
fact that during the collision, the hard ball is
deforming because it elasticity resulting a print
that deviate from a spherical cap (as considered)
Taking into account the elements of the local
strains theory for massive bodies collision is
justified the reduced roughness values increasing
parts diameter.
Some deviations from theoretical values are due
to other factors of influence (initial surface
roughness, longitudinal advance of the disk tool
reporting to piece).
Also different values were obtained and
variation due to plasticity of tested materials. Note
that the processing of material with high plasticity
(OL37), we obtain a higher roughness. More
tenacious material (elastic plastic) with a higher
content of perlite (OLC45, 17MoCrNi14) allows
obtaining lower roughness.
4. Conclusions
The smoothing method presented in this paper
is a processing way of the workpiece cylindrical
surface which is based on the impact effects of the
balls found under the action of the centrifugal force
with the workpiece surface. Also as result of the
smoothing by mechanical shocks, the workpiece
surface layer is affected by a hardening process.
Theoretical considerations were used to obtain
different mathematical relations to characterize the
superficial surface roughness of the work piece
resulted after the impact phenomena. Some
experimental researches were finalized to offer an
image concerning the influence exerted by certain
work conditions on the surface roughness
parameter R
a
. Because each encounter is repeated
in the immediate vicinity of micro crater above is
possible that by increasing the superficial surface
layer micro hardness the plastic deformation
approach to partially elastic deformation.
References
[1] Dieter, G., Mechanical metallurgy (in
Romanian), Bucureti: Editura Tehnic, 1970.
[2] Dietrich, R., Garsaud, D., Gentillon, S.,
Nicolas, M., Precis methodes dusinage.
Methodologie, production et normalisation,
Afnor, Nathan, Paris, 1981.
[3] Grmescu T., Sltineanu L., Upon some
theoretical aspects at metallic surfaces
smoothing with centrifuged balls, Iai:
Buletinul Institutului Politehnic, Tom LII
(LVI), Fasc.5A, 2006.
[4] Korzynski, M., Dzierwa, A., Pacana, A., and
Cwanek, J ., Surf. Coat. Technol., 204 (5), pp.
615-620, 2009.
[5] Lin, Y.C., Wang, S.W. & Lai, H.-I., The
relationship between surface roughness and
burnishing factor in the burnishing process. Inl
J of Advanced Manufacture Technology,
Vol.23, pp. 666-671, ISSN 1433-3015, 2004.
[6] Lupescu, O., N Surfaces smoothing by plastic
deformation (in Romanian), Chiinu: Editura
Tehnica Info, ISBN 9975-910-67-X, 1999.
[7] Sltineanu, L, Tanasa, R., Gramescu, T,
Coteata, M., Academic J ournal of
Manufacturing Engineering, 6 (4), 109-113,
2008.
[8] Tanasa, R., Researches on finishing
technologies by mechanical shocks, PhD
Thesis, Iasi, 2010.
277
EXPERI MENTAL RESEARCH ABOUT I NFL UENCE OF
I NCLI NATI ON ANGLE DI RECTI ON UPON THE SURFACE
ROUGHNESS I N BAL L END MI LL I NG OF OLC45 (C45) MATERI AL
PACA I oan
1
, LOBONI U Mircea
2
1
phd. student, North University of Baia Mare, Str. Dr. Victor Babes, nr.62 A , Baia Mare, Romania,
e-mail: pascaing@yahoo.com.
2
phd. prof., North University of Baia Mare, Str. Dr. Victor Babes, nr.62 A , Baia Mare, Romania,
e-mail: mircea.lobontiu@ubm.ro .
Abstract: Machining with ball end mills become a very usefull process to obtain pieces with complex
surfaces. Possibilities of machining with ball nose end mills represent a adequate process to achieve
surfaces with a good quality. That fact lead to exclude some finishing operations that was necessary if the
surfaces was machined with a different process and tools. Elimination of this finishing operation have a
great economic advantages, because milling with ball nose end mills offer a adequate surface roughness
and precision. But, all of this are possible only with a good choise of different values of cutting
parameters and with an optimum inclination angle represented by value and direction of inclination. The
present paper includes an experimental investigation of surface roughness depending on the value and
direction of tool axis inclination in raport with surface machined.
Keywords: roughness, surface, inclination angle, direction.
1. Introduction
Obtaining a high quality of surfaces machined
with ball end mills present a great interest for
reserchers and persons involved in machining of
die and mould surfaces. The surface quality in ball
end milling is the results of influence of a number
of factors such as: the stability of machine tools,
the correct choosing of cutting tools and cutting
parametters, properties of material machined and
the tool angle inclination upon the surface normal.
To have good results all these factors should be
chosen to have a minimal negative influence in the
process.
Machining of dies and moulds surfaces in
milling process involved polishing operations that
represent sometimes approximately two thirds of
the total manufacturing costs [4]. In this way the
milling process with ball nose end mills represent a
solution to reduce the costs of process and of final
product but only if the values of parameters
involved in process are adequate. Chosen of
optimum value of different parameters is dificult
because it is necesary to reduce the time of
machining and to obtain surfaces with good
roughness. In international context a great number
of researches is dedicated to surface quality
prediction. Some of them is dedicated to
experimental research, but a lot of them are
proposing some mathematical models. That
mathematical models proposed take into account
the cutting speed, feed value variation and the
following parameters: cutting depth [8], cutting
depth and radial depth [5], hardness of the material
and tool radius [7].
By using a cutting tool angle between the axis
of tool and machined surface can contribute to a
higher quality of machined surface [1].
The paper [9] and [6] suggested that a tool
inclination in the range of 1020 degrees
represents the optimum machining strategy for
high speed milling in the die and mould making
industry, but in other paper like [10] the optimal
value of inclination angle in ball end milling of
block materials was found at 15 degrees. Another
reserch show that when piece was machined with
inclination of tool axis of -17 degrees following Y
axis in one way the profile of the machined work-
piece, show the improvement of the machined
surface texture quality [2]. In paper [3] it is noted
that in practice the angle of 45 degrees for
inclination of tool axis the quality of surface is
low.
278
Because surface roughness have an important
effect upon product quality and economicity of
process, was necesary to study the influence of
inclination direction of tool axis upon surface
normal and try to compare the surface roughness
obtained in some situation.
2. The analysis of tool axis inclination
Machine tools with five axis lead the possibility
to work with one or two angle of tool axis
inclination in relation with normal to the surface,
situation that allow to achieve better cutting
conditions and therefore a high quality machined
surface. Working with this inclination angle the
contact zone between tool edge and material is
changing that involve another condition with
modification of chip thickness and area of contact.
Also, the cutting forces are different and the
energy consumed are different. Using this
inclination angle we have possibility to reduce the
costs of process and to obtain a higher quality.
Inclination angle can be applied to tool around
the axis X,Y or Z corresponding to the structural
characteristics of machine tools and to
characteristics of surfaced machined.
Inclination angles proposed for analysis in this
paper are around the X axis by angle A positive
(Fig.1) or negative (Fig.2) and around to Y axis by
angle B positive (Fig.3) or negative (Fig.4).
3. Effects of inclination angle upon effective
cutting speed
By using an inclination angle between tool axis
and normal to surface we change some
characteristics of process. One of them is effective
cutting speed thus value depending by effective
diameters. In order, the effective diameters depend
by cutting depth, radial depth and inclination
angle. Increasing of effective diameters involve the
increase of effective cutting speed.
To establish the value of effective speed
milling V
c-eff
coresponding to inclination of 15
degress angle value, we used relation (1) [3]:
1000
R
p
a R
arccos
n
sin D n
ef c
V
(
(
|
|
.
|
\
|
+ u t
=
(1)
with the following condition:
Figure 1: Inclination of tool axis in B negative direction
around Y axis
Figure 2: Inclination of tool axis in B positive direction
around Y axis
Figure 3: Inclination of tool axis in A negative direction
around X axis
279
Figure 4: Inclination of tool axis in A positive direction
around X axis
|
|
.
|
\
|
s u ( |
.
|
\
|
R
p
a R
arccos 90
n
R 2
e
a
arcsin , (2)
where:
-D nominal diameter of tool [mm];
-a
p
cutting depth [mm];
-n spindle speed [rot/min];
n
inclination of tool axis [grade];
-R tool radius [mm].
Geometrical parameters values used in our
experiments are presented in Table 1.
Table 1:Geometric parameters used
No. Geometric
elements
Units of
measurement
Values
1 Diameter of
ball end mill
mm 16
2 Cutting depth mm 0,2
3 Axial depth mm 0,2
4 Feed per
tooth
mm 0,1
5 Tool angle
inclination
degrees 15
The results obtained for two values of
speendle speed used in our experiments are
4. Experimental work and conditions
Machining by using a cutting tool angle
between the axis and machined surface can
contribute to a higher quality of machined surface
[1]. Analyzing the indication presented in paper
[9],[6],[10], [11] and [2] about optimum value of
inclination angle we choose to apply in our
experiment value of 15 degrees in raport with
normal to the surface.
Table 2:Effective cutting speed values
No. Speendle
speed n
[rot/min]
Inclination
angle
[degrees]
Effective
cutting
speed
[m/min]
1 10000 15 234,6
2 15000 352
Once identified the optimum value of
inclination angle we found that is necesary to
extend the studies by analysing the influence
of inclination angle direction upon surface
roughness. We want to establish the best
direction of inclination angle that lead to high
quality of machined surface, but according
with some defaults machining conditions.
4.1 Experimental setup
To make the experimental analysis of
inclination angle direction upon surface roughness
was necessary a number of experiment that was
made by using a OKUMA MU400VA five-axis
CNC milling machine equipped with a maximum
spindle speed of 15,000 rpm (Fig.5.), that was
propriety of S.C. RAMIRA S.A. We used a tool
holder SRM2160SNM and inserts SRG16C-
VP15TF made by Mitsubishi Carbide that is
presented in Fig.6.
Figure 5: Five-axis milling machine
280
Figure 6: Tool holder SRM2160SNM
4.2 Workpiece configuration
The material used was OLC-45 (C45) with
the following caracteristics: 0,420,50% C,
0,500,80% Mn, 0,170,37% Si, maximum
0,040% P etc. The workpiece had a economic
shape because we can machined six surfaces
with the same quatity of material (Fig.7.)
Figure 7: Workpiece configuration
4.3 Surface roughness measurements
In the technical literature and industry are
many parameters used to defined the surface
roughness, but in generally is defined as the
inherent irregularities of the workpiece affected by
machining process. The paper [3] show that the
most popular of the 2D parameters is the average
roughness Ra. In order to obtain more information
about surface machined we made the
measurements of surface roughness in feed
direction and perpendicular to it. We used the
tester TR200 that we applied on new device that
provides the conditions of perpendicularity and
parallelism between roughness tester feeler and
feed direction. That measurements stand is
presented in Fig. 8.
4.4 Experimental conditions
In order to establish the conditions utilized
in our experiments we must say that all
experiments were conducted using in feed milling
directionand with parameters indicated in Table 1,
2 and 3.
Figure 8: Roughness tester TR200 with adaptor device
Table 3: Experimental conditions
Nr Inclination
angle
[degrees]
Spindle
speed
[rot/min]
Inclination
direction
1
15
10000
A0B+15
2 A0B-15
3 A+15B0
4 A-15B0
5
15000
A0B+15
6 A0B-15
7 A+15B0
8 A-15B0
5. Experimental results and discussion
Experimental data on surface quality of the
measurements obtained were analyzed in feed
direction and perpendicular to feed direction for
two values of effective cutting speed and were
centralized in Table 4.
Table 4: Experimental data
Nr.
D
i
r
e
c
t
i
o
n

o
f
i
n
c
l
i
n
a
t
i
o
n
E
f
f
e
c
t
i
v
e
c
u
t
t
i
n
g

s
p
e
e
d
[
m
/
m
i
n
]
Surface
roughness [m]
I
n

f
e
e
d
d
i
r
e
c
t
i
o
n
C
r
o
s
s

t
o
f
e
e
d
d
i
r
e
c
t
i
o
n
1 A0B+15 234,6 0,176 0,317
2 352 0,191 0,341
3 A0B-15 234,6 0,189 0,435
4 352 0,195 0,498
5 A+15B0 234,6 0,335 0,357
6 352 0,367 0,491
7 A-15B0 234,6 0,317 0,360
8 352 0,192 0,411
281
Direction of inclination angle of tool axis
relative to the surface normal is important because
siginificant differences can arise in terms of
quality. Both, absolute value of the angle of
inclination and direction of this angle are very
important.
Variation of surface roughness when machined
OLC 45 at 15 degrees inclination angle value
applied in different direction is presented in Fig. 9
for measurements made in feed direction, and in
Fig.10. for measurements made perpendicular to
feed direction.
Figure 9: Variation of roughness for measurements
made in feed direction
made perpendicular to feed direction
The results releave the fact that it is
recomanded to avoid situation of inclination when
results contact between material machined and tool
edge were the effective cutting speed is reduce.
This situation appear when we inclined tool axis
around X axis in direction A positive and around
Y axis in direction B negative. In this case when
inclination angle value is 15 degrees it is
recomanded to not increase the cutting speed
because the surface roughness is worse, fact
observed for situation when we made
measurements in feed direction and perpendicular
to feed direction. It is possible that for another
value of inclination angle the surface roughness
variation to be different.
6. Conclusions
Experimental results show that direction of
inclination in ball end milling process have a great
influence upon surface roughness.
Using the direction of inclination angle around
the X axis with angle A positive or negative must
be avoid because the surface roughness is weaker
comparativ with direction in inclination made
around Y axis. Best results of surface roughness is
obtained when we incline tool axis around Y axis
with angle B positive. Difference between surface
roughness obtained with inclination angle in B
positive direction, whose the best situation, and
inclination in A positive direction is about 90% for
measurements made in feed direction and 40% for
measurements made perpendicular to feed
direction.
References
[1] Antoniadis, A., Bilalis, N., Savakis, C.,
Maravelakis, E. and Petropoulos, G., Influence
of machining inclination angle on surface
quality in ball end milling, Proceedings of
AMPT2003, Dublin, Ireland, 8-11 J uly 2003.
[2] Boujelbene, M., Moisan, A., Bouzid, W. and
Torbaty, S., Variation cutting speed on the five
axis milling, J ournal of Achievements in
Materials and Manufacturing Engineering,
Vol. 21, Issue 2, April 2007.
[3] Daymi, A., Boujelbene, M., Linares, M.,
Bayraktar, E. and Ben Amara, A., Influence of
workpiece inclination angle on the surface
roughness in ball end milling of the titanium
alloy Ti-6Al-4V, J ournal of Achiements in
Vol.35, Issue 1, J uly 2009.
[4] Fallbohmer, P., Rodriguez, C., A., Ozel, T. and
Altan, T., High-speed machining of cast iron
and alloy steels for die and mould
manufacturing, J ournal of Materials
Processing Technology 98, pp. 104-115, 2000.
[5] Kadirgama, K., Noor, M., Rahman, M., Rejab,
R., Haron, C. and Abou-El-Hossein, K.,
Surface roughness prediction model of 6061-
282
T6 Aluminium alloy machining using
statistical method, European J ournal of
Scientific Research, Vol.25, No. 2, ISSN 1450-
216X, 2009.
[6] Ko, T., J., Kim, H., S. and Lee, S., S.,
Selection of the machining inclination angle in
high-speed ball end milling, International
J ournal of Advanced Manufacturing
Technology 17, pp. 163-170, 2001.
[7] Krizbergs, J. and Kromanis, A., Methods for
prediction of the surface roughness 3D
parameters according to technological
parameters, 5
th
International DAAM Baltic
Conference ,, Industrial engineering-adding
innovation capacity of labour force and
entrepreneurs, Tallinn, Estonia, 20-22 April
2006.
[8] Rashid, M., Gan, S. and Muhamad, N.,
Mathematical modeling to predict surface
roughness in CNC milling, World Academy of
Science, Engineering and Technology 53,
2009.
[9] Schultz, H. and Hock, S., High-speed milling
of dies and moulds-cutting conditions and
technology, Annals of the CIRP 44, pp. 35-38,
1995.
[10]Tonshoff, H., K., and Hernandez, J ., C., Die
manufacturing by 5 and 3 axes milling, J ournal
of Materials Processing Technology 20, pp.
105-119, 1989.
[11]Vidakis, N., Antoniadis, A., Savakis, C. and
Gotsis, P., Simulation of ball end tools milling,
Proceedings of International Conference ICPR
2001, Prague, Czech Republic, 29.07-
03.08.2001.
277
EXPERI MENTAL RESEARCH ABOUT I NFL UENCE OF
I NCLI NATI ON ANGLE DI RECTI ON UPON THE SURFACE
ROUGHNESS I N BAL L END MI LL I NG OF OLC45 (C45) MATERI AL
PACA I oan
1
, LOBONI U Mircea
2
1
phd. student, North University of Baia Mare, Str. Dr. Victor Babes, nr.62 A , Baia Mare, Romania,
e-mail: pascaing@yahoo.com.
2
phd. prof., North University of Baia Mare, Str. Dr. Victor Babes, nr.62 A , Baia Mare, Romania,
e-mail: mircea.lobontiu@ubm.ro .
Abstract: Machining with ball end mills become a very usefull process to obtain pieces with complex
surfaces. Possibilities of machining with ball nose end mills represent a adequate process to achieve
surfaces with a good quality. That fact lead to exclude some finishing operations that was necessary if the
surfaces was machined with a different process and tools. Elimination of this finishing operation have a
great economic advantages, because milling with ball nose end mills offer a adequate surface roughness
and precision. But, all of this are possible only with a good choise of different values of cutting
parameters and with an optimum inclination angle represented by value and direction of inclination. The
present paper includes an experimental investigation of surface roughness depending on the value and
direction of tool axis inclination in raport with surface machined.
Keywords: roughness, surface, inclination angle, direction.
1. Introduction
Obtaining a high quality of surfaces machined
with ball end mills present a great interest for
reserchers and persons involved in machining of
die and mould surfaces. The surface quality in ball
end milling is the results of influence of a number
of factors such as: the stability of machine tools,
the correct choosing of cutting tools and cutting
parametters, properties of material machined and
the tool angle inclination upon the surface normal.
To have good results all these factors should be
chosen to have a minimal negative influence in the
process.
Machining of dies and moulds surfaces in
milling process involved polishing operations that
represent sometimes approximately two thirds of
the total manufacturing costs [4]. In this way the
milling process with ball nose end mills represent a
solution to reduce the costs of process and of final
product but only if the values of parameters
involved in process are adequate. Chosen of
optimum value of different parameters is dificult
because it is necesary to reduce the time of
machining and to obtain surfaces with good
roughness. In international context a great number
of researches is dedicated to surface quality
prediction. Some of them is dedicated to
experimental research, but a lot of them are
proposing some mathematical models. That
mathematical models proposed take into account
the cutting speed, feed value variation and the
following parameters: cutting depth [8], cutting
depth and radial depth [5], hardness of the material
and tool radius [7].
By using a cutting tool angle between the axis
of tool and machined surface can contribute to a
higher quality of machined surface [1].
The paper [9] and [6] suggested that a tool
inclination in the range of 1020 degrees
represents the optimum machining strategy for
high speed milling in the die and mould making
industry, but in other paper like [10] the optimal
value of inclination angle in ball end milling of
block materials was found at 15 degrees. Another
reserch show that when piece was machined with
inclination of tool axis of -17 degrees following Y
axis in one way the profile of the machined work-
piece, show the improvement of the machined
surface texture quality [2]. In paper [3] it is noted
that in practice the angle of 45 degrees for
inclination of tool axis the quality of surface is
low.
278
Because surface roughness have an important
effect upon product quality and economicity of
process, was necesary to study the influence of
inclination direction of tool axis upon surface
normal and try to compare the surface roughness
obtained in some situation.
2. The analysis of tool axis inclination
Machine tools with five axis lead the possibility
to work with one or two angle of tool axis
inclination in relation with normal to the surface,
situation that allow to achieve better cutting
conditions and therefore a high quality machined
surface. Working with this inclination angle the
contact zone between tool edge and material is
changing that involve another condition with
modification of chip thickness and area of contact.
Also, the cutting forces are different and the
energy consumed are different. Using this
inclination angle we have possibility to reduce the
costs of process and to obtain a higher quality.
Inclination angle can be applied to tool around
the axis X,Y or Z corresponding to the structural
characteristics of machine tools and to
characteristics of surfaced machined.
Inclination angles proposed for analysis in this
paper are around the X axis by angle A positive
(Fig.1) or negative (Fig.2) and around to Y axis by
angle B positive (Fig.3) or negative (Fig.4).
3. Effects of inclination angle upon effective
cutting speed
By using an inclination angle between tool axis
and normal to surface we change some
characteristics of process. One of them is effective
cutting speed thus value depending by effective
diameters. In order, the effective diameters depend
by cutting depth, radial depth and inclination
angle. Increasing of effective diameters involve the
increase of effective cutting speed.
To establish the value of effective speed
milling V
c-eff
coresponding to inclination of 15
degress angle value, we used relation (1) [3]:
1000
R
p
a R
arccos
n
sin D n
ef c
V
(
(
|
|
.
|
\
|
+ u t
=
(1)
with the following condition:
Figure 1: Inclination of tool axis in B negative direction
around Y axis
Figure 2: Inclination of tool axis in B positive direction
around Y axis
Figure 3: Inclination of tool axis in A negative direction
around X axis
279
Figure 4: Inclination of tool axis in A positive direction
around X axis
|
|
.
|
\
|
s u ( |
.
|
\
|
R
p
a R
arccos 90
n
R 2
e
a
arcsin , (2)
where:
-D nominal diameter of tool [mm];
-a
p
cutting depth [mm];
-n spindle speed [rot/min];
n
inclination of tool axis [grade];
-R tool radius [mm].
Geometrical parameters values used in our
experiments are presented in Table 1.
Table 1:Geometric parameters used
No. Geometric
elements
Units of
measurement
Values
1 Diameter of
ball end mill
mm 16
2 Cutting depth mm 0,2
3 Axial depth mm 0,2
4 Feed per
tooth
mm 0,1
5 Tool angle
inclination
degrees 15
The results obtained for two values of
speendle speed used in our experiments are
4. Experimental work and conditions
Machining by using a cutting tool angle
between the axis and machined surface can
contribute to a higher quality of machined surface
[1]. Analyzing the indication presented in paper
[9],[6],[10], [11] and [2] about optimum value of
inclination angle we choose to apply in our
experiment value of 15 degrees in raport with
normal to the surface.
Table 2:Effective cutting speed values
No. Speendle
speed n
[rot/min]
Inclination
angle
[degrees]
Effective
cutting
speed
[m/min]
1 10000 15 234,6
2 15000 352
Once identified the optimum value of
inclination angle we found that is necesary to
extend the studies by analysing the influence
of inclination angle direction upon surface
roughness. We want to establish the best
direction of inclination angle that lead to high
quality of machined surface, but according
with some defaults machining conditions.
4.1 Experimental setup
To make the experimental analysis of
inclination angle direction upon surface roughness
was necessary a number of experiment that was
made by using a OKUMA MU400VA five-axis
CNC milling machine equipped with a maximum
spindle speed of 15,000 rpm (Fig.5.), that was
propriety of S.C. RAMIRA S.A. We used a tool
holder SRM2160SNM and inserts SRG16C-
VP15TF made by Mitsubishi Carbide that is
presented in Fig.6.
Figure 5: Five-axis milling machine
280
Figure 6: Tool holder SRM2160SNM
4.2 Workpiece configuration
The material used was OLC-45 (C45) with
the following caracteristics: 0,420,50% C,
0,500,80% Mn, 0,170,37% Si, maximum
0,040% P etc. The workpiece had a economic
shape because we can machined six surfaces
with the same quatity of material (Fig.7.)
Figure 7: Workpiece configuration
4.3 Surface roughness measurements
In the technical literature and industry are
many parameters used to defined the surface
roughness, but in generally is defined as the
inherent irregularities of the workpiece affected by
machining process. The paper [3] show that the
most popular of the 2D parameters is the average
roughness Ra. In order to obtain more information
about surface machined we made the
measurements of surface roughness in feed
direction and perpendicular to it. We used the
tester TR200 that we applied on new device that
provides the conditions of perpendicularity and
parallelism between roughness tester feeler and
feed direction. That measurements stand is
presented in Fig. 8.
4.4 Experimental conditions
In order to establish the conditions utilized
in our experiments we must say that all
experiments were conducted using in feed milling
directionand with parameters indicated in Table 1,
2 and 3.
Figure 8: Roughness tester TR200 with adaptor device
Table 3: Experimental conditions
Nr Inclination
angle
[degrees]
Spindle
speed
[rot/min]
Inclination
direction
1
15
10000
A0B+15
2 A0B-15
3 A+15B0
4 A-15B0
5
15000
A0B+15
6 A0B-15
7 A+15B0
8 A-15B0
5. Experimental results and discussion
Experimental data on surface quality of the
measurements obtained were analyzed in feed
direction and perpendicular to feed direction for
two values of effective cutting speed and were
centralized in Table 4.
Table 4: Experimental data
Nr.
D
i
r
e
c
t
i
o
n

o
f
i
n
c
l
i
n
a
t
i
o
n
E
f
f
e
c
t
i
v
e
c
u
t
t
i
n
g

s
p
e
e
d
[
m
/
m
i
n
]
Surface
roughness [m]
I
n

f
e
e
d
d
i
r
e
c
t
i
o
n
C
r
o
s
s

t
o
f
e
e
d
d
i
r
e
c
t
i
o
n
1 A0B+15 234,6 0,176 0,317
2 352 0,191 0,341
3 A0B-15 234,6 0,189 0,435
4 352 0,195 0,498
5 A+15B0 234,6 0,335 0,357
6 352 0,367 0,491
7 A-15B0 234,6 0,317 0,360
8 352 0,192 0,411
281
Direction of inclination angle of tool axis
relative to the surface normal is important because
siginificant differences can arise in terms of
quality. Both, absolute value of the angle of
inclination and direction of this angle are very
important.
Variation of surface roughness when machined
OLC 45 at 15 degrees inclination angle value
applied in different direction is presented in Fig. 9
for measurements made in feed direction, and in
Fig.10. for measurements made perpendicular to
feed direction.
made in feed direction
made perpendicular to feed direction
The results releave the fact that it is
recomanded to avoid situation of inclination when
results contact between material machined and tool
edge were the effective cutting speed is reduce.
This situation appear when we inclined tool axis
around X axis in direction A positive and around
Y axis in direction B negative. In this case when
inclination angle value is 15 degrees it is
recomanded to not increase the cutting speed
because the surface roughness is worse, fact
observed for situation when we made
measurements in feed direction and perpendicular
to feed direction. It is possible that for another
value of inclination angle the surface roughness
variation to be different.
6. Conclusions
Experimental results show that direction of
inclination in ball end milling process have a great
influence upon surface roughness.
Using the direction of inclination angle around
the X axis with angle A positive or negative must
be avoid because the surface roughness is weaker
comparativ with direction in inclination made
around Y axis. Best results of surface roughness is
obtained when we incline tool axis around Y axis
with angle B positive. Difference between surface
roughness obtained with inclination angle in B
positive direction, whose the best situation, and
inclination in A positive direction is about 90% for
measurements made in feed direction and 40% for
measurements made perpendicular to feed
direction.
References
[1] Antoniadis, A., Bilalis, N., Savakis, C.,
Maravelakis, E. and Petropoulos, G., Influence
of machining inclination angle on surface
quality in ball end milling, Proceedings of
AMPT2003, Dublin, Ireland, 8-11 J uly 2003.
[2] Boujelbene, M., Moisan, A., Bouzid, W. and
Torbaty, S., Variation cutting speed on the five
axis milling, J ournal of Achievements in
Vol. 21, Issue 2, April 2007.
[3] Daymi, A., Boujelbene, M., Linares, M.,
Bayraktar, E. and Ben Amara, A., Influence of
workpiece inclination angle on the surface
roughness in ball end milling of the titanium
alloy Ti-6Al-4V, J ournal of Achiements in
Vol.35, Issue 1, J uly 2009.
[4] Fallbohmer, P., Rodriguez, C., A., Ozel, T. and
Altan, T., High-speed machining of cast iron
and alloy steels for die and mould
manufacturing, J ournal of Materials
Processing Technology 98, pp. 104-115, 2000.
[5] Kadirgama, K., Noor, M., Rahman, M., Rejab,
R., Haron, C. and Abou-El-Hossein, K.,
Surface roughness prediction model of 6061-
282
T6 Aluminium alloy machining using
statistical method, European J ournal of
Scientific Research, Vol.25, No. 2, ISSN 1450-
216X, 2009.
[6] Ko, T., J., Kim, H., S. and Lee, S., S.,
Selection of the machining inclination angle in
high-speed ball end milling, International
J ournal of Advanced Manufacturing
Technology 17, pp. 163-170, 2001.
[7] Krizbergs, J. and Kromanis, A., Methods for
prediction of the surface roughness 3D
parameters according to technological
parameters, 5
th
International DAAM Baltic
Conference ,, Industrial engineering-adding
innovation capacity of labour force and
entrepreneurs, Tallinn, Estonia, 20-22 April
2006.
[8] Rashid, M., Gan, S. and Muhamad, N.,
Mathematical modeling to predict surface
roughness in CNC milling, World Academy of
Science, Engineering and Technology 53,
2009.
[9] Schultz, H. and Hock, S., High-speed milling
of dies and moulds-cutting conditions and
technology, Annals of the CIRP 44, pp. 35-38,
1995.
[10]Tonshoff, H., K., and Hernandez, J ., C., Die
manufacturing by 5 and 3 axes milling, J ournal
of Materials Processing Technology 20, pp.
105-119, 1989.
[11]Vidakis, N., Antoniadis, A., Savakis, C. and
Gotsis, P., Simulation of ball end tools milling,
Proceedings of International Conference ICPR
2001, Prague, Czech Republic, 29.07-
03.08.2001.
283
PARTI CLES DIMENSI ONAL ANALYSI S AND MI CROSCOPI C
CHARACTERI ZATI ON OF HYDROXYAPATI TE POWDER
Aurora Anca Poinescu
1
, Rodica Mariana I ON
2,3
1
Valahia University of Targoviste, 2 Regele Carol 1 Street, poinescua@yahoo.com,
2
Valahia University of Targoviste, 2 Regele Carol 1 Street, rodica_ion2000@yahoo.co.uk
3
ICECHIM, Bucharest, Analytical Department, 202 Splaiul Independentei, Bucharest-060021,
Romania.
Abstract: The studies conducted for this paper we chose the wet precipitation synthesis method at
room temperature to obtain hydroxyapatite powder. Size particle analysis of hydroxyapatite was made in
isopropyl alcohol. The surface study on nm scale and the surface topography were evaluated by using
AFM with tapping mode. The SEM pictures of the grown spherulite crystals shows many spherical
agglomerations and few crystallites of 0.1 m in size with pores.
Keywords: SEM and AFM microscopy, hydroxyapatite powder, wet precipitation.
1. Introduction
A challenging problem in health care is the
regeneration of partially or fully lost organs.
There are two types of regeneration:
physiological and reparative.
Physiological regeneration is a natural
process of disintegration and restoration of
molecules, cells, and tissue (e.g., bone).
Reparative regeneration in the broad sense
is the restoration of parts of body lost as a result of
traumas, healing of defects of tissue and organs,
etc.
The use of calcium-phosphate-based
biocompatible materials close in composition to
the inorganic component of bone tissue contributes
to the creation of favorable conditions for the
reparative functions of bone tissue.
In the past several decades, a number of implant
materials based on calcium hydroxyapatite
(Ca
10
(PO
4
)
6
(OH
2
), calcium phosphate (-
Ca
3
(PO
4
)
2
, ceramics of these phosphates,
bioglasses, and composites have been used in
orthopedics, neurosurgery, and dentistry.
In the case of fractures, the restoration of bone
tissue is a complicated biochemical process, which
includes the synthesis of proteins and nitrogen-
containing polysaccharides and deposition of
calcium salts and is accompanied by
hypermetabolism.
A serious problem related to the
biocompatibility of materials is the creation of
implants that would ensure the delivery and in situ
release of a therapeutic agent with osteoinductive
and anti-inflammatory effects [1].
Hydroxyapatite surface coating applications
include metallic orthopedic and dental implants
where osseointegration HA promotes both
processes and reduce the release of metal ions
acting as a physical barrier, preparation for the
replacement of bone fragments. [2].
2. Materials and methods
The synthesis methods of hydroxyapatite are
many, the literature abounds in this area. The
studies conducted for this paper we chose the wet
precipitation synthesis method at room
temperature.
Figure 1 shows the flow of technology of wet
precipitation of hydroxyapatite powder by Sung
[3].
284
Figure 1: Modified chemical precipitation route for HA
powder preparation. (Adapted fromSung) [3]
Chemical reagents used in this paper: Ca
(NO
3
)
2
.4H
2
O and (NH
4
) 2HPO
4
were separately
dissolved in deionized water. He added Ca(NO
3
)
2
in aqueous solution (NH
4
) 2HPO
4
, agitated
vigorously for about 1 hour at room temperature
and a precipitate was obtained milky, somewhat
gelatinous, which was then shaken for 1h.
The mixture was sintered at 1000
0
C for 24
hours. After having been washed and filtered.
After filtering the sticky product was compacted
dry oven at 800
0
C. Dried powder was crushed in a
mortar with pestle and then calcined in an alumina
crucible at three different temperatures 800
0
C,
1000
0
C and 1200
0
C for 1h.
Table 1
Method Sampl
e
Temp.
(
O
C)
Concentr.
(M)
Reflux after
mixing
HA1 800 0.1:0.06
Reflux after
mixing
HA2 1000 0.1:0.06
Reflux after
mixing
HA3 1200 0.1:0.06
3. Results
The objectives of this work were:
to synthesize biocompatible hydroxyapatite;
to analyze the size particle of hydroxyapatite;
to characterize by SEM and AFM
microscopy hydroxyapatite powder obtained
by wet precipitation.
3.1. Size particle analysis of hydroxyapatite
in isopropyl alcohol
Determination of average particle size and
particle size distribution will be done using a laser
diffraction particle size analysis by Malvern
Mastersizer Type S-type. It offers a wide range of
measurement in the range 0.05 to 3500 microns,
excellent data reproducibility and flexibility in
handling evidence.
In the case of HA-A powder disperses not
enough to be mechanically made a DLS
measurement type. Apparent dispersions are
composed of a particulate phase and no detectable
perfectly clear, that the precipitating phase.
Precipitation is too fast and does not allow the
measurement.
So they resorted to grinding (in agate) to
precipitate directly in the presence of alcohol.
After the samples were ultrasonic in an ultrasonic
bath for 5 min (35kHz). Dispersions have
developed an opalescent appearance of white
precipitate respectively.
After several minutes, large aggregates have
settled, and then were collected samples from the
metastable phase (upper).
That is metastable for about an hour the
precipitate as well. The average size (math) is
about 1800nm (1,8 microns).
But the physical interpretation is quite different,
namely: Record several populations of particles.
0
5
1 0
1 5
2 0
2 5
0 . 1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0
I
n
t
e
n
s
it
y

(
%
)
S iz e ( d . n m)
S iz e D is t rib u tio n b y In te n s it y
R e c o r d 1 : H A - A 1 R e c o rd 2 : H A - A 2 R e c o r d 3 : H A - A 3
Figure 2: Size distribution of the three HAp samples
after intensity
285
Considering the first instability manifested by
a set of measurements to another (see the three sets
of measurements per color), but also change the
number of population (number of peaks) is very
likely that the system be reviewed format of
elementary particles in different forms of
aggregation result from the process of
sedimentation. The system contains over 6000 nm
particles (the ultimate maximum).
0
1 0
2 0
3 0
4 0
0 . 1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0
V
o
l
u
m
e

(
%
)
S iz e ( d . n m)
S iz e Dis tr ib u t io n b y V o lu me
R e c o r d 1 : HA - A 1 R e c o rd 2 : H A - A 2 R e c o r d 3 : HA - A 3
Figure 3: Size distribution of the three HAp samples by
volume
In the figures 2, 3 and 4 are illustrated the
distributions of the three Hap samples after
grading intensity, volume and number
respectively.
0
1 0
2 0
3 0
4 0
0 . 1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0
N
u
m
b
e
r

(
%
)
S iz e ( d . n m)
S iz e Dis tr ib u tio n b y N u mb e r
R e c o rd 1 : H A - A 1 R e c o rd 2 : H A - A 2 R e c o rd 3 : H A - A 3
Figure 4: Size distribution of the three HAp samples by
number
After global analysis of the system (above) has
developed software filtering of unit interval
analyzed in view of highlighting the possible
elementary particles between 0.6 - 100nm which in
the global analysis cannot be registered because of
the large aggregates shield. Such particles have
emerged about 40 nm single mode. These in a
different associations with a different kinds of
aggregates generates large majority in the global
analysis above.
0
2 0
4 0
6 0
8 0
10 0
1 1 0 1 0 0 10 0 0 10 0 00
I
n
t
e
n
s
it
y

(
%
)
S iz e ( d . nm)
S i z e D i s tr i b uti on b y Inte n s i ty
R e c or d 6 : H A - A ( 0 . 6 - 1 0 0 nm) 3
Figure 5: Size distribution of the HA-A samples by
intensity
0
1 0
2 0
3 0
4 0
5 0
1 1 0 1 0 0 1 0 0 0 1 0 0 0 0
V
o
l
u
m
e

(
%
)
S iz e ( d . n m)
S iz e Dis tr ib u tio n b y V o lu me
R e c o r d 6 : H A - A ( 0 . 6 - 1 0 0 n m) 3
volum
0
1 0
2 0
3 0
4 0
5 0
1 1 0 1 0 0 1 0 0 0 1 0 0 0 0
N
u
m
b
e
r

(
%
)
S iz e ( d .n m)
S iz e D is tr ib u tio n b y N u mb e r
R e c o rd 6 : H A - A ( 0 .6 - 1 0 0 n m) 3
number
For HA-A has conducted a global analysis.
Reproducibility is better between the sets of
286
measurements, but measured particles seem more
independent particles and not aggregates.
3.2. SEM microstructural characterisation
Microstructural observations were performed
with a Quanta 200 scanning electron microscope
equipped with software analyzer for quantitative
elemental analysis. For analysis by scanning
electron microscopy, hydroxyapatite powders were
placed in LR White resin.
Figure 8: Electron microscopy SEM for Hap powder
calcined at 1000
0
C, 2000 X.
In Figure 8 shows electron microscopy of
hydroxyapatite sample calcined at 1000
0
C with
2000x magnification and in the next figure same
sample but at a 1000x magnification. We can
distinguish micro pores (<10 m), allowing
diffusion of ions and fluid macro pore (100-
600m) to allow cell colonization. The micro
pores give the ceramics osteoconductive
properties.
calcined at 1000
0
C, 1000 X
SEM images of crystals grown shows several
spherical clusters and few crystals of 0.1m.
calcined at 1000
0
C, 10000 X
Subgrains size it is about 70 nm which
corresponds to the synthetic HA powder size [5].
For a good correlation, SEM images show several
clusters of high spherical spherulites and
crystallites less than 0.1m in size.
3.3. AFM microstructural characterisation
Investigations by Atomic Force Microscope
(AFM) were performed with an Agilent 5500 SPM
system, described by PicoSPM controlled by a
MAC Mode module and interfaced with a
controller PicoScan from Agilent Technologies,
Tempe, AZ, USA.
All AFM measurements (256 samples / line
256 lines) were made by scanning the surface at a
rate of 0.8 - 1.2 lines per second and were
conducted at room temperature, the mode of
palpation.
For investigation with the atomic force
microscope, HA solutions were prepared fresh
before each experiment by suspending an
appropriate amount of each sample in ethanol.
Scanning movement is led by a piezoelectric
scanner which scans the tip in a raster pattern on
the sample (or scans the sample on top).
AFM revealed a rough surface architecture for
HA, the predominant size of grains being in the
range of 90 - 100nm.
287
Figure 11: AFM image of the HA powder calcined at
1000
0
C
AFM is conducted in an ambient gas or liquid
medium and ethanol in our case. AFM shows the
architecture for the HA surface, the grain is large
in the range 70-100nm. Crystal size and size
distribution reached critical nucleus depends on the
size, in conditions of super saturation, rather than a
standard crystal growth, because aggregation of
very small particles was observed.
Atomic force microscopy (AFM) in comparison
with the (SEM) can be measured in all three
dimensions (x, y, and z) with a single scan. 3-D
surface topography was recorded on an area of 0.5-
0.5 mm
2
.
Figure 12: 3-D representation of the HA crystals
obtained by wet precipitation
Figure 13: AFM analysis of hydroxyapatite powder
4. Conclusions
Hydrothermal techniques give hydroxyapatite
powders with a high degree of crystallinity and
better stoichiometry having a wide distribution of
crystal sizes. Nanometer sized crystals can be
obtained at temperatures lower than 100 C with
precipitation techniques.
The chosen methods was the wet precipitation
because the wet-chemical precipitation route is the
most talented route owing to its ease in
experimental operations, low working temperature,
high percentages of pure products and inexpensive
equipment requirement.
In the SEM images of HA, was depicted small
crystals (<100 nm) in the agglomerated particles
and the uniform grain size with a narrow size
distribution corresponding to an improved
crystallinity of HA powders, especially for that
sample after calcination at 1000C for 1h. For a
good correlation, the SEM pictures of the grown
spherulite crystals show many spherical
agglomerations and few crystallites of 0.1 m in
size.
The determination of size distribution of the
grown spherulite and sintered HA materials were
carried out by atomic force microscopy.
At higher temperature the deagglomeration of
bulk phases and agglomeration of nano phases
leads to the nano crystalline HA in this present
study.
The crystal size distribution attained depends
on the size of the critical nucleus under the
supersaturation condition, rather than on standard
288
crystal growth, since the aggregation of very small
particles was observed.
In 3-D graphical representation of
hydroxyapatite crystals (Fig. 13) can be seen on
the grain growth direction z due to the sintering
process (wet precipitate) and particularly the
calcination temperature.
One of the main attractions at the AFM is the
ability to show high-resolution surfaces in the
liquid insulation. Since AFM does not rely on the
conductivity, and image scanning mechanism will
not be disturbed by the presence of liquid
References:
[1] Zakharov N. A., Polunina I. A., Polunin K.
E., Rakitina N. M., Kochetkova E. I.,
Sokolova N. P., and Kalinnikov V. T.,
Calcium Hydroxyapatite for Medical
Applications, Inorganic Materials, Vol. 40,
No. 6, 2004, pp. 641648.
[2] Monmaturapoj N., Nano-size
Hydroxyapatite Powders Preparation by
Wet-Chemical Precipitation Route, J ournal
of Metals, Materials and Minerals. Vol.18
No.1 pp.15-20, 2008
[3] Sung, Y., Lee, J . and Yang, J .,
Crystallization and Sintering
Characteristics of Chemically Precipitated
Hydroxyapatite Nanopowder. J . Cryst.
Growth. 262 : 467-472, 2004.
[4] Lazic, S., Zec, S., Miljevic, N., Milonjic, S.,
The Effect of Temperature on the
Properties of Hydroxyapatite Precipitated
FromCalciumHydroxide and Phosphoric
Acid, Thermochim.Acta, Volume 374,
Issue 1, pp. 13-22; 2001;
[5] Poinescu, A.A., Ion, R.M., Trandafir, I.,
Bacalum, E., Radovici, C., Obtaining and
characterization of a calcium
hydroxyapatite, The XV-th International
Scientific Conference Tehnomus, May 8-
9, (2009), Suceava Romania;
[6] Santos, M.H., Oliveira, M., Palhares de
Freitas Souza, P., Sander M.H., Gerais,
W.L., Synthesis Control and
Characterization of Hydroxyapatite
Prepared by Wet Precipitation Process,
Materials Research, Vol. 7, No.4, 625-630,
(2004).
289
OVERVI EW OF LATEST MI NERAL CARBONATI ON
TECHNI QUES FOR CARBON DI OXI DE SEQUESTRATI ON
Bodor Marius
1
, Vlad Maria
1
, Balttefan
1
1
Universitatea Dunrea de J os din Galai,
bodormarius@yahoo.com, maria.vlad@ugal.ro, stefanbalta@yahoo.com
Abstract: This relatively new domain, of carbon dioxide sequestration for mitigation purposes, is one
of the carbon dioxide capture and storage (CCS) alternatives. CO
2
sequestration could be used in the
future where geological storage, for example, is not an option due to the lack of suitable locations or
could represent a hazard to the surrounding areas. Also could be used in those locations where the needed
material (raw minerals or industrial wastes) are abundant.
The interest in this technique of CO
2
storage is growing fast and results of researches on this topic are
coming from a larger number of countries every year. What is more interesting is that the research of the
industrial wastes carbonation is developing, giving in the future new purposes for these wastes.
Keywords: mineral carbonation, carbon dioxide, calciumcarbonate, storage
1. Introduction
The greenhouse effect is known to be created
by different gases in the atmosphere. One of these
gases is CO
2
and because its concentration is
increasing dangerously for the past one hundred
years is now considered to be the main cause of the
temperature rise on Earth.
In order to avoid the potentially devastating
consequences of global warming and climate
change, the CO
2
emissions into the atmosphere
caused by human activities should be reduced
considerably [1]. For this goal to be reached some
carbon dioxide capture and storage (CCS)
technologies were proposed but for the moment
none of them were efficient enough in reducing the
CO
2
emissions. Despite the negative results,
researches are still ongoing and in some cases the
results are promising as for example the storage in
underground cavities. The same is hoped to be
happening with the CO
2
storage in minerals and
wastes and hopefully all the CCS technologies will
contribute together to a more efficient CO
2
reduction.
2. Carbon dioxidesequestration techniques
2.1. Geological storage
One of the technologies, that have already been
employed on a significant scale, but not large
enough to have a global CO
2
emissions mitigation
impact, is storage of CO
2
in underground cavities.
This includes the so-called Enhanced Oil Recovery
(EOR) and also Enhanced Gas Recovery (EGR),
which are concepts aimed at improving the oil/gas
recovery potential of an oil/gas field by flooding it
with CO
2
[2].
Perhaps the greatest problem related to
underground storage is the permanency of the
solution, as there will always be a risk of leakage.
Therefore, this solution would require continuous
monitoring of storage sites for thousands of years.
2.2. Ocean storage
Another widely studied option for CO
2
sequestration involves injecting CO
2
into the ocean
at great depths, where the gaseous CO
2
reacts to
form carbonic acid (H
2
CO
3
). The carbonic acid
then dissociates into a (bi)carbonate ion and
hydrogen ion in accordance with the equation
below: [2]
CO
2
(g) +H
2
O(l) H
2
CO
3
(aq) HCO
-
3
+H
+
CO
2-
3
+2H
+
(1)
Although ocean storage could provide a fast
and relatively easy alternative for CO
2
emissions
reduction it has lost its appeal in recent years,
largely due to the uncertainty when considering the
290
environmental consequences (e.g. decreasing pH
of ocean water) and the lack of permanency [1].
2.3. Storage below sea bed
An alternative to both geological storage and
ocean storage described above is CO
2
storage
below the ocean floor at depths of at least 3000 m
of ocean and several hundred meters of marine
sediment. In contrast to the previously mentioned
options this option does not suffer from lack of
permanency (ocean storage) or the demand for
monitoring the storage site (geological storage).
The idea is based on the fact that CO
2
becomes
denser than water at sufficient depths (3000 m),
but that it still needs to be trapped in order to
prevent it from being released by ocean currents or
e.g. earthquakes. Therefore, it should be stored
below the seabed [3].
This alternative is still new and further research
is ongoing in order to verify the theories.
2.4. Mineral carbonation
The reaction between a metal oxide bearing
material and CO
2
is called carbonation and can be
expressed by the following reaction:
MO +CO
2
MCO
3
+heat (2)
where in practice M describes a (metallic) element
such as calcium, magnesium or iron. The reaction
in Equation (2) is exothermic and the heat released
is dependent on the metallic element bearing
mineral at hand (for the magnesium- or calcium-
based silicate minerals- olivine: 89 kJ /mol CO
2
,
serpentine: 64 kJ /mol CO
2
and wollastonite: 90
kJ/mol CO
2
at 298 K). [2]
One major benefit of CO
2
sequestration by
mineral carbonation consist of the environmentally
benign and virtually permanent trapping of CO
2
in
the form of carbonated minerals by using abundant
mineral resources such as Mg-silicates [2]. Unlike
other CO
2
sequestration routes it provides a
leakage-free long-term sequestration option,
without a need for post-storage surveillance and
monitoring once the CO
2
has been fixed.
In addition to the benefits of mineral
carbonation, this option is the only CO
2
sequestration option available where large
underground reservoirs do not exist and ocean
storage of CO
2
is not feasible.
Another benefit of mineral carbonation is that,
at least theoretically, the carbonation process could
proceed without energy input, but this has not yet
been accomplished.
Attempts to speed up the carbonation reaction
include using both dry and wet methods, additives,
heating and pressurizing the carbonation reactor,
dividing the process into multiple steps,
pretreatment of the mineral source and more
(fig.1).
There are several different elements that can be
carbonated, but alkaline earth metals, calcium and
magnesium, have proven to be the most suitable
due to their abundance and insolubility in nature.
Figure 1: Main carbonation processes and variants.
291
In addition to the abundant magnesium and
calcium containing minerals, there are also
industrial solid residues that contain large amounts
of Mg, Ca and even Fe.
Currently the most investigated mineral
resources are olivine, serpentine and wollastonite.
From the solid residues, steel slag has recently
received a lot of attention but other industrial
residues also have been studied: asbestos-mining
tailings, electric arc furnace (EAF) dust, cement-
kiln dust, waste concrete, coal fly ash etc.
Figure 1 displays the various carbonation routes
that are currently being investigated.
3. Directions of mineral carbonation
Trapping carbon dioxide in carbonates can be
achieved through various process routes as
described further, ranging from the most basic
accelerated weathering of limestone to advanced
multi-step processes.
3.1. Direct carbonation
Direct carbonation is the simplest approach to
mineral carbonation and the principal approach is
that a suitable feedstock, e.g. serpentine or a
Ca/Mg rich solid residue is carbonated in a single
process step. For an aqueous process this means
that both the extraction of metals from the
feedstock and the subsequent reaction with the
dissolved carbon dioxide to form carbonates takes
place in the same reactor.
3.1.1. Direct gas-solid carbonation
Gas-solid carbonation is a simple approach
towards mineral carbonation. Here particulate
metal oxides are brought into contact with gaseous
CO
2
at a particular temperature and pressure. The
dry process has the potential of producing high
temperature steam or electricity while converting
CO
2
into carbonates.
Unfortunately, the reaction rates of such a
process have been too slow and the process suffers
from thermodynamic limitations and further
studies around this alternative have mostly been
abandoned.
One of the important benefits of using industrial
solid residues as feedstock for carbonation
compared to the carbonation of mineral ores is the
possibility of utilizing a waste stream. The
possibility of simultaneously binding CO
2
and
lowering the hazardous nature of e.g. municipal
solid waste incinerator (MSWI) ash makes this
carbonation route interesting [4]. However, the
potential CO
2
storage capacity for this option is
limited, simply because the amounts of material
that may be carbonated are too small [4].
Nevertheless, where these wastes are present
the carbonation should be taken into consideration
because as shown by R. Santos et al. [5] during
the carbonation the strength of the stainless steel
slag is increased. This will give an added value to
the carbonated stainless steel slag wherefore the
material can be used in secondary applications.
The sale price of this product might increase.
3.1.2. Direct aqueous carbonation
The direct aqueous mineral carbonation-route
referring to carbonation preformed in a single step
in an aqueous solution, appears to be the most
promising CO
2
mineralization alternative to date
[6]. High carbonation degrees and acceptable rates
have been achieved but the process is (still) too
expensive to be applied on a larger scale [7].
Ranging from 4080 /t CO
2
mineralized (includes
energy use) compared to 0.46 /t CO
2
[2] stored
for geological storage.
Direct aqueous mineral carbonation can be
further divided into two subcategories, depending
on the type of solution used. Studies focusing on
carbonation in pure aqueous solutions have quickly
made way for additive-enhanced carbonation
experiments and a common solution type used
today, originally presented by OConnor et al.
(2000) [8], consists of 0.64 M NaHCO
3
+ 1.00 M
NaCl. However, it has been reported [9] that there
are still improvements to be made regarding the
above mentioned solution.
When it is necessary to use additives in
carbonation processes it is extremely important to
recycle these, due to the large scale of any
industrial application [7].
Work on finding optimal aqueous carbonation
conditions is ongoing and even though it has been
studied extensively, some questions still remain
unanswered. For example increasing the
liquid/solid ratio (L/S) has been reported to have
both a positive and a negative effect on CO
2
conversion.
The advances made to aqueous solution
chemistry by McKelvy et al. [9] were significant,
292
but unless the (expensive) additives used cannot be
recycled the process route becomes unattractive.
Nevertheless, the studies conducted on direct
aqueous carbonation have improved the overall
knowledge of aqueous carbonation reactions
considerably.
3.2. I ndirect carbonation
When the mineral carbonation process is
divided into several steps it is classified as indirect
carbonation. In other words, indirect carbonation
means that the reactive component (usually Mg or
Ca) is first extracted from the feedstock (as oxide
or hydroxide) in one step and then, in another step,
it is reacted with CO
2
to form the desired
carbonates.
3.2.1. Multistage gas-solid carbonation route
Direct gas-solid carbonation of silicate minerals
has been shown to be too slow for any large scale
implementations, but a staged gas-solid
carbonation process could overcome the slow
reaction kinetics. The process involves extraction
of magnesium (oxide or hydroxide) in an
atmospheric pressure step followed by a
carbonation step at elevated temperature (>500 C)
and pressure (>20 bar) [10].
It has been found that the carbonation of MgO
is significantly slower than the carbonation of
Mg(OH)
2
[11]. Using this observation Zevenhoven
et al. [10] suggested (noting that Mg(OH)
2
production from serpentine in one step cannot be
done because of thermodynamic limitations) that
the direct gas-solid carbonation process should be
divided into three-steps: 1) MgO production
(Equation 3) in an atmospheric reactor followed by
2) MgO hydration (Equation 4) and 3) carbonation
(Equation 5) at elevated pressures:
Mg
3
Si
2
O
5
(OH)
4
(s)
3MgO(s) +2SiO
2
(s) +2H
2
O (3)
MgO(s) +H
2
O Mg(OH)
2
(s) (4)
Mg(OH)
2
(s) +CO
2
MgCO
3
(s) +H
2
O (5)
In addition to the faster carbonation kinetics in
the three-step gas-solid carbonation route
described above, the process is also preferable
from an energy efficiency point of view compared
to the two step carbonation of MgO [10]. However,
the three-step process is still too slow for large
scale implementation, as preliminary tests
performed in up to 45 bar pressures have shown.
Progress has been made in enhancing the direct
gas-solid reaction rates, primarily by increasing
pressure. Experiments showing that all three-steps
in Equations 35 are fast enough for industrial
implementation are still required.
Dividing the gas-solid carbonation route into
several steps could be beneficial, but there is not
enough evidence yet for industrial viability.
3.2.2. Acetic acid route
In order to speed up the aqueous carbonation
process, the use of acetic acid for the extraction of
calcium from a calcium-rich feedstock has been
suggested by Kakizawa et al. [12]. In principal it
consists of two-steps as given in Equations 6 and
7:
CaSiO
3
+2CH
3
COOH
Ca
2+
+2CH
3
COO
+H
2
O +SiO
2
(6)
Ca
2+
+2CH
3
COO
+CO
2
+H
2
O
CaCO
3
+2CH
3
COOH (7)
Equation 6 describes the extraction step and
Equation 7 the precipitation step. In principal the
acetic acid used in the extraction step could be
recovered in the following precipitation step.
Inspired by the concept of binding CO
2
in
calcium extracted from a calcium silicate such as
wollastonite using acetic acid [12], Teir et al. [13]
investigated the possibility of producing a high
value PCC material from calcium silicates and
later from other calcium-containing materials [14].
Steelmaking slags then became the centre of
attention as they can contain significant amounts of
both CaO and MgO. Eloneva S. [15] reported that
8090 % pure calcite was produced from blast
furnace slag using acetic acid. However,
significant amounts of sodium hydroxide were
required for promoting the precipitation of
carbonates from the acidic solution.
Research in Finland has focused on steel slag
carbonation and especially the possibility of
producing valuable precipitated calcium carbonate
(PCC) [14, 15]. The annual CO
2
binding potential
of steelmaking slags is in the order of 70 180 Mt
CO
2
. Globally the CO
2
sequestration potential for
this option is small, but for individual steel plants,
293
however, the method could provide significant
economical benefits.
3.2.3. Two step aqueous carbonation
Two-step aqueous carbonation has been
investigated because the overall carbonation
reaction can easily be divided into two-steps:
extraction and precipitation that may be
investigated and optimized separately.
By upgrading a waste product into a product of
high commercial value, expensive CO
2
sequestration processes could become
economically feasible. One such approach has
been investigated by Katsuyama et al. [16] who
studied the use of waste cement for the
development of high-purity CaCO
3
by CO
2
carbonization.
Katsuyama et al. [16] studied the feasibility of
producing CaCO
3
from waste cement by first
extracting calcium from pulverized waste cement
in a water slurry at high CO
2
pressure (several
MPa), followed by the precipitation of CaCO
3
from the extracted solution at lower CO
2
pressures,
producing high purity CaCO
3
(up to 98%) from
waste cement at relatively high reaction rates. The
estimation of the cost of producing high-purity
CaCO
3
could be as low as 105 /m
3
when
compared to the commercial price of 154269
/m
3
. In addition, if the produced CaCO
3
could be
purified to meet the requirements of ultra-high
purity CaCO
3
(>99% CaCO
3
) the potential profits
could increase substantially. The current cost of
ultrahigh purity CaCO
3
is around 7,700 /m
3
, while
Katsuyama et al. [16] estimated a production cost
of only 250 /m
3
.
Gorset et al. [17], describing a way of
producing pure MgCO
3
from olivine, claims that
the process consisting of one dissolution step and
two precipitation steps is rapid enough for large
scale implementation. The process does not require
the use of strong mineral or organic acids even
though the dissolution step requires an acidic
environment. The required acidity (pH 35) is to
be achieved using pressurized CO
2
(50150 bar)
and a temperature around 100170 C, while the
following step consisting of MgCO
3
precipitation,
takes place in another reactor with preferably a
lower CO
2
pressure (50 80 bar) and a higher
temperature (140 250 C) favoring the
precipitation of carbonates. Experimental results
showed a high degree of purity, between 99.28 and
99.44% MgCO
3
, of the precipitated carbonate.
The pH-swing process developed in Japan is
another two-step aqueous carbonation process
where at first the pH of the solution is lowered
thereby enhancing the extraction of divalent metal
ions. In the second step the pH is raised to enhance
the precipitation of carbonates.
The principal reactions taking place inside the
extractor (Equation 8) and the precipitator
(Equation 9) are:
4NH
4
Cl +2CaOSiO
2
2CaCl
2
+4NH
3
+2H
2
O (8)
4NH
3
+2CO
2
+2H
2
O +2CaCl
2
2CaCO
3
+4NH
4
Cl (9)
Equation 9, taking place inside the precipitator,
consists of both CO
2
absorption and CaCO
3
precipitation. In their study, Kodama et al. [18]
investigated a CO
2
sequestration process that
utilizes pH swing using NH
4
Cl. The energy input
requirement for the investigated process using steel
making slag as the mineral source was estimated at
around 300 kWh/t CO
2
, but the loss of a chemical
additive (NH
3
) was considerable.
Different approaches of a two-step aqueous
carbonation process have been presented. All
options are good in theory, but it remains uncertain
whether or not these processes could lead the way
to any significant scale long-term storage of CO
2
in
the future. More experiments for large scale
viability are required.
3.3. Other processes of carbon dioxide
sequestration
Beside the process routes mentioned above,
there are other processes and applications that
resemble mineral carbonation and are described in
the following section.
3.3.1. The precipitated calcium carbonate
production
The production of valuable products (e.g. PCC)
by utilizing CO
2
was studied extensively. An
example of this concept is Two-step aqueous
carbonation of solid residues. Various methods to
obtain a product of desired properties have been
used and one of the simplest methods is that of
direct aqueous carbonation without the use of
additives.
294
3.3.2. Brines used in carbonation
The saline-based solution formed as a waste
product during the oil or natural gas extraction is
called brine and it is found in large quantities in
storage tanks. Due to its large amounts available
and a high concentration of elements fit to form
carbonates (mostly calcium and magnesium) this
brine represents an option for carbon dioxide
storage by a carbonation process. Although this
brine is capable of forming carbonates, its use in a
large scale process is for the moment not
appropriate due to the slow kinetics. If the pH of
the brine is raised the carbonation process is faster,
but unknown factors about the parameters
(temperature, pressure, brine composition and pH)
need to be further studied.
3.3.3. Accelerated weathering of limestone
Another option is carbon dioxide capture and
storage by accelerated weathering of limestone
(AWL). This option imitates the natural carbonate
weathering according to the following reaction:
CO
2
(g) +H
2
O(l) +CaCO
3
(s)
Ca
2+
(aq) +2HCO
3
(aq) (10)
The product of an AWL plant would be a
calcium bicarbonate solution that could readily be
released and diluted into the ocean with a minimal
environmental impact [19]. However, there are still
many issues to deal with, such as the energy
demand of transporting large amounts of calcium
containing (waste or mineral) material to the AWL
plant that preferably should be located near a CO
2
point source as well as a possible disposal site (e.g.
the ocean). In an ideal case (with access to free
limestone, e.g. waste fines, and a free water
source, e.g. power plant cooling water) the CO
2
mitigation cost by means of AWL could be as low
as 2.33.1 /ton CO
2
. Rauet al. [19] suggests that
some 10-20% of the United States point-source
CO
2
-emissions could be mitigated this way.
Despite the potential positive effect of
bicarbonate disposal Rau et al [19] concludes that
further research is needed to fully understand the
impacts of AWL effluent disposal in the ocean.
Some of the most important applications
and techniques for mineral storage of CO
2
are
described above, yet almost all of these
options need further research before starting a
large scale process. Despite the inconvenience
given by results lacking, all of them should
represent an option when CO
2
emission
reduction is taken into account.
4. Conclusions
It can be concluded that the carbonation
techniques using additives for reactivity rising
gave better results than those not using additives
and for that, studies were performed with the
purpose of a better understanding on the reactions
complexity regarding the carbonation. Despite
some good results the performed studies have not,
for the moment, came with significant discoveries
regarding the reactivity in mineral carbonation
process.
For the techniques using additives in
carbonation the biggest problem represents the
recycling of these additives and for the moment no
such breakthrough was reported and if this will
never be accomplished, maybe methods that dont
need them will be implemented.
Besides the cost of the process that is for the
moment larger than of those other carbon dioxide
capture and storage methods, the direct aqueous
carbonation seems to be the most promising option
of mineral carbonation. Other studies, however,
show that the dissolution and the precipitation
steps should be separated, even if the costs will be
higher. This separation would take out the need to
balance these reactions that are opposite. So,
taking that into consideration, the indirect aqueous
carbonation is the most attractive option.
Despite the good results on the high purity
precipitated calcium carbonate process, for the
moment, no large scale production was realized. If
this will be accomplished it is expected for the
price of carbonation process to be acceptable even
if is at a higher rate, due to the great value of the
high purity PCC.
Before one or more mineral carbonation
techniques could be implemented on an industrial
scale, much more research should be done and
important improvements of existing options are
needed. Another way for the large scale
carbonation to become reality would be
represented by completely new techniques.
5. References
295
[1] Huesemann, M.H., Can advances in science
and technology prevent global warming?
Mitigation and Adaption Strategies for Global
Change. (11), pp. 539-577, 2006.
[2] [Metz, B., O. Davidson, H. C. de Coninck, M.
Loos, and L. A. Meyer (eds.)], IPCC special
report on carbon dioxide capture and storage,
Prepared by Working Group III of the
Intergovernmental Panel on Climate Change,
Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA,
IPCC, 2005.
[3] House, K.Z., Schrag, D.P., Harvey, C.F.,
Lackner, K.S., Permanent carbon dioxide
storage in deep-sea sediments, 8th International
Conference on Greenhouse Gas Control
Technologies, 19-22 J une, 2007, Trondheim,
Norway.
[4] Rendek, E., Ducom, G., Germain, P., Carbon
dioxide sequestration in municipal solid waste
incinerator (MSWI) bottom ash, J ournal of
Hazardous Materials. (128), pp. 73-79, 2006.
[5] R. Santos, D. Franois, E. Vandevelde, G.
Mertens, J . Elsen, T. Van Gerven,
Intensification routes for mineral carbonation,
2010.
[6] Lackner, K.S., CLIMATE CHANGE: A guide to
CO
2
sequestration, Science. (300), pp. 1677-
1678, 2003.
[7] Huijgen, W.J .J., Comans, R.N.J . Carbon
dioxide sequstration by mineral carbonation:
Literature review update 2003-2004, Energy
Research Centre of The Netherlands, Petten,
The Netherlands, 2005.
[8] O'Connor, W.K., Dahlin, D.C., Nilsen, R.P.,
Turner, P.C., Carbon dioxide sequestration by
direct mineral carbonation with carbonic acid,
Proceedings of the 25
th
International Technical
Conf. On Coal Utilization & Fuel Systems,
Coal Technology Assoc. March 6-9, Clear
Water, FL, Albany Research Center, Albany,
Oregon, 2000.
[9] McKelvy, M.J., Chizmeshya, A.V.G., Squires,
K.D., Carpenter, R.W., Barat, H., A novel
approach to mineral carbonation: Enhancing
carbonation while avoiding mineral
pretreatment process cost, Arizona State
University, Center for Solid State Science,
Science and Engineering of Materials Graduate
Program, and Department of Mechanical and
Aerospace Engineering, J une 21-22, 2005.
[10] Zevenhoven, R., Teir, S., Eloneva, S., Heat
optimisation of a staged gas-solid mineral
carbonation process for long-term CO2
storage, Proceedings of ECOS 2006, 12-14
J uly 2006, Crete, Greece, pp. 1661-1669.
[11] Butt, D.P., Lackner, K.S., Wendt, C.H.,
Conzone, S.D., Kung, H., Lu, Y.-C., Bremser,
J .K., Kinetics of thermal dehydroxylation and
carbonation of magnesiumhydroxide, J ournal
of the American Ceramic Society. (7), pp.
1892-1898, 1996.
[12] Kakizawa, M., Yamasaki, A., Yanagisawa,
Y., A new CO
2
disposal process via artificial
weathering of calciumsilicate accelerated by
acetic acid, Energy. (26), pp. 341- 354, 2001.
[13] Teir, S., Revitzer, H., Eloneva, S., Fogelholm,
C.-J ., Zevenhoven, R., Dissolution of natural
serpentinite in mineral and organic acids,
International J ournal of Mineral Processing.
(83), pp. 36-46, 2007.
[14] Teir, S., Eloneva, S., Fogelholm, C.-J .,
Zevenhoven, R., Dissolution of steelmaking
slags in acetic acid for precipitated calcium
carbonate production, Energy. (32), pp. 528-
539, 2007.
[15] Eloneva, S.,Reduction of CO
2
emissions by
mineral carbonation: steelmaking slags as raw
material with a pure calcium carbonate end
product, PhD thesis, 2010.
[16] Katsuyama, Y., Yamasaki, A., Iizuka, A.,
Fujii, M., Kumagai, K., Yanagisawa, Y.,
Development of a process for producing high-
purity calciumcarbonate (CaCO
3
) fromwaste
cement using pressurized CO
2
, Environmental
Progress. (24), pp. 162-170, 2005.
[17] Gorset, O., J ohansen, H., Kihle, J ., Munz,
I.A., Raaheim, A., Method for industrial
manufacture of pure MgCO
3
froman olivine
containing species of rock, Patent,
WO/2007/069902, 21.6.2007.
[18] Kodama, S., Nishimoto, T., Yogo, K.,
Yamada, K., Design and evaluation of a new
CO
2
fixation process using alkaline-earth metal
wastes, 8
th
International Conference on
Greenhouse Gas Control Technologies, 19-22
J une, Trondheim, Norway, 2006.
[19] Rau, G.H., Knauss, K.G., Langer, W.H.,
Caldeira, K., Reducing energy-related CO
2
emissions using accelerated weathering of
limestone, Energy. (32), pp. 1471-1477, 2007.
Acknowledgements
Bodor Marius would like to acknowledge
the support provided by the European Union,
296
Romanian Government and Dunrea de J os
POSDRU 107/1.5/S/76822.
tefan Balt would like to acknowledge the
support provided by the European Union,
Romanian Government and Dunrea de J os
POSDRU 6/1.5/S/15.
297
DEVELOPMENT OF MEANS AND A PLATFORM
FOR RESEARCH
Liliana Georgeta POPESCU
1
, Mihai Victor ZERBES
1
, Radu Vasile PASCU
1
,
Lucian LOBON
1
, Livia Dana BEJ U
1
1
"Lucian Blaga" University of Sibiu, email: liliana.popescu@ulbsibiu.ro
Abstract: The paper describes the operation of a web platform made for an e-research center. The
platform includes a number of procedures performed and the specific tools of research process designed
to facilitate online research management, to facilitate dialogue between clients and professionals, enhance
teamwork, encourage exchange of ideas and the construction and design of prototypes of new products.
Keywords: e-research centre, platform
1. Introduction
PCs can create virtual spaces that are used in an
innovative manner, spaces for training,
information, communication, collaboration,
exploration, documentation, multimedia, word
processing, illustration, simulation and virtual
reality spaces.
In this work is presented the development and
testing of a general software infrastructure for a
virtual research center. The infrastructure must
support all activities concerned with technical
virtual research, especially in problem solving,
correcting and assessing solutions. To give a
modern interpretation of distance research, it was
primarily the relationship between the structure of
a center for research and dialogue can be
developed through it.
2. Development and characteristics of e-
research platform
2.1. Development of e-research platform
The main phases of the development platform are:
- Developing Virtual Research Centre - creation,
consolidation and expansion (infrastructure,
operational methodology);
- The initiation, conducting joint research and
development;
- The creation, development and maintenance of
facilities and research platform;
- The study of models of product development
and manufacturing process;
- The knowledge management;
- The research activities - joint design;
- Dissemination of results and knowledge
transfer to industry;
The input data refers to the participating
universities and research centers, human resources
and materials, the existing knowledge base. The
output which are resulting from research and
design activities, are the products used in industry,
the knowledge that will increase database and
feedback which improve continuously the business
center.
Efficiency and viability of such a platform is
based on systematic and systemic structure that
allows continuous improvement, which could be
extended.
The platform also allows the dissemination of
knowledge through databases.
2.2 Characteristics of e-Research Centre
platform
The platform includes a number of procedures
performed and the specific tools of research
process designed to facilitate online research
management, to facilitate dialogue between clients
and professionals) enhance teamwork, encourage
exchange of ideas and design prototypes new
products.
The website was done using Microsoft Visual
Studio 2010, free version of the code editor,
editing applications are of the HTML and ASP
programming language, the applications for
managing databases are build in Access, and
298
statistical evaluation is accomplished through
Excel applications.
3. Platform functionality
Depending on the issues raised and shaped
theme, specific steps will be taken. Since design
research is carried out by research teams in
different locations, it is necessary to communicate
a precise methodology supported by special
software and a phased deployment of the research
methodology to a fully efficient.
3.1 Access
The access can be made from the Internet or
intranet from any computer connected to the
Internet. The access to the general information and
those related to projects within the center is free.
3.2 Menu Sections
The main window of access to the site, which is
open on the activities menu and the center where
you can access the pages related to: Product
Design, Design Services, Product Design, Process
Design, Maintenance is shown in Figure 1. In these
pages have general information about products,
processes and services designed engineered by
Research Centre.
Figure 1: Access page to the Center Activities
The access to the members' personal
information: professionals, associates or customers
are secure. Each specialist, customer or employee
shall enter into the database by completing an
application that is generated by accessing the
"Application". Each user receives from the
platform administrator an username and initial
password, which can be customized after the first
login. On acceptance of entry in the database, it
generates a message of acceptance, the new
member receives the right connection space (log-
in).
3.3 Data Bases
Databases are build in Microsoft Access and
contains tables that store information about
customers, employees, professionals, projects and
newsletters.
Figure 3: Tables fromAccess Data Base
3.4 Secure Data
The platform management of the database, ( the
add / change user names and / or passwords, the
access to pages listing and changing the lists of
customers, experts and partners, the adding pages
and newsletters and projects) is secure and can be
accessed and performed only by the administrator.
3.4.1 Specialists Data
Information about the identification of members
can be viewed only by the site administrator.
Figure 4: Specialist Detail Page
This page allows you to add a new member,
following receiving the of application by the
platform manager. The window that can be viewed
behind the application, by the administrator of the
platform is shown in Figure 4. On the basis of
299
completed applications received from new
members, new information is added by pressing
the "Add New Member" button. The input data can
be modified or deleted by pressing the "Save
Changes" and "Delete Member" buttons.
3.4.2. The Specialists Section
This section provides information on
research areas and centers that can provide
advice in the field. Platform database contains
data related to research centers or universities
in Romania. Customer registration database
platform is build based on the activity field, so
they can be easily accessed by selecting the
industry. Specialist is a generic term used for
individuals (experts, researchers, teachers, etc..)
and for institutions, research centers, etc.. as they
have skills and can work in an e-research project.
3.4.3 Customers Section
In this section you can access to view / change
personal information stored in the database.
Customer registration in the platform's database is
carried out by their field activity, so they can be
easily accessed by selecting the industry ("Select
Category" - Figure 5.
Figure 5: Window add new customer database
3.5 Communication Module
Communication includes discussion forum
tools needed, on-line sessions, internal messaging.
The forum tools were developed to support the
listing of asynchronous discussions, the project
manager coordinating discussions based on the
project development. There are also possibilities
for chats video on-line sessions, internal
messaging (custom work groups). Addition page
on the server for the name and URL address of the
.pdf files, which are newsletters (informing letters)
for storage in the database.
4 Using e-research Platform
A general process for conducting network
research involves the following steps:
- Establishing customer contact;
The responsible for this section is the platform
administrator. When the client is seeking or
making a product order, process, or improve its
new project, the first stage of the dialogue can be
achieved by sending a text message, in an e-mail
form. On the home page is the "Home" menu,
which contains all the contact details of the center.
- Registration by the administrator of the
proposal;
The page allows the administrator to add a new
project based on what the client requested. The
project may be based on designing a new product,
or it can be based on improving a product or an
activity which is part of a bigger process (for
example, determinating the internal tensions for a
particular item
- Designation of project manager;
Depending on the category in which the project
(field) fits, the platform administrator, sends the
proposal of the customer to the specialists that are
added in that field related database. Depending on
the capacity and availability of specialist, the
project manager is appointed. Note that the
administrator can be also the project manager, at
the same time.
- Composition of the Working Group;
The forming of the working group is a task of the
project manager and of the platform manager,
which will invite potential specialists in the field to
be part of the team. The specialists database of the
platform enables fast selection and information of
the specialists that may form the group.
- Registering platformproject teamand client
project;
Table 'Admin users list is completed by the
administrator, with the team selected on a specific
project - a project that will receive a code. On the
platform can be run in parallel several projects, and
each of which can run separately. Also in this
section can be completed the database with
customer data and potential new specialists added
to the project.
- Establishing the schedule and procedure for
teamwork;
Useful methodology to be used is that of
Project Management. The main stages that go
through when applying this methodology are:
- The defining of the project theme in detail:
specify the problem, purpose or major
objectives, deliverables elements with
deadlines (times required), technical
specifications, costs, management support, etc.
- Planning the project
a. Defining the structure of work activities (activity
name, duration);
300
b. Implementation of network-type project plan
will specify the order and durations of each
activity;
c. Allocation of resources to each activity (human,
material, equipment);
d. Optimization of planning in terms of duration;
e. Estimated cost calculation.
The project may contain in addition to the actual
research or the optimization design modules of
products or technologies that achieve the design
prototype and prototype implementation.
- Establishing, analysis and completion basis;
In order to establish research directions,
grading criteria, the project manager analyzes the
client wishes, according to the questionnaire that
he completed. To create questionnaires or tests, the
platform program contains Excel files. On the
platform are already built predefined questionnaire
types to be completed by the clients and
specialists. Depending on the theme of the project,
these questionnaires can be changed or new ones
can be designed.
- Establishing product requirements
(classification criteria);
The platform administrator is responsible for
the registration of a new project or proposal. The
page allows the administrator to add a new project
based on what the client requested. The project
may be to design a new product, improving a
product or an activity from a process.
After outlining the problem that is intended to
be solved and the models to be developed, it is
necessary a complex task of planning and tracking
a project that aims to solve the problem. Useful
methodology to be used is that of project
management. The main stages that go through
when applying this methodology are:
- Defining the theme of the project in detail:
specify the problem, purpose or major objectives,
deliverables with deadlines sites (times required),
technical specifications, costs, management
support;
- Planning the project:
a) Defining the structure of work activities
(activity name, duration);
b) Implementation plan network type of project
that will specify the order and durations of each
activity;
c) Allocation of resources to each activity (human,
material, equipment);
d) Optimization of planning in terms of duration;
e) The calculation of cost estimates.
The project may contain in addition to the
actual research or the optimization of module
design and products and technologies that achieve
the design prototype. Project activities will be
added to the calendar center (figure 5), so, the
team/research team to be informed of the date and
time of deployment and type of action.
Figure 5: Add new action projects in calendar
There are cases when certain actions are
repeated, they recorded as repetitive actions (for
example - "Repeat Every Monday from 01 to 25 of
October" - "the action is repeated every 01 months
from October 25), (figure 6).
Figure 6: Model of schedule worksheet
The project manager sets the team's working
procedure, inform and instruct the team in this
regard. Also in this stage, it lay down the
responsibilities of members. The administrator
manages the material received and sends the
information and their agenda to the professionals.
The module allows managing the application-level
platform structures in the database. The platform is
composed of four administrative parts: system
administrator or project manager, specialists,
employees and customers. Between these
structures there are specific relations of the
organization of the program.
The project manager sets the working
procedure in the system depending on the type of
project in order to choose or to repeat certain steps,
holding several meetings (synchronous or
asynchronous), requests for opinions. The
following presentation is a general procedure for
301
conducting an online research. The platform is
generally available for the following steps:
- The proposal to address theme ideas;
At the request of project manager, each team
member will complete in the assigned field (figure
7) three proposals for resolving the theme.
Figure 7: Work sheet (synchronous-asynchronous)
project team
- Improve and complete the initial ideas;
Since the working sessions can be run both
synchronously and asynchronously, the project
manager sets a deadline date and time when
proposals are loaded by each participant. If this
session takes place synchronously, this stage
overlaps with the previous stage, so the
participants in this session can make changes
immediately after the initial data were added. In
the event that not all team members are logged
synchronously, it is recommended that the project
manager to provide a period of several days to
analyze and make possible amendments to the
proposals.
- Analysis of proposals and determining
potential improvements;
Ideas are centralized by the project manager and
the site manager is posting at the "Project
Description" section that is allocated to the project
in progress, the solutions that match the theme.
Solutions that do not correspond to the theme or
are not feasible are eliminated with the agreement
of members.
- Comparison of proposals for selecting the
optimal solution;
The project team is the one that notes from 5 (the
maximal value) to 0 - zero (minimum) proposed
solutions. The proposal ordering application after
scoring may be accessed by the platform manager,
from the platform. Scores are awarded depending
on the requirements to be met by product / service /
project. Data is loaded by each party specialist in
its field, and the centralization is made by the
project manager.
- Determination of the optimal solution;
The Solutions hierarchy is done by collecting the
scores given by team members. The field that is
assigned to the project description is named
"Project Description" on the "Project Information"
page (fig.6.18) and it allows insertion of multiple
data, as the window is provided with the cursor.
All final data will be posted from shares concluded
in this field manager. The fields assigned to the
specialists will be emptied (blank) for the next
steps.
- Review the solution by each teammember;
The review of the solution through the model takes
into account the highlighting of possible errors or
omissions have not been studied or have not been
considered in previous steps and that can be
corrected at this stage. For this action the project
manager may convene an online meeting to get a
fairly and quickly feedback before a decision is
being taken. In this case on the center calendar will
appear a note named as a "Project ID xzy -
Meeting September 15 at 10:00 am"
- Adopt constructive solution;
At this stage it is decided if the resulting solution
can be adopted and implemented. Again, the team
members "vote" and each will sustain its point of
view. The meeting is Synchronous , under a joint
meeting.
- Modelling constructive solution/concept
design;
Modelling can be done by one team member
designated for this purpose. It is indicated to be
used a software that allows viewing and making
changes to the design. The image is transmitted on
a modelled solution. jpg,. pdf or. tif to the team
members, for analysis and evaluation.
- Improvements/changes in the constructive
solution of the model;
Team members are given a deadline to review and
propose improvements to the model. They may
post their opinions on the "Project Information", in
the field assigned to each of them and then, during
a synchronously meting, the person in charge with
302
the modelling will present the improved variant.
The final model image is sent by mail to the whole
team.
- Acceptance of Product;
Team members will submit their final agreement in
terms of modelling. Agreement must be given and
the client, before the team ceased the activity. If
the customer finds design deficiencies of the
product, can make some suggestions to improve
the maintainability design. The client can actively
participate - on-line at the stage of conception.
Consent given by the team is practically ending the
activity of the design team.
- Transmission of client project;
Project Manager presents to the customer, the
solution / project that is the subject of the research
contract.
- The audit project;
The project audit is highlighting the strong parts of
the project and of networking team work and also
the weak parts, to be clearer.
- Update data base;
Platform manager responds of this activity. The
accepted solution will be registered in the data base
and presented under "Research Center" which is
under "Product Conception" section, or "Product
Design" section, depending on the type of research.
The completed Project will be presented to the
center members and prospective clients (if any)
through the newsletter sent and posted on the
platform. The virtual product design involves not
only the virtual, but also the maintenance of
process modelling, human interaction - modelling
and application object model. Virtual product
maintenance, analysis and testing can take place on
real products. To maintain the virtual
maintainability analysis system it has to be
developed key technologies so that the platform
can perform maintenance of virtual prototyping,
modelling and simulation of maintenance actions
based on failure analysis.
E-Research Centre should be seen as a living
thing that needs to be improved and adapted with
the evolution of business processes and
technologies.
5. Conclusions
To achieve fair and effective planning of a
project within a virtual research center is necessary
to have databases that contain:
- Database that aims general Engineering
knowledge, basis and specialty and liaise with
experts in the field;
- Database with information on staff: qualified
people, location, contact address, scientific
skills, managerial skills, labour costs;
- Database which includes laboratories and
opportunities available to investigate, research,
experiment, and model in these laboratories;
- Database on equipment, licensing, precision
level, input parameters, output parameters,
user training required, running costs,
consumables: characteristics, quantity, unit
price;
- Database which contains the existing software.
- This information will be used to complete each
activity needs of research projects to be carried
out based on the methodology of project
management.
References
[1] Brndasu, P.D., Dezvoltarea unui sistemde e-
learning i e-creativ design n domeniul
sculelor alchietoate, Grant CNCSIS, A. Cod
105/2006, 2007, 2008.
[2] Brndau P.D., Beju D.L. s.a., E-learning i e-
design n domeniul sculelor achietoare,
Editura Universitii Lucian Blaga, Sibiu,
2008.
[3] Popescu, L.G., Contributions on the efficiency
research activity in industry, Phd Thesis,
"Lucian Blaga" University, Sibiu, 2010.
303
I DENTI FI CATI ON AND QUANTI FI CATI ONS OF CHEMI CAL ELEMENTS
AND MI CROSCOPI C CHARACTERI SATI ON OF THE COPPER BASE
ALLOY AT DI FFERENT TEMPERATURE OF MOULDS
Cristina CAMBER I ORDACHE, Maria VLAD, Simion BALI NT,
Vasile BASLI U, Mihaela GHEORGHE
Dunarea de J os University of Galati, e-mail: Cristina.Camber@ugal.ro
Abstract: In this research paper, the structure identified by microscopic analysis for one copper base alloy at
two different moulds are comparing between them because the solidified conditions was different. The samples
were prepared by induction furnace using electrolytic copper and zirconium sponge and the ingots were cast into
the steel moulds where solidifies its. Copper alloys are an knew group of casting Cu alloys, well-known as
corrosion resistant materials, which are widely used as the devices for their antibacterial properties. Depending
on the combination of alloying elements and other factors, like solidification temperature or micro alloying
elements, the alloys with different mechanical and physic-chemical properties are obtained.
Keywords: copper alloys, structure identification, optical microscopy imaging, chemical elements.
1. I ntroduction
Devices copper alloys are a special group of
industrial copper alloys which have antibacterial
properties at normal room temperatures
(approximately up to 25C).
These properties of the alloys are resistant to
sudden temperature changes. Due to this, in the
design of this type of alloys, their mechanical and
thermal strains have to be critically considered
without ignoring their environment
aggressiveness during exploitation. In order to
accomplish such set of requirements, copper
alloys assigned for devices production need to
have an appropriate microstructure. Together with
chemical compositions and process parameters,
microstructure is as important parameter that has a
significant impact on the mechanical properties of
cast parts. [1]
Copper alloys contain various contents of
major and minor alloying elements. Cr-Zr has
stable equilibrium phase diagram with a
maximum solid solubility of ~0,12 at%Zr at
972C and it reacts with copper and aluminium to
form a thermodynamically stable phase. [2] One
of the major difficulties of copper alloys with
zirconium is that the last can lead particular
intermetallic effect with deleterious effects on the
physical and mechanical properties of the cast
parts. Compounds are very hard with the result
that machining cast parts with relatively high
zirconium contents can be difficult, resulting in
high casting finishing costs. Heat treatment
processes do not change the size and distribution
of this phase.
Are known that zirconium is desirable element
in copper alloys which improves the temperature
properties and thermal stability of such alloys.
Now, all the efforts are being made in attempts to
modify the adverse effect of intermetallic
zirconium phases, e.g., decreasing their size and
modifying them into a less harmful morphology.
Besides the major alloying elements that have a
huge impact on the solidification path of these
alloys, there are also some minor elements that
significantly change their solidification path. It is
known that the addition of elements such as Al,
Fe and Mo can modify the zirconium phase
morphology into less harmful shapes. This means,
that there is a possibility of controlling the cast
components by optimizing Zr with usually Al
additions. [3].
The typical copper devices are very complex
regarding their chemical compositions and
obtained structures. There are alloying elements,
Al, Fe, Mo, which have a significant impact on
the solidification path of these alloy Interactions
among them create different phases and
intermetallics, the shape and distribution of which
in the as cast and heat treated alloys depend on the
304
corresponding process parameters. At elevated
temperature, thermally stable intermetallics
should stop or reduce the movement of
dislocations and increase the mechanical
properties of alloys at elevated temperatures. The
strengthening effect of such intermetallics
depends on their stability at elevated
temperatures. The more stable intermetallics
achieved the better strengthening effect.[4]
2. Experimental procedure
2.1. Sample preparation
The structural matter of a material has a direct
impact on the physical and mechanical properties
of a cast. For this reason, the control of the
manufacturing process and the attainment of the
desired material properties require that the macro
and microstructures present must be defined and
described, both qualitatively and quantitatively
(Fig.1). [1]
Casting was performed directly into the steel
moulds. Alloy cast into the steel moulds (two)
were prepared by induction furnace type into the
graphite mould at a high temperature max.
1300C.
The mass of the cast sample was generally 750g.
The shape of the chill-mould was circular with
cone form without corrugate sides. The teeming
was performed into different ways. The melt was
teemed from above into moulds, down-hill
casting, at the different temperature of moulds.
One problem with downhill casting is
producing a good ingot surface. The quality of
surface depends on how well the moulds have
been prepared before the casting. In order to study
the surface quality of the ingot it was proceed to
heat one of mould before to be used. The second
mould was used at the room temperature.
The stream of melt has a large area exposed to
the air. The exposure greatly increases the risk of
oxygen and nitrogen absorption by the melt.
Those gases react chemically with alloying metals
in the melt, in our case with Al and micro alloying
elements, Fe and Mo in our case, which may
occur in steel, and lead to the formation of oxides.
The samples to study were cut horizontal from
A1
A2
B1
B2
Fig. 1: Opti cal micrograph(x=400) of thejointpreparedat 1300C and cast intwo si milar moulds of steel beeingat differenttemperatureof mould.
A-hot moul d(100 C). B-could mould (10C). A1: mi crograph of thedown part of the sampl e of the hot mould ; A2: micrograph of the middle part of the
sample of the hot mould, B1: micrograph of the down part of the sample of the cold mould; B2: micrograph of the middle part of the sample of thecold
moul d.
x400
x400 x400
x400
1
2
3
1
2
3
3
2
1
3
2
1
305
different parts of shape and after that were
prepared for analyses at the metallographic lab
and vertical form different part of shape to
analyse the internal character of crystals after
solidification process.
Optical analysis was used for microstructure
and phase identification. The measurements were
performed with OLYMPUS BX 51M instrument.
The elemental analyse was investigated with
spectrometry of X ray fluorescence (FXR).
Table 1 presents a comparative view of the
elemental composition of the tested copper alloy:
Chemical
elements
Cold mould Hot mould
Zr 14.76 15.29
Cu 78.46 77.91
Fe 0.53 0.31
Mo 0.070 0.081
Al 6.15 6.38
All the investigated alloys of the given
elemental contents were prepared under
laboratory conditions at Faculty of Metallurgy,
Material Science and Environment, University
Dunarea de J os of Galati. Romania by the internal
procedure, our own recipe for melting and
preparation of copper alloy.
2.2. Sample characterization
The optical microscopy imaging is the first
method used for investigating the solidification
path of metals and alloys. The resolution of the
optical instrument is generally very high, which
makes this instrument useful for many research
applications. A search of the available literature
showed that optical microscopy imaging is a
powerful tool that has been successfully used to
determine characteristic of solid phase during
solidification processes. [5]
To determinate quantitative and qualitative
elemental analyses by FXR of the chemical
composition of metallic samples which ware
investigated for differenced temperature of
moulds was used the portable spectrometer.
The main aim of the present work was to
characterize the solidification path of copper alloy
CuZrAlFeMo and especially the homogeneity and
the assimilation of the chemical elements during
the solidification process in two similar moulds
but being at different temperature.
Optical investigations
Optical micrograph imagines of samples casts
by different heating and cooling rate of steel
moulds, i.e., presented in Fig. 1, relived that the
solidification of the alloy is significant depending
by the temperature of moulds.
The microstructure of the samples is shown in
Fig. 1. (A1, A2-B1, B2). As shown in Fig. 1 at
both samples casts in cold and hot mould the
microstructure is clearly defined with three micro
structural constituents: crystalline grains and
mechanical mixture of phases. Two micro
crystalline constituents: 1-black grains with
polygonal form, 2-blau grains uniform distributed
with round form. The third micro structural
constituent is copper based matrix phases.
Solidification as encountered in common
processes does not occur at equilibrium, since
during solidification of most castings,
The major differences between the shapes
and the sizes of micro constituents, and their
arrangement/morphology in multiphase
system are because both temperature and
composition gradients exist across the casting
between the moulds walls and alloy.
The four right-hand terms are the change
in free energy because of temperature,
composition, curvature, and pressure
variation, respectively.
Table 1 shows obvious differences between the
chemical elements as a result of the of the alloy
cast at the different temperature of moulds.
Copper alloy of the hot mould A, contained
high percentages of the alloying elements Zr, Al
and less percentage of Fe as micro alloying
element while copper alloy of the cold mould B
contained low percentage of alloying elements Zr
and Al but a high percentage of Fe.
The temperature interval of the solidification
process was also shorter and it commenced at the
cold mould.
Changes on chemical composition of alloy
during the solidification process can be the result
of different internal reactions.
306
Is known that the solid shell varies with the
casting method and the shapes and sizes of the
castings.
In our case the samples casts into steel moulds
at of the different temperature if moulds have the
columnar character of the internal crystals. The
cross section area of the columnar crystal
increases with the distances from the cooled
surface of mould.
The growth rate of the solidification front
decreases with the distance from the ingot surface.
The growth conditions and consequently also
the structure morphologies are not constant.[6]
The dendrite arms and the crystal cross
sections seem to grow in two or three different
steps,
Eq (1) and (2).
The relationship:
growth
2
den
=const, (1)
is valid in this case, but with different
constants for each steps. At the constant eutectic
temperature precipitation of the eutectic alloy
occurs.
Ingot with long solidification times the number
of new floating crystal decrease during the
solidification process and the central zone of
branched columnar crystals is extended over the
whole section in the upper part of the ingot.[5]
Is known that that the necessary condition for
solidification is undercooling:
dr/dt=(T
L
-T
crystal
)
n
(2)
The temperature of the melt and the solid
phase of an ingot as a function of time.
4. Conclusions
This paper presents the results of research on
copper based alloy in the different temperature of
moulds cast.
The structure can be characterised as a very
fine surface zone that consists of a fine network of
tin dendrite arms and not, as in ingots, of a great
number of fine-grained crystals.
At the ingot from into the cooled mould we
can observe the less quantity of micro alloying
elements assimilate comparing with the chemical
elements from the second ingot where the micro
alloying elements are assimilated more.
In the same time, the dendrites from the
cooling mould became tin comparing with the
second ingot.
Due to the high temperature gradient the thin
dendrite crystals grow inwards and form columnar
crystals. In the cooled mould its are stopped by
the formation of the equiaxed crystals that have
grown freely in the melt.[1]
With the aim of improve the corrosion
resistance of the antibacterial copper based alloys,
a number of copper alloys can be prepared by
changing the concentrations of alloy elements
and micro alloying elements in order to increase
the each possible phase or change of its shape in
which the alloy element is dominant, since the
phenomenon of change of the mean concentration
in the actual case, due to redistribution of liquid
enriched or depleted by the alloy elements, leads
to a change of mean concentrations to values
significantly higher or lower than the nominal
value, and to the possibility of the occurrence of
other phases.[4]
When a larger portion of micro alloying
elements exists in the matrix phase the
solidification occurs at different temperatures,
which is good. However, a further increase in the
content of micro alloying elements may lead to
the formation of a new phase, the strengthening
effect of which is poorer. This means, the micro
allowing elements has to be added in a proper
amount to the alloy, depending on the Copper
content [2], [3] and [5].
The results presented in this paper are only an
introduction to further ongoing research aimed at
obtaining the required combination of the further
antibacterial copper alloy with the goal of
providing optimal characteristics for mechanical
and corrosion resistance as well as defining a
mathematical model that can predict the different
cases on the micro and macro levels of
segregation.
Performed thermal analysis and microscopic
characterization of Cu-enriched phases can be
determined using thermal analysis. The
importance of the analysis of these phases is due
to their strengthening effect that is usually
enhanced and controlled by applying heat
treatments that promote the precipitation of
coherent or incoherent alloying or micro alloying
elements that improve the properties of Copper
based alloy.
307
Copper alloy with better properties than those
achieved using conventional casting method can
be produced with the rapid solidification methods.
Acknowledgement
We would like to thank to the Project SOP
HRD EFICIENT 61445/2009 for financial
support and we also thank to Prof. Anisoara
Ciocan and Prof. Florentina Potecasu of the
Faculty of Metallurgy, Material Science and
Environment, University Dunarea de J os of
Galati for providing facilities.
References
[1] Fredriksson, H., Akerlind, U., Materials
Processing during Casting, J ohn Wiley
&soons, Ltd, England, 2006.
[2] D.Arias, J.P. Abriata, Buletin of alloys phase
diagrams, Vol.11 No.5, 1990.
[3] Moise Ienciu, Nicolae Panait, Petru
Moldovan, Mihai Buzatu, Elaborarea si
turnarea aliajelor neferoase speciale, Editura
Didactica si Pedagogica Bucuresti.
[4] Kebbache I, Debili, M.Y., Separation of
Aluminumand Copper by Intermetallic
Compounds after HF Induction Fusion,, J OM,
2010
[5] ASTM
[6] Scandinavian J ournal of Metallurgy,
Processing and Materials Engineering Ed.
Wiley, Ltd, England, 2006.
[7] Effenberg, G., Ilyenko, S., Al-Cu-Zr
(Aluminium - Copper - Zirconium),
SpringerMaterials
{http://www.springermaterials.com).
308
309
MANGEMENT OF TECHNI CAL DATA I N THE STUDY OF THE
BRAKI NG SYSTEM OF CARS
CI UBOTARI U DANUT
1*
, NECULAI ASA VASI LE
2
1
Ministry of Domestic Affairs and Administration, SPCRPCIV, Botosani, Romania,
email danutciubotaru@yahoo.
2
Gheorghe Asachi Technical University, Iasi, Romania
Abstract: Braking process of a vehicle is complex, giving possibility to the driver to reduce cars
speed or to stop it at big speed in a short distance. Using the diagrams registered for the different
working conditions, defined accordingly with the experimental plan, there have been analyzed the
values of deceleration, variation of speed and the covered space for each specified moment of
braking. As result of the experimental researches, we have reached the conclusion that a good
behavior at cars braking is registered when the diagrams of deceleration variation in relation with
braking time have a symmetric shape.
Keywords: vehicle stability, experimental braking process, ABS.
1. INTRODUCTION
Braking process of a vehicle is complex, being
defined by factors leading to different effects from
the point of view of impact, intensity and
duration.
The braking system gives possibility to the
driver to reduce cars speed or to stop it at big
speed in a short distance.
Last generation cars are endowed with:
-Service brakes;
-Backup brakes;
-Handbrake;
-Auxiliary brake.
From the point of view of the vehicles control
and safety, the most important is the service
braking system (the foot brake).
The longitudinal deceleration a
x
is one of the
main parameters of the braking system, which can
be calculated with formula:
(1)
where
x
is the longitudinal coefficient of
adherence of the tires with the running path and g
the gravitational acceleration (g =9,81 m/s
2
).
Because, in reality, a car is not perfect
from the point of view of the technical
characteristics, the effectiveness of braking is
lower.
The Directive of EU 71/320 [1] shows that this
criterion has to satisfy the following condition:
(2)
In the specialty literature, the given value for
the adherence of the tires with the running path on
a dry asphalt is
x
=0.8.
Introducing this value, which applies only to
the vehicles and tires produced until 1980, in
relations (1) and (2), there results the deceleration
of a vehicle in good technical conditions, at
braking on a dry asphalt
a
x
=6.0 7,85 [m/s
2
]
In the situation of actual cars, the maximum
coefficient of adherence has the value
x
=1 1,2
[2] , if braking takes place on a running path with
dry asphalt, situation when the deceleration of old
cars with tires produced at the present moment
can reach values in the range 7,359,3 [m/s
2
].
The modern cars are endowed with systems
which prevent wheels blocking (ABS) and the real
braking distance is quite close to the calculated
value obtained with the maximum values of the
coefficient of adherence. In this situation, the
deceleration can be close to g =9,8 [m/s
2
].
310
2. ANTI -BLOCKI NG SYSTEMS (ABS)
Anti-blocking System is used at cars for
avoiding the blocking of wheels during braking,
for providing the stability of the vehicle and an
optimal deceleration.
Anti-blocking of wheels is realized by adapting
the braking force at each wheel, function of the
wheels adherence to the running path.
The advantages of a system [3] assure keeping
the vehicle under control, although the braking
and the side forces are increasing, on a minimum
braking distance with a reduced wear of tires.
Some of the conditions to be accomplished by
an ABS are following presented:
1. ABS must function on all the range of
speeds;
2. ABS must correspond to the conditions of
adherence with the running path;
3. To assure the steering of vehicle at
slippages appearance;
4. To recognize aquaplaning and to respond to
it;
5. To rapidly and automately correlate the
hysteresis of tires at specific conditions;
6. The option of engine brake to be possible;
7. If an error at the ABS functions occurs,
during braking, the braking system must
function normally on classical principles.
t
S
KG
DG
KP
DP
1
2
3
Vehicle speed
Figure 1Variation of speed in relation to the breaking
time
The functioning principle of the anti-blocking
system is based on rapid braking without wheels
blocking, by applying short and rapid actions
having as objective the delay of wheels blocking.
This process is realized by the action upon the
pressure of brake fluid from the brake installation,
regulated to avoid the wheel blocking regardless
the adherence to the running path.
In fig.1 [4] there is presented the diagram of
variation for the speed of a vehicle with ABS in
relation with the braking time, for the four wheels
of the car.
When the brake pedal is acted, the value of
wheels rotational speed is decreasing due to a true
value depending to the vehicles speed (pt.1 from
fig. 1.). When the force which the brake is pressed
with is bigger or the running path is slippery, the
rotational speed of wheels significantly reduces
(pt.2). In pt.3, ABS is activated and releases the
brakes. When the rotational speed of wheels
increases again, the brakes are once again
activated, the process being repeated until the
vehicle is brought to the desired speed.
2.1. Braking of vehicles without wheels
blocking
If the car is not endowed with a regulator of
the braking force of the ABS, the total braking
force is differently distributed between the
vehicles bridges, from which reason the wheels
of the front bridge and those of the rear bridge are
not simultaneously blocked.
In the situation that the wheels have a good
adherence to the surface of the running path, the
rear wheels are first blocked, context in which the
vehicle may lose transversal stability and if the
front wheels are blocked, the control of the
advancing direction can be lost.
A good behavior during braking is met when
the motion of the car is continued in straight line
at brake action, even for great speeds, condition
imposed accordingly to the Directive of EU
71/320. So, in the moment when the emergency
brake is acted, the motion of the car must continue
in straight line, it means a bigger relative braking
force upon the front wheels must be provided, in
comparison with the force acting on the rear
wheels.
The coefficient of distribution of the braking
force [5] represents the ratio of the braking force
acting on the front wheels, to the total braking
force, acting on all the wheels.
T =
P
ST1
/ (P
ST1
+P
ST2
) (3)
P
ST1
and P
ST2
represent the braking forces
upon the front wheels and respectively upon the
rear ones and they depend on structure of the
brake system. This coefficient (
T
) can be
311
calculated for any type of vehicle [5], accordingly
to the known parameters of the brake system.
l l
L
h
P
P
in
z
st
P
G
z
P
st
P
1
Figure2Scheme of forces in the process of braking
without wheelss blocking
In the situation of neglecting the loss of energy,
for overcoming the resistance forces, the
maximum of the deceleration for a vehicle at the
limit of rear wheels blocking will be:
a
x
= (4)
Where: l
1 =
the distance in longitudinal direction,
between the mass center of the vehicle and the
front wheel, L represents the distance between
wheels (the wheelbase) and h
c
is the height of the
mass center.
The maximum deceleration of a vehicle at the
limit of front wheels blocking is:
a
x
= (5)
Where: l
2
=the distance between the rear wheel
and the cars mass center.
The optimum coefficient of adherence
opt
, will be given by relation:
opt
=(L
.
T
l
2
)/ h
c
(6)
In the conditions when the correct value of the
coefficient of adherence exceeds
opt
, the rear
wheels will be firstly blocked by braking and
relation (4) allows calculation of deceleration. In
the situation when the value of the coefficient of
adherence is smaller than the optimum value
opt
,
the front wheels will be the first blocked when the
vehicle braking and the expression (5) will be
used for calculate the deceleration.
When
opt
has a value close to zero or is
negative, only the rear wheels will be blocked in
any situation of motion.
In the situation of a vehicle with unblocked
wheels, the braking distance is calculated with
formula:
S
0
=(t
1
+t
2
+0,5
.
t
3
)
.
v
0
+v
0
2
/2a
x
(7)
In the situation when the initial speed of the
vehicle (v
0
) is not known, the braking distance of
a vehicle with unblocked wheels (unlike the case
of blocked wheels, when the slip length can be
measured) can be determined with certain errors
and this circumstance can influence the
correctness of the obtained results.
In the situation of emergency braking, the
braking forces acting on the vehicles wheels
should not exceed the forces of adherence of the
tires at the running path, the calculated value of
the deceleration of a vehicle with blocked wheels
will be smaller and the braking time and braking
distance will be bigger, in comparison with the
case when the coefficient of adherence is known.
2.2.Comparison of the braking processes for
vehicles with ABS and without ABS
From the study of vehicles theory, the
deceleration of a vehicle in good technical
conditions without ABS reaches the top, fig.3, just
at the beginning of thebraking process and then
some decreasings are happening, due to the fact
that the top of deceleration is reached before
blocking the vehicles wheels.
Figure 3 Diagramof braking with ABS-phase of
starting the braking process
After the wheels blocking, a decrease of the
deceleration takes place, because braking of a
blocked wheel is less efficient [6].
Maximum of deceleration lasts more during
the braking of a vehicle without ABS at a lower
speed; at a bigger speed, when the braking time is
greater, the top of deceleration lasts less. This fact
explains the decrease of speed for vehicles
312
without ABS when a certain displacement speed
is reached.
In fig.4 the diagram of variation of
deceleration in relation with time is presented,
during the period of speed variation of vehicle
without ABS.
0
,
8
1 0
D
e
c
e
l
e
r
a
t
i
o
n

[
m
/
s

]
2
2
4
6
8
0
0
,
6
0
,
4
0
,
2
0
1
0
1
a) b)
Figure 4Diagramof braking without ABS: a)- phase of starting the braking process; b)- phase of real braking
In specific road conditions, the value of the
vehicles deceleration may have a bigger
correctness, which imposes an experimental
research.
The deceleration of a vehicle on a running path
can be measured using a special device
(decelerometer) or can be calculated with relation
[6], if the initial speed v
0
and the braking distance
are known.
a
xn
= (8)
3. EXPERI MENTAL RESEARCH UPON
VEHI CLES BRAKE
From those above presented, the necessity of
developing some experimental researches upon
the process of cars braking has resulted. For
experiments, two Daewoo cars in good technical
shape and with the following characteristics have
been used:
1) for vehicle without ABS
a) Vehicle category M 1;
b) Wheelbase of the vehicle, at dynamic test:
2360 mm;
c) Mass: nominal: 1050 kg, maximum: 1550
kg;
d) Maximum speed: 190 km/h;
e) Tires dimensions: 185/65 R 15;
f) Service brake with two independent circuits
;
1. Brake type : front: disk and rear : drum ;
g) Brake with double circuit;
h) Without ABS;
i) Mass of vehicle to be loaded:
-Test I: 1103 kg (667 kg +436 kg)
-Test II: 1558 kg (798 kg +760 kg)
2) for vehicle with ABS
a)Vehicle category M 1;
b) Wheelbase of the vehicle, at dynamic test:
2360 mm;
c) Mass: nominal: 1050 kg and maximum:
1600 kg;
d) Maximum speed : 190 km/h;
e) Tires dimensions: 185/65 R 15;
f) Service brake with two independent circuits
,in X;
g) Mark and type of brake lining: front: ferrodo
182 ; rear : DOW 8273 ;
h) Brake with double circuit;
i) Brake type : front: disk and rear: drum ;
j) Vehicle endowed with ABS;
The experiments have been done in the testing
ground of Daewoo Craiova cars factory, on
different running paths with surfaces covered
with: asphalt dry - wet, gook, glaze, snow, car
313
empty loaded, engine clutched declutched, and
also in normal conditions of circulation on
national roads in Botosani County, with the
Testing Laboratory owned by SC
ROMTURINGIA.
In the experiments, a device for measuring
vehicles deceleration (decelerometer) type
MAHA VZN 100 was used. The device is
appropriate for such measurements both for
vehicles with hydraulic and pneumatic brake
system and for those with ABS, as in fig.5.
The device is compound of: control board, for
data displaying and registering; transducer wheel
with its fixing support; flexible drive cable and
dynamometric pedal (specific for braking tests).
The transducer wheel transmits rotational
motion by the flexible cable to the control board,
for displaying and registering, where the
displacement speed can be read at any moment. In
the registering apparatus, a time base is
embedded, which electrical signals, at a period of
one second, are marked on the registering paper,
moving by rolling on the internal drums, with a
speed proportional with the speed the wheel is
running on the ground.
The used device measures the process
parameters and draws the corresponding diagrams
for speed, deceleration and the covered space,
function of the type of braking.
The conformity of the measuring device VZN
100 with the standards stipulated for the devices
checking is certified by the German Association
for Technical Inspection (TUV) [5]. This device is
also used by the Lithuanian centers for technical
inspection for establishing the efficiency of the
brake system, basing on the maximum
deceleration.
Using the diagrams registered for the different
working conditions, defined accordingly with the
experimental plan, there have been analyzed the
values of deceleration, variation of speed and the
covered space for each specified moment of
braking.
As result of the experimental researches, we
have reached the conclusion that a good behavior
at cars braking is registered when the diagrams of
deceleration variation in relation with braking
time have a symmetric shape.
Based on the found parameters of vehicles
motion, we can provide the variation of the
maximum values of deceleration for specific
situations.
Figure 5. Principle scheme of a car endowed with the 5
th
whell
314
5. CONCLUSI ONS
Based on the bibliographical researches in the
considered technical field and on the experimental
researches, certain conclusions of undoubted
importance can be presented:
I. In the developed experiments, the
deceleration established for vehicles with ABS
have varied from 8,05 m/s
2
(at the speed of 40
km/h) up to 8,45 m/s
2
(at the speed of 80 km/h),
and often the maximum values have been close to
9 m/s
2
.
For vehicles without ABS, the deceleration
have varied from 7,05 m/s
2
(at speed of 40 km/h)
up to 6,87 m/s
2
(at speed of 60 km/h) and up to
6,65 m/s
2
at a speed of 80 km/h. When v
0
increases, the difference of deceleration or
vehicles with and without ABS increases with
11,1%, 18% and respectively 24%.
II. There was calculated that the average of
braking distance for a vehicle with ABS is smaller
with 4,25% in comparison with those for a vehicle
without ABS.
If the initial motion speed is 40 km/h, and if v
0
= 80 km/h, approximately the deceleration is
modifying up to 15,59%;
If v
0
= 60 km/h, the braking distance for
vehicles with ABS was of 20,80 m, and for
vehicles without ABS it was of 22,42 m, the
difference being of approximately 7,35%.
III. The value of the braking distance can be
calculated based on the braking process
parameters, such as: braking time t
3
, the value of
deceleration a
xn
and the initial speed v
0.
There is proposed the use of the values of the
parameters determined within the experimental
research for the examination of situations specific
for traffic accidents, for a more correct defining of
the methods for calculating the braking
parameters for cars with or without ABS.
IV. As a result of research accomplishing,
there has been noticed that the used cars are
corresponding and respect the stipulations of the
Regulations of braking R-13ECE ONU. That is
why the regulations of EU impose for all the cars
built since 2011 the obligation of endowment with
ABS.
REFERENCES
[1]. ***, Council Directive 71/320/EEC of 26 J uly
1971 on the approximation of the laws of the
Member States relating to the braking devices
of certain categories of motor vehicles and of
their trailers.
[2]. Mitunevicius, V.-1999-Application of braking
responses of vecles for experts examinations
of vehicles.Im Transbal-tica-99:Collection of
scientific reports of the International
Conference (Transbaltica-99;tarptautines
konferencijos moksliniu pranesimu rinkinys,
Vilnius, 8-9 April 1999).Vilnius: Technika,
1999, p.221-226(in Lithuanian).
[3]. Neculaiasa Vasile-1996 Vehicle Dynamics,
Polirom Publishing House, Iasi, 1996;
[4]. Gillespie,T.D.-1992 Fundamentals of vehicle
dynamics.Society of Automotive Engineering,
Inc.400 Commonwealth Drive Warrendale, PA
15096-001.1992.250 p.
[5]. Illarionov, V.A.-1997-Experts examination
of traffic accidents Moscow;Transport,
1997.255 . (in Russian).
[6]. Sokolovsku, E.-2004- Investigation on
interaction of the wheel with road its elements
in the context of examination of traffic
accidents: Doctor;s thesis:tehnological
sciences:Transport engineering (03T) Vilnius,
2004.147 p.
315
COMPOSI TE SHAPE MEMORY ALLOYS FOR METALLI C
STRUCTURES
A. Enache, I. Cimpoeu, S. Stanciu, I. Hopulele, R. M. Florea, N. Cimpoeu*
The Gh. Asachi Technical University of Iasi Faculty of Materials Science and Engineering,
Prof. dr. docent Dimitrie Mangeron Rd., 61 A, 700050, Iai, Romania
Abstract: Shape memory alloys exhibit nice and interesting properties based on them internal solid state
transformation from martensite to austenite under temperature modifications. Damping capacity represent one of
the latest attraction of shape memory alloys based on them big internal friction values having applications in
many domains as dissipaters, dampers or structural elements for noise and vibrations attenuation. Composite
with metal matrix materials represent interesting materials that use both materials properties. In this study a
shape memory alloy composite material is analyzed by his interface between SMA matrix material and a steel
reinforcement element point of view. Using scanning electrons microscopy and EDAX analysis few results were
obtains and comment to establish the interface nature.
Keywords: shape memory alloy, composite, diffusion
1. Introduction
Among the various types of composites, the
family of discontinuous metal matrix composites
(MMCs) containing particulates, whiskers, wires,
precipitates, fibers, nodules and platelets, are
favoured because they offer improvements to the
mechanical properties of the monolithic alloys
while remaining relatively easily deformable.
Metal matrix composites are compound materials
whose microstructure consists of a metallic alloy
into which a particular reinforcing component is
introduced. MMCs offer advantages in applications
where low density, high strength and high stiffness
are a primary concern. The availability of various
types of reinforcements at competitive costs, the
feasibility of mass production and high damping
capacity [1,2] make this type of MMC, metallic
matrix and wires as reinforcement elements, more
attractive. However, these materials may suffer
from inhomogeneous distribution, size or shape of
wires, low ductility, inadequate fracture toughness
and inferior fatigue crack growth performance
compared to that of the matrix [3,4].
To optimize the mechanical and physical
properties, in particular the damping conditions
and damage tolerance of such materials one can
utilize shape memory alloys as reinforcement.
Shape memory alloys (SMAs) have received great
attention because of their shape memory effect
(SME) and many investigations are conducted on
their basic performance and applications.
The transformation under compression can
result in stress in the matrix, which in turn
enhances mechanical properties such as yield stress
[510], fracture resistance [11,12], an capacity of
suppression of crack growth [20] and thermo-
mechanical fatigue [9,13]. Two SMAs which
generate large amounts of strain and are capable of
generating a large force upon transformation back
to the austenitic phase are NiTi alloys and Cu-
based alloys. Copper-based SMAs are particularly
interesting because of their low-cost and relative
ease of processing. On the other hand, NiTi alloys
tend to be more thermally stable and to have a
lower density, higher yield and ultimate tensile
strength; they are also more resistant to corrosion
than Cu-based alloys [14].
Traditional restoration techniques often do not
give structures sufficient resistance against
maximum expected earthquakes and/or might be
too invasive. Therefore, there is an ongoing effort
to find techniques that can guarantee structural
stability and at the same time respect the integrity
of the structure. Special devices that exploit the
superb damping properties of Shape
Memory Alloys (SMA) are under development.
SMAs have found applications in many areas due
to their high power density, solid state actuation,
high damping capacity, durability and fatigue
resistance. When integrated with civil structures,
316
SMAs can be passive, semi-active, or active
components to reduce damage caused by
environmental impacts or earthquakes.
Though most of the research activities of
SMAs applications in civil structures are still in
laboratory stage, a few have been implemented for
field applications and found effective [15-19].
Having a composite material with a metallic
matrix of a copper based shape memory alloy in
this study were analyzed some concerning about
the metallic interface nature. The results present a
highly well form interface between the matrix and
reinforcement element, both metallic materials that
improve the mechanical behavior of the material.
To obtain the shape memory alloy based on
copper for the composite matrix was used a
laboratory furnace with graphite crucible using
copper, zinc and aluminum high purity materials
with reduce percentage of iron [21].
After the matrix were melted, the liquid alloy,
was poured in a special semi-form prepared for
arch material reinforcement element stabilization
with the geometrics and main elements presented
in figure 1.
Figure 1: Composite realization semiformhaving the
elements: 1-metallic body form; 2-cylindric cavity for
reinforcement element fixation and matrix body form; 3-
articulation; 4-rod suport for arch fixation at extension;
5- rod used for supporting the extension arches for
wires; 6- hand support, 7-reinforcement fiber; 8-arch
used for metallic fiber extension
Each melting provide material for nine samples
with different materials as reinforcement element if
is necessary or similar like in this case.
In figure 2 is presented the geometrical shape of
the composite samples obtained and further
analyzed.
Figure 2: Catia software design of the shape memory
based composite
Chemical composition was determined through
spark spectrometry analysis using Foundry Master
equipment (for matrix and reinforcement elements
chemical analysis) and EDAX analysis as well for
interface study. In this study different EDAX
software applications were used to determine the
chemical variation of the elements on Line,
Mapping or Point mode with automatic or element
list considerations.
Microstructures of the composite in different
heat treated states of material were obtained with a
scanning electron microscope (SEM) LMH II by
Vega Tescan brand using a secondary electrons
(SE) detector at a 30 kV lamp power supply.
A composite material made by CuZnAl shape
memory alloy as matrix and arch wire steel as
reinforcement element was obtain by classical
melting method.
Composite material surface was analyzed using
a scanning electrons microscope to observe
especially the interface between those two metallic
materials that form the composite. From
microstructure images, figure 3 a) and b), can be
observe that a thick interface is formed in
composite after the material suffer a 30 minutes
homogeneous thermal treatment.
A proper chemical attack of CuZnAl and Fe-C
alloys to evidence both microstructures is difficult,
as can be seeing from figure 3 were only the
interface is marked better, interface based on
aluminum compounds. On the matrix some
formation of martensite variants can be seeing and
on reinforcement element no microstructure is
evidenced.
Microstructural analyses represented in figure 3
evidence an interface formed between the shape
memory matrix and the elastic element with
different dimensional areas and relatively straight.
317
a)
b)
Figure 3: SEM microstructure of the composite
material a) 1000 x power and detail b) at 2500x
The interface dimension is around 1-2 m with
areas modify as inhomogeneous.
In figure 4, on a selected line presented in a), is
represented the chemical elements copper, zinc,
aluminum, iron, manganese and silicon
distributions. The selected line represents the
elements distributions on matrix, the reinforcement
element and interface as well in the same time. The
line selected is of 20 m perpendicular on the
border area and evidencing the mass transfer type
at the interface.
In figure 4 b) is presented a complete chemical
analysis of the elements variation between matrix
and reinforcement element with accent on the
interface. Can be observed a diffusion type
material transfer between those two metallic
alloys. On the interface area formation of some
new chemical compounds are observed as FeAl
3
and a migration of elements characterize the
interface find a higher aluminum percentage on the
steel part than the shape memory alloy material
were a loss of aluminum is observed.
In figure 4 b) the interface areas is analyzed
through mapping elements technique as well
marking the aluminum conglomeration on the
interface, a separation layer on the interface and
the loss of Al element from the matrix. Silicon
element is also presented on the border forming
different stabilization compounds like AlSi
2
.
a)
318
b)
Figure 4: Composite material interface area analysis a) selected area for analysis and b) distribution of elements
on the interface area
In figure 3 b) two points were selected for
chemical composition investigations the results
being presented in table 1 and table 2 for Pct 1 and
Pct 2. First point is selected on the interface and
the second one near the border on the
reinforcement element part. Even the first point is
situated more to the matrix part the iron element
has a very high percentage presence based on
diffusion process movement of the elements.
Aluminum is also present having a FeAl based
interface. The other elements participate at the
interface formation in smaller percentages and
loosing in front of the iron aluminum compounds
as X-ray signal and quantitative percentages.
Table 1: Chemical composition of the interface in point 1 selected in figure 1 b)
Iron 26 K-series 99081 71,50894 81,84484 75,4816 1,844726
Aluminium 13 K-series 8076 6,7355 7,709049 14,71579 0,381417
Copper 29 K-series 5054 5,971241 6,834324 5,539327 0,197696
Zinc 30 K-series 1144 1,479678 1,69355 1,333941 0,084248
Silicon 14 K-series 1761 1,102398 1,261739 2,31386 0,086212
Manganese 25 K-series 921 0,573593 0,6565 0,615477 0,084937
Sum: 87,37135 100 100
Table 2: Chemical composition of the interface in point 2 selected in figure 1 b)
Iron 26 K-series 120555 80,01884 91,8602 89,77184 2,05754
Copper 29 K-series 3090 3,256238 3,738104 3,210522 0,127146
Aluminium 13 K-series 1858 1,924301 2,209063 4,46842 0,141725
Zinc 30 K-series 701 0,820734 0,942188 0,786391 0,063644
Manganese 25 K-series 1126 0,611817 0,702355 0,697745 0,088394
Silicon 14 K-series 655 0,477439 0,548092 1,065084 0,057611
Sum: 87,10937 100 100
319
On the other part of the interface, at the
reinforcement element side, copper is diffused
in a reduce percentage of 3.738 % wt but
higher than the copper presence in the arch-
wire steel material.
Silicon element decrease as well in percentage
comparing to the other side of the interface fact
that represent a smaller formation of silicon based
compounds. Mechanical tests of the composite
material with present the material behavior,
interface modifications and material properties as
damping element.
4. Conclusions
A composite based on a shape memory alloy
matrix were obtain and analyze by the interface
formation point of view. The reinforcement
element proposes was a steel archwire with 1 mm
diameter. The interface between those two metallic
materials was between 1 and 2 m thick especially
based on FeAl compounds. In formation of the
interface diffusion process interfere with accent on
aluminum and silicon elements decreasing the
copper and zinc percentages. Interesting variations
of chemical elements distribution can be observe
on the border between materials anyway a nice
bond between the metallic materials was created.
Acknowledgement
This paper was realised with the support of
POSDRU CUANTUMDOC DOCTORAL
STUDIES FOR EUROPEAN PERFORMANCES
IN RESEARCH AND INOVATION ID79407
project funded by the European Social Found and
Romanian Government.
References
[1] Cimpoeu, N., Stanciu, S., Meyer, M.,
Ioni, I., Cimpoeu Hanu, R., Effect of
stress on damping capacity of a shape
memory alloy CuZnAl, Journal of
Optoelectronics and Advanced Materials,
Vol. 12, Issue 2, pg. 386-391, IDS 570DE,
ISSD 1454-4164, 2010.
[2] Paun, M. A., Cimpoesu Hanu, R.,
Cimpoesu, N., Agop, M., Baciu, C.,
Stratulat, S., Nejneru, C., Internal friction
phenomena at polymeric and metallic shape
memory materials. Experimental and
theoretical results, Materiale Plastice, vol.
47, nr. 2 , pg. 209-214, 2010.
[3] Davidson, D.L., Eng. Fract. Mech. 33,
965977, 1989.
[4] Davidson, D.L., Campbell, J.B., Page,
R.A., Metall. Trans. A 22, 377391, 1991.
[5] Furuya, Y., Sasaki, A., M. Taya, Mater.
Trans. J IM 34, 224227, 1993.
[6] Hamada, K., Lee, J.H., Mizuuchi, K.,
Taya, M., K. Inoue, Metall. Mater. Trans. A
29, 11271135, 1998.
[7] [7] G.A. Porter, P.K. Liaw, T.N. Tiegs,
K.H.Wu,Mater. Sci. Eng. A 314 (2001)
186193.
[8] Porter, G.A., Liaw, P.K., Tiegs, T.N., Wu,
K.H., J OM, 5256, 2000.
[9] Shimamoto, A., Taya, M., Trans. J SME 63,
2631, 1997.
[10] Tsoi, K.A., Stalmans, R., Wevers, M.,
Schrooten, J., Mai, Y.X., Proc. SPIE 4234,
125133,2001.
[11] Furuya, Y., Taya, M., Trans. JIM 60,
11631172, 1996
[12] Shimamoto, A., Zhao, H.Y., Abe, H., Intl.
J . Fatigue 26, 533542, 2004.
[13] Shimamoto, A., Furuyama, Y., Abe, H.,
Key Eng.Mater. 334335, 10931096,
2007.
[14] Loughran, G.M., Shield, T.W., Leo, P.H.,
Int. J. Solid Struct. 40 271294, 2003.
[15] Saadat, S., Salichs, J., Noori, M., Hou, Z.,
Davoodi, H., Bar-on, I., Suzuki, Y. and
Masuda, A. An Overview of Vibration and
Seismic Application of NiTi Shape Memory
Alloy, Smart Mat. and Struc., , Vol. 11, p.
218229, 2002.
[16] Van der Eijk, C., Zhang, Z.L. and
Akselsen, O.M., Seismic Dampers Based
on Shape Memory Alloys: Metallurgical
Background and Modeling, Proc. Third
European Conference on Structural Control,
ECSC, Vienna, Austria, 12-15 J uly 2004, p.
M1-5.
[17] Song, G., Ma, N., and Li, H.-N.,
Applications of Shape Memory Alloys in
Civil Structures, Eng. Struc., 2006, Vol.
28, no. 9, p. 1266-1274.
[18] Janke, L., Czaderski, C., Motavalli, M.,
and Ruth, J . Applications of Shape
320
Memory Alloys in Civil Engineering
Structures-Overview, Limits and new
Ideas, Mat. and Struc., 2005, Vol. 38, p.
578-592.
[19] Dolce, M. and Cardone, D., Mechanical
Behaviour of Shape Memory Alloys for
Seismic Applications; Austenite NiTi Wires
Subjected to Tension Int. J. of Mech. Sc.,
Vol. 43, no. 11, p. 2657-2677,2001.
[20] W. Huang, Mater. Design 23 1119, 2002.
[21] Mares, M. A., Carcea, I., Chelariu ,R.,
Roman, C., Mechanical and tribological
properties of aluminiummatrix composites
after ageing treatments, Euromat 97 -
proceedings of the 5th european conference
on advanced materials and processes and
applications: materials, functionality &
design, vol. 1 - metals and composites pg.
423-426, 1997.
321
STUDY OF THERMOMECHANI CAL FATI GUE FOR SHAPE
MEMORY ALL OYS TYPE CuZnAl
Achitei Dragos Cristian
1
, Hopulele I oan
1
, Perju Manuela Cristina
1
, Mihai Axinte
1
1
Gheorghe Asachi Techni cal Uni versity of I ai, dragos_adc@yahoo.com
Abstract: The paper presents the design elements of a prototype installation for the study of specific
thermal/mechanical fatigue of shape memory alloys, and fatigue curves obtained respectively.
Keywords: shape memory alloys, fatigue curve, prototype installation
1. Introduction
Memory shape alloys from some installations
are submitted to mechanical stresses and to thermal
stresses so that their integration within a fatigue
calculus system supposes taking into consideration
their function.
Figure1. Actuators
Memory shape alloys have a series of properties
alike the other ordinary metallic materials. Between
those, an important characteristic is their capacity
of changing the geometric shape from a low
temperature to a high one.
Under certain circumstances, shape changing
can be reversible so that the materials can
memorize two geometric shapes such as the high
temperature shape and low temperature shape.
These transformations accomplish because of an
effect named memory shape effect. By memory
shape effect, one can understand that the material
can make labor work when passing from the cold
shape to the warm one.
The pieces subjected to variable (cyclic) stresses
destroy to stresses, which are inferior to static
fracture resistance. The maximum cyclic stress
when a material does not brake calls fatigue
resistance of the material.
In order to dimension the constructive shape
of an element with memory shape from a
mechanic device, a series of numeric values are
necessary, among these values being also fatigue
resistance, labeled by the minimum value of the
deformation recovered after a certain number of
use cycles.
In addition, besides the phenomena met in
classic crystalline materials, memory shape
alloys present supplementary connection
mechanisms at phase change, which are
characteristic only for memory shape alloys.
Taking into consideration a device where
memory element makes a double sense memory
shape effect, within work system (resort
coupling uses for recovery), fatigue resistance
limit defines by the number of cycles until
recovery tension lowers to a minimum value
(approximately 70% from the initial one).
According to the cycling type, a memory
shape alloy can present irreversible deterioration
phenomena of the microstructure defined by
specific categories of fatigue.
If cycling though double effort, fatigue is
thermal and by cycling through simple effort of
memory shape thermo-mechanical fatigue
appears.
The fatigue of metals is the phenomenon that
produces the breakage of different pieces, under
temperature variation conditions and other work
parameters, too.
322
Mechanical fatigue implies breakage production
in the following stages: defects accumulation,
cracks formation and conduction, in stationary
regime at first and non stationary at final breakage.
Figure 2. Fatigue cracks appeared in the areas marked
with 1, they conducted in the interior of the piece and the
final breakage produced quickly in the areas marked
with 2.
The breakages that appear when applying some
variable loads are called fatigue breakages probably
because it noticed generally they appear only after a
considerable period.
A fatigue breakage is very dangerous because it
appears as a preliminary warning. The proper
fatigue breakage is fragile, without important global
deformations. At macroscopic scale, breakage
surface is usually perpendicular on the direction of
the main normal stress.
A fatigue breakage can be easily recognized
based on the aspect of the fracture surface, which
presents a smooth area and a rough area.
Many times, the way breakage advanced is
indicated by a series of concentric circles, which
advance from the initial point of the breakage
towards the interior of the section.
The stresses producing fatigue breakage at high
temperature cannot appear due to some mechanic
causes. Fatigue fracture can be produced by
variable stresses due to temperature, under
conditions where stresses due to mechanic causes
do not produce.
Thermal stresses produce when the variation of
a piece dimension is stopped in a certain way.
If the breakage appears due to a single appliance
of thermal stresses, the stress calls thermal shock.
If the breakage appears after many appliances of
some thermal stresses of low values, the solicitation
calls thermal fatigue. In the equipments working at
high temperatures, there are usually created the
premises for breakages due to thermal fatigue.
Figure 3. Cracks appeared due to thermal fatigue.
2. Experimental installation
In order to study thermo-mechanical
phenomenon of memory shape alloys it designed
and realized a prototype with a complex
configuration.
Figure 4. Standard samples, from memory shape
alloys type CuZnAl, for thermal fatigue test.
The number of cycles necessary for a loading
is 104 105 cycles and functioning time for the
achievement of these cycles is of the order of
tens hours.
Standard sample made of memory shape
alloy will be subjected to traction, being caught
into the dies, loading realizes by a system of
levers, and at their ends, some weights are
attached.
Samples with memory shape alloys were
tested with the chemical composition presented
in table 1.
Table 1
Cu Zn Al Fe Si Mn
75.4 18.6 5.85 0.021 0.026 0.007
Conventional notation: Cu
75.4
Zn
18.6
Al
5.85
.
323
The scheme of the prototype installation
presents in figure 5.
Figure 5. Prototype installation for thermo-mechanical
fatigue testing: 1 command panel; 2 rigid metallic
frame; 3, 4 systemof bearings and levers for load
multiplication on the sample; 5 counter weight; 6
engine assembly, reducer and armfor weight lifting; 7
weights; 8 comparator; 9 shell of the heating
chamber; 10 sample made of memory shape alloy; 11
jigs; 12 fixed die.
The dies used for traction are positions in a
metallic chamber, foreseen with a transparent
visiting cover. The sample will be cyclic heating
and cooled in the range of temperatures 40
100C, by means of an installation that blows
warm and cold air.
Figure 6. Metallic chamber where the jigs are
positioned, measurement system of temperature of
sample.
According to the weights attached to levers
system, the sample will be stressed with a load
whose size is directly proportional with the used
weight. This proportionality achieved by means of
the levers system. For a weight of 5 kg, traction
force on the sample ends will be of 500 kgf.
During experiments, the sample made of
memory shape alloy will suffer an elongation,
which will be determined with a comparator.
The parameters followed on command panel
and a computer using the software named XMEM
can control the entire experimental process. As
well, the number of heating-cooling cycles will be
recorded by XMEM software.
Figure 7. Detail of XMEM software in function.
A very important problem of this installation is
the synchronization of the thermal cycle with
mechanical cycles. Importance that consists in the
fact that memory shape alloys need to determine
thermo-mechanical fatigue during material
education. Material education realizes in double-
effect memory shape materials.
The samples were also analyzed by means of a
dilatometer to observe modifications on the
characteristic points and implicitly if amnesia
phenomenon appears if the alloy losses its
memory shape effect.
The samples are subjected to a variable,
chosen arbitrarily, thermal /mechanical fatigue
cycles. After a certain number of cycles, the
specimen was analyzed dilatometric, following
contraction changes during heating (EMF) and
how the temperature transformation martensite -
austenite varies.
The Cu
75.4
Zn
18.6
Al
5.85
alloy in quenching and
3% range elongated presents the martensite
austenite transformation domain at 65.8C and
102.2C temperatures. After 100 cycles, the
domain have moved at the 40,7C and 99,5C
temperatures.
The maximum of contraction have a growth
from 25 m to 75 m. At 6000
thermal/mechanical fatigue cycles we have a
contraction from 180 m.
0
20
40
60
80
100
120
0 1000 2000 3000 4000 5000 6000 7000 8000
number of cycles [N]
m
a
x
i
m
u
m

c
o
n
t
r
a
c
t
io
n
[
m
]
maximumcontraction
transformation point Ms
transformation point Af
Figure 8. The maximum variation of contraction
function the number of cycles of thermal / mechanical
stress.
324
After 12865 thermal fatigue/mechanical cycles
on the sample surface, subjected to experimental
tests on the prototype installation, have appeared
some cracks (figures 9 and 10).
Experimental tests on this sample have been
stopped. If it had continued, these cracks could be
moved inside the sample, and finally they would
have broken it.
Following shows the appearance of
microcracks micrographs at 500X magnification.
Figure 9. Microcracks on surface sample from
Cu
75.4
Zn
18.6
Al
5.85
alloy, magnification 500 X.
Figure 10. Microcracks on surface sample from
Cu
75.4
Zn
18.6
Al
5.85
alloy, magnification 1000 X.
4. Conclusions
The applications of memory shape alloys
impose the determination of functioning period
for pieces that work under the conditions of under
load cyclic heating.
The determination of the characteristics for a
memory shape alloy allows components producers
to assure a certain guarantee on these products.
The achievement of the prototype installations,
as well the creation of a standard of thermo-
mechanical fatigue for a series of analyzed alloys
would be a solution for the producers of memory
shape components.
The development of thermal/mechanical
fatigue process is characterized by a diminution of
contraction, and finally the destruction of the
sample.
4. References
[1] Stefan, M., Vizureanu, P., Bejinariu, C.,
Badarau, Gh., Manole, V., Studiul
proprietilor termice ale materialelor,
Editura Tehnopress, Iai, 2008, ISBN 978-
973-702-566-1.
[2] Stanciu S., Materiale cu memoria formei
metode de investigaie si aplicaii in tehnica,
2009, editura Universitas XXI.
325
STUDY REGARDI NG THE AUDI T OF MANAGEMENT PRI NCI PLES
Costel Mironeasa
1
, Silvia Mironeasa
2
, Georgiana Gabriela Codin
3
University tefan cel Mare, Suceava, costel@fim.usv.ro
1
, silviam@fia.usv.ro
2
Abstract: In this study we developed a check list and we applied a scale to measure the
implementation quality of the management principle, as set in the standard ISO 9000. The audit
was for the organizations that are included in small and medium organizations category, from
North Moldavia and Banat.
Keywords: principles, audit, discriminate analysis, standards.
1. Introduction
Management must fulfill quality criteria.
Applying the principles of quality management to
management practice is the base transition to the
total quality of applying the management functions
[1]. Modern management systems must answer the
challenges linked to quality assurance, employee
safety, information security, environmental
protection or social responsibilities and for some
specific organizations, to the insurance of food
safety.
The term Total Quality Management (TQM)
seems to be an universal remedy that can be
applied to all organization regardless of the profile
processes which take place in the organization. In
practice, this concept is transposed in a new culture
[10] with values [11], principles and procedures
[12]. The management principles can be found in
the TQM and are especially adopted by the
organizations which implement the requests
specified in the standard 9001 [7].
The standard 9000 sets forward the eight
principles regarding:
1. Customer focus
2. Leadership
3. Involvement of people / stakeholders
4. Process approach
5. System approach to management
6. Continual improvement
7. Factual approach to decision making
8. Mutually beneficial supplier relationships
The implementation of the standards has as
purpose the management of the management
systems, regarding the compliance of these
principles in order to create a culture oriented
towards the quality of the processes from the
system [3]. The requirements expressed by the
standards SR EN ISO 9001 [9]; SR EN ISO 14001
[5]; SR OHSAS 18001 [8]; SR EN ISO 27001 [6]
etc. are implemented in order to answer two
important directions: ensuring satisfaction of
clients/interested parties and ensuring compliance
with the requirements of the standard. Requirements
specified in table 1 reflect the management
principles from the analyzed standards.
The management systems are implemented in
organizations following the quality assurance
cycle: plan, do, check and act. Client orientation
and the interested parties involve a strategic
approach of all the processes that are part of the
system. The evaluation of the functionality of the
system can be obtained through monitoring,
control and audit [4].
The management of the organization uses the
audit as one of the instruments which can help
evaluate the status of the audited entities. Audit, as
an instrument used by management determines: if
the management principles are complied to, if there
are established objectives for the process
development, if there are performance criteria
which establish targets for the development of
activities which account as a process, which are the
results of the processes and if the results
correspond to the pre-established criteria, the level
of compliance of the requests to the objectives, etc.
The audit has the great advantage that it allows
obtaining much more complete information than
the administration of questionnaires to determine
the status of a system or process.
326
Table 1. Management principles reflected in the requirements from standards ISO 9001; ISO 14001; ISO
18001and ISO 27001
Crt.
no.
Management
Principles
ISO 9001
ISO 14001 /
OHSAS 18001
ISO 27001
1. Customer-Focused
Organization
5.2; 7.2 (7.2.1; 7.2.2; 7.2.3); 7.3.6;
7.5.2; 7.5.4; 8.2.1
4.3 (4.3.1; 4.3.2);
4.4.3; 4.4.6
5.1;
2. Leadership 5.1; 5.3; 5.4 (5.4.1; 5.4.2); 5.5
(5.5.1; 5.5.2; 5.5.3); 5.6 (5.6.2;
5.6.3); 6.1; 6.2.2; 6.3; 6.4; 7.3.1;
7.3.4; 7.4.1; 7.5 (7.5.1; 7.5.2;
7.5.3); 8.2 (8.2.3; 8.2.4); 8.3; 8.4;
8.5 (8.5.1; 8.5.2; 8.5.3)
4.2; 4.3.3; 4.4
(4.4.1; 4.4.2;
4.4.3; 4.4.6;
4.4.7); 4.5 (4.5.1;
4.5.2; 4.5.3); 4.6
4.2 (4.2.1; 4.2.2;
4.2.3); 5.1; 5.2.1; 8.1
(8.2; 8.3)
3. Involvement of
People
5.5 (5.5.1; 5.5.3); 6.2.2 4.4 (4.4.1; 4.4.2;
4.4.3)
5.1; 5.2.2
4. Process Approach 02; 7.1; 7.5.2; 8.2.3; 8.4 4.5.1 02;
5. System Approach to
Management
02; 4.1; 5.1; 5.4.2; 5.5.2; 5.6
(5.6.1; 5.6.2; 5.6.3); 6.1; 7.1;
7.4.2; 8.1; 8.2 (8.2.1; 8.2.2; 8.2.3);
8.4
Introduction; 1;
4.2; 4.3 (4.3.1;
4.3.2); 4.4 (4.4.1;
4.4.2; 4.4.3); 4.6
4.2 (4.2.1; 4.2.2)
6. Continual
Improvement
1.1; 4.1; 5.1; 5.3; 5.6 (5.6.1; 5.6.2;
5.6.3); 6.1; 8.1; 8.4; 8.5.1
4.1; 4.2; 4.3.3; 4.4
(4.4.1; 4.4.2); 4.6
4.2.2; 7.2; 8.1; 8.2;
8.3
7. Factual Approach to
Decision Making
01; 4.1; 5.6.3; 6.2.2; 5.1; 5.3; 5.5
(5.5.1; 5.5.3); 6.3; 6.4; 7.2.3;
7.3.1; 7.4.2
4.4; 4.4 (4.4.1;
4.4.3; 4.4.3;
4.4.6); 4.6;
4.1; 5.2.1; 7.2
8. Mutually Beneficial
Supplier
Relationships
02; 5.1; 5.3; 5.5 (5.5.1; 5.5.3); 6.3;
7.2 (7.2.2; 7.2.3); 7.3 (7.3.1;
7.3.3); 7.4.2; 7.5.1; 8.2.1; 8.4
4.2; 4.4; 4.3.1;
(4.4.1; 4.4.3;
4.4.6; 4.4.7); 4.5.1
5.1;
2. Methodology
We focused on two regions from Romania:
Banat (Timioara and Rea counties) and North-
Moldavia (Suceava and Botoani counties). In total
34 organizations have been audited from which 19
in the Moldova area and 15 in the Banat area. In
each organization all top managers and 82
intermediary managers were audited: 46 in the
Moldova area, 36 in the Banat area. The number of
management systems implemented in the
organizations was different, between one system
and up to four systems. Quality management
systems (QMS) were implemented in all the
organizations. Besides this system the following
systems were also implemented: Environmental
Management System (EMS); Occupational Health
and Safety Management Systems (OHSMS);
Information Security Management System (ISMS).
The audited management systems were found in
the following combinations: QMS +EMS; QMS +
EMS + OHSMS; QMS + EMS + OHSMA +
ISMS.
Information gathering was achieved by direct
interview with each process owner found in the
organizational process of the organization. They
were addressed questions which had the purpose to
obtain answers referring the way in which the
management principles are implemented.
3. Results and discussions
The evaluation of the way in which
management applies the management principles
was achieved by following two directions: the
evaluation of the way in which management
perceives the level of importance of the
management principles and the evaluation of the
way in which the management principles have
been implemented. The implementation stage of
the principles was done by the following
considerations of the auditor: met, partially met,
and not applied.
The evaluation was done through statistical
modeling and it was applied to the results in table 2
in order to see to which degree are there
differences between the way the organizations,
from the analyzed regions, approach the
327
management principles. The data from table 2 was
adjusted according to the number of participants.
The number of managers which answered was
taken into consideration and also the number of
queries for each principle. The obtained data
represent the average of each considered criterion.
Each principle was assigned a number of questions
which were addressed to the top and executive
managers.
Table 2. The resulting data
Region Management Criteria P1 P2 P3 P4 P5 P6 P7 P8
Moldova Top
M 0,789 0,880 0,833 0,805 0,861 0,900 0,861 0,860
PM 0,114 0,082 0,105 0,128 0,096 0,074 0,090 0,079
NM 0,096 0,038 0,061 0,068 0,043 0,026 0,049 0,061
Executive
M 0,746 0,786 0,594 0,742 0,769 0,815 0,629 0,645
PM 0,132 0,145 0,264 0,099 0,146 0,061 0,182 0,225
NM 0,121 0,069 0,141 0,158 0,085 0,124 0,189 0,134
Banat T
M 0,822 0,926 0,933 0,838 0,873 0,927 0,919 1,056
PM 0,122 0,041 0,089 0,143 0,079 0,053 0,038 0,122
NM 0,078 0,033 0,022 0,019 0,048 0,020 0,043 0,078
E
M 0,343 0,386 0,389 0,349 0,364 0,386 0,383 0,440
PM 0,051 0,017 0,037 0,060 0,033 0,022 0,016 0,051
NM 0,032 0,014 0,009 0,008 0,020 0,008 0,018 0,032
No. of questions 12 18 6 7 11 10 14 6
The obtained data were processed according to
the multiple discriminator analysis method (). The
basic technique of the analysis is to estimate the
position of an element towards a line which better
separates two different population classes. With
the current method, estimations can be made upon
the belonging of certain elements to one or
multiple categories. Application of the
discriminator analysis sought to determine a
minimum number of variables that can express the
separation of behavior which managers have
towards the approach of management principles.
The determination of the associated model follows
the calculation of the sums, the square sums and
the sum of products for the eight characteristics,
without taking into consideration the differences
between the categories of managers from the two
regions. The next step was to determine deviations
from their mean values of the square and product
sums with the relation (1).
12
X X
X X y y
12
1
12
1
j i 12
1
12
1
j i j i

(1)
The obtained data was entered in a system with
eight equations with eight unknowns in order to
determine the value of the coefficients a
i
, i=1...8
(2).
y
1
y
1
y
2
y
1
y
3
y
1
y
4
y
1
y
5
y
1
y
6
y
1
y
7
y
1
y
8
y
1
a
1
z
1
(2)
y
1
y
2
y
2
y
2
y
3
y
2
y
4
y
2
y
5
y
2
y
6
y
2
y
7
y
2
y
8
y
2
a
2
z
2
y
1
y
3
y
2
y
3
y
3
y
3
y
4
y
3
y
5
y
3
y
6
y
3
y
7
y
3
y
8
y
3
a
3
z
3
y
1
y
4
y
2
y
4
y
3
y
4
y
4
y
4
y
5
y
4
y
6
y
4
y
7
y
4
y
8
y
4
a
4
=
z
4
y
1
y
5
y
2
y
5
y
3
y
5
y
4
y
5
y
5
y
5
y
6
y
5
y
7
y
5
y
8
y
5
a
5
z
5
y
1
y
6
y
2
y
6
y
3
y
6
y
4
y
6
y
5
y
6
y
6
y
6
y
7
y
6
y
8
y
6
a
6
z
6
y
1
y
7
y
2
y
7
y
3
y
7
y
4
y
7
y
5
y
7
y
6
y
7
y
7
y
7
y
8
y
7
a
7
z
7
y
1
y
8
y
2
y
8
y
3
y
8
y
4
y
8
y
5
y
8
y
6
y
8
y
7
y
8
y
8
y
8
a
8
z
8
328
The determined coefficients a
i
, i=1...8 offer
information about the relative importance of each
management principle as assessed by the two
studied groups. In order to determine the
importance of one or another characteristic the
coefficients of the associated model are compared.
The calculated values were adjusted to their score
dispersion values, without taking into account the
region and type of manager. The values of the
coefficients are given in table 3. The coefficient
with the largest algebraic value will have the
highest degree of importance, in this case a
2
=
1,397.
Since the difference between the average score
obtained by the managers from Moldova and Banat
for a management principle P
i
, i=1...8, has the
smallest value for P8, this attribute has a less
important discrimination degree for the analyzed
category. The difference between the average
given points is 0,037.
Table 3.
Principle no. P1 P2 P3 P4 P5 P6 P7 P8
Coefficients a
i
1,997 3,83 1,870 2,276 -3,538 -3,781 2,745 -4,540
Standard Deviations 0,311 0,3647 0,3269 0,3205 0,3442 0,3739 0,3355 0,3534
Coefficients a
i
0,6205 1,397 0,6113 0,7295 -1,218 -1,414 0,921 -1,6042
Importance level 4 1 5 3 6 7 2 8
Average difference 0,092 0,097 0,087 0,097 0,097 0,097 0,097 0,037
4. Conclusions
Discriminator analysis provides us with
information on the level of importance which the
management principle has. The use of the audit has
allowed obtaining results through which it can be
established that the leadership principle is the most
important. The second principle, as level of
importance, is linked to a management feature,
Factual Approach to Decision Making. These
findings show that managers believe that in order
to implement a management system, the most
important elements are linked to management. A
surprising element is the location on the second
smallest spot of the principle referring to Continual
Improvement. This is a sign of weakness of the
management system which can be explained by a
lack of understanding of the benefits arising from
reducing the risk of non-conformities.
5. References
[1] Fausto, G., The golden integral quality
approach: From management of quality to
quality of management, Total Quality
Management, J an 1999, Vol. 10 Issue 1, p 17-
35.
[2] Gunaselaran, A., Enablers total quality
management implementation in
manufacturing: a case study, Total Quality
Management, vol. 10, no. 7, 1999, 987-996.
[3] Irani, Z., Beskese, A., Love, P.E.D., Total
quality management and corporate culture:
constructs of organisational excellence,
Technovation, Volume 24, Issue 8, August
2004, Pages 643-650
[4] ISO, 2002, ISO 19011, International Standard:
Guidelines for Quality and/or Environmental
Management Systems Auditing. International
Organization for Standardization, Geneva,
Switzerland.
[5] ISO, 2004, ISO 14001, Environmental
management systems. Requirements with
guidance for use
[6] ISO, 2005, ISO/CEI 27001, Information
technology. Security techniques. Information
security management systems. Requirements
[7] ISO, 2005. ISO 9000, Quality management
system - Fundamentals and vocabulary
[8] ISO, 2007, OHSAS 18001, Occupational
health and safety management systems.
Requirements
[9] ISO, 2008, ISO 9001, Quality management
systems - Requirements.
[10]Matta, K., Davis, J ., Mayer, R., Conlon, E.
Research questions on the implementation of
total quality management, Total Quality
Management, no. 7, 1996, 39-49.
[11]Stephan M. Wagner, S.M, Eggert, A.,
Lindemann, E., Creating and appropriating
value in collaborative relationships, J ournal of
Business Research, Volume 63, Issue 8, 2010,
840-8
329
I NTEGRATI ON OF SI X SI GMA AND LEAN PRODUCTI ON SYSTEM FOR
SERVI CE I NDUSTRY
Militaru Emil
1
Abstract: Six Sigma is one of the most popular quality initiatives recently. Lean Production System is the
world famous production system developed and practiced by Toyota mobile company for a long time. It is
based on two concepts: J ust-In-Time and J idohka. Both are based on varying thinking to improve the
business process, enhance quality, production and competitive position. Besides, the integration of them is
viewed as a new trend in the next management wave.
Keywords: six sigma, lean production system, service industry
1. I NTRODUCTI ON
Nowadays, the development of industrial
core intrinsic technologies is more and more
important due to the dynamic competition in the
global market. Companies have to keep and
continuously upgrade their intrinsic technologies
in the professional field to gain the sustainable
competitive advantage. However, they also have
to continuously upgrade their management
technologies, and keep sensitive to the latest
issues as well as their integration with companys
current system. Otherwise, they still cannot
survive in the market even though their intrinsic
technologies are advanced.
No matter how the management technologies
be developed, they must emphasize their
contribution to business performance, customer
satisfaction and continuous improvement of the
products or services. Moreover, the integration of
different systems is an important issue today and
tomorrow.
This research will focus on Six Sigma and
Lean Production System to discuss their
integration based on the background and thoughts
mentioned above. Six Sigma is one of the most
popular quality initiatives recently. Lean
Production System is the world famous
production system developed and practiced by
Toyota mobile company for a long time. It based
on two concepts: J ust-In-Time and J idohka.
Both are based on the variation in thinking in
order to improve business process, enhance
quality, production and competitive position.
Besides, the integration of them is viewed as a
new trend in the next management wave.
Moreover, regarding the industry
characteristics, service industry is quite different
from manufacturing industry. Even though there
are more wastes and improvement opportunities,
the application of Six Sigma, Lean Production
System or their integration in service industry is
quite few neither in literatures nor practice.
This research proposes the Lean Six Sigma
integration model based on the research gap and
the practical need, and then adapt it for service
industry. The model is named as Lean Six
Sigma for Service (LS
3
) in this research. It
balances the viewpoints of internal and external
customers, and gives consideration to the Lean
speed as well as Six Sigma high quality. Also,
this research tries to contribute to the
enhancement of management technologies.
2. Introduction to Six Sigma
Six Sigma is the major focus of many companies
for its powerful breakthrough performance
demonstrated in GE, Motorola etc. recently. Six
Sigma can help companies to reduce cost, increase
profits, keep current customers and create new
customers. In brief, Six Sigma is a methodology
330
to reduce the variation of every process and their
interfaces to achieve a very high quality level.
In statistical theory, six sigma is an ideal
target value, and expressed as: 6. It means when
the process or product we observed under a
normal distribution, the probability of a specific
attribute value shifts from the mean about positive
or negative six standard deviation would be 0.002
part per million (ppm). Motorola company found
a phenomenon that the process mean would shift
around the center point of specifications in a long-
term processing, and the shifting range would be
about positive or negative 1.5 standard deviations
from the center point of specifications.
Hence, Motorola company modified the
statistical meaning of six sigma. The definition
can allow the sample mean shifts from the center
of the population, and the observed process or
product would out lie the six sigma limits only 3.4
times per million operations under the original
specifications. In addition, the sigma performance
can also be expressed by Defect Per Million
Operations (DPMO) shown as Table 1.
Yield DPMO Shift
from
Mean
Popular
Age
6.68 % 933200 0
30.9 % 690000 1
69.2 % 308000 2 1970s
93.3 % 66800 3 1980s
99.4 % 6210 4 Early
1990s
99.98 % 320 5 Mid 1990s
99.9997
%
3.4 6 2000s
Table 1. DPMO and Sigma Performance
Six Sigma means the world leading quality
level. More and more companies understand to
use Six Sigma to improve the process quality so
as to achieve the business dramatic performance.
This is because Six Sigma requires the
quantitative measurements and analyses of the
core business processes as well as suppliers
involved processes.
Originally, Six Sigma methodology is
applied to manufacturing industries. However, the
applications of Six Sigma are no longer be limited
in manufacturing processes today. Keim (2001)
demonstrated Six Sigma is very suitable to
improve the service performance by two real
cases. Paul (2001) pointed that the recent trends in
Six Sigma are: emphasis on cycle time reduction,
smaller business deployment, and integration with
other initiatives.
As the Six Sigma market grows, so does the
availability of organizations to assist in
deployment and integration. This availability of
technical expertise allows smaller businesses
realistically consider Six Sigma deployment with
minimal economic investment. Besides, due to the
central concern of Six Sigma is to pursue the
customer satisfaction and business performance,
we can view Six Sigma a main structure while
integrating with other initiatives. As for the
integrating initiatives such as Lean Production
System, Total Quality Management or Quality
Costs etc. depend on the different requirements of
each company.
2.1 I ntroduction to Lean Production System
Lean Production System (also called Toyota
Production System) is the world famous
production system developed and practiced by
Toyota mobile company for a long time. It based
on two concepts: J ust-In-Time and J idohka.
This kind of production system is very flexible to
the dynamic change of market demands, and Lean
Production System is established by many small
group improvement activities to eliminate all
kinds of wastes in the business.
An important literature written by Spear and
Bowen (1999) published in Harvard Business
Review pointed that, the Toyota Production
System and the scientific method that underpins it
were not imposed on Toyota they were not even
chosen consciously. The system grew naturally
out of the workings of the company over five
decades. As a result, it has never been written
down, and Toyotas workers often are not able to
articulate it. Thats why its so hard for outsiders
to grasp. In the article, Spear and Bowen
attempted to lay out how Toyotas system works.
They tried to make explicit what is implicit.
Finally, they described four principles three
rules of design, which show how Toyota sets up
all its operations as experiments, and one rule of
improvement, which describes how Toyota
teaches the scientific method to workers at every
level of the organization. It is these rules and not
the specific practices and tools that people
331
observe during their plant visits that in their
opinion form the essence of Toyotas system.
Hence the two authors called the rules as the DNA
of the Toyota Production System.
These rules guided the design, operation, and
improvement of every activity, connection, and
pathway for every product and service. The rules
are as follows:
Rule 1: All work shall be highly specified as
to content, sequence, timing, and outcome.
Rule 2: Every customer-supplier connection
must be direct, and there must be an
unambiguous yes-or-no way to send requests
and receive responses.
Rule 3: The pathway for every product and
service must be simple and direct.
Rule 4: Any improvement must be made in
accordance with the scientific method, under
the guidance of a teacher, at the lowest
possible level in the organization.
All the rules require that activities,
connections, and flow paths have built-in tests to
signal problems automatically. It is the continual
response to problems that makes this seemingly
rigid system so flexible and adaptable to changing
circumstances.
2.2 Four Characteristics of Service I ndustry
Recently, due to the economic and international
trading environmental change, the structures of
many companies are also changed. The growth of
service industries rapidly chases the growth of
manufacturing industries. Especially for the
current situation in Taiwan, many factories are
moving to mainland China. Hence, the needs for
service industries to fill in the space of economic
activities become very huge. Thats why service
industries play an important role in the economic
development recently.
This research concludes the four
characteristics of service industries based on the
literatures written by Kotler (1997), Regan (1963)
and Zeithmal, Parasur& Berry (1985) as follows:
1. Intangibility: It means that services
can be consumed and perceived, but
they cannot easy to be objective
measured like the manufactured
products. Thats why there is usually a
perception gap between the service
provider and consumer.
2. Variability: It means that services are
delivered by people, so the service
quality may change depending on
different time, people and consumer
perception. That is, the variability of
services.
3. Perishability: Unlike the tangible
manufactured products, services cannot
be inventoried. They are delivered
simultaneously while the demands from
consumers appear. Once the demands
disappear, the services perish.
4. I nseparability: Since the delivery and
consumption of services almost be
done simultaneously. Hence the
interactions between servers and
consumers play an important role on
the evaluation of service quality.
Consumers evaluate the service quality
on the moment of consuming the
service.
2.3 MODEL CONSTRUCTI ON
This research proposes an integration model of
Six Sigma and Lean Production System for
service industry called as Lean Six Sigma for
Service (LS
3
). In practice, the first stage is to
Lead the process improvement project by
hearing the Voice of Customer (VOC). The
project identification and its scope must be
clarified so as to serve the customers more
efficiently and effectively by the improvement.
The Lead stage provides the project team a
well-defined scope of the problem they are faced.
Hence the major mission of the Study stage is to
measure the current status or level by quantitative
data, and then to analyze how the problem affects
the process. By the collection of Voice of
Process (VOP), the project team can try to
converge the problem and begin to find out its
root causes.
Moreover, no matter what the process
indicators are, the project team has to well define
them first, and to explain the purpose as well as
the use of each indicator. Most important of all,
the performance indicators definition and their
evaluation methods must be agreed and confirmed
by the people involved. Its very important to
detail record the performance levels and action
332
results for the further enhancement of service
processes.
After studying and analyzing the current
status of the service process targeted, the third
stage is to draw up problem solving
countermeasures. The countermeasures must be
transformed to the Voice of Server (VOS) so as
to Smooth the whole service process. It is hoped
to reduce the defects and speed up the service
delivery by the verification of performance
indicators. Therefore, the project team will
propose some education and training plans for the
coming countermeasures conduction.
The project team can measure if the project
target achieved or not by the proposed
countermeasures so as to continuously monitor
and control the better results. At this time, the
project team has to Sustain the operating
stability of the service process. Therefore, the
purpose of this stage is to confirm the result, the
effectiveness of the countermeasures and if there
is any side effect. Once these things are confirmed
feasible, the project team can view the knowledge
and experiences as the base of knowledge
management and technology accumulation.
Finally, the knowledge and experiences must be
diffused and deployed throughout the organization
so as to be the Voice of Business (VOB).
The LS
3
operating model proposed by this
research shown as follows:
Figure 1. Structure of Implementing LS
3
The key points and tools of implementing
LS
3
are concluded by this research and shown as
Table 2. Moreover, the tools of LS
3
are also
shown as Figure 2.
Table 2. Key Points and Tools of Implementing
LS
3
LS
3
Activities Tools
Lead
Lead the process
improvement
projects by
hearing the voice
of customer
(VOC)
- - Identify the
process
improvement
project
- - Define the
project
performance
indicators
- - Select the
project team
members
- - Accomplish the
project charter
and job
assignments
- - Market survey
- - Project charter
- - Annual police
deployment
- - Quality function
deployment
- - Value stream
analysis
Study
Study and
analyze the
current status of
targeted process
to get the voice of
process (VOP)
- - Observe the
actual process,
and measure
the baseline
- - Analyze the
collected data
to understand
the present
situation
- - Confirm the
problem and
critical-to-
quality
- - Process
mapping
- - Measurement
system analysis
- - Motion and
time study
- - Multi-vari
analysis
- - Cause and
effect matrix
- - Process
capability
analysis
- - Time value
analysis
333
Smooth
Propose the
countermeasures,
and transform
them into the
voice of server
(VOS) to smooth
the service
process
- - Draw up the
improvement
countermeasure
s
- - Implement the
countermeasure
s to accelerate
the service
delivery
- - Confirm the
results by
performance
indicators
- - 5S
- - Operation
balancing
- - Rapid operation
switching
- - Visual
management
- - Eliminate,
combine,
rearrangement,
simplify
- - Process
reengineering
- - Failure mode
and effect
analysis
Sustain
Sustain and
control the
project results,
and spread out
the organization
to be the Voice of
Business (VOB)
- - Standardize the
effective
countermeasure
s to sustain the
results
- - Continuous
control the
improvement
level
- - Design the job
value of
employees in
- - Control chart
- - Check list
- - Process
standardization
- - Error proofing
- - Education and
training
the service
process
- - Knowledge
diffusion and
application
Figure 2. Tools of Implementing LS
3
2.3. CONCLUSI ONS
Due to the limitation of practical resources, the
LS
3
model demonstration by a real case could not
be included in this research. Therefore, this
research used the questionnaire survey to verify
the theoretical logic and feasibility of LS
3
structure. We interviewed several Lean
Production and Six Sigma experts and
consultants, and we expect to provide a base of
verification by their experiences and knowledge.
Finally, we conclude the agreements and
suggestions of the experts as follows:
1. All the experts and consultants agreed
the theoretical structure of the proposed
model by this research, and expressed
the very much agree level on the
logic, implementing steps and their
contents.
2. All the experts and consultants
expressed the very agree level on the
proposed model with PDCA
management cycle.
expressed at least the agree level on
the fitness for use of the proposed tools.
334
Some experts considered that the
names of some tools originally be
applied in manufacturing industry
should be changed to be well
understood in service industry. In
addition, all the experts and consultants
considered that the tools mostly applied
in manufacturing industry should also
be applied to service industry. This is
because there are huge demands and
expectations of these tools in service
industry based on their experiences and
observations. Hence they deeply agreed
the proposed model by this research
contains the tools for reference.
proposed model really integrated the
principles of Six Sigma and Lean
Production System. Some experts
suggested we could emphasize the
principles of Lean Production System a
little more.
feasibility of applying this model to the
service industry.
practical value of this proposed model,
and they also expressed that they will
refer the structure to demonstrate it if
there is any suitable opportunity in the
near future.
Therefore, i conclude the agreements and
suggestions by the experts as follows: all the
experts agreed the proposed model by this
research on the whole structure, implementing
steps and tools planned. They also expressed the
practical value and operational feasibility of this
model is very high. Moreover, all the experts
agreed this model on the fitness for use in the
service industry, and they will refer this model to
demonstrate it when there is any opportunity in
the near future.
References
[1] Breyfogle III, F.W., Cupello, J .M. and
Meadows B. (2001), Manging Six Sigma- A
Practice Guild to Understanding, Assessing,
and Implementing the Stagy that Yields
Bottom-Line Success, J ohn Wiley & Sons,
Inc., pp. 18-20;
[2] George, M.L. (2003), Lean Six Sigma for
Service, McGraw-Hill Inc., pp. 28-42;
[3] J iang, J ui-Chin, Chen, K. H. and Wu, M. C.
(2004), Integration of Six Sigma and Lean
Production , 33rd International Conference
on Computers and Industrial Engineering;
[4] Wu, M. C. (2003), Integration of Six Sigma
and Lean Production , Master Thesis,
Industrial Engineering Department, Chung-
Yuan Christian University;
[5] J iang, J ui-Chin, Fu, K. P. and Ku, S. F. (2002),
Construct the Integration Model of Six
Sigma and Lean Production , 8th Asia
Pacific Quality Organization Conference
Proceedings, pp. 282-289;
[6] J iang, J ui-Chin, Hsiao, F. Y. and Huang L. C.
(2002), Integration Model of Six Sigma for
Hospital 8th Asia Pacific Quality
Organization Conference Proceedings;
[7] Keim, E. (2001), Service Quality Six Sigma
Case Studies, Quality Congress.
335
OPTI MAL PERTURBATI ON FOR MI XED CONVECTI ON HEAT
TRANSFER I N RECTANGULAR CHANNEL
Rachid SEHAQUI
1
Facult des Sciences Ain Chock UFR de mcanique BP 5366 Maarif , Casablanca , Maroc Fax:
(212)(22) 23-06-74 Tel : (212)(22) 23-06-80E-mail: sehaqui@hotmail.com
Abstract: We studied the growth velocity perturbation at the entrance of a rectangular channel for
mixed convection heat transfer. The flow is bidimensionnel and the fluid is newtonian,
incompressible obeying to Boussinesq approximations. Particular forms of entry velocity profile
allowing a better heat transfer inside a channel for a given regime of flow is determined. Optimal
control method based on adjoint equations is used to obtain optimality condition.
Keywords: adjoint equations, convection mixte, optimal contro,l optimal perturbation
1. Abridged English version
The study of the mixed convection in a horizontal
channel, uniformly or partially heated, plays a
significant role in various technological processes,
such as, the chemical vapour deposition of solid
layers, cooling of microelectronic equipment's,
enhancing thermal efficiency of compact heat
exchangers or the calculation of the solar collector
energy. Physical phenomena that appear in mixed
convection, in a horizontal channel are complex
and depend on the Reynolds number Re, the
Rayleigh number Ra, the Prandtl number Pr, and
the geometry of the channel characterised by the
aspect ratio B =l /h. Previous investigations have
essentially focused on the classification of the
different regimes of flow following the values of
dimensionless numbers [1] [6].
In the last decade, increased attention has been
devoted to the development of techniques capable
of enhancing our ability to control the unsteady
flow in a wide variety of configurations such as
engine inlets and nozzles, combustors,
automobiles, aircraft, and marine vehicles. Gadel-
Hak [7] and Gad-el-Hak and Bushnell [8] provide
an excellent introduction and overview of various
control methodologies. In practice, for a given
controlling input, we choose an objective
functional to be maximised, and by a formal
maximisation procedure we obtain a system of
differential equations (state equations and adjoint
equations) and condition of optimality, whose
solution gives the optimal control. The results can
then be applied to develop mechanisms for control.
This approach used in particular by Cathalifaud
and Luchini [9] they attacked the problem of
controlling linear optimal perturbations in spatially
developing boundary layer on flat and concave
plates by considering the wall receptivity. The
optimal perturbations are the result of recent
development in the field of hydrodynamic stability.
For a given definition of disturbance energy, we
can calculate the input disturbance of the boundary
layer that maximises the output disturbance
energy, i.e the optimal perturbation. For the spatial
algebraic growth of steady disturbances leading to
bypass transition in the boundary layer over a flat
surface, this optimal input disturbance was
calculated by Andersson [10].
Many flows exhibit the following scenario: the
flow stars out in laminar state which is instable;
then a linear instability develops, and grows up to
the point where non-linear interactions finally lead
to turbulence. In many applications, it might be
very interesting to delay this transition towards
turbulence and thus to maintain the flow laminar.
The first objective, within this framework, of a
336
system of flow control will be to control heat
transfer by entry velocity profile.
This paper is organised as follows. In section 3, the
method of optimal control is presented. In section
4 the method of adjoint equations and the optimal
control. In section 5, the results are presented and
discussed
2. Introduction
Le transfert thermique par convection mixte
lintrieur dun canal rectangulaire prsente un
intrt considrable en vue dapplications
technologiques telles que : dpt chimique des
couches minces, refroidissement des pices
lectroniques, augmentation de lefficacit dun
changeur de chaleur, calcul de lnergie dun
capteur solaire et plusieurs autres domaines des
sciences de lingnieur. Les phnomnes qui
apparaissent dans le cas de la convection mixte
sont complexes et dpendent de plusieurs
paramtres : le nombre de Reynolds Re, le nombre
de Rayleigh Ra, le nombre de Prandtl Pr et du
rapport de forme B = l / h. Les conditions aux
limites tant celles fixant le profil de vitesse
lentre du tube que celles relatives la gomtrie
de la conduite ont une influence sur le transfert
thermique. De nombreux travaux ont cherch
corrler le transfert thermique aux conditions aux
limites thermiques imposes aux parois du tube ou
la forme gomtrique de la conduite [1]-[6].
Cette dernire dcennie, une attention accrue a t
consacre au dveloppement des techniques
capables de contrler la stabilit des coulement
dans plusieurs configurations. Gadel-Hak [7] et
Gad-el-Hak and Bushnell [8] fournissent une
excellente introduction aux diverses mthodes de
contrle. En pratique pour contrler les
coulements dentre, on choisit une fonctionnelle
objective maximiser, et par un procd formel de
maximisation nous obtenons un systme
dquations diffrentielles (des quations dtat,
des quations adjointes) et la condition
doptimalit, dont la rsolution permet dobtenir la
commande optimal de contrle. Cette approche a
t en particulier employe par Cathalifaud et
Luchini [9] ont trait le problme de contrle des
perturbations optimales linaires dans le cas de la
couche limite pour une plaque plane et concave.
Les dveloppements rcents dans le domaine de la
stabilit hydrodynamique sont lorigine de
lintrt accord aux perturbations optimales. Pour
une dfinition de lnergie de perturbation donne
on peut calculer la perturbation lentre de la
couche limite qui permet de maximiser lnergie
de perturbation la sortie, cest dire la
perturbation optimale. Andersson [10] a tudi la
croissance des perturbations rgulires permettant
dviter la transition, dans le cas dun couche
limite qui se dveloppe au dessus dune surface
plane.
On rappelle dans la section 3 la mthode du
contrle optimum. Dans la section 4 la technique
des quations adjointes. Dans la section 5 on
expose les rsultats obtenus.
3. Mthode du contrle optimum
Lide du contrle optimum est doptimiser la
fonctionnelle reprsentant la quantit physique que
nous cherchons contrler. Nous allons dcrire
dans ce paragraphe la mthode classique
doptimisation base sur les multiplicateurs de
Lagrange, Gunsburger [11]. Soit q la variable
dtat et soit r la variable de contrle ( que nous
voulons contrler ). On suppose que q et r vrifient
le problme suivant:
, ) 0 , = r q F
(1)
Ou , )
r q
F F F , =
. Lquation (1) est appele

quation de contrainte du problme direct ou
systme dtat. F
q
est la contrainte relative la
variable dtat q et F
r
celle relative la
variable de contrle r. On dfinit la
fonctionnelle gain , ) r q, I qui peut reprsenter
par exemple une distribution dnergie, de
temprature ou de frottements visqueuxetc.
Le problme qui consiste en la recherche de q
et r vrifiant les contraintes (1) et dont la
fonctionnelle , ) r q, I est minimale est appel
problme de minimisation . Lapproche
gnrale de la minimisation est de former le
vecteur multiplicateurs de Lagrange. La
fonctionnelle Lagrangienne est ainsi dfinie
par :
, ) , ) , )

, , , , , r q F r q r q L I = (2)
Ou ,. . reprsente le produit scalaire entre les
variables dtat (q,r) et les variables adjointes
337
, )
r q
, . Le problme doptimisation consiste
la dtermination de
, ,r q , pour , )
, r , q L
stationnaire et vrifiant le systme :
s contrainte des quations 0
adjointes quations 0
optimalit ' d conditions 0
=
c
c
=
c
c
=
c
c
=
c
c
r q
L L
q
L
r
L

(3)
4. Mthode des quations adjointes
4. 1. Problme direct
Dans le systme de coordonnes
cartsiennes(x*,y*), on considre lcoulement
plan dun fluide newtonien incompressible de
viscosit cinmatique v, dans un canal
rectangulaire, de longueur l et de hauteur h. La
temprature du plan infrieur (y* =0) est impose
le plan suprieur (y* = h) est adiabatique. On
introduit les quantits adimensionnelles suivantes :
T
T T
T
U
p
p
U
V
V
U
U
U
h
y
y
l
x
x
f
A
= = = = = =
*
,
*
,
*
,
*
,
*
,
*
2
0
0 0
o , , , T p V U dsignent respectivement la
vitesse longitudinale, la vitesse transversale, la
pression et le champ de temprature
adimensionnels et A et ,
0
U , T une diffrence de
temprature de rfrence, la vitesse moyenne, la
masse volumique. Les quations de Navier-Stokes
adimensionnelles scrivent sous la forme :
0 = c + c V U
y x
(4)
U P U V U U
yy x y x
c + c = c + c
1
Re (5)
T
Ra
U P V V V U
yy y y x
2
Re Pr
Re
1
+ c + c = c + c
T Pe T V T U
yy y x
c = c + c
1
(6) (7)
o
v
h U
Re
0
= est le nombre Reynolds,
vo
A |
3
Th g
Ra= le nombre de Rayleigh,
o
v
= Pr
le nombre de Prandtl, o la diffusivit thermique, g
lacclration de la pesanteur, | le coefficient
dexpansion thermique et Pe =Re Pr le nombre de
Peclet. Cet coulement principal fait lobjet dune
perturbation (u,v,p,u) lentre (e), et on tudie
linfluence de cette perturbation la sortie (s). On
cherche maximiser un gain G dfini partir de
lnergie de perturbation. On appelle perturbation
optimale la donne des fonctions u(x
e
,y), v(x
e
,y)
qui conduisent au gain maximum. Les quations
linarises pour la perturbation scrivent :
0 v= c + c
y x
u (8)
, ) u p U u V Uu
yy x y y x
c + c = c + c + c
1
Re v
, ) u
2
Re Pr
v
Re
1
v v
Ra
p V V u U
yy y y x x
+ c + c = c + c + c
u u u
yy x y y x
Pe U T T u U c = c + c + c + c
1
v
(9)
A ce systme on associe les conditions aux limites
suivantes :u(x,0) =v(x,0) =u(x,0) =, ) 0
1
= c
= y x
u ,
u(0,y) =u
e
(y) , v(0,y) =v
e
(y) , u(0,y) =u
e
(y).
4. 2. Problme adjoint
La majoration du gain G(u,v,p,u) moyennant
les quations (8)-(11) conduit ltude de la
fonctionnelle :
, ) , )
, ), ) , ) , ) ,
, ) , ) ,
, ) ,
dy dx
Pe T T u y x d
Ra p V V u y x c
u p U u V Uu y x b u y x a
p u G p u d c b a L
f
e
x
x
yy y y x x
yy y y x x
yy x y y x y x
c c + c + c + c
+ c c + c + c + c
+ c c + c + c + c + c + c
=
} }
1
0
1
1 2 1
1
V v U ,
Pr Re - v Re v v U ,
Re - v , v ,
, , v , , , v , , , , ,
u u u
u
u u
Dans laquelle a, b, c et d sont les multiplicateurs de
Lagrange. La recherche de la solution optimale
revient crire , ) 0 L Grad p u d c b a = u , , v , , , , , . Aprs
des calculs bass sur des intgrations par parties,
on peut crire le problme adjoint en introduisant
338
le groupement A =a+bU. Les variables adjointes
du problme sont : A, b, c et d.
0 c b
y x
= c + c (10)
, ) 0 Re
1
= c c c + c + c c

T d V c b bV U b A
x x yy y x x
, ) 0 Re 2
-1
= c c + c + c + c c c T d c c V cU b U U b A
y yy y x y y y
, ) , ) 0
Re Pr
1
2
= + c + c + c c
Ra
d
Pe
dV dU
yy y x
Aux quations (13)-(16) on associe les conditions
aux limites suivantes, issues du systme (3) :
, ) , ) , ) , ) , ) , )
, ) , ) , ) , ) , ) , ) 0 , , 0 , 0 , , , 0
0 1 , 0 , 1 , 0 , 1 , 0 ,
= = = = = = = = = = = =
= = c = = = = = = = = = =
y A x d y x d y x c y A x c y A x b y x b
y x d y x d y x c y x c y x b y x b
y
4. 3. Mthode de resolution
On superpose lcoulement de base une
perturbation, on associe cette perturbation une
fonctionnelle nergie appele gain que lon
cherche maximiser. Pour se faire on utilise une
formulation variationnelle permettant de mettre en
vidence le problme adjoint assorti des conditions
aux limites adquates. Le problme direct de la
convection mixte bidimensionnelle est de nature
parabolique et on le rsout de lentre la sortie
dans le sens de lcoulement, cest dire des x
croissant. Le problme adjoint consiste en un
systme dquations paraboliques dans le sens
rtrograde rsolu de la sortie lentre ; les
variables adjointes sont appeles variables de
sensibilit. La recherche de la perturbation
optimale, se fait par itrations successives en
initialisant le processus par une perturbation u
e
0
(y)
donne arbitrairement. A la nime itration , le
problme directe sera donn par :
0 v = c + c
n
y
n
x
u (11)
, )
n
yy
n
x y
n n
y
n
x
u p U u V Uu c + c = c + c + c
1
Re v
, )
n n
yy
n
y
n
y x
n n
x
Ra
p V V u U u
2
Re Pr
v
Re
1
v v + c + c = c + c + c
n
yy
n
x y
n
y
n n
x
Pe U T T u U u u u c = c + c + c + c
1
v
Avec les conditions aux limites :u
n
(x,0) =v
n
(x,0)
=u
n
(x,0) =, ) 0
1
= c
= y
n
x
u , dans le cas ou le gain
G est dfinit comme suit :
}
}
=
1
0
2
1
0
2
) , (
) , (
dy y x u
dy y x u
G
e
s
(12)
et la condition dentre est donne par
, ) , )
, ) ,
, ) ,
}
}
=
1
0
2
1
2
1
0
2
1
1
2 dy y
dy y
y y
u
u
A u
n
s
n
e
n
e
n
e
(13)
La rsolution du problme direct permet chaque
itration de dterminer , ) y
u
n
s
, ce qui donne la
condition dentre pour le problme adjoint :
, )
, )
, ) ,
}
=
1
0
2
2
dy y
y
y
u
u
A
n
e
n
s
n
s
(14)
Le problme adjoint litration (n) scrit :
0 = c + c
n
y
n
x
c b (15)
, ) 0 Re
1
= c c c + c + c c

T d V c b V b U b A
x
n
x
n n
yy
n
y x
n n
x
, ) 0 Re 2
-1
= c c + c + c + c c c T d c c V U c b U U b A
y
n n
yy
n
y
n
x
n
y y
n n
y
, ) , ) 0
Re Pr
1
2
= + c + c + c
n n
yy
n
y
n
x
c
Ra
d
Pe
V d U d
(16)
Et la condition dentre crite ci dessus. Une fois le
problme adjoint rsolu litration (n) la
connaissance de la fonction A
n
calcule
lentre permet de passer litration
suivante :
, ) , )
, ) ,
, ) ,
}
}

)
=
+
1
0
2
2
1
0
2
1
2 dy y
dy y
y y
u
u
A u
n
s
n
e
n
e
n
e
339
6. Rsultats
-0,1
-0,05
0
0,05
0,1
0,15
0 0,167 0,333 0,5 0,667 0,833 1
Y U
(
Y
)
Srie2
Srie3
Srie4
Srie5
Srie6
iter =1
iter =7
iter =8
iter =10
iter =11
Figure 1. Profil de vitesse optimumobtenu aprs plusieurs itrations
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
600 1000 1500 2000 2500 3000 3500
Srie2
Courbe de stabilit
Re
G
Figure 2: Variation du gain en fonction du nombre de Reynolds
340
Figure 3: Variation du nombre de Nusselt local(a):profil de vitesse non perturbe,(b):profil de vitesse optimum
Rfrences bibliographiques
[1] J . N. N. Quaresma, R. M. Cotta, Exact
solutions for thermally developing tube flow
with variable wall heat flux, International
Communications in Heat and Mass Transfer 41
(1994) 729-742.
[2] Dae-Young Lee, Sang-J in Park, Sung
Tack Ro, Heat transfer by oscillating flow in
circular pipe with a sinusoidal wall
temperature distribution, Int. J . Heat Mass
Transfer 38 (1995) 2529-2537.
[3] P. Stefano, An analytical approach to fully
developed heating of laminar flows in circular
pipes International Communications in Heat
and Mass Transfer 22 (1995) 815-824.
[4] Tatsuo Nishimura, Naoya Kojima, Mass
transfer enhancement in symmetric sinusoidal
wavy-walled channel for pulsatile flow, Int. J .
Heat Mass Transfer. 38 (1995) 1719-1731.
[5] G. Russ, H. Beer, Heat transfer and flow
field in a pipe with sinusoidal wavy surface - I
numerical investigation, Int. J . Heat Mass
Transfer 40 (1997) 1061 1070.
[6] G. Russ, H. Beer, Heat transfer and flow
field in a pipe with sinusoidal wavy surface-II
Experimental investigation, int. J . Heat Mass
Transfer 40 (1997) 1071-1081.
[7] Gad-el-Hak, M., Flow control, Applied
Mechanics Review, Vol 42, 10, (1989), 261-
293.
[8] Gad-el-Hak, M., Bushnell, D. M., Separation
control: Review, J ournal of Fluid Engineering,
Vol 113, 1, (1991), 5-30.
[9] P. Cathalifaud, P. Luchini, Algebraic growth in
boundary: optimal control by blowing and
suction at the wall. Eur. J . Mech. B- Fluids 19
(2000) 469-490.
[10] P. Andersson, M. Berggren, D.
Henningson, Optimal disturbances in boundary
layer,Physics of fluids 11 (1999) 134-150.
[11] Gunzburger M., Inverse design and
optimization methods : Lagrange multiplier
techniques, Von Karman Institute for Fluid
Dynamics, Lecture Series (1997).
TEHNOMUS - New Technologies and Products in Machine processing Technologies
341
STUDY ABOUT VI RTUAL AND ACTUAL MANUFACTURI NG
PROCESS WI TH THE ROBOT
Traian Lucian SEVERIN
1
, Romeo I ONESCU
1
,
1
Management, University Street, No. 13, 720225, Suceava, Romania, severin.traian@fim.usv.ro
romtit@fim.usv.ro
Abstract: In this paper the authors aimto identify the possibility of broadening the use of industrial
robots for the processing of wood. In the robotics laboratory has attempted a stand where the robot will
be able to process the wood pieces with tool attached to the robot. We try to point out the steps for
application development; our target was cutting process of wooden objects. Industrial applications could
be very many and diverse. The robot programming was developed for working on multiple facets of a
piece type cube it used the robot simulation processing. To realize a comparative study of robot
manufacturing with three axes CNC machines was used NX camsoftware which allowed us to simulate
the same processing.
Keywords: robot, milling, processing, manufacturing, simulation,
I NTRODUCTI ON
In this paper the authors aim to identify the
possibility of broadening the use of industrial robots for
the processing of wood. The wood industry can be
divided into four main areas: the furniture industry, the
construction-related sector, the wooden materials
industry, and other sectors. The jobs that robots are
called on to perform in the wood industry include
painting, handling, grading, and repairing of wooden
parts and products. In the last years, the flat-pack
furniture and construction materials industries are
entering a new era of robotized, flexible manufacturing.
The manufacture wood technology with flexible
technology that can quickly and accurately adapt to
varying surfaces and consistency of materials.
The requirements on personnel and machines are
just as great in high-precision machining applications in
the manufacture of large work pieces. Now we can see
how an industrial robot in wood industry offers
increased productivity at a lower cost.
We can offer an example polishing and buffing of
wood products is an interesting application and many
robots are integrated for wood products. In figure 1 we
can see how a major guitar manufacturer had integrated
a work cell with robot to polish and buff guitar bodies.
In that application, the robot picks up the guitar body
and presents it to sanding and buffing wheels [11].
Material handling of wood products is another
important job that robotics accomplish for end-users. In
many workshops where working with wood, in the
secondary operations of wood product processing for
cabinets, windows and doors manufacturing, robotics
are taking pre-cut parts and assembling them or
presenting them to joining machines.
Figure 1. Details of robotics guitar buffing [11].
End-users ask why they should invest in robotics to
help manufacture wood products when people have
been performing these tasks. Most times the answer
could be: tedium suffered by those in the woodworking
industry is one reason that robots are used but is not
only that. The robots can increase productivity, quality
and efficiency of wood processing objects.
In the robotics laboratory has attempted a stand
where the robot will be able to process the wood pieces
with tool attached to the robot. We try to point out the
steps for application development. Our target was
cutting process of wooden objects. Industrial
applications could be very many and diverse.
342
EXPERIMENTAL
The choice of the parts to be processing is very
important because choosing a piece with a quite
complex geometry can lead to the demonstration of
all robots real skills. For the basic study and the
work one can choose first a relatively simple part,
in our case a cube, on whose sides various
processing can be done. This part allows us to
demonstrate the fact that the robot is not limited by
its freedom degrees, but by the clamping device.
This means that if we could clamp the cube on one
sole corner then we could make processing on all
the cubes sides without any other movement of
the device.
Next step consists of the study of the
positioning and clamping device of the part. This
device has to fulfill the following conditions: to be
robust in order to prevent vibrations; to be
universal in order to allow the clamping of a wide
variety of parts; to be positioned in an optimum
position (height, distance) from the robot to allow
easy processing on all surfaces; to allow a simple
positioning, a good screw on with a reduced
number of operations; to ensure the elimination of
the waste resulted in the processing operation.
The choice of the milling device, its
modification and clamping to the robot form the
next stage of the study within the modeling and
simulation of processing with the help of the robot.
The mill must be mounted on the robot so to
prevent any interference or collision with the
robots parts or with the parts clamping device in
the moment of processing and also must be rigidly
clamped to ensure a quality manufacturing.
After the study of all these problems, we can
start the actual simulation of the robot processing
and we can make its program.
For the stand, parts as well as the splinting
device the software CAD/CAM NX was used,
software which offers a wide range of programs in
one integrated solution, to allow the user to use the
lasted technologies in terms of used tools and
processing processes. NX CAM supports the latest
generation of multiple function tools, including
milling, drilling and turning on a 5 axis machine.
NX allows a wide range of flexible processing on 5
axis machines and with tool control on various
axes.
NX CAM is completely related to the other NX
solutions so that the NC programmers cannot
directly access technical design drawings,
assemblies and tools in a processing environment.
Through modeling by manufacturings
associatively, the changes of the model are
automatically conveyed into processing operations.
Programmers and mechanical engineers can work
with modeling parts, can create and mount devices,
can develop paths and can even mould whole
machines through a 3D simulation of the
processing in this work environment.
For the part to be processed we will start from
the simple cube type part on whose sides various
types of transformations can be done. As it can be
seen in figure 2, on each side of the cube the
different forms from figure 2 can be milled, some
simpler, some more complex. For the
tridimensional parts the 3D, NX CAM software
was used.
After a favorable result, more complex shapes
can be tried, with complex material removal, like
conic wholes, uneven surfaces, with variable
radius. It is to be mentioned that now days there
are specialized software which convert the course
of the tool necessary for the surface processing
from the design software into the robots specific
language.
Figure2. Cube type part with variousmanufacturing
343
Figure 3 processingforms to be done on the cubes sides.
In order to be able to compare the robot
processing with classical methods of processing
from today, we will use the same parallelepiped
used for the presentation of the CAM software,
cube which actually exists.
For the processing device we had to take into
consideration the already existing stand from the
robotics lab. Making a new independent stand
which might be fixed on the floor could lead to an
overload of the robots space of work and therefore
to a reduction in the operation space necessary for
the human factor.
Figure 4 Securing device of the semi-finished
attachable to the already existing stand from the
robotics lab. a) real b ) virtually designed in NX
As for the choice and adaptation of the milling
device to perform the task, we started from a
bought hand mill, specially designed to mill
wholes, slots, edges as well as wooden and plastic
joints.
The specifications of this milling device make it
good for the work parameters necessary for
manufacturing: power 1020 Watts, tension /
frequency: 230 V~50 Hz, weight: 3 Kg,
revolutions: 11500-34000 rpm, mill diameter: 8
mm, adjustment: sliding rods, revolutions
adjustment.
Figure 5. Milling machine STERN ER 1020
For this mill its modification was considered in
order to be adapted and attached to the robot. After
several meetings a certain shape and a clamping
system were adopted, even if this meant the
cancellation of certain mill elements.( figure 6).
Figure6 Milling device attachable to the robot Kuka KR125
344
For the simulation of the robot application the
special software Kuka SimPro was used, fact that
allows the off line development and programming
of robot applications on Kuka robots also offering
the possibility to overlap production times with
training times for the new processing tasks, a better
solution examination and the possibility of
improving robotsfunctioning programmes without
being pressed by time. [8].
For the virtual flexible cell the robots specific
parameters were taken into consideration and also
the clamping device position and the orientation of
the semifinished, meaning an external coordination
system according to which the robot will be
programmed. Figure 7
After programming the robot from this virtual
flexible cell it was noticed that the milling device
can perform any type of manufacturing, within the
limit imposed by the robot. In figure 7 one can see
the robots positioning during the processing of
each proposed shape on each of the cubes sides.
During simulation various aspects had been
followed: accuracy of manufacturing, lack of
interference and / or collisions between various
moving parts, avoidance of impossible to reach
positions from the robot, taking notes and
observations in order to modify and resize some
elements from the final assembly in order to
improve the processing process, obtaining a
program which can be directly transferred onto the
robot for real manufacturing.
Figure 7. Virtual flexible cell manufactured with the
KukaSimPro software.
For the simulation of a classic manufacturing,
NX CAM software was used. It is to be
remembered that in the simulation we will use a 3
axis numeric command machine, which means that
each side of the part to be milled will be processed
separately.
Figure8. Simulation of formmilling on cubes sides
Figure9Simulation of classicprocessingon a CNC machine with the help of NX software.
345
In figure 10 one can see that the program is able
to give also the time for the processing of the entire
assembly, or for each manufacturing. It is also to
be considered that the part must be repositioned for
a new manufacturing, fact that introduces new
times for each manufacturing
Figure 10. processingtimes obtained during classic
simulation with NX software.
Below we present the programming way of the
Kuka KR 125 robot. In order to program the
processing of the intended shapes on the cubes
sides we followed the following steps: we used the
NX software in which we traced with a dotted line
the course to be followed by the tool; to obtain the
coordinates of the points which give us the course
described by the robots tool we used as starting
point the point given by the intersection of the
diagonals.
Therefore, an example of points to be obtained
is the following:
Palpated coordinates for the center of the side
are:
x =16; y =69,5; z =39,5 [1]
Manufacturing is done in the positive direction
of the axis X, so the table below will not show this
coordinate.
Table 1.
Coordinates for U shape manufacturing
Coordinates
contrasted to the
center of the side
Coordinates
contrasted to base
origin
Nr.
crt.
y z y z
1 45 30 114.5 69.5
2 45 -2 114.5 37.5
3 35 -12 104.5 27,5
4 25 -2 94,5 37,5
5 25 30 94,5 69,5
Therefore, for the processing of the U shape 5
points are used, plus the points entering and exiting
the material. A linear movement will be used
between points 1 and 2, respectively 4 and 5 and
between 3 and 5 a circular movement will be used
using point 4 as intermediary.
Figure 11. Design of USV processing through points
Figure 12. Extract from U shape processing
program.
As we can see from picture 11 the robot starts
from the position home, in position P1, then goes
traces a straight line and enters the material,
performs its first linear movement, then the
circular movement, followed by the second part of
the U, and at the point P7 it finishes the
processing and comes out of the material.
This is the procedure which was followed for
the processing on all possible sides of this cube
within the robotics lab.
346
a)
b)
c)
Figura 13. processingof the other three sides
The figure below shows some aspects of the
cube processing with the help of the Kuka
KR125 robot.
a)
b)
c)
Fig, 14. processingexamples on cube sides
The times obtained are similar to the times
obtained in the simulation with the CAM software:
- Loading a program: 32 seconds;
- Star program: 2,53 minutes;
347
- Rectangular program: 6, 10 minutes;
- Circle program: 5, 50 minutes;
- USV program: 2,40 minutes.
It is to remember that having only one clamping
device, the times for placing the cube in the work
position is eliminated fact which leads to a better
processing time.
CONCLUSIONS
After the simulation with the specialized
software the following have been noticed:
Table 2. Comparison between the two processing
methods
Characteristic
With KUKA
Kr 125
With 3 axis
CNC
Work speed
Maximum(of
the milling
machine)
Maximum
Number of
clampings
1 4
Cutting depth 3 mm 3 mm
Strength Medium Maximum
Wotking plan Any plan
Works
vertically
Analyzing table 2 we can draw the following
conclusions:
The CNC machine is sturdier than the
robot, although at the level of the processing
done with the robot we are not talking about
sturdiness as if the mill is properly anchored we
will not get vibrations.
In order to completely process the
parallelepiped part the robot will do this in one
clamping, because the degrees of freedom allows
it. For the 3 axis processing device a new
positioning for each side is necessary, fact which
leads to a longer processing time.
According to the type of mill, on the robot
we can adopt a maximum speed of the milling
device or we can mount a broach with very high
milling speeds on the robot while the 3 axis
machine is limited to the maximum speed
imposed to the machine.
The robot, because of the big number of
freedom degrees can work on several plans,
sidelong or on complex spatial courses, while the
3 axis processing device is limited to the vertical
plan, and therefore in processing the part.
If we took into consideration a 6 axis
processing device then the cost price would be
much higher than of a robot which has the same
amount of work, so it would not be profitable.
The processing device has a protected
volume, so that all waste from processing is not
thrown away, and the robot must be added a
vacuum cleaner for the dust and for the
removed material. Otherwise, it can work in
special arranged rooms, where the removed
material does not harm the environment.
References
[1]. Ionescu R., Semenciuc, D., Roboi industriali,
Principii de baz si aplicaii, Editura OID-ICM,
Bucureti, 1996.
[2]. Ionescu, R., Amarandei, R., Roboi Industriali.
Programe de simulare, Universitatea tefan cel
Mare Suceava, 2003.
[3]. SEVERIN Traian Lucian, IONESCU Romeo
Study about the opportunity in robotic drilling,
in volume of The 15th international conference
TEHNOMUS NEW TECHNOLOGIES AND
PRODUCTS IN MACHINE processing
TECHNOLOGIES, Suceava May 8-9 2009,pag.
195-200, ISSN 1224-029x.
[4]. Romeo IONESCU, Traian SEVERIN, Viorel
PETRARIU, Dumitru AMARANDEI Study
about machining with industrial robots help, The
4th International Conference ROBOTICS08,
Brasov, 13-14 noiembrie 2008.
[5]. Ionescu, R., Severin, T., Petrariu, V., Amarandei,
R., Study about CAD/CAM utilization in pieces
spatial processing with robots help. Iasi, 2008
[6]. IONESCU R., The initial training schemes of the
engineers between Rumanian industrial reality and
European teaching, 3rd Balkan Region Conference
on Engineering Education 2005 LBUS, Sibiu,
Romania, 12 - 15 September, 2005
[7]. SEVERIN Traian Lucian, IONESCU Romeo
Aplicaii robotizate dezvoltate n laborator cu
perspectiva de extindere n industrie, n volumul
celei de-a XIV Conferine tiinifice cu participare
internaional Tehnologii i produse noi n
construcia de maini Tehnomus XIV, Suceava 4-
5 mai 2007,pag. 455-460, ISSN 1224-029x.
[8]. Proiect licenta, David Voiculescu, Prelucrarea
unor piese din industria constructoare de maini cu
ajutorul roboilor industriali,FIMMM, Suceava.
[9]. K u k a GmbH, Programarea robotului KUKA
KR125, KUKA Augsburg, 2005
[10]. Derek K., Rapid Traverse Technology and
Trends Spotted By The Editors of Modern
Machine Shohttp://www.mmsonline.com/.
[11]. Brumsson B., , Is the Wood Industry Ready for
Robotics? Robotics online, Posted 11/14/2007,
http://www.robotics.org/content-
detail.cfm/Industrial-Robotics-News/.
348
349
CONSI DERATI ON OF REFLECTI ON COEFFI CI ENT AS A FUNCTI ON OF
FREQUENCY FOR WEDGES OF DI FFERENT MATERI ALS
Liliana Petre-Cainiceanu, Constantin D.Stanescu, Tudor Burlan-Rotar
Politehnica University,. Bucuresti
Abstract: The wedges made of foamed poly-urethane-ether could withstand large deformations
without any damage. They would show a considerable extension before breaking, but a very low stiffness
made additional support of the wedge tips necessary for horizontal mounting.
1.I NTRODUCTI ON
The wedges made of foamed poly-urethane-ether
could withstand large deformations without any
damage. They would show a considerable
extension before breaking, but a very low stiffness
made additional support of the wedge tips
necessary for horizontal mounting.
Figures. 1 and 2 show the reflection coefficient as
a function of frequency for the different materials.
On the basis of these investigations four different
manufacturers were invited to give in tenders for
the supply and mounting of wedges in the two
planned anechoic chambers, as the opinion was
that the price should also be taken into account in
the total considerations.
Figure1 The reflection coefficient as a function of
frequency for wedges of different materials
Figure2. The reflection coefficient as a function of
frequency for wedges of different materials.
After a thorough testing of wedges from the two
lowest quotations, Sillan wedges of specific
density 120 kg/m', manufactured by GrOnzweig
und Hart-mann, Ludwigshafen (Germany) were
chosen as the best alternative.
Figure 3 shows the reflection coefficient as a
function of frequency for the wedges chosen.
The measurements at frequencies above 250
Hz are carried out in a smaller tube than that
shown in Fig. 3. Both whole wedges and parts of
wedges were investigated.
In order to obtain an estimate of the maximum
departure from a free field to be expected in an
anechoic room lined with these wedges, a digital
computer was used to calculate the sound field for
a point source in rooms of different size and with
different values for the reflection coefficients of
the walls. The calculations were made with the
assumption that the sound waves were reflected in
accordance with Snell's law at the outer walls of
350
the room and that phase coincidence of the
reflected sound existed at the measuring point.
Figure3. The reflection coefficient as a function of
frequency for wedges used in the anechoic rooms
Calculations were carried out for different
positions of the sound source. With these
assumptions and a reflection coefficient of 0.1 one
should have a departure from free field of
maximum 1 dB at 2 m and 3 dB at 11 m in a
room of free dimensions of 12 m X 10 m X 9 m.
3.Sound insulation
As the anechoic rooms were also intended for
a number of psycho-acoustical measurements, it
was desirable to have a background noise level of
about 10 dB below the threshold of hearing at
frequencies above 200 Hz, whereas for
frequencies below 200 Hz a level lower than 10
dB re 2 X 10-5 N/m' was required.
For extraordinary outdoor sound levels, such
as those caused by jet aircraft at low altitudes,
slightly higher background levels would be
acceptable.
It would, however, be an advantage if normal
acoustical measurements were not appreciably
disturbed even in these cases.
These requirements necessitate a wall
construction, which has a transmission loss of
about 80 dB at 200 Hz, as a background level of
60-70 dB re 2 X 10-5 N/m2 must be expected
outdoors during working hours. In order to
achieve this, a double wall construction must be
used where the inner part, the anechoic room
itself, has no stiff mechanical connection to the
outer building.
The outer walls of the building consist of
brickwork and reinforced concrete with a layer of
heat insulation between, total thickness 30 cm.
The roof is made of 20 ern reinforced concrete
with heat insulation. The inner boxes, with the
anechoic rooms, have walls, floor and roof made
af 40 ern reinforced concrete. The space between
the inner and outer walls is 1.1 m and acoustically
damped by covering both the inner side of the
outer shell as well as the outer side of the inner
shell with 5 cm wood wool cement slabs. Also the
inside of the outer roof is covered with this
material.
The large and the small room are placed on 24
and 4 rubber vibration isolators respectively, see
Fig. 4 .
Figure 4. Vibration isolators for the large anechoic
room.
Each vibration isolator is loaded by about 50
tons, and a resonance frequency of about 7 Hz
was intended for the rooms placed on the rubber
pads. The vibration isolators, made from a
mixture of natural and artificial rubber (hardness
50 shore) are protected by a layer of neoprene
and placed so that they can be inspected and
replaced if necessary.
4.Doors
The doors are important parts of the construction
with regard to lining and sound insulation. For the
large anechoic room the entrance, which is 2.5 m
X 2.7 m, is closed by two door sections travelling
on rails and made up from three layers of steel (10
+5 +5 m) with concrete in between and covered
with wedges. When the doors are closed the
wedge lining of the room is thus continuous. For
practical reasons there is also a hinged door
covered with mineral wool, which can replace one
of the door sections for less critical
measurements. The opening in the outer wall is
closed by an air tight steel sliding door and the
351
corridor outside the door has no windows and is
heavily damped.
The 2.2 m X 1.6 m entrance to the small anechoic
chamber is closed by two hinged doors made from
steel plate and concrete. One of the doors is lined
with wedges. In front of the other, which is lined
with a 5 cm layer of Sillan, there is a light, wedge
lined, sliding door which can be slided into the
corner of the anechoic room where there are no
wedges, but only a 5 cm Sillan lining. The wedge
lining of the room is thus continuous when the
doors are closed, and at the same time the
construction is simple and saves space.
5.Floors
The floors of the anechoic rooms should not
influence their acoustical properties appreciably.
Wire netting was decided upon, which has been
used for many anechoic rooms with satisfactory
results. The mesh size was chosen to be 50 mm
using 2 mm diameter steel string for the large
room and 3.5 mm for the small room. The smaller
wire thickness in the large room makes it
necessary for people to use large overshoes with
wide rubber sales when walking around in the
room, in order to protect the strings and to
distribute the weight over a larger area.
In the smaller anechoic room, which is used partly
by the students, the influence of a thicker wire has
been accepted, in order to avoid the use of extra
shoes. The wires are fixed to adjustable tension
grips situated on the outside of the inner concrete
box. Each wire is pre-stressed by 200 kg, so that
the deflection with a person (75 kg) situated at the
center of the floor is about 1 cm.
The maximum load for each netting is 1,000 kg,
however it must be distributed in such a way that
the load per meter of the periphery of the load
does not exceed 100 kg.
The wire nettings are electrically isolated from the
building. This is done in order to reduce the risk
of electric shocks from measuring instruments and
to make it possible to fix the potential of the floor
at any desired level independent of that of the
surroundings.
The nettings are placed at the same level as the
floors of the outside rooms and the central planes
of the anechoic rooms, where most of the
measurements will be taken, are situated about 1.5
m above the nettings. Parts of the rooms which
are below the nettings can be reached through
removable gratings in front of the doors.
A fine mesh perion netting is placed below the
steel nettings, in order to catch small items
accidentally dropped.
6.Lighting
As the linings of the rooms are highly insulating
with respect to heat, the lighting must give a
minimum of heat radiation. At the same time it is
necessary to have quite a high light intensity in
order to be able to work with small microphones,
hearing aids etc. As the wall lining is also very
light absorbing and as no reflectors can be
tolerated around the lamps, which must also be
placed as far away as possible from the measuring
area, the lighting installation must have the
highest possible conversion efficiency.
The main illumination comes from 200 W
mercury vapour lamps, 12 in the large room and 2
in the small room. As these lamps ignite slowly
and are probably not completely noiseless, there
are also a set of ordinary 200 W incandescent
lamps which may be used separately, 8 in the
large room and 2 in the small room. It is a
relatively simple matter to remove the lamps if
this should be necessary for acoustical reasons.
7.Ventilation
The ventilation system is dimensioned with a
view to obtaining a suitable rate of change of air
in the rooms, and at the Same time to avoid any
appreciable increase of background noise level
with the system operating. In both rooms the
ventilation system consists of a row of air inlets
along the bottom of the walls of the room and
outlets along the top of the opposite wall. The
fans are placed in the basement outside the
anechoic rooms, and on its way into the rooms the
air passes through two long, heavily damped
concrete ducts, which are connected by flexible
tubes. The exhaust air is similarly
taken through two concrete ducts connected by
flexible tubes to the fans with outlets into open
air.
For the large anechoic room the ducts are 16 m
and 10 m long with internal cross-section 70 cm X
80 cm. The thickness of the concrete wall is 10
cm. The acoustical attenuation is obtained by
lining one of the inner walls with 50 cm mineral
wool, which is divided into sections by thin metal
plates to avoid sound transmission through the
material along the ducts.
The ducts for the small anechoic room are 7 m
and 5 m long, and their concrete walls are 10 cm
352
thick. The internal dimensions are 45 cm X 60 cm
and the sound absorbing lining consists of a 20 cm
layer of mineral wool along one wall, divided into
sections as that for the large room.
The inlet and outlet nozzles in the anechoic rooms
are made of sheet metal and to avoid sound
radiation from possible vibrations in the nozzles,
these are covered externally with a 5 cm layer of
mineral wool. The inlets and outlets in the large
room have the dimensions 6 cm X 100 cm and in
the small room 6 cm X 40 cm.
Also the ducts leading to the open air inlets and
outlets are heavily damped so that noise from the
fans will not disturb measurements in rooms
situated on the same side of the buildings as these
openings.
The air can be changed 4 to 5 times an hour in the
large room and 12 to 17 times per hour in the
small room. The air speed at the inlets and outlets
into the rooms is about
2.9 m/sec. Starting and stopping of the ventilation
machinery and control of the air temperature is
conducted by control knobs outside the doors for
each room. It is not, however, possible to obtain a
lower air temperature than that of the outside air
as there is no refrigeration system. A refrigeration
system was discussed during the planning stages,
but due to the particular insulating properties of
the wedges, which made calculations difficult it
was decided to leave space for a refrigeration unit
but to delay the procurement and installation until
the room had been used for some time.
Experience from 1966-67 does not indicate any
need for refrigeration of the air.
8. Conclusion
The measurements carried out show that both
anechoic rooms can be used even for very
exacting measurements, the data for the large
room probably representing something near the
limit of what one can achieve today.
From the background noise measurements it is
seen that it has been possible to attenuate the
noise from the ventilation system to such a degree
that the ventilators may be used even during
psycho-acoustical tests.
The insulation against noise in the vicinity of the
buildings is extremely good, in as much as even
with sound pressure levels of the order of 110 dB
re 2 X 10-5 N/m' the sound pressure inside the
room just reaches the threshold of hearing. It is
therefore possible to carry out a large number of
normal acoustical measurements in these rooms
even under extreme outside conditions.
Bibliography
[1]. Rivin, A. N.: An Anechoic Chamber for
Acoustical Measurements.
[2]. Soviet Physics-Acoustics, 7, (2002), No.3.
[3]. Wiuff, A E.: Eksamensprojekt 2005 (Master
Thesis), DTH, Copenhagen,
[4]. 2006 (in Danish).
[5]. Nissen, Preben: Bending Moments in
Members of Tension and Stressed Wires
Subject
[6]. to Concentrated Transversal Loads.
Bygningstekniske Med- delelser, 38 (1997), nr.
1 (in Danish with an English Summary).

Cuprins Tehnomus XVI

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cuprins Tehnomus XVI

Uploaded by

Copyright:

Available Formats

TEFAN CEL MARE UNI VERSI TY of SUCEAVA

FACULTY of MECHANI CAL ENGI NEERI NG, MECHATRONI CS AND

<1, so that the series is convergence.

<1 so that the series to be convergence.

a relation similar to the one form which we

TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

C with an accuracy of 0.2

C. Immediately after heat treatment, samples were

C, keeping time was 40 minutes. As

C/40 minutes Mechanical agitation - -

C/40 minutes Mechanical agitation 150

C/40 minutes Continuous magneticfield300 G - -

C/40 minutes Continuous magneticfield300 G 150

C/40 minutes Ultrasonicfield40,4 kHz and 4W/dm

C/40 minutes Ultrasonicfield40,4 kHz and 4W/dm

is kinematic degree of the monocontour k

C for one hour followed

C for one hour,

C), but times were held at that temperature

C and cooling to -30

ADB arc) can be

. Lquation (1) est appele

You might also like