Acta No54volIV Cuprins Si Rezumate

ACTA TECHNICA NAPOCENSIS
Vol. 54, Issue IV Series: APPLIED MATHEMATICS AND MECHANICS 2011
564
*** COLEGIUL REDACIONAL Redactor ef Prof. Dr. Ing. MARIANA ARGHIR Secretar de Redacie Prof. dr. Ing. Ovidiu TTARU *** Membrii: Prof. Dr. Ing. GEORGE ARGHIR- membru al Academiei Universale din Lausanne, membru asociat al Diviziei de Istoria tiinei Academia Romn Prof. Dr. Ing. PETRU BERCE, Prof. Dr. Ing. NICOLAE BURNETE, Prof. Dr. Ing. Polidor BRATU membru al Academiei de tiine Tehnice, Prof. Dr. Ing. OCTAVIAN BOLOGA membru corespondent al Academiei de tiine Tehnice, Prof. Dr. Ing. Emil COMA, Prof. Dr. Ing. MICHEL CONT, Prof. Dr. Ing. HORIA COLAN membru corespondent al Academiei Romne, Prof. Dr. Ing. Igor DRSTVENEK, Prof. Dr. Ing. VIOREL ISPAS, Prof. Dr. Ing. KRZYSZTOF LIPISKI, Prof. Dr. Ing. RADU MUNTEANU membru al Academiei de tiine Tehnice, Prof. Dr. Ing. IULIU NEGREAN, Prof. Dr. Ing. SORIN MACOVESCU, Prof. Dr. Ing. ILEANA ROCA, Prof. Dr. Ing. IOAN PSTRAV, Prof. Dr. Ing. IOAN VIDA-SIMITI, Prof. Dr. Ing. SORIN VLASE Redactor: MATEMATIC APLICAT I MECANIC: Prof. Dr. Ing. MARIANA ARGHIR Tehnoredactare pe calculator: Prof. Dr. Ing. MARIANA ARGHIR Pagina web: Ing. MARIUS DENE-POP
ISSN 1221 5872
565

OF THE TECHNICAL UNIVERSITY OF CLUJ - NAPOCA 2011 Vol. 54, Issue IV. APPLIED MATHEMATICS AND MECHANICS CONTENTS Mariana ARGHIR, Marius BAIDOC, Florin EPE-BOBESCU, Alina Sabina PACA, Clin Vasile LUMEI, Maintenance Strategies in the Correlation Man-Machine-Environment 569 Mariana ARGHIR, Marius BAIDOC, Clin Vasile LUMEI, Alina Sabina PACA, Florin EPE-BOBESCU, Systems for On-Line Monitoring of Vibrations ..575 Iulian LUPEA, Roxana TIBREA, Acoustic Normal Modes Simulation of a Classroom..581 Iulian LUPEA, Marius MORARIU, Modal Analysis of an Inflated Tire by Simulation..587 Iuliu NEGREAN, Claudiu SCHONSTEIN, Zoltan SZOKE, Contributions to the Dynamics of Mobile Robot PATROLBOT..593 Mihaela SUCIU, Bending Calculus for Circular Tapping Plates with Circumferential Exterior Concentrated Load by Transfer-Matrix Method 599 Gheorghe GLIGOR, Sorin GROZAV, Radu MORARIU-GLIGOR, Influence of the Geometric Form of the Plate upon the Direct Extrusion Force for the Non-Ferrous Metals...605 Liana HANCU, Sorin COMA, Horaiu IANCU, Rare GDLEAN, Rzvan PRUN, Mathematical Simulation of Metal Sheet Bending with Elastomeric Pads .609 Magdalena ORBAN, Cecilia CIORTEA, Diana BOCA, Using the Sphere and the Revolving Cone as Geometric Loci for Graphical Constructions.617 Radu Mircea MORARIU-GLIGOR, The Optimal Design of Rubber Vibration Isolators of a Plate Compactor621 Radu Mircea MORARIU-GLIGOR, Gheorghe GLIGOR, A Method to Determine the Value of the Ground Pressure Developed by One Way Plate Compactors..625
566
Ancua MIRCEA, Kalman KACSO, Iuliu NEGREAN, Ioan VUSCAN, Radu MORARIU-GLIGOR, The Geometric Equations of Two Machine Tool Type 3T3R in Cooperation...631 Marius BAIDOC, Clin Vasile LUMEI, Florin EPE-BOBESCU, Alina Sabina PACA, Mariana ARGHIR, Experimental Simulation for Implementation of Predictive Maintenance........................................................................................................639 Marius BAIDOC, Alina Sabina PACA, Florin EPE-BOBESCU, Clin Vasile LUMEI, Mariana ARGHIR, Methods of the Equipments Alignment645 Lia CHIOREAN, Entry Prediction in the Technical University f Cluj-Napoca .653 Daniela Carmen KOUKACH, Grigore Marian POP, Dan PREJA, Machining Process Modeling..............659 Daniela Carmen KOUKACH, Grigore Marian POP, Dan PREJA, Modeling Manufacturing Processes of Cutting.....................................................................................669 Mihaela SIMION, Adrian-Ioan BOTEAN, Mircea BEJAN, Numerical Analysis of State of Stresses and Deformations of an Industrial Serial Robot 5R...675
567
A
UNIVERSITATII TEHNICE DIN CLUJ - NAPOCA 2011 Vol. 54, Nr. IV. MATEMATIC APLICAT I MECANIC CUPRINS Mariana ARGHIR, Marius BAIDOC, Florin EPE-BOBESCU, Alina Sabina PACA, Clin Vasile LUMEI, Strategii de mentenana n corelaia om mediu - main ..569 Mariana ARGHIR, Marius BAIDOC, Clin Vasile LUMEI, Alina Sabina PACA, Florin EPE-BOBESCU, Sisteme de monitorizare on-line ale vibraiilor575 Iulian LUPEA, Roxana TIBREA, Analiza modal acustic a unei sli de cursuri folosind elemente finite..581 Iulian LUPEA, Marius MORARIU, Analiz modal aplicat unei anvelope folosind metoda elementelor finite587 Iuliu NEGREAN, Claudiu SCHONSTEIN, Zoltan SZOKE, Contribuii cu privire la dinamica robotului mobil de tip PatrolBot ..593 Mihaela SUCIU, Calculul la incovoiere al placilor circulare gaurite cu forta circumferentiala exterioara concentrata prin Metoda Matricelor de Transfer .599 Gheorghe GLIGOR, Sorin GROZAV, Radu MORARIU-GLIGOR, Influenta formei geometrice a placii active asupra fortei la extrudarea directa a metalelor neferoase ..605 Liana HANCU, Sorin COMA, Horaiu IANCU, Rare GDLEAN, Rzvan PRUN, Simularea matematica a indoirii tablelor metalice cu placi active din elastomeric..609 Magdalena ORBAN, Cecilia CIORTEA, Diana BOCA, Utilizarea sferei i conului de rotaie ca suprafee auxiliare n rezolvri grafice 617 Radu Mircea MORARIU-GLIGOR, Proiectarea optimal a elementelor antivibratile din cauciuc din componena unei plci compactoare.621 Radu Mircea MORARIU-GLIGOR, Gheorghe GLIGOR, Determinarea valorii forei de apsare normale dezvoltat de plcile compactoare unidirecionale ..625
568
Ancua MIRCEA, Kalman KACSO, Iuliu NEGREAN, Ioan VUSCAN, Radu MORARIU-GLIGOR, Ecuaiile geometrice a dou maini unelte de tip 3T3R n cooperare .......631 Ancua MIRCEA, Kalman KACSO, Ioan VUSCAN, Iuliu NEGREAN, Radu MORARIU-GLIGOR, Determinarea erorilor de compensare prin utilizarea simulrilor numerice...............................................................................................................................635 Marius BAIDOC, Clin Vasile LUMEI, Florin EPE-BOBESCU, Alina Sabina PACA, Mariana ARGHIR, Simulare experimental pentru implementarea mentenanei Predictive..............................................................................................................................639 Marius BAIDOC, Alina Sabina PACA, Florin EPE-BOBESCU, Clin Vasile LUMEI, Mariana ARGHIR, Metode de aliniere a utilajelor .645 Lia CHIOREAN, Prognoza admiterii n Universitatea Tehnic din Cluj-Napoca...653 Daniela Carmen KOUKACH, Grigore Marian POP, Dan PREJA, Modelarea proceselor de prelucrare ........................................................................................................................659 Daniela Carmen KOUKACH, Grigore Marian POP, Dan PREJA, Modelarea proceselor de prelucrare prin achiere....................................................................................................669 Mihaela SIMION, Adrian-Ioan BOTEAN, Mircea BEJAN, Analiza numeric a strii de tensiuni i deformaii ale unui robot industrial serial 5R.675
569
TECHNICAL UNIVERSITY OF CLUJ-NAPOCA

Series: Applied Mathematics and Mechanics Vol. 54, Issue IV, 2011
MAINTENANCE STRATEGIES IN THE CORRELATION MAN-MACHINE-ENVIRONMENT

Mariana ARGHIR, Marius BAIDOC, Florin EPE-BOBESCU, Alina Sabina PACA, Clin Vasile LUMEI
Abstract: This paper presents the types of maintenances, their applicability, their advantages and disadvantages, and how the rotating machinery can be diagnostics. The fundamental concept of maintenance is getting a high degree of security in the formulation of complex man machine environment. Key words: maintenance strategies, man machine environment.
1. INTRODUCTION Development of the society of consumption are required development has resulted in the development of productions as large, highquality and low cost as possible. In the price of any product, regardless of industry or type of industrial production, maintenance costs are a significant proportion of the price of the products worked. Therefore, a major concern for the management of companies and experts in machinery and equipment is the development of organizational measures and technologies, to reduce as far as the stops accidental of machinery, and the duration of repairs, with the consequence of reducing the cost of maintenance. Also, there are facilities in industrial drives high complexity, technology that works with AIDS to pick up parameters (temperatures, pressures) and dangerous fluids (toxic, corrosive, flammable, explosive lethal, etc). These parameters can diminish the capacity of load-bearing elements and structures of their resistance and can damage the joints and test pieces etc. In this way, providing a total of operation no fuses is possible and, a priori destined, was accepted the notion of risk. To get to the determination of an acceptable risk limits, it was, first, identification of risk and the follow-
up of this risk assessment is to identify and develop the organizational and technical measures necessary to eliminate potential risk is unacceptable. Using such data and of the need to reduce production costs, in time, they carry out studies to determine the technical measures of reliability to ensure safe operation of the plants, leading to the evolution of the different systems and maintenance concepts [Bay 99]. The literature categorizes programs by running the following: 1.1. Corrective Maintenance 1.2. Preventive Maintenance 1.3. Predictive Maintenance 1.4. Proactive Maintenance The concept of maintenance includes all technical and organizational actions on testing, measurement and verification, in the sense of the prophylactic and remedial modules of parts of an installation; in order to maintain or reduce its running status and the operation are at levels of safety and security as performing. Fundamental purpose of comprehensive is to achieve high levels of reliability and mentenability, so by default, an improvement of their cumulativeavailability, and getting a high degree of security system in the formulation of complex man-machine-environment.
570
2. CORRECTIVE MAINTENANCE Corrective maintenance - CORMENT-is the concept that equipment works by stopping the accidentally abrasion of recently established or because of defections. Repair means, as a rule, the replacement of damaged or even the SubAssembly of the whole machine. This situation is most unfortunate and implies, obviously, high maintenance costs. Thus, corrective maintenance includes all work on the equipment in order to remove inappropriate conditions of operation, in terms of operational safety and/or protection of aggregate (to avoid the appearance of defects) or winding faults already incurred. Using this method, machines are left up to the appearance of a major flaw, or until a decline in inadmissible efficiency. Compared to other types of maintenance, corrective maintenance is holding a series of major disadvantages such as: the possibility of damage that can lead to loss of human lives and material goods; the impossibility of planning activity, spare parts, tools and appliances needed repairing machinery; growths the cost of repair, due to the impossibility of assessing the optimal stopping time equipment, etc. 3. PREVENTIVE MAINTENANCE Preventive maintenance - PREMENT represents the concept in which the machinery is stopped programmatically, based on the number of hours of operation for revisions raised, the current were technical and/or capital repairs. In this system, what matters is the number of accumulated hours of operation and the extent of wear is not installed. Preventive maintenance involves performing cyclic-interventions made at the time intervals, specified-including works of maintenance and repairs undertaken, mainly in order to eliminate the effects of wear and tear. The actions taken in respect of prevention, as well as prevent the occurrence of fault conditions, by reducing the risk to occur also States. Thus, while corrective
maintenance operations have a preventive character randomly, those are deterministic. A first important issue that arises when preventive maintenance method adopted is to establish the optimal interval between two revisions planned. If the range you chose is too high, the percentage of fall incidental will increase considerably. On the other hand, shortening the interval between the revisions require extra charges from becoming larger. When you restart to complete an installation, after overhaul, the probability of dropping each individual machine in hand is multiplied by the total number of machines. It is obvious that, in these circumstances, any restart is performed in conditions of considerable risk [Ste 99]. This concept has a number of disadvantages that cannot be ignored, namely: high costs of maintenance and repair activity, thanks to the intervention machinery which does not require such workmanship; the duration of a planned technical revisions due to high volume controlled machines; there is a possibility, as confirmed by practice, the reboot, the plant, which have worked well in advance of the shutdown, to have a malfunctioning as a result of the not good interventions during the revision. 4. PREDICTIVE MAINTENANCE Predictive maintenance is the concept that is non-prophylactic and preventive means that the equipment operates safely until a certain level of wear or a defect. In this system, through its regular monitoring activities and, where appropriate, remedial effects of the various factors are highlighted destructive, which could affect to a large extent technical operational capability of the system and would lead to the initiation of corrective or preventive corrective interventions much more costly. Predictive maintenance highlights the frequency of occurrence of faults and their location on the type of equipment, being a fundamental element to determine whether to carry out complex works of modernization and rehabilitation of industrial installations [Lee 04].
571
The main parameters in the report of the technical condition evaluation are carried out by the operation of technological equipment, as levels of vibrations and noise [Lal 90]. At the time at which it is found that the level of shock resistance came close to the upper limit of the admissibility team should be alert
for intervention by the maintenance at any time, on the equipment [Don 02]. In a situation where it is found that the level of vibration has exceeded the maximum allowable value is necessary to terminate the technological equipment in order to determine the causes of vibrations and then repair malfunctions (Fig. 1).
Fig. 1. Diagnosis of rotating machinery dynamic
Thus, the frequency of the vibration analysis (FFT analysis) is that which makes it possible to obtain information necessary for the time of the industrial equipment malfunctions. Predictive maintenance has numerous advantages, the most important being: reduce costs associated with the stop time of technological equipment and increase profits by increasing the term of production; reduction or elimination of costs of technical incidents or serious damage to the machinery; reduce the costs of maintenance and spare parts stock, many tracks may be ordered even while carrying-out repairs; reduction and/or disposal costs caused by unplanned maintenance, repairs may be carried out with minimum losses; Optimizing the performance of the machines, which often operate without technical specifications; reducing the excessive consumption of electricity; reducing equipment needs and costs related to the phase of "stand-by"; increase in production capacity and reducing capital investment, tools can be used a long time; reducing the risk of unsuccessful repairs, as well as the number of customers dissatisfied with delays in fulfilling orders
or customers dissatisfied with the quality of products supplied; optimization of activity through the use of staff in other activities linked to production; the reduction and/or elimination of scrap or waste resulting from non performances of machinery; remove the additional time and cost of restoration of appearance and the conditions necessary for restarting and optimal functioning; reducing the possibility of sale of defective equipment; increasing safety in operation of the machinery; reducing the risk of penalties for the use of insecure machines; improving the performance of the technological and operational machinery. Predictive maintenance applies perfectly to 85-90% of existing machinery in the plant, machinery may be stopped, without this actually lead to shut down the plant. Statistical analysis of the causes of defects reveals that machinery dynamic 85% of the defects are predictive, 8% is due to the mistakes of operation and 7% is due to random causes (material defects, external causes, etc.). As a result of the research has shown that, as shown in Figure 2, the life curve of a machine is a real rather than theoretical
572
transposition of certain facts relating to the

Mortality
equipment in question.
Period of Period of wear
stable functioning
Probability of
Time (logarithmic scale)
Fig. 2. The life cycle of a machine
Mortality
Period of stable functioning
Period of wear
Annual review
Probability of
Time (logarithmic scale)
Fig. 3. The life cycle of a reconditioned machine
The life cycle of a machine shown in Figure 3 it is noted that immediately after rebooting the machine, there is a high probability that it would be damaged, in particular, because small imperfections in the
business of repairing itself, but also due to the components or sub-assemblies. Many factories respecting theorems reliability can make savings. Thus: If a machine works well, does not intervene on him! or
look but don't touch! We therefore need a predictive maintenance program, which, in addition to the fact that merges the two theorems, detect power failures inherent and substitution of damaged components. In addition, such a programmer reduces the magnitude of many shortcomings by reducing or preventing secondary damage. For example, an engine with a revolving defect can be stopped before damage to the impeller or the shaft. From the economic point of view, it was found that companies that have implemented these predictive maintenance programs have registered an increase of profits by reducing the cost of repair. Predictive maintenance depends on the status of the momentary state of equipment, permanently monitored. In the literature, this monitoring is known as "condition monitoring". In relation to the above indicated, we appreciate that predictive maintenance is a qualitative leap, regardless of the industrial sector in which it is implemented, provides all the information necessary for the detection of early failures, locating and diagnosing them and not least in order to calculate the duration of the safe functioning of equipment. To illustrate the process of monitoring and prediction in his character to detect malfunctions we present processes of oil and vibration monitoring equipment. 5. PROACTIVE MAINTENANCE The concept of proactive maintenance adopted recently by the companies advanced in terms of maintaining dynamic machinery consists of carefully monitoring machinery that works well (the majority), the manager having to deal with them and the tools with problems with specialized staff. Obviously, that proactive maintenance can be applied only in terms of predictive
573
maintenance, maintenance of proactive as a continuation of it. 6. CONCLUSION Followed the theory of this paper, the conclusions are: 1. All maintenance concepts contribute to optimizing the activity recognized companies, regardless of the field of industrial production, but at the moment, predictive and proactive maintenance provides the best conditions of operation, with the lowest cost while at the same time, increased safety in the operation of technological equipment, environmental protection and deserving staff. 2. All this is possible thanks to the comprehensive research in all areas, recognition internationally of the importance of comprehensive industry, so that both the production and all the processes and systems of enterprises, are heavily dependent on trade. 3. In order to streamline maintenance, for which the costs will be reflected in the cost price of the final product, it is necessary to use all the strategies and methods of management, and the first step in this optimization is to use a Computerized Maintenance Management Software followed by the use of Condition Based Maintenance and it only the tools that may be warranted. 4. Another step towards optimization of the industrial activity is thus placing predictive maintenance methods with which you can evaluate appropriate time repairing the (economically or production) and set a time to the possibility of a breakdown that will stop the machine. 5. On the market there are various equipment and software that can be used to streamline maintenance.
574
6. In view of the current state, I consider necessary to conduct additional research (to decrease cost price and increasing the efficiency of comprehensive) in the field of condition monitoring equipment and the methods of prediction based on the status of equipment. 7. REFERENCES
[Bay 99] Bayer, M., Fiabilitatea i mentenabilitatea sistemelor, Editura Bren, Bucureti, 1999. [Don 02] Donald, E. Bently, Charles T. Hatch, Bob Grissom, 2002, Fundamentals of Rotating Machinery Diagnostics, Bently Pressurized Bearing Company. [Lal 90] Lalanne, M, Ferraris, G., Rotordynamics Prediction in Engineering, John Wiley and Sons, New York, 1990. [Lee 04] Lee, J., .a., An integrated platform for diagnostics, prognostics and maintenance optimization, IMS International Conference, Arles, France, 2004. [Ste 99] Steve Goldman, Vibration Spectrum Analysis, Terre Haute, 1999.
STRATEGII DE MENTENANA N CORELAIA OM MEDIU - MAIN Rezumat: Lucarea prezint tipurile de menten, aplicabilitatea acestora avantajele i dezavantajele lor, precum i modul de diagnosticare a mainilor rotative. Conceptul fundamental al mentenanei este de a gsi corelaia ntre om mediu main.
Mariana ARGHIR, Prof. Dr., Mech. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.264.401.657. Marius BAIDOC, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.770.400; Florin EPE-BOBESCU, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.547.258; Alina Sabina PACA, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.756.450.183; Clin Vasile LUMEI, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.742.030.565.
575

SYSTEMS FOR ON-LINE MONITORING OF VIBRATIONS

Mariana ARGHIR, Marius BAIDOC, Clin Vasile LUMEI, Alina Sabina PACA, Florin EPE-BOBESCU
Abstract: The paper presents systems on-line monitoring of vibrations, with the connection parameters for accelerometers: industrial measurements, to align the trees and rotors, for noise measurement, for proximity systems. Key words: on-line monitoring of vibrations, proximity systems.
1. INTRODUCTION Systems for on-line monitoring of vibrations and other industrial parameters are intended for continuous monitoring of all vibrations from industrial equipment and machinery and for the study and analysis of vibration on the stands. Also, the system can be used for on-line monitoring of industrial, other parameters such as temperature, pressure, flow, voltages, etc [Bay 99]. Systems can be used both as a stationary, laboratory apparatus and portable, for measurements in the field. Measurements of vibrations that can be performed are: absolute vibration measurement, measurement of vibration bearings relative to shafts, and they are structured according to the type of machine. These types of measurements can be performed on the following practical applications: vibration control and analysis pertaining to the shafts to steam turbines, hydro and gas from the other machines and compressors, who possess the sliding bearings; absolute vibration monitoring machine which posses camps rollover-bearings, such as electric motors, fans, traps, reducers, etc. The system enables measurements above and on a bench study of vibration, which is composed of both bearings and slide bearings.
Systems for on-line monitoring of vibrations fall into two categories: systems for on-line monitoring of FFT spectral analysis without vibrations; systems for on-line monitoring of vibration with FFT spectral analysis. 1.1. Monitoring systems on-line of vibrations without FFT spectral analysis Monitoring systems on-line of vibrations without FFT spectral analysis are intended to monitor vibrations from a diversified range of dynamic rotating machinery, such as: electric motors has, reducers, pumps, fans, compressors etc. [Don 02] 1.2. Monitoring systems on-line of vibrations with FFT spectral analysis Systems for on-line monitoring of vibration with FFT spectral analysis for continuous monitoring of vibration in a diversified range of dynamic rotating machinery has to be: steam turbines, water turbines, extruded, centrifuges and other machinery, which shall have priority in the production process. The system provides comprehensive solutions in a technique to measure the vibrations and process parameters, like this: the possibility for the diagnosis of vibration fault for setting the machine in its infancy,
576
store data measured in both module hardware acquisition, and on your PC, can view of the process, synchronous acquisition and the simultaneous signals, save directly to your device, process parameters, monitoring values and generating reports. 1.3. Main types of measurements that can be performed with the on-line system
errors of alignment; the weakening system *; eccentricity; mechanical games; the phenomenon of resonance; cavitation; defects of sliding of the camps; defects of the roll-up bearings of the camps; crankshaft; rotor friction; problems of electric motors; defects of gearing etc. 2. CONNECT THE SENSORS TO MONITORING AND ANALYSIS SYSTEM
The on-line systems have the capability to perform the following types of vibration measurements: Magnitudes global vibration time-Trend. Waveform - Oscilogrames vibration. Magnitudes vibration depending on the frequency Spectrograms frequency including type revolving. Types of Windows: Hanning, Flat Top etc", minimum 1600 lines per channel. Magnitudes vibration depending on frequency and time-frequency Spectrograms cascade 3D. XY-the Marshalling all vibrations at the same time, used in the measurement of force and displacements. Orbit diagram of two-composition of waveform at the same time with the vibrations, which is used to measure relative to proximate. Polar diagram - magnitudes vibration depending on the global stage. Possibility of simultaneous visualization of measurement data. The ability to configure different applications simultaneously and view technical drawings, images, graphs, spectrograms with the parameters of vibration or techniques. 1.4. Main defects that can be explored with on-line system The main types of damage that can be analyzed and studied in connection with the online system are: static and dynamic imbalance;
The system is able to perform measurements of vibration with the following types of sensors [Bay99]: compatible ICP and capacitive accelerometers for measuring the absolute, vibrating proximity systems for measuring relative vibrating vibrating velocity sensors, laser sensors for alignment of trees, temperature sensors non-contact, termorezistance, class 1 microphones ICP or class 2. For the monitoring of other parameters, the system supports industrial sensors that have standard outputs: 0 and 4-20 mA, 0-10 V, 10 V For thermocouple and thermorezistance. It is appropriate that they should be able to connect directly to the device. 2.1 Connect compatible ICP and capacitive accelerometers (Fig. 1) In the diagram in Figure 1 is given a single measuring channel. Capacitive accelerometers for block diagram are similar, with the difference that, instead of ICP accelerometer, you can use a capacitive accelerometer and module ICP will be replaced with a capacitive signal processing. The system supports the full range of accelerometers, low, medium and high frequency [Lal 90].
577
2.2. Connecting systems of proximity without contact displacement transducers Measurement of vibrating relative to trees will be achieved with two transducers proximity arranged at 90 to one another (Fig. 2). These types of metrics are used to monitor the machines, who possess the sliding bearings, e.g. turbines, by composing oscillogrames (relative vibration displacement) on two directions X-Y, at the same time, obtaining the
Accelerometer cable
orbit of the time in the camp, as shown in Figure 3 Block diagram of vibration monitoring system with proximity systems is shown in Figure 4. The connection speed of vibration sensors are usually directly in the measuring instrument. Sensors will be with output 4-20 mA and will measure the velocity of vibrating effective rms. Block diagram of the system of monitoring of vibration speed sensor is shown in Figure 5.
Coaxial cable
Analog input Vibration monitoring
Electric motor
Machine
Vibration sensor
ICP module Measuring device
Fig. 1. Monitoring of vibration system with compatible ICP accelerometers
Displacement transducer Arbor Displacement
Fig. 2. Monitoring of vibrations system with proximity systems

Displacement Displacement
Displacement
Fig. 3. The spindle orbit in bearing
578
Sensor
Transducer cable
Proximitor Coaxial cable
Measuring device
Slide bearing
Sensor
-24V Input
Analog inputs
Vibration monitoring
Fig. 4. Monitoring system of vibrations with proximity systems
Sensor
Input
Measuring device
Vibration monitoring
Signal
Fig. 5. Monitoring system of vibrations with speed sensors
2.3. Connecting to industrial sensors for measuring the parameters To connect standard sensors with outputs: 0-4-20 mA, 0-10 V, 10 V is identical to the schema shown in Figure 4.18. To measure temperatures (e.g. bearings, etc.) with thermocouples and thermorezistances is shown that the device can be connected to a greater range of thermocouple and thermorezistances (all types, if possible) directly to the device. Block diagram of the system for monitoring temperatures with thermocouple and thermorezistances is shown in Figure 6. Block diagram for connecting sensors, laser or optical required for measuring revolution and phase, is identical. 2.4. Connecting the sensors for measuring noise Block diagram of the noise monitoring system with microphones compatible ICP is shown in Figure 7.
2.5. Connecting laser sensors to monitoring system on-line for alignment of crankshafts and slide bearings For horizontal and vertical alignment of shafts using laser sensors and onedimensional, for the alignment of the bearings, using laser sensors [Roe 05] twodimensional. Connection scheme of laser sensors to monitoring system on-line is shown in Figure 8. To align the two coordinates (twodimensional), the scheme is similar, with the difference that you will use a single mode laser and a two-dimensional detector. This system has the possibility of rotating machinery dynamic alignment, who works at the cold as well as on-line verification of alignment on dynamic thermal expansions from occurring. The laser system may check the deviation from perpendicularity, parallelism, evenness, horizontality, verticality, straightness etc. of the various pieces and parts from the machines.
579
Balance rope Analog input
Monitoring parameters
Thermocouple or Thermorezistances
Measuring device
Fig. 6. System for monitoring temperatures with thermocouples and thermorezistances

Microphone cable Coaxial cable Analog input Noise monitoring
Microphone
Preamplifier
ICP module Measuring instrument
Fig. 7. Noise monitoring system with microphones compatible ICP

Laser sensor 1 Laser sensor 2 Detector 1 SPC - PSD Cable
Analog inputs
Software for alignment sensor system
Laser module 1
Detector 2 SPC - PSD
Laser beam Laser module 2
Measuring device
Fig. 8. System monitoring with laser sensors one-dimensional
3. CONCLUSION Efficient implementation of predictive maintenance programmed assumes a dynamic monitoring of equipment, both during operation and during periods of repair, but only after they were brought, initially at running fit for prediction, i.e., they must operate within the limits of permissible vibration accepted standards.
Predictive and proactive comprehensive implementation leads to increased safety in the operation of the equipment of dynamic and significant reduction of the accidental shutdowns which optimize the activity of production and reduce the cost of production. To this end, the current trend is introducing effective systems for monitoring of the parameters that affect the operation of the equipment, having an important role in the
580
pursuit of reliability and mentenability dynamic machinery. Therefore, increasing quality and efficiency of repairs and safety in the operation of the equipment may be carried out by the industrial and commercial companies of maintenance policies, particularly in the field of vibrations and noise, thermograph, lubrication, balance and alignment equipment, by equipping with modern equipment and systems, which, in a relatively short time will locate in a lower cost of production. Has an important role in preparing and serving personnel, equipment, requiring little in that sense, this sentence is meant to serve as a practical guide for the implementation of the comprehensive predictive and proactive.
4. REFERENCES
[Bay 99] Bayer, M., Fiabilitatea i mentenabilitatea sistemelor, Editura Bren, Bucureti, 1999. [Don 02] Donald, E. Bently, Charles T. Hatch, Bob Grissom, 2002, Fundamentals of Rotating Machinery Diagnostics, Bently Pressurized Bearing Company. [Lal 90] Lalanne, M, Ferraris, G., Rotordynamics Prediction in Engineering, John Wiley and Sons, New York, 1990. [Roe 05] Roemer, Michael, J., Kacprzynski, Gregory, J., .a. - Advanced Vibration Analysis to Support Prognosis of Rotating Machinery Components, 12th International Congress on Sound and Vibration, Lisbon, Portugal, July 1114, 2005
SISTEME DE MONITORIZARE ON-LINE ALE VIBRAIILOR Rezumat: Lucrarea prezint sistemele de monitorizare on-line a vibraiilor, cu conectarea accelerometrelor: pentru msurarea parametrilor industriali, pentru alinierea arborilor i rotorilor, pentru msurarea zgomotelor, pentru sisteme de proximitate.
Mariana ARGHIR, Prof. Dr., Mech. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.264.401.657; Marius BAIDOC, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.770.400; Clin Vasile LUMEI, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.742.030.565; Alina Sabina PACA, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.756.450.183; Florin EPE-BOBESCU, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.547.258.
581

ACOUSTIC NORMAL MODES SIMULATION OF A CLASSROOM

Iulian LUPEA, Roxana TIBREA
Abstract: The acoustic of enclosed spaces continues to be an important issue in building acoustics, high level of comfort in car habitacles of quality vehicles, industrial facilities and others. The management of acoustic modal parameters plays an important roll in reaching the above targets. In this article the finite element method is used to perform the acoustic normal modes analysis of a classroom, finding out the resonant frequencies and the acoustic modes or the pressure distribution for each mode. The results are compared to a set of values determined by a close related formula. The acoustic modes simulation is an important step for the frequency response analysis and the acoustic optimization. Keywords: room acoustic, finite elements, acoustic normal modes.
1. INTRODUCTION Many practical problems involve the acoustic of enclosed spaces like rooms, auditoria, concert halls, acoustic couplers for calibrating microphones, car habitacles, tanks and others [1], [5]. In such enclosures one can highlight sound resonances at specific frequencies and associated sound pressure fields or acoustic mode shapes. A sound resonance and the associated mode shape are determining an acoustic normal mode of the enclosure. In this situation we have to solve the three dimensional wave propagation equation [1], [2], [3], [4]:
2 p t 2 2 2 = c0 p
(1)
where p(x,y,z,t) is the deviation from the ambient pressure, 2 is the Laplace operator, c0 = K [m/s] is the sound speed or the longitudinal wave speed in air, K [Pa] is the bulk modulus or the modulus of bulk elasticity for gas mediums, measuring the air's resistance to uniform compression and is the air density. For solving equation (1) we need to use the proper boundary conditions imposed by the walls of the classroom under study. The wall
surface acoustical properties are of great interest in room acoustic. These properties are completely characterized by the reflection factor for all incidence angles and for the frequency band of interest. The behavior of the walls is well characterized by the wall impedance, too. Wall impedance is based on the particle normal to the wall velocity caused by the sound wave pressure. The reflection factor and the wall impedance are in general complex values. The Helmholtz equation with the boundary conditions of the rigid walls becomes an eigenvalue problem, resulting eigenvalues or natural frequencies and eigenvectors. For simple regular shapes, the natural frequencies and the associated pressure fields can be determined analytically. For more complicated enclosures only numerical methods implemented in finite element or boundary element solvers can help in solving the eigenvalue problem. 2. NORMAL MODES RECTANGULAR ROOM IN A
For a rectangular enclosure ( l x , l y , l z ) the
wave equation becomes:
582
2 p t2
2 2 p = c0 ( x 2
2 p y 2
2 p ) z 2
(2)
and the solution of the differential equation, assuming separation of variables, takes the form: (3) p ( x, y , z, t ) = p x ( x ) p y ( y ) p z ( z ) T (t ) The boundary conditions of the walls are influenced by the acoustical impedance zw = p / u , where p is the complex acoustic pressure at the wall surface and u is the complex particle velocity at the wall level. In general, the acoustical impedance is complex, having a real or resistive component collecting the energy loss mechanisms and an imaginary or reactive component which expresses the potential energy stored in the air: z w = z R + jz I (4) For rigid walls the impedance is infinite: z w = , resulting zero the normal to the wall (n) sound pressure gradient: p =0 (5) n or: p = 0, for x = 0, x = l x ; (6a) x p (6b) = 0, for y = 0, y = l y . y p = 0, for z = 0, z = l z . (6c) z The air particles velocity is zero at the rigid walls level. By considering this situation on the boundary surfaces we are in an undamped case at which by increasing the impedance real part we are adding damping. One assumes a complex exponential for the time dependent factor of the solution (3): p( x, y , z, t ) = p x ( x ) p y ( y ) p z ( z ) e j t (7) where j = 1 and the circular frequencies have to be determined. By inserting the proposed solution (7) into the equation (2), one get:
2 2 px ( x) p y ( y) pz ( z) c0 d 2 p y ( y ) p x ( x ) pz ( z ) dy 2
d 2 p x ( x ) 1 2 = px ( x ) dx 2 c2 0 d 2 py ( y) dy
2
+
(8)
1 py ( y)
d 2 pz ( z ) 1 pz ( z ) dz 2
and for: 2 / c 2 = k 2 , 0 one get:

d 2 p x ( x ) 1 px ( x ) dx 2 d 2 pz ( z ) 1 pz ( z ) dz 2 d 2 py ( y ) dy 2
(9)
1 py ( y )
+
(10)
+ k2 = 0
The terms of the previous equation are independent each to the other. Each term can be equaled with a constant, in order from the list 2 2 2 kx , ky , kz , resulting:
d 2 px ( x ) 2 + kx px ( x) = 0 2 dx
d 2 py ( y) dy
2 2 + ky py ( y) = 0
(11a)
(11b)
d 2 pz ( z ) 2 + kz pz ( z ) = 0 2 dz
2 2 2 kx + ky + kz = k2 .
(11c)
and the sum of the three constants equals k2: (12)
The solutions of the equations (11 a,b,c) are:

p x ( x ) = C1x e jk x x + C2 x e jk x x
(13a)
p y ( y ) = C1 y e
jk y y
+ C2 y e
jk y y
(13a) (13a)
p z ( z ) = C1z e jk z z + C2 z e jk z z The solution (7) of the equation, becomes:
p = (C1x e jk x x + C2 x e jk x x ) (C1 y e
jk y y
d 2 p x ( x ) p y ( y ) p z ( z ) dx dz 2
2
+ C2 y e
jk y y
(14)
d 2 pz ( z ) px ( x ) p y ( y )
(C1z e jk z z + C2 z e jk z z ) e j t
From the boundary conditions at x=0 indicating a null pressure gradient at the rigid wall, results:
Dividing by p(x,y,z), results:
583
p x x
=0
x =0
=> C1x = C2 x
(15a)
(15a)
p y y
p z z
y =0
From relations (9) and (12) where the eigenvalue k 2 ( k = / c0 ) is defined and relation (18) results:
=0
=0
z =0
=> C1 y = C2 y
=> C1z = C2 z
,n y ,n z
= c0 [(
nx
lx
)2 + (
ny
ly
)2 + (
nz
lz
) 2 ]1 / 2
(15a)
The modal frequencies of the cavity in function of the wave numbers are:
f n x ,n y ,n z = c0 n x 2 n y 2 n z 2 1 / 2 [( ) + ( ) + ( ) ] (21) 2 lx ly lz
Replacing these constant relations in the solutions (13), results:

p x ( x ) = C1x ( e jk x x + e jk x x ) = 2C1x cos(k x x ) (16a)
p y ( y ) = 2C1 y cos(k y y )
p z ( z ) = 2C1z cos(k z z )
(16b) (16c)
For the boundary condition at the opposite wall, direction OX, results:
p x x
x =l x
= 2C1x k x sin( k x l x ) = 0
(17)
From this, and for all three directions, results: k x l x = n x , n x = 0,1,2,... k y l y = n y , n y = 0,1,2,...
k z l z = n z , n z = 0,1,2,...
(18a) (18b) (18c)
The sound pressure expression becomes a sum of terms of the following type: p ( x, y , z ) = Cn x ,n y ,nz cos(
nx
lx
x ) cos(
ny
ly
y ) cos(
nz
lz
z)
(19)
where Cnx ,n y ,nz is the modal amplitude of the acoustic mode, identified by the integer numbers n x , n y , n z and x, y, z variables are confined inside the rectangular room:
0 x lx , 0 y l y , 0 z lz
The acoustic modal analysis of the room generates two parameter types: the mode frequencies and the mode shapes. Modes may be interpreted as sums of interfering traveling waves. Analytical solutions like those derived for rectangular volumes can be obtained for cylindrical and spherical volumes, as well. In case of an irregular enclosure the nodal planes become curved nodal surfaces. For getting the acoustical normal modes of the enclosure, all six walls have been considered rigid. In this situation the particle velocity is zero at the walls surfaces and the pressure variation normal to the wall is null, as well. At the walls level the waves are reflected and standing pressure waves are generated. These standing waves are considered the modes shapes of the air in the cavity. The dissipation of the modes energy, in case of the damping presence, makes the modes amplitudes to decrease and the same for the sound pressure in the cavity. At the resonant frequency determination one can starts with perfectly rigid walls and the air without damping. Later the modal damping is added like in the case of structural modal analysis, frequency response or transient response [7].
3. NORMAL MODE SIMULATION BY USING FEA 3.1. The model set-up The geometry of the enclosure has been designed by using CAD. Based on the surfaces the mesh has been generated [8]. In order to model the system under observation the CAD model of the air envelope has been created first. The classroom interior side of the walls and the
Finally, the pressure expression at the position (x,y,z) and moment t, can be written as a summation of acoustic modes:
p ( x , y , z , t ) = C n x ,n y ,n z
nx =0 n y =0 nz =0
cos(
nx
lx
x ) cos(
ny
ly
y ) cos(
nz
lz
(20)
z ) e j t
584
windows are meshed by using triangular shell elements. These are structural nodes, each one with six degrees of freedom, three translational and three rotational. The interior volume of the classroom occupied by the air is meshed by using tetrahedral volume elements, a fixed mesh proper for the fluids [8]. Each node of air has only one degree of freedom, expressing the pressure at that location. The boundary conditions for the air are imposed by the structural nodes of the surrounding structure. The maximum number of the acoustic modes to be modeled is dictated by the mesh size. At least six elements per wavelength are recommended. The mesh density of the acoustic model should be able to predict modes up to the upper bound of the frequency of interest. The material model MAT10 has been used to model the air. The following parameters are requested in the general case: the bulk modulus, the mass density, the speed of sound, the fluid element damping coefficient and the normalized admittance coefficient for porous material.
3.2. Modal frequencies of the cavity The normal modal analysis, by using Optistruct [8], has been run to identify the structure natural frequencies and the associated acoustic mode shapes. A first set of natural frequencies resulted, are listed in Table 1.
Table 1 Comparison rectangular vs. room (FEA)
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
2 1 2 3 5 3 5 3 4 0 4 1 0 6 2 4 1 2 6 5 3 5 3 5
0 1 1 2 0 0 1 1 2 2 0 2 3 0 2 1 3 3 1 2 2 0 3 1
1 1 1 0 0 1 0 1 0 1 1 1 0 0 1 1 0 0 0 0 1 1 0 1
51,30 52,49 56,09 56,80 57,05 57,29 61,39 61,62 64,33 64,55 64,76 65,55 68,00 68,46 68,46 68,61 68,95 71,73 72,11 72,87 73,06 73,25 76,13 76,68
52,5 52,9 55,2 56,3 56,8 59,5 59,6 62,7 63 63,6 64,4 66 66,9 68,2 68,5 69,1 69,4 70,3 71 71,8 72,9 73,6 73,9 76,7
-2,29% -0,82% 1,60% 0,93% 0,45% -3,64% 3,05% -1,65% 2,19% 1,49% 0,50% -0,75% 1,58% 0,42% -0,03% -0,68% -0,58% 1,99% 1,56% 1,50% 0,26% -0,49% 3,02% 0,02%
The eigenvalues distribution is depicted in Fig. 1. The values are evenly distributed indicating a relatively good room design.
Mode# nx 1 1 2 0 3 2 4 1 5 2 6 3 7 3 8 0 9 4 10 0 11 1 12 1 13 2 14 4 15 0
ny 0 1 0 1 1 0 1 2 0 0 2 0 2 1 1
nz 0 0 0 0 0 0 0 0 0 1 0 1 0 0 1
Rectan gular FEA Diff. 11,41 11,3 1,05% 22,67 22,3 1,55% 22,82 22,4 1,83% 25,38 24,9 1,95% 32,16 31,5 2,14% 34,23 34,8 -1,59% 41,05 41,5 -0,96% 45,33 44,1 2,87% 45,64 44,5 2,49% 45,95 45,8 0,36% 46,75 46,9 -0,39% 47,34 48,8 -3,01% 50,75 49,3 2,88% 50,96 50 1,91% 51,23 51,6 -0,79%
Fig. 1. Modal frequencies distribution
One can observe the axial modes of the types (nx, 0, 0), (0, ny, 0) or (0, 0, nz), the tangential modes of the types (nx, ny, 0), (nx, 0, nz) or (0, ny, nz) and the oblique or tree-dimensional modes (nx, ny, nz) in which all of the indices nx, ny and nz are non-zero. Often it is of interest to find out the density of the eigenfrequencies in a specific frequency band like an octave or an octave fraction.
585
3.3. The modal pressure distribution The first acoustic mode (an axial mode) [100] is along OX axis, obtained at 11.3 Hz (Fig. 2).
Hz associated to the first axial mode (along OY axis [010] (Fig. 3). For the mode 3 [200], the pressure distribution and the two zero pressure surfaces, can be seen in figure 4. These two
Fig. 5. Mode #4 [1 1 0] Fig. 2. Mode 1: 11,3 Hz [1 0 0]
Fig. 6. Mode 10: 45,8Hz [0 0 1]
Fig. 3. Mode 2: 22,3Hz [0 1 0]
This is normal considering that l x > l y and l x > l z . Zero pressure surface is about at the middle section of the length l x of the classroom.
surfaces are deformed because of the details of the classroom. For a parallelepiped enclosure these two surfaces would be planar. The maximum pressure is located at the two opposite walls, following the relation: 2 p( x ) = cos( x) C 2 , 0, 0 lx The first tangential mode [110] (Fig. 5) was obtained at 24.9 Hz. The maximal pressure is located at the two opposite corners. The first axial mode [001] along the OZ axis or the height of the classroom is depicted in figure 6, where one can see the zero pressure surface. The associated frequency was obtained at 45.8Hz. Another tangential acoustic mode has two zero pressure surfaces along OX direction and one null pressure along OY direction [210] (Fig. 7). A much higher oblique acoustic mode [331] is depicted in figure 8. For such a mode the particle velocity has components on all three coordinate axes. The reflections cover all six walls. The higher the mode the more complicated the pressure distribution is.
4. CONCLUSIONS
Fig. 4. Mode 3: 22.4Hz [2 0 0]
The frequency of 22.4 Hz associated to the second axial mode along OX axis [200] shown in figure 4, is very close to the frequency of 22.3
The acoustic modes of a classroom has been observed from analytical and simulation
586
points of view. The values of the fundamental resonant frequencies calculated by using the relation (21) and the values resulted from simulation are in good agreement. The
frequency of the classroom or is close to it. The excitation is depending, as well, on the position of the source in relation to the nodes or the antinodes of the pressure distribution. The acoustic modal analysis by using finite elements is a good and confident approach in observing acoustic systems with applications. Further studies will observe the frequency response and the pressure distribution as a response to the human voice.
5. REFERENCES
Fig. 7. Mode #5, [2 1 0]: 31.5Hz
Fig. 8. Mode [3 3 1]
free oscillations problem or the homogeneous problem has been observed in the paper. In this case we had initial conditions to start oscillations, the boundary conditions were for perfectly rigid walls and no damping in the air and at the walls. Often in the room we have sound sources. In this case forced oscillations are generated. When the source is a pure tone at a specific frequency only that frequency can be observed in the room. An acoustic mode can be excited when the frequency of the source fits a natural
[1] Bdru, E., Grumzescu, M., Bazele acusticii moderne, Ed. Academiei RPR, 1961. [2] Everest F., Pohlmann, K., Master handbook of acoustics, McGraw Hill, 2009. [3] Fahy, F., Foundation of engineering acoustics, Elsevier Academic Press, 2005. [4] Lawrence, E., Austin, R., Coppens, A., Fundamentals of acoustics, John Wiley & Sons, Inc. NewYork, 2000. [5] Heylen, W., Lammens, S., Sas P., Modal Analysis Theory and Testing, Leuven: Katholieke Universteit Leuven, Department Werktuigkunde, 2002. [6] Lupea, I., Vibration and Noise measurement by using Labview programming, Casa Crii de tiin Publisher, Cluj-Napoca, 2005. [7] Lupea, I., Szathmari, R., Vibroacoustic frequency response on a passenger compartment, J. of Vibroengineering, Kaunas, Lithuania vol. 12 (4), Dec. 2010. [8] ** HyperWorks - Optistruct, Altair Eng., Troy, MI, USA.
Analiza modal acustic a unei sli de cursuri folosind elemente finite Rezumat: n lucrare sunt determinai prin simulare parametrii modali ai unei sli de cursuri. Se consider modelarea geometric a slii, discretizarea pereilor i a volumului de aer. Folosind solverul Optistruct, se extrag frecvenele naturale i modul de distribuie a presiunii pentru fiecare mod acustic. Valorile provenite din simulare sunt comparate cu cele calculate folosind o relaie obinut analitic, pentru o sal asimilat cu un paralelipiped. Sunt fcute aprecieri i formulate concluzii asupra fenomenului studiat.
Iulian Lupea, Professor Ph.D., Technical University of Cluj-Napoca, The Department of Applied Mechanics and Computer Programming, iulian.lupea@mep.utcluj.ro, 0040 264401691. Roxana Tibrea, Assistant professor, Technical University of Cluj-Napoca, The Department of Civil Engineering.
587

MODAL ANALYSIS OF AN INFLATED TIRE BY SIMULATION

Iulian LUPEA, Marius MORARIU
Abstract: The automakers are continuously interested in NVH solutions for high quality vehicles. The dynamics of a tire, which is excited mainly from the road, is under observation. In this paper the modal parameters of an inflated tire are determined by using finite element analysis. The natural frequencies and the modal vectors of the prestressed structure for a couple of pressure values are presented. The pressure influence on the natural frequencies can be assessed. Further studies will observe the experimental modal analysis and the frequency response analysis of the structure. Keywords: inflated tire, modal analysis, modal parameters.
1. INTRODUCTION Two important issues that appear when engineers are confronting with automotive design are improving the vehicles dynamics and NVH (Noise, Vibration and Harshness) refining. The wheels are one important target when dealing with these problems because they have the role of sustaining the vehicle and responding the maneuvers of the driver on the road [9]. As a consequence they are, along with the engine, the main sources of vibrations when are dynamically excited by the unevenness of the road, or by braking, steering and accelerating the vehicle [4]. Modal analysis is a good method to formulate a mathematical model for the dynamic behavior of the wheel. It consists on the process of determining its inherent dynamic characteristics like natural frequencies, damping factors and mode shapes in order to build the modal model. There are theoretical models based on the physics of the tire construction, and empirical models which are based on experimental data. One simple, yet reliable theoretical model, was conceived by Zagelaar [11] and it consists of a rigid ring that represents the thread-band, which is suspended on translational and
rotational springs and dampers representing the tire sidewalls and the pressurized air inside, which are linked to a smaller rigid ring representing the rim. This model can only simulate the low frequency in-plane rigid modes of the tire, which means that the threadband keeps its circular shape and the movements consist in rotational (Fig. 1), a horizontal and a vertical displacement of the ring (Fig. 2). Such a simple model is mostly used for studying the ride and comfort of vehicles in the low and medium frequency range (0-100Hz)
Fig. 1. Rotational modes (in plane)
However, when studying the dynamic behavior of the tire at higher frequencies, a model that simulates a more complex deformation is needed. As a consequence, in general a finite element model is being introduced for performing modal analysis.
588
Brinkmeier et al. describe [1] this kind of approach for determining the radiating noise produced by a rolling tire. To predict the dynamic forces exerted on the tire axis, an
Fig. 2. Translational modes (in plane)
analysis of a 3-D tire model with finite elements passing over an obstacle is studied by Cho et al. in their work [2]. The effects of the inflating pressure and rolling speed are taken into account during these investigations. Finite elements analysis is also being used for examining the contact pressures between the tire and the road for different thread patterns [3], or to validate a FEM model of a tire by simulating several test procedures in order to be used in future vehicle dynamics analysis [7]. 2. THE FINITE ELEMENT MODEL Sheet metal wheels are mounted on most of the cars because they are economic, they have a high stress limit and can be easily serviced. They are constructed from two distinguished parts, the rim and the wheel disc (attachment face) and then welded together. The material from which they are fabricated is a cold-formable sheet metal or band steel, with a high elongation and thicknesses from 1.8 to 4.0 mm for the rim and 3.0 to 6.5 mm for the disc, depending on the wheel loads. The metal plate from which the rim is fabricated comes in a rectangular form, and then is bent to produce a cylindrical sleeve with two free edges which will be welded afterwards. To give the sleeve the corresponding thickness profile it follows at least one cylindrical flow spinning operation followed by calibration. The steel disc is obtained by stamping a metal plate. It has to have appropriate holes for
fitting the center hub and the bolts which will fasten the wheel to the hub. The disc can be perforated also for purposes like better brake cooling or just reducing weight. Also many strength requirements are placed on the wheel disc because it has to absorb vertical, lateral and longitudinal forces coming from the road and to transfer them to the wheel through the fixing bolts. Starting from the surfaces of the geometry the mesh, mainly based on shell elements, has been done. The tire size is 185/65/R15 (tire width [mm]/ratio of the tire height to width [%]/ rim diameter [inch]. The height of the sidewall of the tire is 65% of the width (120.25 mm). The cylindrical rim structure is supported by a central disc made by stamping. This part has appropriate holes for the center hub and lug nuts. The outer surface of the wheel disc has a cylindrical portion to support the connection to the rim. Both, the rim and the central disc are modeled by using shells
Fig. 3. The wheel mesh layout
and the material assigned is steel: Young modulus: 2.1e5MPa, Poisson ratio of 0.3, the density 7.9e-9tons/mm3. The rim and the disc thickness values are 2.3mm and 4.2 mm respectively. The rubber material for this study is considered as isotropic E=200MPa, Poisson ratio =0.49, density =2.1e-9 and thickness=6.24mm 3. NORMAL MODE ANALYSIS OF THE PRELOADED SYSTEM 3.1. The prestressed eigenvalue analysis The dynamical equations of some mechanical systems [5], [6] can be written in a matrix form (1), as follows: && + CQ & + KQ = F MQ (1)
where Q is the vector of the system generalized coordinates (system degrees of freedom), M, C, K are the mass (inertia), viscous damping and stiffness matrices and F is the vector of generalized forces. The matrices M, C and K are considered symmetrical, with n x n elements. When doing the normal mode analysis of the system, the damping is neglected and the external forces are not acting, resulting (2): && + KQ = 0 MQ (2) Assuming a harmonic and synchronous motion in the structure, when all coordinates perform in time the motion in phase or out of phase, the following solution is proposed: Q (t ) = u cos( t ) (3) where is considered the natural frequency of the whole system and u is a constant nvector of amplitudes. By replacing the proposed solution in the system of differential equations, results: Ku 2 Mu = 0 or ( K 2 M ) u = 0 (4) The set of the homogeneous algebraic equations (4) has the unknown vector u. Considering = 2 as a parameter, one gets: Ku = Mu (5) known as the eigenvalue problem when trying to determine values for which the system (4) has nontrivial solutions. By solving for and nontrivial solution (u0), the following characteristic equation (6) is obtained: det( K M ) = 0 (6) where r (r=1,2,n) values are the eigenvalues (or characteristic values) of the system. The natural frequencies of the system are derived: f r = r / 2 r = 1,2,...n (7) For each determined eigenvalue r , a vector ur (r=1,2,n) named eigenvector, defining the associated mode shape of vibration is calculated. ur is satisfying the following equation: Ku r = r Mu r r = 1,2,..., n (8) In case of a preloaded structure like the one under observation the stiffness matrix becomes:
K = K 0 K PL
( K 0 K PL M ) u =0
589
The matrix K PL has to be determined in an initial static analysis case of the type: ( K 0 K PL ) U 0 = F (11) where U 0 and F are the static deformation of the structure and the external preloading force respectively. 3.2. Natural frequencies of the system The normal modal analysis, by using Optistruct [12], has been applied to identify the structure natural frequencies and the associated mode shapes. The tire is constrained by using four bolts at the central area. The recommended tire pressures are 1.9 and 2.1 bars for the front and the rear tires respectively. For the structure critical locations the strain energy distribution has been observed. The natural frequencies for a pressure of 0.2MPa are listed in Table 1.
Table 1 Natural frequencies of the tire Mode Frequency Mode Frequency # (Hz) # (Hz)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
18.6 51.5 70.1 70.1 128.9 129.4 142.2 142.2 151.8 151.8 167.3 167.3 173.6 180.2 181.6
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
186.8 186.8 191.3 191.4 206.7 206.8 207.4 207.4 216.0 229.1 229.1 240.8 240.9 250.3 250.3
3.3. Mode shapes The first mode shape, depicted in figure 4, is
(9)
where the preloading is defined or included in the matrix K PL . The relation (5) becomes:
( K 0 K PL ) u = M u
(10)
or: Fig. 4. Mode shape #1
590
characterized by the rotation of the outer zone of the tire (red area) in the wheel plane, around the wheel axis. All figures presented in the sequel are showing the displacement associated to the mode shapes. The steel is much more rigid in comparison to the rubber, consequently the displacement is reduced, the central area being in
In the second mode of vibration the tire is moving out of the wheel plane and along the wheel axis (Fig. 5). Both, the deformed (amplitude) and the undeformed shape of the system for this mode can be observed. In the third and the fourth modes the tire is
Fig. 8. Mode shape #7,#8
Fig. 5. Mode shape #2
Fig. 9. Mode #13
Fig. 6. Modes #3 and #4
Fig. 10. Radial lobes
moving out of the wheel plane, rotating around an axis found in the wheel plane (Fig. 6).
Fig. 11. Multiple radial lobes
Fig. 7. Mode shape #5 dark blue colour.
Figure 7 is presenting the fifth mode from two viewpoints, a side one and a top view. The movement is mainly axial. Two diametric opposite zones are moving out of phase to the
591
other perpendicular and diametric opposite zones. The modes number seven and eight are similar and are depicted in figure 8. Mode #13 is depicted in figure 9. Both the deformed and wire frame undeformed structures are visualized. As long as the mode and the associated frequency are higher, the associated mode shape is more complicated deformed presenting an increasing number of lobes like the modes depicted in figures 10 and 11. 3.4. The pressure variation influence on the natural frequencies The natural frequencies are influenced by the degree of the structure pretension as we can see, as well, from the relations (10) and (11). The finite element analysis has been applied multiple times by considering each time a different tire pressure and hence a different prestressed tire structure. The tire pressure list of values is 0.18, 0.2, 0.22, 0.24 [MPa]. The results are centralized in Table #2.
Table 2 Natural frequencies variation with pressure Mode #
4. CONCLUSION The modal analysis of an inflated tire is important for the tire dynamics, revealing the resonant frequencies and the associated eigenmodes complexity of the structure. For the tire size of 185/65/R15, the variation of these frequencies as a consequence of the tire pressure variation has been observed. This task has been accomplished by using the finite element simulation. 5. REFERENCES [1] Brinkmeier, M., Nackenhorst, U., et al, A finite element approach for the simulation of tire rolling noise, Journal of sound and vibration 309, pp. 20-39, Elsevier, 2008, [2] Cho, J., R., Kim, K., W., et al, Transient dynamic response analysis of a 3-D patterned tire rolling over a cleat, European Journal of Mechanics A/Solids 24, pp. 519 531, Elsevier, 2005, [3] Ghoreishy, M., H., R., Finite Element Analysis of the Steel-Belted Radial Tyre with Tread Pattern under Contact Load, Iranian Polymer Journal / 15(8), pp.667-674, Iran Polymer and Petrochemical Institute, Tehran, 2006, [4] Harrison, M., Vehicle RefinementControlling Noise and Vibration in Road Vehicles, Butterworth-Heinemann, ISBN: 07680-1505-7, Oxford, 2004. [5] He, J., Fu, Z., F., Modal Analysis, Butterworth-Heinemann, ISBN: 0-75065079-6, Oxford, 2001. [6] Heylen, W., Lammens, S., Sas P., Modal Analysis Theory and Testing, Leuven: Katholieke Universteit Leuven, Department Werktuigkunde, 2002. [7] Lee, C., R., Kim, J., W., et al, Validation of a FEA Tire Model for Vehicle Dynamic Analysis and Full Vehicle Real Time Proving Ground Simulations, SAE paper no.971100. [8] Lupea, I., Vibration and Noise measurement by using Labview programming, Casa Crii de tiin Publisher, Cluj-Napoca, 2005. [9] Pacejka, H., B., Tyre and Vehicle Dynamics, Butterworth-Heinemann, ISBN13: 980-0-7506-6918-4, Oxford, 2006.
Natural Frequencies [Hz] 0.18 0.2 0.22 0.24 MPa MPa MPA MPA
17.7 49.8 68.9 68.9 127.9 128.3 140.9 141.0 149.9 149.9 164.6 164.7 171.7 178.7 180.1 18.6 51.5 70.1 70.1 128.9 129.4 142.2 142.2 151.8 151.8 167.3 167.3 173.6 180.2 181.6 19.3 53.1 71.2 71.3 129.9 130.4 143.4 143.5 153.6 153.7 169.9 169.9 175.3 181.8 183.1 20.1 54.6 72.4 72.4 130.9 131.4 144.6 144.7 155.4 155.4 172.3 172.4 177.0 183.3 184.6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
As we can see from the results gathered in Table 2, the variation of the frequency for the same frequency mode is within 2 to 5 Hz for a pressure variation of 0.6 MPa. An experimental modal analysis has to be done for the simulation validation.
592
[10] Reimpell, J., Stoll, H., Betzler, J., W., The Automotive Chassis: Engineering Principles, Butterworth-Heinemann, ISBN: 9780750650540, Oxford, 2001. [11] Zegelaar, P., W., A., The Dynamic Response of Tyres to Brake Torque
Variations and Road Unevennesses, Delft University of Technology, ISBN 90-3700166-1, Delft, 1998. [12] ** HyperWorks - Optistruct, Altair Eng., Troy, MI, USA.
Analiz modal aplicat unei anvelope folosind metoda elementelor finite Rezumat: n lucrare sunt determinai parametrii modali ai unei roi de automobil prin analiz cu elemente finite. Deoarece diferena de rigiditate dintre anvelop i jant este mare, modurile de frecven mai joas vor presupune n primul rnd deplasarea anvelopei. La nceput este determinat starea de tensiuni a roii ca urmare a presiunii aerului din anvelop. In a doua etap sunt calculate frecvenele naturale i modurile de vibraie ale structurii pretensionate. In continuare este observat variaia frecvenelor naturale n raport cu variaia presiunii din pneuri.
PRE-STRESSES TYRE MODAL ANALYSIS SUBCASE 1 SPC = 3 LOAD = 2 SUBCASE 2 SPC = 3 METHOD(STRUCTURE) = 1 STATSUB(PRELOAD) = 1 $ BEGIN BULK GRID 6600040 2724.98639.17379215.7784 GRID 6608089 2903.16496.76384149.8047 $$ $ RBE2 Elements - Multiple dependent nodes $ RBE2 8122 6608084 123456 6600126 6600127 6600199 6600291 6600292 + 6600293 6600337 6600341 6600342 6604529 6605606 6605608 6605610 + 6605614 6605932 6605934 6605935 6605936 6607930 6607933 .... RBE2 8126 6608087 123456 6600413 6606180 6606183 6606839 6607928 + 6607929 CTRIA3 66 3 6607943 6607945 6607944 .... CTRIA3 5206 6 6606141 6602188 6606139 $ CQUAD4 59 2 6607963 6607960 6607961 6607962 CQUAD4 8100 1 6603540 6600043 6603534 6604916 CQUAD4 8101 1 6603539 6603535 6603534 6600042 $ $HMNAME PROP 1"PNEU_D" 4 PSHELL 1 6600126.24 660012 660012 0.0 $HMNAME PROP 2"ZONE_RECOUVREMENT_D" 4 PSHELL 2 6600116.5 660011 660011 0.0 $HMNAME PROP 3"DISQUE_DE_ROUE_D" 4 PSHELL 3 6600114.2 660011 660011 0.0 $HMNAME PROP 4"JANTE_DE_ROUE_D" 4 PSHELL 4 6600112.3 660011 660011 0.0 $HMNAME PROP 5"EVENTS_D" 4 PSHELL 5 6600126.24 660012 660012 0.0 $HMNAME PROP 6"EVENTS_D_2" 4 PSHELL 6 6600126.24 660012 660012 0.0 MAT1 660011210000.0 0.3 7.9-9 MAT1 660012200.0 0.49 2.1-9 EIGRL 11.0 300.0 MASS SPC 3 6608087 3 0.0 SPC 3 6608086 3 0.0 SPC 3 6608089 3 0.0 PLOAD2 20.2 3037 PLOAD2 20.2 623 PLOAD2 20.2 $$ ENDDATA
3 6608084 1245 0.0 6350
SPC
3 6608088 3
0.0
SPC
Iulian Lupea, Professor Ph.D., Technical University of Cluj-Napoca, The Department of Applied Mechanics and Computer Programming, iulian.lupea@mep.utcluj.ro, 0040 264401691. Marius Morariu, Ph.D. student, Technical University of Cluj-Napoca, The Department of Applied Mechanics and Computer Programming, 0040 264401691.
593

CONTRIBUTIONS TO THE DYNAMICS OF MOBILE ROBOT PATROLBOT

Iuliu NEGREAN, Claudiu SCHONSTEIN, Zoltan SZOKE
Abstract: In keeping with the fact that the robots during operations are performing moving trajectories situated in the configuration space, or in the Cartesian space, its imposed a continuous control of kinematic parameters and respectively the generalized force from every driving joint, in order to achieve a proper control of the dynamic parameters. In keeping with this, the purpose of the paper is to determine the dynamic equations for the mobile robot PatrolBot. The motion equations will be determined on the basis of the new concepts in advanced mechanics, such as Variational Principles in Analytical Mechanics, concerning the acceleration energy, part of important scientific researches deployed by the main author. Key words: mobile robot, variational principles, motion trajectory, control, acceleration energy.
1. INTRODUCTION In the paper will be analyzed the dynamics of the mobile robot structure called PatrolBot, which is a product of the American company Mobile Robots Inc., located in the laboratory "Mechanics of Advanced Robotics", in the Department of Mechanics and Programming, from Technical University of Cluj-Napoca. The mobile platform is a system able to integrate in a fairly wide class of general applications [1].
and independent each other. In this feature, it appears that mobile robot cannot move laterally with a speed that is transverse to the axis of the wheels; this constraint is called nonholonomous constraint. Hence, the robot can reach any position and orientation, moving forward or backward, followed by a rotation around its central axis [2], [3]. 2. THE DYNAMICS OF PLATFORM PATROLBOT MOBILE
In keeping with the Fig.2, the kinematical differential restrictions of the platform are presented in accordance with the Table 1.
y0
{Mrs2 ; r2}
yR
y2
q5
y 4 q8
x2
R
x4
O3
{Mrf 4 ; r4 }
xR
q9
O2
= q3
q3
xC
2b y 3 q6 C
x3
O4
{M rf 3 ; r3 }
2l
y1
q4
O1
x1
q3
q7
Fig 1 The mobile robot PatrolBot
q2
rP
PatrolBot structure has two independent driving wheels and two driven wheels, so it is included in class mobile systems with differential orientation. The robot executes a plan-parallel movement so in the configuration space, each of the three movements are parallel
{M rs1; r1}
x0
O0
q1
Fig.2 The independent parameters of PatrolBot in finite displacements
594 Table 1
s q3 dq1 + c q3 dq2 = 0 c q3 dq1 + s q3 dq2 + l dq3 rrs dq 4 = 0 c q3 dq1 + s q3 dq2 l dq3 rrs dq5 = 0
(1) (2) (3) (4) (5) (6) (7)
s ( q3 + q7 ) dq1 + c ( q3 + q7 ) dq2 + + ( L c q7 b s q7 ) dq3 = 0
s ( q3 + q9 ) dq1 + c ( q3 + q9 ) dq2 + + ( L c q9 + b s q9 ) dq3 = 0

c ( q3 + q7 ) dq1 + s ( q3 + q7 ) dq2 +
The generalized expression (2) represents the acceleration energy for any robot kinetic link, where M i is the mass corresponding to each kinetic link of the robot, i Ii* which is the axial centrifugal inertia tensor and i Ipi the inertia tensor planar centrifugal that characterizes the entire kinetic assembly ( i ) , relative to the frame {i} , applied in the mass center of each link Ci . In the same expression,
i
vCi
and
i& vCi
are
+ ( L s q7 + b c q7 ) dq3 rrf dq6 = 0
+ ( L s q9 b c q9 ) dq3 + rrf dq8 = 0
c ( q3 + q9 ) dq1 + s ( q3 + q9 ) dq2 +
In the previous expressions from Table 1, there have been introduced the following notations: {c qi = cosc qi ; sqi = sinc qi , i = 1 n} . The moving differential equations for any mechanical system can be obtained by means of the DAlembert-Lagrange principle considered as fundamental equation in the study of the dynamics of mechanical systems with links. In the study, to determine the dynamic equations of the serial structure there are used fundamental notions of advanced mechanics like acceleration energy, [2], [3] [4] which is integrated in the following, expressed:
EA && i i i + Qg + QSU = Qm ;& ; &&i q
the velocity and the acceleration of mass center, i & i and i are the angular velocity and i acceleration of the kinetic link ( i ) relative to the moving frame {i} attached to the robot. Using as fundamental expression [4], the relation (2), the expression of acceleration energy for mobile platform PatrolBot is [5]:
tot EA 2 2 &&1 + l q &&3 c q3 l q &2 s q3 + q &&2 + l q &&3 s q3 + l q &2 c q3 + Mrs q 3 3 = 2 2 + 2 && &&3 c q3 + l q &2 s q3 + q &&2 l q &&3 s q3 l q &2 c q3 + q1 l q 3 3 2 Mrs rrs 3 3 &&2 &&2 &4 &4 2 2 2 4 2 2 &3 q &4 + q &&5 + q &5 + q &3 q &5 + + q4 + q3 + q4 + q3 + q 4 2 2 2 Mrf 2 2 && L q &&3 s q3 + q &3 c q3 + b q &&3 c q3 q &3 s q3 + + q 1 2 2 2 2 && && & && & + q2 + L q3 c q3 q3 s q3 + b q3 s q3 + q3 c q3 +
) ( ) (
(1)
i Qg
where,
QSU
is the generalized handling force,
i and Qm are generalized gravitational and generalized driving forces, which are characterizing every driving joint of the general robot structure. The expression (1) presented above, represents a generalization of the variational principle based on acceleration energy which is further used for determining the differential equations of motion for any type of mechanical robot structure. In relations (1) the i &k ; q &&k ) represents the following [4]: term E A ( qk ; q
2 Mrf rrf &&2 1 &&2 &4 1 &4 3 &2 &2 &&2 1 &&2 &4 1 &4 3 &2 & 2 q 6 + q7 + q6 + q7 + q6 q7 + q8 + q9 + q8 + q9 + q8 q9 + 4 2 2 2 2 2 2 1 2 &&2 &&1 &&3 ( q &&1 s q3 q &&2 c q3 ) + M q + q2 M R xC q 2 1 && &&2 &4 &2 ( q && M R xC q 3 1 c q3 + q2 s q3 ) + 2 IP q3 + q3
) ( ) ) ( ) ( 2 2 2 &&1 L ( q &&3 s q3 + q &3 &&3 c q3 q &3 + q c q3 ) b ( q s q3 ) + 2 2 2 && & && && & + q2 + L ( q3 c q3 q3 s q3 ) b ( q3 s q3 + q3 c q3 ) + ( ( ) ( )
where,
{q j ; j = 1 9}
(3) represents the generalized

R
coordinates,
are the undetermined Lagrange parameters, xC is the mass center of the mobile robot, and IP represents the inertia
i ; i = 1 7
1 i &T iv & + &k ; q &&k ) = M i i v EA ( qk ; q Ci Ci 2 1 i &T i * i & i + i Ii i + i i Ii* i i + 2 1 i &T i + i i i Ii* i i + 2 1 T i T + i i Tr i I pi i i i iT i I pi i i i i 2
(2)
( )
moment with respect to Cz axis. In the same expression, M rs and M rf are the masses of the back and front wheels; rrs and rrf are the radii of the wheels, and M is the total mass of the mobile structure PatrolBot [2]. The differential moving expression of mobile robot PatrolBot are as presented in Table 2.
595
&&1 q &&3 s q3 + q & 2 c q3 M pl R xC + 2 Mf L = M q 3 = 1 s q3 + c q3 ( 2 + 3 ) 4 s ( q3 + q7 ) 5 s ( q3 + q9 ) + 6 c ( q3 + q7 ) + 7 c ( q3 + q9 )
)(
Table 2
(1)
&&2 + q &&3 c q3 q & 2 s q3 M pl R xC + 2 Mf L = M q 3 = 1 c q3 + s q3 ( 2 + 3 ) + 4 c ( q3 + q7 ) + 5 c ( q3 + q9 ) + + 6 s ( q3 + q7 ) + 7 s ( q3 + q9 )
)(
be analyzed independently. Hence, to realize a straight line motion, according to its design, an essential condition is that the driving moments of 4 5} the wheels must be equal {Qm = Qm , so in keeping with this it can be deduced that [5]:
&&4 = q &&5 ) , q3 = cst., ( q &3, q &&3 ) = 0, ( q &&6 = q &&8 ) , ( q7, q9 ) = 0} (4) {(q
(2)
&&3 I + 2 Mrf b2 + L2 q P &&1 sq3 + q &1 q &3 c q3 q &&2 c q3 + q &2 q &3 sq3 ) + Mpl R xC + 2 Mrf L ( q
= l ( 2 3 ) + 4 ( L c q7 b sq7 ) + +5 ( L c q9 + b sq9 ) + 6 ( L sq7 + b c q7 ) + 7 ( L sq9 b c q9 )

2 M rs rrs 4 &&4 + Q rs sgn ( q &4 ) = Qm q 2 rrs f
(4)
&3 Mpl xC + 2 Mrf L ( q &1 c q3 + q &2 sq3 ) = +q
(3)
According to Table 2, it can be seen that, ( 2 = 3 ) respectively ( 6 = 7 ) .There it is introduced the notation:
&&4,5 = P = q && && 1 q 1 q 1, 2 cos q3 rrf sin q3 rrf
(5)
2 M rs rrs 5 &&5 + Q rs sgn ( q &5 ) = Qm q 3 rrs f 2 2 M rf rrf &&6 + Q rf sgn ( q &6 ) = 6 rrf q f 2 2 M rf rrf &&8 + Q rf sgn ( q &8 ) = 7 rrf q f 2 2 M rf rrf &&7 + Q 7 sgn ( q &7 ) = 0 q f 4 2 M rf rrf &&9 + Q 9 sgn ( q &9 ) = 0 q f 4
(5)
with the observation that { P } takes the first form from (5) in the case of translation along q1 , respectively the value of the second term in the case of displacement along q2 . The expressions of the driving moment of the robot PatrolBot in order to realize a translation are:
r 4 5 &&1 c q3 + q &&2 s q3 ) + M ( q Qm = Qm = rs 2 + P ( M rf rrf + M rs rrs ) + M g
(6)
(6)
(7)
(8)
(9)
To realize the orientation, the following 4 5} condition has to be accomplished {Qm = Qm . The conditions for orientation of the mobile robot PatrolBot are:
&1 ,q &2 ) = 0,( q &&1,q &&2 ) = 0} { ( q1,q2 ) = cst ., ( q
(7)
The moving of the mobile structure should be studied considering the fact that the structure, to achieve the target point, cannot simultaneous do a positioning (translation) and an orientation, so the two movements will be analyzed independently [5]. 3. DETERMINING OF PATROLBOT`S WHEELS DRIVING MOMENTS The dynamics equations presented above are highlighting the complexity of the dynamic control. In order to express the dynamic control functions of the mobile structure, on an established trajectory, the actuating motors must overlap the generalized forces , as inertial forces, or gravitational forces (of the handled object). The moving of the structure has to be analyzed in keeping with the fact that the robot PatrolBot, due to its structure in order to achieve the goal point can`t realize simultaneous a translation and a orientation, hence the two displacements will
This in keeping with the restriction conducts to:

&&5 = q &&4 ) , ( q &&6 = q &&8 )} {( q {( 2 = 3 ) , ( 6 = 7 )}
(8)
After a few transformations, there are obtained the expressions of the driving moments for the two wheels in the case of orientation, as having the following form:
I + 2 M b 2 + L2 P rf L2 + b 2 4 5 = Qm = + M rf + Qm 2l 2l M g rrs l s &&3 + + M rs rrs q Qm 2 2 ( rrf rrs ) + L
(9)
where
L2 + b 2 s R = + L R xC + rrf Qm xC rrs l
) .
The relations (6) and (9) are representing the expressions of the driving moments of the mobile robot wheels. Analyzing the above expressions, it can be seen that the final form of the driving moments containing a static and a dynamic component.
596
4. THE WORKING PROCESS OF THE MOBILE ROBOT PATROLBOT Further is considered a working process, presented in Fig.3, where is integrated the mobile platform PatrolBot. The process contains ( j = 1 9) working sequences, consisting in five translations and four orientations.
xR
yR
After a few transformations, are obtained the functions for generalized velocities and coordinations as shown: ( )2 && ( i 1 )2 && & ji ( ) = i q q ji 1 + q ji + 2 ti 2 ti (12)
q ji t q t &&ji ji 1 i q &&ji 1 + i q 6 6 ti ti
P 2
ji + 6 ti q ji t q t &&ji ( i 1 ) + ji 1 i q &&ji 1 ( i ) (13) + i q 6 ti 6 ti
q ji ( ) =
(i )3 q &&
6 ti
ji 1 +
( i 1 )3 q &&
yR
P 1
xR
3
xR
4
xR
yR
The input parameters for study are presented as follows in the Table 3:
Table 3 Seq. Config. Time j=19 k=027 jk s 1 0 1 2 3 6* 7 8 9 0 2 4 6 18 18,5 18,75 19,25 Duration Coordinates values
ti s q jk m , rad
6 8
yR
y0
P
9
xR
P 7
x0
yR
O0
Fig.3 The working process imposed to PatrolBot
0 2 2 2 0 0,5 0,25 0,5
0 2 1,571 0
Further, from the working process of mobile robot described by Fig. 3, will presented the sequences ( j = 1,3 ) , which representing a translation ( q2 ) , and orientation ( q3 ) with = 2 .
the be are an
On the basis of the parameters presented in Table 3 in keeping with (11)-(13), are determined the expressions for coordinates, velocities and accelerations as in Table 4:
Table 4
Secq. j=19 Interval j=13 Coord.* Expressions for generalized coordinates, velocities and accelerations
4.1. Polynomial functions of ( 3n ) type with restrictions The interpolating functions are consisting in generation of linear functions with respect to time for generalized accelerations from every driving joint belonging to the robot. According to [4], [5] it is generated a linear function with respect to time as follows:
i 1 &&ji ( ) = i &&ji ( i 1 ) + &&ji ( i ) q q q ti ti
q jk m , rad
0,0417 3
3 2
& jk q & q jk m s , rad s

0,125 2
& jk m s2 , rad s2 q
&&jk q
1 2
0,25
1,5 - 0,5 0,25 - 1,5
q2
0,1 + 0,75 0,25 2 + 1,5 ( 1,5 + 1
0,0417 ( 6)3 +
0,125 ( 6)2
(10)
3
1,571 3,351 ( 10,053 ( 18)2 10,1 (2 - 36)

3 40,2 2 q3 13,4 748,9 +13945,7 865 1497,8 + 13945, 80,4 - 1497,8
where t i = i i 1 represents the duration of each ( i = 1 3 ) segment of the trajectory. The unknowns for the generalized accelerations at i 1 and i are defined as:
&&ji ( i 1 ) = q &&ji 1 ; q &&ji ( i ) = q &&ji . q
3,351 ( 19,25 10,053 ( 19,2 10,1 (2 - 38,5)
(11)
On the basis of Table 4, are represented the time variation of generalized coordinates, velocities and accelerations.
q2 ( ) m
597
On the basis of expression (6) and (9), are represented graphically the variation for the driving moments, on the sequences ( j = 1,3 ) .
Qm
4.0814
N m
s
&2 ( ) m s q
2.8435
1.6055 1.2342
s
0
s
&&2 ( ) m s2 q
1.2342
Fig. 6 The variation of driving moments (sequence j = 1 )

Qm
5.3799
N m
s
2.8424 2.5375
0.3049 0 0.3049
Fig. 4 The kinematical parameters on sequence j = 1

q3 ( ) rad
2.5375 2.8424
5.3799
Fig. 7 The variation of driving moments (sequence j = 3 )
The previous graphics are representing the variation with respect to time of the driving moments, in the case of the two independent movements: linear translation and rotation for orientation. Their determination was based on the conditions necessary to achieve a straight shift or orientation of the robot and the kinematic constraints. 5. CONCLUSIONS In the above presented paper there have been developed the dynamic control functions for the mobile robot PatrolBot. Using as starting point the dynamics equations, in keeping with the conditions which must be taken into account for straight line displacement or orientation, there have been deduced the driving moments of the wheels. The dynamic model of the structure is based on acceleration energy, which substituted
&&3 ( ) rad s2 q
Fig. 5 The kinematical parameters on sequence j = 3
598
in the specific expression of mechanical systems with nonholonomous links are leading to differential equations of motion for the mobile robot PatrolBot. Having the driving moments of the wheels, the structure was integrated in a technological process. The process has been modeled using ( 3n ) .type with restrictions polynomial functions, in keeping with the conditions necessary to achieve a straight shift or orientation of the robot and the kinematic constraints. For two sequences of the working process, consisting in a linear translation and a rotation for orientation, there were represented graphically the variation in time of the driving moments. As a final conclusion, the graphical representation is in accordance with the real working process of the mobile robot PatrolBot. 6. REFERENCES [1] ActivMedia Robotics. Mobile robots. http://mobilerobots.com. [2] Negrean, I., Schonstein C., Kacso, K., Negrean, D., Formulations about Dynamics of Mobile Robots, Proceedings of 2010 International Conference on Robotics 23-25 September 2010, ClujNapoca, Romania [3] Negrean, I., Schonstein C., Vuscan, I., Kacso, K., Negrean, C., Dynamics of
[4] [5]
[6]
[7]
[8]
[9]
Hybrid Robot Structures using Variational Principles, Proceedings of 2010 International Conference on Robotics 2325 September 2010 Cluj-Napoca, Romania I., Negrean, Mecanic Avansat n Robotic, Editura UT PRESS, ISBN 978973-662-420-9. Cluj-Napoca, 2008. Schonstein C., Contribuii in dezvoltarea unei structure robotizate hibride, Tez de Doctorat, Universitatea Tehnic din Cluj-Napoca, 2011. Negrean, I., Negrean, D. C., Albeel, D., The Kinematic Control of the Mobile Robots, Proceedings, of the 7th International Conference MTeM, 2005, Cluj-Napoca, ISBN 973-9087-83-3. Negrean, I., Vucan, New Formulation about Dynamics of Robots, Proceedings of AQTR 2008 IEEE-TTTC (THETA 16), International Conference on Automation, Quality and Testing, Robotics, Tome II, pp. 284-289, Cluj-Napoca, Romania, (2008). Nehmzow,U., Mobile Robotics: A practical introduction, second edition, SpringerVerlag, London, 2003. Dudek, G., Jenkin, M., Computational Principles of Mobile Robotics, Cambridge University Press, 2000.
Contribuii cu privire la dinamica robotului mobil de tip PatrolBot Rezumat: n conformitate cu faptul c roboii n timpul operaiunilor efectueaz traiectorii de micare situate n spaiul configuraiilor, sau n spaiul cartezian, se impune un control permanent al parametrilor cinematici, respectiv al forelor generalizate din fiecare cupl motoare, n scopul obinerii unui control precis al parametrilor dinamici. Aadar, scopul lucrrii este determinarea ecuaiilor dinamicii pentru o structura mobil de tip PatrolBot. Ecuaiile de micare vor fi determinate pe baza unor noi concepte n mecanica avansat cum ar principiile variaionale, cu privire la energia acceleraiilor, care reprezint o parte important din cercetrile tiinifice desfurate de ctre autorul principal.
Iuliu NEGREAN Prof. Univ. Dr. Eng., Manager of Mechanical Systems Engineering Department, Technical University of Cluj-Napoca, Department of Mechanical Systems Engineering, iuliu.negrean@mep.utcluj.ro, Office Phone 0264/401616. Claudiu SCHONSTEIN PhD student, Technical University of Cluj-Napoca, Department of Mechanical Systems Engineering, schonstein_claudiu@yahoo.com, Office Phone 0264/401750. Zoltan SZOKE PhD student, Technical University of Cluj-Napoca, Department of Mechanical Systems Engineering, szzoli69@yahoo.com, Office Phone 0264/401750.
599

BENDING CALCULUS FOR CIRCULAR TAPPING PLATES WITH CIRCUMFERENTIAL EXTERIOR CONCENTRATED LOAD BY TRANSFER-MATRIX METHOD
Mihaela SUCIU
Abstract: The bending calculus for the circular tapping plates is very important for its applications in the mechanical, robotic, medical, military and aerospace industries. We present an application for this calculus: a circular tapping plate, embedded at the interior boundary and free at the exterior boundary, charged with a circumferential exterior concentrated load, using the Transfer-Matrix Method. That is an important possibility to result the circular plates, with the opportunity to program this algorithm and to obtain a simply code of this calculus. Key words: tapping plate, circumferential load, state vector, Transfer-Matrix, Diracs function, Heavisides function.
1. INTRODUCTION The bending calculus for the circular tapping plates is very important for its applications in the mechanical, robotic, medical, military and aerospace industries. We present an application for this calculus: a circular tapping plate, embedded at the interior boundary and free at the exterior boundary, charged with a circumferential exterior concentrated load, using the Transfer-Matrix Method. This method is based on the theory of Diracs and Heavisides and the calculus with their functions and operators. After the analytical calculus we obtain the four elements of the exterior circumference state vector and for the interior circumference state vector of a tapping plate. Now, we can know for all the sections r0<r<R, all the elements for the all state vectors. That is an important possibility to result the circular plates, with the opportunity to program this algorithm and to obtain a simply code of this calculus. For our studies, we define a transfer-matrix, that we should integrate in a general manner for the differential equation, witch gives the deformed of some plates loaded with an exterior density q(x,y), written by means of the Diracs and Heavisides functions and
operators. After, we can express, for each case, the boundary conditions for the displacements calculus or for the stresses calculus in any plate points. 2. CALCULUS PREMISES The bending plates theory has some bases: the curvature of the plate varies when the exterior load is perpendicular to the median plan; this curvature variation is made in two plans and they form one surface with double curvature, called elastic surface; the elastic surface form is characterized by the arrow variation low, w(x,y), in Cartesian coordinates; we suppose that the numerical values of the function w(x,y) are very small compared with the plate thickness, h. The some simplifying work assumptions [1] are: we admit that the points aligned on the same normal line at the median surface before deformation remains aligned on the same normal line at the deformed surface after deformation; the normal stresses in the parallel sections at the median plan are negligible compared with the bending stresses; it not exist crush for the plate layers some more than others, besides that local crash under a concentrated load; the sign convention is as follows: the positive arrow is to up and the
600
positive angle is counterclockwise; when the arrow w decreases, the angle is negative and dw is negative too.
dr
r+dr
r dr
3. THE GENERAL CIRCULAR PLATE EQUATIONS For the circular plates loaded with an exterior density q(x,y), the general equations for the deformed of any plate are: 1 (1) w ( x , y ) = q ( x , y )
D is the double Laplacian operator, with the
+d
Fig. 2. Axial section of the plate
developed expression:
4 24 4 = 4 + 2 2 + 4 x x y y
dw dr
(4)
(2) (3)
and:
D = Eh 3 12 1
with E as longitudinal elasticity modulus, is the Poissons coefficient and h is the plate thickness. D is called also plate bending rigidity. The contour integration domain is essential for the integration of the equation (1). It is useless to determine a transfer-matrix for some plates, because the problem hade not explicit solutions. The theoretical solution can be completed only for the field of the circular plates loaded with axially symmetric charges. We can be inferred the equations which allow to calculate the efforts and the deformations in all points of the circular plate, using the Transfer-Matrix Method. We note the arrow w(r) of the circular plate at the distance r from the axis and (r)-the angle around which it spins the normal line (Fig. 1. and Fig. 2.), given by the relation (4):
In the Fig. 2. shows one axial section of the plate and two normal sections at distance between them of r-before deformation and r+dr-after deformation. We have the relative elongation of segment AB at the rate z of the middle fiber:
A' B ' AB = z ( + d ) z = zd
(5)
and the relative radial elongation is:

r = z
d dr
(6)
The point A' is on the circle of radius r+z, and the tangential relative elongation is:
t =
2 z =z r 2 r
(7)
The elasticity theory links the deformations with the radial and tangential stresses by the formulas (8) and (9), ([2] and [3]):
1 ( r = r E = 1 ( t t E
t
) )
(8)
E ( r + t ) r = 1 2 = E ( + ) t t r 1 2
(9)
With (6) and (7), the expression (9) becomes:

Fig. 1. Circular deformed plate
601
Ez d r = 1 2 dr + r d = Ez + t 2 1 r dr
(10)
(Mr + dMr )(r + dr)d Mr rd + q(r) rdr d dr 2 Mt dr d + (T + dT)(r + dr)dr d = 0 (17)

Neglecting the small higher order term, we obtain:
For a sectoral prism element on the plate [1], we write the balance equations. The moments are applied on the element faces. They are taken per unit length after one radial axis. Mr is the radial moment and Mt is the tangential moment. If we known the stresses r and t, is easy to calculate the resultant of these moments on the faces:
h Eh3 d 2 + rd M r rd = rd h r zdz = 2 12 1 dr r (11) 2 h 3 d M dr = dr 2 zdz = Eh + dr h t 2 t dr 12 1 r 2
Mt
d (rMr ) = rT dr
(18)
and replacing Mr and Mt with (13): dres 1 d 2 d rT = r 2 = r D r dr dr dr We note: (19)
or, with the notation (3):
1 rT D
(20)
d M r = D dr + r d M = D + t dr r
We note:
(12)
The differential base equations assess the deformations and the efforts in a circular plate loaded with an axially symmetrical force:
1 d dr = D rq (r ) d res = dr r d (r ) = r res dr dw = dr
1 (Mr + Mt ) res = D(1+ ) and we obtain: d 1 d d 2 1 d (r ) res = + = + = r dr r dr dr 2 r dr
(21)
(13)
(14)
The cut force T is per unit of length on a circumference of radius r. Mres is not a bending moment, but is proportional at a bending moment. The balance equation after the vertical axis for the radial prism element is:
(T + dT)(r + dr)d Trd + q(r) rdr d = 0

where:
(15)
In the equations (21) are inserted four sizes [1]: w and , which have a physic significance-they are the displacements, w is the arrow and is the angle; Mres and , which are: Mres is proportional at a bending moment and is related at the cut force by the relation (20). After integration, the equation system (21) give four constants: wR, R, MresR and R and:
= R q1 (r ) r res = resR R log + q 2 (r ) R
(22)
d (rT) = q(r) r dr
(16)
With writing the sum of the moments in report with the tangential axis of the arc circle of radius r, at the level of median plane, the balance equation for the element is:
with:
q 1 (r ) = q 2 (r ) =
r R r R
D 1
q ( )d q 1 ( )d
(23)
We can write:
602
d (r ) r = resR r R r log + rq 2 (r ) R dr
(24) After integration, we have:

= R
R2 1 R 1 r R2 r 1 r r + resR + + q3 (r ) log R r 2 r R 2r 2 2 r
CIRCUMFERENCE AND FREE AT THE EXTERIOR CIRCUMFERENCE, WITH CIRCUMFERENTIAL CONCENTRATED EXTERIOR LOAD We consider a tapping circular plate, embedded at the interior circumference and free at the exterior circumference (interior charged with a circumferential concentrated exterior load (Fig. 3.).
ro ro
(25) and the arrow w is:

r 2 R2 r 1 r w = wR + R R log + resR R2 log R 2 R 2 r r 2 R2 1 r 2 + R2 + q4 (r) R log + R 2 2 2
(26)
h
4. THE CIRCULAR PLATE AND ITS TRANSFER-MATRIX At the radius r, we define for a circular plate, a state vector with four elements:
Fig. 3. A circular tapping plate embedded at the interior boundary, with a concentrated vertical exterior load.
{V (r )}r = { res (r ), (r ), w(r ), (r )}

The state vector at the radius r0 is:
(27)
After the calculus with Diracs and Heavisides functions and operators, for this charge, the density associated function is:
{V (r0 )}0
= {
res 0 , 0 , w 0 , 0 }
(28)
q(r ) = F (r R)
(33)
and at the radius R is:
We calculate the functions qi(r), i=1,4, for the uniform load at the interior circumference:
q1 (r ) = FR [1Y (r R)] D
FR [1 Y (r R )]log R D r
{V (R )}R
= { resR , R , w R , R }
(29)
(34) (35) (36) (37)
We can write for the radius r:
{V(r)}r = [T]r {V(R)}R +{Ve }r
(30)
q 2 (r ) =
q 3 (r ) =
with: [T]r is the Transfer-Matrix for the circular plate at the radius r:
r 1 0 0 log R 0 1 0 0 2 2 r R r r r 1 1 [T ]r = R2 log r 2 + R2 log + R2 r 2 1 R log 2 2 R 4 R R 2 2 R r R r R 1 1 r r log + 1 r R 2r 2 r 2 2
FR R R 2 r 2 1 + log [1 Y (r R)] 4D r
q4 (r) =
FR R 2 [1Y(r R)] R2 r2 R0 + r2 log 4D r
For r=r0, the matrix relation (33) becomes:
{V(r0 )}0 =[T]r {V(R)}R +{Ve}r

0
(38)
0 0 0 r 1 Rlog 0 R R 1 r0 0
{Ve }r is the vector for the exterior loads:

1 {Ve}r = q2 (r ),q1(r ), q4 (r ), q3 (r ) r
T
(31)
with:
r 1 log 0 R 0 1 2 2 r R r 1 2 2 r0 [T]r0 = 1 0 r0 + R log + R2 r02 R2 log 0 2 R 4 R 2 2 2 r r 1 1 R R r0 r0 log 0 + 0 2 2 R 2r0 2 r0
(32)
5. APPLICATION: TAPPING CIRCULAR PLATE EMBEDDED AT THE INTERIOR
(39)
603
We put the boundary conditions in the relation (38): - for the interior circumference embedded (r=r0): wr0=0 r0=0 - for the exterior free circumference (r=R): R=0 and: Mr(R) =0 For the matrix equation (38) we obtain: (43) (42) (40) (41)
[1] Gery, M,. Calgaro, J.-A., Les matricestransfert dans le calcul des structures, Editions Eyrolles, Paris, 1973. [2] Tripa, M., Reziztenta materialelor, Editura Didactica si Pedagogica, Bucuresti, 1967. [3] Warren, C. Y., ROARKS Formulas for Stress & Strain, 6-th Edition, McGrawHill Book Company, 1989.
0 M res0 0 0 Fr0 0 = [T ]r0 + D wR 0 0 0 R 0

0
(44)
with [T ]r given by (39) and qi, i=1,4 is given by (34), (35), (36) and (37). We have a linear equations system, with four unknowns. After solving the system we obtain the four solutions for wR, R, Mres0 and 0. Now, we can calculate in all sections of the tapping plate, between r=r0 and r=R the four elements for all the state vectors. 6. CONCLUSIONS This work presents another method for an application of the tapping circular plate calculus, using the Transfer-Matrix Method, embedded at the interior circumference, free at the exterior circumference, loaded with a concentrated vertical charge at the free exterior boundary. The calculus with the TransferMatrix Method is very easy to program and so, that is very important for optimization calculus, function of different optimization criteria, with applications in the mechanical, robotic, medical, military and aerospace industries. 7. REFERRING
604
Calculul la incovoiere al placilor circulare gaurite cu forta circumferentiala exterioara concentrata prin Metoda Matricelor de Transfer
Rezumat: Calculul la incovoiere al placilor circulare gaurite este foarte important pentru aplicatiile sale in industria mecanica, robotica, medicala, militara si aerospatiala. Se prezinta o aplicatie a acestui calcul: o placa circulara gaurita, incastrata pe circumferinta interioara si libera pe circumferinta exterioara, incarcata cu o forta circumferentiala exterioara concentrata, utilizand Metoda Matricelor de Transfer. Aceasta este o importanta posibilitate de a rezolva calculul placilor circulare, cu oportunitatea de a programa acest algoritm si a obtine un cod simplu pentru acest calcul.
SUCIU Mihaela, PhD. Dr.-Eng., Professor, Technical University of Cluj-Napoca, Mechanical Engineering Department, Mechanical Faculty, E-mail: mihaelaica2007@yahoo.fr, Office Phone: 00.40.264.401.752., Home Address: Bd. 21 Decembrie 1989 nr. 23-35, App.14, 400105-Cluj-Napoca, Romania, Home Phone: 00.40.264.591.397., Mobile Phone: 00.40.723.878.112.
605

INFLUENCE OF THE GEOMETRIC FORM OF THE PLATE UPON THE DIRECT EXTRUSION FORCE FOR THE NON-FERROUS METALS
Gheorghe GLIGOR, Sorin GROZAV, Radu MORARIU-GLIGOR
Abstract: In the paper there are presented some sorts of parts in lead, aluminum and copper obtained by direct cold extrusion. In the second part of the work is presented by means of tables and diagrams the force of the parts in lead, aluminum and copper obtained by direct extrusion with more sorts of active plates. Key words: force, lead, aluminum, cooper, direct cold extrusion.
1. GENERAL CONSIDERATION The geometric form of the active plate, that is the angle and the rounding radius in the deforming area have a significant influence upon the technological parameters of the extruding process. Using one set of active plates can get a significant change in the way of the material flow, namely it can be obtained a radial flow, which appear inside the plastic zone. In case of active plates with radius in the deformation area one can see that this leads to a diminution of the size of the dead zone. An increase of the angle leads to a decrease of the dissipated power and of extruded power. In order to make evident the influence of the geometric form of the active plate upon the deformation force, the experimental researches were directed as follows: the influence of the angle of the active plate and the influence of the radius in the deforming zone. For the influence of the angle of the active plate upon the deforming force were used the followings: - half-finished materials of lead 99.5, aluminum 99.5 and copper 99.9, with the following dimensions: - initial diameter: D0 = 24,5mm;
- final diameter: Df = 15mm; 10mm; - length: Ls = 50mm; 30mm; - one set of active plates with =120; 105; 90; 60; 45; - rounding radius of the plane active plate: r=0; r=2; r=4; - deforming degree: =62,51%; =83,34%; - extruding temperature: 20C; - extruding speed: v=100 mm/min; - the INSTRON 1196 of 250 MN Testing machine and the Hydraulic press of 100 ft. The extruded tests in order to measure the force are presented in figures 1, 2 and 3. At the combined extrusion the degree of deformation was different =83.34% (D0=24,5mm and Df=10mm) and =62,51% (D0=24,5mm and Df=15mm). The active plates, which have been used, have different half angle : convex active plate (=120; 105); plane active plate (=90 and the rounding radius r=0; 2; 4 mm); conic active plate (=60; 45).
Fig. 1. Parts in aluminium directly extruded with active plates: a-convex; b-plane; c-conic; =62,51%
606 20 25 30 35 40 45 50 5370 5180 5130 5100 5000 4500 5700 5250 5150 5050 4850 4700 4200 5450 5450 5300 5200 5150 5100 4850 6500 5400 5300 5150 5100 5000 4800 6150 5450 5350 5150 5100 5050 5000 5750
For smaller deforming degree =62,51%, as one can see in Table 2 and Fig. 5, the size of extrusion force for all metals depends on the angle, namely the force is smaller for smaller angles.
a b c
Table 2 Force for direct extrusion of lead, aluminum and copper [N] Type of Maxi force [N] extrusion mum Lead Alumi Copper num ED 15x120x30 2600 22400 56500 ED 15x105x30 2500 22350 56000 ED 15x90r0x30 2900 22150 55800 ED 15x90r2x30 3150 19625 53000 ED 15x90r4x30 3200 19400 51600 ED 15x60x30 2900 19250 50650 ED 15x45x30 2300 16000 38500
Fig. 3. Parts in copper directly extruded with active plates: a-convex; b-plane; c-conic;
2. VARIATION OF FORCE NECESSARY FOR DIRECT COLD EXTRUSION OF NON-FERROUS METALS On the basis of the value obtained and synthesized in the Table 1 there were designed the variation curves of the force depending on the stroke of every angle of the active plate. From Table 1 and Fig. 4 one can see that the size of the force depends on the value of the angle. The greatest value of the extruding force results for =90 and the smallest values are for the conical active plates with =45 and 60.
Table 1 Force for direct extrusion of lead [N], ED Pb 10x50 Angle Stroke mm 0 5 10 15 =45 0 1050 5250 5400 =60 0 1000 5450 5400 of active =90 0 3000 5550 5500 plate [] =105 0 1100 5350 5600
60000
Variation of the force for the direct extrusion
lead aluminium copper
50000
40000
Force [N]
30000
20000 8 10000 0
mate 2 ri 1-lea al: d 2-alu ; miniu 3-cop m per. ;
2 1 0 1 0
=120 0 1000 5400 5650
Fig. 5.Variation of the force for the direct extrusion of non-ferrous metals: =62,51%
In order to make evident the influence of radius in the deforming area upon the force for
A 1- ngle 2- 120 of 1 ac 3- 05 ; tiv 4- 90r ; e pla 5- 90r 0; te 2 9 6- 0r ; : 4 6 7- 0; ; 45 .

6 5 4 3
607
direct extrusion it was used a set of active plates = 90 with the following connecting radius: r =0; 2; 4 mm. The tests were made for lead, aluminum and copper and part of the results have been summarized in Table 3. In Fig. 6 it is presented the variation of force depending on the stroke of the male die for every type of active plate. From this figure one can see that the smallest extruding force is got for the biggest radius, and the biggest force was registered for the flat active place without radius.
Table 3 Force [N] for the direct extrusion with flat plates
From the experimental researches which were made it was found that the extruding force varies for various deforming degrees, namely the more increased the deforming degree is, the more increased the deforming force is. For the direct extrusion of lead with different deforming degrees the results have been summarized in Table 4 and graphically represented in Fig. 7.
Table 4 Force for direct extrusion of lead with different deforming degrees Stroke Force [N] [mm] 0 5 10 15 20 25 30 35 40 45 50 = 62,51% 0 1000 5450 5400 5250 5150 5050 4850 4700 4200 5450 =83,34% 0 1890 1900 1880 1850 1820 1800 1770 1740 1730 1900
Stroke [mm] 0 5 10 15 20 25 30 35 40 45 50
Rounding r=0mm 0 3000 5550 5500 5450 5300 5200 5150 5100 4850 6500
radius of the r=2mm 0 3250 5500 5500 5400 5350 5200 5150 5000 4800 5800
active plate r=4mm 0 3500 5350 5350 5300 5200 5050 4900 4800 4700 6100
6000 5000
Force for direct extrusion ED Pb 10-15x60x50
Force for direct extrusion with plane plate

7000 6000
Force [N]
4000
3000
5000 4000 3000 2000
2000
Force [N]
ED Pb 10x90r0 ED Pb 10x90r2 ED Pb 10x90r4
1000
ED.Pb-10x60x50 ED.Pb 15x60x50
10
20
30
40
50
60
1000 0
Lenght of extruded semi-finised material [mm]
Fig. 7. Variation of the force for direct extrusion of lead with different deforming degrees
0 10 20 30 40 50 60
Stroke of die [m m ]
Fig. 6. Variation of the force for the direct extrusion with flat plates [N]
608
3. CONCLUSION 4. REFERENCES Further to the analysis of the influence factors and of force variation one can draw the following conclusion: - the size of extruding force both for lead and for aluminium and copper depends on the value of angle. The greatest value of the extruding force is registered for =90 and the smallest ones are for the conical active plates with =45 and 60; - the force for the direct extrusion with flat active plates =90 with connecting radius in the deforming area is the smallest for the biggest radius, and the greatest force was registered for the flat active plate without radius; - the extruding force varies for different deforming degrees, namely the more increased the deforming degree is, the more increased the extruding force is. [1] Gligor, Gh., Researches regarding the
workability of the non-ferrous metals by cold extrusion. Doctoral dissertation. The Technical University, Cluj - Napoca, 273 p; 2003. Gligor, Gh., The present stage of cold extrusion technologies. Report No. 1 in Doctorate Work Technical University, Cluj-Napoca, 72 p; 2001. Groza, I and others. The plastic deformation of metals and non-ferrous alloys. Technical Publishing House, Bucharest, p. 251. 1977. Solomon, N., Studies and researches concerning the deformation process by extrusion of special alloys of aluminum. Summary of the Doctorate Work, 57 p. 1997. Tapalaga, I., Berce, P. & Achimas, Gh., Metals cold extrusion, Cluj-Napoca, Dacia Ed., 344 p. 1986.
[2] [3] [4]
[5]
Influenta formei geometrice a placii active asupra fortei la extrudarea directa a metalelor neferoase Rezumat: n aceast lucrare sunt prezentate cteva piese din plumb i aluminiu, obinute prin extrudare directa la rece. n partea a doua a lucrrii sunt prezentate tabelar i n diagrame msurtorile fortei necesare pentru piesele obinute prin extrudare directa la rece a metalelor neferoase cu diferite tipuri de pci active.
Gheorghe GLIGOR Assoc. Prof. Dr. Eng. Technical University of Cluj-Napoca, Cluj-Napoca, Bdul Muncii 103-105, Romania, Department of Manufacturing Engineering, Faculty of Machine Building, ghgligor@ tcm.utcluj.ro, Str. Timuului 14, Cluj-Napoca, 40-364-734829. Sorin Grozav Prof. Dr. Eng. Technical University of Cluj-Napoca, Cluj-Napoca, B-dul Muncii 103-105, Romania, Department of Manufacturing Engineering, Faculty of Machine Building, sgrozav@tcm.utcluj.ro, Str. Vaida Voivod Alexandru 66, Cluj-Napoca, 40-264-411664. Radu Morariu-Gligor Senior Lecturer Dr. Eng. Technical University of Cluj-Napoca, ClujNapoca, B-dul Muncii 103-105, Romania, Department of Mechanics and Computer Programming, Faculty of Machine Building, rmogli70@yahoo.com, Str. Arieului102/107, Cluj-Napoca, 40-264-414380.
609

MATHEMATICAL SIMULATION OF METAL SHEET BENDING WITH ELASTOMERIC PADS

Liana HANCU, Sorin COMA, Horaiu IANCU, Rare GDLEAN, Rzvan PRUN
Abstract: This paper presents a mathematical model and a numerical simulation for metal sheet bending with elastomeric pads made of plasticized polyvinyl chloride. Compression tests were performed in order to determine some characteristics for the materials and the results were used in the simulation program. A correction for the geometry of the elastomeric pad was proposed and the results for the simulation are also presented. Key words: mathematical model, simulation, elastomer, bending, compression.
1. INTRODUCTION It was initially believed that elastomers were a part of the rubber group, due to their similar properties. In the 80s, this opinion has changed so that the term of elastomers also included natural or synthetic rubbers and not vice versa. An elastomer is defined as a macromolecular material which at room temperature is capable to return substantially to its original shape and size after the tension removal. A rubber is defined as a macromolecular material which is able to return in one minute to less than 1.5 of its initial length after it had been stretched twice than its initial length and maintained so for one minute. Therefore, considering these definitions, we can say that all the rubbers are elastomers, but not all the elastomers are rubbers. The group of elastomers does not include the following materials: natural rubber, synthetic rubber and thermoplastic elastomers. The use of elastomers in the design of the sheet metal forming tools has evolved in the last years, because parts of different shapes and sizes can be achieved using the same active elements. Sheet bending in dies with elastomeric pads is a modern process used often in manufacturing because it offers a wide range of advantages as compared to the classic processes [3]. The precision of the parts is
higher than that obtained by using classic procedures because the elastic material of the pad responds with an opposite force that will create a calibration pressure on the part. The price of the tool is lower because the elastic pad can be used for different parts and the preparation time of the workplace is shorter. 2. CONSTITUTIVE MODEL OF THE ELASTOMERIC MATERIALS The constitutive model of elastomeric materials is based on the assumption that their mechanical response can be described using the complementary strain potential as a state function. This parameter generally depends on the components of the stress tensor:
U = U ( ij | i, j = 1, 2, 3)
(1)
Once admitted the existence of the scalar potential the strain components can be computed as partial derivatives with respect to the corresponding variables ij :
ij =
U , ij
i, j = 1, 2, 3
(2)
Equation (2) represents the general formulation of a hyper-elastic constitutive
610
model. In practice, the only problem consists in adopting a convenient expression of the strain potential. In the particular case of the rubber-like materials, this expression should be compatible with two characteristics of their mechanical behaviour: a) Isotropy b) Isochoric character of the strain field. Isotropy represents the invariance of the mechanical response with respect to the rotation of the coordinate frame. The isochoric character of the strain field ensures the volume conservation even if large distortions occur (this is the typical behaviour of the elastomeric materials). In order to fulfil the requirements mentioned above, the strain potential can be expressed as follows:
U = U (I 2, I 3 )
12 13 11 22 23 I 3 = 21 32 33 31
(7)
In this paper, the dependence of the strain potential U on the third invariant is considered negligible. As a consequence, U will be defined as:
U = U (I 2 ) .
(8)
For a better agreement with the experimental data, this dependence will be expressed as a polynomial:
2 U = a 0 + a1 I 2 + a 2 I 2 +K+ n k + an I 2 = ak I 2 k =0 n
(9)
(3)
where I2 and I3 are the second and the third invariants of the deviatoric stress tensor
= ij ij kk ij
k =1
1 3
(4)
The first invariant is not included in Eq. (3) because it is always zero. The Kronecker symbol has been used in Eq. (4):
1, ij = 0, if = j, if i j.
The polynomial order n is chosen according to the volume of experimental data and the desired accuracy of the constitutive model. Generally, the flexibility of the model is increased when higher values of the exponent n are chosen. The coefficients a0 ,K, a n are material constants and result from an identification procedure. In general, the strain potential U must be zero when the second invariant I2 vanishes. This constraint is fulfilled only if:
a0 = 0
(5)
(10)
The second and the third invariants of the deviatoric stress tensor are defined in the following manner:
12 23 I 2 = 11 22 33 22 21 32 31 33 22 22 33 = 11 11 13
Under such circumstances, Eq. (9) becomes:

2 n k (11) U = a1 I 2 + a 2 I 2 + K + an I 2 = ak I 2 k =1 n
(6)
After replacing the expression of U given by Eq. (11) in Eq. (2), the following of constitutive equations are obtained:
11 + 12 21 + 23 32 + 31 13 33
ij =
where:
U dU I 2 = , ij dI 2 ij
i, j = 1,2,3, (12)
611
dU = a1 + 2a 2 I 2 + K + dI 2 + na I and I 2 , = ij ij i, j = 1,2,3.
n 1 n 2
= ka k I
k =1
(13)
k 1 2
2 1 = 33 = 11 , 11 = 11 < 0, 22 3 3 0 , , 1 , 2 , 3 , = i j = i j ij (18) Under these conditions, the second invariant I2 becomes (see Eq. (6)):
(14)
1 2 I 2 = 11 3
(19)
By combining Eqs. (12) (14), the constitutive equations can be put into the explicit form:
k 1 ka k I 2 ij = ij , k =1 n
i, j = 1,2,3. (15)
Because in the particular case of the axial compression test, 11 is the only stress component different from zero, the constitutive model defined by the equation (15) reduces to the following relationship (see also Eq. (19)):
k 1 ka k I 2 = 11 = 11 k =1 n
The main steps of an identification procedures based on the results of axial compression tests will be presented below. The experimental tests are made on cylindrical elastomeric samples. If A0 is the of crosssection area and h0 is the height of the undistorted sample, while A and h are the crosssection area and the height of the same sample at a certain moment during the compression test, the condition of volume conservation can be expressed as follows: A0 h0 = Ah (16)
2(k 1) 2 = 11 ka k 11k 1 = 3 3 k =1
n
(20)
= 2 ka k
k =1
2 k 1 11
3k
The raw data provided by the axial compression test is represented by a table of values [hl , Fl | l = 1,K, N ] , where: hl = h0 hl (21)
Considering that the direction of compression represents the axis 1 of the coordinate frame, the strain components can be evaluated using the formulas:
h 11 = ln h 0 1 1 h 1 A0 (17) = ln 22 = 33 = 11 = ln h A 2 2 2 0 i, j = 1,2,3, i j ij = 0,
is a current variation of the height and Fl represents the associated axial force acting on the specimen. These raw values can be easily transformed into pairs of strain and stress components. More precisely, each pair [hl , Fl ] is converted into a pair [ 11,l , 11,l ] using the formulas:
11,l = ln
hl hl h hl = ln 0 = ln 1 h h0 h0 0
, (22)
The components of the deviatoric stress tensor also have the following expressions:
11,l =
Fl F h = l l = Al A0 h0
F h hl F hl 1 = l 0 = l A0 h0 A0 h0
(23)
612
The identification procedure used in this paper is based on the least-square method. The coefficients a1 , K , a n are evaluated by minimizing the error-function:
Tijkl =
ij kl
, i, j , k , l = 1,2,3.
(27)
The quantities Tijkl (i, j , k , l = 1,2,3) define the
11,l n F (a1 , K , a n ) = 2 ka k k 11,l (24) 3 l =1 k =1

N
relationship between the perturbations of the strain components and the perturbation of the corresponding stresses:
D ij = Tijkl D kl , i, j = 1,2,3. (28)
k =1 l =1 3 3
The minimum constraints:

F = 0, a m m = 1, K , n,
(25)
generate the following set of linear algebraic equations having the coefficients a1,...,an as unknowns:
k N 2(m+ k 1) 1 N 2 m 1 a = k 11,l k 2 11,l 11,l , (26) l =1 k =1 3 l =1 m = 1,K, n.
n
Both the stress components and the tangent elastic modulus associated to the current moment of the analysis are calculated using the constitutive equations presented in the previous section of the paper: equations that define the relationship between the deviatoric stress and strain components: = S ij , i, j = 1,2,3 ij where:
k 1 k 1 S = ka k I 2 = a1 + ka k I 2 k =1 k =2 n n
(29)
Due to its linearity, the solution of this set is straightforward (the authors have used Gauss elimination procedure).
3. IMPLEMENTATION OF THE CONSTITUTIVE MODEL IN THE FINITE-ELEMENT PROGRAMME ABAQUS/STANDARD
(30)
I2 =
1 3 3 pq pq 2 p =1 q =1
(31)
In order to prove its applicability and performances, the constitutive model presented above has been implemented as a UMAT routine in the finite-element programme ABAQUS/Standard [1],[2]. In general, the UMAT routine must perform two specific tasks: Using the strain components ij ( i, j = 1, 2,3) associated to the current moment of the analysis as input data, the UMAT routine must evaluate the corresponding stress components ij (i, j = 1,2,3) ;
equation that defines the relationship between the hydrostatic components of the strain and stress tensor:
(32) where > 0 is a very large penalty constant.

The determination of the current stress state consists in the solution of the nonlinear set , i, j = 1,2,3 as given by Eq. (29) considering ij unknown quantities. Any appropriate numerical procedure can be used for this purpose (e.g., Newtons method). The authors have used the HYBRD1 routine from the mathematical library MINPACK-1. These routine implements a solution strategy based on Newtons method. As soon as the components of the deviatoric stress are determined, Eq. (32) provides the
For the same moment of analysis, the UMAT routine must also evaluate the components of the so-called tangent elastic modulus,
hydrostatic part . Finally, the current stresses are calculated using the additive decomposition: + ij , i, j = 1,2,3 ij = ij (33)
613
After combining Eqs. (35) (42), the following relationship is obtained: 1 [ A] + S [B ] + {D } = (43) {D }, + S { }{ }T [C ] The following matrix notations have been used in Eq. (43)
The evaluation of the tangent elastic modulus starts from the additive decomposition of the strain tensor: + ij , i, j = 1,2,3 ij = ij (34)
{D } = [D 11 , D 22 , D 33 ,2 D 12 ,2 D 23 ,2 D 31 ]T ,
(44)
By perturbing this decomposition, one obtains the relationship: + ij D , i, j = 1,2,3 D ij = D ij (35)
{D } = [D 11 , D 22 , D 33 , D 12 , D 23 , D 31 ]T ,
(45)
, 22 , 33 , 12 , 23 , 31 ]T , { } = [ 11
(46)
The strain perturbations in the right-hand side of the above formula result from the constitutive equations (29) (32). More precisely, the perturbation of equations (29), (30) and (31) leads to the formula: + ij DS , D ij = SD ij where,
DS = S DI 2 ,
k 2 S = k (k 1)a k I 2 , k =2 n
(36)
1 1 1 1 [A] = 3 0 0 0
1 1 1 0 0 0
1 1 1 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 , 0 0 0
0 0 0 0 2 3 0
(47)
(37)
(38)
DI 2 = pq D pq ,
p =1 q =1
(39) (40) (41)
= D ij ij D , D ij 1 3 3 D = rs D rs . 3 r =1 s =1
2 1 1 0 1 2 1 0 1 1 2 0 2 [B] = 1 0 0 0 3 3 0 0 0 0 0 0 0 0
0 0 0 0, (48) 0 2 3
In the same way, the perturbation of the hydrostatic stress is obtained from the constitutive equation (32):
D = 1
(42)
2 1 1 0 0 0 1 2 1 0 0 0 1 1 2 0 0 0 1 [C ] = 1 0 0 0 3 0 0. 3 0 0 0 0 1 0 3 1 0 0 0 0 0 3
(49)
614
Eq. (28) is equivalent with the matrix relationship:
{D } = [T ]{D }.
(50)
By comparing Eqs. (43) and (50), one may easily notice that:
machine (Figure 1) has been converted into a diagram relating the sectional stress and the axial logarithmic strain. In the next stage, the experimental data resulted from the conversion have been used for calibrating the constitutive model of the elastomeric material.
1 T [T ] = [A] + S [B ] + S { }{ } [C ]
1600
PVC+30% plasticizer
(51)
1200
Eq. (51) defines the tangent elastic modulus of the elastomeric material in accordance with the constitutive equations (29) (32). The calculation of this matrix represents the second task that should be performed by the UMAT routine.
4. NUMERICAL SIMULATION OF A SHEET METAL BENDING PROCESS WITH ELASTOMERIC PAD
F [N]
800
400
0 0 1 2 3 4 5
h [mm]
Fig.1. Compression curve for PVC with 30% plasticizer
The capabilities of the constitutive model presented above have been tested by numerical simulation of a sheet metal bending process with elastomeric pad. The calibration of the material model (defined in the UMAT routine) is made using the experimental data provided by a standard compression test, which allows the determination of the dependence between the height variation of the specimen and the compression force. The material coefficients resulted from the identification procedure presented in the previous section have been passed as input data to the finite element ABAQUS. The material of the deformable pad is polyvinyl chloride (PVC). Its elastic characteristics were obtained by introducing the plasticizer dimethyl phthalate, which is the methyl ester of phthalic acid. The specimens used in the compression test have been prepared according to the prescriptions of the standard SR ISO 815A1:1995, in the shape of circular disks with 29,0 0,5mm diameter and 13,0 0,2 mm thickness. These samples were subjected to compression on the universal testing machine INSTRON 1196. The axial force (F) vs. height variation (h) diagram recorded by the testing
The metallic sheet used in the numerical simulation is aluminium AA 1050-O. The mechanical behaviour of this material has been described using the elastoplastic constitutive model defined in ABAQUS/Standard. The following material parameters characterizing the elastic / plastic behaviour of the AA 1050-O sheet metal have been adopted [5]: Youngs modulus: 69 GPa Poissons ratio: 0.33 Conventional yield stress: 30 MPa Ultimate strength: 50 MPa Total elongation: 40 %. The dimensions of the blank subjected to bending are as follows: thickness 1mm, length 70 mm and width 55 mm. For the U bending simulation, a punch with 10 mm radius was used. The dimensions of the deformable pad are the following ones: length 150 mm, height 110 mm and width 55 mm. After repeated simulations, it was established that the elastomeric pad should have the geometry presented in Figure 2. This geometry is able to achieve a full bending of the metallic blank around the punch profile. The finite-element simulation provided numerical values of the stresses and strains in the bent regions of the blank, as well as in the pad. The values were obtained for different stages of the bending process. Figure 3 presents
615
the distribution of the von Mises equivalent stress associated to the final stage of the forming process.
deformable pad with the following composition: PVC 70% and 30% plasticizer, and a punch having the fillet radius R = 10 mm. One may notice the accelerated evolution of the force in the second stage of the forming process, when the metallic sheet is fully embossed in the pocket of the elastomeric pad.
16000
12000
numerical simulation
8000
F [N]
Fig. 2. Proposed geometry of the elastomeric pad
4000 R=10mm U bending
Table 1 presents the minimum and maximum values of the von Mises equivalent stress () in the elastomeric pad and in the bent part. These stresses have maximum values when the punch stroke also has the maximum value.
Table 1 Von Mises equivalent stress () Bent part Equivalent stress [N/mm ]
2
10
12
14
punch stroke [mm]
Fig. 4. Variation of the bending force during the bending process (punch radius R=10 mm)
5. CONCLUSIONS
Minimum +2,310-5
1
Maximum +5,510
Elastomeric pad Equivalent stress [N/mm ]

2
Maximum +6,8100
Minimum +2,9710-5
Fig. 3. Distribution of the equivalent stress
The diagram from Figure 4 shows the variation of the bending force as resulted from the finite element simulation, in the case of a
Using the finite-element programme ABAQUS, the simulation of the bending with an elastomeric pad is possible. The characteristics of the sheet material must be determined, as well as those of the elastomeric material used for the pad. By introducing this data in the program, the numerical analysis of the bending process can be performed. Unlike other theoretical approaches found in published papers worldwide, the constitutive model developed by the authors has the advantage of simplicity and also the advantage of the possibility of identifying the material parameters using only the experimental results provided by standard compression tests. The proposal of a plasticizer polyvinyl chloride as a material for the deformable pad is a challenge in this area, if taking into account that PVC is a thermoplastic material which is friendlier with the environment than rubber. It can be also observed after the simulation, that this material seems to have the same advantages for sheet metal bending as compared to the classical technology.
616
6. REFERENCES
[1] *** Abaqus Analysis Users Manual (release 6.7). Electronic Documentation. Dassault Systmes, 2007. [2] *** Abaqus Theory Manual (release 6.7). Electronic Documentation. Dassault Systmes, 2007. [3] A. von Ende, Untersuchnungen zum Biegemformen mit elastischer Matrize, Diss. Universitat Erlangen Nurnberg, Munchen, Wien, Hanser, 1991 [4] N. Panc, L. Hancu, N. Balc, Research Regarding the Improvement of the Performance of Rubber Dies, Proceedings of
the World Congress on Engineering, 2011, Vol I, WCE 2011, p. 783-786 , ISBN: 978988-18210-6-5; ISSN: 2078-0958 (Print); ISSN: 2078-0966 [5] Colombi, M. Matriaux mtalliques. Paris, Dunod, 2008 (ISBN 978-2-10-053918-5)
ACKNOWLEDGMENT
This work has been supported by the Grant PN II CNCSIS IDEI 205 (contract number 655/2009)
SIMULAREA MATEMATICA A INDOIRII TABLELOR METALICE CU PLACI ACTIVE DIN ELASTOMERI Rezumat: In lucrare se prezint un model matematic i simularea numeric a indoirii tablelor metalice utiliznd plci active elastice realizate din policlorur de vininil plastifiat. Teste de compresiune au fost realizate pentru determinarea unor caracteristici de material iar rezultatele au fost utilizate n programul de simulare. In final s-a propus o geometrie a plcii din elastomer iar rezultatetele simulrii sunt de asemenea prezentate.
Liana Hancu, PhD Eng., Professor, Technical University of Cluj-Napoca, Manufacturing Department, Liana.Hancu@tcm.utcluj.ro, Office Phone: 0040264401710, Home Phone: 0040745780603, Cluj-Napoca, Romania. Sorin Coma, PhD Eng., Reader, Technical University of Cluj-Napoca, Manufacturing Department, dscomsa@tcm.utcluj.ro, Office Phone: 0040264401731, Cluj-Napoca, Romania. Horaiu Iancu, PhD Eng., Professor, Technical University of Cluj-Napoca, Manufacturing Department, Iancau.Horatiu@tcm.utcluj.ro, Office Phone 0040264401709, Home Phone 0040741924231, Cluj-Napoca, Romania. Rare Gdlean, Engineer, SC Napomar SA Cluj, grnytro@yahoo.com, Office Phone: 0040264401710, Home Phone: 0040742645443, Cluj-Napoca, Romania. Rzvan Prun, PhD. Student, Technical University of Cluj-Napoca, Manufacturing Department, Razvan.Pruna@tcm.utcluj.ro , Home Phone: 0040745773098, Cluj-Napoca, Romania.
617

USING THE SPHERE AND THE REVOLVING CONE AS GEOMETRIC LOCI FOR GRAPHICAL CONSTRUCTIONS
Magdalena ORBAN, Cecilia CIORTEA, Diana BOCA
Abstract: The graphical solving of some problems in spatial geometry is facilitated by the use of some auxiliary surfaces, as geometric loci, which comply with the requirements of the problem. The paper proposes some ways of utilizing sphere and revolving cone, as auxiliary surfaces, for graphically determining the positions occupied in space by some geometrical elements, materialized by parts in machine building, of distances between them or of angles they form or have to form with other elements, within the requirements imposed by some concrete design problems. Key words: geometrical loci, sphere, revolving cone
1. INTRODUCTION The use of some geometric loci for graphical solving of some problems in spatial geometry, such as problems of distances and angles, most often leads to the simplification and shortening the time of designing work [1]. Among these, the frequently used in graphical constructions are: a) the sphere as locus of the points equally distanced from a given point; b) the circular straight cylinder, as a locus of the points equally distanced from a given line; c) the revolving cone as locus of the lines intersecting in the same point and which meet a plane at a given angle; d) the bisector plane of the dihedral angle formed by the two given planes. Knowing that, the locus of the points equidistant from a plane, a sphere, a cylinder or a revolving cone is, as the case, a system of two parallel planes to the first, two concentric spheres with that considered, two cylinders and two cone having the same axis as the considered surface, it results that the problems may have one or more solutions [2]. In the present paper there are analyzed some modalities of using the sphere and revolving cone, as geometric loci, for graphical determination, on the basis of some initially
known data, of the positions in space of some geometric elements, materialized by parts in machine building, as well as of some distances or angles needed in designing process. The geometric elements are assimilated with vertices, edges and plane faces, respectively, of some machine parts, exemplified within the paper. 2. RESULTS AND DISCUSSIONS 2.1. The auxiliary surface is the spherical surface A. Determining the vertex coordinates of some prismatic parts, lying on axis Ox, at a given distance from the known point The vertex of part from Figure 1, looked for, belongs to axis Ox and to a sphere described by a given point A(a,a) considered as center, with radius equal to the given distance. The level plane [N] drawn through a determines one of the big circles of the sphere whose projection is obtained describing the circle from point a as center, with radius l. The circle, thus drawn, cuts axis Ox in M and N. MN is the diameter of the section in the sphere effected with the vertical plane [V]. For its determination it is necessary to draw the circle with MA as radius. Points B and C are the solutions of the problem.
618
n' x nC
a' D
m' N' Bm l O
a)
A x C B O
and with center in the middle of this segment. The straight line D(d,d) meets the sphere in points M(m,m) and M1(m1,m1) which are the solutions of the problem, as each of them, together with points A and B, determines a big circle in the sphere. In order to intersect line D with the sphere, the auxiliary plane [P], drawn through line D, is used and a rotation of this plane and line is effected to the level plane [N] which passes through the center of the sphere. By this rotation, m0 and m10 are obtained which then, by elevation of rotation, become M(m,m) and M1(m1,m1) respectively. C. Determining a vertex lying in a given plane, at known distances with respect to two points The point looked for will be placed, in the same time, on a sphere with center A(a,a) and radius r, on a sphere with center B(b,b) and radius k and on the given plane [P] (Fig. 3). In each sphere one determines the section effected by plane [P], and points M(m,m) and N(n,n) common to the two circles thus obtained, are the solutions of the problem.
P'
b)
Fig. 1. Points lying on Ox, at distance l to A(a,a).
The actual size of segment AB(ab,ab) results by transformation of AB AM [H], [3]. B. Determining the vertex coordinates of a prismatic part with base a right angled triangle There are known the vertexes A(a,a), B(b,b) of the base with the right angle in point M looked for (Fig.2).
c' m' a' x c
l
'
b' m1' d' b1'
N'
a'
k b' m'
n'
e'
N' O
x y z
a m m0
m 1 d0 b m10
b1 d P e
F
r
n b a P m
z
O Px
c0
a)
a)
b)
Fig. 2. Determining the vertex of a triangular prism.
One determines the actual size of segment AB and draws the sphere with diameter 2R= AB
b)
Fig. 3. Point lying in a given plane, at known distances with respect to two points.
619
2.2. The auxiliary surface is the conical surface A. Determining a straight line which passes through a point S contained in plane [P] forming an angle with plane [H] The auxiliary surface employed is a revolving cone with vertex in point S(s,s) whose generatrices form angle with plane [H] (Fig. 4). SA(sa,sa) is the frontal generatrix of the cone. Plane [P] cuts the cone through lines SM1 or SM2. The condition that the problem be possible is that sa sm1 or a m1, respectively, ctg ctg or . Thus, it follows that: a) the problem has two solutions for < ; b) a single solution for = and c) is impossible for > .
i' P' v' v
the horizontal circle with diameter mn = mn. The two circles have points (,) and (,) common, which together with point V(v,v) determine the required generatrices.
3' r' x v' 2' s' O
' ' n' m' 1' 4' 3 2 s r v n m 4 1
a)
b)
Fig. 5. The line forming known angles with planes [H] and [V].
P1'
i 1' xP x h' 2
s'
O1
' i i1
The problem admits for + < /2, four solutions, for + = /2, two solutions, and no solution for + > /2.
m' 1 a'O m1 a
h2
<
x1
P x1
P P1 h1
a)
b)
Fig. 4. Line forming a given angle with plane [H].
B. Determining a straight line passing through a given point forming known angles with planes [H] and [V] The line looked for is the common generatrix of two revolving cones with the common vertex in the given point, whose generatrices form angles and with the two projection planes (Fig. 5). In order to intersect the two cones, a sphere with variable radius with center in the given point V(v,v) is used. The sphere intersects the cone which has the end axis through the frontal circle with diameter rs = rs and the vertical cone through
C. Determining a plane containing the given point, inclined with respect to [H] and [V] a) A line D(d,d) is drawn, in accordance with the method previously presented, which passes through the known point A(a,a) and which forms, with the projection planes [H] and [V], the complementary angles (/2-) and (/2-) (Fig. 6-a). The plane drawn through point A(a,a), perpendicular to line D(d,d), [P] D, is the required plane. The plane [P] is defined by the horizontal line G(g,g) and frontal line F(f,f) and forms angles and with planes [H] and [V]. b) A sphere of arbitrary radius and the center in point A(a,a) (Fig. 6-b) is used. The required planes are parallel to the tangent plane to the sphere forming with planes [H] and [V] the dihedral angles and respectively. Two cones with vertical axis are circumscribed to the sphere, respectively the straight end axis and vertexes V(v,v), S(s,s) respectively. The generatrices of the two cones form with the projection planes the angles (/2-) and (/2-) respectively. The cone V(v,v) meets the sphere through the horizontal circle bc, and cone S(s,s) through the frontal circle de. The contact points between the sphere and the cone are the contact points between the sphere and the tangent planes looked for, i.e. (1,1)
620
v'
P'
v' a' 1'
d' 2'
b' 1' 4'
c' e'
T' 2 g' P' 2
v'
g' f' Px
x
d' a' s'

5' 8' 4 8
/2 /2 v f d 1 a
h' O h
2 P
3 7 b a v c
f' v f
d 1 5
e 2 6
P 2
T2
c)
a) As other two cones, symmetrical to the former ones can be also circumscribed to the sphere, with respect to the center of the sphere, it results eight tangent planes, two by two parallel. In order that the problem be possible, it is necessary that ak ab. The problem admits: four solutions if + > / 2 (Fig. 6-a,b), two solutions if + = / 2 and no solutions if + < / 2. 3. CONCLUSIONS The paper analyzes and presents concrete cases of establishing positions, in space, of some geometric elements using the sphere or revolving cone as geometric loci containing the respective elements. The constructions
b)
Fig. 6. Inclined plane with respect to projection planes.
developed are relatively simple, sufficiently accurate and, in some cases, can eliminate certain analytical calculations, reducing on this way, the time for designing. 4. REFERENCES [1] Tnsescu, A., Geometrie descriptiv, E.D.P., Bucureti, 1975. [2] F. G.-M., Gomtrie descriptive. Elements et exercices, Paris, 1920, Ed. J. Gabay, Reprint (1996), ISBN 2-87647-170-1. [3] Orban, M., Proiecii i metode de transformare a proieciilor, Ed. Casa Crii de Stiin, Cluj-Napoca, 2008, ISBN 975-973-133-393-9.
Utilizarea sferei i conului de rotaie ca suprafee auxiliare n rezolvri grafice Rezumat: Rezolvarea grafic a unor probleme din geometria spaial este facilitat de utilizarea unor suprafee auxiliare, ca locuri geometrice, ce rspund cerinelor problemei. Lucrarea prezint cteva modaliti de utilizare a sferei i conului de rotaie, ca suprafee auxiliare, pentru determinarea grafic a poziiilor pe care le ocup n spaiu unele elemente geometrice, materializate pe piese din construcia de maini, a distanelor dintre acestea sau a unghiurilor pe care acestea le formeaz sau trebuie s le formeze cu alte elemente, n condiiile impuse n cadrul unor probleme concrete de proiectare. Rezultatele, suficient de precise, obinute exclusiv prin metode grafice simple i rapide, elimin n cele mai multe cazuri, necesitatea efecturii, n prealabil, a unor calcule analitice.
Magdalena ORBAN, PhD. Eng., Associate Professor, Technical University of Cluj-Napoca, Descriptive Geometry and Engineering Graphics Department, E-mail: magdalena.orban@gdgi.utcluj.ro, Office Phone: 40 264 401780 Cecilia CIORTEA, Student II-Ind.Eng., Technical University of Cluj-Napoca, Machine Building Faculty, E-mail: eliza.cecilia@yahoo.com Diana BOCA, Student II-Ind.Eng., Technical University of Cluj-Napoca, Machine Building Faculty, E-mail: diaboca01@yahoo.co.uk
621

THE OPTIMAL DESIGN OF RUBBER VIBRATION ISOLATORS OF A PLATE COMPACTOR

Radu Mircea MORARIU-GLIGOR
Abstract: This paper presents a procedure of optimization that can be applied for vibration isolators made of rubber, part of a plate compactor. The dimensioning of rubber vibration isolators is based, first on the calculation, then on choosing of some values or roundness of them, after which the final calculation is made. The proposed dimensioning method removes the drawback of the classic method being also faster. Keywords: optimization, vibration isolators, MathCAD.
1. INTRODUCTION It is a fact that the vibrations affect the proper functioning of any machine or installation from the working process. Hence, vibration isolation is required for obtaining the optimal functioning parameters, for reducing the wear of the components as well as for ensuring the safety during the working procedure of the human operator. Starting from the design stage it is necessary to provide a package of technical prescriptions (measures) that ensure that the vibrations regime is kept between the allowable limits imposed by the current regulations. The 2000/14/EC, 98/37/EC directions (applied in EU countries) the law 90/1996 (republished) and Work safety rules (published in 2002) are establishing the admissible levels for noise and vibrations in case of industrial installations. Likewise, the Romanian standard SR EN 1299 assesses a method for isolating the vibration source. The vibration isolators from isolation systems are: the metallic springs and elastic elements made of rubber, the last ones offering a better damping, especially for passing through the resonance. In this paper is presented an optimization method that can be applied for the vibration
isolators made of rubber, with round section, submitted to compression (figure 1).
Fig. 1. Vibration isolator submitted to compression
2. ISOLATION SYSTEMS BASED ON RUBBER VIBRATION ISOLATORS Isolation systems are made of rubber vibration insulators, which can be connected either in series or in parallel so that both, the prescribed geometrical and mechanical functioning conditions are fulfilled. The selection of one isolation system or another must take into account the following aspects: the mounting scheme, the magnitude of the acting forces and first of all, the isolation rate.
622
The figure 2 presents an assembly which consists of following components: the bed plate of the plate compactor (1) on which are attached (by means of chocks) four vibration isolators (3) that are mounted in the horizontal plane and are subjected to shear. In order to absorb (undertake) the vertical vibrations, was picked a vibration isolator (2) mounted in vertical position relative to the bed plate (1).
P (3) 2 3 xmax E st + 2 3 The obtained value is rounded to integer. d= D. The height h0 of the vibration isolator is determined in the following way: d (4) h0 = 4 E. The admissible strength of the vibration isolator element is verified using one of the following mathematical relations: 4 xmax E st 2 st st = + 2 3 a d 3 (5) for static loading, respectively: 4 xmax d E st 2 din din = + 2 3 a d 3 (6) in case of dynamic loading. F. Check the specific admissible deformation: P x st = h0 a 2 3 d E st + 2 3 (7) in static conditions, or dynamic conditions: P x din = h0 a 2 3 d d E st + 2 3 (8) G. If one of the conditions, mentioned above is not verified, the diameter d, the height h0 or the shape coefficient as well as the type of rubber must be reconsidered and the calculations are done again.
4. THE OPTIMAL DESIGN OF A RUBBER VIBRATION ISOLATOR, USING MATHCAD
Fig. 2. Isolation system based on four vibration isolator elements with horizontal disposition and one vibration isolator in vertical plane.
3. THE CALCULUS, DIMENSIONING AND TESTING ALGORITHMS The parameters which are considered in dimensioning and checking of the vibration isolators made of rubber are: the modulus of elasticity (E) and the shear modulus (G), the admissible compression stress () and shearing (), the allowable unit strain (), the dynamic multiplication coefficient (d), the internal energy loss factor, the mode of operation. It is considered a vibration isolator made of rubber, of full circular section, height h0 and diameter d. The input data are: the axial force P, the maximum deformation xmax, the hardness of the rubber, the operating conditions. The dimensioning to compression implies the following steps: A. The selection of the shape coefficient from one of the values: 0,25; 0,50; 0,75; 1,0; B. The assignment (based on tables) of a value for the modulus of elasticity (E), ; C. The calculus of the preliminary diameter d of the vibration isolator, using the formula:
For the optimal design of rubber vibration isolators with rounded section, subjected to virtual compression using MathCAD software, the mass of the isolator element was considered the objective function that must be minimized.
The constraints that are imposed refer to check on the admissible strength and admissible unit deformation. A supplementary condition requires that the deformation of the vibration isolator to be less than the maximum deformation. Hence, the implementation of the problem of the optimal dimensioning in MathCAD is:
Input data: axial force, maximum deformation, rubber strength, working conditions, modulus of elasticity, admissible resistance, admissible unit deformation, shape coefficient, rubber hardness (). Objective function:
16 is subjected to the following constrains: M (d ) =
d3
(9)
4 x max d E st d
2 3 din 3 + 2 a (10)
1,385 [N/mm2], as well as the admissible unit deformation a = 0,15 (or 15%). After that the dimensioning calculus was conducted, it resulted a diameter of 49,515 [mm], the determination and testing for all considered elements, this value being rounded to an integer value, that is 50 [mm]. It can be noticed that the obtained element do not verify nor the strength condition neither the admissible unit stress, so, the dimensioning calculus must be performed from the beginning. The optimal dimensioning, using MathCAD software, assumes making of the objective function and restrictions as previously described. So, it result an optimum diameter of 54,591 [mm] By performing a test for the vibration isolator element, using the rounded value of 55 [mm] it resulted that it keeps in with all the constraints. In Table 1 are presented the dimensional and design characteristics of the vibration isolators:
Table 1. The dimensional and design characteristics of the vibration isolators
623
4 x max a d
P 1 x max 2 d E st + 2 2 3 The called function Minimize is: F:= Minimize(M, d)
(11)
Item
(12)
The returned value of the Minimize function represents the optimal value of the diameter d.
4. NUMERICAL EXAMPLE
Preliminary diameter [mm] Selected diameter [mm] The height of vibration isolator [mm] Strength test Admissible deformation test
Classical design calculus 49,515
Optimization calculus
54,591 55 27,5
YES YES
50 25
NO NO
For exemplifying the optimal dimensioning method, it is considered a vibration isolator element that is subjected to an axial force P = 1960 [N], the maximum deformation having the value of 4 mm, the rubber hardness is 60 [Sh (A)], and the work regime is dynamic and short termed. Are also selected: the value for the shape coefficient = 0,50, the modulus of elasticity, E = 5,4 [N/mm2], admissible strength a =
It is noticed that after performing the dimensioning calculus, it was obtained a diameter that dont verify the admissible strength and compression tests. Thereafter, the calculus must be performed again by considering higher values for the vibration isolator diameter or modified values of shape coefficient or rubber strength (high strength rubber).
624
This type of calculations are time consuming and are purely empiric, the designer dont have the certainty that he selected the best variant of vibration isolator element.
5. CONCLUSIONS
unknown: the diameter of the vibrating isolator element. This paper can offer a simple alternative to those who desire to dimension vibration isolators but dont own enough knowledge about other optimization techniques.
6. REFERENCES
Due to elastic properties, characterized by equivalent stiffness coefficient, these elements are determining the working regime as the vibration isolator regime for the components that must be protected against vibrations, especially in case of passing through resonance regime. The vibration isolator elements are very important in the well functioning of machines and installations. The dimensioning of these elements can became very difficult and may require a lot of calculus. The author propose a method of dimensioning that bases on the optimization techniques and uses a dedicated soft, the results obtained validate the proposed method. The optimization process, by introducing some constrains, leads to a solution that satisfies the strength and deformation requirements, by a reduced material consumption, the total time necessary to design the vibration isolator being shortened. Although, this method presents a great disadvantage: the selection of a large number of input data that implies a problem with a single
[1] Bratu, P. Sisteme elastice de rezemare pentru maini i utilaje, Editura Tehnic, Bucureti, 1990, 288 pag. [2] Buzdugan, Gh. Izolarea antivibratorie a mainilor, Bucureti, Editura Academiei, 1980. [3] Davey, B.A., Payne, R.A. Rubber in Engineering Practice, New York, 1984. [4] Gafianu, M., .a. Organe de maini, Volumul I, Editura Tehnic, Bucureti, 1981. [5] Lambrescu, I. Calcul numeric ingineresc MathCAD, Editura Universitii din Ploieti, Ploieti, 2004. [6] Lindley, P. - Engineering design with natural rubber. Tehnical Bulletin The Malayesian Rubber Producers Research Association MRPRA, 1973. [7] Maxfield, B. Essential MathCAD For Engineering, Science and Math, Academic Press, 2009, ISBN: 978-0-12-374783-9. [8] Ursu-Fischer, N. Vibraiile sistemelor mecanice. Teorie i aplicaii, Editura Casa Crii de tiin, Cluj-Napoca, 1998.
PROIECTAREA OPTIMAL A ELEMENTELOR ANTIVIBRATILE DIN CAUCIUC DIN COMPONENA UNEI PLCI COMPACTOARE Rezumat: Lucrarea prezint o modalitate de dimensionare optimal a elementelor antivibratile din cauciuc, supuse la compresiune centric, din componena plcilor vibratoare. Modalitatea de dimensionare propus nltur dezavantajele metodei clasice, care se bazeaz pe un calcul de predimensionare urmat de alegerea unor valori standardizate sau rotunjirea valorilor parametrilor obinui.
Radu Mircea MORARIU-GLIGOR, Lector, Technical University of Cluj-Napoca, Department of Mechanical Systems Engineering, Faculty of Machine Building, rmogli70@yahoo.com, Ariesului 102/107, Cluj-Napoca, 0743-120463.
625

A METHOD TO DETERMINE THE VALUE OF THE GROUND PRESSURE DEVELOPED BY ONE WAY PLATE COMPACTORS
Radu Mircea MORARIU-GLIGOR, Gheorghe GLIGOR
Abstract: One of the most important parameters that are influencing the vibrating plate operation is represented by the ground pressure. The determination of this parameter from the design phase is an important objective for any designer. This paper came with a mathematical model, based on which the value of the ground pressure can be easily determined.
1. INTRODUCTION The plate compactors are equipments used to compact the soil, broken stones (non-cohesive materials), concrete or asphalt coatings, as well as residuum at the ecological ramps.
form the vibrating generator. The resulted vibrations generate the equipments advancing movement. In order to get compaction it is necessary to reach a certain mode of operation mainly characterized by the frequency and amplitude of vibrations due by the vibration generator. 2. MATHEMATICAL MODEL In the Figure 2 and 3 is represented the mechanical model and the forces that act upon the components of the compacting plate. The forces which act on the plate compactor are represented in figure 3. Starting from this, a mathematical model was developed the study proposed in this work. In the mechanical system, subjected to study, one of the components, rather the bed plate, performs a translational motion, along the surface of the ground, on which it acts with a force, normal on the ground, denoted with N and whose value can be determined. The supporting plate and the driving engine are involved in a plane-parallel motion, the entire system is characterized by 3 degrees of freedom (two translations along horizontal and vertical directions and a rotation around the horizontal axis).
Fig.1. Plate Compactor
From constructive point of view the plate compactor (Figure 1) consists of: base plate (1), vibrations generator (2) (fitted on the base plate) driven by a combustion engine (4) by means of V-belt gearing (3), the equipment being oriented by a handle (5). The plate compactors use a mechanism (driven by motor oil or a diesel one) which creates a descending force added to the equipments static weight. One or two eccentrically weights turning around usually
626
y1 y2 a rsm O5 C3 h 4 rsc r mctg O4 y3(in) y 3 m 3 ,Jc3
t
L2 O3L 6 D1 D3 D2 L1 d3 d2 L5 h3 L4
x3
(in) x3
h2 h1 O1
d1 O2
L3 m2
x2
x1
Fig. 2. Mechanical model

y1 y2
rsm O5 F 0
Ff O1
t T rsc r mctg L2 O3L 6 O4 F L2_L1 N L5 L1 F L6_L5 O2

m2g
L4
F L4_L3 L3 m2 x 2
x1
Fig. 3. Forces on the components of the plate
In order to develop the mathematical model that implies the write-in of the equations of motion, have been taken into account the following reference frames (figure 2): O1x1y1 a fixed reference frame, attached to the ground. The differential equations that characterize the motion of the two analyzed bodies are written relative to this system; O2x2y2 represents a mobile reference system, attached to the bed plate, which performs translation motion along the horizontal axis. To simplify the equations, it was considered that the O2y2 axis is passing through the center of the rotation axis of the inertial vibration generator. O3x3y3 a mobile reference system, connected to the frame, involved in a plane-parallel motion. On the basis of the mechanical model a mathematical model it was developed composed of four non-linear differential
equations corresponding to the movements performed by each component part of the mechanical system, as follows: - an equation corresponding to the plates movement along the horizontal direction; - three equations corresponding to the rotation of the frame around horizontal and vertical directions; The generalized coordinates that characterize the mechanical system subjected to study, are: - xO2_1 the displacement along the horizontal direction of the origin of the system referring to the vibrating plate; - xO3_1, respectively yO3_1 the displacements along the horizontal respectively the vertical direction, of the O3 origin of the mobile system referring to the frame; - , turning angle of the frame (in trigonometrically sense) due to the horizontal line; The quantities used by the mathematical system have the following significance: m0 eccentric mass; m2 plate mass; m3 mass of the frame and motor; k1 stiffness; k2 axial factor of rigidity; xA_B, respectively yA_B distances on the axis O1x1, respectively O1y1 between points A and B. In order to solve the mathematical model previously presented it was written a program in C language. Resolution on the system of differential equations is found with the help of RungeKutta fourth order method. The study conducted on the proposed mechanical model relieved the fact that the plate is subjected to the following forces: (figure 3): the sole weight m2g; normal reaction N; the friction force N, which acts counterwise to the displacement of the bed plate; the rotational inertia caused by the inertial vibration generator; the stress exerted on one strand of the transmission belt, whose position (along the upper or lower strand) depends on the direction of rotation of the driving motor; the elastic forces that appear in the rubber elements; The differential equations are established based on the momentum theorem, projected on the horizontal and vertical axes, as presented bellow: - horizontal projection:
627
( m2 + m0 )
d xO 2 _ 1 2 = m r cos t N sign 0 dt d t2 + ( x L 2 _ 1 xL1 _ 1 ) k1 + ( xL 4 _ 1 x L 3 _ 1 ) k1 + T cos
d 2 xO 2 _ 1
(1)
vertical projection:
0 = m 0 r 2 sin t m 2 g + N + ( y L 2 _ 1 y L1 _ 1 ) k1 + ( y L 4 _ 1 y L 3_ 1) k1 + 0 + T sin + k 2 d max y L 6 _ 1 y L 5 _ 1
daca y L 6 _ 1 y L 5 _ 1 > d max , daca y L 6 _ 1 y L 5 _ 1 d max .
(2)
From equation (2) can be determined the normal reaction N, as it results from the following expression:
The expression (3) is used to determine the values for the ground pressure (force)
N = m0 r 2 sin t + m 2 g ( y L 2 _ 1 y L1 _ 1 ) k1 ( y L 4 _ 1 y L 3_ 1) k1 0 T sin + k 2 d max y L 6 _ 1 y L 5 _ 1
daca y L 6 _ 1 y L 5 _ 1 > d max , daca y L 6 _ 1 y L 5 _ 1 d max .

(3)
From equation (1) can be determined the expression for horizontal component of the acceleration that characterize the motion of the bed plate along the horizontal direction.
The equations that describe the motion of the bed plate are determined based on the theorem of the mass point motion (5).
d 2 xO 2 _1 d t2
d xO 2 _1 2 + (xL 2 _1 xL1_1 ) k1 + m0 r cos t N sign 1 dt = m2 + m0 + ( xL 4 _1 xL3 _1 ) k1 + T cos
(4) (5)
m3
d 2 xC 3 _ 1 dt
2
F ,
x
m3
d 2 yC 3 _ 1 dt
y1
2
= Fy
y2
By projecting these equations on O1x1 and O1y1 directions, are obtained two differential equations of second order. Another fundamental theorem that can be applied for the same purpose is the theorem of relative angular momentum; the result is also a differential equation of second order. The forces that are acting upon the supporting plate are graphically represented in the figure 4. The differential equations that define the motion of the supporting plate, based on the equation of the mass point movement are represented in (5), where the right term of the equation represents the horizontal and vertical projections of all forces exerted on the chassis.
rsm O5 T t rsc rmctg L4 F L6_L5 L2 O3 O4 L 6 F m g L2_L1 3 F L4_L3 L5 L1 L3 m2 x2 O2 O1
x1
Figure 4. The forces exerted on the supporting l
628
d 2 xO3 _1 d2 ( ) + + ( xC 3 _ 3 cos + yC 3 _ 3 sin sin cos m3 x y C3_ 3 C3_ 3 d t2 d t2 = ( xL 2 _1 xL1_1 )k1 ( xL 4 _1 xL3 _1 )k1 T cos ,
d 2 yO 3 _ 1 m3 + d t2 d2 + ( xC 3 _ 3 sin yC 3 _ 3 cos sin ) d t2 daca y L 6 _ 1 y L 5 _1 > d max daca y L 6 _ 1 y L 5 _1 d max
d ) dt
=
(6)
x C 3 _ 3 cos yC 3 _ 3
= ( y L 2 _ 1 y L1_ 1 )k1 ( y L 4 _1 y L 3 _ 1 )k1 T sin m3 g + 0 + k 2 d max y L 6 _1 y L 5 _ 1
d ) dt
(7)
JC3
d2 = ( x L 2 _ 1 xC 3 _ 1 ) ( y L 2 _ 1 y L1 _ 1 )+ ( y L 2 _ 1 y C 3 _ 1 )(x L 2 _ 1 x L1 _ 1 ) k1 + d t2 + (x L 4 _ 1 xC 3 _ 1 ) ( y L 4 _ 1 y L 3 _ 1 ) + ( y L 4 _ 1 y C 3 _ 1 ) (x L 4 _ 1 x L 3 _ 1 ) k1 +
d max | y L 6 _ 1 y L 5 _ 1 | k 2 + (x L 6 _ 1 xC 3 _ 1 ) sau + 0 (x 03 _ 1 + rsm sin cos (h4 rsm cos ) sin xC 3 _ 1 ) sin + + T + ( y 03 _ 1 + rsm sin sin + (h4 rsm cos ) cos y C 3 _ 1 ) cos
The generalized coordinates in this case, are represented by: xO3_1 i yO3_1 the displacement along the horizontal and vertical direction of the origin O3 of the mobile system attached to the supporting plate, respectively the rotation angle of the supporting plate. After replacing the expressions for both, the second degree derivative and bounding forces, the differential equations defined with (5) became (6) and (7). It can be noticed that the fourth of the differential equations, above written, is obtained by applying the theorem of the angular momentum, relative to the mass center C3 whose final expression is given by (8). The system of four differential equations (4), (6), (7) and (8), presented above, characterizes the motion of all the elements contained by the vibrating plate. 3. NUMERICAL EXAMPLE
(8)
The study was conducted for the case of a compacting plate having the following characteristics: the engine speed: 3600 [rot/min], the vibration generator speed: 5400 [rot/min], the mass of the bed plate: m1 = 70 [kg], the mass of the frame and driving engine: m2 = 55 [kg], the mass of the eccentric: m0 = 3,465 [kg], the eccentricity r = 0,0134, the constant of elasticity k1 = 58110*2 [N/m], the constant of elasticity k2 = 608250 [N/m], the friction coefficient between the ground and the base plate: = 0,35. All these elements resulted from designing a variant of compacting plate, using the application Solid Edge, in this way, were obtained values for masses, displacements, mass centers, inertia moments, etc.
629
Time [s] Fig. 5. The diagram of variation in time for the normal ground pressure
Time [s] Fig. 6. Diagram of variation for the ground pressure for different values of the vibration generator speed
The solving the differential system, allows to deduce the variations in time for the plate (linear) and for the frame (linear and angular), as well as the time variation for the ground pressure (force). In figure 5 is illustrated the time variation for the ground pressure exerted by the compacting plate, the maximum value is of 18.213 N. This force is responsible for the process of ground compacting and this is the reason for knowing the way different process parameters are influencing its variation. One important parameter that affects the ground pressure is represented by the speed of the vibration generator. For the considered vibration generator, the following values were taken into study: 4200, 4600, 5000, 5400, 5800, 6200 [rot/min]. Due to the fact that the rotation speed of the driving engines take different values, in order to attain the imposed values is necessary to
select the transmission ratio corresponding to the V-belt cord drive. In the figure 6 is represented the variation in time of the ground pressure, for different values of the vibration generator speed. It is noticed that increasing the value of the ground pressure causes an increase in the speed of the vibration generator. A possible explanation for this can be the mathematical expression used for determining the value of the ground pressure. 4. CONCLUSIONS The increase in speed of the vibration generator has a positive effect on the ground pressure in the way of increasing, but it can also have a bad influence on other parameters (for example: the displacement speed) or even on the functioning of the compacting plate.
630
The value for the ground pressure depends on a lot of factors that also influence directly the well functioning of the equipment. The mathematical model developed in this paper represents a very useful instrument in determining the values of the parameters that affect the functioning of the equipment as well as of the ground pressure exerted by the compacting plate. Thereby, are obtained the values for the functioning parameters that correspond to a high value of the ground pressure, high speeds and reduced frame vibrations. 5. REFERENCES [1] Buzdugan, G., Fetcu, L., Rade, M. Vibrations of mechanical systems (in Romanian), Academy Publishing House, Bucharest, 1975; [2] Bratu, P. Elastic systems for machines and machinery (in Romanian), Technical Publishing, Bucharest, 1990; [3] Harris, C.M. - Shock and Vibration Handbook, Fourth Edition, New York, McGraw Hill, 1996; [4] Mihilescu, S., Vlasiu, G. Civil engineering machines and working methods (in Romanian), Educational and Pedagogical Publishing, Bucharest, 1973;
[5] Munteanu, M. Introduction in dynamics of vibrating machines (in Romanian), Academy Publishing House, Bucharest, 1986; [6] Shigley, J.E., Mischkle, Ch.R. Standard Handbook of Machine Design, Second Edition, McGraw-Hill Companies Inc., 1996, ISBN 007-056958-4; [7] Ursu-Fischer, N. Vibrations of mechanical systems. Theory and applications (in Romanian). House of Science Book, ClujNapoca, 1998; [8] Ursu-Fischer, N., Ursu, M. Numerical methods in engineering and programming in C/C++, (in Romanian), vol. I., House of Science Book, Cluj-Napoca, 2000; [9] Ursu-Fischer, N., Ursu, M. Programming in C in engineering (in Romanian), House of Science Book, Cluj-Napoca, 2001; [10] Ursu-Fischer, N., Popescu, D.I., Haiduc, N., Morariu-Gligor, R., Ursu, M. Contributions in modelling and simulation of the vibrating plate compactors movements. tiin i Inginerie, vol. II, Editura AGIR Bucureti, 2002, pg. 669-676, [11] Ursu-Fischer, N., Popescu, D.I., Haiduc, N., Morariu-Gligor, R., Ursu, M. Study of the operation at the plate vibrating compactor, tiin i Inginerie, vol. III, Bucureti, Editura AGIR, 2003, pg. 213-220.
DETERMINAREA VALORII FOREI DE APSARE NORMALE DEZVOLTAT DE PLCILE COMPACTOARE UNIDIRECIONALE Rezumat: Unul din parametri cei mai importani n funcionarea plcilor compactoare l reprezint fora de apsare asupra solului. Determinarea acesteia nc din etapa de proiectare constituie un obiectiv pentru fiecare proiectant de plci compactoare. n lucrare este prezentat un model matematic pe baza cruia se poate determina valoarea forei de apsare asupra solului funcie de caracteristicile constructive ale plcii compactoare.
Radu Mircea MORARIU-GLIGOR, Lector, Technical University of Cluj-Napoca, Department of Mechanical Systems Engineering, Faculty of Machine Building, rmogli70@yahoo.com, Ariesului 102/107, Cluj-Napoca, 0743-120463. Gheorghe GLIGOR, Assoc. Prof. Dr. Eng., Technical University of Cluj-Napoca, Cluj-Napoca, Bdul Muncii 103-105, Romania, Department of Manufacturing Engineering, Faculty of Machine Building, ghgligor@ tcm.utcluj.ro, Str. Timuului 14, Cluj-Napoca, 40-264-417882.
631

THE GEOMETRIC EQUATIONS OF TWO MACHINE TOOL TYPE 3T3R IN COOPERATION

Ancua MIRCEA, Kalman KACSO, Iuliu NEGREAN, Ioan VUSCAN, Radu MORARIU-GLIGOR
Abstract: In this paper are presented the geometric equations for the most and complex structure type 3T3R 3T3R. The structure is compound from a positioning and orientating structure of the TOOL and the other is afferent of the PIECE. Both structures are in cooperation on a platform commune. Starting from the general cooperating model type 3T3R 3T3R the equations can be applied for any type of kinematic structure. Key words: kinematic structures, applied mechanics, cooperation, machine tool, 3T3R structure.
1. INTRODUCTION The cooperating and general structure type 3T3R3T3R is compound from a positioning and orientating structure of the TOOL and the other is positioning and orientating structure of the PIECE (fig. 1).
In the action of cooperating of both structures, it follows that, the workpiece, represented by system OP xP yP z P { P} symbolized according to figure 2, to be machined by the tool represented by system OS xS yS zS {S } after a certain technological process. By a series of homogeny transformations, as follows, the cooperating tool-piece noted with TSP can be expressed with relation:
TSP = (TPO ) TS 0
1
(1)
where
Fig. 1. The cooperation of the positioning and orientating structure of the TOOL and of the PIECE
represent the system of the tool OS xS yS zS {S } in rapport with the system of the piece, OP xP yP z P { P} ;
TSP
1 (TPO ) represent the transposed matrix of the system of the piece OP xP yP z P { P} in rapport with the fixed system O0 x0 y0 z0 {0} ;
1.1 The direct geometry equations To achieve the equations of the cooperating general model of structure type 3T3R3T3R are facilitated by the fact, for both structures, of the TOOL and the PIECE was determinate independently the direct and inverse geometry equations (see Acta Technica Napocenis no. 54, vol. III, 2011). Having for both structures the equations, it can be determinate the direct geometry equations for the structures type 3T3R3T3R.
For that, are necessary the equations of the last system of the positioning and orientating structure of the TOOL and of the PIECE.
represent the system of the tool OS xS yS zS {S } in rapport with the fixed system O0 x0 y0 z0 {0} . The cooperation of the tool and of the piece can be expressed by the next relation:
TS 0 TSP00 TSP10 = TSP20 T SP30 TSP01 TSP11 TSP21 TSP31 TSP02 TSP12 TSP22 TSP32 TSP03 TSP13 TSP23 TSP33
TSP
(2)
632
Fig. 2. The kinematical structure type 3T3R 3T3R in the zero configuration (the cooperation of the structures 3T3R 0 type portal and 0 3T3R type table)
The components of the equation (2) are obtained with the aid of orientating matrix RSP which express the orientating of the tool system {S } from system { P} . In this way it is use the relation: RSP = R ( P zS ; P xS ; P yS ) :
RSP R PzS ; PxS ; P yS = szS sxS s yS + czS c zS szS cxS szS sxS c yS + czS s yS = czS sxS s yS + szS c yS czS cxS czS sxS c yS + szS c yS cxS s yS sxS cxS c yS
(3)
From relation (5) result the column vector of the positioning and orientating parameters, define with the next calculus expression:
633
P S p P pzSP P pxSP P p ySP SP = ........ = .................................. . (4) P S P P P xSP ySP SP zSP P
Having all parameters, the positioning vectors

P
X SP
p SP can
be wrote as:
pzSP
b1 c ( q10 ) s ( q12 ) b2 c ( q10 ) s ( q12 ) + d 2 s ( q10 ) s ( q12 ) d 3 s ( q10 ) s ( q12 ) d 4 s ( q10 ) s ( q12 ) d6 s ( q10 ) s ( q12 ) d7 s ( q10 ) s ( q12 ) + c ( q11 ) c ( q12 ) q3 c ( q11 ) c ( q12 ) q8 + c ( q10 ) s ( q12 ) q1 c ( q10 ) s ( q12 ) q7 + s ( q10 ) s ( q12 ) q2 s ( q10 ) s ( q12 ) q9 d 2 c ( q10 ) c ( q12 ) s ( q11 ) + d 3 c ( q10 ) c ( q12 ) s ( q11 ) + + d c q c q s q + d c q c q s q + d c q c q s q ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 10 12 11 6 10 12 11 7 10 12 11 4 = c ( q10 ) c ( q12 ) s ( q11 ) q2 + c ( q10 ) c ( q12 ) s ( q11 ) q9 + c ( q12 ) s ( q10 ) s ( q11 ) q1 c ( q12 ) s ( q10 ) s ( q11 ) q7 + b1 c ( q12 ) s ( q10 ) s ( q11 ) b2 c ( q12 ) s ( q10 ) s ( q11 ) d 5 c ( q4 ) c ( q11 ) c ( q12 ) s ( q6 ) d 5 c ( q5 ) c ( q6 ) s ( q10 ) s ( q12 ) d c ( q ) s ( q ) s ( q ) s ( q ) d c ( q ) c ( q ) c ( q ) c ( q ) s ( q ) 5 10 4 6 12 5 5 6 10 12 11 d 5 c ( q4 ) c ( q6 ) c ( q10 ) s ( q5 ) s ( q12 ) + d 5 c ( q6 ) c ( q11 ) c ( q12 ) s ( q4 ) s ( q5 ) d 5 c ( q12 ) s ( q4 ) s ( q6 ) s ( q10 ) s ( q11 ) + d 5 c ( q4 ) c ( q6 ) c ( q12 ) s ( q5 ) s ( q11 )
b1 c ( q10 ) c ( q12 ) b2 c ( q10 ) c ( q12 ) + d2 c ( q12 ) s ( q10 ) d3 c ( q12 ) s ( q10 ) d4 c ( q12 ) s ( q10 ) d6 c ( q12 ) s ( q10 ) d7 c ( q12 ) s ( q10 ) + c ( q10 ) c ( q12 ) q1 c ( q10 ) c ( q12 ) q7 + c q c q q c q s q q c q s q q + + ( ) ( ) ( ) ( ) ( ) ( ) 10 12 2 11 12 3 11 12 8 c ( q12 ) s ( q10 ) q9 + d 2 c ( q10 ) s ( q11 ) s ( q12 ) d3 c ( q10 ) s ( q11 ) s ( q12 ) + d4 c ( q10 ) s ( q11 ) s ( q12 ) d6 c ( q10 ) s ( q11 ) s ( q12 ) = d7 c ( q10 ) s ( q11 ) s ( q12 ) + c ( q10 ) s ( q11 ) s ( q12 ) q2 c ( q10 ) s ( q11 ) s ( q12 ) q9 s ( q10 ) s ( q11 ) s ( q12 ) q1 + s ( q10 ) s ( q11 ) s ( q12 ) q7 b1 s ( q10 ) s ( q11 ) s ( q12 ) + +b2 s ( q10 ) s ( q11 ) s ( q12 ) d5 c ( q5 ) c ( q6 ) c ( q12 ) s ( q10 ) d5 c ( q10 ) c ( q12 ) s ( q4 ) s ( q6 ) + d5 c ( q4 ) c ( q11 ) s ( q6 ) s ( q12 ) d5 c ( q4 ) c ( q6 ) c ( q10 ) c ( q12 ) s ( q5 ) d5 c ( q5 ) c ( q6 ) c ( q10 ) s ( q11 ) s ( q12 ) + + d5 c ( q6 ) c ( q11 ) s ( q4 ) s ( q5 ) s ( q12 ) + d5 s ( q4 ) s ( q6 ) s ( q10 ) s ( q11 ) s ( q12 ) + d5 c ( q4 ) c ( q6 ) s ( q5 ) s ( q10 ) s ( q11 ) s ( q12 )
(5)
p xSP
(6)
p ySP
s ( q11 ) q3 s ( q11 ) q8 b1 c ( q11 ) s ( q10 ) + b2 c ( q11 ) s ( q10 ) + + d 2 c ( q10 ) c ( q11 ) d 3 c ( q10 ) c ( q11 ) d 4 c ( q10 ) c ( q11 ) d6 c ( q10 ) c ( q11 ) d7 c ( q10 ) c ( q11 ) + c ( q10 ) c ( q11 ) q2 c ( q10 ) c ( q11 ) q9 c ( q11 ) s ( q10 ) q1 + = +c ( q11 ) s ( q10 ) q7 d 5 c ( q4 ) s ( q6 ) s ( q11 ) d 5 c ( q5 ) c ( q6 ) c ( q10 ) c ( q11 ) d 5 c ( q6 ) s ( q4 ) s ( q5 ) s ( q11 ) + d 5 c ( q11 ) s ( q4 ) s ( q6 ) s ( q10 ) d c q c q c q s q s q ( ) ( ) ( ) ( ) ( ) 5 4 6 11 5 10
(7)
1.2 The inverse geometry equations In this case the input data are the position and the orientation of the piece in rapport with the system of the tool. Are wrote a series of equations between the systems {0} {S } { P} , system noted with TSP , and meaning:
634
R p SP = SP 0 0 0 1
P (0 )
TSP
S Px S = S Py Pz 0
Px S Py S Pz
S
Px S Py S Pz
S
P P
pxSP p ySP P pzSP 1
(8)
Having the final results of the positioning and orientating coordinates relations (10) and (11) in the case of the tool-piece cooperating, these through particularization can be applied to any type of kinematic structure. 2. CONCLUSIONS Mathematical modeling of the structure type 3T3R3T3R represent an important realization, because starting from this it can be determinate the generalized coordinates for any type of kinematic structure. Also open new research directions about the modeling and simulation of machine tools by applying on that of locating matrix, usually apply on robots. 3. SELECTIVE REFERENCES [1] Negrean, I., Vuscan, I., Haiduc , N. (1998), Robotics: Kinematic and Dynamic Modeling, Didactical and Pedagogical Publishing House, Bucharest, Romania; [2] Negrean, I., Duca, A., Negrean, D.C., Kacso, K. (2008), Advanced mechanics in robotics, UT Press, Cluj-Napoca, Romania;
The positioning vectors with the relation:
p SP
(0 )
are explained
p SP
pxP cos ( Px ) + b1 p yP cos ( Px ) + + ( d1 + d 2 d 3 d 4 d 5 pzP ) cos ( Px ) P p xSP ............................................................. .......... P pxP cos Py + b1 p yP cos Py + p ySP = = + ( d1 + d 2 d 3 d 4 d 5 pzP ) cos Py .......... S ............................................................. p zSP p cos ( Pz ) + b1 p yP cos ( Pz ) + xP + ( d1 + d 2 d 3 d 4 d 5 pzP ) cos ( Pz )
( ) ( (
( ) ( )
(9)
The coordinates for positioning are:

TSP = P pxSP q j ; 03 P TSP13 = p ySP q j+ 1 ; P TSP23 = pzSP q j+ 2 .
(10)
The coordinates for orientating are:

TSP21 = sin P xS qi ;
P P TSP20 = cos xS sin yS qi+ 1 ; (11) P P T = cos cos PS xS yS 22 P P TSP01 = sin zS cos xS qi+ 2 . P P TSP11 = cos zS cos xS
Ecuaiile geometrice a dou maini unelte de tip 3T3R n cooperare Rezumat: n aceast lucrare sunt prezentate ecuaiile geometriei directe i inverse pentru cea mai complex structur de tip 3T3R 3T3R. Aceasta este compus din o structur de poziionare i orienatare a SCULEI, cealalt fiind aferent PIESEI. Ambele structuri sunt n cooperare pe o platform comun. Pornind de la acest model general de cooperare de tip 3T3R 3T3R ecuaiile geometriei directe i inverse pot fi aplicate pentru oricare tip de structur cinematic.
Ancuta Mircea, Eng., PhD Student, Technical University of Cluj-Napoca, Machine Building Department, aecrimancuta@yahoo.com, str. Brates no. 5, Cluj-Napoca, 0748/362266. Kalman Kacso, Assistant, Technical University of Cluj-Napoca, Department of Applied Mechanics and Computer Programming, kacsokalman@gmail.com, Office Phone 0264/401616. Iuliu Negrean, Prof. Dr. Eng., Head of Department of Applied Mechanics and Computer Programming, iuliu.negrean@mep.utcluj.ro, Office Phone 0264/401616. Ioan Vuscan, Prof. Dr. Eng., Technical University of Cluj-Napoca, Machine Building Department, Gheorghe.Vuscan@tcm.utcluj.ro, str. Aviator Badescu, no. 24A, Cluj-Napoca. Radu Morariu-Gligor, Lecturer Dr. Eng., Technical University of Cluj-Napoca, Department of Applied Mechanics and Computer Programming, rmogli70@yahoo.com, 0743/120463.
635

DETERMINATING THE ERRORS COMPENSATIONS BY USING NUMERICAL SIMULATIONS

Ancua MIRCEA, Kalman KACSO, Ioan VUSCAN, Iuliu NEGREAN, Radu MORARIUGLIGOR
Abstract: In this paper, are presented numerical simulations results regarding the compensation of the errors. To realize that, the start point it given by the errors, which appear inevitably during the machining processes. This approach it was determinate a general and complex mathematical model for the most complex structure type 3T3R3T3R. The model is able to calculate all generalized coordinates knowing only the initial linear and angular errors. The initial errors can be imposed or measured; it must know the size of these. This thing allows that, the next parts that will correctly and precisely machine even those was improperly positioned into fixing devices. Key words: kinematic structures, error compensation, machine tool, 3T3R structure.
1. INTRODUCTION The most and complex structure type 3T3R 3T3R is characterized by that, the all motions are attributed simultaneous, both positioning and orientation structure of the tool and of the piece. This model in variant without errors and with errors allows to be applied for any type of structure (728 variants possible) and is calculating the positioning and orientation of any point from the surface of the part. Also, into the mathematical model are introduced the initial errors. The model is calculated the generalized coordinates to allow to machine correctly and precisely the parts with corrections. 1.1 Cooperating general model 3T3R 3T3R applied for a machine tool type 3T2R 0 A major advantage of the general cooperating model with errors compensations is that, it can be applied to any kinematic configuration, with other words it can be particularized for any type of machine tools. In this case, it is about a machine tool type 3T2R0 (fig. 1). This is having three
coordinates linear (q1, q2 and q3) and the other two is for orientation, noted with q4 and q5 as we can see in figure 1. The main stages of the mathematical model with errors compensations are: to process a part correctly without initial errors; to process the same part with initial errors.
Fig. 1. The cooperation kinematic structure 3T2R 0 in the zero configuration
636
In the both stages, the output dates are the coordinates generalized. In the case of machining without errors the input dates are the positioning and orientation vectors of the workpieces. In the other case, that of machining with errors beside the positioning and orientation vectors of the workpieces appears the errors. These can be imposed or measured. The phenomenon of errors compensations is given by the fact, that knowing the initial errors, for the next parts which are machined, it will be realize correctly and without errors. The coordinates generalized are given the linear and orientation displacement along the X, Y and Z axis, necessary to machine correctly. Also, in the both cases, the reporting of the systems is different. In the case of the calculating of coordinates generalized without errors the reporting is made at the nominal system of the piece, noted with SP. In the other case, with initial errors, the reporting is made on the real system of the piece, noted with SPreal. Are use different type of systems because, the last one, SPreal is affected by errors. These systems can be overlap or not, depending by the machining conditions.
pxSP , p ySP , pzSP the positioning coordinate of the

tool-piece cooperation along the X, Y and Z axis; zSP , xSP , ySP the orientation coordinate of the tool-piece cooperation along the X, Y and Z axis; As a remark, the positioning and orientation vectors ( pxSP , p ySP , pzSP ) are equivalents with
dX , dY and dZ the afferent notations of the SolidWorks software. The coordinates generalized results from the general cooperating model 3T3R3T3R are: Table 2 The generalised coordinates (displacements on the motor couples) mm o
q1
6,66
q2
-179
q3
24,99
q4
0
q5
0
1.2 The case of machining of workpiece without errors

To accomplish this thing, in a first step are machined some points from the piece without to introduce initial errors.
Fig. 3. The validation of the generalized coordinates for the machining point 1
To machine another point from the surface of the piece is illustrated bellow. In this case, is about a inclined hole, meaning that are necessary orientation movements.
Fig. 2. Positioning and orientation coordinates for the machining point 1
Table 1
(0 )
P
Fig. 4. Positioning and orientation coordinates for the machining point 2
X SP = p SP
P
SP
p SP mm
SP o
pxSP
24,99
p ySP
6,66
pzSP
5
SP
SP
The input parameters are listed in the table 3. Table 3 P P 0 ( ) X SP = p SP SP

SP
The signification of the notations are:
pxSP
23,79
p SP mm p ySP pzSP
25 13,79
zSP
0
SP o xSP
45
ySP
0
637
The output parameters are presented in the table 4. These represent the new generalized coordinates obtained from the mathematical model.
Table 4 The generalized coordinates

mm
o
q1
-7,5269
q2
-183,6830
q3
23,79
q4
0
q5
45 Fig. 6. The values of the initial errors for the machining points 1 and 2
As we can see from the figure 6 the new system of the part is the real system, noted with S Preal . In this case, the system is overlap on nominal system of the workpiece, noted with S P .The values of the imposed errors are listed in the table 5.
Linear errors mm dx
0,38
Table 5 Angular errors o
As we can see, from the figures 2 and 3, respectively 4 and 5 the positioning and orientation vectors are the same after the results obtained from the general cooperating model. The generalized coordinates calculated with the general model 3T3R3T3R are introduced into the SolidWorks where are given the same values to allow to the tool to process the machining points desired. Having identically results, it meaning that the general mathematical model 3T3R3T3R is validate in the case of machining without introducing initial errors (fig. 3 and 5). 1.3 The case of machining of workpiece with errors Like in the precedent paragraph, here the input parameters are remain the same, with difference that in the mathematical model are introduced the initial errors noted with d x , d y , d z the linear errors and with x , y , z the angular errors. These errors can be imposed or measured; a fact important is that to know the size of the errors to determinate the generalized coordinates. The points which it will be machined are remaining unchanged. The parameters which must be introduced into the model are the positioning and orientation vectors and the linear and angular errors. These are represented in the figure 6 and in the table 5.
dy
0,52
dz
0,28
x
0
y
0
z
0,5
These errors (tab. 5) and the positioning and orientating vectors (tab. 4) are introduced in the cooperating general model 3T3R3T3R resulting the compensating generalized coordinates, which allow to accomplish a machining process correctly and precisely. The coordinates generalized obtained from the mathematical model are presented in the table 6. These value of q1 , q2 , q3 , q4 and
q5 represent the displacements necessary on X, Y

and Z directions so the workpieces to be processed correctly even these are wrong positioned into fixing devices. Table 6 The compensating generalized coordinates mm o
q1
-0,738
q2
-0,28
q3
-0,3218
q4
0
q5
0
638
For the machining point 2 the positioning and orientating vectors remain unchanged (tab. 1). Additionally into the cooperating general model 3T3R3T3R are introduced the errors, presented in table 5. Behind the calculus the new generalized coordinates are introduced into SolidWorks which are give the new displacements along the X, Y and Z axis in such manner that, the workpiece to be machined correctly.
3. CONCLUSIONS
The numerical analysis, in this case a particularization of the general cooperating model with errors compensation for a structure type portal 3T2R 0, by the results obtained confirm that the general cooperating model type 3T3R3T3R, first is work properly and the second it can be applied for any kinematic structure. The remark regarding this affirmation is to argument the dimension of values taken on the numerical simulations. In conclusion, the general model type 3T3R 3T3R with error compensation represents an important way into the manufacturing processes because knowing the size of initial errors, even them, imposed or measured can be determinate the auxiliary displacement for the motor couples. Knowing this, the next parts, which must be process, will be correctly manufactured. Also, is allowing to realize parts correctly, precisely even these are incorrectly, and imprecise installed on fixing devices.
The compensating generalized coordinates mm o
q1
-0,728
q2
-0,28
q3
-0,4456
q4
-0,5
q5
0
In figure 8 these are identically with the positioning and orientating vectors of the workpiece, meaning that the cooperating general model 3T3R 3T3R with errors compensation is validated (fig. 7 and 8).
4. SELECTIVE REFERENCES [1] Negrean, I., Vuscan, I., Haiduc , N. (1998), Robotics: Kinematic and Dynamic Modeling, Didactical and Pedagogical Publishing House, Bucharest, Romania; [2] Negrean, I., Duca, A., Negrean, D.C., Kacso, K. (2008), Advanced mechanics in robotics, UT Press, Cluj-Napoca, Romania.
Fig. 8. The validation of the generalized coordinates for the machining point 2 Determinarea erorilor de compensare prin utilizarea simulrilor numerice Rezumat: n aceast lucrare sunt prezentate rezultatele simulrilor numerice privind compensarea erorilor. Pentru a realiza acest lucru punctul de pornire este dat de erori, erori care apar n mod inevitabil pe durata proceselor de prelucrare. n acest sens a fost determinat un model matematic pentru cea mai general i complex structur de tipul 3T3R 3T3R. Modelul matematic este capabil s calculeze toate coordonatele generalizate cunoscndu-se doar erorile ini iale de pozi ionare i orientare. Acest fapt permite, ca urmtoarele piese s fie prelucrate n mod corect i precis, chiar dac sunt pozi ionate imprecis n dispozitivele de fixare.
Ancua Mircea, Eng, PhD Student, Technical University of Cluj-Napoca, Machine Building Department, aecrimancuta@yahoo.com, al. Brates no. 5, Cluj-Napoca, 0748/362266. Kalman Kacso, Assistant, Technical University of Cluj-Napoca, Department of Applied Mechanics and Computer Programming, kacsokalman@gmail.com, Office Phone 0264/401616. Ioan Vu can, Prof. Dr. Eng., Technical University of Cluj-Napoca, Machine Building Department, Gheorghe.Vuscan@tcm.utcluj.ro, str. Aviator Badescu, no. 24A, Cluj-Napoca. Iuliu Negrean, Prof. Dr. Eng., Head of Department of Applied Mechanics and Computer Programming, iuliu.negrean@mep.utcluj.ro, Office Phone 0264/401616. Radu Morariu-Gligor, Lecturer Dr. Eng., Technical University of Cluj-Napoca, Department of Applied Mechanics and Computer Programming, rmogli70@yahoo.com, 0743/120463.
639

EXPERIMENTAL SIMULATION FOR IMPLEMENTATION OF PREDICTIVE MAINTENANCE

Marius BAIDOC, Clin Vasile LUMEI, Florin EPE-BOBESCU, Alina Sabina PACA, Mariana ARGHIR
Abstract: In this paper are prezented two experimental stands: the study of vibration and mechanical and
electrical defects simulation, and the study of shaft horizontal alignment. The implementation is given in the field of predictive and proactive maintenance in the industry. Attainment of these stands is necessary to simulate actual conditions of occurrence of failure for the application, the impairment corrections of the manufacturing process.
Key words: experimental stands, implementation, predictive maintenance.
1. INTRODUCTION One of the most important objectives of the implementation of the comprehensive qualitative predictive and proactive industrial companies in the technical equipment is the equipment for measuring vibration simulation and experimental stands for fault in different operating regimes, necessary training of personnel. To this end have been designed and executed two experimental maintenance implementation stands necessary predictive and proactive as well as for staff training in the sector of activity: Stand for simulation study of vibration and mechanical and electrical defects; Stand for the study of shaft horizontal alignment These stands can be used in laboratories for studying the dynamics of the technical universities of machinery fault diagnosis and rotor. 2. STAND FOR SIMULATION STUDY OF VIBRATION AND DEFECTS Analysis of defects by studying dynamic rotating machinery vibration is a complex phenomenon that involves both theoretical
knowledge and practical knowledge, acquired over time. In this respect, for the study of vibrations and simulation of mechanical and electrical defects was designed and executed a stand experimental, need preparation and training for maintenance engineers [Lee 04]. The stand is designed for accelerated testing by simulating the mechanical and electrical faults, the induction of disturbing forces or external requests, resulting in the reduction of dynamic equipment reliability. The chassis was designed on the basis of experience gained from the measurements, are monitoring and analysis of vibration in rotating machinery dynamic. The stand is shown in Figure 1 and can be simulate the following defects: static and dynamic imbalance in 2 or more planes; errors of alignment parallel or angular; the weakening system resistance; defects of transmissions by straps defects of gearing; eccentricity; the phenomenon of resonance; defects in the camps of rolling bearings;
mechanical clearances;
defects of sliding of the camps; curved shaft;
rubbing rotor; problems of electric motors gearing faults. The stand can be a range of diverse sensors such as: accelerometers uni and triaxiale, medium and low frequency welting, for absolute measuring of vibration; proximity systems for measuring relative vibration; the speed and motion transducers for vibrations; vibration transducers with output 4-20 mA and dynamic output; transducers with TEDS function;
640
phase and speed sensors; temperature sensors with or without contact (thermo resistance, thermocouples, infrared); sensors for measuring the torque flow; force transducers; impact hammers for resonant frequencies determination; laser distance sensors for measurement of relative motion; laser sensors for measurement of realtime 2D and 3D.
Fig. 1. Stand for vibration study
Senzor de tura ie Speed sensor
Piezoelectric Accelerometre accelerometers piezoelectrice Proximity sensor Traductoare de proximitate
Fig. 2. Montarea Fig. 2. Mounting pe stand stand a traductoarelor for vibration de transducers vibra ii and i a senzorului speed sensors de tura ie
In Figure 2 is given the position on the stand of the displacement transducer, piezo contact accelerometers with magnet and speed sensor. Main technical characteristics of the stand:
Electric motor power: 0.75 kW; Power supply: 220 Vac stand; Power frequency: 50 Hz; The magnitude of the frequency domain [Lal 90]
641
generated by fundamental-stand [Vac 06]: 0,5 to 1 x 30-400 Hz (12000/24000 rpm); Minimum resolution: 0.01 Hz frequency, adjustable from operator interface; Overall dimensions: approx. L800xl400xh200 mm; Shielding protects: Adjustable, 150% of current roll call for 1 minute; Control mode: linear v/f Control; Optimized Vector Control; PID (positive, negative); PWM; Frequency settings: Potentiometer, external analog signal (0 ~ 5V, 0 ~ 10V, 0-20mA); Terminal keys, control logic from the PLC, PWM; START/STOP Control: Keyboard or Terminal; Possibility to change the time period for raising electric motor revolution up to fundamental frequency (frequency set): 0.1 ~ 3000 s; LCD Monitor: Display: frequency, revolutions of the engine or the 12. elastic pads, etc.
study of the rotor (rpm), output current, output voltage, past alarms, the system parameters; Simulation tools: elastic and rigid foundations; Transmission: by elastic coupling and rigid, belts, toothed gears. Main parts: 1. driving Motor 2. devices for simulating drag between rotor and stator, or in camps 3. uniform rotors 4. nonuniformity rotor 5. the order, measure and control box 6. motor shaft coupling elastic 7. bearings of sliding 8. rollover camps 9. motherboard 10. strap wheel 11. toothed gears
Stand
Laptop used for data analysis and interpretation Control panel and data recording
Fig. 3. Study of vibration using on-line monitoring systems
Depending on the system chosen and by connecting the sensors corresponding, on the stand can be conducted the following studies [Bay 99]: - influence of temperature on the roll at the fiability camp thermodynamic machines; - loading the bearings by simulating a friction in bearings or raising the upper casing; - influence of unbalanced rotor, static or dynamic; - determination of resonant frequencies of different types of rotors and their influence on cars;
- calculation of the thermal head and their influence on cars; - study of motion and vibration of the time in the camps of the camp and slip-Orbit's precession time zone in camp; - influence of sliding and dezaxed camps of axes on the consumption of electricity; - measuring the torque in the case of the bearings axles; - Study reaction on elastic foundations machines, namely rigid. Determination of critical speeds; - Relaxing stiffness of the system and its influence on the reliability of dynamic rotary machines;
642
- study of the reaction of the flexible rotor based on their gear and measurement of the arrow; - alignment of trees with the help of sensors or laser displacement transducer; - static and dynamic load balancing on the stand (1 plane-2 planes), etc. In order to achieve the proposed themes, vibrations sensors and those for technical
Electric motor Accelerometer cable Piezoelectric accelerometers
parameters fitted stand can be connected to different measurement and diagnostic equipment, such as: vibration analyzers, data collectors, monitoring systems, etc. online. In Figure 3 the study is presented on the stand by means of vibration monitoring systems, multichannel, online and Figure 4 shows the components necessary to carry out dynamic equilibration in two planes.
Speed and phase sensor Acquisition Unit Balancing software
USB cable
Control panel
Note:
Counterweights set
Electronic scale
Key kit
Measuring tools
Laptop
- On the stand is presented only the dynamic balance option, - The system includes up to 6 rotor correction method with the addition or removal of material, - The range of speeds is up to 12,000 rpm or 24,000 rpm with 1:1 transmission ratio in sheaves, - For higher speeds the transmission ration can be changed.
Fig. 4. Stand for vibration study balancing option
3. STAND FOR THE STUDY OF SHAFT HORIZONTAL ALIGNMENT Aligning rotating machinery dynamic owning the following mechanical couplings is one of the most important operations that apply to them in order to eliminate the defects of misaligning of shafts. In practice it was found that the operation of a rotor with defect of nonlinearity of shafts a long period of time lowers the service life of the bearings or sliding of the camps, the fatigue damage of trees, couplings, and the increase in energy consumption, excessive
[Ste 99] internal heating or mechanical destruction of tests, in the case of pumps, etc. and, consequently, damage the equipment. To study the phenomenon of nonlinearity of shafts horizontal was designed and made a stand shown in Figure 5. This stand is designed for studying methods of aligning with different alignment kits, such as: laser kits, kits with analog or digital clocks comparators, travel kits with sensors with or without contact, etc [Wic 05]. Main parts: motherboard; bearings with bearings; mechanical coupling with bolts; device for adjusting the camps in the vertical and horizontal direction; mounts bearings
643
with or without electrical resistors for the simulation of the thermal head, etc. On the stand can simulate different defects, such as [Don 02]: - nonlinearity radial - nonlinearity axial - combined nonlinearity - influence of thermal head on alignment - the influence of gaming on the alignment
Rolling bearing cases
correction of "soft foot" (the feet of the machine are not in the same plane with pedestal) radial and axial check heartbeat to elastic couplings.
Mechanical coupling with bolts
Bearings regulating device in the horizontal and vertical direction
Motherboard
Bearing supports with or without electrical resistors to simulate thermal expansion
Fig. 5. Stand for horizontal shaft alignment study
3. CONCLUSION Analysis of defects by studying dynamic rotating machinery vibration is a complex phenomenon that involves both theoretical knowledge and practical knowledge, acquired over time. This theme issues are of concern not only at the theoretical level, but also practically the front trying to contribute to the deepening of knowledge and issues related to the analysis of vibration fault, by studying and simulating mechanical and electrical defects in rotating machinery dynamic as well as horizontal shafts on misaligning, which were designed and executed two stands trial, necessary in the first place training, preparation and maintenance engineers and study the
dynamics of rotor machinery within the undergraduate laboratories. Recognizing the importance of preparing personnel serving equipment maintainability by making the two stands has sought to deepen, especially by maintenance engineers, to work on issues related to the analysis of vibration fault, by studying and simulating mechanical and electrical defects in rotating machinery dynamic as well as concerning horizontal misaligning shafts, which are particularly useful to study the dynamics of rotor machinery within the undergraduate laboratories. 4. REFERENCES
[Bay 99] Bayer, M., Fiabilitatea i mentenabilitatea sistemelor, Editura Bren, Bucureti, 1999.
644 [Don 02] Donald, E. Bently, Charles T. Hatch, Bob Grissom, 2002, Fundamentals of Rotating Machinery Diagnostics, Bently Pressurized Bearing Company. [Lal 90] Lalanne, M, Ferraris, G., Rotordynamics Prediction in Engineering, John Wiley and Sons, New York, 1990. [Lee 04] Lee, J., .a., An integrated platform for diagnostics, prognostics and maintenance optimization, IMS International Conference, Arles, France, 2004. [Ste 99] Steve Goldman, Vibration Spectrum Analysis, Terre Haute, 1999. [Vac 06] Vachtsevanos, G., .a. - Intelligent Fault Diagnosis and Prognosis for Engineering Systems, ISBN: 0-471-72999-X, Wiley, USA, 2006. [Wic 05] Wichtendahl, A., Ron Denton, Cost Effective Predictive Maintenance Solutions, Wilcoxon Research, Inc, 2005.
SIMULARE EXPERIMENTAL PENTRU IMPLEMENTAREA MENTENANEI PREDICTIVE Rezumat: n lucrare se prezint dou standuri experimentale pentru: studiul vibraiilor i simulrii defectelor mecanice i electric, precum i studiul alinierii arborilor orizontali. Lucrarea se ncadreaz n domeniul implementrii mentenanei predictive i proactive n industrie. Realizarea acestor standuri este necesar pentru simularea condiiilor reale de apariie a defeciunilor n vederea aplicrii corecilor ce se impugn, fr afectarea procesului de producie.
Marius BAIDOC, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.770.400; Clin Vasile LUMEI, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.742.030.565; Florin EPE-BOBESCU, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.547.258; Alina Sabina PACA, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.756.450.183. Mariana ARGHIR, Prof. Dr., Mech. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.264.401.657.
645

METHODS OF THE EQUIPMENTS ALLIGNMENT

Marius BAIDOC, Alina Sabina PACA, Florin EPE-BOBESCU, Clin Vasile LUMEI, Mariana ARGHIR
Abstract: The sculpture presents theoretical concepts and process-orientated, for the implementation of the programme required machinery alignment, in keeping 3 alignments and the calculation of "soft leg" types of maintenance, their applicability of rotary machineries. Key words: machinery alignment, soft leg, maintenance.
1. GENERAL KNOWLEDGES The alignment plays an important role in operation dynamic rotary machinery [Lal 90]. The alignment unfit to lead the machinery bearings deterioration, fatigue trees, damaging connections, consumption of energy increased, longer excessive internal destruction mechanical interspacing and, implicitly, damage equipment [Bay 99]. Not alignment represents departure from the same axially to two or more cars, which are coupled with one another using connection rigid or elastic (axis of rotation machines are not the same line). Execution with a quality show following advantages: cut stops - output growth; reducing the consumption of energy energy expenditure reduction; forces cut the ball bearings to increase service life of camps; decrease use vibration - reducing expenses for maintenance. reducing wear joint-increasing duration of availability of machinery. Not alignment is decomposed into two components: Not alignment radial - or the parallelism, representing the distance between those two axes of symmetry of trees and is measured in [mm] or [m].
Not alignment angular [Don 02] represents the angular relative position of symmetry axes and measured in [mm/100mm] or [m/mm]. 2. THE MAIN FACTORS INFLUENCE THE ALIGN THAT
- The alignment may be affected by the way of fixing the foundationless of machinery [Ste 99]. A foundationless weakened, dirty, cracked or damaged may lead to given the "leg soft", and hence the not alignment. - A thermal is another factor that determined not alignment and to be taken into account when he performed matching warm. These growers influence in direct alignment, i.e. temperature rise machinery due their heating during the operation or due to factors such as: ambient environment, agents working generation with these machines [Lee 04]. - The pipelines link, entering the car component, is another factor which may affect the position of trees. - The use of modern methods of matching the machinery dynamic rotary leads to increased production, the length of operating the plant, reducing costs of maintenance and increase safety in operation.
646
3. THE METHOD RADIAL
OF
MATCHING
In calculating parameters of matching the radial method is considered the next configuration of cars with axle horizontally, as shown in Figure 1.
MF MM
Fixed Machine Masina fixa
Masina mobila Mobile Machine
D E F
B C
Fig. 1 Scheme of Fixed Machine Mobile Machine System
Sensors needed carrying alignments are members over the mechanical machinery or the axes, i.e. one sensor on each half linkage. Consider the following notations technical for Figure 1: The sensor mounted on half linkage MF fixed machine-sensor mounted on mobile half linkage MM A - machine-distance sensors [mm] B between - the distance from the sensor Mf the soles of the machine mobile face (incident of fixing the machine on pedestal) [mm] C - distance from the sensor Mf soles to the rear of the machine mobile [mm] D - distance from the sensor Mf the soles of the [mm] fixed the car's E - distance from the sensor Mf soles to the rear of the machine fixed [mm] F - distance between rear feet fixed the car and feet in the back of[mm] mobile machine for matching cold machine is sufficient introduction program in the distances A, B and C. 3.1. Parameters calculation for matching Relations for calculating alignment in the cold in direction vertical and horizontal are listed below:
Df =
Ds =
B (Om + O f ) A O f 2A
C (Om + O f ) A O f 2A
(1.1) (1.2) (1.3) (1.4) (1.5) (1.6) (1.7) (1.8) (1.9)
V f = VMF + S f
Vm = VMM + S m
Af =
B (Vm + V f ) A V f 2A
As =
C (Vm + V f ) A V f 2A
d (V f + Vm ) A V f 2A d (O f + Om ) A O f 2A
Rv = Ro =
Uv =
(V f + Vm ) 2A
2A
100
100
Uo =
(O f + Om )
(1.10)
where: MF - notation car fixed. MM - notation mobile machine;
Vf - measured in vertical direction on the clock comparator mounted on the car fixed MF (mm); Vm - measured in vertical direction on the clock comparator mounted on mobile machine MM, (mm); VMF - measured in vertical direction to the sensor mounted on the car fixed MF (mm); VMM - measured in vertical direction to the sensor mounted on mobile machine MM, (mm); Oh - measured in horizontal to the sensor mounted on the car fixed MF (mm); Om - measured in horizontal to the sensor mounted on mobile machine MM, (mm); SF - arrow rod rack fixed to the Finance Ministry, where alignment with watches comparator, (mm); SM - Arrow rod rack set in mm, where alignment with watches comparator, (mm); As - movement in vertical direction of part of the machine mobile MM, (mm); Af - displacement in vertical direction of the party in the back of mobile machine MM, (mm); Ds - displacement horizontal part of the car back mobile MM, [mm]; Df - displacement horizontal part of the car back mobile MM, [mm]; Ro - not alignment 3 values horizontal, (mm); Rv - not alignment 3 values the vertical direction, (mm); Uo - value not alignment angular horizontal direction, (mm); Uv - value not alignment angular in vertical direction, (mm). In the case alignment machinery with sensors laser support the fixing of their couple or axes is different from the point of view geometric and arrow is considered to be zero. 3.2. Soft leg calculating The programme also includes the option of calculating the leg (stalk) soft and the relationship to it are listed below:
Pfd =
B ( PfdMM + PfdMF ) A
647
PfdMF
(1.11) B ( PfsMM + PfsMF ) Pfs = PfsMF A
Psd = Pss =
C ( PsdMM + PsdMF ) PsdMF A C ( PssMM + PssMF ) PssMF A
(1.12)
(1.13) (1.14)
where: Pfd - Addition of material for correction stalk soft soles MM face the right, (mm); Pfs - Convention-Addition of material for correction stalk soft soles MM face the left, (mm); Psd - Addition of material for correction stalk soft soles MM back at right, (mm); Pss - the addition of material for correction stalk soft soles MM back to left, (mm); PfdMM-value sensor indication for the calculation MM stalk soft face right (mm); PfsMM-value sensor indication for the calculation MM stalk soft face left, (mm); PsdMM-value sensor indication for the calculation MM stalk soft back right, (mm); PssMM-value sensor indication for the calculation MM stalk soft back left, (mm); PfdMF-value sensor indication for the MF calculation stalk soft face right (mm); PfsMF-value sensor indication for the MF calculation stalk soft face left, (mm); PsdMF-value sensor indication for the MF calculation stalk soft back right, (mm); PssMF-value sensor indication for the MF calculation stalk soft back left, (mm).
3.3. The heat growing calculation
For matching warm reasonable machinery is necessary for each calculation thermal growing soles to the machinery, according relations: d s1 = (Tar T0 ) H MF 106 (1.15) 6 d f 1 = (Tar T0 ) H MF 10 (1.16)
648
d s 2 = (Tar T0 ) H MM 106
(1.18) where: ds1- thermal particular back; Mf, (mm); df1- thermal face particular Mf, (mm); ds2- particular heat back mm, (mm); df2- particular heat back mm, (mm); Tar- mean temperature corresponding to each soles in hand, (C); T0- ambient temperature, (C); - coefficient of particular heat (m/mC); HMF- height center axis bar Mf the pedestal, (mm); HMM- height center axis bar mm from the pedestal MM, (mm).
a) Indication sensor MM for matching hot and cold in the simulation MM
VF 2 = 2 (E ds2 C d f 2 )
d f 2 = (Tar T0 ) H MM 10
(1.17)
MM displacement for the thermal growing and simulation is: B (VF 1 + VM 1 ) A VF 1 d fc1 = 2 A (1.25) C (VF 1 + VM 1 ) A VF 1 d sc1 = 2 A (1.26)
d fc = d fc1 + d fc 2
d sc = ( d sc1 + d sc 2 )
(1.27) (1.28)
VM 2 =
(1.19) 2 ( A B ) d s 2 + ( C A ) d f 2 )
CB
CB
where: VF1- MM values indication sensor in the simulation, (mm); VM1-Values indication Mf sensor in the simulation, (mm); dsc1- MM movement back in the growing thermal (mm); dfc1- MM movement towards the thermal growing, (mm); dsc- MM movement back in the simulation, (mm); dfc- MM movement towards the simulation, (mm).
c) Final senzor indication, in case of alignment with comparative watches, for the hot align MF-MM
VFTC = VF 2 + VF 1 VMTC = VM 2 + VM 1
(1.20) MM displacement for the thermal growing is:
d fc 2 = d f 2
d sc 2 = d s 2
(1.21)
(1.29)
(1.22) where: VF2- MM values indication sensor in the simulation, (mm); VM2- Values indication Mf sensor in the simulation, (mm); dsc2- movement back mm, (mm); dfc2- moving front mm, (mm).
b) Indication Mf sensor for matching hot and cold in the simulation Mf 2 ( E d f 1 D d s1 ) VF 1 = ED (1.23) 2 ( A + D ) d s1 ( A + E ) d f 1 ) VM 1 = ED (1.24)
VFT = VFTC V f
(1.30) (1.31) (1.32)
VMT = (VMTC Vm )
MM displacement for the thermal growing and simulation is: A fc = A f + d fc (1.33) Asc = As + d sc (1.34) where: VFTC- MM values clock indication the simulation, (mm);
649
VMTC- Values indication clock Mf the simulation, (mm); VFT- real values of the deviations Mf the simulation, (mm); VMT- MM values real irregularities in the simulation, (mm); Afc- MM movement towards the simulation, (mm); Asc- MM movement back in the simulation, (mm).
4. THE METHOD OF RADIAL-AXIAL MATCHING
In case the machinery alignment method radial-axial a sensor will be mounted in the axial direction and another in the radial direction as in the 1 figure.
4.1. The alignment parameters calculation for
Relations for calculating parameters are listed below:

D f = Oa B Or + A 2
B + C Or Ds = Oa + A 2
(1.35) (1.36) (1.37) (1.38)
A- sensor mounted in axial direction; d- diameter assembly horizontal sensor; Ds- displacement horizontal part of the real of the machine mobile MM, [mm]; Df-displacement horizontal part of the car back mobile MM, [mm]; Or- measured in horizontal to the sensor mounted radially (mm); Oa- the measured in horizontal to the sensor mounted axially (mm); Vr- measured in vertical direction to the sensor R depending on the arrow (mm); Va or VA- measured in vertical direction to the A sensor, (mm); Vr- measured in vertical direction to the sensor R, (mm); S- arrow rod rack of fixing the sensors, [mm]; As- movement in vertical direction of the party in the rear of the machine mobile MM, (mm); Af- displacement in vertical direction of part of the machine mobile MM (mm); Ro- no alignment 3 values horizontal, (mm); Rv- no alignment 3 values vertical direction, (mm); Uo- value no alignment angular horizontal, (mm); Uv- value no alignment angular in vertical direction, (mm).
4.2. Calculating "soft leg"
Vr = VR + S
Va = VA
V B Vr Af = a + A 2 As = V a ( B + C ) Vr + A 2 Rv = O Ro = r 2 Vr 2
The relations for the soft leg are:

Pfd = B (Pfd 1 + Pfd 2 ) A Pfd 1 Pfs1
(1.39) (1.40) (1.41) (1.42) (1.43)

Pfs =
B (Pfs1 + Pfs 2 ) A
(1.45) (1.46) (1.47) (1.48)
B (Psd 1 + Psd 2 ) Psd = Psd 1 A B (Pss1 + Pss 2 ) Pss = Pss1 A
Uv =
Uo =
Va 100 A
Oa 100 A
4.3 Calculating of the thermal growing
The corresponding relations are: (1.44)
where: R- sensor mounted in the direction radial;
d s1 = (Tar T0 ) H MF 106
(1.49)
650
d f 1 = (Tar T0 ) H MF 106
(1.50) (1.51) (1.52)
d fc = (d fc1 + d fc 2 )
d sc = (d sc1 + d sc 2 )
(1.63) (1.64)
d s 2 = (Tar T0 ) H MM 106
d f 2 = (Tar T0 ) H MM 106
a) Indication of sensors R and A for the alignment at hot and cold following the MM simulation:
(1.54) The MM displacement following cold thermal alignment: d fc 2 = d f 2 (1.55) d sc 2 = d s 2 (1.56) where: Vr2- values indication sensor R the simulation, (mm); Va2- values sensor indication of the simulation, (mm); dsc2- movement back MM, (mm); dfc2- moving front mm, (mm).
b) Indication sensors R and a matching hot and cold in the simulation MF
B (d s 2 d f 2 ) Vr 2 = 2 d f 2 C A (d s 2 d f 2 ) Va 2 = C
(1.53)
where: Vr1-Values indication sensor R the simulation, (mm); Va1-Values sensor indication of the simulation, (mm); dsc1-mm movement back in the thermal growing (mm); dfc2-mm movement towards the thermal growing, (mm); dsc- MM movement back in the simulation, (mm); dfc- MM movement towards the simulation, (mm).
c) Indication final sensor, the alignment with comparator clocks, for matching warm MF- MM
VrTC = Vr 2 + Vr1 VaTC = Va 2 + Va1 VrT = (VrTC Vr )

VaT = (VaTC Va )
(1.65) (1.66) (1.67)
D (d s1 d f 1 ) Vr1 = 2 d f 1 E D (d s1 d f 1 ) Va1 = E
(1.59) (1.60)
The MM displacement following cold thermal alignment: B V d fc1 = Va1 + r1 A 2 (1.61) B + C Vr1 d sc1 = Va1 + 2 A (1.62)
(1.68) The MM diplacement following the thermal growing and simulation: A fc = A f + d fc (1.69) Asc = As + d sc (1.70) where: VrTC- values indication clock R the simulation MM, (mm); VaTC- values clock indication of the simulation; MF, (mm); VrT- real values of the deviations R the simulation MM, (mm); VaT- real values of the deviations of the simulation; MF, (mm); Afc- MM movement towards the simulation, (mm);
Asc- MM movement back in the simulation, (mm).

5. ALIGNMENT OF THE HORIZONTAL SHAFTS WITH THE PROGRAMM POCKETPC ALINIERE
For carrying out the radial method costs by horizontal shafts was carried out a program installed on POCKETPC, named ALINIEREA, which includes the following options: - given the detection leg soft; - matching cold; - calculation thermal growing; - matching hot; - alignment simulations in vertical and horizontal directions; - alignment quality of elastic connections, function by number of machine revolving; - graphic design of align; - generate automatic report or technical documents of matching. Program has been tested successfully on various machines align shafts rotary who is inside of mechanical chain, as have: electric motor - reducing, engine - pumps, groups acting bands, engine-fan etc. A special case represents align rotary machines who work at the temperatures where appear heat growings. The heat growings appears, generally in vertical, but they may also occur in horizontal because of machine configuration, lateral pipelines to link heated, no uniform heatingmachine due process technology etc. The programme PocketPC-ALINIEREA shall have the opportunity to carry out both matching warm in vertical and horizontal.
6. CONCLUSIONS
amplitude function of frequency and function of time; figure Lissajous (orbit movement of precession); amplitude phase depending on speed of shaft. To that end, it requires careful analysis of phase, and as much as possible, should be determined form distorted the tree. Unlike the vibration analysis systems fever, liabilities, situation in which he gave up easily to the analysis phase, where cars with rotor, this parameter is the most important, because he is the link between the movement of rotation and the vibration (movement of precession). Also, the variation phase and amplitude of vibration, in the field of interest, helped to determine critical revolving of a rotor, speeds that are extremely important for the smooth operation of a car dynamic. Practice has shown that the main defect appeared in the dynamic rotary machines are disturbing the balance static and dynamic rotors. To eliminate this defect imposing balancing is carrying out in situ or sitting, which, in fact, is one of the most important action that is taking place in the compartments for maintenance. Equilibration means by which action corrective mass distribution of the verified rotor and, if necessary, is corrected, so that the vibration had one and forces from camps to learn the limits prescribed for appropriate alone functional revolving speed.
651
Another defect often encountered in dynamic rotary machines is no alignment, which represents a departure from coaxiallity two or more shafts, which are coupled among them using connections rigid or elastic. The corrective action regarding this type of defect is matching. Implementation of matching (alignment) tools are a necessity for any enterprise or industrial company, regardless of what category is part, this concept the opportunity to predict defects and to intervenes promptly on machinery, avoiding the kind of failure, with direct effect on production.
In order to achieve a comprehensive study of dynamic a car with rotor measurement is needed following parameters: amplitude function frequency; amplitude function of time (the form of wavelength);
652
7. REFERENCES
[Bay 99] Bayer, M., Fiabilitatea i mentenabilitatea sistemelor, Editura Bren, Bucureti, 1999. [Don 02] Donald, E. Bently, Charles T. Hatch, Bob Grissom, 2002, Fundamentals of Rotating Machinery Diagnostics, Bently Pressurized Bearing Company. [Lal 90] Lalanne, M, Ferraris, G., Rotordynamics Prediction in Engineering, John Wiley and Sons, New York, 1990.
[Lee 04] Lee, J., .a., An integrated platform for diagnostics, prognostics and maintenance optimization, IMS International Conference, Arles, France, 2004. [Ste 99] Steve Goldman, Vibration Spectrum Analysis, Terre Haute, 1999.
METODE DE ALINIERE A UTILAJELOR Rezumat: Lucrarea prezint notiunile teoretice i aplicative, pentru realizarea unui program necesar n alinierea utilajelor, n meninerea alinierii radiale i a calulului de picior moale pe tipurile de mentenan, aplicabilitatea acestora mainilor rotative.
Marius BAIDOC, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.770.400; Alina Sabina PACA, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.756.450.183; Florin EPE-BOBESCU, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.745.547.258; Clin Vasile LUMEI, PhD Stud. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.742.030.565; Mariana ARGHIR, Prof. Dr., Mech. Eng., Department of Mechanics and Computer Programming, Faculty of Machines Building, Technical University of Cluj-Napoca, Tel: (+)40.264.401.657.
653

ENTRY PREDICTION IN THE TECHNICAL UNIVERSITY OF CLUJ-NAPOCA

Lia CHIOREAN
Abstract: The present paper approaches an important component of university management, which is represented by university admission. Starting from the number of places provided for the entrance examination in the Technical University of Cluj-Napoca and the extent to which they were occupied in the interval 2005-2010, prognoses regarding the number of candidates for the year 2011 were developed, using specific methods belonging to the theory of probability and mathematical statistics. The study refers only to the bachelor level admission for the entire university. The results of the prognoses were validated by comparison with the data collected for the entry in September 2011. Key words: entry, prediction, statistics, regression, candidate
1. INTRODUCTION The present paper aims at making use of methods found in mathematical statistics to make a prediction of the number of candidates to take part in the entrance examination in the Technical University of Cluj-Napoca. The analysis of the data regarding the interval of years 2005-2010 with respect to the number of entry candidates and the number of places provides through comparative studies, syntheses and regression functions other data of interest. The results of the prognoses were compared to the actual data collected after the 2011 entry to demonstrate the method validity as well as its advantages and disadvantages. The conclusions of the study can bring forth solid information that could be used in defining further strategies developed by the management of the university with respect to planning future number of places as well as education marketing activities. 2. MATERIALS AND METHODS Starting from the premise that the objective of the strategic management in the universities
lies in reaching competitive advantage in a given competitive context [1], the university has to pay much attention to attracting students. Shattock states that: the best guarantee for the financial security may be an academically successful university which attracts large numbers of students, educates them well and has good results in[2]. John Naisbitt [3] has the opinion that people survive only through his capacity to act now, in the present moment, based on former experience and thinking of potential future consequences. Assuming future, people make the present bearable and give significance to the past. Past, present and future alternatives interconnect in anticipating and predicting would-be actions. This paper presents a number of analyses and comparative evaluations regarding the number of candidates enrolled for admission in the period under investigation and the number of places for entrance examination, as well as the extent to which these places were occupied both for budgeted places and places with tuition fees. Fort the study of the territorial distribution of the candidates, the authors used the Pareto analysis, a classification statistical method also
654
known under the name of rule 80/20. This rule states that 20% of the resources generate 80% of the most important results. The rule also mentions the great lack of balance between inputs and outputs, i.e. between cause and result. The Pareto diagram is a tool that helps distinguish significant issues from possible issues that ranks information function of priority and importance and provides actual information that can be used in decision making while clearly and suggestively indicating the direction of action. Prediction as support of the decision-making process represents a process through which the development of specific future events is anticipated and indicator evolution can be estimated. Prediction has to be but at the basis of finding out the needs, potential means and further development of the university. The university strategy needs to be defined capitalising the data used from the prognoses related to domains of interest and data obtained from own system along an interval of time that should be as long as possible. Prognoses are based on specific studies related to the theory of probabilities and mathematical statistics. The present study made use of the regression functions and probabilistic calculations. The regression function is a mathematic expression deduced after applying experimental data that approximate the dependence between two or more variables. In this case, the variable is represented by the number of candidates for admission in every academic year. Given a set of known data for a number n of years, simple mathematical expressions are looked for as well as a related graphical representation, trying to find the curve passing nearest to the defined points. The expression deduced in this manner is called regression function. In cases where the relationship depends upon only one variable, the shape of the regression function can be determined with the method of the least squares. Using the dedicated software CurveExpert, for the case of a single variable, we looked for the regression function that best approximates the set of experimental data, so that the values calculated for the prediction year shall be as close as possible to reality. The
software specialised in curve processing gives the possibility of selecting the function giving the best approximation from the functions obtained. It can use a variety of function types covering the majority of mathematical models found in practice and applied in the Windows operation environment. This software was developed by Daniel Hyams [4]. The approximation quality is given by the estimation standard error and the correlation coefficient. The estimation standard error is determined with the relationship:
S=
[y
i =1
f (x i )]
n nc
(1)
where: n - the number of experimental points; xi, yi the coordinates of the experimental points; f the shape of the regression function under study; nc - the number of function coefficients. The value of the correlation coefficient yields from the relationship: S t Sr (2) R= St where: S t = (y y i )2
i =1 n
(3)
y being the arithmetical average of the ordinates of the experimental points yi, and Sr the fraction numerator from relationship (1). The standard error (S) defines the spreading of the experimental points around the regression function line. The more the points come closer to the diagram, the more the standard error reaches zero as value. The correlation coefficient (R) represents a more accurate measure of the regression function quality, found when the standard deviation is relatively large. The more its value nears 1, the more the regression function is seen to better approximate the set of experimental points. After introducing the experimental data, the CurveExpert program displays the graph of the regression function and the user has more options to choose from:
to visualise the mathematical expression, the coefficients of the regression function and intermediate results; to make calculations using the determined relationship; to save the coordinates of the calculated points from the line of the regression function. The user can request the performance of the calculations for all the types known to the software and then, can select the line and the numerical data corresponding to the studied functions to be visualised. In case none of the regression function types known to the software fits, the user has the possibility of defining the own type of function. After fin ding with the regression method the estimate value for the current academic year, a correlation coefficient was applied by making the ratio between the potential candidates from the current academic year and the preceding academic year, that is value v. Going on with the extrapolation of the dynamic series with value (v) previously found and using the student distribution, we calculated the size of the prediction field, its minimum and maximum values respectively:
655
3. RESULTS AND DISCUSSION
The analysis made in the Technical University of Cluj-Napoca with respect to the degree of enrolling students after the admission, highlights a special performance of the university concerning the places financed by the budget. In Figure 1, it is evident that the degree of enrolment is 100% all along the period of time under investigation, with a very insignificant difference for the years 2002 and 2003.
3000 2500
2000
1500
1000
500
Year Admitted
2002 1820 1802
2003 1830 1824
2004 1820 1820
2005 1835 1835
2006 1930 1930
2007 2010 2010
2008 2400 2400
2009 2500 2500
2010 2600 2600
2011 2683 2683
Number of places
Fig.1. Degree of budget places enrolment in the period 2002-2011
max = v + t q
; min = v t q
(4)
where: - tq is the corresponding value from the table regading student distribution for=0,05 and a number of degrees of freedom of f=n-1, - n is the number of experimental points; - =
n
As for the tuition fee places, they were occupied to various extents. The analysis demonstrates that in the years 2003, 2007 and 2008 most places with tuition fee were occupied, while, the minimum number of tuition fee places is found in 2010, with a percentage of 11.87% (Figure 2).
% occupancy
60,00
51,91
50,00
50,70 46,18
(dif1 u )
+ ... + (dif n u ) ; n 1
2
49,15
(5)
41,97
40,00
39,82 34,62 27,67 27,16
n - dif i is the difference between the experimental value and the calculated one, by means of the regression function at point i. The Student distribution is normally used to check the statistical hypotheses based on results obtained from small volume samples, n 30, or when the approximation with one single normal distribution does not fit.
- u=
dif i
i =1
30,00
20,00
(6)
11,87
10,00
0,00 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011Year
Fig.2. Degree of tuition fee enrolment in the period 2002-2011
Interesting to notice the distribution as per counties; the admission candidates come from all the 42 counties in the country, in various percentages, most options originating in the counties that neighbour Cluj County.
656
As there is a number of significant counties providing a large number of admission candidates, we tried to use a Pareto diagram for this purpose (ABC). In Figure 3, the diagram for year 2010 is given, in this respect. A similar distribution was found for the rest of the years in question.
1200 1000 Candidates 800 600 400 CLUJ 200 0 BISTRITA-NASAUD 60% 100% 80% Cumulative %
previous years (Figure 4) and the CurveExpert software, we looked for estimate results for the tendency of becoming candidates for the admission in our university, taking into account both an optimistic and a pessimistic scenario. On the basis of these data, the CurveExpert software led to a regression function, of Gauss (x b )2 2 type y = ae 2c , which graphically represented most faithfully the given points (Figure 5).
470 450 0
MARAMURES
HUNEDOARA
SATU MARE
40% COVASNA PRAHOVA SUCEAVA BRASOV NEAMT 20% 0%
HARGHITA
MURES
SALAJ
BIHOR
ALBA
SIBIU
0 0
Vital Few
Cumulativ %
Cut Off %
[42]
Fig.3. Diagram ABC for the admission candidates in 2010/counties of origin
CANDIDATES
County Useful Many
430 410 390 370 350 330
0 0
Analysing the Pareto diagram plotted for year 2010 (fig.3), one can notice that the majority of the admission candidates (82,93%) come form eight candidates: Cluj, BistriaNsud, Alba, Slaj, Maramure, Satu-Mare, Mure and Sibiu, representing 21,05% of the candidates county of origin. Except for the county of Cluj, with a normal first position, counties Bistria-Nsud, Slaj, Satu-Mare and Alba are also included here, as extensions of the Technical University of ClujNapoca can be found there. In this way, it is demonstrated that the strategy of establishing extensions in these places contributed to enrolling many students from these areas.
CANDIDATES
0 0 0 2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
Fig. 5. The regression function corresponding to the evolution of the number of admission candidates in the Technical University of Cluj-Napoca
5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0
YEAR
The calculated coefficients are: a = 4285,4841; b = 2007,0741; c = 7,8421526; The standard error is: 317,1626039; The correlation coefficient is: 0,6971313. The value calculated for the year 2011 is y=3780,76. By means of the CurveExpert software, the calculated values for every point on the line are highlighted, so that the differences difi between the experimental and calculated values can be obtained (Table 1). In order to calculate the width of the prediction field, relationships (5), (6), respectively (4) are applied, with a correction coefficient of 0,9456. This coefficient was calculated dependent on the number of potential candidates, according to the statistical data related to the number of children born between 1985 and 2003. For the year 2011, a
2002 3648
2003 3693
2004 3509
2005 4629
2006 3997
2007 4266
2008 4477
2009 4144
2010 3908
CANDID.
Fig.4. Evolution of admission candidates number for the Technical University of Cluj-Napoca
With the statistical data regarding the number of candidates in the admission in the Technical university of Cluj-Napoca in the
657
Wide field of forecast: 422
211
number of 260.393 children born in the year 1992 was considered and reported to the number of children born in 1191, i.e. 275.275.
Tab.1 Calculated values for the number of candidates Year 2002 2003 2004 2005 2006 2007 2008 2009 2010 Candidates 3648 3693 3509 4629 3997 4266 4477 4144 3908 Calculated 3476,1 3744,5 3968,56 4138,19 4245,48 4285,29 4255,72 4158,18 3997,35 Diff. 171,90000 -51,50000 -459,56000 490,81000 -248,48000 -19,29000 221,28000 -14,18000 -89,35000
Predicted maximum level: 3787
probability of employment: = 224 100 = 53,08 %

422
Projected average: 3576
Number of places offered for admission : 3563

211
vacancy risk: 46,92 %
Minimum expected level: 3365
Fig.6. The probability to occupy the admission places in the admission session 2011/2012
For q = 0,05 and a number of degrees of freedom f = n 1 = 8 , one obtains from the student distribution table, tq=2,306. With these data, one finds the width of the prediction field calculated for the year 2011: max.=3787, and min.=3365. If one takes into account the number of places offered for the year 2011/2012, and defined by the University Senate according to the demands of the faculties in the university and approved by the Ministry of Education, Research, Youth and Sports, namely 3563 (of which 2683 places financed from the government budget and 880 financed by the students themselves), one will notice that the number of places is situated below the average value predicted, though very close to it, which could lead to a probability of occupying the places for admission in a proportion of 53,08 % from admission places offered (Figure 6). The probability of occupy the admission places (PO) is calculated with relationship 7: max cs PO = 100 , (7) L where: max, is the maximum value of the prediction field; cs the number of places offered; L the width of the calculated prediction field.
If one takes into consideration only the admission number for the budget places, it is visible that the number of predicted candidates exceeds the figure offered, meaning that all the places will be covered in the 20111 admission After the admission in September 2011, it was shown that predictions were right, all the budget places were fully occupied, the tuition fee places being covered in a proportion of 27,16 %, which is beyond expectations. At the level of faculties, the prediction can be used as an instrument only if the admission conditions remained constant for many years in line, and this is the case of the Faculty of Automation and Computer Science. For the rest of the faculties, the predictions do not exhibit the same relevance because of various reasons. The most significant cause lies in not observing the rule that past trends extend in the future, interruption elements appearing due to the fact that they gave up an entrance examination based on a written test in favour of documents provided by the candidates as well as to the conservative opinions of some of the faculties with respect to the predicted number of places. Among these, one can enlist the Faculty of Architecture and Town Planning, where there has been a high level competition every year, though the number of places offered has remained relatively small, and the candidates main concern lies not in occupying the places, but in getting a good entrance score or obtaining a budget place or a tuition fee place, as a candidate.
658
4. CONCLUSIONS
Statistical processing cannot take into account all the variables that influence candidate decisions. In this kind of prediction, the study of the budget places degree of occupation seems to be nearer to reality as in the case of tuition fee places subjective factors and financial means of candidates need to be taken into account especially in relation to present-day economic situation in the country. Prediction cannot, then, take into account unpredictable conditions such as changing of entrance examination requirements, introduction of new specialisations, changes in legislation, changes in advertising educational programs, economic crises, overturning of circumstances, as well as relevance of considering only the fist option expressed by the candidates. Other aspects that were not considered are the baccalaureat scores, which actually have quite a big influence for the candidates decision. However, experience has proved that there is no direct connection between the two components mentioned above. Hence, the conclusion could be that better trained highschool graduates make for the university, in our case those who pass their final highschool examination during the first session of exams. But it was proved that the concrete situation produced after the entrance examination for the academic year 2011/2012 came very close to the predictions presented here. The budgeted places were occupied in full, those with tuition
fee in a proportion of 27,16 %, a figure that exceeded expectations. The aim of the prognoses should consist in stimulating those entities whose prediction and forecasting is negative, to enable them develop strategies that could cancel negative trend producing factors; in the case of entities that possess a positive trend, the prediction could help them plan future number of places and capacity of education devices, in correlation with market demand. Our university strategy consisted in attracting as many students as possible and of occupying first the budget places every year. At the end of the admission session, in case some faculties were left with vacant places, these vacant places were redistributed according to the students demands to other faculties. The study here concerns only Bachelors level. The study may not have the same relevance for Masters students, first because the latter do not have a history as the second bologna cycle began only in 2009 in the Technical University of Cluj-Napoca.
5. BIBLIOGRAPHY
Nicolescu, O. (coord.), Strategia universitii. Metodologie i studii de caz, Bucureti, Editura Economic, 2007. [2] Shattock, M., Managing successful universities, Society for research Higher education & Open University, New York, 2003. [3] John Naisbitt, Megatendine, Bucureti, Editura Politic, 1989. [4] http://www.curveexpert.net
[1]
PROGNOZA ADMITERII N UNIVERSITATEA TEHNIC DIN CLUJ-NAPOCA Rezumat: Lucrarea abordeaz o component important a managementului universitar, admiterea n universitate. Pornind de la informaiile referitoare la numrul de locuri scose anual la concursul de admitere n Universitatea Tehnic din Cluj-Napoca i gradul lor de ocupare din perioada 2005-2010, prin metode specifice din teoria probabilitilor i statistica matematic, s-au efectuat prognoze cu privire la numrul de candidai care se vor prezenta la admitere n anul 2011. Studiulde fa s-a ocupat doar de ciclul de licen, la nivel de universitate. Rezultatele prognozelor au fost validate prin comparaia acestora cu datele culese n urma admiterii ncheiate n luna septembrie 2011.
Chiorean Lia, engineer, University chief secretary, The Technical University of Cluj-Napoca, The Rectors Office, E-mail: Chiorean.Lia@staff.utcluj.ro, Office telephone: 0264-202209, Address: Calea Dorobanilor nr.89, bl.X3, Sc.1, Ap.8., cod:400609, Cluj-Napoca, Home telephone: 0264410876
659

MACHINING PROCESS MODELING

Daniela Carmen KOUKACH, Grigore Marian POP, Dan PREJA
Abstract: In the first part of the article on modeling of machining process with the finite element method (FEM) as a modern approach in investigating the processes of cutting, is presented the current state of research on thermo-mechanical materials used for modeling the cutting process, breaking criteria and fracture modeling methods, meshing and remeshing strategies of the finite element (FE) networks, modeling friction processes between piece and tool, modeling heat transfer between tool and chip and the tool wear model. The second part of the article presents the results of the research, performed during an internship in the Machining Institute at Stuttgart University, Germany. Key words: modeling, finite element method (FEM), breaking criteria , FE networks
1. INTRODUCTION The role of modeling manufacturing processes becomes universally recognized in the industry, with increasing of the competition in the area of product and process design. To achieve high productivity in the cutting processes there is important, besides the selection of the machine, tool, cutting conditions and process sequence, a complete analysis of all the requirements and parameters needed to develop and optimize the processes and the cutting tools which can be made only by modern methods of simulation. The modeling is mainly based on the finite element method, for which there is a wide range of software on the market. These programs provide the calculation and simulation of the cutting process both 2D and 3D in a satisfying time. However, so far some did not succeeded in developing a comprehensive calculation model, generally valid, for the cutting process cutting, given the complexity of the cutting process.
2. MODELING WITH THE ELEMENT METHOD (FEM) 2.1 Overview of the cutting process
FINITE
The plastic deformation process for the orthogonal cutting is complex, on an extended temperature range and high levels of plastic deformation coeficients, creating special modeling difficulties in establishing a viable model. At first, the cutting layer suffers strong plastic shear deformations in the primary deformation zone I (Fig. 1) around the OC shear plan, additional shear occurs in the smaller dimension area II, at the tool-chip interface OE. The finite element method (FEM) allows the modeling of the chip forming mechanism with complex problems relating the temperature distribution, strain distribution and the field of plastic deformation produced by thermo-elasto-plastic loads specific to the orthogonal cutting.
660
Fig.1. Orthogonal cutting model with continuos chip forming
The temperatures are very different in the primary and secondary cutting areas. The temperature field is the result of the thermal conductivity process, started from heat sources formed in the deformation process and friction on the contact surface (isothermal deformation) or adiabatic heating (adiabatic deformation). At the breaking forming proces witch generates the chip at small and medium cutting speeds, the temperature in the primary cutting area is relatively low (for steel AISI 1045-DIN 1.1191, is about 200C). This is also characteristic for other materials. At high cutting speeds, the heat generated in the deformation process will not disperse, resulting in a considerable increase of the temperature in the primary cutting area, and a considerable drop of the material strengh. The result is a chip forming characteristic to the adiabatic deformation. The secondary cutting area temperature is higher then in the primary area due to heat generated by friction in the tool-chip interface. In this area the temperature can reach, during the cutting process, the melting temperature of the processed material Tm . In general, we can confirm that during the cutting process, the materials to be processed go through visible changes. These changes differ very much in the cutting areas and lead to a mechanical qualities change of the material. The mathematical description methods of flow curves can be classified into: tabular, with neural networks, empirical and semiempirical. 2.2 Thermo-mechanical models of materials used in the cutting process modeling [3] Stress and deformation fields describe separatly one dynamic and one cinematic state of the particles of a continuous environment. So
far, however, there is no universal theory that would merge tensions, deformations, time and temperature to achieve a complex charging scheme to reach a specific concordance between calculations and experiments. The material model for metals is typically determined by the additive decomposition of elastic and plastic deformations having an associative law. The law imposes the requirement of orthogonality of the vector that refers to the plastic deformation growth into the flowing area. This requirement is achieved by the flow condition by v. Mises and by two scalar ecuations, that connect the strains and deformations of the material. The first of the two scalar equations, called the equation of state, connects the hydrostatic presure:
p = 1 3 tr ( )
(1)
with relative volume change:
p = p ( , T )
(2)
The second scalar equation, called the equation of determination or flow curve, contains a part of the material flowing conditions. It connects the yield point S with the hardening parameters q , the deformation speed intensity & p , temperature T and time :
& p , T , ) S = S ( q, (3)
To establish the determination equation, are enough experimental data that can be obtained from a simple charging scheme, such as tensile testing on a single axis. The obtained curves in the coordinates , can be used as determination equation at the cutting process modeling. In practice, though, this requires a large computing effort. For a reduction of computing time and ease of use, the curves will be approximated through analytical equations. The mechanical deformation complexity and the visible differences in the shape of the curves for different materials, hinder the decisive choice of the approximation equation.
661
The use of tables became known, as in [7] neural networks in [1]. The empirical and semiempirical methods have the biggest spreading. The analyses made for developing and empirical and semiempirical use for de flow curves can be found in Table 1. The semiempirical ecuations can be presented with the following equations, based on research and considering the thermally activated plastic deformation method:
S = aterm + term
& ) + (T , & )). (T ) S = ( Sa ( ).U ( term
p p p
(4)
in which Sa ( ) is a rigid-plastic model of the plastic deformation considering the hardening process and the speed effect, & p ) considers the ductile resistance of term (T , the material deformation, (T ) considers the temperature effects. The analysis of the table 1 shows that the evolution of the flowing curve is based on very simple models: the rigid-plastic model of v. Mises and Newtons linear-ductil model. A continuous development of the models was produced considering the gardening processes, deformation speed and thermal resi stance decrease.
p
Table 1 Development analysis of the semiempirical equations of flow curves Author, Publication Equation de stare years 0 1 Mises =
S T
Newton Ludwik P., 1909 Perzyna P., 1966 Sellars C.M., Tegart W.J., 1972 Litonski J., 1997 Vinh T. U.a., 1979 Johnson G.R. Cook W.H. Forma iniial, 1983 Altan T., 2000 Ee K.C. 2004 Usui E., 1987, Maekawa K. u.a., 1991
&p S =
S = T + A
& p* ) n S = T (1 +
Q
& p = A(sinh( a. mepM )) m e RT
& p* ) m (1 + C ) S = T ( 0 p + p ) n (1 + B
& p* ) m e mT S = T ( p ) n (
& p* ))(1 (T ) m ) S = ( A + B ( p ) n )(1 + Cln( & p* ))(1 (T ) m + a.e m (T S = ( A + B( p ) n )(1 + Cln(

o o
T0 ) 2
& p* + a.e m ))(1 (T ) m ) S = ( A + B( p ) n )(1 + Cln(

k .T m p* m n p* N N & & ) 1 . Ai .e ki .T + B.e k (T T0 ) 2 S = T . e .( ) d .( i =1 T , (& )
&
p*
Klopp, R. W.; Clifton, R.J., 1985 Zerilli F.J., Armstrong R.W., 1987 Follansbee P.S., Kocks U.F., 1988
& p* ) m T m S = T n (
S = T + A( p ) n + B0 e (
p n 0
0 + 1ln(
)T
m
Oxley P. L. B., 1989
T .k ln & p* m1 B S = T + A( ) + . 1 G 0 k B -constanta Bolzmann & p* ))T S = T (Tm )( p ) m (Tm ) , Tm = (1 Aln(
( )
662 Hensel A., u.a., 1990 Childs T.H.S., 2000
& p* ) m e mT xx = T ( p ) n e n (
1 2
& p* ) m (T ) ciT i S = T ( p ) n (T ) (
i =1
0 Marusich T.D., 1995 El-Magd E., Treppman C., 2000 El-Magd E., 2001 El-Magd E., 2003
& p* ) m ciT i S = T ( A + p ) n (1 + B
i =1
& )e mT S = ( T ( A + ) +
p n p
& p* ))e S = T ( p ) m (T ) (1 + Aln( 0 S = 1 0 T & 1 + e T &

0
T T2
T 1 T T * p n p m & ) + & ) e 1 + Be T2 0 = (( T + A( ) )(
m
2.3.Breaking criteria modeling methods [4]
and
fracture
From experimental investigations it was concluded that the cutting process has a cyclic character, which is based on the formation of chip elements and finally the formation of the flowing chip. In this case the occurrence stages and crack development are repeated continuously. Because of this the breaking model should foresee the continuity interruption moment as well as the crack propagation direction. Given the above, the moment of destruction starting is physically correct determined, by comparing the state of stress and deformation with the limit value, and it is used in the so called breaking criteria. To achieve the numerical algorithms using these criteria requires complex resources. Also, these algorithms are not stable enough and require the use of aids, in fact breaking geometric criteria. In a research project J. M. Huang and J.T. Black, show that using breaking geometric criteria leads to stable numerical solutions, altohught they do not contain informations about the physical behaviour of the mechanical breaking. The most known informations about the breaking criteria are presented in the table 2.
If the nature of the crack is known in advance, you can apply simple breaking criteria, based either on the intensity of the plastic deformation or the amount of tension ( see criteria 2.1 and 2.5 of table 2). At the modeling of the steel chip formation with the surface separation method, Zahhbani S.A. used the criteria with the tension parameters as follows: f = n f + f
(5)
n = max( 2 ,0)
where the crack occurs at f 1 . Wince used the same method at the aluminium alloys processing modeling. The values f and f were determined from the tension-compresion diagram for a given strain, where the amount of plastic deformation has been defined after the Oxley cutting model.
663 Table 2 Breaking criteria [4] Breaking criteria 0 1 1. geometric 1.maximum distance from the nod to the cutting edge 0 2. physic 1 1.Plastic deformation intensity 2. strain energy density 3. cumulative work done at breaking 4. Cumulative sensitivity to disturbance 5. maximum main stress 6. resistance to crack generation 7. cumulative damage Conditions 2
d = d cr
I cr = p
I cr = : d I cr = <
1 >d p Y
I cr = < p + b1 H + b2 >d p
f =I f =
p
K IC (2l )0,5
0
D=
f = ( D1 + D 2 . exp( D3 0 )).(1 + D4 ln )(1 + D5T 0 ) ; 0 = M p & p f = 0 Vc &p p H T + 0,09 exp f = < 0,075 ln >
100 37,8 293
2 H 2 = A Y (1 + v ) + 3(1 2v ) 2 Er 3 Y p f
f T
(T T0 )
As a general criteria for material processing modeling using plastic deformation parameters, can be used the criteria first proposed by Hancock and Mackenzie and later by Johnson and Cook, namely:
D=
p
(6)
0
f = ( D1 + D 2 . exp( D3 0 )) (1 + D4 ln )
(1 + D5T 0 )
0 = M
(6a)
where is the hydrostatic pressure, M is v. Mises tension, D constants are determined experimentally by the Johnson-Cook method. If the crack nature is not known whether before or at last is known that the crack has a combined character, there have to be used combined breaking criteria. In modern practice the combined criteria are used by the following two assumptions: -for the tension and strain state a full value is being chosen, which depends on both the size of plastic deformation and the size of maximum tangential tension; -between two criteria, which independent of each other contain the tension and plastic deformation characteristic, will be
(6)
664
a kind of competition. It is assumed that the brake will take place at that point where at least one of the two criteria is met. Hashemi (1994) used a combination between the equivalent plastic request and the maximum main pressure for the simulation of rapid processing. The ductile breaking concept was introduced by Itawa. He calculated (in 1984) the braking criteria criteriile de rupere by the time step, based on the state of final demand. Marusich and Ortiz proposed in 1995 to use the brittle braking criteria, which is based on KIC toughness about breaking the material or ductile breaking criteria:
max (1, ) = f
(7)
f =
K IC
2l
The criteria is based on pure growth and melting and uses a version of the plastic criteria of Rice and Tracey:
max p (1, ) = fp
(8)
fp = 2,48e 1,5 p
Owen and Vaz (1999) admitted a chip breaking criteria which was based on considerations of damage related to an adaptive remeshing and element wear. Important for the working of FE cutting models,in Langrange`s formulation, is the deployment of the method for achieving that break to interrupt the material continuity. The tension values in the primary and secondary deformation areas, chip shape, cutting force components and especially the force at the free surface and other characteristics of the cutting process and cutting model validity are closely related to the properties of the breaking method.Till now there are known four different types to achive the main break: -division of the nodes on a default border; -division of the nodes on any border; -division of the nodes by deleting elements; -FE network remodeling.
The method of nodes division on a default border consists in dissolving the link between the processing surface nodes and the allready processed surface, when a geometrical or physical breaking criteria is met and the nodes are on the processing surface. Basically this is done either by breaking overlapping nodes or by introducing elasic elements. With such a method a development of a break is beeing modeled, but the direction of propagation of the crack will pe introduced artificially. This leads to errors in determining the tension and makes it impossibile to model the state of tension and strain and implicitly the cutting forces near the cutting edge. The method of nodes division by any border uses a node division in a point and on one direction in which one of the physical breaking criteria is met. In this case it is necesary to compress the network in the crack tip, and a finit element analysis for choosing the direction of the crack propagation. The method of nodes division by deleting the elements in the integration point where the breaking criteria is met, consists in losing the material masses with the deletion of elements. At a small size of elements, the mass of material that is lost is small, so that this does not affect experimantal results. In the cutting finite element model can only small elements on one surface be deleted, in this case the method will run similar to the method of nodes division on a default border. In the last years gained popularity, as a breaking modeling method, the method that does not consits in instantaneous removal of elements to achive predefined criteria, but it reduces the load on the element and the crack changes within certain limits. If the intensity of deformation of a finite element reaces the value 0 , the development of the crack starts. When it reaches the value f , the finite element will dissipate, losing its capacity to resist to a load. In mid-1990 network restructuring algorithms have been developed (adapting R), which compensate the need of a limit zone and an continuous finite element network level support introduction. It was shown that using the criteria that uses a permanent renewal of the finite element
665
network, although requiring more resources, it describes better the cutting process than the method that uses the nodes division on an default border criteria.
2.4. Meshing and remeshing FE networks strategies [5]
Failure indicator rate
2 Y & = I D 2E
2 H (1 + v ) + 3(1 2v ) 3 Y
&p
2.5.Friction at the contact between piece and tool modeling [6]
One of the biggest disadvantages of breaking criteria for tension parameters is the need for adaptive compression network (Hadaption and/or R-adaption). This occurs in the cutting edges to eliminate errors of the tension gradients in the shear area. As a possible alternative there can be used finite elements with high rank (P-adaption), single elements or a mediation of the stress after a finite element nearby. The remeshing is important for an approximation on an lagrange basis, due to the sensitivity of the method of solving a highly deformation grade gearing. To resolve this problem there has been addressed two ways, one active and one passive. A passive approach results in a predistorted grid. It must compensate the deformation resulting from cutting simulation. Such a method has been used by Strenkowski and Carrol in 1985 . The active approach consists of a statement that triggers a remeshing subroutine. Such a statement may be geometric, like those used by Sekhon and Chenot in 1993. Marusich and Ortiz used as a statement the plastic energy application rate or an element of estimation based on consideration of material failure like the one used by Vaz, Jr to simulate the rapid processing.
Table 3 Meshing and remeshing statements Method Observations PreNo remeshing distorsional network Element Geometrical considerations deformation Plastic &p p = Y energy application rate
The friction between tool and piece has an important role in the process of cutting. It is an important source of heat and the most important factor in tool wear.
Table 4 Method 0 Coulomb`s law Exponential law Friction models Expression 1
f = c n
Iwata et. al.
0,07 n H f = v tanh H 0,07 v
f = e 1 exp C
n e

Eldrige et al Wu et al Sekhon Chenot Amontons Coulomb Zorev and and
f = f (T0 ) exp f = eq f = K v f
p 1
A T vf
kf 3
R = N ; R = m R = N ; = max
For friction, generally is used Coulombs model, although the model itself does not fully describe the friction phenomenon. This is why many researchers have studied different experimental models. Shirakashi and Usui used in 1974 an exponential law, which was first developed for non-ferrous materials and was based on the active contact surface concept. Itawa developed a model that also takes into account the Vickers herdness of the workpiece. It was later considered in the models also the influence of the temperature as was introduced by Eldridge in 1991. Wu assumed that the friction is proportional to the equivalent pressure.
666
In 1993, Sekhon and Chenot admitted Nortons friction law, based on the relative speed of sliding between tool and chip. Newer models tend to improve Coulombs law. Amontonss and Coulombs approximation reports friction at the shear load from the flowing pressure. Zorev divided the friction area in two: displacement and adherence. Dudzinski and Molinari paid attention to the influence of friction coefficients of Coulombs law and their dependance on cutting speed and temperature. Tyan used the law of friction factor and analyzed the influences of speed and rate of application of the load on the friction factor m. Today friction cand be simulated in an acceptable way, but it depends on coeficients very hard to calculate.
2.6.Heat transfer between tool and chip modeling [8]
To simulate the heat transfer between tool and chip, Usui and Shirakashi introduced the transit heat transfer concept. It involves a heat flow on the side of the chip and of the tool interface. The two infinite half-bodies approximate the friction, loaded on the chip and the tool, energy estimation Another hypothesis to the chip-tool heat transfer problem, based on the first law of termodynamics, named thermal balance method, was introduced in 1996 by Obikawa and Usui.
2.7.Tool wear model [9]
The simulation of the tool wear phenomenon was stimulated by the need to continuously increase the precision of the FEM metal cutting models and their applicability in industry. The main types of tool wear are considered such as abrasion, adhesion and diffusion.
Table 6 Tool wear model Takeyama and Marata Usui Tool wear model Equation
Behaviour at different temperatures began to be analyzed with the introduction of high speed cutting. There are several approximations for modeling behaviour at different temperatures. The adiabatic warming assumes that heat from mechanical work of unelastic deformations and from friction inside an element is beeing generated and there is no disposal through heat transfer. This was experimentally verified for the processes of manufacture of materials with low diffusion (spray) and at high cutting speed. Combining fully consider heat transfer within the workpiece and tool.
Table 5 Heat transfer models Method Equation Transit heat T T kw transfer = k t Constant factor Two infinite halfbodies Heat balance
dW E = G (V , f ) + D exp dt RT dW B = A nVS exp dt T
n w qw = f c q f
n t
Takeyamas and Maratas model consider the wear through abrasion and difussion. Usuis model for adhesion wear consider the factors of temperature dependency, normal pressure and the speed on the contact surface. Mathew showed that the influence of the abrasive wear according to Takeyama and Marata can be neglected. Molinari and Nouari proposed a model and established a direct temperature control into the diffusion rate.
q w t0,5 k w = 0,5 ; qf=q+qt qt w kt
q
w
w w
c Tw dv = qt ct Tt dv
t
3. CONCLUSIONS
The cutting process modeling by finite element method (FEM) involves four steps:
667
1.Model creation 2.Model resolving 3.Simulation results analysis 4.Possible model changing or optimization The most important step of the four mentioned above, is the model creation. The best way to demonstrate the accuracy of a virtual model is to compare it with experimental data. Given the above, in the second part of the article on cutting processes modeling, is presented an 2D numerical simulation of a orthogonal cutting process modeling using the commercial software SFTC DEFORM-2D V 9.1 [10] for steel SI 1045. The cutting force will pe experimentaly determined for different cutting conditions and the friction coefficient for the developed model will pe chosen in order to minimize errors in the prediction of cutting force value.
4. REFERENCES
[1] Gonzalo, O.; Cerro, I.; Lamikiriz, A.; Etxeberia, I.; Lpez de Lacalle, L.N.; Rivero, A.: Prediction of milling forces from an oblique cutting FEM model, Proceedings of the 8th CIRP International Workshop on Modeling of Machining Operations, Chemnitz, 2005, pp.235-242 [2] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.: Cause Analysis of Errors in FE Prediction Orthogonal Cutting Performances, Proceedings of the 10th CIRP International Workshop on Modeling of Machining Operations, Calabria, 2007, pp.141148 [3] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.:
Thermomechanische Materialmodelle zur Modellierung von Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.24-33 [4] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Bruchmodelle fr die Modellierung von Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.14-21 [5] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Die FEM-Modellierung als moderner Ansatz zur Untersuchung von Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.34-46 [6] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Thermomechanische Wechselwirkungen beim Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.4-13 [7] Kahlori, V.; Lundblad, M; Lindgern L.: Numerical and experimental Analysis of Orthogonal Metal Cutting, J. Manuf. Science and Engieneering: Trans. Of ASME, 1997, pp.1-10 [8]Vaz Jr.,M.:On the numerical simulation of machining processes, Mechanical Science 2000, vol 22, no.2, pp.179-188 [9] Yung-Chang Yen, Jrg Shner, Blaine Lilly, Taylan Altan :R&D Update Machining Estimation of tool wear in orthogonal cutting using the finite element analysis, www.ercnsm.org 2002 [10] DEFORM Users Manual: Version 9.1. DEFORM Inc., 2008
MODELAREA PROCESELOR DE PRELUCRARE Rezumat: n prima parte a articolului pe tema modelrii proceselor de prelucrare prin achiere prin metoda elementului finit (FEM), ca i abordare modern n cercetarea proceselor de achiere, se prezint stadiul actual al cercetrilor privind modelele termo-mecanice de materiale utilizate la modelarea procesului de achiere, criterii de rupere i metode de modelare a ruperii, strategii de
668
discretizare i rediscretizare a reelelor cu elemente finite (FE), modelarea proceselor de frecare la contactul dintre pies i scul, modelarea transferului termic dintre scul i achie i modelul de uzur a sculei.n partea a doua a articolului sunt prezentate rezultatele cercetrilor efectuate n cadrul unui stagiu la Institut fuer Werkzeugmaschinen, Universitatea Stuttgart, Germania.
Koukach Daniela Carmen, Eng., Eng., Romanian Bureau of Legal Metrology, DA (Direcia Autorizri), dana_koukach@yahoo.com, 0040264431366, Cluj-Napoca, Str. Micu, Nr.8, Bl. A10, Ap.16, 00407 48916349 Pop Grigore Marian, PhD.Eng., PhD Assistant, Technical University of Cluj-Napoca, DIF (Departamentul Ingineria Fabricaiei), grigore.pop@tcm.utcluj.ro, 0040264401634, ClujNapoca, Str.Milcov 10, 004075678357 Preja Dan, Dipl.Eng., Manager OEM, S.C. Ghring S.R.L., OEM Romania, dan.preja@guehring.de, 0040755088336, Cluj-Napoca, Str. Dimitrie Gusti, Nr.7/1, 0040723951424
669

MODELING MANUFACTURING PROCESSES OF CUTTING

Daniela Carmen KOUKACH, Marcel POPA, Grigore Marian POP, Dan PREJA
Abstract: In the second part of the article about modeling manufacturing processes of cutting, is prezented a numerical 2D simulation for the modeling of orthogonal cutting processes by using the commercial software SFTC DEFORM-2D V 9.1 for AISI 1045 steel (DIN 1.1191). At the same time it will be shown the experimental determinations for selecting an appropiate friction coefficient, that minimize errors in predicting the cutting force. For the chosen friction coefficient, it will be determined the maximal chip temperature with the help of the created model, whici will be compared with the experimental values presented in the literature. The researches have been conducted in an intership at the Institut for Machining Tool, University Stuttgart, Germany. Key words: cutting force, friction, temperature chip
1. INTRODUCTION The purpose of the research is focussed on achieving a 2D numerical simulations to model the orthogonal cutting process using commercial software SFTC deform-2D V 9.1 [11] for AISI 1045 steel (DIN 1.1191), and experimental deterining of the cutting force in different cutting conditions. In this context, theoretical and experimental research concerning the orthogonal cutting process have been directed towards the following objectives : 1. the development of a model for the numerical 2D simulation of the orthogonal cutting process ; 2. the experimental determination of the cutting force in diferent cutting conditions and the selection of frictin coefficient for the developed model, in order to minimize errors in the prediction of the cutting force value; 3. the determination of the maximal chip temperature for the friction coefficient obtained under paragraph 2, and the comparison with the experimentally determined value reported in the literature. 2. THE DEVELOPMENT OF A MODEL FOR THE NUMERICAL 2D SIMULATION OF THE ORTHOGONAL CUTTING PROCESS Thaking into account the previous theoretical considerations, with the help of the commercial software SFTC DEFORM-2D V 9.1 [11], there has been created a model in whici the tool is considered as a rigid body made of the hardened material TiC and the part as a body whici undergoes elasto-plastic derformations and is made of the material AISI 1045. These materials come from the database of the software SFTC DEFORM-2D V 9.1. The mesh network of the tool contains initialy 875 quadrilateral elements and that of the part contains 1917 quadrilateral elements (as shown in fig. 1 and fig. 2). This networks are in the contact area tool-part more refined than in other areas.
670
The constitutive equatuions for the tool are in accordance with the law of power, this beeing considered rigid. The termical parameters of the tool material have been defined as a function of temperature, and come from the database of the software SFTC DEFORM-2D V 9.1 [11]. The angle of approach of the tool is =5.
Fig.1. The mesh network of the tool
Fig.4. Modeling of the tool material (TiC)
3. EXPERIMENTAL RESEARCHES 3.1 Description of the experimental equipment (fig.5) The simulation does not take into consideration several phenomena such as: cutting edge wear, residual internal stresses in the material and machine oscillations and also the simulation model containes certain simplifications whici had to be introduced in the experimental model as well. The influence of the cutting edge wear was reduced by changing the cutting edge of the tool for every test and by using tools with very high hardness. The used tool is a square cutting plate of the type Walter Tiger Tec WKP 35, cod ISO SEHW1504AFN. The tested parts are rectangular plates 170/50/5 of AISI 45 steel. In order to avoid residual stresses, the tested parts were subjected to the annealing process for stress relief. The effect of oscillatory behavior of the machine were studied using a prototype testing equipment. The testing equipment was designed to study the cutting process and the structure. Its ability of controlled flexibility is useful when it is wanted to reduce the influence of oscillations on the cutiting process. The simplification consistes of the assumption that the tool is a rigid body without deformation. This is achieved by using a
Fig.2. The mesh network of the part
The constitutive equatuions for the part are in accordance with the law of Oxley, which was implemented to describe the flow of material as a function of strain, strain rate and temperature. The termical parameters of the parts material have been defined as a function of temperature, and come from the database of the software SFTC DEFORM-2D V 9.1. [11] (as shown in fig. 3).
Fig.3. Modeling of the parts material (AISI-1045) using the Oxley equation
671
special tool holder and through the hardness difference between the tool and the workpiece. User interface Its role is to create a user-friendly interface with the Siemens control panel. The connection with the control panel is based on PROFIBUS technology. The user interface is build using Visual Basic. The interface is used to control the machine position and the speed of the machine table. The Siemens command The Siemens comand provides the actual control of the linear motor. It descibes the interference of motion and alerts the user about the unacceptable discrepancies for the movement of the machine table. In accordance with the given commands, for example when, during the cutting process, the speed of the table is slown under the influence of acceptable interferences, the table is stopped and a failure display announces the failure of the experiment. The testing equipment The cutting process whici is submited to the simulation is planing, in whici the tool is stationary, and the part performs a translation movement on the x direction. The framework of the testing equipment consistes of solid concrete with polymers. To support the tool there is used the method of a portal construction. The resistance is controled by adjustable massive traction anchors, whici link the portal to the framework. The liniar motor is used for the motion of the table. This allows a continuous change without steps of the speed, from 0 to 200 m/min. Piezoelectric sensors The sensors are displayed as follows:two on the tool holder and two on the measuring plate from underneath the table. There role is to measure two main forces of the planing process: the cutting force and the passive force. These forces are compared with the FEM simulation data. Before the start of the experiment, a calibration of the sensors is performed. The alternating current amplifier The current coming from piezoelectric sensors is too small for the accurate measurement value. The amplifiers role is to provide a controlled increase of the signal, so
that an accurate reading can be accieved. From here, the loaded signal is sent to a National Instruments data registration system. Data registration computer With the help of the data registration system and the Labview 7 program, data from the 4 sensors is saved and analyzed. The data flow of the sensors passes 2 Labview modules. The first module saves the signal from the sensor in .txt format and connect it with the information from the amplifier and the calibration test.
.Fig.5. Experimental equipment The forces have been measured using the SlimLine type 9133B on the Z axis and 9134B on the X axis, both sensors are produced by Kistler Instrumente AG. These sensors are made of quartz with an extremly flat design and are used to measure dynamic and quasi-static forces. The sensors have a welded frame, where the embedded connexion cable is integrated in the body with a non-detachable connection socket.
Table 1 Type SlimLine Sensor techincal data Domain Over- Sensitivity [pC/N] Fz [kN] load Fz [kN] 14,0 17,0 -3,8 26,0 30,0 -3,8 Rigi dity [kN/ m] 2,5 5,6
9133 B 9134 B
672 Table 2 Type SlimLine Sensor techincal data External Internal Height diameter diameter H D [mm] d [mm] [mm] 9133B 16,0 6,1 3,5 9134B 20,0 8,1 3,5 Mass m [g] 6,0 6,0 Table 4 The cutting parameters an the results obtained from the simulation No Cutting Feed Cutting Tan Maximal Fric test speed rate force gential tem tion [m/min] [mm/ Fx [N] force perature coeffici rot] Fz [N] [C] ent 1 50 0,1 895,13 358,68 476,17 0,8 817,76 290,80 460,41 0,7 2 100 0,1 871,96 343,26 586,13 0,8 811,15 285,04 559,38 0,7 3 150 0,1 847,40 314,24 646,09 0,8 787,16 267,02 620,46 0,7
The force can be measured by the sensors through an assembly or through the preloading elements. When the sensor is loaded, this produces an electrical charge proportional to the applied force. This load conducted to the exterior by an electrod or an integrated cable. The cable is integrated in the body and this assembly is shielded. The cutting parameters and the results obtained for the cutting force and the passive force are summarized in table 3.
Table 3 The cutting parameters an the results obtained from the determinations No. Cutting Feed rate Cutting Tan test speed [mm/rot] force Fx gential [N] force [m/min Fz [N] ] 1 50 0,1 849 417 2 100 0,1 886 468 3 150 0,1 745 433
In Fig. 6 the experimentaly determined results were represented graphically and those obtained from the simulation for the cutting force Fx [N] according to the feed rate [mm/min] for a feed of 0,1 mm/rot, using the value 0,8 for the friction coefficient .
Cutting force Fx [N]
Cutting force Fx [N] 1000.00 500.00 0.00 0.00 50.00 100.0 150.0 200.0 0 0 0 Feed rate [mm/min]
experiment simulare 0.8
3.2 The simulation results The simulation started with the entering of the tool in the workpiece at the specified speed. The chip is detached from the machined surface when the surface connections are unfolded and a contact with the placement surface of the tool is achieved. To avoid the numerical instability caused by the imbalance of forces from the surface nods, that are being separated, a distribution was specified that allows the gradual decrease of their value to zero. Due to the strong deformations of the elements around the tip of the tool, a further remesh of the model was needed for a more advanance analysis. For the friction between tool and chip there has been selected a friction coefficient so that the errors in obtaining the cutting force Fx are minimized. It was concluded that the values 0,7 and 0,8 for the friction coefficient are the more satisfying.
Fig.6. Cutting force Fx [N] according to the feed rate [mm/min] for a feed of 0,1 mm/rot
For the obtained model regarding the above, there has been studied how temperature varies across the three areas when using the value 0,8 for the friction coefficient : I the primary shear plane; II the area where fiction occurs between the chip and tool rake face; III the area where friction occurs between the cutting surface and the tool placement surface
673
4. CONCLUSIONS The thermo-mechanical model developed to simulate orthogonal cutting, which bounds the thermal-structural effect with the consideration of temperature-dependent material characteristics, is acceptable. The model calibration regarding the cutting force experimentally determined results, led to obtaining some values for the temperature field according the experimentally determined values presented in the specialty literature [1]. So the application of such an working algorithm where the experimental data is used to calibrate the simulation models, is very usefull. The research undertaken contrbutited to the foundation of an thermo-mechanical model for the orthogonal cutting that can be calibrated regarding the experimental data of determining the cutting force. 5.REFERENCES [1]Adibi-Sedeh, A.H.; Vaziri, M.; Pednekar, V. et al.: Investigation of the Effect of Using Diferent Material Models on Finite Element Simulations of Machining, Proceedings of the 8th CIRP International Workshop on Modeling of Machining Operations, Chemnitz, 2005, pp.215-224 [2] Gonzalo, O.; Cerro, I.; Lamikiriz, A.; Etxeberia, I.; Lpez de Lacalle, L.N.; Rivero, A.: Prediction of milling forces from an oblique cutting FEM model, Proceedings of the 8th CIRP International Workshop on Modeling of Machining Operations, Chemnitz, 2005, pp.235-242 [3] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.: Cause Analysis of Errors in FE Prediction Orthogonal Cutting Performances, Proceedings of the 10th CIRP International Workshop on Modeling of Machining Operations, Calabria, 2007, pp.141-148 [4] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Thermomechanische Material-modelle zur Modellierung von
Fig.7. Area I (PAS 1150) ,Cutting speed 50 m/min, feed 0,1 mm/rot, friction coefficient 0,8
Fig.8. Area II (PAS 1150) , cutting speed 50 m/min, feed 0,1 mm/rot, friction coefficient 0,8
Fig.9. Area III (PAS 1150) , cutting speed 50m/min, feed 0,1 mm/rot, friction coefficient 0,8
The maximum value for the temperature of 476C obtained in area II was compared with the values experimentaly determined and reported in the literature in the interval of 400...450C and found that the difference is acceptable (fig.7-9)
674
Zerspanprozessen, ZWF-Sonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.24-33 [5] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Bruchmodelle fr die Modellierung von Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.14-21 [6] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Die FEM-Modellierung als moderner Ansatz zur Untersuchung von Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.34-46 [7] Heisel, U.; Krivoruchko, D.V.; Zaloha, V.A.; Storchak, M.; Stehle, T.: Thermomechanische Wechselwirkungen
beim Zerspanprozessen, ZWFSonderpublikation, 2009, Carl Hanser Verlag, Mnchen, pp.4-13 [8] Kahlori, V.; Lundblad, M; Lindgern L.: Numerical and experimental Analysis of Orthogonal Metal Cutting, J. Manuf. Science and Engieneering: Trans. Of ASME, 1997, pp.1-10 [9] Vaz Jr.,M.:On the numerical simulation of machining processes, Mechanical Science 2000, vol 22, no.2, pp.179-188 [10] Yung-Chang Yen, Jrg Shner, Blaine Lilly, Taylan Altan :R&D Update Machining Estimation of tool wear in orthogonal cutting using the finite element analysis, www.ercnsm.org 2002 [11] DEFORM Users Manual: Version 9.1. DEFORM Inc., 2008
MODELAREA PROCESELOR DE PRELUCRARE PRIN ACHIERE ABSTRACT: n a doua parte a articolului pe tema modelrii proceselor de prelucrare prin achiere, se prezint realizarea unei simulri numerice 2D pentru modelarea unui proces de achiere ortogonal, utiliznd soft-ul comercial SFTC DEFORM-2D V 9.1, pentru oel AISI 1045 (DIN 1.1191). Totodat, se prezint determinrile experimentale efectuate n vederea alegerii unui coeficient de frecare adecvat, care s minimizeze erorile n prezicerea forei de achiere. Pentru coeficientul de frecare ales, se determin cu ajutorul modelului creat, valoarea maxim a temperaturii achiei, care se compar cu valorile determinate experimental prezentate n literatur. Cercetrile au fost efectuate n cadrul unui stagiu la Institut fuer Werkzeugmaschinen, Universitatea Stuttgart, Germania.
Koukach Daniela Carmen, Eng., Eng., Romanian Bureau of Legal Metrology, DA, dana_koukach@yahoo.com, 0040264431366, Cluj-Napoca, Str. Micu, Nr.8, Bl. A10, Ap.16, 0040748916349 Popa Marcel, Prof. Dr. Eng., Technical University of Cluj-Napoca, Faculty of Machine Buiding, Department of Manufacturing, marcel.popa@tcm.utcluj.ro, 0040264401634, Cluj-Napoca, Str. Baia Mare 21, 0040722365475 Pop Grigore Marian, PhD.Eng., PhD Assistant, Technical University of Cluj-Napoca, Department of Manufacturing, grigore.pop@tcm.utcluj.ro, 0040264401634, Cluj-Napoca, Str.Milcov 10, 004075678357 Preja Dan, Dipl.Eng., Manager OEM, S.C. Ghring S.R.L., OEM Romania, dan.preja@guehring.de, 0040755088336, Cluj-Napoca, Str. Dimitrie Gusti, Nr.7/1, 0040723951424
675

NUMERICAL ANALYSIS OF STATE OF STRESSES AND DEFORMATIONS OF AN INDUSTRIAL SERIAL ROBOT 5R

Mihaela SIMION, Adrian-Ioan BOTEAN, Mircea BEJAN
Abstract. In this paper is presented a numeric analysis to determine the displacements and stresses for the strength structure of an industrial serial robot. This numerical analysis with finite element method consist in study of the robot when is loaded with maximum handling force, to realize a simple task like pick&place. Keywords: analysis, finite element, displacements, stresses, robot.
1. INTRODUCTION The numerical study has the purpose to determine the displacements and stresses for the strength structure of the industrial serial robot Fanuc LR MATE 100iB. To determine the displacements and stresses on entire structure of the robot, has realized a numerical study in static mode, with Finite Element method (FEM) using Ansys 12.1 software (working module Static Structural). To perform this numerical study has chosen a working position of the robot, where arm 1 and rotation module, respectively arm 2 and arm 1 form a 90o angle. In figure 1 is shown the CAD model of the studied robot with considered working position and the constructive elements of this. The strength structure of the robot has established starting from structural study of this from technical documentation, more exactly the study of internal construction of the robot, the motion transmission mechanisms from driving motors to modules, of assembly of different mechanical, electrical and pneumatic parts, etc. Thus, was observed that the internal parts are attached with housing by screws, and the rotation motion of the modules is transmitted from driving motors to modules through gears and reducers.
Fig. 1. CAD model of the robot
The efforts that arise inside of the robot from driving motors are taken, almost entirely, by housing (thought screws), so we can say that the housing is the strength structure of the robot. The material of the strength structure is iron alloy, having the following mechanical features: - Density : = 7200 kg m 3 ; - Youngs modulus : E = 1,1 10 5 N / mm 2 ; - Poisson ratio: = 0 ,28 ; - tensile ultimate strength: ri = 280 N / mm 2 .
676
2. NUMERICAL STUDY STRENGTH STRUCTURE OF INDUSTRIAL ROBOT
OF THE
The numerical study was realized using the values of efforts obtained analytically in the paper [4], the robot being loaded with forces and/or moments corresponding to nodes 2, 3, 6, 7 and 8, which represents the gravity center of driving motors plus the maximum handling force F3 = 50 N; To perform the numerical study it was taken into account the following simplifying hypothesis and constrains: - were not taken in study the rotation joints, gears and bearings and screws; - the covers and their corresponding modules of the robot was considered a single body; - the robot was fixed in the feet from the robots base; - the mesh of the structure was done automatically, choosing the dimension of finite element 30 mm, resulting 170.241 elements and 288.561 nodes. In figure 2 is presented the mesh of strength structure of the robot realized with Ansys 12.1.
module at a time, finally obtaining the global distribution of displacements and stresses. Was approached such a procedure to observe the behavior of the strength structure under the loading action, and also to observe the evolution/variation of the distribution of displacements and stresses, from the base module to orientation module. In figures 3 and 4 are presented the obtained distribution of displacements and stresses using FEM in static mode, for the chosen particular working position, and for a better view of the areas where the equivalent stresses (von Mises) are maxim has done details.
a)
Fig. 2. The mesh of the strength structure of the robot Fanuc using Ansys 12.1
To perform the numerical study with FEM its starts from the base of the robot, adding one
b) Fig. 3. Distribution of displacements obtained with FEM
677
c)
a)
d)
b)
e) Fig. 3. Distribution of displacements obtained with FEM (continuation)
b) Fig. 4. Distribution of von Mises stresses obtained with FEM
678
e) Fig. 4. Distribution of von Mises stresses obtained with FEM (continuation)
c)
In table 1 has given the numerical values of maximum total displacements (by axis x, y and z) and of maximum von Mises stresses, and in figure 5 a) and b) are given the variation graphics of displacements and stresses according to application nodes of efforts.
a) Fig. 5. Variation of total displacements (a) according to application nodes of efforts [4] d)
679 Table 1 The maximum numerical values of total displacements and von Mises stresses
b) Fig. 5. Variation of von Mises stresses (b) according to application nodes of efforts [4] (continuation)
not have a great influence on the structure of resistance of the robot; equivalent von Mises stress values are influenced by the CAD model of the structure and construction of various elements such as beams connection etc. the obtained reduced numerical values distinguish that the studied particular working position not raise problems from strength structure and accuracy point of view, even when the robot is loaded with maximum load handling. Future numerical studies will be made deeper (static and dynamic) and experiments to obtain information as true on the state of stresses and deformations for the Fanuc LR Mate 100iB robot. ACKNOWLEDGMENT: This paper was supported by the project "Doctoral studies in engineering sciences for developing the knowledge based society - SIDOC contract no. POSDRU/88/1.5/S/60078, project co-funded from European Social Fund through Sectorial Operational Program Human Resources 20072013. 4. REFERENCES [1] Bejan, M., Rezistena materialelor, vol. 1 (ediia a V-a) i vol. 2 (ediia a IV-a), Editura AGIR, Bucureti i Editura MEGA, 2009, Cluj-Napoca, Romnia [2] Creu, A., Rezistena materialelor, vol. 1, Editura MEDIAMIRA, 2005, Cluj-Napoca, Romnia [3] Simion, Mihaela, Tensiuni i deformaii n structura de rezisten a unui robot industrial serial, Raport de Cercetare nr. 2 n
3. CONCLUSION AND OBSERVATIONS In this paper has presented: a numerical study in static mode to determine the displacements and equivalent stresses in the case of strength structure of the industrial serial robot Fanuc LR Mate 100iB, for a particular chosen working position; the maximum value of displacement has obtained for the rotation module (figure 3 b)); the interested value of displacements is the one obtained over the entire strength structure (figure 3 e)), which is quit low so that will not influence the accuracy of the robot; the maximum values of stresses has obtained to the base module, but the von Mises equivalent stress corresponding to the entire structure is quite low, so it will
680
cadrul doctoratului, iunie 2011, ClujNapoca, Romnia [4] Simion, Mihaela, Botean, A.I., Bejan, M., Analytic study of efforts from the structure of a industrial serial robot 5R, Acta Tehnica Napocensis, Series: Applied Mathematics and Mechanics, vol. 54, Issue 1, ISSN 12215872, pag. 225-228, 2011, Cluj-Napoca, Romnia
[5] Simion, Mihaela, Botean, A.I., Bejan, M., Studiul analitic i numeric al deplasrilor pentru o grind articulat, tiin i Inginerie, vol. 19, Editura AGIR, Bucureti, 2011, ISSN 2067 7138, pag. 657-666, 2011, Cluj-Napoca, Romnia [6] tutoriale ANSYS 12.1 [7] *** www.fanucrobotics.com
Analiza numeric a strii de tensiuni i deformaii ale unui robot industrial serial 5R Rezumat. Lucrarea de fa prezint o analiz numeric pentru determinarea deplasrilor i tensiunilor n cazul structurii de rezisten a unui robot industrial serial 5R. Acest analiz numeric cu metoda elementului finit cuprinde studiul robotului atunci cnd acesta este ncrcat cu fora maxim de manipulare, pentru a realiza o micare simpl de pick&place.
Mihaela Simion Eng., PhD. Student, Technical University of Cluj-Napoca, Department of Strength Materials, email: mihap_elly@yahoo.com. Ioan-Adrian Botean PhD.Eng., Professor Assistant, Technical University of Cluj-Napoca, Department of Strength Materials, email: Adrian.Ioan.Botean@rezi.utcluj.ro Mircea Bejan PhD.Eng., Proffesor, Technical University of Cluj-Napoca, Department of Strength Materials, email: Mircea.Bejan@rezi.utcluj.ro

Acta No54volIV Cuprins Si Rezumate

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

Acta No54volIV Cuprins Si Rezumate

Uploaded by

Copyright:

Available Formats

ACTA TECHNICA NAPOCENSIS

Vol. 54, Issue IV Series: APPLIED MATHEMATICS AND MECHANICS 2011

ISSN 1221 5872

ACTA TECHNICA NAPOCENSIS

ACTA TECHNICA NAPOCENSIS

TECHNICAL UNIVERSITY OF CLUJ-NAPOCA

ACTA TECHNICA NAPOCENSIS

MAINTENANCE STRATEGIES IN THE CORRELATION MAN-MACHINE-ENVIRONMENT

Fig. 1. Diagnosis of rotating machinery dynamic

transposition of certain facts relating to the

Time (logarithmic scale)

Fig. 2. The life cycle of a machine

Period of stable functioning

Time (logarithmic scale)

Fig. 3. The life cycle of a reconditioned machine

TECHNICAL UNIVERSITY OF CLUJ-NAPOCA

ACTA TECHNICA NAPOCENSIS

SYSTEMS FOR ON-LINE MONITORING OF VIBRATIONS

Analog input Vibration monitoring

ICP module Measuring device

Fig. 1. Monitoring of vibration system with compatible ICP accelerometers

Displacement transducer Arbor Displacement

Fig. 2. Monitoring of vibrations system with proximity systems

Fig. 3. The spindle orbit in bearing

Proximitor Coaxial cable

Fig. 4. Monitoring system of vibrations with proximity systems

Fig. 5. Monitoring system of vibrations with speed sensors

Fig. 6. System for monitoring temperatures with thermocouples and thermorezistances

ICP module Measuring instrument

Fig. 7. Noise monitoring system with microphones compatible ICP

Detector 2 SPC - PSD

Laser beam Laser module 2

Fig. 8. System monitoring with laser sensors one-dimensional

TECHNICAL UNIVERSITY OF CLUJ-NAPOCA

ACTA TECHNICA NAPOCENSIS

ACOUSTIC NORMAL MODES SIMULATION OF A CLASSROOM

For a rectangular enclosure ( l x , l y , l z ) the

wave equation becomes:

and for: 2 / c 2 = k 2 , 0 one get:

and the sum of the three constants equals k2: (12)

The solutions of the equations (11 a,b,c) are:

p z ( z ) = C1z e jk z z + C2 z e jk z z The solution (7) of the equation, becomes:

Dividing by p(x,y,z), results:

Replacing these constant relations in the solutions (13), results:

(18a) (18b) (18c)

Fig. 1. Modal frequencies distribution

Fig. 5. Mode #4 [1 1 0] Fig. 2. Mode 1: 11,3 Hz [1 0 0]

Fig. 6. Mode 10: 45,8Hz [0 0 1]

Fig. 3. Mode 2: 22,3Hz [0 1 0]

Fig. 4. Mode 3: 22.4Hz [2 0 0]

Fig. 7. Mode #5, [2 1 0]: 31.5Hz

TECHNICAL UNIVERSITY OF CLUJ-NAPOCA

ACTA TECHNICA NAPOCENSIS

MODAL ANALYSIS OF AN INFLATED TIRE BY SIMULATION

Fig. 1. Rotational modes (in plane)

Fig. 2. Translational modes (in plane)

Fig. 3. The wheel mesh layout

3.3. Mode shapes The first mode shape, depicted in figure 4, is

or: Fig. 4. Mode shape #1

Fig. 8. Mode shape #7,#8

Fig. 5. Mode shape #2

Fig. 9. Mode #13

Fig. 6. Modes #3 and #4

Fig. 10. Radial lobes