Professional Documents
Culture Documents
Structures Laboratory
4
jQ
Final Report
.fr CT I 71984D
84 10 16 156
Destroy this report when no longer needed. Do not
return it to the originator.
I
I
The contents of this report are not to be used for
advertising, publication, or promotional purposes.
Citation of trade names does not constitute an
official endorsement or approval of the use of such
commercial products.
TECHNICAL REPORT SL-84-13
July 1984
Final Report
Approved For Public Release; Distribution Unlimited
This report has oeen reviewed in the Federal Emergency Management Agency
and approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Federal Emergency Management
Agency.
Prepared for
Unclassified
1S5. DECLASSIFICATION/DOWNGRADING
SCHEDULE
i1. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, It different from Report)
Available from National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22161.
19. KEY WORDS (Continue on reverse side It necessary and Identify by block number)
Unclassified
SECURITY CLASSIFICATION OF THIS PAGOE(3hen Dole Entered)
PREFACE
NTIS R&I
DTIC TABA
UuannouncOe
justification -
WTI-
_Distritu~ton/ _
AvailabilitY Codes
Avail and/or
ýDist Special
CONTENTS
Page
PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
LIST OF FIGURES .................. ............................. 3
LIST OF TABLES ................... .............................. 3
CONVERSION FACTORS, NON-SI TO SI (METRIC) UNITS OF MEASUREMENT . . . 4
CHAPTER 1 INTRODUCTION .............. ......................... 5
1.1 Background ............................. 5
1.2 Objectives and Scope ................ .... ..................... 6
1.3 Procedure ....................... ... ........................... 6
CHAPTER 2 STRUCTURAL DETAILS AND MATERIAL PRCPERTIES ......... 7
2.1 Entryway ...................... .... ........................... 7
2.1.1 Shape ....................... ... ........................... 7
2.1.2 Airblast Prediction ................... .................... 7
2.1.3 Structural Details ................ ... .................... 8
2.2 Closures ...................... .... ........................... 8
2.2.1 Configurations Examined ................. .................. 8
2.2.2 Analysis of Various Configurations ........ ... ............ 9
2.2.3 Type 3 Door Details ................... .................... 10
2.2.4 Projected Cost .......................................... 12
2.2.5 Commercial Door and Supports ...... ............... .. 12
2.3 Models .................... ............................ 14
CtLAPTER 3 EXPERIMENTAL PROCEDURE .......... .................. .. 29
3.1 DIRECT COURSE ............................................... 29
3.2 Test Plan ................. ........................... .. 29
CHAPTER 4 RESULTS ................. .......................... .. 36
4.1 Data Discussion ............. ........................ .. 36
4.2 Structural Response ........... ...................... .. 36
4.3 Static Test ................................................. 37
CHAPTER 5 ANALYSIS.... ...... o... .......... ......... ... 45
5.1 Verification of Predictions................ .................... 45
5.1.1 Airblast ................ ......................... .. 45
5.1.2 Blast Door Response ......... .................... .. 45
5.2 Predicting Response to Other Weapons/Pressures ........ .. 46
5.3 Pass-Through Entryway ........... ..................... ... 47
CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS ..... .............. .. 53
6.1 Conclusions ................. .......................... .. 53
6.2 Recomuendations ............. ........................ .. 54
REFERENCES ..................... .............................. .. 56
APPENDIX A PRES,3URE DATA ............ ...................... .. 59
2
LIST OF ILLUSTRATIONS
Figure Pg
2.1 Elevation view of entryway ....... ................ .. 16
2.2 Plan view of entryway .......... .................. .. 16
2.3 Charge tower as seen from entryway interior ....... 17
2.4 Pretest predicted loading history for blast door ..... 18
2.5 Typical slab reinforcement details .... ............ .. 18
2.6 Concrete placement in stairs ..... ............... .. 19
2.7 Tying reinforcement in end wall ..... .............. ... 19
2.8 Reinforcement of stairwell walls; underground
chamber complete ........... .................... .. 20
2.9 Finishing concrete in stairwell roof ... ........... ... 20
2.10 Backfill placement around completed structure ...... 21
2.11 Blast door configuration concepts .... ............ .. 21
2.12 Blast door strap hinge details .... ............... .. 22
2.13 Blast door shell prior to concrete placement ....... 23
2.14 Pretest blast door response prediction ............ .. 24
2.15 Cross section of commercial door and supports ...... 25
2.16 Commercial door with support beams in place ....... 26
2.17 Commercial door with support plates in place ....... 27
2.18 1/10-scale models ............ .................... .. 28
3.1 DIRECT COURSE charge tower assembly ... ........... .. 30
3.2 Pretest photograph of site layout .... ............ .. 31
3.3 Surface view of full-scale entryway ... ........... .. 32
3.4 Reinforced concrete blast door in place .. ......... .. 33
3.5 Instrumentation of full-scale structure .. ......... .. 34
3.6 Instrumentation of dead-end tunnel 1/10-scale model . . . 35
3.7 Instrumentation of pass-through tunnel 1/10-scale
model ................ ......................... .. 35
4.1 Posttest surface view of entryway .... ............ .. 39
4.2 Damage to stairwell roof immediately above landing .... 40
4.3 Damage to stairwell roof ....... ................. .. 40
4.4 Closures; reinforced concrete blast door on left, open,
commercial door on right ....... ................ .. 41
4.5 Closures; reinforced concrete blast door on left,
commercial door on right ....... ................ .. 42
4.6 Damage to commercial door support plate .. ......... .. 43
4.7 Failure of test slab; static test .... ............ .. 44
5.1 Comparison of measured to calculated pressure data
for blast door loading ....... ................. .. 49
5.2 Calculated response of blast door to DIRECT
COURSE loading ........... ..................... .. 50
5.3 Calculated loading history of blast door for a I-MT,
50-psi event ..................... ............ 51
5.4 Calculated blast door response for 1-MT, 50-psi event 51
5.5 Pass-through entryway system ..... ............... .. 52
5.6 Comparison of incident to peak pressure on closure
in dead-end and pass-through entryway systems ..... 52
LIST OF TABLES
Table
2.1 Blast door analysis results ...... ............... .. 15
3
CONVERSION FACTORS, NON-SI TO SI (METRIC) UNITS OF MEASUREMENT
Multiply By To Obtain
feet 0.3048 metres
inches 0.0254 metres
megatons (nuclear equivalent 0.004184 terajoules
of TNT)
kilotons (nuclear equivalent 4.184 terajoules
of TNT)
pounds (force) per square 0.006895 megapascals
inch (psi)
pounds (mass) 0.45359237 kilograms
pounds per cubic foot 0.1601846 grams per cubic centimetre
psi per inch 0.271447 megapascals per metre
tons (nuclear equivalent 4184.0 megajoules
of TNT)
4
BLAST DOOR AND ENTRYWAY DESIGN AND EVALUATION
CHAPTER I
INTRODUCTION
1.1 BACKGROUND
5
1.2 OBJECTIVES AND SCOPE
The objectives of this project were to design a walk-in, reinforced con-
crete entryway and blast door and evaluate the design in a I-KT simulated air-
blast environment in the DIRECT COURSE event at White Sands Missile Range
(WSMR), N. Mex. The DIRECT COURSE event is described in Section 3.2. Various
blast door configurations were considered and from those analyzed, a prototype
door was selected for use in the entryway. A commercially available fire-
resistant door with special supports was also tested. By using 1/10-scale
models, airblast loading data for a I-MT weapon were obtained for both a
single-tunnel dead-end entryway system and a double-tunnel pass-through entry-
way system. These data will be used for the analysis of entryways and blast
doors for the key worker blast shelter.
1.3 PROCEDURE
-16:1
CHAPTER 2
2.1 ENTRYWAY
2.1.1 Shape
7
In examining the configuration of the entryway, it is apparent that with
the given orientation of the stairway, the wall and roof slabs will be loaded
primarily from the inside. However, it is unlikely that the backfill sur-
rounding the structure would allow enough deflection in the slabs to cause
failure. Hence the worst case loading for the wall and roof slabs would be
with the open end of the entryway facing away from ground zero, so that the
slabs are initially loaded by the soil-transmitted pressures only. Since the
survival of personnel inside the protective shelter is more dependent on the
vulnerability of the closure itself, the entryway was tested for a worst case
loading on the doors, i.e., with the entryway opening facing ground zero.
Obviously, in the case of a real-world blast shelter, if the probable burst
point of a nuclear weapon can be determined with any degree of accuracy, the
shelter should be oriented in such a way that the entryway opening is facing
away from ground zero.
2.2 CLOSURES
Prior to the design of a blast door, tentative design criteria were es-
tablished. A minimum static ultimate resistance of 150 psi was selected for
the door based on the predicted peak pressure at the door opening. In order
to minimize the cost of the hinges and to facilitate handling, the weight of
the door was limited to a maximum of 1,500 pounds. The level of protection
from prompt radiation provided by the door was also taken into consideration.
8
Three concepts for a reinforced concrete door configuration were con-
sidered (Figure 2.11). In each case the clearance provided by the door
opening was taken as 3 feet wide by 7 feet high, with the door overhanging
this opening by 4 inches on all edges.
The Type 1 door (Figure 2.11) was a conventional reinforced concrete slab
with steel channels running the length of the door to provide a bearing sur-
face for the hinges and latch mechanism. The Type 2 door was of a sandwich-
type construction, with steel plates of equal thickness on each side of the
door, anchored to the concrete with welded shear studs. The Type 3 door in-
cluded a steel plate on the tension face (inside) of the door with deformed
bars across the short span near the compression face. Hooks welded to the
steel plate fix the deformed bars in place and anchor the steel plate to the
concrete. High-density concrete was examined in each configuration for its
radiation attenuation potential.
Each door configuration was examined for its flexural capacity, weight,
and radiation attenuation. Door thickness and steel ratios were varied to ob-
tain a "best case" for each door type. Equations for flexural capacity from
Reference 14 were used to calculate the maximum resistance of each door.
Results of this evaluation are given in Table 2.1. The value labeled "%
rad" indicates the percentage of the radiation level outside the door which
will be transmitted through the door. Exact figures for radiation levels
penetrating the door are not given, because this phenomenon depends not only
on weapon size and range, but also on entryway orientation and weapon type.
Reference 10 was used for radiation attenuation calculations.
As indicated in Table 2.1, the Type 1 door met neither the flexural
capacity requirements nor the weight restrictions. While the flexural
capacity could be raised by constructing a thicker door, this would also raise
the weight, which is already higher than the imposed limit. The Type 2 doors,
by weight, are much stronger than the Type 1 doors because of the much higher
steel ratios. However, the Type 2 doors would present a problem for concrete
placement since each outside face is covered by a steel plate. Also, since
the rebound force on the door is in the range of 25 percent of the primary
load, less steel is needed in the compression face than in the tensile face.
The Type 3 door, while not as strong as the Type 2 doors, does meet the
9
tentative flexural capacity requirement, is lighter than Type 2, and does not
present the construction problem posed by a Type 2 door. Pased on this evalu-
ation, a Type 3 door, 3 inches thick with 11-gage sheet steel in tension and
No. 4 bars at 6 inches on center in compression was selected for further
examination.
It will be noted from Table 2.1 that in each case the high-density con-
crete gave lianited additional radiation protection while raising the door's
weight considerably. For this reason, high-density concrete was not con-
sidered for use in the blast door beyond this evaluation.
The welded hooks in the Type 3 door serve as shear reinforcement and pre-
vent the steel plate from separating from the concrete during loading. Shear
calculations were based on the dynamic reactions at the edges of the door
opening, obtained from Reference 15, Table 5.4. The shear reinforcement se-
lected consisted of No. 3 bars at 6 inches on center.
The door is designed to overhang the opening by 4 inches on all edges;
therefore, hinges must support the weight of the door and only some of the re-
bound force. The hinges are subjected to very little stress caused by the
primary loading of the door. The predicted rebound force was determined by
the use of Figure 9-1.4 in Reference 16 with the following input parameters:
"T= natural period of door = 5.5 ms
td = predicted duration of load = 120 ms
Assuming a ductility of 10, the ratio of rebound force to primary load
is approximately 0.15, which yields 21 psi on the door in rebound. It was
assumed that the hinges would carry about one-half of the rebound load, the
other half being carried by the latch mechanism. The hinges used were typical
heavy-duty strap hinges with 1/2-inch-diameter pins and 18-inch-long, 1/4-inch-
thick straps. The four hinges used satisfied the rebound requirements of the
L .door and gave a high factor of safety for the weight of the door alone. The
hinge straps were bent by the manufacturer to accommodate the 3-1/8-inch off-
set of the door. The hinges were fastened to the door with 3-3/8-inch-diameter
V- bolts and anchored to the door frame and supports with 3/8-inch-diameter bolts
(Figure 2.12).
The door frame was designed to prevent excessive concrete cracking in the
door supports, which are required to carry the entire load placed on the door
10
in addition to their own load. It was also considered desirable for the door
frame to provide a smooth surface which the door could close against, so that
a good seal between door and frame could be more easily attained. The door
frame was fabricated from 4- x 4-inch steel angles with a 7- x 4-inch angle on
the hinged side of the door. Shear studs were welded to the inside of the
frame every 6 inches to provide anchorage in the door supports. The door
frame was anchored in place before the door supports were formed, so that the
frame and supports were an integral unit. The frame may be seen in construc-
tion photos (Figures 2.6 and 2.7).
As stated previously, the door supports were required to withstand the
total load carried by the door plus the load placed directly on them. With
4 inches of door overhang, 4 inches taken by the hinges, and a minimum of
5 inches of clearance to the end wall from the open door, the door supports
were required to project at least 13 inches from the end wall. The dynamic
reactions around the edges of the door were obtained from Table 5.4, Refer-
ence 15. Based on design calculations, a depth of 15 inches was selected for
the support on either side of the door, and the support above the door sloped
from 15 inches deep immediately behind the door to 30 inches deep at the ceil-
ing. Principal reinforcement ratio throughout the supports was 0.36 percent.
Strips of the 11-gage sheet steel were cut 3 inches wide and welded to
the sides of the door's rear plate before concrete placement to aid in forming
the door. The hooks, made from No. 3 reinforcement bars, were cut 5-3/4 inches
long, bent through a 135-degree arc, then welded to the steel sheet. The
11-gage sheet was fixed on all edges during welding to prevent excessive warp-
ing. The No. 4 bars were fastened to the welded hooks with wire, but were not
welded to the side sheets. Prior to concrete placement, the hinges were fas-
tened to the door shell as shown in Figure 2.12.
The blast door was constructed at WES and transported to WSMR prior to
construction of the entryway. All steel used in the door was ASTM
Grade 50. The concrete placed in the door had an average 28-day strength of
4,210 psi. The complete door shell, prior to concrete placement, is shown in
Figure 2.13.
The resistance function of the door was considered bilinear with a
maximum resistance of 180 psi (Table 2.1) and a stiffness given by: 1 5
K
201 El
ba = 657 psi/in
K
ba 3
11
where
b = long unsupported dimension of door = 84 inches
a = short unsupported dimension of door = 36 inches
The elastic deflection is 180 psi/(657 psi/in) = 0.27 inch.
12
beams would not be set in place until the blast shelter was occupied, at which
time they would be mounted in the boxes. The beams, W6 x 12's, weighed approx-
imately 44 pounds each.
Although the beams were capable of supporting the predicted load on the
door, there was doubt as to whether the door was capable of supporting the
load between each span (24 inches). Spreader plates were designed to transfer
the load from the door to the beams. Two different concepts for these plates
were examined. First, six separate plates would be hung, two from each beam,
to completely cover the door surface. This would in effect make the plates a
double cantilever, fixed in the center. Second, two continuous plates, each
covering one-half of the door surface, would be placed between the door and
the beams. However, if the assumption of negligible flexural capacity for the
door is correct, a 1-inch-thick plate would be needed for the smaller canti-
lever plates, or a 7/8-inch-thick plate for the continuous concept. Each
1-inch steel plate would weigh about 110 pounds, making it unmanageable in
the confines between door supports. Also, the cost of the plates would elimi-
nate any economic benefit gained by using a commercially available door. The
size of steel plate needed to insure elastic response of the door was imprac-
tical and costly. Therefore, a smaller, more manageable plate was selected
and a high level of damage was predicted. Six cantilevered plates, 3/16 inch
thick, weighing about 20 pounds each, were used to reinforce the door. The
plates were supported on the wide-flange beams with steel angles (Ll × 1 x
1/4 inch).
Rebound forces acting on the door were anticipated to be about 25 percent
of the primary load, assuming the door survived. To prevent the door from
opening during rebound, it was necessary to anchor it to the support beams.
L-shaped anchor bolts served this purpose with the small leg of the "L" hooked
i' • to the back sides of the support beams. Three 3/8-inch bolts were used at
each beam. Photographs of the commercial door and its supports are shown int
Figures 2.16 and 2.17.
Approximate cost of the commercial door and its supports is given below:
Approximate
Materials Cost
Steel fire door, fire rating B $200
Door frame 67
(Continued)
13
Approximate
Materials Cost
Hinges 17
W6 v 12 beam, 11 feet 108
3/16-inch steel plate & angles 120
1/4-inch plate 45
Total Materials $557
only;
Labor (Fabrication
does not include installation costs) $200
Total Cost $757
2.3 MODELS
14
cii
00 co
E-4-4
-*4 -0
C4 L,
Ndii
A* 00 00 -
-44
cc N - 4.i
0 .0 C
Cu '-44 -4 (
41. 4J
to V UV
U) 0-
C.) P4 cn0 \s0o
" Vn "4 WLn
rii
v-a Cu 0 C 0
,0-41
~ic
4J
En
(n..'i, 00
\0 N4
.0 00
ON
r-4
V)
r-4
$4 r- N' -
F- I -44
5- '4 N 0 0
00 C n00
-r4
- r
-4 -u
.,4t .- 0) 4
N -~ N 5
,BLAST
i6
16
Fiur 2..Cag-oera enfo
enrwyineir
I1
PEAK PRESSURE =142.38 (PSI)
-100
us
c-
C0.50
0-
0 20 40 60 80 100 120 140 160
TIME, MSEC
18
41
0I~
'4.4GJ'
1.4
.9-4
0-@
41 o
0 4)
G00
19
.
Figure 2.8. Reinforcement of stairwell walls;
underground chamber complete.
20
Figure 2.10. Backfill placement around completed structure.
STEEL CHANNEL>
TYPE 1 ,
p = p = 0.5%
TYPE ,STEEL
3 PLA TE
21
'AMM.
*
q * -.1
* .9 . L&J
1
* *.. 4J
I'. * (4
.b.'. h
I-.
uJ
'4 **'*
* * C')
* .A .
.4.'.. c
14: * **
U)
-4
A
0*
4.. 4A* . * * 00
A '':...,.-
A ' . * ** S.i
V * 4
St k-.5 '.5
U)
'-4
0
A 0
6
b * *?* .. :..' .45U)
A" ***''
..* Z * .
b.' * . ,.. * * CJ
* £
4 * (4 e.J
I-.
. -J 4)
A. 0 * 1-4
.1
IL.
b
.4
.1
44 * 9.*
.9..
1
Jr
22
Li
0
-
*
V
23 V.
ImI.-
0.4
Y (MAX) 0.36"
95• 0.24
0 00.3
0. YPRM=0.0
TIME, SEC
24
•; • 3/16"STEEL PLA TE
COMMERCIAL
DOOR
W 6x 12
"38"
ANCOR LT
25-
25
II
OO
W N.v.. .
26
•. .. .. , • ,,
' ,
- " ". - : . .- -• , •'aL".,
- " ' ,.A , '. sA '• : -" •
"X• is• ; ' -
IfI
if 27
,j- ,
'9.
~Uri
I
"dl~
71N
* 28
CHAPTER 3
EXPERIMENTAL PROCEDURE
The DIRECT COURSE event was a high-explosive (HE) test sponsored by the
DNA in October 1983. A simulated nuclear weapon airblast and ground motion
environment were provided by the detonation of 609 tons of an ammoniun nitrate-
fuel oil (ANFO) mixture at a height-of-burst of 166 feet. The airblast ef-
fects from the detonation of 609 tons of ANFO are approximately equivalent to
the airblast effects from the detonation of a I-KT nuclear device or the deto-
nation of 500 tons of TNT.
The explosive charge was placed in a spherical fiberglass container,
35 feet in diameter, and suspended from a rigid steel column. A photograph of
the charge tower assembly is shown in Figure 3.1.
The entryway test structure was constructed and instrumented during June-
August 1983. After construction and instrumentation were completed, backfill
was placed in 1-foot lifts around the structure up to the existing ground
level. Backfill material consisted of a sandy clay. A photograph of the
entryway and models in place prior to thL test is shown in Figure 3.2. The
closures are shown ia Figures 3.3 and 3.4.
Figures 3.5-3.7 are instrumentation plans showing all airblast gages used
in the test. The airblast gage behind the commercial door, No. P-8, was ren-
dered inoperable prior to the test when its cable was severed during construc-
tion. Hence no airblast data are available for the area behind the commercial
door. The airblast gages were Kulite V5 92 pressure gages with a maximum
range of 200 psi.
29
t ý
3:
-4-
30
.-4
C
4W
-u
Cd)
'4 0-
C
c-c
-C-
Cd)
-C
CC C-
4 0
vi
'1 -'0
4 -C
.4 4 C
4'
I Cc
j t;
31
C
.4
-t
A
(closures
of full-scale entryway
Figure 3.3. SurfaCe view4
tunnel).
at bottom Of
on left arnd right
-32
m
MW-i:
0m
0-M
Cad
li
1All
33
6'
P2
23'-6"
-r
34
IPll
GZ
P13
35
CHAPTER 4
RESULTS
All airblast data were recorded on magnetic tape and later reduced to
digital format. A "sample and hold" technique was used to digitize the data.
For digitizing, data records were started at the time of detonation. Each of
the pressure record3 was divided into 200,000 uniform time steps per second
and sampled at the end of each step.
The location of the test structure was at a higher-than-predicted pres-
sure level. Two surface-flush gages at ground level, P1 and P2, recorded an
average maximum overpressure of 69 psi. Hence the pressures recorded inside
the entryway were also higher than those predicted. The maximum pressures
recorded on the entryway end wall and on the blast door were 180 psi and
159 psi, respectively.
Complete pressure records for each of the airblast gages are shown in
Appendix A. Gage P8 was inoperable prior to the test. Zero time on each of
the records is the time of detonation.
Although the roof and walls of the entryway suffered cracking and some
permanent deflections, there was no structural failure. The stairwell walls
had large cracks along the roof and floor joints, and smaller vertical cracks
along their height. The end wall, which received the most severe loading,
suffered the most damage. Large cracks formed from the floor to the roof
along each wall/door support joint. The end wall was displaced a maximum of
about 2 inches on the east edge and 1-1/2 inches on the west edge. The roof
exhibited minor cracking along its entire length but was largely undamaged. A
posttest surface view of the entryway is shown in Figure 4.1. Wall and roof
. ,damage are shown in Figures 4.2 and 4.3.
Upon examination, the blast door showed no signs of damage. The exterior
concrete face of the door was intact and had no cracks. The seal surrounding
the door's edges was slightly deformed but was not torn. The pressure record
from gage P7, immediately behind the door, indicates that there were no pres-
sure leaks at the door's edges. The door's hinges were undamaged and the door
36_
could be easily opened and closed again, as shown in Figure 4.4. The center
of the door had a permanent deflection of about 0.18 inch (the door was
checked for distortion prior to the test).
The commercial door was completely destroyed during the test. The door
and each of the steel plates supporting it were bent around the wide-flange
beams, so that large gaps were left between door and frame at either end of
the door. As shown in Figure 4.5, the door was obviously inoperable. One of
the steel support plates is shown in Figure 4.6. The pressure records for
gages P5 and P6 should have been nearly identical, had the commercial door
survived. However, comparison of the pressure records for these gages seems
to indicate that the commercial door failed very early (about 10 ms after
initial loading).
The 1/10-scale models were intact after the test and showed no signs of
displacoment or rotation. The amount of dust and dirt that accumulated in the
;•models
J during the test was insignificant.
loading.
A longitudinal crack formed from the left edge of the test section to the
left load point on the top surface of the test specimen very early in the
test. When loaded to approximately 24,000 pounds, a diagonal crack formed on
the left end of the test section, and •he slab abruptly failed (Figure 4.7).
As seen in Figure 4.7, the slab failed in diagonal tension, an undesir-
able mode of failure. The static test indicated that additional shear rein-
forcement near the door supports would be needed to insure ductile behavior at
high overpressures.
The equivalent uniform load necessary to cause the same moment at the
center of the test section due to a two-point load of 24,000 pounds is approx-
imately 49.4 psi. The equivalent uniform load necessary to cause that same
moment at the center of the full-scale door, simply supported on all four
sides, is only 57 psi. The calculations are obviously not in agreement with
37
the measured results of the DIRECT COURSE event, where the door survived a
pressure of 159 psi.
The initial concrete cracking in the test slab was probably due to a lack
of confinement at the edges of the slab, which is not a problem in the full-
scale door. How much influence, if any, this initial crack had on the final
results of the static test is hard to determine. The conclusion drawn from
this test is that the shear reinforcement should be modified in the full-scale
door to insure a ductile mode of failure.
38
38
Figure 4.1. Posttest surface view of entryway.
39
Figure 4.2. Damage to stairwell roof immediately above landing.
.4 ..... ....
-. ! ~
40
0
-40
41.
.1Z4
4
-0
1 4 .J
0
u- 0
~c3
WWI~
42
Fiur 4..Dmg ocmecaldo upr.pae
44
771.
lii
44
AA
"CHAPTER 5
ANALYSIS
5.1.1 Airblast
Since the pressures recorded at the tunnel entrance were higher than
those predicted, the measured internal entryway pressures were also higher
than the initial predictions. Revised internal pressures were calculated
based on the measured pressure data at the tunnel entrance with the following
input parameters:
Peak pressure = 69 psi
Positive phase duration = 107 ms
Impulse = 1.48 psi-s
Comparisons of posttest pressure calculations with actual recorded data
are tabulated below:
Airblast Actual Peak Calculated Peak
Gage Pressure, psi Pressure, psi
P4 180 166
P5 159 158
PlO 185 235
PI1 153 235
P13 65 75
The results of these calculations show close agreement between measured and
calculated pressures for the full-scale structure and the pass-through 1/10-
scale model, while the calculated pressures for the closed-end model were
overpredicted. Calculated positive durations were also reasonably close to
the measured values (Figure 5.1).
A• Comparison of calculated and measured pressures inside the entryway gave
assurance that input parameters for the design event (1 MT, 50 psi) would
yield a reasonable estimate of the airblast record inside the entryway during
that event.
Since the peak pressure on the blast door wou higher than that predicted,
the response of the blast door would be expected to be slightly more severe
than pretest calculations indicated. A single-degree-of-freedom response
45
15
analysis was again performed for the blast door with the airblast record
from the center of the door input as the loading function. For this analysis,
the door was considered lightly damped (3 percent of critical damping). Re-
sults of this analysis are shown in Figure 5.2. The calculated maximum re-
sponse of the door was 0.45 inch at 8 ms after the door was initially loaded,
resulting in a ductility of about 1.7 and a permanent deflection of about
0.18 inch. Recall that the measured permanent deflection in the door was also
about 0.18 inch.
Since both the internal pressure calculations and the blast door response
predictions compare favorably with the measured data, response predictions due
to other weapons and/or pressures should approximate the response with little
error.
The DIRECT COURSE event provided a I-KT simulated nuclear airblast en-
vironment from which response and pressure predictions could be verified.
However, survival of structural elements in a l-KT environment does not guar-
antee survival in a 1-MT environment at the same overpressure level. Airblast
predictions for a I-MT, 50-psi event were necessary to verify the surviva-
bility of the blast door in that environment.
Internal tunnel pressures were calculated for a 1-NT, 50-psi event with
the following input:
Peak pressure = 50.0 psi
Positive phase duration = 964 ms
Impulse = 12.4 psi-s
The predicted loading history for the blast door is shown in Figure 5.3. The
calculated peak pressure on the door is 165 psi. Note that this peak pressure
is substantially higher than the peak pressure of 142 psi predicted for the
I-KT, 50-psi overpressure level at the DIRECT COURSE event. This is because
the airblast associated with a 1-KT event decays much more rapidly than the
airblast from a 1-MT event at the same overpressure level. Hence, the inci-
dent pressure at the closed end of the entryway will be higher for a 1-MT
event than for a I-KT event, and the reflected pressures at the same point
will be higher by an even larger margin.
Response calculations for the blast door were performed with input from
Figure 5.3, with the following results:
46
IMaximum deflection = 0.55 inch
Ductility, p = 2.0 (see Figure 5.4)
Permanent deflection = 0.28 inch
These calculations indicate that the blast door will sustain very light damage
and should remain completely serviceable when subjected to a I-MT, 50-psi air-
blase environment. (Recall that the calculated ductility from the DIRECT
COURSE event was about 1.7.)
Although the blast door will survive the airblast effects of a 1-MT,
50-psi event, a slightly higher surface overpressure with a worst case orien-
tation will cause substantially greater damage. For example, a 1-MT event
with a surface overpressure of 60 psi will result in a peak pressure of
204 psi at the blast door with a duration of about 900 ms. As the ultimate
resistance of the door is about 180 psi, this load would very probably result
in the complete collapse of the door.
When a shock wave strikes the end wall of the dead-end entryway, a re-
flected pressure is instantly developed on that surface. The magnitude of the
reflected pressure varies from about double the applied pressure to greater
than a factor of 8 times the applied pressure, depending on the magnitude of
the incident pressure. Since the peak reflected pressure acting on the door
in a dead-end tunnel is so sensitive to changes in the surface overpressure,
it may be desirable to consider a pass-through tunnel system, in which the
shock wave is allowed to flow through the tunnel past the closure.
47
Calculated Maximum Response of Blast
P psi Peak Pressure Door in a Pass-Through Tunnel,
so, on Closure, psi in.
50 77 0.15
100 142 0.29
130 180 0.44
135 186 1.2
140 192 4.5
150 204 14.5 (probable collapse)
These calculations show that the blast door, when placed in a pass-through
entryway, should survive the airblast effects of a peak surface overpressure
(P so) of about 135 psi. Survivability of the shelter closure at 135 psi
yields a decrease in the lethal radius from ground zero of about 35 percent,
or about 1,500 feet. While this difference may seem insignificant, inclusion
of a pass-through entryway would obviously provide a beneficial factor of
safety for a 50-psi-rated shelter.
A rough cost analysis of the two entryway systems examined indicates that
the pass-through entryway could be constructed for an additional 60 percent
above the cost of the dead-end entryway. This figure would approach 50 per-
cent if the stairwell formwork is reused on the mirror-image second stairwell.
4
4.5 180 -
4.0 160 -
-2
3.5 140 -
3.0 - 120 -
U 2.5 -,100
xIJ
' 2.0 -I 80
-J
- \ -CA L CULA TED DATA
• 1.5 60 -
II
1.0 40 -
0.5 20 - \
. 0 0- • _-.-
-0.5 -20 I I I I I I I I I
100 120 140 160 180 200 220 240 260 280 300
TIME, MSEC
49
0.5
Y (MAX) 0. 45"
0.4
-0.3
0.
0~
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
TIME, SEC
50
200
w
D 100
w
50
50
_,0 -II I I I
0.8
z
2
•'. • 0.4
LU
04
Y (PERM) =0.28"
0.2
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
TIME, SEC
51
Figure 5.5. Pass-through
entryway sYstem.
400
DEAODE,cVD
300
0
0
0
K wj20
0.
20 406100002
10 10 8 0
SURFACE OVERPRESSURE,
PSI
Figre
.6. Co m ari ~
0 Of incident to
in dead-end peak
and pass-through pressure on Closure
entrrway sYstems.
5n
CHAPTER 6
6.1 CONCLUSIONS
53
were unable to support the load between each beam span. A much thicker, and
thus heavier and more expensive, steel plate would be required to insure sur-
vival of this closure concept. Placement of the supports prior to the test
was a cumbersome process and required an undesirable amount of time, espec-
ially when considered in the context of an imminent nuclear strike. The cost
of the commercial door, door frame, and fabrication of the supports was about
$750. With thicker, more expensive, steel plates the door might be made to
survive, but the problem of placing the supports would be much worse. At the
pressure levels examined in this test, standard commercial doors are not a
practical alternative for blastproof shelters.
The entryway wall and roof slabs survived the DIRECT COURSE event with
only minor cracking, and the structure remained completely intact. As ex-
pected, the greatest damage occurred at the slab joints, since the net effect
of the interior and exterior loads was to punch out each corner of the tunnel.
In no case was the damage extensive enough to warrant changes in the slab de-
sign. The slab design tested, which was 6 inches thick with principal rein-
forcement consisting of 1/2-inch bars spaced at 12 inches on center, is ade-
quate for a 1-NT, 50-psi event. Higher overpressures would require slightly
thicker walls, but this parameter is not as sensitive to changes in peak pres-
sure as might be expected because of the additional strength provided by the
soil backfill.
Calculations of airblast in the tunnel were performed with the ANSWER
computer code, which was developed in the Explosion Effects Division of the
Structures Laboratory at WES. Predicted peak pressures inside the full-scale
structure were remarkably close to those measured at the DIRECT COURSE event,
and predicted durations and impulses were equally accurate. The ANSWER code
has proven to be a valuable tool for predicting airblast characteristics in-
side a tunnel.
6.2 RECOMMENDATIONS
54
be used when the design overpressure is greater than about 25 psi. The cost
of upgrading a standard door can be larger than the cost of a reinforced con-
crete door.
3. Additional static tests of the blast door should be conducted to deter-
miae the minimum amount and spacing of sheai reinforcement needed to insure
ductile behavior.
4. The blast door evaluated is recommended for use in a 50-psi shelter with
modifications to the shear reinforcement. For reasons of economy, welded
shear studs should be used in place of bent reinforcing steel. These shear
studs should be spaced at ne> more than 1-1/2 inches on center near the door
supports to insure ductile behavior at higher overpressures.
5. Shear stirrups may be safely omitted from the entryway walls, but should
be included in the roof slabs. A minimal number of stirrups should be used in
the walls to keep the principal reinforcement properly aligned during concrete
placement.
6. Backfill placement around the entryway should be tightly controlled to
prevent excessive outward wall mcvement.
55
REFERENCES
.56
13. G. A. Coulter; "Blast Loading in Existing Structures-Basement
Models"; Memorandum Report No. 2208, August 1972; Ballistic Research Labora-
tories, U. S. Army Research and Development Center, Aberdeen Proving Ground,
Md.
14. Civil Nuclear Systems Corporation; "The Air Force Manual for Design
and Analysis of Hardened Structures"; Technical Report AFWL-TR-74-102, October
1974; Air Force Weapons Laboratory, Kirtland Air Force Base, N. Mex.
15. J. M. Biggs; "Introduction to Structural Dynamics"; 1964, McGraw-
Hill Book Company, New York, N. Y.
16. Committee on Structural Dynamics, Engineering Mechanics Division,
American Society of Civil Engineers; "Design of Structures to Resist Nuclear
Weapons Effects"; Manuals of Engineering Practice No. 42, 1961; American
Society of Civil Engineers, Now York, N. Y.
I
57
APPENDIX A: PRESSURE DATA
5,
59..
[DC ENTFRYNRY
P 1 / 4 0g95
I2OO0O HZ CRL_- 87•50
LP'3/O "707/ CUTOFF= I 1000 Z
Cl
L. Q /'c
Ll -
C,3
-I
60
DC ENTRYNRY
P2/4095
200000. HZ CRL= 94.70
LP3/O 70X CUTOFF= 11000. HZ
-
-
(Ii CD
/
Lo
.. 0(L ..
L.J
W w
0,..
C;.
•4 tO 1• . liO. 16,0 150. 42.00. 22,0. 240. 26O. ZOO. 300.
--- - .- - 0
TIME IN MSEC - - -_'1
61
DC ENTRYNFIY
P3/4095
200000. HZ CPL- 76.40
"LP3/0 7OX CUTOFF= 11000. HZ
S_=_= -
- 0)
(0
0- W0
:f_ =___,
m
C,.
LU- 2
CD- ocw"
to.C1 . 10 6 . 0 . 2 ` 4 , ' O 3 0
TIE INME
--- •
0~ 62
DC ENTRYWRY
P 4/4095
200000. HZ CRL= 191.5
LP3/0 70% CUTOFF= 11000. HZ
U,,- - -...
LUn
U)l
LUn
>(N
•o V)
0-
C, 0
• ° I
100. 120. 140. 160. ISO. 200. 220. 240. 260. 280. 300.
TIME IN MSEC
63
DC ENTRYNRY
P5/4095
200000. HZ CRL= 196.-3
LP3/O 707. CUTOFF= 11000. HIZ
0 o
c?.
V)
Lo
4 (,
V)
- U)
64 ki
DC ENTRYVRY
P7/ 4095
200000. HZ CRL= 9.980
LP4/0 70Z CUTOFF= 9000. HZ
i; °
I ~0€
UJ
i "i
C,) ___a-
0-n
u')
0--
iX: c-°
::7
0 ~C
i00. 1-,0 - 140. 160. 1,80. 200. 220. 2,4O. 260. 9-80 300.
TIME IN MSEC
65l I
DC ENTRYNPY
P9/4095
200000. HZ CRL= 86,00
LP3/0 70% CUTOFF= 11000. HZ
WINl WINl
'iC
. , C,_ _ _ _ _ _ _ _ _ _ _
L,
LL;
CD
a- C;~
Lo M
i Ul
=D (013C
(i2
0
CD.
o00. 120. 140. 160. lfJU 200. 220. 240. . 280. 300.
TIMF IN MSEC
66
'V
DC ENTRYWRY
P10/4095
200000. HZ CRL- 235.4
LP3/O 707, CUTOFF= 11000 H7
2U,
"L m
o'
L a
0- (fj
._J 106
U).
-$. Lfl
i_ "~TM
I 0, i2 ] . iO. '"
i6 , IN
60. 20 . -,0."' 40 •60. 29 . 00
_.._iTIME N MLE
'1I__
g -7
--- ---
X a
"0-
0- CI
!•
'I0 |2 0i. 14,0, t6O. 180.
TIME IN MSEC
200. 20.- 240, 260. 290. 300.
68
DC ENTRYNRY
P 12/4095
200000. HZ CRL= 87.30
LP3/0 70Z CUTOFF= 11000. HZ
C'
6,t
L 0 -
in 0 :D-
0-~ __ - - *
L)L
M IN M
z i
0. 20. 40. O. 0. 200. 0. •0. '60. 280. 00
TIME IN MSEC
/69
DC ENTRYNPY
P1 3/4095
200000. HZ CRL.- 85.40
LP3/0 7,X CUTOFFz 11000. HZ
EtR 1111
9676 - i1/14/ql3
1 r.911
Er.)
U,
U:C; - __ _o _
,- U
* 07
...
0- C3
(J2 :•
,,(0 0' Q.
u . S..
I ,oj
[C ENTRYNRY
P14 /4095
2000006 HZ CRL_- 87.60
LF3/O 70Z CUTOFF= 11000. HZ
* ,,=____ -.
C3
* /,
* 0 .
(') 0' CL
LO Cl..
,; -J
U)
EL
C)
° 1/" ,-,,,
C, , •
0
100. i20 i40. iS . 19c• 2J0 220. 40. no- 2:10 l00.
"TIME. IN M-r"EC
71
-- = = =-= -
DC ENTRYIRY
SP1 5/4095
200000. HZ CRL- 88.30
LP3/0 70Y% CUTOFF= 11000. HZ
EUl WE
C3
*000
Li..
0- U/
U) /\0.
0,lil
-• UJ 100'tO
. 120. 140. 160. ISO. ZZ00. 2.20. 240. 1160. 1180. 300.
•,:;TIME
.PEAK VLUE 15 3 IN "MSEC.
OVER CALIBRATION
72 ,•
DISTRIBUTION LIST
35807 Commander
Huntsville, Ala.
Harry Diamond Laboratories
Commander r Division, Middle East (REAR) ATTN: AKXDO-NP
US Army Engineer iv DELHD-TA-L/Technical Library
ATTN: Mr. A. Oliver Werner/MEDED-TS DELHD-N-P/Mr. j. J. Meszaros
Po Box 2250 2800 Powder Mill Road
Winchester, Va- 22601 Adelphi, Md. 20783
Commander Director
Engineer
US Army ATTN: District, Omaha US Army Ballistic Research Laboratories
Hr. Tim Knight
Tames L.nATTN:
HTTN:Mr. Mr. J. H. Keefer
Mr. James L. Sack VLD/Mr. Steve Zardas
6014 USPO & Courthouse Technical Library/Mr. Edward Baicy
215 N. 17th St. DRDAR-BLT/Mr. C. W. Kitchens
Omnaha, Nebr. 68102 Mr. George Coulter
Aberdeen Proving Ground, Md. 21005
Commander
US Army Engineer Division, Ohio River
Comoander Research Center
ATTN: ORDAS-L/Technical Library US Army Materials & Mechanics
PO Box 1159 ATTN: Technical Library
Cincinnati, Ohio 45201 Watertown, Mass. 02172
Center Commander
Defense Technical Information Jr. US Army Nuclear and Chemical
Agency
12 cys ATTN: DDOB/Mr. Harry A. Schrecengost, ATTN: MONA-WE/Mr, John Uecke
Cameron Station Technical Library
Alexandria, Va. 22314 MONA-WE
7500 Backlick Road, Bldg 2073
Commander Springfield, Va. 22150
(BRADC)
US Army Belvoir R&D Center
Technical Library Commander
ATTN:
US Army Materiel Development and
Fort Belvoir, Va. 22060 Readiness Commacnd (DARCOM)
adin: Co D-WN
MAJ G. Gocsaneanu (STRBE-CFLO) ATTN: ORCRD-wN
Canadian Forces Liaison Officer 5001 Eisenhowed Ave.
Center
US Army Belvoir R&D 22060
Va.
Fort Belvoir,
nAcustoChief of Research,
Office, Development,
C aa
Commandant
and Acquisition
US Army Engineer School DAACS
Army
ATZA-DTE-SUATTN: Department of the
ATTN:
Fort Belvoir, Va. 22015 DC 20310
F v , 2Washington,
(R&D)
Assistant Secretary of the Army
Commander ATTN: As-stant for Research
US Army Missile Command
Info Command
Redstone Scientific Washington, DC 20301
ATTN: Documents
Pedstone Arsenal, Ala. 35898
4dk%1M J~
Northwestern University
Command and Control Technical Center
Department of Civil Engineering
Department of Defense
ATTN: Dr. Ted Belytchko
Room 2E312 Pentagon
Evanston, Ill. 60201
Washington, DC 20301
Agbabian Associates
Headquarters, Department of Energy
ATTN: Dr. M. S. Agbabian
ATTN: Library/G-49
250 N. Nash Street
Washington, DC 20545
El Segundo, Calif. 90245
f
Superintendent, US Naval Postgraduate School Ballistic Missile Office
ATTN: Library Code 2124 ENSN
. I
-z
Kaman Sciences Corp. Scientific Services, Inc.
ATTN: Gunning Butler, Jr. ATTN: Dr. B. L. Gabrielsen
PO Box 7463 Mr. C. Wilton
Colorado Springs, Colo. 80933 517 E. Bayshore
Redwood City, Calif. 94063
Mason & Hanger, Silas Mason Co., Inc.
ATTN: A. T. Papp Southwest Research Institute
Larry K. Skeen ATTN: Mr. Phil Nash
Pantex Plant, PO Box 30020 Mr. A. B. Wenzel
Amarillo, Tex. 79177 6220 Culebra Road
San Antonio, Tex. 78284
H. L. Murphy Associates
Box 1727 Stanford Research Institute
San Mateo, Calif. 94401 ATTN: Dr. G. Abrahamson
Dr. A. Florence
Physics International Co. Mr. Jim Gran
ATTN: Technical Library 333 Ravenswood Ave.
2700 Merced St. Menlo Park, Calif. 94025
San Leandrc,, Calif. 94577
TRW Defense Systems Group
Pacific-Sierra Research Corporation ATTN: Mr. Norman Lipner
ATTN: Dr. Harold L. Brode Dr. Paul Chen
12340 Santa Monica Blvd Tech Info Center/S-1930
Los Angeles, Calif. 90025 One Space Park
Redondo Beach, Calif. 90278
R&D Associates
ATTN: Technical Library Terra Tek, Inc.
Mr. J. G. Lewis ATTN: Mr. S. J. Green
PO Box 9695 University Research Park
Marina Del Rey, Calif. 90291 420 Wakara Way
Sait Lake City, Utah 84108
R&D Corporation
ATTN: Document Library Weidlinger Associates
1700 Main St. ATTN: Dr. Melvin L. Baron
Santa Monica, Calif. 90401 333 Seventh Ave.
New York, N. Y. 10001
Research Triangle Institute
ATTN: Mr. Edward Hill Weidlinger Associates
PO Box 12194 ATTN: Dr. J. Isenberg
Research Triangle, N. C. 27709 3000 Sand Hill Road
Menlo Park, Calif. 94025
I u 4) 93 1 u 4) x cl La
ýu 4J $4 .14 a) -H 4j 4) 0 Aj $4
4 JU 4 -1 .4 M Z 4J 4J 4ý 60 (U
P-4 = w 401 4-4 00 4) u 0 4v 0 Q 41 41 u 0 1 0
4 " (U 0 0 .93 A lu A 41 0 (a 0, 0 cl (U .0 0
3t 0 m 41 1) cc 14 41 414 to W P. 3ý 0 m 1 41 to H ýA LH El ca
-0 -4 -4 C) ýj a 4-4 0
Ln cc 4) ýo jj 4) Lr.l m 10 -4rj =U 4) M .0
'1 \0
0 .0
ý) r. 4J
Qj 4-4
Q 0
Aý
r, 41 C3 w r- 4j m $4 54 Ow cc
4J t6 '14 4) 4) -H 14W 4, coc -1 w > 4) 4 0
4) 0 ft 4 1 3t ?4 0 0 0 4) to 0 -4 4) . 'o to m 1 3: C-0 m 0 0) 4)
(U -4 (n 0 J3 4) w d3 -4 4J .0 (d 4 IT w 14 Cl) 0 X1 Qj W .0 m
44 00 Aj .0 >1 m 93 .0 0 4 4J fA 44 00 41 .0 >1 0 cc w cl .0 :3 4j W
W IX M 0 M 4J (A .0 (o 4) (U 44 4) 0% 1.4 m 4) C4 co '0 m 41 (A Q 0 4U 4) 44 Gj
t4 1.4 V $4 oj Q) -,4 Ea w to
U5 m 0 r-4 41 44 6j 41 H 4) [a - as 10 = (U -4
0 vj m ftl -,
A 'o
93 44 .1 u m 0 m w m 0 w w 'A to
1 J3
41 41 0 00
m Q) "a
to pu
vo ý4
.14 0 ta u to 0 do W M 0 Q $4 0W M
-, I u m 0 m 2 a CA, 44 v " w 14 u to o -m w
w
0 ý4
t3 U) W
4.e =ýj 4J
U " 00
W
u C3 >, w 14 to 0 0 v Q 4J 0 U u C3 >1 $4 .4 0 a a V 0 Q w 4.; u
60 M 4U - U th 4J $4 W 0 4) 4) to 0 4h U M ýj w W V A 10 4) 4)
(a -m V 01 2 0 W 4J 04 U 44 $4 -A -- -0 w :3 m (U 41 00 u 4.4 w
41 ýj ul u c 4, 0 [a 4J (U '%14-1 0 4-0 41 r. 44 =1
C3 4) w 0 41 x to V C3 10 Q 'A C3 W 0 w C3 " x 0 u "o C3 -H 'a j (A
;A 0 -0 4J V 0 M W 40 >1 -r4 4) 4) 4) w 04 0 -0 4J U M 'A M >, -M 4) 4) 4) 4) 1. a) 'A
-,i C3 10 -4 .0 ý4 Cl U 4J 4) C3 sw -4 > m u 41 4)
4j 0 4) V t Dh C3 4 -0 $4 14 W 41 0 lu 0 0.44 "0 14 14 W
41 cc 0 as L24 lu 0 Ln,
-C 41 $4 th Cn 44 w A Cn -4 a 44 r-4
w 4) 4) cu v
-0 In qj W 4) 0 4 W
41 4) m ., 49 (A ta 1% .14 > 41 1.4 .1 .0 J4 CA U3 0. >, -4 -H
6 >
4) 0
41 w 4j > w 4 C4 a 4) 0 41 Oj 4 Oj 'M 4J 4J > $4 M 0.
0 31 4) 4 C3 -1 0 0 u $4 4j w m m 39 4) 4J 4 C3 -H 0 0 lu
m u 4) 1 $. m 0 .0 41 -H C4
4J q) w - 0. r-4 r4 Le) 0 00 W w IV (A 41 A(A to
41) U
4) ýi4 W U
Qj C3ýLn
M 0 400 ta 4V CU go
4-, P. 4) '"' "' a - "D 4) J". d 0. .'4 U W .0 - .0 Q 04
4) 41 to .4 to 41 0 0 PO 4j 4j 1 w w J 4j ýJ 4 4J 0 93 41 li I
.11 'A 4) E, m C3 A4 4 to (u 0 0 w .1 0 v m C A4 m 0 0) V3
ow 0 Ili sw = 'A I M ; 4J >, ý) 10 U) 93W 0 m w (a W 1-4 to to
w 1-4 4) 41 41 to cu ci -4 m 0 z x w 1.4 w j 41 m 1 41 w 39 W >1 lz
-4 'd to (U w m 4)
CID w ".a ýl -H 93 31 0 :3 0, La o w 0ý .0 3ý .14 12 3; :3 " V3
31 4) 44 C6 u m 41 41 = - 29 w 41 0 44 0, W u 0 m ýj 4-) il
U3 to (a 4) CO 0 > 4J I 4j 0 0 0 'A 41 cn w rA 4) m 0 p &) w I U 0 F4 0 4J (a 41
4.) 0 W 0 4) a ID 0% 0 0 .,q 41 0 w 0 Gj C3 , 0 .0 0ý a 0 0 4) =
41 o 0 -0 0 > 41 ýo u 44 ý4 d v 0 0 10 4) > 41 10 A 4.) ýj ýj 41
w .14 0 > 0 m to 4j W - 0 > 0 m 41 M -H 0
u 41 10 lu 0 a) W fA .14 (a u 3; u 41 V w 0 w v w to 41 .0 ..-1 ra u 3t
4) 44 r. M 44 0 0 Q m -rý 93 do j 9 " '* W " 'o .. 0
41 0 0 w $4 *j
w 4 3c m t, 3t V >% 41 0 tl W 10 cd tj 3r 0
'D u 0 0 w 0 w 0 a Q ýq w C3 - .. m u 0 0 w 0 0 0 C: w
ot 0 00 4 u U I ., 4) ;., .0 0 V) 6 -4 4j U) .0 00,4 U U g -M Qj ýQ 0 En W -,4 IJ
r.3
w
gy
4)
>
-r4 u
a 114
4) 44
0
w
w
41
cl
0
u
41
1
10
44
CC
41
.0
`a'
-4 0
6r m8 0
4w am
0
w
4)
(U 4u -H tý
>
-,4
44 41
U 93 4J U L) U
0
0 E.
4J
0
8
4.
4 'D
m
.0
w
44 P4
cl
0
0
:3 m
Ea +4 r.
0 4) (a 0
4) 41 w 0 x C3 10 6 +j 4) u ta ca W Q C 10 (1) U 0 ?A 4 1
0 u 0 u 4) -0 lqx H 10 0 - 41 w - >, 4) W C4 C3 'Lid 10, 10, om 11 .0 0 - 4j $4 4; 0)
CS 1+4 M 14 W -0
-0 4V Q U 44 m a U 00 -W" 404 W -0 4u) C-) 44 0 '*04 0) EU -Mm Wu -:$
ýq 14 A u 0 0 w cl .0 -A 0 W W u z w
4) 0 w 4) L. F4 1ý 'I W m 0 4) 0 m 0 4) 0 W W m ra m
W 10 44 .0 60 0 IJ .0 0 0
m w 41 41 w - w lu 10
C3 44 0 0 4j 4j
In 0 93 41 00 0 41 = q) I
m , .0 64 C (a 'I u 0 m .0 w 0 . -4
41 W 4J 0 44 0 O.J 44 0 4) .ý :3 Cý ýg ýj 4) 41 .0 C3 1" 0 0 4j 0 W uz
CIM E. 0 W to 10 0 41 (j 0 u 41 3t a pq -H E. to w m I- m 0 .0 li u 0 u li 0
J4
0 v W x 93 In
v 41 0 0 *j w qý -M 4j 4, 0 j w
Q 41 00 w u 0 1 0 A w = 41 u I
0. m 0 .0 Aj 61 (4 0 93 .4 Qj A 41 0 92W It 0 A J W IA 0 4d 4 ,
3; o to I j ol (21 m 41 %. wV do 4ý w J
10 A 0 ýl a 1w 0 q 0 4. . , %ý.:
Ln 0 . 11 w .0 w W ej I m U 4 10 .0 Q W A4
r- 'i ed w co 0. m IJ 4 a 14 0. M tA 4
V 'A > o 0 ta > .,4 aj 4j 0 Id r. V
w .0 4 31 60 4 0 4) w 'A 0 ow m m 1 74 00 a 0 v 4) La 0
4) -4 m ej w to A m iu 11 4J w m 41 .0 m
4w 00 j j t6 44 00 j .0 "1 1. r. A = - w
CY, to w 04 9 0 4 41 W AZ 4) 4) w w x m 0 a j ;a .0 M W IU 4ý 4)
6 Q li 0 cc w 0 04
to 93
so 4) M M 93 W U M M fe w go 0 lu w to fa 1, U .0 "1 .1.0 FýA ýu Iffli 41
u m 0 2 cl 4. 4 u j $. u m 0 :3 a c u ýj w
u cl >. w 1ý 'A Sl 0 " w w " = u C3 >, w - = 0 w v m w j Q ýj
00 u 'A j w w .0 .0 0 4j v 40 'a u 0 j w 6 c c .0 0 w 41
; 13 w U 44 W (; ., .0 41 p :3 lu j cc u 4. w
. = 44 V r, 4) -0 j 0 j , a 1.4 0
41
'A 1 li 13 - Q
W 10 M to 0 C,
a I- v lu 'A
0 -0 41 (d 0 fa (4 40 ýH 4) V 6 c 44 z 0 Vcu j u 0W ma w ,
m x 041 4)u V
0 C3 ..w
v 'a
v w la
0
. 0 0. 10 - a - > z u " Q 0 .. 0 " -C - > a 41 w
j 0 4) w . - j 0 4) 19 " >% 0, ý - m L4 6
11 14 W , 4.0 -00 00 0
4A
do I., On
C)
Q > 0 0 6
.0 . li
to) (a 41 V L X, 4d (n qj 64 v j
6. > %. .1 - Ij .. V (A Ift ý w
.. 0 : 41 - 0) q j j > $. (4 0 V 0
11 *Ci I.J >0 10, 141, 1. j W m > r. 31 W 4j 0 0 w
A to u 41 6d u A ch 0 c 10 U 4ý I W U 9:6 ON Q X
0 60 G c 0
CL W -00 .60 41 14 m
11 IL -0 tC to
w w j ta A li a 41 0 1 0 C3 "a
W ." w w .1 n m 0 w w Go 0 -A 'k - 0 0 11 1-4
0 do 'A - m , j ,
$. j 4 W W 4) 4 Q 0 z 0 a w A - ; -c - >, - - m 0
31 31 2 0. La :ý - C3 0 Q.
31 w 4j C U 0 q j .4 31 w 41 0 CL.6 u C3 4 J
(n 14 4) w , 0 > 4, 0 c 0 j'a tn ce 44 -4 0 > j 6 1 40 0 a 0 j w
91 0 Gi Q 0 0 0 6. 0 t r - 4) CY, 0 0 0
0 0 v 0, > 2 v uG j .1 j Q c 0 0 10 v > 10 u ý4 "
0
.0 4) w
p > a4) 'o j a . W .. 0 W > C Is a le
li to ,j IV .. 0 Q c . 0
4d I. C 41 , . a 4 0) 31 4 4 'A 4, ýl 4 r
0 0 :D w Q 31 Is u 3. 0 0 w 6. a -J m V 0
u p 0 0 w 0 0 0 a, WJ r.
0 W.2 u u -,.o J; 4 40 u w .0 Ln w
-0 .. w C: , 0 4d "V 0
w - v () 4 , 0 C3 Lj > w 'o t 6
u a .4 u r, a C) 10 41 IA 0 41 u c u u z 07 Z '14' '0
6j w 0 x
I j tj U M 'A v 6 0 x %i v
CK a 0 u w CK c L, 0 t, qd v it Qý .0
Q .0 .
a c 6 u ca 14 9 1 '. r, , 4
iclu r, u
0 § 4, 0 w
12 o 0 am 0 41 r 40 6
6 V be C r cc q 4 41
0 :3 64 .1 0, C r
a ft 0 .0 u o go 0 -0 L,
-All %I