ML20151A121

Effect of Natural Phenomena on Existing Pu Fabrication Facilities,Response of Structures to Wind Hazard at Atomics Intl Nuclear Matls Development Facility,Santa Susana,Ca, Vol I

ML20151A121
Person / Time
Site:	07000025
Issue date:	08/31/1980
From:	Alikhanlou F, Mcdonald J ARGONNE NATIONAL LABORATORY, TEXAS TECH UNIV., LUBBOCK, TX
To:
Shared Package
ML20151A116	List: ML20151A114 ML20151A117 ML20151A121
References
REF-PROJ-M-3 NUDOCS 8010240558
Download: ML20151A121 (40)

Text

.-.

O EF ECT OF s ATURAL 3-ENOMENA ON EXIST NG 3LUTONIUM FAB;lCATION FACILITES Response of Structures to Wind Hazard at the Atomics International Nuclear Materials Development Facilliy Santo Susano, Californio Volume !

-v.

ns':i:u:e fo Dises':er Researc,

~~EXAS ~~EC-Us VE AS Y

_ubboc s,~~ exes 792.09 8010240 5 6 8

i L

i THE-EFFECT OF NATURAL PHENOMENA t

ON EXISTING PLUTONIUM FABRICATION FACILITIES-k

. RESPONSE OF STRUCTURES TO EXTREME WIND HAZARD at the ATOMICS INTERNATIONAL NUCLEAR MATERIALS DEVELOPMENT FACILITY Santa Susana, California i

VOLUME 1 h

1 by Kisher C. Mehta James R. Mcdonald

.Faribor: Alikhanlou Institute for Disaster Research Texas Tech University Lubbock, Texas August 1980 i

t FOREWORD-The U.S. Nuclear Regulatory Commission has undertaken a oroject to analyze tne effects of natural phenomena upon' existing plutonium fabrication'

]

facilities.

The work is being accomplished by a task force of experts who are contributing to the various. phases of the project.

This report is one of a series of reports, to be' oroduced by Texas Tech University, which examines the response of structures and-the damage consecuences to an existing plutonium fabrication facility caused by severe wind.

The Atomics International Nuclear Materials Deveicoment Facility (NMOF) located at Santa Susana, California is the subject of this report.

Volume I of this report presents the methodology, the basic data, the results and the conclusions of the study.

Volume II contains the structural calculations on which the results are based.

The project tasks are performed by Texas Tech University under sub-contract from Argonne National Laboratory (Con' tract Number 31-109-38-3712).

Mr. James E. Carson, Division of Environmental Impact Studies, Argonne National Laboratory, is the project manager.

Dr. James R. Mcdonald and Dr. Kisher C. Mehta of Texas Tech University are the principal investigators for the project.

The project is coordinated through the Department of Civil Engineering anc the Institute for Disaster Research, Texas Tech University.

l I

i

.1-4 TABLE OF CONTENTS Pace I.

. INTRODUCTION 1

II.

STRUCTURAL SYSTEMS AND MATERIAL PROPERTIES 3

A. -General Layout of the NMDF Building.

3 B.

Structural Systems 3

C. -Vault 13 D.

Material Properties 13 III'. -STRUCTURAL RESPONSE ANO. DAMAGE CONSEQUENCES 16 A.

Threshold Windspeeds to Produce Damage 16

)

3.

Atmospheric Pressure Change (APC) 20 C.

Combined Wind and Atmospheric Pressure Change 22 D.

Windborne Debris 23 E.

Damage Consequences 23 IV.

THRESHOLD WINDSPEEDS AND FAILURE MODES 27 A.

Potential Damace to NMOF Facility 27-B.

Atmospheric Pressure Change Effects 30 C.

Potential Damace from Windborne Debris 30 D.

Summary of Potential Failure Modes 31 V.

DAMAGE SCENARIOS 32 A.

Damage Scenario for Nominal Windspeed of 110 mph 32 B.

Damage Scenario for Nominal Windspeed of 130 mph 33 C.

Damage Scenario for Nominal Windspeed of 150 mph 33 D.

Damage Scenario for Nominal Windspeed of 170 mph 33 VI.

REFERENCES 35 ii

LIST OF' TABLES

. Table Page I:

Material Propercies-15 II

. Tornadic Windspeeds; Atmospheric. Pressure Change and Requirements 'for Venting.

21 III Windstorm. Generated Missile Velocities

$4 IV Damage Consequences 26 V.

Threshold Failure Windspeeds for:NMDF 28 LIST OF FIGURES Figure Page 1

General Layout of.the NMDF Building 4

2 Roof Framing Plan of the NMDF. Building 5

3 Typical Section - East and West Walls 7

4 Typical Roof Diaphragm Connection at Walls 8

5 Opening in Exterior Walls Exposed to Outside 9

6 Typical Precast Concrete Wall Panel 10 7

Shear Wall System 12 8

Section Through Vault 13 g

Wind Pressures on a Building 18 n

iii

j I. ' INTR 00'CTION U

This report is cart of a study sponsored by the U. 5. Nuclear.Regula-

. tory Commission to assess.the cotential radiological consequences of natural

~

phenomena--(flood, earthquakes, and severe. winds) on existing plutonium fabrication _ facilities. ' The study involves determination of hazard risk, structural response, source term, dispersion, demographic patterns and dose levels.

The paper by J. A. Ayer and.U. Burknardt,'" Analysis of Effect of Abnormal Natural Phenomena on Existing Plutonium Fabrication Plants" [1]*

provides background en the overall hazards' evaluation.

The response of-structural systems and components to wind hazard at the Atomics International Nuclear Materials Development Facility (NMDF) located at Santa Susana, California is the subject of this report.

The windstorm risk assessment was made by Fujita [2] based on tornaco and other severe wind recorcs from the geographical region' surrounding the plant site.

The windstorm hazard at the site consists of straight line winds or tornadoes and is expressed in terms of expected value of windspeed for a given probability of occurrence.

Associated with tornadic windspeeds are implications of atmospheric pressure change and windborne debris.

Structural response of the building and the potential damage from windborne debris are expressed in tems of threshold values of windspeed to produce postulated damage to the building enclosure.

The damage postulation is based on ten years of windstorm damage investigation experiences involving more than fifty windstorm incidents by the senior authors.

The structural response and missile impacts are subsequently translated into consequences of

" Numbers in brackets pertain to References 1

damage to glove boxes and filters.

These consecuences then provide infor-mation to the source term evaluators who, in turn, determine the amount and-form of plutonium that would be available for dispersion into the atmosohere.

The type of structural systems and construction material properties at NMDF.are discussed in Section II of this report.

The structural systems and the material properties are documented from the plant drawings and specifi-

.catiens, the EDAC Task I report [3] and a s te visit.

A generic discussion of structural response to the windstorm hazard,: including the effects of wind, atmospheric pressure change and windborn debris, is containec'in Section III.

The consequences of damage to glove boxes and filters are also defined in Section III.

Section IV contains postulated failure modes, calculated threshold windspeed values, and a sumary of postulated damage for NMDF.

Actual calculations of the values presented in Section IV are contained in Volume II of this report [4].

Scenarios of expected structural damage and the consequences of damage to plutonium containment for selected windspeeds are presented in Section V.

l l

2

.g r

II.

STR.TURAL SYSTEMS'AND. MATERIAL PROPERTIES In this section, the structural systems emoloyed in the NMDF building are. described'and material properties common to them are defined.

Only those features of the structure,tnat are critical to wind hazard assessment-

~

are presented herein.

A.

General Layout of the MNOF Building The HMOF building is located in Santa Susana, California.

It is

~

a one-story,'rectangula'r building with plan dimensions'of'62 ft x 202 ft.

The roof.of.the building is 17 ft above ground level.

As shown in Figure 1,.

the building is civided into an office area, support laboratories, a glove box room, a vault, an exhaust filter room, air. lock and other equipment. rooms.

Mishima [5] has identified the critical areas of. concern.

They are indicated in Figure 1.

The building has no windows; there are, however, several, exterior doors.

The doors in the exterior walls of the glove box room have enclosures around them to prevent exoosure of these doors to impact from windborne debris.

The cast-in-place reinforced concrete vault is located on the west' side of the building.

It has heavily reinforced walls and roof and a heavy vault-type door.

B.

Structural Systems The roof framing plan for.the NMDF building is shown in Figure 2.

The primary structural system consists of steel roof girders that span 61 ft 8 in. in the east-west direction at 20 ft on centers.

The roof girders connect to the columns at the exterior walls with simply supported bolted connections.

Channel sections span between the columns at roof level.

Pre-cast concrete wall panels are supported by the columns and the channels to form the wall enclosure.

The roof is constructed with a cellular-type metal 3

~

201' 1/2" 70' 1/2-90' - 0" 40' - 10" ly1 II I

I

/[

Glove Box Area A/C {

Chemical

/

Support Equip. Room ll l/,

l i

Exlians t Fil ter Room

_f

/

' Office Area f

[

(

Gl_ove Box Room-Al lo General T

Support Electric Emergench Enoip. RoomS"['

p Va 1 Til t-up concrete (6") exteriorwallsh

/\\

o

~

( 31 4

kk

~

~

Filter Platform

\\h

/

General Support

$I y

CRITICAL AREAS N

Area A/ U quip.

l/

IUi fl0TE:

Interior partitions are steel s tuds wills gypsum board facing.

_ME ZZANiilE FIGURE 1.

GENERAL LAYOUT OF Ti!E NMDF BUILDING

a

l. ftAt4 Roof MCk IS li.ll. R09f tT5sH Iku 16-16 Decar WeIH 2" TO 3** TsetCM 20td8Le f t CON (MTE

2. Fest wrto 5tE Horrs oe4 FIGun[ 4 8

D E

F H

• to SPACE $ @ 2 0'- 0" ' = 2 0 0 '. 0 "

• 01'-80 "

I i

_;{

_.J 3

i i

i i

t

-S C l3.15 M

2r-4's5'ayd

• I '

8C:3.15 g gyp) __

N 7

(Tve.)

IO Wf EI

__$[El L ROOF vet K (Anor(

(SEEports)

O pecx 4

1 4-0 $

3 U

^ -

gb

[

sitEL STACK O

O tie Z<'AWlHI 3

o $-

e-eo vvf 21 (IT'""

G ms. [ ie

2. ~O3

/

SE'C 12 O

O

()

ersesif tsf t)

G A.16 i M_

im em 7-6" ~ro'-e# g( _* 6 "

Kit 2-I f

T T

i r

i T

T T

?

VA UL.T PnEcAST Wall. PANCl$

FIGURE 2.

ROOF FRAMIilG PLAN OF Tile NMDF BUILDING [ 3 ]

4 deck wnich is covered with a ligntweight concrete topoing.

A built-uo tar

. anc gravel roof comoletes the. roof system.

A typical section through an east or west exterior wall is shown in

- Figure 3.

The W27 x 114 wioe flange roof. girders are connected to the W8.x 31 columns with eight 7/8 in, diameter bolts.

Because of the type of connections, the' steel frames provide very little resistance to lateral loads.

Lateral lead resistance is provided by shear wall action.

The columns are supported on reinforced scread footings.

They are secured to the footings with two 3/4 in, diameter anchor bolts that are 1 ft 3 in. long..

The roof system is constructed with H. H. Robertson FKX 16-16 cellular steel roof deck.

The deck spans 20 ft. between the roof girders.

The roof

- deck sheets are 24 in wide.

The longitudinal seams are joined by a 1-1/2 in.

long weld.,at 20 in. - on center.

The ends of the she,ets are fastenec to the flanges of the roof beams with 3/4 in, diameter (1/2 in. diameter effective) fusion welds. A weld is placed in each valley of the deck, giving five welds in each 24 in. width.

At the walls running oarallel to the deck sheets (east-and west walls), 3/4.in. diameter fusion welcs are soaced at 4 ft on center.

The cellular deck is covered with a :enolite concrete topping which varies in depth from two to three inches.

(See Fig. 4 for typical roof diaphragm con-nection details).

The exterior walls of the building are constructed of precast reinforced concrete panels.

The panels are typically 6 in, thick and 18 ft 2 in, high.

Their widths are nominally 20 ft.

Openings in the exterior walls are shown in Figure 5.

The panels are reinforced with #5 bars at 18 in, olaced in center each way.

Details of a typical wall panel are shown in Figure 6.

Since the panels with openings are more heavily reinforced than the panels 6

4 SEC TIGURE 4 70R i

' DI A PHRA 6M Cor1 N LC T ION

.5 ~ s s

.\\

l r

l

.l I

\\

._/

\\

\\ -

,, ; /

27 W 114

/

s b

. Oi

- 8 C 13.75 d

%,,r.x,e i

4 7/g") BOLT 3 l

{

3d;[

o-I r-,/l p tro 9

l t A

-a 2-) W pigg;,

h'

&W Si ceLuMW (TTP.)

( A,_

'L ;

f4 /

9 k

2tlNT;2:!N G %

Jg pgyu 6

--8 4 7Ca(LJ $ '4'O s.$ D5"! *a0 k(

l3 J

Nf 37 WALLS BfMid 006. O A4 h-,.

H. '4 20ntL f G i2" o, c.6 ALL I a

?h"ELb j' I

C*LUMN CAP O

'3 [sb. v. way I

- 2 4' = W G "

b s'.o'(tyrgj m n46 n w.

t i

y 9

l CowSTAVC?ioni Jo4=T.

g,

t

+ ___ 9. c..a~- 0. 2 DETAIL A-A

'u q j

i q

1 l t I

!I

!.,, l, /*

Po rst FiLitD wi w i

c w m n v:nc.w u 2 4*f Mf5 1

-i "

(.s'Lond =/poutu c)l -

,i

^-

7 WALL SECTION i

FIGURE 3.

TYPICAL SECTION - EAST AND WEST WALLS 7

g

C.

S O

L

'o L

A W

>4

. T T

oA

,3 S

s eH c.

a A

E s

c r o

(

u5 f

i H

e

• o f

t k 0

_"2 ?)

WoA t

p1 D

u(

7 9

7 J

(

P s

3 N

w C.

8 s t e

C oE A

ne p

l P

g i

A

=

t o

t t

i d 's B

2 gD nr v A

t f

e*-

M j

a s e.

t T

ol r

w l

a C 4 S

a f

t S

A, f@

A t

E s

s a $.

L E

Ho E

c (

V L

G%

"3w

'* ?

T nt ar O

W ad n P

A t

t s

7

'g A

at t

.T a

ar L

(, vn sv t ooT T

s, u

I N

W(

J' a

t.

Wt O

.sp t

A A

i prjwP eu s

o I

l L

)

t 8c T

t i

C r

a H f O

f zcf

(

i g

A W a (i E

I

\\

E H S mrs i

T X

e5d4t 4

r N

C f

r W 8

s N

'g 4

s Ncl E

o e

b O

od N

h i

"o C

t iI EP 's N

t ew I

5d2s wSA 6l a

O H

C 2 0 t s G

t A

A R

M R

l G

i s

A P

n R

A l

I

e iP s

D A

I D

FOOR L

A

~8 S

C F,

4

~

I Vs

,t L

,e L

P 8

2 r

.n A

Y e

L

= c f

W T

1. '/

/

3,a e

h,\\

4 a

G

r. e 3

^

E s

l

(

3%L. wu i

c T

s t

i a" U

- A L*

w 2 7E t6 O

4 u, ev *4 E

S o

V s

en

[,

0 R

y U

. I' f

nE"s l

U i

fA G

n "f

I l

o F

+.A,.

^

t c

i n

T i

R

/*-

ON d

+

+AM

[

T

[

l 3b N '"

q A

O C'

_e1 I

~

O.

N d

f O

s I

<r I

X'b

}

4. r T

)

C

1. 2 E

6 4

NN m'

O C

s L t

+

p i

nu ~

Yo-L t

A a o

l e

G Ww Av t

s ar A

S n s

/s w

c*

s <

0.

o.

2 R

)

l r.

o

[

6d r e

i 5wm

.s a 3

P r

l

~-

i o

A 3

ic I

L D

oc

r a

t s

tr.

we

_$I, k-L~

tI

=

t i

erw, d

EV f

l' e

e

.e R

N'

]R({

k' 5

jP1~

8 o.

i Ig

.? E l[

~

E D

lJ I

~4

,s L

S L

I

(-

n A

U 3

~'q 8

W O

gl o

'I

't 1

O G

x 1

t 1

T C

0 l

0 D

R S

E fi SO s

P X

E a

)i.

5 S

L

'3 L

T%3I A

W f

R

)9 O

73 I

R l

E o

L I

1 L

L X

A L

E W

AW N

T I

S T

A S

G E

E s

N W

{

I n

N L

• 6 E

'8 L

P X

6 A

O C

W

't

~

t*

l n

6 i

i T

5

• i R

O E

4 N

R 5

S 6

UG

'o I

H~4 F

, E. W

.e r

4x

~

a si.

f-l 3

b

{*.,

W l'

f$

N

t

,;, s

,i 1

r

.s

')e r

)-

5 a

w

,S 4 r) m Drth L

w l

) U[

[

r_

_.de-

{

i e,

g v

g

(

g e

L y,

E N

"g g

A g.

g P

p L

u L

p A

=

IL W

g E

TER

~

C g

a N

5 O

C 4

[

y p.

T 3

T S

=2 g

\\

C A

s U

R e

gg O

4 E

=e

\\

H "

R 2V 5

P I

8 p

5 t

e 5

L M4 9

"6

\\

e oA 4 A a'

r I

. A G

A 5 8 rrv r

s C

1 4

.s is W8 )$r e P

t n

E t

tr P Y

K es e r. oOe l

s t

no ut e

f r

e

]'

s e T O r

]

A..st vA Te re oC H

l j

S n.p n T

~

6 o

c. s A. N g

t T

t ti t

F aao a

.t E

f.

E T

a cmCt

(

w n "f S

t i 1 p W

R i lA 5eoFb s

U y

8 t r e e

t G

s 2/

i@s6$AA t

w

~

M A

I 6

a F

j 6

' y.?

4 se

=

~

=

r y

f f

it

~

f

'/ _

~,3 s

b

]

3" k

3 -

l e

I 4

E 8

n 8

t y

5 w.jm "

1 L

' 7, k *.4

,e A-e 8

P 4' e d' s.

i l

L

, J.

Y

.l lwithout < openings, it is assumed that all panels have essentially the same' strength.--The panels are succorted onJall'three sides through. connections to tne: columns on sides'and the' channel at top which spans between columns.

At the cottom,. the > panels are' connected ito the floor slab with #4, dowels at' i

'24 in'.orlat 12 in. nn center'(Ref. Figure 3).

Along the. vertical sices, r

the panels are provided with.6 x'l/2 x 6'in. insert olates_(Ref. Figures 6 and.7).

These-plates are field welded to the columns using 3/16-x 6 in. long -

welds!

Similarly, five insert plates __are provided at the top of'the panels.

These are field welded to the channel.section'(Ref. Figure 3).

The roof deck is welded to the channel section with 3/4 in, diameter 2

4 fusion welds at 4 -ft on center, as mentioned previously (Ref. Figure 4).

Since the roof metal deck's corrugation runs carallel to the' channel, the roof deck.is not expected to provide any significant restraint _to lateral deformation 'of. the channel.

Hence, the channel section is assumed to act independently of roof diaphragm restraint.

Failure of the wall canels b related to the' strength of the channel member.

i The exterior walls act as a shear wall system to resist. lateral loads, as shown in Figure 7.

The roof ' deck, along with the zonolite concrete tooping acts as 'a diaphragm.

Because the roof deck is so securely welded to roof girders (more than 100 welds per girder), these connections should provide sufficient stiffness to the roof deck so that it can develop diachragm action.

A me::anine is located inside the building along the west wall over the J

general. support area (Ref. Figure 1).

Approximate dimensions of the. mezzanine are 20 ft x 40 ft.

The mezzanine structural system is essentially independent of the overall building system.

The filter platform, which is a part of the mezzanine, is an area of concern.

The me::anine floor system is supoorted on pipe columns and ties into the west wall framing system,

^

11

6.

,1 NM d ' VT

^^",5 Not 5w owa l

BUT Het C34T2IBUTC 73 R!sistene s,

200F CIAPMAAGt1 %. -

$0UTM SAEAR h

f W ALL (6"TWitr)

L

- HotTH S H LA R, knALL (7& " ry ggs)

)

2 0 t '- 9 V "

1 l

EAST - WEST RESISTING S STEM f-- EAST $ Mist wn* ( g" pa,gu,)

200F pia N eag g 1

/~ Wf.JT SWEAR WALL ((' Tw,c e)

NORTH - SOUTH RESISTING SYSTEM FIGURE 7 SHEAR WALL SYSTEM i

12

1

.i -

-e

-C.

Vault-i The' vault is-located on the' west side cf the building (Ref. Figs..

.'l an'd 2)..Its roof. is 10 ft 9 in..above tne finished floor.

The plan dimensions of the vault are 10 ft 9 in. x 20 ft.

A section through the-t vault is shown in ~ Figure.8.

i The vault walls and roof are g in, thick.

f They are reinforced with #4 bars each way at each face as shown in Figure 8.

The. vault door opens into the glove' box room-of the main building.

4 D.

Material Properties Properties of the building materials that are 'significant to wind damage assessment are listed in Table 1.

The table lists median values of material properties, and a range of low and-high values.

The variation of the material property values is assumed to be log-normal; the magnitudes of I

the ranges of strength are based on judgment.

The crimary source of material property values is (OAC Task I Recort [3].

In cases where material properties are not available in docu ats, judgments based on standard professional practice are made.

'For steel'and weld metals, the ultimate shear strength is taken as

{

1//T times the ultimate tensile strength of the material.

This relation-1 ship is based on the maximum distortion energy theory for ductile material [6).

I i

13

f 1

4

)

i i

DonnM m.

a cAsr m rtAcs 4

j.

10'

• 8 *

.- s 4 powgLs @ t 8 "o.c.

Dict t

r o-

_ll y_3-o' g

m,,M\\

@r a

h

%r ! ',

" = 4 w e

n

_L js4 $ t&*3cTw we WALL F4NEL h

(,.

u.rus

\\

I

• e

',-q" a

q

^

vauti l-CAST iH Ptut WALL

,a 84 CowtLJ Oo*Lono -

. " ~1

^~

I j-)

l e is o.c.

l

=+

l, Al ji ;.__

)

g i

~ -

i t '3-

~e u.

ii

.kC FIGURE B.

SECTION THROUGH VAULT 14

^

9 TABLE I i

IMIEftl AL Pit 0PERTIES Median Hange Source Material Property Value

' low liiglT_

Structural Steel ASTM A36 Yield Strength 44 ksi 40 ksi 48.5 ksi EDAC [3]

Hetal Deck Yield Strength 38 ksi 33 ksi 45 ksi EDAC [3]

Ultimate Tensile Strength 46 ksi 40 ksi 54 ksi EDAC and Judgment:

Reinforcing. Steel Yield Strength 48 ksi 44 ksi 53 ksi EDAC [3]

o Ultimate Tensile Strength-80 ksi 76 ksi 85 ksi EDAC [3]'

Welding E70 Shear Strength 47 ksi 40 ksi 56 ksi EUAC [3]

Structural Bolts & Anchor Bolts Ultimate Tensile Strength 68 ksi 64 ksi 73 ksi FDAC [3]

Shear Strength 48 ksi 42 ksi 56 ksi

'EDAC [3]

Structural Concrete e 28 days Compressive Strength 4.0 ksi 3.4 ksi 4.7 ksi EDAC 3]

Bolt Bearing Strength 3.4 ksi 2.9 ksi 4.0 ksi EDAC 3 3]

E 4

--e y

III.

STRUCTURAL RESPONSE AND DA" AGE CONSEQUE3CES The effect of wind loacs on a building and its components:is referred to herein as-structural resconse.

This section presents a generic discus-sion of structural response and damage consecuences.

In order to predict camage to glove boxes containing plutonium as well as to filters, the struc-tural response of the building and its ccmponents due to three' effects of-windstoms, namely wind, atmospheric pressure change (only in case of tor-nadoes), and windborne debris must. be. evaluated.

The wind and atmoscheric pressure enange effects may be combined under specific circumstances.

The

. general analytical aoproach for detemining a threshold value of windsoeed that will produce significant damage to a building or its comoonents is

-oresented in this section.

In addition, discussions concerning damage.from windborne debris is also presented.

The structural damage to the building and its components is then translated into subsequent damage to glove boxes and filters.

Secause the consequential damage to glove coxes and filters is random, rational judgments regarding glove box and filter damage are made.

Fire, as a consequence of windstom damage, does not appear to be a pertinent ha:ard.

In more than 50 major wincstom events investigated by the authors, not a single one produced a fire as a consequence of windstorn damage.

A.

Threshold Windspeeds to Produce Damage Threshold values of windspeed to produce damage to a building and its components are obtained by applying basic techniques of structural analysis.

These technioues are utilized by the authors to determine wind-speecs in tornadoes (7).

Damage, as used here, imolies the removal of cladding due to outward acting forces or the collapse of a member or the structure due to outward or inward acting forces.

4 i

16

- ~

3

~

Wind interacts with.a flat-roofed. building.and produces inward-acting external pressures.on the windward wall and cutward-acting external pressures On the sidewalls, tne leeward wall, and the roof.(Ref. Figure '9 ).

In addition, relatively-high outward acting external pressures are oroduced on localized areas at wall corners, roof corners and eaves (Ref. Figure 9 ).

In cases where tnere are coenings in the walls or the' roof.of a building, internal pressures are also oreduced.

These internal pressures may comoine with external pressures to produce 'a more ' severe. loading condition on a builcing component.

Since wind can come.from any direction, the failure.

f I

mode of a building ccmponent should be evaluated for the inward acting ores-sure as well as the outward acting pressures.

Knowing the strengths'of the materials and the type of structural system, principles of mechanics are apolied to determine structural response and the wind pressure to produce a postulated failure.

The structural response of a building component is made up of a static ano a dynamic part.

For low-rise buildings and relatively stiff components, the contribution of.the dynamic part of the response can be neglected.

The fundamental frequencies of low-rise buildings or their components such as masonry walls or metal roof decks have fundamental frecuencies greater than 3 H:, while most of-the free field wind gust spectrum energy is in the frequency range that is less than 0.5 Hz

[3,9].

The discarity between fundamental frecuencies of building components and gust frequencies of the wind suggests that the dynamic.oart of the t

response is negligible for ordinary structures.

Once the wind pressure recuired to produce the postulated failure mode is obtained, the corresponding windsoeed V is calculated using aporocriate equations that relate windsoeed to aerodynamic pressure.

The general form j

of the equation is p = 0.00256V C (T) 17

t i

, y-

+

/

@@ed Q* A-

/

4 N

EXTEEN AL WINO FREssugEs

/b w'!S LI As? ws:-q

<f 0\\.

CF BUICING c5 h

e 4

9-4 t

f s-

!!!k G

e P: ;q g

,.fsi: f ij:I f::b g%b c

'!!!.?

ihi f f::::::::@l.ll.gh

,N N

:+:.

e3 *

~

I l

LOCAL!!ED WINO FRE53URE FIGURE 9.

WIND PRESSURES ON A BUIi.0 LNG 1

18

l where.

1e is the wind cressure in psf

'V is the windsceed in moh C is a' shape factor or pressure coefficient Equation =(1)Lis the stagnation pressure lmultiolied by an accroariate cressure l

. coefficient, pressure coefficients are.obtained primarily from wind tunnel-tests of model structures.

Coefficients from the American National Standards-Institute Standard A58.1-1972 [10] are used in this study.

The ANSI A58.1 Standard [10] defines three types of pressure coefficients:

(1)

External pressure coefficient, C.

3 (2)

Internal pressure coefficient, C pg (3) Net pressure coefficient,'C f

External' pressure coefficients are applicable for' external wind pressures.

-acting on enclosed buildings.

The equation for externally acting wind pressure is:

2 p = 0.00256V (C )

(2) p If the building has windows, doors or other openings that allow the wind to get inside the building, internal pressures act on the walls and roof in addition to tne ' external pressures.

The equation for combined external and internal wind pressure acting on a building component is:

2 p = 0.00256V (C -Cp4)

(3) p The sign of the. internal pressure coefficient C is a function of wind direc-p4 tion and opening locations in a given building.

Het pressure coefficients are used for structures such as chimneys or towers.

The wind pressure is the net horizontal pressure and is obtained from the equation:

o p = 0.00256V'(C ).

(4) f 19 j

, 3.

'With' knowledge'of the wind oressure.-c calculated from structural mechanics l

'procecures, and with aporopriate cressure' coefficients determined from the.

literature, the thr'eshold windsoeed V can;be calculated utili:ing the above equations.

L The threshold windspeecs-that produce damage as determined using' the

'above ecua. ions include wine gusts.

The calculated wincsoeeds are ecuivaient-to." gust speed" given in. Column 3, Table 14 or " tornado windspeed" given in.

7 Column C, Table -of1 Reference.2.

Whether the threshold windspeeds are straight line winds or tornadic winds depends upon the probability of occur--

.rence of that intensity wind.

B.

Atmospheric Pressure Change (APC)

If a tornado is the windstorm hazard, then the effect'of atmospheric change (APC) may also contribute to the' damage.

A region of reduced pressure I

exists near the core of a tornado.. As the tornado passes over a building, the pressure inside the building becomes greater than that on the outside, thus producing a differential pressure across the walls and the roof of the building.

Table II gives the APC values associated with tornadic windspeeds

~

for.different probabilities of occurrence at the NMDF site.

The probabilities

-of occurrence of tornadic windspeeds at the NMDF site are obtained from F

Reference 2.

The APC values are calculated using the cyclostrophic equation

[11].

If a building is sealed, it will experience the effect of APC as the t nado passes over it.

However, most industrial buildings are not totally I

sealed (air tight).

If there are enough openings in.the walls or the roof to allow air inside the building to escape, the differential pressure will be equali:ed. The venting areas necessary per cubic feet of building volume are given in Table II for values of APC.

i 20 I

(

[

I i'

l TABLE II Tornadic Windspeeds, Atmospheric -Pressure Change and Requirements for Venting a

b c

d

. Probabilities of Straight Line Torradic Atmospheric Pressure Venting Arca Occurrence per Year Windspeeds, mph Windspeed, mph changenpsf sq. ft/cu. ft 10 43

-I 10 54

-2 10 65 10~

74 10' 83

-5 S

10 90 1

-6 10 123-49 0.07 x 10-

-7 10 178 104 0.22 x-10~

-8

-3 10 230" 173 0.47 x 10

" Includes gusts; Column B of Table 14 from Reference [2]

bColtann C of Table 14 from Reference [2]

CDetermined using cyclostrophic equation Reference [11]

Escaping air is limited to 25 mph

" Extrapolated from Figure 6 of Reference [2]

~

,1.

i 1

C.

,Comoined Wind and Atmospheric pressure Change j

For ouildings wnich are sealed, the comoined effects of wind and atmospheric pressure change may' produce the most critical loading condition on the building components.

The highest load could be'due to outward acting pressure caused by.the maximum windspeed in a tornado and the associated-atmospheric pressure change at the location of the maximum windspeed.

It is possible to express the value 'of-ti e atmospheric pressure change in terms of maximum windspeed if certain assumptions are permitted.

This information f

is given below.

The maximum windspeed, V, in a ternado is a combination of the tangential, V and translational V windspeeds:

g tr 7

V=V

+V (5) tr Fujita [2] assumes that translational windspeed is 20", of the maximum wind-speed, hence V,= 0.8V (6) j The cyclostrophic equation suggests that atmospheric pressure change (APC) at the point of maximum windspeed in a tornado is:

2 APC = 0.5 o V (7) where ; is mass density of air. The total outward acting pressure at the point of maximum windspetd in a tornado due to combined effect of wind and APC on a building component would be 2

2 p = 0.00256V C,* 0.5 a V (g)

Substituting the value for p and utilizing V = 0.8V, tne total outward t

acting pressure will be p=V2 (0.00256C + 0.00164)

(9) p The value of C would depend on the type of comoonent such as side wall, p

roof, roof corner, etc.

For example, the pressure coefficient for the 22

w roof is C = 0 7, hence, the' uplift pressure would be p

2 p = 0.00343V (10)

A tnreshold value of a tornadic windspeed can be determined that would fail a building component.

Two requirements are essential to consider combined-effects of wind and ApC; they are (1) the building is sealed, and (2) the threshold windspeed is in tornadic windspeed range as specified in Table II.

D.

Windborne Debris Windstorms tend to pick up and transport various types of loose debris.

The kinds of debris range in si::e from roof gravel to automobiles.

Most of the' debris consists of objects such as sheet metal, timber _ from damaged houses or other lightweight objects.

In a very intense tornado (windspeeds greater than 200 mpn), debris can be propelled to high velocities to become damaging missiles.

Velocities attained by typical pieces'of debris whien can cause damage are shown_in Table III.

Missiles which impact exterior walls may not pose danger to glove box integrity of to HEpA filters 'if much of the missile energy is absorbed by the wall. The walls of the NMDF are reinforced precast concrete panels.

Hence, windborne debris damage is not likely to be critical at this facility.

E.

Damage Consequences The building damage and damage consequences discussions oresented here are generic in nature.

Structural response, component damage and missile impact translate into damage to glove boxes, filters, or other containments of plutonium.

The damage to glove boxes and the subsequent plutonium releast potential are defined as /cIlous:

Crushino of Glove Box:

If a heavy object falls on the glove box, structural members of the box may collapse resulting in the glove box being crushed.

This event could occur if a load-bearing wall or building frame should collapse, thus allowing the roof structure to fall downward.

In this case, the roof integrity of the glove box would be violated.

The material inside the glove box would be exposed to the atmosphere.

23

qk TABLE III Windstorm Generated Missile Velocities [11]

Missile Vel cities, mph Impact Weight Area Windspeed, mph (V) 2 Missile

_(lbs).

(ft )__

F00 150 200 250 300 Timber Plank 28 0.04.

70 98 124 160 2 li x 4 in. x 15 ft Timber Plank 115 0.29 60 90 100 125 4 in. x 12 in. x 12 ft Standard Steel Pipe 76 0.067 65 65 110 3 in. dia x 15 f t Utility Pole 1490 0.99 80 100 13.5 in. dia x 35 ft Automobile 4000 20 25 45 Interpolation of windspeed is reasonable and consistent with the current state-of-the-knowledge on missile generation.

i m.

t Perforation of the Glove Box:

Pieces of-timoer, concrete clocks, loose pieces of pioe'or equicment Could strike a glove box, causing an opening.in:the. glove box window.

Plutonium stored in canisters is not likely to be released -

in this case, but. loose material' in powder form could-possibly escape the confines of:the. glove box.

Failure

-of an exterior wall. could allow the wind to circulate throughout the build,ing, causing loose oojects to be thrown against the glove boxes..Windborne debris could cause missile impact on the glove.bcx and may cause per-foration-of.the glove box.

Tear in Glove: 'The' gloves are the weakest elements with respect to tne glove box integrity.

Flying or moving debris could strike and. tear a glove.

Soma of the material in powder form could tHe pulled. or blown frem the glove box should the ventilation system be altered by'the effects.of

.the wind. Containerized material or material in: pellet form is not likely to escape.

P These three definitions of. glove box damage are correlated with extent of damage tc the building and it's components, and are shown in Table IV.

Damage scenarios in Section V present actual damage _ consequences.

F I

f i

l l

l 25

TABLE IV Danage Consequences Building or Component Damage Glove Box or Filter Damage-Remarks-1.

Mechanical equipment on roof Filters.in ventilation equipment topples but does not penetrate crushed roof 2.

Collapse of mechanical-equipment Glove box crushed under equiguest; Only the box beneath.the through roof filters crushed

. equipment damaged i

3.

Uplif t of small portion of roof perforation of a few glove boxes; Items may fall.through-ccener or eave a few filters crushed

' roof opening 4.

Uplif t of entire roof deck perforation of a few glove boxes; Items may fall through a few filters crushed opening g

4 5.

Failure of doors or windows lear in several' gloves; glove box Wind and win'dborne debris close to opening may be perforated, cause damage a few filters outside of boxes crushed 6.

Wall corner failure perforation of. glove boxes located near the wall corner 7.

Loss of wall siding.

filters c'rnshed; perforation of a Windborne debris can enter few glove boxes the building 8.

Outward collapse of non-load-filters crushed; perforation of a Windborne debris can enter bearing wall few glove boxes the. building 9.

Inward collapse of non-load-Glove boxes in the vicinity of walls bearing masonry wall crushed; filters crushed 10.

Collapse of load-bearing wall Glove boxes crushed; filters crushed-Roof collapses downward '

11.

Lateral collapse of building Glove-boxes crushed; filters crushed i

h.1

. fe a

' IV.

Threshold Windspeecs and Failure Modes i

iThe threshold values of windspeed that cause. failure of building com-ponents have-been calculated. ' Detailed calculations are' contained in V'olume-

II of. this" report [4].. The failure mode that occurs at the lowest windspeed is the~ critical failure mode for'a given building component.

Critical failure modes of components and their associated windspeeds are summarized in this section.

These data are then used to formulate camage scenarios'in Section V for selected windspeeds and associated probabilities'of occurrence.

Calculated threshold windspeeds are considered gust speeds or tornadic windspeeds.which. include gusts.

In. addition, the thresho".d windspeeds to' produce damage are also considered.to be nominal windspeeds since they are based on median strengths of materials.

Windspeed ranges are provided for-each calculated threshold windspeed 'to reflect variation in material proper-

. ties.

In cases where the material properties are not the governing. failure criteria, windspeed ranges.are. based on subjective engineering judgments.'

All windspeed ranges are assumed to have log-normal distributions.

)

Critical failure modes, threshold windspeed values for wind damage, atmospheric pressure change effects, and missile impact damage are described below.

A.

Wind Damage to NMDF i

The framing and construction details of the NMDF. building are oescribed in Section II.

Calculated threshold windspeeds to fail. roof components, wall'-

components, and the structural frame are shown in Table V and discussed below.

The exterior doors 'in the glove box room are protected from direct exposure to wind by enclosures.

The glove box room is relatively sealed, hence its components could experience atmospheric pressure change (APC) effects,in case

,f al tornado.- The doors to the exterior of the glove box room could fail at 1

27

\\

i l

TABLE V Threshold failure Windspeeds for NMDF Naninal Threshold Windspeed Range, mph Building Component Windspeed, mph

. Low liigh

. failure Mode.

Doors:

Glove Box Room 110 105 IIS Outward failure due to APC Others 125 120 131 Outward failure due to wind Roof:

Roof deck at north or 274 255 298 Yielding of deck, but no removal south eaves Roof deck at cast or 190 177 207-Yielding of deck, but no renoval west caves Roof deck at corners 173 161 188 Yielding of deck, but no removal Secondary roof beams over 138 126 151 Lateral-torsional buckling and ro exhaust filter room subsequent collapse downward Main roof girders 162 148 177 Lateral-torsional-buckling and subsequent collapse downward Walls:

Corners 199 190 209 Outward collapse Intermediate 230 224 247 inward or outward collapse Other Areas:

Lateral collapse of frame

>300 Collapse of shear walls Vault

>300 Outward collapse of walls Stack

>300 Buckling of support angles

=

thresnold.windsoeed.of.110 mph.

Failure of'the doors will alleviate APC.

j

~

-l effects on other components of-the glove box room.

The exterior doors in j

others parts.of the building do not experience APC' effects because of the'

~

' pre'sence of the grill which provides sufficient venting.

These doors-to.the-exterior could fail inward or outward at the threshold windspeed of 125.moh.

The metal roof deck in the corner areas of the building would experience sufficient uplift to cause yielding of the material at the threshold windspeed i

of 173 mph;Lhowever, the substantial welding around the perimeter would pre-vent removal ~of:the roof deck.

The roof deck in other areas of the building would also yield at threshold windspeeds of 190 mph for east or west eaves, and.

274 mph for north or south eaves, but removal of the deck is not expected.-

~

The uplift forces on the roof deck are transferred to secondary roof beams at north and south ends of the building, and to main roof girders in other parts of the building.

The uplift forces cause the bottom flange of the beams and girders to experience compressive stresses.

Lack of adequate l'ateral supoort for the bottom flanges of the beams and girders induces lateral-torsional buckling failure mode.

The beams and girders would twist as the bottom flange would deflect laterally.

When the aplift forces reduce, the downward roof loads are again transferred to the roof beams and girders.

The twisted beams and girders tend to resist gravity load by bending about their weak axis, rather than the strong axis, which is normally the case.

Plastic hinges would develop with respect to weak axis bending and mechanisms would form, which would result in downward collapse of the beams and girders.

The secondary wide flange beams could collapse at a threshold windspeed of 138 mph.

The main roof girders are likely to collapse at a threshold windspeed of 162 mph.

Collapse of the main girders w Jid collapse the entire roof system.

The precast concrete wall panels located at the corners of the building could collapse at a threshold windspeed of 199 mph. The failure mode is 4

29 n

r 4

./*

yielding of the reinforced. concrete panels.subsecuent.to-the failure'of:the-top. support channel.

The corner panels'would collapse outward.

The' remaining _

-wall panels could collapse -inward or outward at the threshold. windspeed of 230 mph.-

The shear. walls'are able to resist wind pressures associated with wind '

speeds 'of more than 300 moh.'

Since the roof and' wall-components fail below this windspeed, lateral collapse'of the building failure mode.is not... feasible.

Calculations-show that no structural damage of significance occurs to vault structure for.windspeeds up to 300 mph.

The rest of the building could be lying in rubble, but the vault-is 'likely to remain intact.

The 42.ft high exhaust stack located near the south.end of the building is substantial.

Calculations indicated that it is able to resist windspeeds

. of 300 mph.

3.

Atmospheric Pressure Change Effects Doors in the glove box room could fail at 110 mph due to APC effects because this room is tightly sealed.

The exnaust filter room, however, has sufficient venting to negate APC effects.

The vault is capable of with-standing the APC pressures.

Roof failure in tne glove box room does not

=

depend on APC forces, because the doors will fail at lower windspeeds than threshold windspeed for roof failure.

The APC pressure is relieved due to venting through door openings.

C.

Damage from Windborne Debris Windborne debris is of secondary concern to the NMDF building.

The energy that a missile possesses as it approaches this building will be dissipated upon impact with the exterior concrete walls.

If the exterior wall has failed prior to the occurrence of a missile (199 mph), ecuipment j

30

s C

4 9

inside of.the building. is likely to alrehdy be crushed by the walls.

There-fore, a missile entering the facility subsequent'to a wall collapse wi?.1 cause little additional damage.

The walls and' roof of the vault can with-

. stand missile impact postulated in this sttdy.

0..

Summary of Failure Modes Calculations of threshold values of windspeed that cause damage suggest the following sequence of failure modes:

110. mph' The exterior door in glove box room fails due to APC in a tornado

'125 mph Other doors in the building fail due to wind alone.

Failure of these doors can subsequently result in internal pressure development because of air flowing in or out of building, depending on wind direction.-

138 mph The secondary wide flange beams that span-between the roof girders over the exhaust filter room collapse downward.

As the beams fall downward, they could strike and damage the filter enclosures.

162 mph' The main roof girders collapse downward.

The roof deck attached to the girders will also collapse downward.

Once the roof has collapsed, the steel columns and precast concrete wall panels also collapse; they could collapse inward or outward.

Essen-tially, the major part of the building would be destroyed; except vault and exhaust stack which are capable of resisting windspeeds of more than 300 mph.

31

~5 4 --

e V..

DAMAGE SCENARICS t

Damage scenarios for selected probabilities of occurrence of windsoeed are formulated from the calculated threshold windspeeds presented in Section-

~

IV.

The damage scenarios.are used for subsequent identification of source terms.

Four damage scenarios for'. selected windspeed values are presented to'

' formulate.a trend of increasing damage with reduced probability of occurrence.

Fujita [2] developed tne relationshipL between windspeed values and their

- probability of occurrence at the Atomics International Nuclear Material.

.- i Development Facility.

The values used here and presented in Table IIt are taken from curves B and C of Figure 6 in Reference [2]. 'The windspeed values-are gust speeds in the case of straight line winds and maximum tornadic windspeeds in the case of tornadoes.

Damage causing threshold windspeeds are either. gust speeds or maximum tornadic windspeeds.

Since damage is based on median material strengths,.the threshold windspeeds are termed nominal windspeed.

Variation in material croperties, or subjective engineer-ing judgment, based-on the type of damage, establishes the windspeed range for each damage scenario. These windspeed ranges may be used to provide error bands on potential damage to the facility.

A.

Damage Scenario for Ncminal Windspeed of 110 mph

-6 probability of Occurrence:

2 x 10 Windsoeed Rance:

105 mph to 115 mph, based on failure of doors Glove Box Room: One of the two doors in the east or west exterior wall could fail due to APC.

Subsequent to failure, wind could circulate through the room, but the demage to glove boxes and filters would be minimal.

The enclosures around the doors will protect the inside from

.windborne debris.

32

4

'.)

Exhaust Filter Room and Air Lock:

No damage Vault:

No damage B.-

Damage Scenario for Nominal Windspeed.of 130 mph orobability of Occurrence:

7 x'10

~

Windsoeed Rance:

120 mph to 141 mph, based on failure of doors Glove Box Room:

No additional damage Exhaust Filter Room and Air Lock:

Exterior doors could fail due to wind forces.

Wind circulating through building could cause interior

. partitions in the exhaust filter room to collapse.

Debris.from the collapsed wall could strike the exhaust filter enclosures.

Vault:

No Damage C.

Damage Scenario from Nominal Winospeeo of 150 mph probability of Occurrence:

3 x 10~

Windsoeed Rance:

130 mph to 173 mph, based on roof failure Glove Box Room:

Interior partitions at north or. south end of' glove cox com could collapse from wind circulating through the building.

One or two glove coxes near the partition could be crushed by the collapsing wall.

The roof and the exterior walls remain intact.

Exhaust Filter Room and Air Lock:

The secondary wide flange beams fail due to lateral-torsional buckling.

As the beams twist, the dead load is transferred to bending about their weak axis.

A mechanism forms and the beams collapse, possibly striking the filter enclosures.

Twisting of the beam will likely create a tear in the roof deck at points of welds.

A small portion of the roof deck may tear and collapse with the beams.

Vault:

No damage D.

Damage Scenario for Nominal Windspeed of 170 mph Probability of Occurrence:

1 x 10~7 Windsoeed Rance:

150 mph to 193 mph, based on roof failure 33

r

.. ~ '

r Glove Box Room:

Main girders _of.the roof-undergollateral-torsional-L buckling with subsequent ' development of -a.mechanis: wnich would cause ti Leollapse of the. roof.

In addition to.the roof the concrete wall panels will also collapse. All ' glove boxes'are likely tu be crushed under the weight of the-roof -and wall panels.

Collapse of the roof could cause-tearstin the roof deck, thus providing openings through which scurce material could be transported by the winds.

Exhaust Filter Room and-Air Lock:

Collapse of the roof in a manner similar toi hat at the glove box room could crush the. filter enclosures, f

t i

- Any material available for transport could be carried through openings-in the roof oeck.

Vault:

No damage' 4

34

O p

_,o a

VI.

REFERENCES 1.

Ayer, J. A., and W. Burkhardt, " Analysis of the Effects of Abnormal Natural Phenomena on Existing Plutonium Fabrication Plants," United

.l States Nuclear Regulatory Corrmission, -Wasnington, DC,1976.

2.

Fujita, T.

T., " Review of Severe Weather Meteorology at Rockwell International, Chatswor:h, California," Technical Report prepared for Argonne National Laboratory, Argonne, IL,1977.

3.

EDAC, " Structural Condition Documentation and Structural Capacity Evaluation of the Atomic International Nuclear Material.Develocment-Facility at Santa Susana, California," Task I--Structural Condition, Tecnnical Report prepared for Lawrence Livermore Laboratory oy Engi-neering Decision Analysis Company, Inc., Irvine, CA,1978.

4 Mehta, K.

C., J. R. Mcdonald, and F. Alikhanlou, " Response of Structures to Extreme Wind Hazard at the Atomics International Nuclear Materials Development Facility," Vol. II, Technical Report prepared for Argonne National Laboratory by Institute for. Disaster Research, Texas Tech University, Lubbock, TX.

5.

Mishima, J., " Identification of Features Within (AI-NMDF) Plutonium

. Fabrication Facilities Whose Failures May Have a Significant Effect on tne Source Team," Working Paper on Increment of Analysis orepared for the U.S. Nuclear Regulatory Commission by the Battelle Pacific Northwest Laboratories, Richland, WA.

6.

Higdon, Ohlsen, Stiles, Weese and Riley, " Mechanics of Materials," Third Edition, published by John Wiley and Sons, Inc., Somerset, NJ,1976.

7.

Mehta, K.

C., J. E. Minor, and J. R. Mcdonald, "Windsoeed Analyses of April 3-4, 1974 Tornadoes," Journal of the Structural Division, ASCE, Vol. 102, No. ST9, September, 1976, pp. 1709-1724.

8.

Kim, Soo-II, " Uplift Wind Loads on Flat Roof Area," a

' D. Dissertation, Department of Civil Engineering, Texas Tech University, uMk, TX, August 1977.

9.

Morris, Nicholas. F., " Wind Effects on Air-Supported Structures," Preorint 2860, ASCE Spring Convention and Exhibit, New York, NY, April 25-29, 1977.

10.

American National Standards Institute, " Building Code Recuirements for Minimum Design Loads in Buildings and Other Structures, ANSI A58.1-1972, 60 pp.

11.

" Design Guidelines for Wind Resistant Structures at the Argonne National Laboratory Site," Institute for Disaster Research and Department of Civil Engineering, Texas Tech University, Lubbock, TX, 1975, 68 pp.

l 35

..