III

ý'- 3
-/--- WASH- 1255

.1

K
~.

MASTER
A STUDY OF VERTICAL AND HORIZONTAL EARTHQUAKE SPECTRA

44

ill
ft.4 ~

I j

44;

April 1973

Directorate of Licensing
U.S. Atomic Energy Commissium
Washington ,D.C.

I kII
 WASH-1255
.~~~~t.*~~~ Ir4* Itt~ V;.!
Cat(.b n tic- ii
"h t4t~q st 1rt. 'I . nll

A STUDY OF

VERTICAL .AND HORIZONTAL EARTHQUAKE SPECTRA

prelpared by

NATHAN M. NEWMARK CONSULTING ENGEEHING SERVICES

for the

DIRECTORATE OF LICENSING
UNITED STATES ATOMIC ENERGY COMMISSION

UNDER CONTRACT NUMBER AT(49-5) -2667

April 1973

t. ~fl 4. Prn !1 r
 TABLE OF CONTENTS

I. INTRODUCTION.

L.I Object and . . . . . . .

1.2 Earthquake Records Studed . 2

2. PROCESSING OF DATA ............................ .......

2.1 Time-history ............. ........................ .

2.2 Response Spectra ... ............ .......... '..........
2.3 Amplification Factors_......................................
2.41 Statistical Analysis ......

3. MRESENTATION AND 1iS:uUSxI Oi QF P[ESULTS.

3.1 Ground Motion ....... . .............

3.2 Response Aripli cation ........................

I4. DESIGN RECOME1JDA lIONS AND CONCLUSICS...

4 .1 Genera l ................................ lot!

14.2 Ground Motions .............
It. 3 Amplificatlon of Ground mot.tIon ....................... .- ?
S4.4 Design Spectra ...............................................
4.5 Combined Effect5 ........ ................
S4.6 Reconmmendations for rutvre Studies .........................

REFERENCES ............

APPENDIX

A. SEGMENTAL BASELINE ADJUSTMENT .................

31. METHODS OF NGO.KALIZING RESP3NSE SPECT•A ..........-............... 14
 I

I IN OWMtOVCI ION

I.) Objecct and cp

The objec of t1he study. a, a.

ou invd in the, wirk dir ,-tivI" dated

is August1 197.1 f ron' the Div !' on oi Rcoto.~cr S tandards , U.S. Atcvnfic Cnc rgy

Cc'visnswas~ Ce,.,ential Il. f

, ,n1rC.1y
11ol as fol low-::

(a)0e ve Io j ve r t i c a a nd ho r i :on t aI f t%o c o~ipooen t!'.) resPons v,

tspec Ira fcoi a ser ie,ý of 14 %tuonj motion

*0t irthtsa.kc, rCcor~ds , i ociudi n,,

foj S Yt
frnanrdo rood, over thc. raj o( (rc4uenV1 int: in~terest, (or

0 !i, ý2 , and 10( picr~cn t. o t i. i t i ýa I dwipi nc. The records which were

iw- I sdt-d ab a pairt. -)f tht- study -.,re dt-scr ibnd in Section I .2.of this report.

cl
(b) Cornpire the virrt ir and horiiiantal retiulnses obtained from

the,-' tsprrctra.

(c) Ortermine the shapte of itI'c spc~tra in the high frequency

ronc Iron 6 t.o 50 he rtz~, -ith ': a crnpha* is on t i tc-dependent

parmntrwt . ~fet

(d) ýuqqcet procedurr!ý for comtining simic stresses dnrived

frryn hvt i.ontaI anid vertical rp se

(r) Sw'~mar lr~ the above-tioted studies In the ligh

194f

rcfrenZcr-datp~rnh for dr!slign.

1411 of the objtcrive~t noted have been accompl ished with two

f-.1nor. q7-ctptions, nmrseI' that after undertaking a comprehensivic study

of the. alailable cround motion dnta, It was decided that the ground motion

data wiere 9cneraIly valid only in the frequvency ralnge of- 0.05 to 30 hertz

an'd accordlngly the response ,.pectra were plotted only for this range.-

#5e result of further studies on the site Properties, and based

Secnd

on inform-ation that we had available from Previous studies, we felt that.

it wsould be difficult to drr aboat~si~m
sitt eaon.

wi nce only 4 p: ri fvs strong~noo ~reod cn roo munaala~ýA

of the availabl records atrc.=n ksi . Ncc~ees som aber'vtion.t

pe~rtinent to Wetd p:r rncc "ire Wtaisre ani a part of ou stua&5 aind

crin ~en
t it, :- ,re %,r to ! . * c* tw

studies~were mad inc I di recrds, corroi2ndinr to peak grin acee

re!por ted fully. The 11012~ re-pni ine trj va~t reI;e heci Ar

bal~ed 'in the "boundWd dat i,

m"ethod-; used in procesin th at n Seto 2; 121 a~ p~e tor! nA"

Sect ion 3 ; and (31 Ast~ 3r On;lsin in %e ctin 4.

The procedure~ empla(ved fo adasin lo ~* P -ýTl~

is pros'rtrd a% Appendix A, Th mrho rdfrorati en

1.2 Earthquaike Reri(rs Trdr'

The %tudie% we-re carr~r uu !r Me Wx~ .ol-n Vn

San Fernanido &seismic hwent

Pacoima~
Da,297. ~~~~~r 60O Rcodt 4%
 3

Other sv;.:mic event5

El Cerntro, 5-18-40, 2037 PSi (Riecord IA 1)

El Centro, 2-9-S6, 0633 PST (Record It, It)

[I Centro, 4,-8-68, 1830 PSI (Record IA 19)

Hollywood Storaqe Oasementp 7-21-52, 01453 POT (Record IA 6)

Hollywood Storage PE Lot, 7-21-52ý 01453 PDT (Record IA 7)

San Fraicisco Golden Gate Perk, 3-22-57, 111" PST (Record IA 15)

fe"rndale, 1O-7-5l, 2011 PST (Record 1A 2)

F#rndale, 1ý-21-54, Il$6 PST (Reccord IA 9)

Cureka, 12-21-5b, 1156 PSi (Pecurd IA 8)

tiollistev, 4-13-6), 2323 PSI' (Record IA 18)

The accelerogram records used in this study were digitized by

the Earthquake Engine.erlng Retenarch Laboratory of the California Institute

of Technology; the CIT designation of the earthquake record Is Indicated

In parentheses at the end of each 1Isting. The records are av,.Ilable

through the Orartiment of Convwrce or the California JnLtitute of

Technology.
 4

2. PROCESSING OF DATA

2.1 Time,.hi story

Since the dig9t;zed record from the instruments are in terms of

acceleration time-histories, the corresponding velocity anrd displacernt

are obtained by integration. Adjustments in the records are nsrm.a.ly

required as illustrated by the following exanple. The acceleration,

velocity, and displacement timc-histories for a typical record, San

Fernando earthquake, N69W coiponent of Ca-staic, 2-9-71, 0600, PST are shown

in Figs. 2.1-2.3. Since the motion during on earth ,5ka is a to and fro

type motion, the velocity time-history, Fig. 2.2, is obviousty Incorrect,

and the corresponeing displacement tine-history, Fig. 2.3, also is in crror.

The figure indicates a displacement of 130 In. at 31 seconds. Thfi5

displacement would have Increased further had the •.o.putaticiN .•t•tnui.d.

The errors In earthquake racordc may arise frr'• •*y nt'.•ber of

sources such as (a) the, instrutment errors, in,.ludinq cffcects a.ss iat.d

with mounting and instrument housing; and (b) the proce5';,.Yq of tfic ret.ord

where the initial conditions (sc,.e r'otion is re.Qt;red to tr;-:k.r the

mechanism) and the zero acceleration line (baseline) are not kn.n. The

errors In the velocity and displacement tim'e-histories arising frco-

integration of the accelerogram are largely associated with the latter

category. In order to minimize the record processin_ errors, the tr7t0al

conditions are taken as zero, and a baseline correctik is a.p;.ed t.c VSe

accelerogram record. (1) Armong various baseline adjust-e.nt proc-edures,

one which minimizes the square of the error in the velocity is nost

coarionly used. This procedure assurves a polynomial, usually a seco-nd

degree, for the correct acceleration baseline, In some instances the

record processin•r.•.,rors are further reduced by using different polynomi a I

for accelecra tion baseline In different port ons of the record. Mhis

procedure is explained in detail in Appendix A.

Each component of the 14 earthquakes considered in this study

wan adjusted by the procedure given in Appendix A. A second degree

polynomial was used as the acceleration baseline in ali cases, with

"hesgmental" adjustments used occasionally as deMcrlbed in Appendix A.

The adjusted .elocity a•rd displaucoen~nt time-histories were computed by

integratinq the adj us ted acce leration t ime-history and they were plotted

on A Calconi)p Plotter. An examination of the accc lerogram records used In

this study indicated that the strong motion portion of the record Is

included within the first 20 seconds of the record. For consistency the

computntion and plotting of time-histories were carried out for the first

30 seconds Oa each record,

An a general illustration of the effects of adjustnant., the

unadjusted and the odjusted acceleration, velocity, and displacement ground

motion time-hi.tor i.e', for a typical record of the San Fernando erthquake,

N69W cotanent of Costaic, 1.9-71, 0600 PST are shoawn in Figs, 2.1-2.6.

2.2 erpon Spectr

After the dAjusted time-history was obtained for each rccord,

the time-history was used as a base motion Input to a single-degree-of-

freedom damped oscillator and the response of the oscillator as a function

of the frequency of tht oscillator was computed. In this way a response

spectrum is generated. The computation of response spectra Is well

documented in the literature(2,3) and Is not repeated here. Io computing

 6

the response spectra, the adjusted time-history was used because the

spectral ordinates computed from the unadjusted record. may be *n error

for low frequencies. The response sýpectra for 0,5 percent of critical

damping for both the adjusted and unadjusted records of the San. Fernando

earthquake, N69W component of Castaic, 2-9-71, the example used in Section

2.1, are shown in Fig. 2.7. It is seen that adjustingv thd record affects

the spectral ordinates for low frequencies. Although as a Serera! rule

the adjustment has little effect for frequencies of interest for nuclear

reactors, in order to provide a better basi5 for comparisons reported later

herein, the computations of response spectra were carrivd out usting3 the

adjusted time histories,

The response computations were carried out for .38 trequercies

having a range of 0.05 to 30 hertz. In computation of tIe response spectra,

the interval In the frequency range over Ah;ch the cotut3at~ons .are carried

out influences the shape of the spectra. Genorally this urlln.e It not

tafrin,., A c a rsIvr.
large if small intervals in frequency O.re used in co

between two response spectra. for El Centro, Caillforniat S-20-1f944, N&

conponent, one computed by the California Institute of Techrolog~y q usi ng

91 frequencies and the other computed by the authors -;ing 3f .rce-.n-ies ,

is shown in Fig. 2.8. In qIeneral the aqre.et,'nt between th:e 1" s ctra ;s

q.oite c;ood. The difference between the tw ,pec¢.a for fre~qu-cie"s hel

0.3 hertz Is due to differt.nt Procedures used for baselnee adju-st-nt of

the accelerogram. Another comparison between the two soectra for the sato

a coarse frequency inter-.al (38 frequenc~es), and a fre freqe-cy

earthquake,

Interval (81 frequencies), both comoutatic.ns te;n carried CA by tve auth•.'rs,

is shcwn In Fig. 2.9.

 7

Response con.uitint iOts at hiVjher frequoncies were not carried

out since the validity of ground motion data for frequencies greater than

30 hertz is questionable. In addition, the response computation at high

frequencies (small periods) require very small integration time steps

which are time-consuming computatioially. For each component of the record,

the response computations were carried out for 0, 0.5, 2, 5 and 10 percent

of critical danping. After the re.ults from the response computations were

examined, they w;ere plotted on a Calcomp plotter. Plots of the response

spectra are included as Figs. 2.10-2.51; the maximum ground motions are

shown by three straight-l ine se(pgments in each figure, and correspond to

the values nivun in Tablu 3.1.

2.3 Amplification Factor•-

Since the ground motions for earthquake records differ from each

other, the computed response quantities cannot be conpared on an absolute

bo!;s. Ore method of making a meaningful comparative study involves

n.)rmaliZation of the respon;re spectra by equating the area under the

reponrse !pectra between any two frequencles.,5) Another method of

nontnalizing the re.sponse spectra is to compute the ampllficotion factor --

the ratic, of the computed response to the maximum ground motion -- for

diplacemner:, velocity, and acceleration at each frequency for the range

of interest. With this procedure, the amplification factors can be used

to cievelop design response spectra for a given ground motion. This

procedure wc.s followed and is described in Chapter 3. The procedure for

norrnmlizing the response spectra is explained in Appendix B.

 8

2.4 Statistical Analysis

For d.np;nq coefficients of 0.5, 2, 5 and 1) percent of critical,

and for each frequency, the riean and the standard deviation for both thu-

horizontal and vertical components, ot the 14 cartilquakes weret corlputed.

Since there are twu horizontal components for each earthquake, the samdle

size for the horizontal components is twice that for the ve.rtical ccx..por.ents.

Therefore, on a numerical basis the statistical value:; for the horizontal

components are miore reliable than the corresponding values for the vertical

components. Initially the statistical studies were carried out as -,

function of frequency bands; subsequently it was found nect:ssary to carry

out the ttatistical cre-putation at e.ach frequency inscead of within fre"I4uen'"y

bands in order to account accurately for the variation of ;iriplification

factors as a function of frequency.

With the ,eans (or averaqes) and the standard deviations at

various frequencies, the percentile amplifications frori a normal distribution

curve %..ere com-puted from

where *, and 0 are , respect i ye ly, the, prcentile arplification, the

. . - and the standard deviation at a given frequency. The

vY3lues of the zi.',:ficient c for various probability levels are given in

table 2.1. The peurcentile amplifications at each frequency vwre used to

obtain average r:iplification coefficients within frequency bands of interest

(see S.vction 3.2).

7h,,rct!.tile ampl ification computations ,crer,2 also ,-.irried ou.t b~

ran"-inl the dat3. For large sample sizes the two procedures give results

which are penerally in close agreement with each other.

 9

Table~ 2.1 Nor-ma1 Deviationis from Mean, in a Normal Distribution,

in Terms of Standard Deviation, as a FuLnction
of Cumulative Probability P

P deviation from mean

C tandard deviation

O 999 3.09023
0.998 2.87816
0.995 2.57503
0.990 2.32635
0.980 2.05375
0.97U 1 .8079
0.950 1.64485
0.900 1.28155
0.6S0 1.03643
0.600 0.84162
0.750 0.67449
0. 700 0.52440
O.-650
0.38532
0.600 0.25335
0550 0.12566

Sourcf: hable 11,2, .0. 670' "lstati~tical Methods in Quol ity Control "
by Dudley J. lowdle. Vrenticr'-+tal I, Inic. , Englewood C IIffs, New Jersey,
1957.
 IMI

-0.

-0.5~

-*1). '4)~1*. "!N F RY4P . C L iIJ

CWIMfIC WN3W
________________ WI: I I mmmi ii jiii * mi.-. ______- ______

~
-~~ , "4- - ; i 1ir- * Il ~- 1 .1 1

40

04

4. a 1216

ic

FI.2.2VFLOICITY - TIME H1115F0Ry. l94N F~f3Nc4NDO. Cq1Li-.. //i7

CPSrqic NG~3w - UNPDJUSTEV RECORD
 * 'a - -I-- -4- -4- -4 - -4- -4 - -4- + - i-j

100 +

NI.-
0

*1
C&

-100 +

I io1
~L4 it

M.~2.3 D17v'L;4lcUMFNT - ITIMIFI H11f0flY. 'iCN FF c~4010~ C;4J *~~I7

CA51qIC 1469wa - uN40Rju3?(1~ WRDTf~

 30
K 4 -4-4--4 -11-44 -4--4-I.-4. . -.
; S

z
0' 4--

-10

.. L, Q +

I C. I. "i I i

! I It 7 ?~
10
I I;-----4--~ S * S

4 I "I Z'4 it

F!G.2.4rJI-)FLqCF.Mf:NT - T!e 4iiTO~RY. 74N~ FFNwqNDO. CRLU.

CA4SNIC N469W - `3&F(K:MFhTq;LL PIOJiJ9TF[l~RF~CCRP
 so I I I I I I I I I -- 4----4-----4-
i +-i- +I

40+

20 +

Los 0 -
I AA.A A. A A%P
--
t.
- ..
*
~
A ~
AAAA A A -A PA'\ ..-.-- .. ~- U ~I - I 1 I
r, vv 1v, v-V -VV Vv V'I4o J ~VVV V-

-L40 +

-s---t- i ~ ; t ~ , * ~ I-I I 4 1 I * I I a iw I .1 *

.4 ) i ýl IC5 ?.11 3'

F! G..25 VCLOC IT Y TME

I MiTORT.
I 54N FT$S'4NGi. CkLJI ..

CL4*Srq!C N639W - 'WCf*MNT4LLY ROJU9TF.(? Rt'CC'ON2

 U. r -4
-4-1

0.4

ci:

-J
t~9
Ii
-0.XZ

-0.4

V. r,
-1 IS

F!C.2.6 PCU.(.LE.Rci4T!ON T!MF- ISTV~ORT. 'jc;N F;F.INC;NOO. rrfL' C,/3-/12-71

CqFrqTc; 14GR9W - ¶GMFNTcLLY PO)Ja'3IEU RELC!RU,
 I DD. 0
50-0 rill"> + ,lly*c N

2LJ. 0

"IX
<\
10,0

Li
LA

s0 AN,

I-
LA
C
-J
2'.0

1.0

O. C.

0.01 0. 0*11 Q. M 0.1 0.2 S.11 10.0 20.0 5c.C( 100.0

FRCQjf~cps~'
FIG.2.7 RiESPONSE 5FEC1RP FOR~ SPN FEHNPND0. CPL IF. . 02/9/1971 - C15P!PC N59W
UNADJUSTED RECORD - 0.5 f'FRCENI OF CA'7!CAL OPmP)NG
 100.0

50.0

I1
20.0 i
10.0

Li

S.0
z

A-
-J

-J
2.0
LaJ

0.5i

0.2 0.0Ci
0.02 0.05 0.1 0.2 0.c 1.0' 2.0 5.0 1U. 0 20.0 50.I0 D00.n
F9EMtENCY. CPS
FiG. 2B RESPONSE 5FECTFIA FOR EL CENT90. CALIF.. 5/18/19LNo - NS
2.0 PERCENT OF CRITICAL DAlMPING - CA'.TECt1 5'PCT9A
 100.0

50.0

20.0

10.0

(I)
S.0

p.-
-4

2.0
LAJ

0.s

0.01A
0.1 4O.
0,0Olt
0.05 0.1 0. 0.5 1.0 ?.0 5.0 10.0 20.0 50.0 100.0
VFrEOVEcy. CP
FIC.2.9 RE~SPONSE SPECTiRA F'CR EL CENiRO 16940 - NS - 2.0 PER CUNT OF CARI1ICAL
DAMP ING. USING 81 EOUPLLT 9PMCUI FHEUUEt4CIfv ON THiF PilOCT
I - I Aý- 1 .1- mandommumom

500.0

200.0

100.0

.50.0
K\K
4-)

z
20.0

4-)
-j
10.0
L*J

5.0 \/>

2.0

1.0

0.5
0.01 0.02 0.0M 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUEr4CT. CPS
FIG.2IQ0RESPONSE SPECTRA ..JAN FERNANDO.
CALIF... 2/9/1971 -PRt-OJMR 574W
0..5. 2. 5. & ',0 PFRCENT CRITICAL OAMPING
 him [JIILIIIiU I ~ . 1
500.0 I I

I ',XS.KX) \\/
if
~ / ~.
\ ~
I
I
S

4 N~14. ~\~7t
/ .7'
200.0 \ 7 I'4~ \
-
'-I
~bK' ~ k-K. 5,1

100.0
/
~ \ /'

K 7 \ ~ I
/

K',
501.0
/
N.
Li

K
Cl)
/
'.4.

z If N
20.0
// \

N
Li 1
0
-J
10.0 I
~JJ /

/
.14 I
'II
S.

/ K.'
//
N~
O%~
~ ~'
A4
\

2.0 / S. '?~~'?'i'K.t~ S. *4~~ j/

1.0
~ /1'} V N
\\~~*\ I,.: 'El-I -
~'. I

p
-. ~

( L ,
Is 't
U .'.0
0.01
I

0.02 0.05
~fli\
0.1
~
0.2
:iL~~.~~I<
0.5 1.0
AK
2.0
A '.2

5.0 to 0 20.0 50.0 100.0

FHEQUENCT, CPS
FIG.2.11 RESPOME5 F'ECTRR -SAN FERNANDO0. CRL IF.S. /9,f 19- PCO1 SG
0. 5S.2. 5. & 10 PERCENT CRITICAL DAMPING
_ _ _ _ _ _ _ _ _.. _. _ _.. ... .. .. .. w .... . . , . .. .• . .

500.0

200. 0

100.0

$0.0

L)

z
20.0
1-

V
-J
10.0

5.0

?. 0

3.01. 0.05 . 0.2 0.5 i.0 2.0 5.0 10.0 20.0 S.0 0.
FREj2UENCY. rpS

FIG.2.12FFS-PON'SE SPECTRA - SAN FERNANOC. CALIF., 2/9/19'71 -PP~flM'-1 VER11CAL

0. .5. 2. S. & 10 PEFRCENT1 CRITICRL 0PMpiNr
 200.0

I
• 2 A I

/ \

Ni
. *1~

100.0 7 I
2 /

7j

50.0
/
~ ) // -/ /
f N-
I
/
/
j /1
/1 ~ ! LIX

0
I
1/
A' -
I
1

>.
/

..
7"'

7'

K / .,...\•...
20.0
/ 1/ -- -

K
.1

Li
L.J
\ /
I // ~ k~A A

If) V\ \.MI ~A'iNI~ A

z <N AK?
a \I /J I' '

p.-
/ F. I I

4 X
/
K
N
7"
/.
7
Li -9
/
-J

K.
LA
/-•J /
/

I
le

/* N /
, ,

1.0 ", T
\ /

/ /
~yxx
0.5 -I,
/
\
./.\1
I'

\1/ \ N. /
./

// \ /

~2 /
0,? / i-I
I K','
X
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FFRQ"JDCY. CPS
FIG.2.03RESPONSE $PECTRAl - SAN FER .NRNOO. CPLIF.. 2/9/1971 - CRSTPIC N21E
0. .5. 2. S. & 10 PERCENT CRITICAL DPMPING
 100.0
IxI
-"- ai j
1A\
50.0

\ k \.*x

20.0

L)J

10.0.
Lu
/>
5.0
ID

2.0

1.0

0.5

0.2
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 5O.0 100.0
FROEUENCY. CPS

FIG.2.14 RESPONSE SPECTRR - SAN FERNANOO. CRLIF.. 2/9/1971 - CA5TRIC N69W

0. .5. 2. S. & 10 PERCENT CRITICAL ORMPING
 200.0

100.0 I ~~~~il/\-\ 1

/
50.0

N
K
20.0 •mm•

/
u

10.0
\~ /f' ~a\
-1d

-jJ

5.0
•
~

a
"1"
-
/ ha

LAi

N
zi /*
2.0 "'I
/

1.0
/
-(
x 1/
0.5
S.

/
~j \I

-. I \ .AA'\
0.01 0.02 0. Ds 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FPCOOLENCT. CF5

FIG2,15 RESPONSE SPIECTRq - SAlN FER~NC)O. CALIF,. 2/9/1971 - C05TAIC VERTICAL

0. .5. 2. S. 4 10 PERCENT CRITICPL DPMPIWG
 200, 0 'V ~ I ~
i
P -

./ \
/

K>
4/
-. ,---~-

Y
4
'N /
L

4%\
N
2: 2

50.0 F / 4,s, *Ii

20.0 'a
7 -I - --

K
*
9
a

/•

/
Lai •.:- • .
10.0
/
5.0 K / K'
½-'
2 I i

N".
Lu /
/
2.0
/ - . .. _
! ./x. i I i"•r I |

1.0 /~> /
I I
I._t

/0
T

/
t t

-N 4/i V.
,/[/
-'I

/ / //
0.5 \ / Xi J
/ I | f A
All -Z~/
/
/ I'
/ IY\I
V
/
~'
- I / !
.1. -
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FRQUENCT. CPS

FIG.2.16 RESPONSE SPECTRA - SAN FERNANDO. CALIF.. 2/9/1971 - HOLIDAY INN NS

0. .5. 2. 5. 4 10 PERCENT CRITICAL ORMPING
 2001.0

100.0 r

50.0

20.0

Li
L&J
In 10.0
z

a-
5.0
-J

2.0

2.0 -•

0,5 -~

0.2 -
0.01 0.0? 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 M..0 50.0 100.0
FFCGUENKv. CPS
F!G.2,17 RESPONSE SPECTRIP - 5N VFERNANQO. CRLIF. . 2/9/19-71 - ilOLiGOP INN EW
0. .5. 2. S. & 10 PEFACENI CfUiICRL OPMPINC
2•0. ru

100.0

50.0

20.0

10.0

?.2

5.0

//

2.0

1.0

0.5

n n

0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 X..0 50.0 100.0
FREQUENCY. CPS

FIG.2.I8RE5PONSE SPECTRR - SAN FERNANDO. CALIF.. 2/9/1971 - HOLIDRY INN VERTICAL

0. .5. 2. 5. & 10 FE3CENT CRITICAL OAMPING
 200.0

50.0 - T-,---
/X\/

/ />

0. " 0.0 .S Or 0
,/5,
ECN Ar1R ~P
i. 0 "-iE-U_.C-. U--

0.01 0.02 O. ors 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 -50.0 100.0
FIREWENDC T. CPS

F IG.2,19 RESPON5E SPECTRAF - 5AN FERNRNNOO. CRLIF.. 2/9/1971 -VENTURA BLVO NIlE
O. .5. 2. S. ý 10 PEFRCEN7 CRITICPL OWqMING
 200.0 7\ j~\
7K1
T

'7
, N.

~v1 N.. / \ - \ I
/ 9.
/ K~ \7\ S
N
.9
!\, 'a N..?'.
/

~2

A \ K'\
\~/
I
\~,/ X\ N, ,, A. i
I ~
0-.' .4'
a....
'a, \ / -- /
(K, K / K %X ~.2\
50.0 I

'1 '9
'
/
/ -\
'a a' / 7
'a
a-
/
/ N '%~~\\X \
20.0 A /lk \/' Lz'~i
I 7 ~L' \
- N 9. N j : 4
Li
U, /9. .~1 *4/ K
N
L
Nt
10.0 Z
NO
z / 9.,

/ N /
N.
/

K K
/ 'a' \4. 9

N
/
-

5.0 ,' a-' N 'a 4

.9.

/ ,
'a
.
1/
-7, 9.
//
K
~LJ
o~~*
/ a) S NI
2.0
Ilk-

I.0

// \\ e
0.5

N., I

0.01 0.02 0. iS 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREOUENCY. CPS

FIG.Z20ESPONSE SPECTRA - SAN FERNANO0. CALIF.. 2/9/1971 - VENTURA BLVD N-Nw

0. .5. 2. 5. & 10 PERCENT CRITICAL OqMPING
 C00. 0
/ N
100.0

50.0

20.0
i V, i\. V

N 10.0
z

N.y

5.0 0
Li
0
-J

''jo
(1
JN

0.

0.2
liI 0. QX? 0.5 i .0 2.0 S. 0 10.0 20.0 50.0 .100.0

!.
F[O1 .2,21 fEPONi E SECI FA - 5RN CAiIF . . t--/19/1 971
F1F NPNCCI. - EThTURP BLVO VE RI ICRL
0. .5. 11. I PERCEiN1 CHITIMCPI. OD;'-Mf I
(3.4 10St
 500.0

200.0 -~-j

100.0

50.0

LSJ

z
20.0

-j 10.0
-W

5.0
N,

2.0

1.0

0.5
3. 0.02 0. 0 0.1 0.2
0i 0..S 1.0 2.0 ~0 iO.0 20.0 50.10 100. 1;
FRrEtU ENC Y. CPs
F1;-2.22RE5PfINGE 3PECTR'7 - El- C~-NI'N*. Cq:- NS
 MR10.01, IN IF, W qwý

200.0

100.0 .
N.
.. ~. I

4 4
50.0

ý0. o

t-j
L..j

10.0

>). U

I.fl

I-.
Q.~}$K~ 'A 0.2 s.1 n~r !nOJ.O

F-IG.2,23 H 5 p O'ý
CRLp 1 0

Am
 IXII

100.0 .... i./i, . o -.

\i.,",,i7
X,: " '~

,-, "- -, " .? \ i A

20.0 1~'

'cl'

I
x " " 7 •
I
•0.0 "-.. ,,..' '- I
N.,--
-" " '- .L '-- i \ , ".v t-

-- 2 "-'\. .. " , 4 " . "s )<.i ' •N .

y"fK" ...

r0.5

• 1.,.
2FEPNE -.
SETR - L .
TRO ... .R.. 5/8/:tf -'/ C . , /-•
C', .5. & 1
,,. PECN C1IR DM'
1 .- 0 .0 o .0 0 . 0 .'. . . . 1 . 0 0s . 0 .

• .REQUENCY.'CPS

10PRETC IIA"APN
4t
0 .5.2...
 50.0
,
, '
x\ 7~YN7zT77
-
-
'-
,
I.
/
f N~I ! >( ~N 1 y

-4+~; r~I~
~s
'1 .i~ x

20.0
N.

N.
N. / _ V
i \o

KiN A
.~

\ ,i~\ /
.~\ \
/ I "-p
~-

F \\/ I
/ . 1\
10.0 E

~N7J 7 - -.

// K l/N.j~ -p

5.0 2 I.-

'I ,~. /1':2 -~ -~-

(7< /
\ /

/ ~. 4'>6 ~< N.
N
~ N

2.0
LAJ - / \ I ....>

1.0
( / --

\\ ,.// *~
-A -p.
A.
I i
'p
/
/

I J~I
0.5
7K ( >K /\

1
I- N //

-7 / -p

0. 1 E
/
2"'- F W
'7
1?)
N
¶*
-V I
(/1K \ /
/ 'N
-p
I'
/
' 'p.

/
x N.

0.05
1' / / AK/N K I
K>
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FPEOUNC•T. CP5

FIG2,25RESPONSE 5PECTRI - EL CENMRO. CALIF.. 2/9/1956 - NS

0. .5. 2. 5. & 10 PERCENT CRITICRL DRMPING
so.0

IN
20.0

10.0

5.0
Nl-

2.0

1.0

0.s -,X

0.2
VI1/
0.1

0.05
0. 01 0.02 0.Q5 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 SG.0 100.0
FREQUENCY. CPS

FIG.2SRESPONSE SPECTRA - EL CENTRO. CALIF.. 2/9/1956 - EW

0. .5. 2. 5. & 10 PERCENT CRITICAL OAMPING

S -.
 50.0 V ~ * ~W * ~ V - V

/ x \
\/# X- x~ /
N I ,!XI "../ X
N
2 xj
/ / /
20.0
/

10.0
.t >I P.

VP
/

/ ,
/"

*
, " / \" •l "i." ",
5.0 K
/* Ko Ie
LAI /,\ %/ V
N. I~ /
N, >/ ".. "/' 4

-
z /
2.0

1.0 \ /
K l W•
, --- -.... ...-.> - 0%

-AJ
x Y1r~ r- ....... I.
N

0.5 -~---~- 4 -. ~...----..'-- 4

/
I.~-.-~--- I
/ R i, /*.. ' ,,
XV N /
- 'I

'N>

N
AN' N x I' / K"
"I

K
0.2
N
~<

~' \\
N /
/ "
,, ..•.\ !-j, /
N
7 \,ix,
,e N. ./
/

N)
l
><
N
Is
0;) •, ', .,
-½
*1
Al
\/
N
/-, -,,. 7'-.
'K; "K
<N>
/ \.~I/
'I" a

0. O5 L /1K> / i /
0.01 0.02 0.OS 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 So.0 #100.0
FP4EQUENiCY. C.PS

FIC.2'27RESPONSE SPECTRA - EL CENTRO. CALIF.,. 29/1956 - VER1ICAL

0. .5. 2. 5..4 10 PERCENT CRITICAL OAMPING
100.0

41 -.Y.-

20.0 * v4

XjIt Ix

0.10

0. 1 0.0 0.0 0. 0.2 05 2. C 00

__.0 2c.0. 0.

~4. ~ . \.Ej \. CPSV

F.Ir-~
G.7CF7NS PCR -E ETC.CLF.4//98-N

0. .5. 2. 5. & 10 PERCENT CR]TICRL ORMPING

100.0 T Z I ' ' I~

A'

20 . I 0

b
'kKxl

-v0. 2
% ~- i2Z'

0.0 0.2 00 01 ot05 10 2050 1. 00 5. 0.

FI.229ES~oD SPCTR -- EL CETO //98 -I.

0. .5 .5 41 E EN RTCA RP
 ion c
0
..................
7 .'

,V CA'x

20.0

T7~

Li
-. V' \~'\ >
LJ
U,
N. S. 0

Li
D
-J
L~J ~ ,I.. .t-

1.0

O.S

0.2

fl.1
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FFPUENCY. CPS

FIG.23ORESPONSE SPECTBR - EL CENTRO. CRLIF.. 4/8/1968 - VERIICAL

0. .5. 2. S. & 10 PERCENT CRITICRL ORMPING
 MoG. 0

50.0 >

20.0
/4/
10.0

5.0
\ A\/
3-

Li
0
-J
U.'
I I
0

N' 4 \
1.0

0.2
r4-
0. 1
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 S.O i0.0 20.0 50.0 100.0
FMfQUENCT. CPS

FIG.2,31RESPONSE 5PEC1Rf - HOLLYWOOD. CALIF.. 7/21/1952 - HLWO 5TG 85MT NS

O, .5. 2. 5. & 10 PERCENI CRITICAL OAMPING
100.0 Ni
I 7-.

A-
K
N
N '-1 4 2
- -~
77 * -S.-.-

y
~
C
4

I .. N, I ~ /
.1
N
~ 1/
\ \ 0

llý
x \
f~o~-~~fl
S0.0 \7 I\/. /
.~' A

10.0
/5\
, X!
5.0 j .OF

I-

1.0

0.5 K\2

qb!
X/&b <\
"Sý

0.2 I
N!

0.1
0.Ol 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 18.0 20.0 50.0 100.0
FREQUENCY. CPS

FIG.232RESPONSE SPECTRA - HOLLYWOOD. CALIF.. 7/21/1952 - HLWD 5TC, BSMT EW

0. .5. 2. 5. , 10 PERCENT CRITICRL DAMPING
 50.0
7 1~'<I2K~'~x~/~\
\'
t
/-, " ' i.o .
VI. V
\/2
_
20.0

10.0
IV~/1
IA
~ r -- - -- -~
~1* "kN

/ \~ §/ 'INC
K
K
.
I

5.0
' , \ 7F F'
y \' - A

Li / y

r\
I
.1 '.1

z
U,
S.. N -,
K 4- I.

-J
Li
2.0

1.0
'.5

'S

/
/
-

j~iXj\ /

/ / > xx/K
1', N
*0
0.5
No
/ \ ~i/ ~
N 4 tN N
V 7, K jX~ , /

0.2
N
-
/

/
~ I t~1
'I

7>
/

0. I
4
\,

% k\ ,xI (\ ~ K

7' / \ --.
N
'N
/ N
;x.
/ N
0.05
f
/1
(
/ .5.
N. I K2 K
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS

-[G.2.,33ESPONSE 5PECTRA - MOLLYWOOO.'CALIF.. 7/21/1952 - HLWD STG 8SMT VERTICAL

0. .5. 2. 5. & 10 PERCENT CRIIICAL DRMPING
 5-0. c
~? '~ I 1 'N INe
7 X/ ly
I
- J.
~/' \; /
/\
~/ \)
11
~/
I
~ ~97\ N. ,' -o
..
\ 7. \ ,, H- ~" 7
20.0 yI ~*1

N I.
/\ /N.I7\A/ \V\o / N.

I V ~ ~ 7
V

10.0 . 4f

5.0
IV ~ I A*~1i~/
kJ
In
N
z

1.0
~hJ

OF$
,// ~T
k//xr %'

0.5

0. 2
V
7~~~~~/ KIv x>i\ 1

0. 1

0. 0$
0.01 0.02 0.05 0.1 0.2 0.9 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FRFDLJENCY. CPS

FIGZ-34RESPONSE SPECTRA - OLLYWOC3. CAPLIF.. 7/21/1952 - HLW-9 STG PE LOT W~5

0. .5. 2. S. &10 PEtOCFE.!T CRITICAL DA1MPING
 50. c
.. - 7. ....... ......

t. 0

20.0

N1 A

10.0 A2

5.0

1.J
A
z

U 'S
C
-J
1.0 C'

0.s

'K1

0,2
4NI

0.1

0. 05
U.U0 U0U2 0.05 0. I a..2 0.5 1.0 2.0 5.0 10.0o 2oo SO.0 100.0
FfiEQUEWCYf. cps
FIG.2.3580PONH5 SPECIMPý -- HONOLOO0.
CPlLIF. . 7/21/1952 - HLWO 5IC PE LOT EW
U. -5- 2- S. it M0 PERICENT CRIT ICAL DAMPING
50.0 1. "
f I . -'T-

/ "I ~ t~
20.0
'I
' 4' ~
jill ~. A J
.5

,5.
1 '5

7 \I~/ 0's
it ~
I"' ... \~..
1~
10.0u I7.
~ 14\ I xiy
,/ 1' NA
S.0

4~' I jKs5
i~ii+; K1 V ~A.
A' '-'4 4 -Y
2.0 I.

V.
K, j ~
1.0 ~4X
// \~/ '. 'I 555 .5\

0.s %\ 1K / I -~---s~.-.~---~
/r
I'

0.2 4~ I -~~-~--- 4. - m 'A

'~i
! I

//N\%
.N
A
,/K?>

0.1 i~i~zi~
1,, T N,

0.05
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CP5

FIG.2.36RESPONKE 5PECTRA - HOLLYWOOD. CALIF.. 7/21/19S2 - HLWO 5TG PE LOT VERT.

0. .5. 2. S. 4 10 PERCENT CRITICAL ODMPING
 50.0

/N

20.0

10.0
x, \ .N, . ' I 77,7 77[
- 7 4o•
5.0 X,\o/

z
2.0

1.0

0.5
\.,

LL
'x_.-_ , , • _
,"
• - ,i1..~
• i"" .•' X - '- .. .,".I¢7I4, .

i\ . ......
-*

T •,~~~ 'I .V T ..

I "" "% . ! I " , lV! .""' i " -, i • ' I ¢.

/ ", /'° I/ \t d. \ i, . ,",...,"
0.1 ,

0.050.
01 0.02 0.05 0.1 0.2 0.5 1.0 ?..0 5.0 10.0 20.0 50.0 100.0
FRE0UENCY. CPS

FIG.237RESPONSE SPECTRA - SAN FRRNCISCO. CALIF.. 3/22/1957 - GLON CATE PK NIOE

0. .5. 2. 5. & 10 PERCENT CRITICA.L DARPING
Sc. 0

Xa te .' ' 7 ~
wS x x~7 .~

20.0
K / ' -\ ' Nt'

'a3'A'"t .' 3

.0.0

5.0

/ ' ~~~ X ~~/ .* 1' "a/ .

1.0

10
0.5

0.2

0. 1

0.05
0. 01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS

FIG.2I38RESPONSE SPECTRA - SAN FRANCISCO. CALIF.. 3/22/1957 - GLON GATE PK SBOE

0. .5. 2. 5. & 10 PERCENW CRITICAL DAMPING

-1 ý .. .., ý -- "'-:'.7:'-- - - 7.:

I G.I
0it

fl, •
\J
' f t "
2.0 v
1.0 L <2
\
",,. , I , "

.0.5
,

/ \,>

0.• .- • _

/ ,.. . .. I.'
- "../ \
'- S.- . .i V. '

0.01 "K ""\ " : " x. , / i• /

0.01 0.02 0.05 0.1 0.2 3.5 1.0 2.0 S.0 10.0 '20.0 50.0 100.0
FFEQUENCT. CPS

FIG.2.39RESPQNSE SPECTRA - SAN FRANCISCO. CALIF., 3/22/1957 - GLON GATE PK VERT.

0. .5. 2. 5. & 10 PERCENT CRITICAL ORMPING
 200.0

100.0

50.0

20.0

LU
U,
10.0

I-.
-4 5.0
-J

2.0

1.0 I I

0.5

0.2 -
0.01 0.02 0.05 0.1 0.2 O.S 1.0 2.0 5.0 10.0 20.0 SO.0 100.0
FREQUENCY. CP'S

FIG.24CRE3PONSE SPECTRA - FERNDALE. CALIF.. 10/7/1951 - N46W

0. .5. 2. 5. & 10 PERCENT CRITICAL ORMPING
50.0

20.0

///

2~1.0

I / \ I\I

0.5 /---

0.? ' 1 K",1 \ z__.\

0.03 0.02 0.0$ 0.1 0.2 0.S 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FF"OUE. Y. CPS

FIG.2A..IESPONE 5PECTRA FERQPLE. CPLIF.. 1.0/7/1951 -54W

0. .5. 2. 5. & !0 PERCENT CRITICPL .,NG

 1U0.(.,
,.N.."
"i"i,
., ".
_ I, 'i' --,', . ".'
. ', "\., ',.- '\ .'I.,
,. " ". -
,
'
4
i.-", .
\ . • . ... S .\ . . ",-
S0.0 •.. .",I.r,
N,.
-44
__-, *-KI,,"' N ",,3 . \i. ,l.'
( ), -•. ,.-• N "t. ,% , 7?..,
20.0
"v , \
"-.. -• ×- -,. ., • I-XV N,.-" • . I N ,

/\
10.0
/I ~x/ x
"I
.- / 4f \l
5.0 /

I-'
_ _..N
'JJ
/ N
\ 5
a
'N
~, N

7
Y,
1.0 5.
/
j5" II \ F', I
U A>~~ 1 1-11 7'
/
/ L/N\ AV
0.S /

0.2
S
7,
/

5,
'a
\//
V N
.1Y

L
~/L/4)<
n i LN I I

0.Oi 0.02 O.OS 0.1 1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FFEQUENCY. CPS

FIG2.42RESPONSE SPECTRA - FERNORLE, CALIF., 10/7/1951 - VERTICAL

0. .5. 2, 5. & 10 PERCENT CRITICAL OAMPING
 100.0 I
N N" I
7' A
'I -. / / K~K LX
N
S0.0 '---A
I
C', /A\X1 ,X .\ /1/
/
'5

K" )* V
lely
NA-P
144V
A ,><
4/ -. Ca
.5 '5

20.0 41< * A
-im T W
'N
1

K
0
4
\' ~ N
>1 L
I
\ ,~\/' \ .,, >
10.0

W 5.0

U'

2.0
LAJ

1.0 r74 i
X\ /,\I-7 '
0.s

0.2
A.A
7 "x/
n iL
0.01 0.02 0.05 0.1 0.2 0.5 1.0 Z.0 S.0 10.0 P.0.0 50.0 100.0
F~REQUEJNCYr. CPS

FIG-2-43RESPONSE SPECTRP FERiNOflLE. CRLIF.. 1.2/21/1954

- - N'46W
0. .5. 2. S. & 10 PERCENT CRITICAIL DRM14pl~
Icc. o I " "/
-:A' ./2 I , . , .i

" ., iV" ';/ /

X ,,-- 7,

20.O
""~ ..o I " ,! - .T '-, 3"'
' .
!'vl ol""
10. C

K
V
--. -' " ,;/ ., .2 i ,. /
/ /
5.0
( -\-
//
K

,5
Li'
W..
2.0 .' " -" // -. 1~-~.---
/ \
/

1.0 i I f

K IY -l
x
N

/\ N

/
0.5
>IZT ~K) ;~
\/

/1
4

' t':~ / K '4/

~1~
I l

I *~
'A S..
N, A> 1 K>
A' VI 2< /
/ S.

UI. (
~./t ~i /

/
n1I'l
0.01 0.0O 0.05 0.1 0. , 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FFEUENCT. CFS

FIG.244RESPONSE SPfECTRR - FERNDRLE. CALIF.. 12/21/1954 - N44E

0. .5. 2. 5. & 10 PERCENI CRITICAL DAMPING
 SO. 0
/ / k\

20.0

!0.0

Lu
5.0

2.0
Lai
0//x

1.0

0.5

0.?

0.t1
0.01 0.02 O.OS 0.1 0.2 O.S 1.0 2.0 S.0 10.0 20.0 50.0 100.0
FtfOUENCY. Cl'S

FIG2.45HESPONSE SPECFRA -FERINDALE. CF.IF.. 12/21/1954J - VERT.ICPL

0. .S. 2. S. 4 10 PERCENT CRITICAL. DAMPINJG
 100.0
I
f
50.0

20.0
I
10.0 I
ui
t&j
(In
5.0

U'
2.0

1.0

0.5

0.2 -

i
0. 1
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0t 5.0 10.0 20.0 50.0 100.0
FFiEOENCY. CPS

FIG2a4ERESPONSE SPECTfRA - EUREKA, CRLIF., 12/21/19541 - NJ11W

0. .5. 2. 5. & 10 PERCENT CRITICAL OAMPING
 100.0

50.0

20.0

10.0

L~j

2.0

1.0

0.

0.Ol 5.O 10.0 20.0 50.0 100.0

F IG.2,4MEPONE PFTC.TPA EU3Eep.1 CPLI.. /12'2`195~4 -. N7QC7

0. .. ~ ~2. S. 4 10- PERrE.N'l CfIiICPI' DQPIPING
 ýý'.T- V '-

( 9> A. A

.41

-K 4 <r~
~-<I
10.0
N _ A

-. ~ /Irv, ~
U0

N~I L w ,.f i k / x. / I •.•;• /' !,.\ 1 N' --- 7'.--

'7' -T 'N

0 '~ / / ~t ~':>yl
V 7> /

K/
-N.
N *"
NJ. /
\~I .
A

*
>'&/ I 'N
/1
'N 1'
\{/ \~
/~\ ,, >lx /jK.\ / K _
1~ ~.- / \(

0.c
3i O.05 0.1 0.2 0.5 1.0 2.0 5.c 10.0 20.0 so.c 100.0
FFITIEC'. CPS

FIG2A48RESPONSE SPECTRP - EIJREKP. CPLIF.. 12/21/1954 - VERTICRL

0. .5. 2. S. Al 10 PEPICENr CWIT1CL QPMPNG

7 Z7
 _ .
---.-. * \ /'

20. . .

0, j. » ŽK .0 1

.5 ~.

C.

S '4~~~*~* ~

--- 'I
4 ¶.
4.
*

N. oH- I
------
""Ni i
'4-'
"4 '4
<I
A '
5 '1

* -.. ~ -1-.*~,*
-- 4.-..

.5
~4.

-f
.4.. 1
A
.7.
p V ,..•,

(
4

p ~- ~F. 4 A. .. '- .-.. -

-4 -".

~ /
4, ~, 4' 4

'1
'4
1. 11L4
t

r' .1 ,~
•,),0 ;9n t3 50.0 100.0

*8. 9

.2 0~S4l 49\
 (2
01G. n

[ t. A t

'. N .~

L~J

2.0

I.
.4
'a

/
Ln
1ý0

VI
r\ \~/' I
,.0

o.s

0. N
C.0 0. 0%21,
.. . 1.,0. 0. 0- 20.0 50.0 100.0
FREAENtCT. CPS

FIG-2.,%n-EPONSE SPECTRRA - H-CILLISTER. CALIF.. ~4/8/1961 - t8:

0. .5. 2. 5. & 10 PiECENT CRiTTCRL ORMPING
....
 100.0

SO. 0

et.. 0

10.0c

Li
(n
S.0
z

0
-J

I.0

GAL 0. 0o2 0.5 it' 2.,0 .0 10.0 'CI.0 50.0 IOu. 0

FRfQUICN-~Cf C-Ps

F I C-2-51 Pr"_' SNINSE' c)-PE T-ýP FR.¶~. CCTF-I . 14/8/ 196 1

FtC2.5115$Tf. - VERI ICRL
0. , 5. " e , ý.. k IC) 1. Cfi
. PEfi(:FNI - I I ICAIA-OP101P
I I N('
 61

3. PREl ANKADISCUSSION 1f: r.[SbLTS

3.1 Ground Motion

The n imum va I ucs of ground iot ion obtained from the adj usted

acceleration, velocity, and displacenwrnt tinx!-h;storits for the 3 components

of thbe 14 earthquake records considered are givt:n in Table 3.1 along with a

bri,. f t.ite de'scription. It -.Ahould be r:'coynied that although an atteupt

podt
p.ss tc charactcx.ric the., titv: descriptions ax rock, alluvium, or

othe•rwise., t:Gvese dotcriptio,% are not. cunplettly d-pendable and accurate

owinq to a lack of satitfactory information aoout the geologic condition.

at •" tc:, of the .5ites if,,re

f- ,tr ,nynm1tion I It, rumn (5 have bee n lIcated.

Th, rtod i P i-;ona I rat ; o o! v wo r 4., , v aand i r.

ce 1hvaii Y mum grounmd

aCC.tl&ra.t i<' lo, l tY.

; l di ý,p ku: nikent
, oiiy., resp :ct Iwv: I y, ki also gi ,.en in
"I

TAI t 3.. Ow rot iQ Od/v' ;i at Ieast in part a function of tIe focal

dI a n , op! the. 1 a ,ke.. Ind the attccnuntion

a.,Icthrt Of iwltion in the ground. (6)

c.f very largre local dis larr., Vh;vre. the ground motion opproaches; a hari-monic

rotion, c o d.v" .a:pproch.es 1.0. ]he rati o Incr ise; " ... the.. focal (. ,tarlace

.,. [ .. ra ~r.polq
O ,i / y f r thie e.•rt houal we,c~.c
ords cOn!rV id(Irc• i) Ohw.,

,tild! , 1 81,, to 17,61 for, Ow horioruna arnd 2.66 to 30.58 IOr the vertical
2
C, Cr..ry-.m t rwc'5pc t ivel y. Thc,- a'vrage ad/v for tilt, 23 liorl tontal conpotlen.u

a. .Ohe It' vorrtical conpon:!nt's are 5.c Aind 10.6, rA:.spert ye y. Aong, the

42 -o:ne; reatorrs cunsidfr,,'d in thi s tudy, only one coripon,,.t (vertica'

ceývonret l U 1-rl.rn, 1f 2037 P. T ) ha., an Wd/v 2 which iv

Is xxtrcýnxly
ah.-t,
hiO• ad/v' 'or this ri 1cs 6 peilrcet l tjher th;an the riex1.

hiqh-est value of tO) vertical colJonmne.it If this hitilhett ad/v2 value is

n.ot imc -inAd, thI average a4d/v2 for the reroi'inirg 13 vertical contrionc,-ts

2 and
redwcc:.. to *. 1. A o) vucrriry ad/v
the average values for horizonta|l
vertical comrponents for jlluvium and rock is presented in Table 3.2.

Further discussion of the dot4 presentad in Table 3.2 and particularly

with reference to classification by acceleration level appears in the

next section.

Since the Ahrno couponents of the gjround tro~t ion in a given

record for a giv en 5 ito W.~e the s ane: focal d. tacztht- crence in
diIf If

jad/v v.aIum~ p robablIy a r iwe fromr at tenua ti un'of t he g round nti)on aod loc al

g~oloI I c alI, n P c t v on t r ann m ~I

a 1 ion no( t ho t r Our 11 1T)t, I oIN .

T rto(I Pd 1 u2 slo
POIM Suo,01 hE4iIOn
d ler'tt ' sj'

of: qvi.i5,l I. IIU%~nto nor'tIhsea W. .e'edN tho r-at&4 ad/'d 2~K~* tOWC1

Q1 15 A) fIhIN- I1,.
.ruc 1i"ViVlao.t Ins i h '
locjto 0 ip osiuri.1. 1e On no

In5 tX~Iý 1" t. I1th" 44. O ~I

n q .pyr a a%, ill , .e.-d OV-1 a" t Iator 10. W o 4.

I~is*'t..I.Ii~!..~. (i.sf o.w Iy I .iappe ca

:t.~1 l " r th.t¶ " afp
dI/tV 2 ja tu i%

"Ifor . AIYUM
IWO ""11-0 1541 Wel, *~~*a '04.0y A. W. i

""UpIo 10, al IssQ'> mv-0 s"-' it A v .1

it is normully castonwry to work with the auart~tits of welocit.y an

 i onsh it) b
acc c Ic ro t ionn.. and the re IIt. wczu-A.n themi. Ono. reat on I or d

is that very fev..

good rucasuremtents of d isplacemont hav'e been recorded,

and the least is known about this quantity. tn an effort to arriv.e at

the intel'rolat onship bet~wcn vaitis of vcfloc; ty anid acctecrat ion, a

number of studicies were f-,dv of th. di(a ta. rho i fcati lt-ir of .h~

sl ud ius 'or,4 uecocwed in thte tlb lr!ý which orcl df:!.ýCr ihtv hr icefly ill tht,

t Ib v
) T01 3 3 thI rt, I 4 f!1t~ I.ed *:
p 1, lnmit.t I I tho ni it) ) 1 i1 t

v' 1c l icO
11. y ti:o t.n Ifirt. Ito i) 141W t.lit! oI~, .cmI I .41d V,'ilIt I c c.( ttiit )Iiet!

ki tlit, IL ttt.
Io,0 I e- v c I w! (. tol
at lit otl o I(iie f (IIIti ( t,~Iv
y '.i I lit' (10I,
1 I1

I.I
Itli' f r ~ Cit I4 col1 hij~ilwhý cit iit 4I'%(e iti ! 41tiI' I) I I nillY,

h~n ht'HirQ Ie.

' uu
I
4et~ s .~ . co. , /it at-I i "Ic tek, ii

- f -51 li I 1 " 11 1-opi I~it I it ýA,1 10 ii~oýb1 1.tt.t W I

. ll: f~1 1 eu
jj 10s11 "t nlf1
. f.,

OP -1 1 i fnt t ato. I .fit . w;

I t- I iper I
It I I44t iji I pt
I III
rei w ile 5 .. 4111 itl~ . Ihi.

*, pt $i Iej~ii4 IiI~ l'p

I lt
in~t i woI k/sif I,)fj ~it leI I'IIII Il Ii d( 1,

5
I I pi~3.- ) 1H.' oe - fý t it5(11 1 j ' I I.,
t id #Itit litI. I tenV I fi' F.iiioisst it W
 Af

va I&% **)*Id "nLt bef o~~bt.ainmhd by a' (V

dCa in *unhu*~rh? WI I typ Nh~ Q

Table~ 3.4, but. instead individual trace data, as shown in Table 3.3,

waert avuraq*ed. On this bas is, aN wrd in Table 3 the~ ratilo of ne*

vo~loc it y to the' owccu I'rlt. o" for thet've~rtical t~races convar~ad with tQ1

of the hior izontal t roces is on thv order of 0,90,) andi ranpgs fro U
0,h

to 0.99.

Similarly, in Table 3.3 in shown the ratilo of the dea ground

acceleration from the vertical traces to that for th.e hocriz~ont~al tr4accu;

these values are shown in the rihjtt-hanird too cou'rns of TalsIe 3..In

order to obtain a measure of those ratio~s, which are nwhedter

c sca*ling
purposes fa~r the vertical spectra, studies of the data for various clea5es

of traces based on site charocte.ristis are orcienfltd in Title 3.6. -

will be nioted that the vaIucS. rorp, from 0.40~ to 0,7 dopondrS on I

accelerationr level included and th'e typc of Sit~e, otc. on the basit of

these Iata, and real izing that only a .few.s-pkwevilt>ed for APh rock:
sites, we selectcd for use later o value of 2/3 as rWil of verl1caI to

horizontal acceleration.

3.2 RepneApiiain

Studies of the response ampifitation in thr va~rious ranges of'

frequencies %yere made by studying plots on the fur-way Inarthric plot,

with the imput groun~d mot ion rormal~zed to the folcing va~vs for all

of the earthquaJle motions st~udied:

(1) A mxnuiirm. ground dmsplacer~ent -of f0irch

(2) A maximum ground veloity'of 10 ;Vflec

 W1

Hil, mk I i l d vi) I m - I i w I It c ilm

I',p I f-I .C11fi* 14!. 1 wr%,,
I IC,1

calca Ica Wre i.kde ore shovri it Fitis. 3.1 V

t-o . SeparLc plots

o re drawn for vyertic al i nd tot hori.-onlital conm:.on'.rnI for 1t,,' eflanpInri

ofva

of 0.1;, 2, s d I1( pri'1..,41t

011 f1 1*1111
rlw f .11,. v ,ivo 1(1c, ( he

11ira p (rls' l I dft ; t r, i tuit.) I on) 111t 10. 15 01 d t.ht- 7ýý per. en,.l 11.ý
Vc tr k 1ag) of
re .,pon.se. For ot-he r probab I iit y I :,ve IS the !I Iots wou I d have genera I I y

the same charactcri!stics.

It can be seen that the anlplifications for d plac.enwrt are for

a pract ical purposes cons tat -/or the range of frcquencies from about

0.10 hert.r to 5onem upperi limit af the order of: u.4

O r 0,5 hert?. It is

a. .so ciearý frorr t.ho Ii gurj.s thnt the arqmlification factore' for acc.leration

ilr: v irtuailly consli.. n t for the raice. of frequencies from about 2 hertz to

;about 6 hert.7 in the hCr iontal responsees and Irom about 3 hert2' to about

10 hertz in the vertical responses, and thcrn decrease fairly u iformly to

intersect the grounm' (rotion accelerations at frequencle.s of the order of

30 to -0 het'Lz, varying with the direction and the damping factors.

For the interrxidiate range of velocities, the horizontal

ang I ; icat ioris appeýar 'n genera I to be s t-t IstI cal I y pr a t. ica ly cons tant

,n th•e rijnqe of f re~qutnc iev from about 0.5 to 2 hertz or more, al though

t hc,'• ,•,re orca! I ona.I I i gh t round'i lncg! w ith a hig her vfli,1n it, the central

pOrtion, but there is no bow is for using other than a !iiinrile constant

ar.ilification level it) this range. For t'he vertical corqponent, there is

evidence of a slight drop In amplification factor as the frequency

increa•es in the internp.-diate range, but the drop is only of the order

of about 10 percent over the frequency range for which "velocity

 66

arrplification" is valid, and it therelore appears reasonable for practical

purposes to use altu a conitant a.li fa',tion for the veloci ty ran.qp. ;s

well as for the displacement and acceleration ranges.

An examination of the response spectra sho•,..n in Figs. 2. W-2.5)

indicates that for most cases the response spectra for vertical ground

motion are flatter (broader) than those for horizontal motions. Careful

examination will reveal that the response spectra for ground motion in the.

vertical direction approach the ground acceleration at higher frequencies

than is the case for the horizontal direction. This pher.or.:-ron is due,

nmainly to presence of higher frequencies in vertical co.Toor:.en!s of

ground motion. Therefore, in constructing a design secrtrun-, it Seerfs5

reasonable to amplify the vertical ground acceleration over a larger

frequency dormnin than for the horizontal ground )cceIerati;n,,

An eantinat ion of rf.sponse ampl ifications, es.pecia iy the

amplifications for velocity and accelerat ion, indicated that arth.quakes

with low ground acceleration have grea.t.ur anplifications than with,

a high ground acceleration. The prese~nct, of sharp pieaks in ai acrcra-

reduces the response anplification for high frequencies. Th,, is

in Figs. 3.17 ano 3.18. Figure 3.17 .ho-s the adjusted accc13rlon-t,-

history for El Cent:ro, 5-18-19140, INS coprronent with the fr.s,,. o n h, sth

peaks scaled such that the n\ixlmum-round :acceleration i! i.O.. Thr(

response spectrum for 2.0 perc ent of critical c-) I ndr i..-en to cround

acceleration for this accelerogram is corrared .with the; er- itrout

fictitious high peaks, Fig. 3.18. It is seen that fot hich freqtuncies

the arrolification for acceleration is lckier for accelerogr.:4••oth hi-h

peaks. The difference between tMe tw'. curves for inaerndate a.d low
 67

frequencies ar;ses from normal izinq tht response spcctra to only tihe

ground acceleration over the entire range of frequenciess (see Appendix B

for further discussion of this point).

For this reason, the. statistical valuks of response, anpIification

were cormputed for two groups of data. One qroup included all data,

whercas the other 9roup included records with peak accelerations greater

than 0.lg and 0.05q for the horizontal and vertical conponents, respectively.

Th:, r'cords which were included In the lattUr group are as fol lowS (sce

Table 3,1):

Pacoitmr Dam, 2-9-71, 0600 PST (al-l con.)onents)

I
Cas ta; c, 2-9-71, 0600 PSI (all convonents)

Holiday Inn (lfirs t Floor), 2-9-71, 0600 PST (alIl conponents)

15250 Ventura Boulevard (Ba.suillnt), 2-9-71, 0600 PST (all comF~onents)

£l Cent ro, 5-18-4,0, 2037 PSI (all components)

EI Centro, 4-8-68, 1830 PST (NS con)onent only)

San Urnncisco Golden Gate Park, 3-22-57, 1144 PSI (all coilnonents)

F!:rndale, 10-7-51, 2011 PST (Nl46W and S414W componornts only)

FerndAle, 12-21-5 4, 1156 PST (N46W and N44E convonents only)

(ureka, 12-21-514, 1156 PST (all romponents)

Hali ister, L-8-61l, 2323 PSI (N89W and verticail component!, only)

The ý,tatistlcol valuefs of response anmplifications corresponding

to $0, 75, 90 and 95 rpercentllc: levels obtained from a nornvil dlstribution

curve for 0.5, 2.0, 5.0 and 10.0 percent of critical daiqmng are presented

in Tables 3.7 and 3.8 for all 28 (horizontal) or VI (verticai) coq)onent.:;

along with corresponding amp1lificatlon values for those horlzontal arnd

vertical comlonrnt.ý with peak ground accelerations greater than 0 1g and

0.059,respectivelt. These amplifIcatIons were obtained by first confutlnq

 68

the mean and the standard deviation of amplifications for displacement,

velocity iod acceleration at each frcqucrncy for whi.ch resporse co-.putations

niad been performed. From the mean and the standard deviation at each

frequency, amplifications corresponding to various percentile levels wre

determined. The displacement, velocity and acceleration amplifications

for a nunmer of frequencies (within a frequency bond) were then averaged

and are presented in the tables. The frequency bands used for averaging

the amplifications, which were establ ished, through examination of the

spectra, are as follows:

Horizontal displacement 0.2 to 0.4 hertz

Horizontal velocity 0.4 to 2.0 hertz

Horizontal acceleration 2.0 to 6.0 hertz

Vertical displacement 0.1 to 0.3 hertz

Vertical velocity 0.3 to 3.0 hertz

Vertical acceleration 3.0 to 10.0 hertz

Amplification values similar to those just described 1ýut c.tarw

from ranking the data are presented in Tables 3.9 ond 3.10.

The amplifications given in Tables 3.7 to 3.10 lndicate that for

records with horizontal and vertical grounr' acceleration greater thAn .1g

and 0.059, respectively, the velocity and acceleration a.rmplificatiors are

lower than tha corresponding values obtained by includini a); data. In

other words, the strong motion data clearly indicate a decrease ;n

amplificatlon, especially for the velocity and acceleration r"iors,

as compared to the case whkre low intensity excitation is im•cluded.

 Table 3.1 Ka€imium Ground PlottO:ns and Site Description

ad decrimtxion Remarks
Record Maximum
descrlpton qround Acc. ground vel. grournd displ. 2
a, 9 v, ;n/sec d, in v (See Sec. 3.1)

Pacoima Dam, 2-9-71, 0600 PST (Record M 041) Highly jointed Small building
diorite gneiss houses the
S 74 W 1.250 22.49 5.11 4.&A 4 km from instrument
S 16 E 1-241 43.70 23.18 5.82 surface faulting
Vertlcal 0.718 23.06 13.75 7.17
Ref. (8)
Castaicy 2-9-71, 0600 PST (,Record TO 056) Sandstone Small building
houses the
N 21 E 0.333 6.7) 2.05 5.82 instrument
N 69 W 0.23? 10.55 3.22 3.14
Vertical 0.138 2.7-5 1.42 13.13
Ref. (8)
Holiday Inn (First Floor), 2-9-71, 0600 PST (Record IC 048) Alluvium Instrument on the
8 km from, first floor of a
NS 0.258 12.13 8.70 5.90 surface faulting 7 story RC building
EW 0.137 9.68 6.37 3.60 structure
Verlical 0.177 12.81 5.37 2.66
Ref. (8)
15250 Ventura Boulevard (Basement), 2-9-71, 0600 PST (Record I1 115) Alluvium Instrument in the
water tabie at 55' basement of a
N I1 E 0.234 10.96 7.07 5.32 12 story RC building
N 79 W 0.154 7.88 4.48 4.2c structure
Vertica) 0.108 4.77 3.09 5.67
Ref. (8)
El ceitro, 5-18-40, 2037 PST (Record IA 1) Alluvium to Instrument on the
about 5000 ft first floor of a
NS 0.352 1:.b88 4.74 3.35 2 story massive
EW 0.223 112.2 6.58 4.13 concrete, heavily
Vertical 0.280 3.95 4.41 30.58 reinforced structure
Ref. (8)
 Record max imum Maximum Max i mum ad Site Remarks
description ground acc. ground vel. ground displ. -2 description
a, 9 v, in/sec d, in v (See Sec. 3.1)
El Centro, 2-9-56, 0633 PST (Record IA 11) Alluvium to Instrument on the
about 5000 ft first floor of a
iS 0.036 1.52 1.27 7.65 2 story massive
EW Q.C55 3.11 2.48 5.45 concrete, heavily
Vertical 0.016 0.75 0.55 6.05 reinforced s:ructure
Ref. (9)
El Centro, 4-8-68, IPIO PST (Record IA 19) Ailuvium to Instrument on the
about 5000 ft first floor of a
NS 0. I-12 10.49 3.68 1.84 2 story massive
EW 0.058 4.72 4.68 4,71 concrete, heavily
Vert ;ca I 0.036 K16 1.36 14.06 reinforced structure
ýj
Ref. (9)

H•Olywood Storage Base.ment, 7-21-52, 0453 POT (Record IA 6) 700' * of alluvium Instrument in the
basement of a
NS 0.059 2.58 1.41 4.83 14 story RC building
EW 0-046 3,74 2.73 3.47
Vert ca 1 1, 0323 1.12 0.85 6.02
Ref . (8)
Hollywood Storage PE Lot, 7-21-52, W453 POT (Record IA 7) 700' of alluvium

NS 0.063 2.60 1.26 4.54

Ew 0. 043 4.11 2.89 2.84
Vertic,'l 0.023 1.22 C 81 4.818
Ref._(8)
Son Franc;Sco ý,Oldern Gate Park, 3-22-5?, 14'4 PST (Record IA 15) Siliceous sandstone Instrument in a
1-2 miles fr sm ImalI
ahick used to
fi 10 F 0. JOS 1.0I 0.20 6j*o San Andreas Fault house electrical
I. 6( 17?7 1,26 0.18 5.56 equ Iprent
Vertical a, 051 0.41 0,16 1 . 26
________
____ ___
________ _______Ref. ~)_ _ _ _
 Ma itI Mi aed
Record qr--.jed -veI 5 ; te
Srour.d acc. qr~.iund dst~1. RIe4 ri, s
descript ion V , ;n/sec 2 descrip,;On
At 9 d. r~ V (5e e . 1,

Ferndale, 10-7-51, 2011 PST fRecerd IA 21 40-80 ft of Instrument on the

Alluvium over ground floor of a
N 46 w 0.120 2..86 0.95 5.39 100 ft of 7 story fr;ime
S 44 w 0.121 3'3 1.07 16.99 sardstone o,,vr structure
vertical 0.032 'C2 1 l,2 1•7.4 si4t7sto4
,r" .... • *" f. (9)

Ferndale, 12-21- 1,156 •ST (Record 1A 9) 40-80 ft of Instruneit on. the

alluvium over ground flocer of a
N 46 W 0.209 9. 79 4.,2 4.15 100 ft of sandstone 2 sto.sry frane
N 44 E 0.166 41.O 9.O9 2.61 over siltstce structure
Vertical 0.045 3.13 2.49 4,42
Ref. (9)
Eureka, 12-21-54, 1156 PST (Record IA S) 100' sanidstone Pt ;.
In tr-,tent
ýpoorly onusolidated) basempnt of a br;cL.
N 11 W 0.189 5.92 S.45 17.61 over 360 ft cf and stone hu;ldinq
N 79 E 0.271 9.23 3.14 3.86 siltstone over
Vertical 0.110 2.64 2.22 13.54 sandstone
Pef.(9
Hollister, 4-8-61, 2323 PST (Record IA 18) 500 ft of a1luvium Instrument .)n the
over cenozoic rocl first floor of
S 01 W 0.076 3.10 3.03 9.2E water table at 50 ft the public library,
N 89W 0.189 6.45 1.97 3.46 a 2 story structure
Vertical 0.056 1.7-3 1.03 7.45
Ref. (9)
 72

Table 3.2 Surriary of Average ad/v2 Values

No. of
ite Directi on No.do ad/v2
R~ecords

alluvium &.rock horizontal 2b 5.6

alluvium horizontal 22 5.7
rock horizontal 6 5.4

alluvium & rock, a>0.19 horzontal 20 5.7

alluvium, a>O.I¶j horizontal 14 5.9
rock, a>0.19 (same as above) horizontal 6 5.4
alluvium, a<O.lg horizontal B53

alluvium G rock vertical 14 10.7

ai•hvium & ror.k vertical 13 9.1
alluvium vertical I 10.0
alluvlu, vertical IC 7.9
rock verticAl 3.

alluvium & rock, C-0.OSg vertical 12.4

alluvium, a>0.059 vertlcl S 12.0
alluvium, a•0.OS-9 vrticol 4 7.3
rock, &>O.0S9 (same as above) vertical 3 13.0
alluvium, Z0.0Sg vertical 6 9.4

Not including the one extrerr value, C1 Centro, 5-8-0 2037 PST,
vertical corponqnt ad/v 2 - 30,58
 Table 3.3 Ratios of .ax;'cr. Ground Motions

Record v/a v/a v/a - vertical a - vertical

descrfptlon horizontal vertical v/a - horizontal a - horizontal

Pacoima 0am
2-9-71, 0600 PST 17.99 35.21 32.12 1.78 0.91 0.57 0.58
(Record IC 041)

Castaic
2-9-71, 0600 F:T 20.21 37.54 15.19 0.75 0.40 0.54 0.64
(Record I0 056)

Holiday inn (First Floor)

2-9-71, 0600 PST 47.02 70.65 72.37 1.53 1.02 0.69 1.29
(Recor%' IC M48)

15250 Ventura Boulevard (Basement)

2-9-71, 0600 PST 46.84 51.17 44.17 0.94 0.86 0.46 0.70
(Record IH II5)

El Centro
5-18-40, 2037 PST 39.43 52.56 14.11 0.35 0.26 0.80 !.25
(Record 14, 1)

El Centro
2-9-56, 0633 PST 42.22 56.55 46.88 1.11 0.82 0.29 0.44
(Record IA 11)

El Centro
4-8-68, 1830 PST 73.87 81.38 32.22 0.43 p.39 0.25 r..62
(Record IA 19)

Hollyw*od Storage Basement

7-21-52, 0453 PDT 43.73 81.30 48.69 1.11 0.59 0.39 0.50
(Record IA 6)
 Record v/a v/a v/a - vertical a - vertical
description horizontal vertical v/a - horizontal a - horizontal

Hollywood Storage PE Lot

7-21-52, 0453 POT 41.27 95.58 53.04 1.28 0.55 0.37 0.54
(Record IA 7)

San Francisco Golden Gate Park

A 3-22-57, )144 PST 9.92 M028 8.04 0.61 0.78 0.40 0,48
(Record IA 15)

Ferndale
10-7-5) , 2011 PST 14.07 23.83 31.87 2.26 1.33 0.26 0.27
(Record IA 2)

Ferndale
12-21-54, 1156 PST 46.64 84.94 69.56 1.48 0.81 0.22 0.27
(Record IA 9)

Eureka
12-21-54, 1156 PST 31.32 34.06 24.00 0.76 0.70 0.41 0.58
(Record IA 8)

Holl ister
4-J-61. 2323 PST 134.13 U0.79 30.89 0.90 0.75 0.30 0.74
(Rocord IA 1H)
 75

Table 3.4 Suwnary of Average v/a Ratios

No. of v/a
Site Direction or n/s
records (In/sec/g)

alluvium & rock hor I zontal 28 45

alluviumi & rockC hori .zon tal 26 48
al luviumi horizontal 22 52
rock horizontal 6 22
rock hori zontal 4 28

alluvium 1.rock, a>0.19j horizontal 20 39

alluvium & rock, a>0.19g hor zontal 18 42
alluvium, 6>0.l9 hor Izontn l 14 47
alluvium, e<0.lg horl ontal 8 60

alluviun &; rock vert i calI 14 37

al luviutn 1 rock vertical 13 40
al luviumi vertical II 43
vertical 3 18
rock:
vertical 2 24

allviun f.rot..I',
a>0.059 vert I cal a 30
Alluviumi & rock~, aPO.(0Sqj vt i ca)
Vwe.. 7 33
vert Ical 5 37
alluviumr, ;r.'O.O59
vet t!Cal 6 47

Wot imcludiri thr ex~treme ratios) SOn Francisco Golden Ga~te Parkt
3-22-57, 111Th PST.
 76

Table 3.5 Summary of Vertical to Horizontal v/a Ratios

No. of v/a - vertical

S; te records Average v/a - horizontal

alluvium & rock 28 0.92

alluvium 22 0.92

rock 6 0.91

alluvium
n, rock, ah>O.1g. a >0.O5q 15 0.85

alluvium alone, ah.•-Olg, avO.059 9 8,I

alluvium G, rock, Ah<0.19, a <OO5g 13 0.99

Actually alluvium values only since all rock co.orwcritt had peak groond
accelerations > 0.1g (horizontal) and 0.059 (vertical).
 77

Table 3.6 Summary of Vertical to Horizontal Acceleration Ratios

S ,No.
of a - vertical
Site records Average a - horizontal

alluvium & rock 28 0.53

alluvium 22 0.53

rock 6 0.54

alluviun & rock, ah>Q.1g, a v>0.059 15 0.65

alluvium, a h> 0 .I, a v>0.059 1.9 0.72

alluvium 1, rock, ah<0.Ig, av<O.O.5g 13• 0.40

A Actually alluvium values only siince-all rock components had peak ground
acc,4lerations ", O. Ig (horizontal) and 0.059 (vertical).
 78

Taole 3.7 Statistical Values of Response Amplifications

(Normal DIstribution - Horizontal Components)

All Records (28) Records with a > 0.1g (20)

Percentile Damping
D V A D V A

0.5 1.98 2.86 4.00 1.97 2.58 3.67

2.0 1.69 2.23 2.91 1.68 2.06 2.76

50
5.0 1.39 1.74 2.20 1.40 1.66 2.11

30.0 1.13 1.38 1.72 1.15 1.34 1.65

0.5 2.66 3.81 5.U2 2.66 3.41 4.65

2.0 2.23 2.89 3.52 2.224 2.68 3.36

75
5.0 1.80 2.19 2.59 !.8'3 2.10 2.48

10.0 1.43 1.69 1.97 1.47 1.66 1.89

0.5 3.27 4.67 5.95 3.-A8 4.16 5.53

2.0 2.72 3.48 4.06 2.74 3.23 3.90
90
5.0 2.17 2.60 2.93 2.21 2.51 2.82

10.0 1.71 1.98 2.20 1.75 1.94 2.11

0.5 3.64 5.19 6.50 3.65 4.60 6.05

2.0 3.02 3.81. 4.39 3.04 3.57 4.22

95
5.0 2.39 2.84 3.14 2.44 2.75 3.03

10.0 1.87 2.15 2,33 1.•1 2.11 2.24

 79

Table 3.8 Statistical Values of Response Amplifications

(Normal Distribution - Vertical Components)

All Records (14) Records with a > O.05g (8)

Percentile Dampingv
D V A D V A

0.5 1.77 2.74 4.22 1.86 2.52 4.02

2.0 1.57 2.10 2.86 1.65 1.97 2.80

50
5.0 1.33 1.56 2.08 3.40 1.51 2.05

10.0 1,09 1.22 1.62 1.16 1.17 1.59

0.5 2.33 3.67 5.47 2.48 3.39 5.46

2.0 2.04 2.77 3.60 2.17 2.61 3.70
75
5.0 1.70 2.06 2.52 1.81 1.97 2,57

10.0 F.38 1.55 1.91 1.47 1.49 1.92

0.5 2.83 4.51 6.59 3.04 4.17 6.76

2.0 2.46 3.37 4.27 2.63 3.18 4.51

5.0 2.04 2.47 2.92 2.18 2.37 3.04

1ý.ý 1.63 1.84 2.17 1.75 1.78 2.22

0.5 3.13 5.02 7.26 3.37 4.64 7.53

2.0 2.71 3.73 4.67 2.91 3.52 4.99
95
5.0 2.24 2.73 3.16 2.40 2.62 3.32
10.0 1.79 2.01 2.32 1.92 1.95 2.40
 80

Table 3.9 Statistical Values of Response Amplifications

(Ranking - Horizontal Components)

All Records (28) Records with a > O.1g (20)

Percent? le Damping
D V A 0 V A

0.5 1.82 2.63 3.74 1.77 2.34 3.30

2.0 1.54 2.10 2.82 1.50 1.88 2.63

50
5.n 1.32 1.66 2.14 1.30 1.54 2.08

10.0 1.10 1.35 1.71 1.08 1.29 1.64

0.5 2.49 3.50 4.74 2.43 3.17 4.37

2.0 2.11 2.73 3.42 2.06 2.59 3.23

75
5.0 1.74 2.14 2.55 1.75 2.07 2.45

10.0 1.42 1.65 1.96 1.45 1.61 1.89

0.5 3.27 4.64 5.93 3.14 4.07 5.51

2.0 2.69 3.39 4.04 2.68 3.05 3.82

90
5.0 2.10 2.56 2.86 2.11 2 .4 3 2.7!

10.0 1.69 1.94. 2.15 1.68 1.84 2,04

0.5 3.64 5.03 6.43 3.59 4.45 6.10

2.0 3.11 3.78 4.30 3.07 3.47 4.01

95
5.0 2.41 2.82 3.09 2.47 2.73 2.79

10.0 1.85 2.16 2.29 1.89 2.14 2.12

 81

Table 3.10 Statistical Values of Response Anplifications

(Ranking - Vertical Components)

All Records (14) Records with a > 0.059 (8)

Percentile Damping
D V A D V A

0.5 1.65 2.40 3.86 1.54 2.04 3.25

2.0 1.44 1.89 2.63 ).41 1,58 2.37
50
5.0 1.21 1.44 1.96 1,21 1.21 1.84
10.0 0.97 1.14 1.60 1.02 0.98 1.54

0.5 2.15 3.34 5.07 2.07 2.95 4.79

2.0 1.88 2.58 3.36 1.79 2.33 3.25
75
5.0 1.57 1.9' 2.40 1.52 1.82 2.40
10.0 1.28 1.48 1.82 1.27 1.40 1.83

0.5 2.53 4.37 6.14 2.64 3.81 6.06

2.0 2.22 3.31 3.99 2,31 2.94 4.16
90
5.0 1.86 2.41 2.80 1.92 2.24 2.80

10.0 1.48 1.78 2.07 1.54 1.68 2.06

0.5 2.97 4.94 6.96 3.20 4.33 6.98

2.0 2.59 3.66 4.47 2.77 3.29 4.68
95
5.0 2.15 2.63 3.03 2.28 2.43 3.09
10.0 1.72 1.94 2.23 1.83 1.81 2.20
 -----------

500.0
XK i ,1\X1 .2ý
-1 ý.:>/
I K
I/
200.0
<
100.0
\I A 7 . ./\ 4 4 4b7

50.0

"A

z .2~ ~
V.- -' ~.2jI
20.0 \~
, /, N .-. N
"Xi N-)
.
--'4. 2
-. ~
11
N /

-A
"I
10.0
I- ~Ki
5.0
S.-
N,
/

N-
N-

N
N- 2' (UN
N--
N- ' p
.1 '-S

7 *2
>~If
N
L
T N ./~

f7~\
2.0 A5

N K>-
/ 1~ N

1.0 r4 ~
I-,
\ ~/ S /

'>7>
iI
4e x
I
6 -v
,Ž1>
( K) N
V./

N
K 1~

0.,
0.O1 0.02 0.05 a.) 0.5 1.0 12.0 5.0 10.0 20.0 50.0 100.0
MFQUENCY. cp's

F IG 3.1 MEAN FESPONSt P"PL IF ICRT ION - HOR I ZON1'A COMtPONENTS

0.5 PEFRCINT OF CRITICAL ORf'IING
 SOT4. 0 -, 7 -~ I \* ,r
I I III I I

/
I N
A' \ *I
K 'II
>1 j

7 - I
I
77 NIP 1
MoO. 0 K 4-
~

7 \7<s Ar+
I K,
K €¢J
\~/ I\7~~Nq bw,
100.0 • • R• 'S ~
,.===._._..L.

r~K
-, 'r I I

7
l r

.#

SO.0 KY K J
J•
~ ir. 2 * /
~AI / , NN

20.0
I-

-J
10.0
'U

s.0

1.0

0.5
0.01 0.02 0.0s 0.1 . 0.5 1.0 2.0 , 5.0 10.0 20.0 50.0 10.0
FlEUDCY. CPS

FIG 3,2 MEAN RESPONSE R:MPLIFICATION - IORIZONTAL COMPONENTS

2.0 PERCENT OF CRT I1CAL DAWPING
 1 *X I -7 I ~ J X . r
S00.0 ' -

~
I
S
~
NJ
~ II i7T7X
~ I 'N.
AJ '011'

~/I'v. i N r~
j ~ '4
/! / /\

100.0
( K,> I
r
X . . UI1
AV

.I
i
/ \ N' ,.. I'

I' \ NK /
II./ N /iN~ 'i\
50.0
I

N
I

~1
/ y >7
U0,0 /sx )IL Ih~,j' N N.NIX
I'
I
Li
uj 0

O's !
KY ~KA \Ir

10.0
/ le'"

5.0

N
Q
N /
'N.

1.0 /71/f .- '!~~ - . N 2k~

I(j)Kj27<~
n~ I

0.01 0.02 O.0S 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENC. CPS

FIG 3.3 MERN RESPONSE AMPLIFICPTION - HORIIZONTAL COMPONENTS

5.0 PERICENT OF CRITICAL DAMPING
 ...........

Soo.u

7 'OK
/-
.. j./x
* *

; ''4,.
A-'

I
/
- P

200.0 \
•' • -- v w _
I..,
\. /1 \ Z~i~x""NJ
-J
A
-t
A,

-. - I
-
/

\/ -
0/ ' "s 7'>N
K / / /, .. ~/
,1
x
§,.~ y~
1O0.a

50.0
RAA

1.lk ca
10.0

5.0
A-
A
2.0

1.0

0.01 0.02 0.0s 0.1 01? 0.s 1.0 2.0 5.0 10.0 20.0 50.0 100.0
VF9UCT.M' CPS
FIG 3.4 MEAN RESPONSE P$PLJFICR71ON - HORIZONIAL. CO*'ONENTS
10.0. PERCENI~ OF CRITICAL CAWING
 500.0

200.0

100.0

50.0

z
• 10.0
2O.0
x/ >

5.0

/1) 1

2.0

1.0

0. S
0.01 0.02 0.0S 0.1 0.2 0.5 1.0 2.0 S.0 10.0 20.0 50.0 100.0
FIPMEWY'E. CPS
FIG3.5 75 PERCENT]LE RESPONSE AMPLIFICEItON - HORIZONTAL COMPONENTS
0.5 PERCENI .OF CRITICAL DAMPING
 500.0

100.0 //

1O0.0
NN

sQ.0

--"20.0
/x/
0

10.0

5.0

7 --

A,

1.0
'00.7

0.5
0.5 0.01 0.02 0.05 0.) 0.a 0,5 1.0 2.0 5.0 10.0 ?0.0 50.0 100.0
FFEgU(CY. CPS

FIG 3.6 75 PERCENTILE RESPONSE Ft4WLIFICA71ON - HORIZOWJAL COMP'ONENTS

2.0 PERCENT OF CRITICAL. OMtING
 500.0,

200.0

100.0 IN
I'F~ 7
\>< T
joS.

50.0

00,,i

>< Vr xý N
X/,
V,/

J ..

Z.

2,0.

5.0

n rz
. 0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 S.0 10.0 20.0 50.0 100.0
FRE0UEK-r. CP'S
FIG 37 75 PERCENTILE RESPONSE AMPIAFICATION - H8R1ZONTRL COMPONENTS
5.0 PERCENT OF CRITICAL ORMPING
SOD.A

1Go.o!
-, .-
... ,,P AK J,,

5o. 0

2.06

t.0o 0.02 c.0s 0.1 0.2 0.5 1.0 2.0 s.0 10.0 20.0 50.0 100.0
Fi=JENC'. cPS

FIG 38 75 PERCENTILE RESPONSE RMPLIFICRTION - HORIZONTAL COMPONENTS

10.0 PERCENT OF CRTTICPL DiMPFING
 500.0 I ,/ " IN . 7"
\7.- '
~2K i €
,,1
i"

I/I I
> .. ./
/ /.' _.: . ix • ,~ . .: \ .
KK / >~ s\ /
200.0 m 4

I- ,, " ", I : ". • - ' , - ., IA ." ! '< . ",. \

100.0
- /
K' -. i -;."-%- " ".

. lr*x/... . /\ , .o'-
/
/
50. n
I,
4

Li
/ I S

sdJ
(n
- /--- \
z
20.0

4-
-4
>< \.
\~~1 ,-
-I)
>4, \4~/ 'iN.
'N '.0
0
Li I - -. ~----. -' --- p-- 4 ~ I I - I S ~ - I

K
-J N
t'J /
/7 / /
/
*1
N V I
/

5.0
x.
r --

A
I

NX~\ N
/
N-

~3
1'

N \

/'

N E
Ks 7
N.-

'N
• i A .T r
7,

T
K2< T
N..
I N
/
N

>/
/

/'
/
K N.
I'
/
N

A
7

K /
/ Cx 7 K\
1.0 - t - I -~-#d- 4 -%--~.-

7>
4 -~---- I -%--~-* -%--.r- 4 I \
I
I

ti/
K~
,~

//
L.
k... f \. ,
N
2 7 (7 I

0.01 0.02 O.OS 0.1 0.2 0.s 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FvEWo.T. cPS
FIG3.9 MEAN RESPONSE AMPLIFICA71ON - VERTICAL COM¶PONEN7S
0.5 PERCENT OF CRITICAL OPMPING
 500.0 i~ ~/
\I. ,
- - I
4

I
7~~~ I 0 ,~
\ 7
~0t3. 0
*17
~N~- I
'•-0I \J., N- .. N "-... .'
K ~ J
/
~7 .-- I " " 'f \ i •
100.0 I"
x

A
50.0 \
7{\
,
X
'.

2K
Li

- p0.0
7
K "N ,.
i-
/
,..
iii /
b
'-'N
\ r 4-
U

Y
I-' /
i
) •0

U
L / '.-
&
I
,..
~
-J
I0.~
IhJ

s.0

":X--A • "....-
~. 0 ,"(% Io> It.

1.0
X. 00__1\•:

I.S
0.01 0.02 0.05 0.1 0.2 0.S V.a 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS

FIG3JOMEPN RESPONSE RMPLIFICRTION - VERTICAL COMPONENTS

2.0 PERCENI OF CRITICAL D•MPING
 500.0 V NJ I ~V'S.
zr
/ *I /\
K ~'K \ J/ *~4'
200.0 I I

'I

I,

100.0 ~. ,,if> / I
N

-
7 / /
/
50.0
/ ~\. T77'K~~) 4- Ne1

p0.0
2-
/ 5%

'-a / /
'0
t.a
10.0 K f
/ Y I
i

7'
ii.

K
;x.
2•.l 2"%
K"!.
' ,:L/, N
x.
5.0 K)
I
T

,/\ ~/> 2'K~

.5 \
I /\
4." A
/
I- x
//
\ /
}.0 11 I'
-x
/
/
/ 'I
1K' .~>,

~-1
X~~LI I
'A
5'- -~
N
'

.~.L
/
'K
k >~:: ~
I lor,,

I.,
,. ~ J ~v

:4' "I, .5

0.5
.0.01 0.02
,1

0.05 0.1
5%.

0.2
I.
0.5
A
1.0
~rJx ,rIx ( ~-E
-- . . . .. , , dr

5.0 10.0 20.0 50.0 100.0

FRriOUENCY. CPS
FIG 3.11 MEAN RESPONSE PMPLIF!CTION - VERTICAL COMPONENT5
5.0 PEHiCENI OF CfIICAL ORPWING
 500.0 S -

>
I

IKX,1. )K,
'/4
200.0 \/1
~A4-
/ A
Io0. 0
[~7\ xv /1 /'%~
\\~~/f
~( ! ~< p
ii~
..

50.0 ./\ t, •.\ ,/

/
~.
~4/ /IX / A-
)
/ >
2
-~ ~ -
¢/ , ,I/
'/, I , / ~ >1
u /
7. /
"./ ' ". " 1,., "/ /
\/
20.0
U, '0
Nf

10.0
/x

5.0 X,~f 2?~

2.0
N, x '0000

rl •., t,
0-0?1 0.02 0.05 0..1 0.S 1.0 2.0 5.0 10.0 20.0 SO.0 100.0
FF~QEMK. CPS

FIG31 12 MEAN RESPONSE AMPLIFICATION - VERTICAL COMPONENTS

10.0 PER~CENT OF CRITICAL QfA¶PING
 - ..-. -~ ~ EuJU~3j mini I -~

500.0

200.0
xi\K./ *0)

100.0
.' s

so.0
LI
hi

z 20.0
I- tO

10.0 7'x
'hi

5.0

2.0

1.0
N

n %
. 0.0t 0.02 0.05 0.1 0.2 0.5 1.0 ?.Q S.0 10.0 M0.0 50.0 100.0
. ftg4)NCY. CP'S

FIG3.13 75 PERCENTILE RESPONSE W~4LIFICAT ION - VERTICAL COMPONCNIS

0.5 PERCEN7 OF CRITICqL CAMPING
500.0
.,I "X~, N ~ Jr - ."--
- y

f~tLY
~K I

'~I~ ~'j~:o
*1~ i*'
>44 1'
200.0
~ \*

{'*. 1~

7K
r6

V i u~~

~A( )
1' / -1 a /

I
04 'S / /
5.0
/V
.- &~,7
SO. 0 L

5,0 07'

A ll,
SDI

2.0

1.0 * .oz

71

0.5
0.01 0.02 0,3s 0.1 0.5 1.0 2.0 S.0 10.0 20.0 50.0 1 o00.0
FFRQUEENL. CIS

FIG3.14 75 PERCENTILE RESPONSE AMPLIFICATION - VERTICAL COMPONENTS

2.0 PERCENT OF CRITICAL DAMPING
 so0.0

200.0 itt
I00.0

/. IN
50.0

I.,

~420.0

0'

10.0

25.0
9 >

N\>/ IN
1.0

.n cz
0.01 0.02 0. 0ý 0.1 0.2 O.S 1.0 2.0 S.0 10.0 20.0 50.0 10010

FIG3-157S PERCENTILE RESPONSE AMPL[FICATIONb - VEM~ICAL COMPONENTS

5.0 PERCENT OF CRITICAL ORMPJNG

...... •. .. •
SQO. 0

100.0

N' /IN,~*

ao. o

p I l~a
%0

10.0

:p -y/

5.0

2.0

1.0

0O.
(a01 0.02 0.0 0.1 O.S 1.0 2.0 S.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS

F1G3.1675 PERCENTILE RESPONSE ::PLIF[CA7ION - VERTICRL COMPONENTS

10.0 PERCENT OF CRIT[IRL 0 W[ING
 z1N0zN.z0

10
~If

2.0 >

0.! 0.02 0.S 0. 1 C..5 1.0(1 .0 S~.I 10.0ri2.r 50.0 100.0

FREQUENCY. %r.P3

FI.C.3,I8 RESPFONSE $.PECTRR NORMqLJZEC TO 09OUNG RCC1[LFRTIGN FOR EL CE:NT91

FO 1940 - NS - 2.0 PFRCUNT OF CRITICRL ORMFING
 100

4. DESIGN RECOMMENOATIONS AND CONCLUSIONS

4.1 General

On the basis of the information obtained as a result of the

studies sunrnarized in Chapter 3, it is possible to construct design response

spectra.(10) In general, the procedure employed is to select the maximum

values of the ground motions (acceleration, velocity and displaceiftnt) and

to plot these on a four-way log plot of frequency, relative displacement,

pseudo-velocity and pseudo-acceleration. Next, for each degree of damping

and as a function of the percentile ampllfication limit selected for design,

each ground motion bound is multiplied by the appropriate *maptfication

coefficient to obtain the controlling bound in vach of the displacernto,

velocity and acceleration regions. These amplified values are applicable

in certain frequency domains. For extreme low and high frequency ranges

the response is fared back to the maximum ground motion values.

On the basis of the findings in this study, the recomendations

for arriving at the parameters necessary for cotnstructing response spectra

follow.

4.2 Ground Motions

ad/v Values

By way of the review, in Table 3.1, there are presented tfva

values of maximum ground acceleration, velocity and displacerent as

observed from the balanced time-histories for each of the three corPor.ents

of motion of the seismic event studied. In addition, there is recorded

for each component of motion the value of ad/v2 , a nondioensionai quantity

which server as a good Index for monitoring the shape -- especially the

breadth -- of the spectra. The values of ad/v2 range fron 1.84 to 30.58
 101

but as noted in Ref. (6)for most cases of seismic events, the values of

ad/v2 are found to range from about 5 to 15, which is the case also for

this study. In general it will be noted that the ad/v2 values are higher

for the vertical excitation, as would be expected in part because the high

frequency components In the vertical direction are more pronounced than

in horizontal directions.

The procedure that was employed to arrive at the average or

representative ad/v2 values was one of computing the value of ad/v2 for

each individual component of a given record and, taking the mean of these
2
values of ad/v . Another approach which wAs examined was that of averaging

the values of acceleration, velocity and displacement for the horizontal

components and similarly for the vertical components, and then computing

the ad/v2 values from the average values computed. For the horizontal
,motions, the ad/v2 values computed by either of the two methods were

essentially the same, namely falling between values of 5 and 6. In the

case of the vertical components, the ad/v

2 values computed
by the former

method was about 10.7, whereas the value computed by the latter method was

about 7.9. It is believed that the study based on the characteristics of

the individual traces Is much more meaningful than that based on the

characteristics of overall averaging of data without any particular

attention to the trace by trace characteristics.

As may be deduced from the discussion In Appendix A, the maximum

value of the ground displacement Is very sensitive to the adjustment

procedure used since It is obtained through double integration of the

accelerograms. For this reason, In arriving at recommended bounds o.f

maximum ground motions (acceleration, vrlocity and displacement) as

mentioned in the previous chapter, it is felt that, on the basis of the

 102

data in hand and with appreciation of the manner in which they were

obtained, the primary emphasis in arriving at the maximum ground motion

bounds should be placed upon the acceleration and velocity bounds and a
2
representative value of the ad/v . At this point it is desirable to give

some attention to the site characteristics, and especially with respect to

the ad/v2 values. For the most part, the data studied were those obtained

from alluvial sites and studies of the ad/v2 values showed (see Table 3.2)

that for all sites taken together the ad/v2 value for horizontal motions.

was about 5.6; for the alluvial values only (with the three rock sites

removed) it was 5.7; and for the three rock sites it was 5.4. In the case

of the vertical values, for all 14 sites the average was 10.7; for the

alluvial sites excluding three rock sites the value was 10.0; and for the

three rock sites it was 13.0. Obviously, on the basis of the number of

sites studied and the number of traces involved in drawing conclusions,

much more confidence can be placed on the values for the alluvial sites

than for the rock values.

As noted earlier in the report, after the study of all the

data from the 14 records, it was ascertained that the low intensity data

apparently had a significant influence on the amplification values obtained

and possibly on the other characteristics associated with the development

of design spectra. On this basis it was decided to re-examine the data,

casting out all traces which had accelerations less than O.lg in the case

of horizontal motion and O.05g in the case of vertical excitation. The

summary of the ad/v2 data computed on this basis is presented in Table 3.2,

along with the data from the original study. It suffices to say that In

the case of the horizontal excitation, the values of ad/v2 were not changed
 103

markedly by the restructuring of the data. In the case of the vertical

excitation, the restructured analyses did exhibit a slightly wider

variation in values of ad/v2 from 7.3 to 13.0 as a function of site

characteristics, but on the whole were not greatly different than what

had been observed earlier.

In summary then, on the basis of the studies just noted for the
2
strong mot;on data, it would appear that the ad/v values of 6 for the

horizontal direction and 10 for the vertical direction are reasonable

for both rock and alluvial sites. It is believed that the value of 13

for ad/v2 in the vertical direction for the three rock sites studied may

be high in view of the limited rock data; in any event the "rock sample"

is small and a better basis for judgment can only be obtained when more

strong motion rock data can be considered.

v/a Ratios

In order to arrive at a basis for estimating the ratio of the

velocity to the acceleration, the data summarized in Table 3.4 were

studied. It may be observed that the ratio v/a for all data for the

horizontal excitation is about 45 in/sec/g for all sites, 52 for just

the alluvial sites, and about 22 for the rock sites. Again, it is believed

that the limited rock data provide a highly biased value and it would not

be recommended that the design values be based on these, at least for the

present. On the basis of the study of the data In hand, and reflecting

the re-examination of the truly strong motion data as described previously,

a v/a value of about 48 in/sec/g is suggested for alluvium and about 28

for rock. The ratio of v/a for the vertical data to that for the

horizontal data, Table 3.5, is about 0.90 regardless of site.

 104

av/ah Ratios

A study of the ratio of the ground acceleration in the vertical

direction to that in the horizontal direction, Table 3.6, shows a value

of about 0.53 for all the data, and there seems to be little difference

between that for the alluvial sites alone or the rock data alone. However,

for alluvium this ratio increases to 0.65 when earthquake records with a

horizontal acceleration greater than O.1g and a vertical acceleration

greater than 0.05g were considered in computations. On this basis then,

the acceleration ground value for a given site for the vertical spectrum

should be based on an acceleration value of roughly 2/3 that for the

horizontal motion and thereafter the velocity bounds should be taken as

2/3 times 0.9 or 0.6 times that of the corresponding value for the

horizontal case.

4.3 Anplification of Ground Motion

The expected responses to a given ground motion of a single-

degree-of-freedom damped oscillator as a function of frequency, in other

words, the response spectra, may be obtained if the amplification values

as a function of frequency are known. Amplification data are presented in

Tables 3.7-3.10 for various probability percentile values based on normal

distribution calculations and also on ranking of the actual data. It will

be noted that the differences between the normal distribution calculations

and the ranking of data are quite small. A value of 75 percent for ranking

means that 75 percent of the values fall at or below that particular

amplification velue. It is necessary to point out that the amplification

values given were calculated on a frequency-by-frequency basis and varied

 105

considerably as a function of frequency as discussed in Chapter 2. The

frequency band over which the amplifications are presented was selected

from a study of the plotted response spectra plots of Figs. 2.10 through

2.51. Each value given was considered to be constant over the frequency

range used. In particular, the horizontal accelerations were averaged

over a frequency band of 2 to 6 hertz, the horizontal velocity over a

frequency range of 0.4 to 2.0 hertz and the displacements over a frequency

range of 0.2 to 0.4 hertz. In a like manner, for the vertical acceleration

the frequency band over which the amplifications were calculated were 3 to

10 hertz, for the velocity 0.3 to 3 hertz, and for displacement 0.1 to 0.3

hertz.

On the basis of the study of the amplification values and the

spect'ra, it was ascertained that on the high frequency (acceleration) side

of the spectrum, for horizontal motions the drop-off of acceleration beyond

6 hertz down to the maximum ground motion value should occur over a

frequency span that would lead to a faring to ground acceleration at about

40 hertz for 0.5 percent damping, 30 hertz for 2 percent damping, and 20

hertz for 5 and 10 percent damping. It is appreciated that the data in

this high frequency region is not as reliable as might be desired, and

these bounds probably require further study.

In the case of the low frequency spectral domain, i.e., in

the displacement region, the data indicated a slight decrease as the

frequency decreased, as would be expected, because eventually the

displacement spectra must fare back to the ground displacement.

However, so little data exists on which to base decisions here that

for purposes of design at the moment, it is suggested that the constant

 106

displacement bounds be carried to 0.05 hertz; such an approximation will

not be greatly In error based on our studies to date. In any event, the

faring into the ground displacement apparently takes place at a very low

frequency.

In the case of the vertical data, it was found that in the high

frequency (acceleration) region, carrying the drop-off to 50 hertz for all

4 values of damping (0.5, 2, 5 and 10 percent of critical) seemed to be

representative of the data studied. As noted earlier In this report,

the data studied did not extend out to a frequency of 50 hertz, but the

50 hertz faring point for the ground acceleration was an extrapolation of

the trends of the decreasing acceleration amplification as a function of

frequency in the ranges in which the study was conducted.

In the case of the low frequency end of the spectrum for the

vertical spectra, just as in the case for the horizontal spectrao it Is

recommended that the displacement bounds be carried out without decrease

to the 0.05 hertz level.

The design amplification factors were computed for damping values

of 0.5, 2, 5 and 10 percent of critical. Plots of acceleration, velocity

and displacement design amplifications versus percent of critical damping

for both the horizontal and vertical components and for various probability

levels are shown In Figs. 4.1 through 4.6. The design amplification for

any value of damping can be obtained easily from these figures.

4.4 Design Spectra

2
concerning ad/v
On the basis of the summary presentations given

values, end acceleration and velocity values for both horizontal and

vertical :.pectra for sites on alluvium and rock, the ground displacement,
 107

velocity, acce~leration and ad/v2 values are shown in Table 4.1. The
22
displacement values were calculated on the basis of the ad/v2 (6 and 10

for horizontal and vertical effects, respectively), the a and v values,

and were then rounded to a representative number. For purposes of

completeness at this point, the actual ad/v2 value was computed and is

tabulated; It will be seen that for the horizontal case these are nearly

6 and for the vertical case nearly 10.

The next step in arriving at response spectrum bounds is to select

the appropriate amplification value and multiply this by the appropriate

ground motion parameters. The amplification values derived as a part of

this study are presented in Tables 4.2 and 4.3 for horizontal and vertical

components, respectively. The amplification values in Tables 4.2 and 4.3

are the same as those presented in Tables 3.7.and 3.8 for accelerations

greater than 0.1g horizontal and 0.05g vertical. Tables 4.2 and 4.3 also

include amplification factors for mean plus one and plus two standard

deviations based on a normal distribution. The spectrum bounds given in

Tables 4.2 and 4.3 are obtained by multiplying the ground motion values

of Table 4.1 by the appropriate amplification values. The faring frequencies

given in Tables 4.2 and 4.3 are in accordance with those discussed in

Section 4.3. The response spectra based on the tabulated values for 50,

75, and 90 percentile are presented in Figs. 4.7 through 4.18.

4.5 Combined Effects

The studies Indicate that the design spectrum has an equal

probability of occurrence in any horizontal direction, and the records

show that earthquake.motions occur in all three directions simultaneously,

without consistent relations among the motions in the various directions.

 108

Hence it Is recommended that the effects of earthquakes on structures,

components, or elements be computed by taking the square root of the sum

of the squares of the particular effects or responses at a particular

point caused by each of the three components of motion (two horizontal

motions at right angles to one another, and one vertical motion).

4.6 Recommendations for Future Studies

Although the data studied in this project are believed to be

representative, it should be obvious that there was a limited amount of

data available for rock sites. Also, we have indicated the importance

of studying high intensity "strong motion" data and not having these

intermixed with lower intensity data. For these reasons, It would be our

recommendation that there be a further study of a broader spectrum of data,

for both low and high intensities, to define more clearly the ranges In

ground motion bounds and amplification factors that may occur as a function

of the intensity of the ground motion and the site characteristics.

 109

Table 4.1 Horizontal and Vertical Design Ground Motions

2
Site Direction Acceleration Velocity Displacement ad/v
a, g v, in/sec d, In

alluvium horizontal 1.0 48 36 6.04

rock horizontal 1.0 28 12 5.92

alluvium vertical 2/3 29 33 10.10

rock vertical 2/3 17 11 9.80

 110

Table 4.2
Horizontal Design Spectra Amplifications and Bounds
Spectrum bounds Spectrum bounds
Percentile Damping Amplification frequency (alluvium) (rock)
percent 0 V A hrtz
hertz V
InD In/sec A
q V
In0 In/sec A
g

0.5 1.97 2.58 3.67 40 71 124 3.67 24 72 3.67

2.0 1.68 2.06 2.76 30 60 99 2.76 20 58 2.76
50
5.0 1.40 1.66 2.11 20 50 80 2.11 17 46 2.11
10.0 1.15 1.34 1.65 20 41 64 1.65 14 38 1.65
0.5 2.66 3.41 4.65 40 96 164 4.65 32 95 4.65
2.0 2.24 2.68 3.36 30 81 129 3.36 27 75 3.36
5.0 1.83 2.10 2.48 20 66 101 2.48 22 59 2.48
i0.0 1.47 1.66 1.89 20 53 80 1.89 18 46 1.89

0.5 2.99 3.81 5.12 40 108 183 5.12 36 107 5.12

84.1 2.0 2.51 2.93 3.65 30 90 143 3.65 30 83 3.65
(I o) 5.0 2.04 2.32 2.67 20 73 111 2.67 25 65 2.67
10.0 1.62 1.81 2.01 20 58 87 2.01 19 51 2.01
9%.a ." 1Z , 1. IS A 1
.7.4". 'U.

2.0 2.74 3.23 3.90 30 99 155 3.90 33 90 3.90

90
5.0 2.21 2.51 2.82 20 80 120 2.82 27 70 2.82
10.0 1.75 1.94 2.11 20 63 93 2.11 21 54 2.11

0.5 3.65 4.60 6.05 40 131 220 6.05 44 129 6.05

2.0 3.04 3.57 4.22 30 109 171 4.22 36 100 4.22
5.0 2.44 2.75 3.03 20 88 132 3.03 29 77 3.03
10.0 1.91 2.11 2.24 20 69 lo 2.24 23 59 2.24

0.5 4.01 5.04 6.57 40 144 242 6.57 48 141 6.57

3.34 3.89 4.54 30 120 187 4.54 40 109 4.54

2.0
(2 O) 5.0 2.67 2.98 3.23 20 96 143 3.23 32 83 3.23
10.0 2.08 2.28 2.37 20 75 109 2.37 25 64 2.37

Ground motions a, g v, In/sec d, In

alluvium 1.0 48 36
rock 1.0 28 12

I.
 ill

Table 4.3
Vertical Design Spectra Amplifications and Bounds

Faring Spectrum bounds Spectrum bounds

Percentile Damping
percent 0Amp0 fation
VV A frequency
hertz D (alluvium)
Y A 0 (rock)
V A
in in/sec g In In/sec g

0.5 1.86 2.52 4.02 50 61 73 2.68 20 43 2.68

50 2.0 1.65 1.97 2.80 50 54 57 1.87 18 33 1.87

5.0 1.40 1.51 2.05 50 46 44 1.37 15 26 1.37
10.0 1.16 1.17 1.59 50 38 34 1.06 13 20 1.06

0.5 2.48 3.39 5.46 50 82 98 3.64 27 58 3.64

75 2.0 2.17 2.61 3.70 50 72 76 2.47 24 44 2.47
5.0 1.81 1.97 2.57 50 60 57 1.71 20 33 1.71
10.0 1.47 1.49 1.92 50 49 43 1.28 16 25 1.28

0.5 2.78 3.81 6.15 50 92 11o 4.10 31 65 4.10

2.0 2.41 2.91 4.J3 50 80 84 2.75 27 49 2.75

84.1
(I a) 5.0 2.01 2.18 2.82 50 66 63 1.88 22 37 1.88
10.0 1.62 1.64 2.08 50 54 48 1.09 18 28 1.09

0.5 3.04 4.17 6.76 50 100 121 4.51 33 71 4.51

9D 2.0 2.63 3.18 4.51 50 87 92 3.01 29 54 3.01
5.0 2.18 2.37 3.04 50 72 69 2.03 24 40 2.03
10.0 1.75 1.78 2.22 50 58 52 1.48 19 30 1.48

0.5 3.37 4.64 7.53 50 Itl 134 5.02 37 79 5.02

2.0 2.91 3.52 4.99 50 96 102 3.32 32 60 3.32
5.0 2.40 2.62 3.32 50 79 76 2.21 26 45 2.21
10.0 1.92 1.95 2.40 50 63 57 1.60 21 33 1.60
0.5 3.70 5.09 8.29 50 122 147 5.52 41 86 5.52
97.7 2.0 3.18 3.86 5.46 50 105 112 3.64 35 66 3.64
(2 cr) 5.0 2.62 2.85 3.60 50 87 83 2.40 29 48 2.40
10.0 2.09 2.11 2.58 50 69 61 1.72 23 36 1.72

Ground motions a, g v, In/sec d, in

alluvium 2/3 29 33
rock 2/3 17 11
 112

a V I Y I

71

Percent ile
97.7 (2w-)
6 95
C
A02 90
0
84.1 (luo)
75
5 50
AE

3
5
////_____
0 /2//

0 2 4 6 8 I0

Percent of Critical Damping

FIG. 4.1 DESIGN AMPLIFICATIONS

 113

0 C .84.1
75 0I

13

0 2 4 6

Percent of Critical Damping

FIG. 4.2 DESIGN AMPLIFICATIONS

 114
m
SI I V

7 4 1. 6. a

Percentile
97.7 (20")
95
90
0
84.1 (1I')
U 75
*4.-

750
E
4 Zl I 4 ~ I I

£
0
4-

0
0
0
U
U 4F- /Ir
4
0
4-
C
N ~1~
0
N

0 2 4 6 8 I0

Percent of Critical Damping

FIG. 4.3 DESIGN AMPLIFICATIONS

 115

C
Q

C.

M0

A0

Percent of Critical Damping

FIG. 4.4 DESIGN AMPLIFICATIONS

 116

Percentile
97,7 (2a")
/95
90
a 84.1 (to-)
0
-750
55

.2

01

U U I lU

Percent of Criticol Damping

FIG. 4,5 DESIGN AMPLIFICATIONS

 117

Percentile
97.7 (2o-)
6 95
£
90
84.1 (Ia-)
a 75
I.-
50
E
C
0

0
0
0
U
U

0
U
4-

ni
0 2 4 6 8 I0

Percent of Critical Damping

FIG. 4.6 DESIGN AMPLIFICATIONS

 s00.0

200.0 111 - *1z 1

so 0

- 0.0 O

U.p

20.0

10.0

0. Sr

0.01 0.02 0.os 0.1 0.2 o.s 1.0 2.0 S.0 1o.0 20.0 so.0 100.0
FREQUENCY. CPS

FIG. 4.7 DESIGN SPECTRA. HORIZONTAL DIRECTION. RLLUVIUM, 50 PERCENTILE

0.5. 2. S. & 10 PERCENT CRITICAL DAMPING
500.0

200.00
100.0

so. a

20.0
p

2.0

1.0

O.tl1 0.02 0.0S 0.1 0.2 0.5 1.0 2.0 S.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS

FIG.4.8 DESIGN SPECTRA. HORIZONTIAL DIRECTION. HLLUVIUM. 75 PERCENTILE

0.5. 2. S, 4,10 PERCENT CRITICAL DAMPING
 -00.0

20,0 .

0o.0
Ilo,

I ~~, or) 0 4ý0 Y

-20.0

Lip

-o 10.00

S.0

2.0 ll

1.0

0.5
0.01 0.02 0.05 0.1 0.2 o.5 1.0 2.0 s.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS

FIG.4.9 DESIGN SPECTRA. HORIZONTRL DIRECTION. ALLUVIUM. 90 PERCENTILE

0.5. 2. 5. & 10 PERCENT CRITICAL DAMPING
 500.0

I <

L&J
(n

/,, \c\ /\
N.0

20.0

FREUE.Y.aP

1.0

0.5 ýn./X
0.01 0.02 0.05 0.1 0.2 0.S 1.0 2.0 S.0 10.0 20.0 s0.0 100.0
FREQUENCY. CPS

FIG.4.10 DESIGN SPECTRA. HORIZONTAL DIRECTION, ROCK. 50 PRCENTILE

0.5. 2. 5. &.10 PERCENT CRITICAL DAMPING
 500.0

200. 0

100.0

z NON
/X\
so.0
.0
Li
LiJ

z
20.0 /X
10.0
L~J

s.0

2.0 5/X\/
1.0 AV<

0.5
0.01 0.02 0.05 0.1 0.2 0.S 1.0 2.0 5.0 10.0 20.0 SO..0 100.0
FREQUENCY. CPS

FIG.4.1I DESIGN SPECTRFIR. HORIZONTRL DIRECTION. ROCK. 75 PKCENTILE

0.5. 2. 5. t 10 PERCENT CRITICAL ODMPING
 200.0

SO.0 .
FEU.\P

zN
o" 20.0

Li

S.0

2.0

1.0

0.5 L
0.01 0.02 0.Os 0.1 0.2 o.S 1.0 2.0 S.0 10.0 20.0 60.0 100.0
FREQUENCY. CP'S

FIG.412 DESIGN SPECTRR, ?IORIZONTRL DIRECTION. ROCK. 90 FRCENTILE

0.5. 2. 5. 4 10 PERCENT CRITICRL ORMPING
 500.0

I
200.0

I
100.0

SO.0

U
~aJ
C',
2
20.0
a-
U
-j
10.0
Lu

S.O6

2.0

1.0

I
o.s
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FRiEQUENCYr. CPS
FIG.4.13 DESIGN SPECTRA. VERTICAL DIREClION. ALLUVIUM, 50 PERCENTILE
0.5. 2. 5. & 10 PERCENT CRITICAL DAMPING
 500.0

200.0

100.0

so.0

LiJ
20.0

I-
10.0

S.0

0.01
I
0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 2Fl.0 50.0 100.0
FREGUENCY. CP'S
FIG.4.14 DESIGN SPECTRA# VERTICAL DIRECTION. ALLUVIUM. 75 PERCENTILE
0.5. 2, 5, 1 10 PERCENT CRITICRL DAMPING
 SO0.0 7 Y :)K. K x \'/ \
( 1* / \N N
I

'K ,/N
/ /(\ • g"
/
S/ \\
200.0 • N.
xA
/ 0
/
K N / "

>~
' nnl n A
i IUU. U
'7>
K ( 31<- Y. N\ .. '\ ,/ (
*
SO. 0 K--A I
.i•. _ i__,..-..•-.dLL--,• --,dmIm•U
d
I
-\ 7! - /

K
\-° .
/1
*0 /

/~ -\
/ d .~ 4. -4
A / *
\ / X/
z /
20.0 / /
4 >K1
3.-

10.0
/~
N

K 7
/

s~
N
/1 \~ /
/

\/ x T• T
'-a

0'

K½
I I

K;
-J
/
/ /
N /
/
/

N/ N /
9;
/

/
N /
5.0 N ,' /
0 yN
"N /
/
\/x "K /
X
/
/

K /
/ /
'N

7 /~NY7K7i% 7N777~7K
V

1.0 7 /

I/ /
/ / N
/

/ \ /i / / 'I
-..-...- ~ 4 -~-~ I ~ 4 ~ ~ I 4 ~ I

0.01 0.02 0.0S 0.1 0.2 0.5 1.0 2.0 S.0 10.0 20.0 50.0 100.0
FREGUENCY. CPS

FIG.4.15 DESIGN SPECTRA. VERTICAL DIRECTION. ALLUVIUM. 90 PERCENTILE

0.5. 2. 5, 4 10 PERCENT CRITICAL DAMPING
 500.0

200.0

100.0

50.0

Li
(11

z
20.0

I-

Li
-J
10.0
.uJ

0.5
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CP5

FIG. 4.16 DESIGN SPECTRA. VERTICAL DIRECTION. ROCK. 50 PERCRNIILE

0.5. 2. 5. & 10 PERCENT CRITICAL OAMPING
 500. G
I T7 I~7!
,..~ 77V777
200.0 /
I.
V-
I /% V\ ~>~'*

- N

100.0 N~- /
Iv
77/
A y
/'S /

;•. 7"
50.0
o.• -

IN Y. -

20.0 -- V
L&J

N
# A- -

5.0 N
X
/ >0 ,/
\. %.
/

2.0
xN

N / N

1.0
\ / \
/
N
nr L /
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 s0.0 100.0
.FREQUENCY.CP
FEGA417 OES(CN SPECTRFI. VERTICAL O1REr-f-~'lN. ROCK. 7S PERCRNTILE
0.5. 2. S. & 10 PERCENT CRITIUfIL DAMPING
 I~77~71~7TV
so0.0
7~7VT Y1 x ~

4 \ A
7 S..

\\~ ,kii~ jN/

/'

/ / ~b.. >'
200.0 AI

'~0 /
/1

6I 9'
%
x,. NN
,
100.0
~Y
- c~ /
/1' * /
so.0 \/T~ .. L - ~.- -f.-
*N 7
/

2
/ S /
/ / \ ~/,
-~
\V'~ ,/ \ a,,

20.0
LU

K7~ 'a, \~
P
4-
~A. ... /
4
I--'

I0.Q
A
a-
/ \ a,
K
, N,
lix a,
/
a,

.0 N
S.0 K/K
7'
/2K
/~ 0
*8;~ /
L
Y/ N

2.0
\ ~IK , 7'
'.0 \,~/
K a.
,,,I•.
/

- .5

/ < S
a, /
/
'•/
is / 2K
,7
/ / /
a, a.

nr l K
0.01 0.02 0.05 0.1- 0.2 0.S 1.0 -- 2.0 S.0 10.0 20.0 so.0 100.0
FHEQUENCY. CPS

FIG.4.18 OE5IGN 5PECTRA. VERTICAL DIRECTION. ROCK, 90 PERCRNTILE

0.5. 2. S. & 10 PERCENT CRITICAL OAMPING
 130

REFERENCES

I. Amin, M. and A. H.-S. Ang, "A Nonstationary Stochastic Model for

Strong-Motion Earthquakes", Civil Engineering Studies, SRS 306,
University of Illinois, Urbana, Illinois (1966).

2. Blume, J. A.. N. M. Newmark, and L. H. Corning, Design of Multi-Story

Reinforced Concrete Building for Earthquake Motions, Portland Cement
Association, Chicago, Illinois (1961).

3. Veletsos, A. S., N. M. Newmark, and C. V. Chelapati, "Deformation

Spectra for Elastic and Elasto-Plastic Systems Subjected to Ground
Shock and Earthquake Motions", Proceedings of the Third World
Conference on Earthquake Engineering, New Zealand (1965).

4. "Analysis of Strong Motion Earthquake Accelerograms", Volume IlIA -

Response Spectra, California Institute of Technology, EERL 72-80,
Pasadena, California (1972).

5. Housner, G. W. and P. C. Jennings, "Generation of Artificial

Earthquakes", Journal of the Engineering Mechanics Division, ASCE,
Vol. 90, EMI (1964).

6. Newmark, N. M. and E. Rosenblueth, Fundamentals of Earthquake

Engineering, Prentice-Hall, Inc., Englewood Cliffs, New Jersey (1971).

7. Housner, G. W., "Earthquake Ground Motion", ASCE-IABSE International

Conference on Planning and Design of Tall Buildings, Vol. Ib-6,
Bethlehem, Pennsylvania (1972).

8. "Strong-Motion Instrumental nlata on the San Fernando Earthquake of

Feb. 9, 1971", edited by D. E. Hudson, Earthquake Engineering Research
Laboratory, California Institute of Technology, Pasadena, California
(1971).

9. Wiggins, J. H. and W. J. Hall, "Effect of Soft Surficial Layering

on Earthquake Intensity", Civil Engineering Studies, SRS 216,
University of Lllinois, Urbana, Illinois (1961).

10. Newmark, N. M. and W. J. Hall, "Seismic Design Criteria for Nuclear

Reactor Facilities", Fourth World Conference on Earthquake Engineering,
Santiago, Chile (1969).
 131

APPENDIX A

SEGMENTAL BASELINE ADJUSTMENT

A.1 Introduction'

As was mentioned in Chapter 2, the errors in earthquake records

may arise from any number of sources. The errors in computation of velocity

and displacement time-histories which arise from the integration of

a ccelerogram records are caused by record processing where the initial

conditions and the zero acceleration line (baseline) are not known. The

initial conditions are not known since some motion is required to trigger the

mechanism to start the accelerograph. Baselines of recordings not on film

may be slightly in error due to warping of the paper. Although the error

in selecting a baseline for calculational purposes may be very small in

computing the acceleration time-history, the corresponding errors in velocity

and dispidcement may be very large. For example, for a record with duration

of 30 seconds, a constant shift of O.OOlg in the acceleration baseline

results in an error of 11.6 in/sec in terminal velocity and 162 in in the

final displacement. In addition, as was mentioned in Chapter 2, the motion

during an earthquake is a to and fro type motion and a velocity time-history

which is either positive or negative for an extended period is likely to

be in error.

While the initial conditions are usually assumed to be zero

displacement and velocity, various methods have been employed for obtaining

the acceleration baseline. One method which has been used often is

to assume a polynomial, usually a second degree, for the acceleration

baseline, obtain the velocity time-history and use a least squares procedure

to minimize the error in it within the duration of the record. For some
 132

records, applying this procedure within several portions of the record

instead of the total duration of the record further reduces the errors

in the velocity and the displacement time-histories. This procedure is

explained in the following sections.

A.2 Parabolic Baseline Adjustment

in the form(2)
Assuming a parabolic acceleration baseline

VoW = C + C2 (t - to) + C3 (t - to)2 to < t < td (A-l)

where Yo(t) is the acceleration baseline in the interval to and td, and the

C's are unknown coefficients, the corrected (adjusted) acceleration W(t)

is obtained by subtracting from the unadjusted or the "as-read" acceleration

/u(t), the assumed acceleration baseline Yo(t). Therefore, one obtains

- CI - C2 (t - t0 ) - C (t - t)2
Yc(t) = Yu(t) - Yo(t) =u W(t)
or (A-2)
Yc(t) = Yu(t) 3 Cj (t
"jE-I -t)- " l to0 < t :ý t d

The method of least squares is then used to determine unknown coefficients

C.. Thus,
J
t d [yc(t)1 2 dt = 0 for J =, 2, 3 (A-3)
jto

Equation (A-3) gives three simultaneous equations in terms of 3 unknown

coefficients. Once coefficients C. ,'re determined, the corrected velocity

and displacement are obtained by integrating Eq. (A-2). Therefore, one

obtains

'cit = Yu u t(t
0t
- - to) v
0 (A-4)
(t TC
 133

3I
it)2
YcM =yu(t) -. J"
Tj- 7 Ci (t - to)Jil - Vo(t - to) - d (A-S)
j=I

where v0 and d0 are determined from the corrected and uncorrected

velocities and displacements for the initial or starting time to.

vo= (to) - 9c (t0 )

(A-6)
do = YU (to) - Yc(to)

Having the coefficients C1 , C2 , and C3 , the velocity and displacement

baselines are computed froc. Eqs. (A-4 and A-5), respectively. Thus,

3
M 3Z I C( - to)J +v (A-7)

3
YO(t) =F
j=1 7 ITo1 ) Cj(t
j
- 0) + V0 (t - t 0+ d0 (A-8)

A.3 Segmental Adjustment

The interval for which the square of the velocity is minimized for

determining unknown coefficients C., is usually taken as the duration of the

record. Once coefficients C. are evaluated, the shape of the velocity and
J
displacement baselines is determined within the interval of the record.

The expression for the velocity baseline, Eq. (A-7), is a cubic equation.

In cases where the velocity baseline cannot be approximated by a cubic

equation, the minimizing of the square of the velocity within the duration

of the record does not completely reduce the record processing error. To

illustrate this point, the unadjusted velocity time-history for the North-

South component of the E) Centro, 5-18-40, 2037 PST is presented in

" The record was digitized by the Earthquake Engineering Research Laboratory
of the California Institute of Technology.
 134

Fig. A.M. This record was adjusted using a second degree polynomial and

minimizing the error in square of the velocity within the first 30 seconds

of the record. The velocity baseline for the nonsegmental adjustment, Fig. A.l,

underestimates..the correction in the initial portion of the record (0 to

10 seconds), and overestimates the correction for the subsequent portion

(10 to 30 seconds). In order to further reduce the error in the adjustment

procedure, the baseline correction was applied in several segments of the

record. Three segments of approximately 10 seconds each were selected.

The three 10-second intervals were selected on the basis that within the

initial portion, the unadjusted velocity is mainly negative; in the middle

portion, it is both negative and positive; and in the final portion of the

record, it is mainly positive. A procedure similar to that for a nonsegmental

adjustment was used to obtain coefficients C1 , C2 and C3 in Eq. (A-1)

within each segment. For the first segment, both the unadjusted and the

adjusted initial displacements and velocities are assumed to be zero.

However, for the subsequent segments, the unadjusted and adjusted initial

displacements and velocities are taken to be the final unadjusted and

adjusted displacements and velocities of the previous segment. The velocity

and displacement baselines for both the nonsegmental and segmental

adjustments are shown in Figs. A.] and A.2. The adjusted velocity and

displacement obtained by using the two procedures are shown in Figs. A.3

and A.4, respectively.

The unadjusted and adjusted maximum accelerations, velocities

and displacements are presented below:

 135

Maximum Maximum Maximum

acceleration velocity displacement
Case a, g v, in/sec d, in ad/v 2

Unadjusted 0.359 22.59 91.91 24.80

Nonsegmental adjustment 0.355 16.14 24.29 12.79

Segmental adjustment 0.352 13.88 4.74 3.35

The maximum ground motions indicate that the displacement is

most sensitive to adjustment procedure. The quantity ad/v2 which provides

a bound on relative magnitudes of acceleration, velocity, and displacement

is also given in the above table. It is seen that the adjustment procedure
2
.affects ad/v2 to a large degree. The segmental adjustment results in a

more realistic ad/v2 than the nonsegmental adjustment.

The segmental adjustment procedure, while assuring the continuity

of velocities and displacements from one adjustment interval of the record

to another (Figs. A.1 and A.2), introduces a discontinuity in the adjusted

acceleration. The error due to this discontinuity is of high frequency

nature which does not i.Ffect the ground motion response spectra in the range

of frequencies of interest for this study (see Fig. A.5). However, the

error due to the baseline adjustment procedures is more pronounced in the

displacement, causing unusually large response values in the low frequencies.

This is to be expected since the difference between the two baselines in

Figs. A-] and A.2 represents a low frequency error.

A.4 Summary and Conclusions

A segmental baseline adjustment procedure for determining the zero

acceleration line of earthquake records is presented. The procedure is

similar to the well known polynomial baseline adjustment where the square
 136

of the error in the velocity time-history within the duration of the

record is minimized. In the present procedure, different polynomials,

usually a second degree, are assumed for the acceleration baseline in

different portions of the record and the square of the error in the

velocity time-history within each portion is minimized. The following

procedure is recommended when applying segmental baseline adjustment-

a. Obtain the unadjusted velocity and displacement

time-histories by integrating the accelerogram record.

b. From the plot of the unadjusted velocity time-history,

determine the intervals over which the record is to be adjusted. These

intervals can be determined by visual inspection, and by drawing a series

of curves which would roughly represent the velocity baseline. Experience

with a series of strong motion earthquake records has indicated that

generally two or three adjustment intervals are sufficient.

c. Apply the polynomial, usually a second degree, baseline

adjustment within each interval. For the first interval the unadjusted

and the adjusted initial displacements and velocities are assumed to be

zero. For the subsequent intervals, the initial displacements and

velocities are taken to be the final displacements and velocities of the

previous interval. This gives coefficients C.J in Eq. (A-]) for each

interval.

d. Knowing coefficient C., obtain the acceleration baseline,

Eq. (A-]), for each interval. The adjusted acceleration can then be

obtained by subtracting the acceleration baseline from the unadjusted

acceleration, Eq. (A-2).

e. Once the adjusted acceleration is obtained, the adjusted

velocity and displacement time-histories are obtained by proper integration.

 137

REFERENCES

1. Trifunac, M. D., "Low Frequency Digitization Errors and a New Method

for Zero Baseline Correction of Strong Motion Accelerograms",
Earthquake Engineering Research Laboratory, California Institute of
Technology (1970).

2. Amin, M. and A. H.-S. Ang, "A Nonstationary Stochastic Model for

Strong-Motion Earthquakes", Civil Engineering Studies, Structural
Research Series No. 306, University of Illinois, Urbana, Illinois
(l966)

3. Brady, A . G., "Studies of Response to Earthquake Ground Motion",

Earthquake Engineering Research Laboratory, California Institute of
Technology, Pasadena, California (1966).
 8)L

>elcity
-10- c

Boseline For Segmental

-20 Adjustment

--- -Boseline For Nonsegmentol

Adjustment

_3o i iiI I I I i I I1i1 !11

12 is 20 24 28 32
Time, sec.

FIG. AI VELOCITY-TIME HISTORY, EL CENTRO, CALIF., 5/18/1940-NS UNADJUSTED RECORD

 I

CL

Time, sec.
FIG. A-2 DISPLACEMENT - TIME HISTORY, EL CENTRO, CALIF. 5/18/1940 - NS UNADJUSTED
RECORD
 30 i 1 1-1 1 1-4--A-4-4-1 + 30~~-4
4 1 I I I I I I I

20 .4-

101o. .1..
Li
Lfl

z
0
0
Li

4.
0
-J
LLj

-10 -

Segmentol Adjustment
Nonsegmental Adjustment

-20-b
4.
"---U I I I I I I I I I -4---4-i~~
0 4 a 12- 16 20 24 29 32
T'IME. SEC

FIG.A,3VELOCITY - TIME HISTORY. EL CENTRO. CALIF.. 5/18/1940 NS

SEGMENTRLLY AOJUSTED RECORO
 ~30 - ... I t I I I I-+--4--4----I I

SeoentI Adiustment
Nonsegmental Adjustment

10
4-

I- 0
CL.

Cl

-10
I
/
/ 1.
-20
/

-30
0 4 8 12 16 20 24 28 32
7IME. SEC
FIG.AA DISPLACEMENT - TIME? HISTORY. EL
CENTRO. CALIF,. 5/18/1940 - NS
SEGMENTALLY ROJUSIED R~ECORD~
 200.0
I
100.0 I

50.0

20.0

LU.

10.0
z

I-m
"S4
5.0

2.0 -

1.0

0.5 T

0.2O
0.01 0.02 0.05 0.1 0.2 O.S 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS
FIG, A-5 RESPONSE SPECTRA, EL CENTRO, CALIF., 5/18/1940-NS 0 AND 10 PERCENT. CRITICAL
DAMPING
 143

APPENDIX 8

METHODS OF NORMALIZING RESPONSE SPECTRA

81. Intrcduction

The procedure employed for obtaining the amplification factors

used in the statistical study was to normalize the response spectra for each

component of the earthquake and for each damping value to one of the three

ground motion parameters. The amplification factors were obtained by

normalizing the response spectra to maximum ground acceleration for high

frequencies, to maximum ground velocity for intermediate frequencies, and

to maximum'ground displacement for low frequencies. This procedure leads to

a miniml.ation of the standard deviation in each of the three frequency

ranges, as explained in the next section.

82. Differences in Normalization Procedures

When a single response spectra is normalized to one of the three

ground motion quantities, the spectral ordinates at various frequencies are

divided by a constant. The shape of the normalized response spectra is,

therefore, identical to that of the response spectra itself. At any frequency,

the ordinates of the response spectra obtained by the three normalization

procedures are proportional to each other -- the proportionality factor being

the ratio of two of the three ground motion quantities.

The mean response amplification (acceleration, velocity, or

displacement) at each frequency is obtained by first normalizing the individual

response spectra to the appropriate ground motion and then computing the.

mean amplification. Since for different earthquakes the maximum ground

acceleration, velkclty, and displacement and their ratios vary, the three

normalization procedures do not give identical shapes for the mean response
 144

amplification. If the ratio of the ground motion quantities were the same

for each record, the three normalization procedures would give identical

shapes.

The mean value and the mean plus one standard deviation value of

the complete horizontal response spectra for two percent of critical damping

normalized to 1.Og horizontal ground acceleration are shown in Fig. B.I.

Figures 8.2 and B.3 show the complete horizontal spectra normalized to 1.0

in/sec and 1.0 in horizontal ground velocity and displacement, respectively.

The lower curves in these figures are similar to those in Fig. 3.2 with the

exception that in these figures each curve is plotted for the entire range

of frequencies. In each figure the difference between the two curves at.

each frequency is the standard deviation, c, of the amplification factor.

The figures indicate that normalization to maximum ground acceleration gives

a standard deviation which increases rather uniformly from high to low

frequencies, whereas normalization to maximum ground displacement gives a

standard deviation which increases from low to high frequencies.

Normalization to maximum ground velocity results in a standard deviation

which is nearly uniform over the entire range of frequencies. Thus, it

seems reasonable that normalization should be made not to a single ground

motion parameter, but to all three parameters for different ranges of

frequencies. Thus, as may be seen from the figures, for high frequencies,

normalization to maximum ground acceleration gives the smallest standard

deviation while for low frequencies normalization to maximum ground

displacement yields the smallest standard deviation. For intermediate

frequencies, normalization to maximum ground velocity gives a standard

deviation which is nearly constant and it is smaller than those obtained

 145

by normalizing the spectra to either ground acceleration or ground

displacement. Similar results are obtained for normalized spectra for

vertical components, Figs. 8.4 to B.6.

The design amplification factors presented in Chapters 3 and 4

were obtained by normalization to the three ground motion parameters as

just described. If normalization to a single ground motion parameter were

to be used, then the maximum ground velocity would be the best of the

three ground motion parameters, since it gives a standard deviation which

is nearly constant over the entire frequency range.

 oll 1 1.
-gw

S00.0

200.0

100.0 tcy

5o.0 /X

Zn

-\,20.0

>-

ED 10.0

5.0

2.0 Y

1.0 ___

0-. \ /\ n/\ / /Y

0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY, CPS

FIG.B.I MEAN AND MEAN PLUS ONE STANDARD DEVIRAION ACCELERATION AMPLIFICATION
HORIZONTAL COMPONENTS - 2.0 PERCENT OF CRITICRL DAMPING
 50.0

20.0

10.0

5.0

L.J

2.0

D 1.0 271ýrs" X

0..

0.01 0.02 0.0s 0.1 0.2 0.5 1.0 2.0 S.0 10.0 20.0 so.0 100.0
FREQUENCY. CPS

FIG. 8.2 MEAN qNO MEAN PLUS ONE STANODARD EVIATICN VELCCITY RM.LIFICPATIN
HORIZONTAL COMPONENTS - 2.0 PERCENT OF CRITICAL ORMPING
 so. 0

10.0

U,
~" s.0
z0

<>-

o.s

0.2 >

0.1
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQUENCY. CPS
FIG.B.3 MEAN ANO MERN PLUS ONE STANODAO DEVIATION OISPLRCEMENT RMPLIFICATION
HORIZONTAL COMPONENTS- 2.0 PERCENT OF CRITICAL OAMPING
 5oo0.

200.0 5

100.0 -A\

50.0

oo20.0
I-v

-.J

2.01

0.01 0.0? 0.05 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 50.0 100.0
FREQOENCY. CPS
F!G.B.4 MEAN AND MEFh1 PLUS ONE 5TFINOARO DEVIRTION: ACCEILERAT]ON qMLIFICATION
VERTICRL COMPONENt5 - 2.0 PERCENT OF CfRi:.PL OPIRM-NG
s0.0

20.0

2.0

0.2a

0 .05
0.11 0 5 01 02 05 1 20 50 1. 00 5. 0. 0.2

1P; FRQENY

FIGB.5 MEAN RND MEqN PLUS ONE STRNOIR0L OIEVIRTION VELOCITY RMFLIFICRTION
V.RTIJCRL COMPONENTS - 2.0 PER9CENT OF CRITICRL ORMF!NC
 100.0

II
50.0

20.0

10.0

(n
5.0
z

I- I-.'

2.0

1.0

0.5

o °i 0.0
0.01 - 0.02 0.1
0.05 0.2 0.5 !. ( 2.0 5. 0t.0 201.0 50.0 i00.0
FREQUJENCY. U'S
FIG.8.G MEARN AND MEAN FL.US. ONE STANDA~RD DEVIFRTION DISPLACEMENT ~~~O
VERTiCRL COMPONENTS - 2.0 PERCENT OF CR'tTCPL OPm;:INC