Sie sind auf Seite 1von 6

A R e-dellnition ofthe Stiffness ofR einforced

C oncrete E lem ents and its lm plications in Seism ic D esign

Sum m ary

': #&

f .

f ,,,
Tom Paulay
Prof.Emeritus
University ofCanterbury
Christchurch.New Zealand

Tom Paulay.born in 1923 in Hungary.


received hisengineering education in

HungaryandNew Zealand.-lXemajor
partofhisresearch oq'
er4O yzarsat
the University ofCanterbury addressed
the seismicdesign ofreinforced concrete
structures.He ispastpresidentofthe
lnternationalAssociation ofEarthquake

Zngindering.

ltispostulated thatforthe purposesofseism ic design the ductile behaviourof


lateralforce-resisting structuralcom ponents,elem ents and indeed the entire
building system .can be satisfactorily sim ulated by sim ple bi-linear force-displacementrelationships.This enables the displacementl-elationships between
thesystem and itslateralforce-resisting elem entsataparticularlim itstatetobe
readily evaluated.To thisend som e widely used fallacies.relevantto the transition from elasticto inelasticbehaviour.are exposed.A re-definition ofyeld displacem entsandconsequentlystiffness.allowsm uch morerealisticpredictionsof
the m ostimportantfeature ofseismic response.elementdisplacements.to be
m ade.The conceptsintroduced are rationalyetvery sim ple.Theirapplications
are closely interwoven w ith the designer-sintensions.Thestrategy providesthe
designers with unexpected freedom in the assignment of strengths to lateral
force resisting elem ents.such asfram esorstructura)walls.Contrary to currenl
design practice.whereby a specificglobaldisplacem entductility capacity isprescribed for a particularstructuralclass-the designercan determ ine the acceptable displacem entdem and to beim posed on thesystem .RX isshouldprotectcriticalelem entsagainstexcessivedisplacem entdem ands.

jjyjyodud j
*on

D esign Criteria

A study ofthe assessm entofthe structuralperform anceofexistingbuildings


with earthquake risk triggered inquiriesadressing thelikely responseof
buildings as constructed.rather than
their com pliance w ith a particular

The prim ary purpose ofthisstudy was


to address m eans by Nvhich perform ance criteria. conform ing w ith the
appropriate design lim itstate.m ay be
rationally executed.'
Thesecriteriaare:
-

Earthquake-induced deform ations


should lim it the expected displacem ent ductility dem and on any elem entto itsductility capacity.l
-taimax.
stipulatedin codes.

sz
laxim um m agnitudesofinterstorey
displacem ents.to be expected atlocations rem ote from the centre of
m ass.should notexceed those considered acceptable for buildings.
typically 2-2.5% of the storey
height.

code.A maiorperceivedneed wasthe


estim ation of torsion-induced displacem ents ofelem entsofductilesys-

tems g1.2-3j.In the processseveral

issuesem erged with apparentconflict


w ith ingredientsofourcurrentdesign
pratice. The description of progressively emerging fallacies. firmly entrenched in ourroutine seismicdesign

techniques (4
'1.is the subject of this
presentation.

Peer-reviewedbyinternationalexpertsandacceptedforpublication
byIABSE PublicationsCom m ittee
Paperreceived:August21.2000
Paperaccepted:Nosem ber23.20(-('1

36 Science and Technology

Them otivation forthe introduction of


som e unfam iliar.but not necessarily - Perform ance criteria m ay require
new. principles. relevant to ductile
displacem entsassociated with aspestructural response. was the need to
cificlim itstate.to be lessthan those
em phasize the im portance of earthallow ed by the lim itslisted above.
quake-induced displacelnents. rather
than a particularm ethod ofassigning
strengths to elem ents of a system . Term inology used
ldentification of structuralbehaviour.
rationale and transparency ofa viable - ln the study ofearthquak'e-induced
design strategy. com bined with sim displacem ents of buildings. referplicity of application. were central
ence willbe m ade to the sntcntral
issuesofthism otivation.
5').'5'f?#??
StructuralEngineering International 1/2001

Nvhereaselem entswith excessstrength


willyield a little later.These changes
in the onset of yielding were notexpected to change com ponentductility
mentsofthesl'stem nl?
13*besubiect dem ands significantly enough to be a
to different displacements.Typical causeofconcern.
elenAentsare bentsofductile fralnes Instead ofabstractderivationsofrt
cleorinterconnected w alls in the salne vantrelationships.som e sim ple exam 1:lane.
ples.connnxonly encountered in design
zz
N lateral force-resisting e1em ent practice.willbe used to show the apm ay conlprise severaliom ponents. plicability ofcertain principles.Fig.2(a)
Components will be subjected to showsfourrectangularreinforced conidentical displacenAents. Typical crete canlileq'er w alls of identical
COm 1)OnentS are be1m S OF C()1um nS heigths and widths.The interconnecOr1&a1lS.
tion ofthese w allcom ponents issuch
These term sareil1tlstrated in Fig.l.
thatatany stage ofseisnnic response.
identical displacennents will be in-lposedonallfourelenaents.The1e1gths
Wallcomponents
ofthewallcolnponents.lwi.issuch that
-- - j
therelativesecond I'
nonzentsofareaof
j*. *
l
z
N structuralsystelm connpriseslateralforce-l-esistingelem ents.generally
arranged in tw o ortllogonaldirections.D ue to torsionaleffects.ele-

'

A wallelement

I
I

l .- -x
%

#.

I
A frameelement
,

l
l

.* xx

&

%'
.. .>'#

w.

..
e

'
!

Frame component
s
cv
$/
E

1
%
k
> .*'

.e,, j %
o

! *
t c'
l 2
l rzl

--

j A frameelement
l
!

I
!

m wal
letement
l

l G,.

.m xx

l
-

FiL
.l''
h()llellc/tl/'
lfF'
ta
.;

TraditionalConceptsofthe
Theory ofElasticity

ofthese tw o com ponents willchange.


A departure from the strength assigned in proportion to component
stiffnessimplies thus.as shown in Fig.
2(c).a corresponding decrease or increase.respectively.ofthe com ponent
yield displacem ent.The fallacies resulting from this traditionalusage of
com ponentstiffnessareexam inedsubsequently

A D elinition ofY ield


D eform ations

To allow theT'
er3'convenientuse o1'bilinear m odelling of ductilt
l structural
respense.it is essentialto define the
transition pointfrom linear elastic to
linearpost-yield behaviour.Tw o types
the sections-being proportionalto l).
i- of illterrelated deform ations.such as
are 1.2.4and 8.respectively.R-l
ne total yield cun'ature and l'
ield displacerelative stiffnessofthe elem ent.com m ent.need to be considered.
prising fottr com ponent cantileNrer
walls.is thus 15.The relative componentstrengths-based on traditionally kield Curvature
defined component stiffness- are To i11ustrate an acceptable sim ulation
show n in Fig.2(1)).ltalso show slhat.
of the non-linear m om ent-curvature
according to traditional assun-lplions. relationship.anexam pleofarectanguallelem entsw illconAm ence yielding at lar rzinforced concrete u'all section.
arelatiA-e laleraldisplacementof.
1.
shou 1)i1
7Fig.3.w illbe used.ThetechFig. 24c) show'
s the innplications of nique associated with the anall'ses is
1,
'l5 ofthe totalstrength being redis- wellestabli
.
shed j5).Approximations

tributed froln component(4)to con'


l- (61.mostusefulforseismicdesi
gn purponent(3)withoutreducing the total poses.are lessw ellknown.The strain

strength of the four-com ponent elem ent. The traditional definition of


com ponentstiffness.based on the se'
T'
he requiremcnts for static equilibri- cond m om ent ofarea of the section.
um and deform ation com patibility in suggests that the yield displacem ents
statically indetcrm inate structure. is
wellestablished.These principles are
stillwidely used w hen strength.to latComponentnumber:(6 Xt @)
' (@)
eralforce-resisting ductile elem entsof
Latera/force: r
12% 7
15*:
C.J.
J-.
(#..
asystem are assigned.

patte1'11 associated Nvith the onsdt of


yieldingofthel-einforcem entattheextrem e lension fibre enables the corresponding curvature to beexpressed as:

Brith the introduction of eqtliN'alent


lateralstatic seism ic design forcesand
the acceptance ofductileresponse.the
sam e technique continued to be widely used. lt im plied the notilln that
strength assigned proportionally to
elementstiffness willresult in the sim ultaneousonsetofyieldingin allele-

m ents.

Q!10
@

4k

Forllne purpose ofbi-linearm odelling


ofthe tlexuralstrength-curvaturerelationship at the criticalsection of the
wallcom ponent-itisconvenientto introduce the term <xnom inalyield curA'aturep.subsequently referred to sim ply asyield curvature.by linearextrapolation to the nom inal flexural
strength ofthe section.5.
11
,-thus

i;5

K.

V'
s

l'=lwi: .1

e ?)

2fzh.

VE=1.(K)

(a)InterconnectedWallCompoflenlo
15

15

S
t:?o
e
e)
*

y4.

ur
here 6y.isthe l'ield strain ofthesteel

usedand(1Uisthelengthoftheflexuraltension zoneofthe elasticsection.

Subsequentll'itwas realized that.be.k


c
-- -.9
cause.of expected significant inelas4
%% 5 .. 4
.4D s
tic earthquake-im posed deform ations.
3
3
- -'
som e deviation from strength distribu2
g
1
1
tion according to elasticbehaviour.is
quite acceptable.Thus the practice of
lgm-spfacement (c)Strength-displacement
strenglh re-distribution u'as adapted. (b)Slrer
relationships
relationshipswith so/'ne
strength retfjstzfbfgf/bn
ltw asassum ed tlatforcom pflnentsa
reduction offlexuralstrength willresult in an earlier onset of yielding.
StructuralEngineering International 1/2001

(1)3
'= 6!
''(l-k)la=6.
y.
'((lu)

(1
:
))=(5,
'
1u.
/N1))d))=((5,
1n,
'Nly).
'
(lw1ejp:
(6)./1k.)
.

(2)

w here 5'
1) is the m om ent associated
,
Nvith theflrstl'
ield.i.e..(b,.given by eq.
(1),Frol'
n extensiN'e paraluetric analq'
-

sesg61ithasbeenestabl1shedthatfora
iven cross sectitln. the value of

(51,
:'NI)).
't
.rennainsessentially constant
Science and Tccllnology .
37

2.0
A =0.58

11

1.0

:
B) 1.s
U
p=ttoz/Ag
Q
C!
r
%
M
%
I n

%
*
Q:o s
.

'
2)06

Y
'

1/2,0

y
rw

g=5.0

r. .

)$O.4

( My
I
j
!

'
:
E
2A 0.2

pm;

Component

um=3.97
-

phz-Ua.$5 #

a
.
1

0.02

'2

15

t/.
/.o
111.26
111.59

i?0.

.
twy
1
0

d 8

z 1,0
Q
1
=%
k
.

Jdea/fzed
elasto-plastic
fesponse
ph=4.31

0.04

0.06

0.52 ; 1.O
c5cs

1.5

1=2.

2,0

E
a
m
m%
.!
:
c)
a:

2
j
.

2.5

Relative LateralDisplacem ent

?w/
(:1)Slrain Patlerns

Fig.4:Bi-lill(
://,'idealis
ztltiol?t?f'(.
't
?/??/Jt???t?/?rtlll(l
.
c/(
'/
,
??c/?/forz'e-detbrnlal
'
iollrt
>/t
.
'
/f/t?F?.
s/?!
'
/p.
.

il-respective of the ratio and arrange- ample.the reference yield di


splace- (2)As eq.(4) states.the yield dism entofreinforcem ent.A san exam ple m entatthetop ofa wallcomponenti.
placem ent is independent of the
fordesign purposes.itm al'beassum ed shown i
strength assigned to acom ponent.
n Fig.2(a).when subjected to
g6jthatthenominalyieldcurvatureof lateralstaticforces-w hen based on eq. (3)svhenthelateralforce-resistingela l'ectangularwallsection is:
(3).canbeapproximatedby
em ent.shown in Fig.2((1).issub(!) =v2e ,
/1.
A.
yi= Csi
hti=lzchs
zoe),
vl/'
lwiz'
il/lwi
jectedtoincreasingdisplacements.
the com ponents w ill com m ence
w here C isa coefficientwhich quantiA higherdegreeofprecision in seism ic fies the distribution pattern oflateral
yieldingin a predeterm ined order.
design isnotw arranted.
ln theexam pleshow n.com ponent
design forces.and hwiisthe heightof
(4)willbethefirstandcomponent
The important m essage of eq.(3) is- the w all com ponent. Because. as
(1)the lastone to yield.Atthis
thatthe yield curvatureofareinforced shown in Fig.2(a).the heights of the
stage the fullstrength ofthe fourconcrete section isproportionalto the wr
alls.hai.and thegrade ofthe tlexural
com ponentelem entis developed.
yield strain of the steel used and in- reinforcementused willbe the same
versely proportionalto the length or fora1lwallcom ponents.the bracketed (4) Asacorollary.andcontrarytotradepth ofthe section.
ditional assumptions.depicted in
term in eq.(4) Nvill be a constant.
The
r
ef
or
e.
yi
el
d
di
s
pl
a
cem
ent
s
wi
l
l
be
Fig.2(b).thesimultaneousyielding
The bi-linear m odelling of flexural
i
nver
s
el
y
pr
opor
t
i
onal
t
o
w
al
l
l
engt
hs
.
of
com ponents with different
strength-curvature relationship. withThi
s
s
i
m
pl
e
r
el
at
i
ons
hi
p
c
an
be
c
onvel
e
ngt
hsisnotpossible.
out post yield stiffness. is presented
niently used in design B'
henever relain Fig.3(b).Itis also found thatm oacementcapacityofalatderate axial com pression load on a tivepropertiesofcom ponentsaresuf- (5) Thedispl
eralforceresistingelementislim itw all.com m only encountered in m ulti- ficientto establish.forexample.ductility relationships.Exam pleswillsubseed by thatofthe com ponentwith
storeybuildings.P =0.07fJAg.doesnot quentlyshow therelevanceofthisvery
the sluallest yield displacem ent.
change the yield curvature to any sigi
m
por
t
ant
r
el
at
i
on.
i.e..largestlength.ln thisexam ple
nificance. H owever- the strength of
i
thas been assum ed thatthe disthe sections m ay be significantly inpl
acem entductility capacity ofthe
creased. The im portant relationships ImplicationsoftheRede
fined Yield
wal
l com ponents is l
-taimas = 5.
arerecapitulated in Fig..:
1.wherefc'de- D isplacem ents
Ther
ef
or
e.
t
he
di
s
pl
ac
el
nentofthis
notes the com pression strength ofthe
The
i
deal
i
ze
d
bi
l
i
near
duc
t
i
l
e
behavel
em
ent
at
t
he
ul
t
i
m
at
e
lim itstate
concrete and A gis the grosssectional
i
our
of
a
f
our
com
ponent
l
a
t
er
al
f
or
c
emus
t
be
l
i
mi
t
ed
t
o
t
he
capaci
tl'of
area ofthe wall.
resisting wall elem ent.shown in Fig.
component(4)withAu4= 5 >:0.5 =
2((1).ispresented in Fig.4.The assign2.5displacem entunits.ln Fig.4the
Yield .
ll/r/tzcczzltlzl/
mentofcomponentstrengthsin accoryield displacem ent of com ponent
(1)waschosen asa displacement
O nce the reference yield curvature of dance with traditional practice. as
J(Jp)-was used.A study
unit.Itfollow sthen thattaheducacom ponent.such asacantileverw all. shown in Fig..
tility dem and on components (1).
is established.the corresponding dis- ofthe relationships dem onstrate that:
(2)and(3)Nvillbenon-critical.
placement for a given set of lateral (1) Therelativeyield displacementof
forcescanbereadily established atany
each w allcom ponentis inversely (6)Because the sequence ofyiel
ding
desil-ed levelofthe structure.For exproportionalto itslength.
is set and is independent of the
Science and Technologl'

StructuralEngineering International 1/2001

strength of the com ponents.strength nents.Thestiffnessofeachcom ponent A .


s.used here-is associated with the
to componentsm aybeassigned in any in Fig.4wasclearly defined byeq.(5). uniform translation of:111its elel-nents
arbitrary manner. This provides the The superposition of the response of and the centre ofn'
lass.The definition
designerwilh considerablefreedom of the four com ponents leads to a non- isthusidenticalto thatrepresented by
choice.ltcan be exploited so asto im - lineartransition from theelasticto the eq. (7). ln this definition. displaceprtlve overallstructuralperform ance. post-yield behax'iour of the elem ent. m C1'
1tSdtlC t0 tOfSiOIRalPhCnOm e11a ar0
aswellasto arris'eatmoreeconomical For design purposes this can again notincluded.ln seism icdesign.the insolutions.Som eexamplesare provided be replaced by bi-linear simulation. fluenceoftorsion on the displacem ent
subsequently.
shownb)'dashedlinesinFf
k.4astotal ofthe centre of m ass m ay be considA
response.A ccordingly the stiffness of ered to representunu'
arranted analytinunrecognisedsignificantadvantage j
ca1refinem ents.
taeojt
zyuuylyjsr
Otthis approach is thatitenables the

designerto controlthe I'


naiorsources
ofthe detrim entaleffects of the t01-sionalresponse ofbuildings.These are
eccentricitieswith respecttothecentre
ofthe m ass of the system .lfdesiredthe strength to elem ents of a systen:
mal'be assigned so thatthe centres 01'

strengthandI
-nasswillcoincideg71.
A Re-definition ofStiffness
ln the contextt)fthisstudl'
-stiffnessrelatesthe totallateralforce to a corresponding horizontaldisplacelzlent.

ComponentSti
ffness
Idealised bi-linear strength-displacem ent relationships are presented in

Ft.4.lnthesepossiblepost-yieldstiff-

This in turn allow s the nolninalyield D uctility R elationships


displacelnentofthe elelnentto be deO nc: the noluinalstrengths and yield
fined as
displacenAents.and hence thestiffness.
of com ponents or eleluents have been
A).
c= XVni,
'
Y k,
set.therelationshipsbetween different
The value of the eleluent yield dis- displacennents and displacem entducplacenAents so derived is shown dis- tilities.to be considered in the designtinctly in Fig.4.Because the sinxulta- arereadily established.
neousyielding ofconlponentswith differentyield displacem ents..
y.is not
possible-son'
le com ponents willyield ComponentsofLateralForceresisting E lem ents
atdisplacem ents lessthan that ofthe
clem ent..yo.LargerdisplacenAentw ill Bi-linearresponsesof4-colnponente1be required to develop the nolminal elrentsw ere presented in Fig.4.ltw as
strength ofthose connponentsforwich stated thatthe displacelnent capacity
..X)
i> -X)o.. -lXe non-linear transition ofthtselelnentiscontrolled by thatof
from elastic to plastic behaviour. its critical component (4) with the
shown by the fullline m arked rzlbtalp sm allestyield displacenlent.The sul'
n
in Fig.,
/-i11tlstratesthisfeature.
of com ponent relative stiffness w as

ness have bct


?n ignored.ltis evident
thata realistic approxim ation ofcom - The abo:e exam ples shou the fallacl'
ponentstiffnessis:
of the com nlon asstlm ption.that the
elementand system orglobaldisplacem entductilitq'factoristhesam easthat
specified in codes for appropriately
nom inal strength/q'ield
detailed com ponents.A displacem ent
ductilitq'capacity off ofthe elem ent
depicted in Fig.4.im plies a ductility
dem and oncom ponent(4)of8
.
t.:4= 5 )x:
0.58/
/0.5= 5.8.
T'
henom inalstrength ofthecom ponentisthechoiceofthe designer.lt
mustbe knou'n before stiffnesscan S-qltem Sti
ffness
be
quant
i
f
i
ed.
stiffnessisproportionalto strenvth! ASFig.1Shou's.a building system Nviil
The nominall.ield displacem entof Comprise asetofparallellateralforceacomponent.aspreviouslv defined. l
-esisting elements.Thc strength.yield
depends on the geometry of the displacement and stiffness of each
com ponent and the relevant vield elem entare determined.asdescribed.
strain of the materialused.fbr de- 'IXC Pl-ocedure applicable to elements
sign purposes nom inal yield dis- 13:13'then be used to estim atethe stiffplacem ents..
l;F.are independentof NCSSand 'ield displacementofthe en,
..
tiresystem .coluprising ofanum berof
strength.
elem ents.

ElementSti
ffness
Thestiffnessofa lateralforce-resisting
elem ent.com prisinga num berofcom -

ponents subiected to identical displacenzents.such as shown in Fig.2 (a).


m aybedefined bythesuperposition of
the bi-linea1*response of the com p0StructuralEngineering International 1/2001

The sole purpose ofthe system nom inalyield displacem ent-thisbeinga reference valtle-isto quantify the systeln
displacenlentductility.Therefore.any
convenient value that is com patible
with overallbehaviour under the action oflateralforces.I'
nal'be used.The
nom inal system yield displacem ent.

tbund to be 52k.= 1.72.Equation (7)allowsthe yield displacelnentofthe elem entin 1ig.4 to bedeterm ined as.
1CL.
=
1/1.72 = 0.58 displacelnent tlnits.
'
rherafore.with the ultim ate displacem enllim ited to :
5)<
:0.5 = 2.j units.the
element displacem ent ductilitq' dem and m ustbelim ited to p..z:
1 2.5.
'().58
.
= 4.
31.

ltwasemphasised thatan entirely different assignm ent of strength to the


fourcom ponentsoftheelem entisalso
acceptable.Such a choice will affect
the stiffness and hence yield displacem ent of the elem ent.These m ay requiregreaterrestriction on theacceptable displacem ent ductility dem and
on the elem ent.This is controlled by
thestrength-independentdisplacem ent
capacitl'ofcom ponent(.
1.).

IuateralForce-resisting F/e/pd?lf.
&of
B uilding Svstem s
Sim itarIim itationsapply when thesystenzdisplacenpentductility isto be lim ited to ensure that the displacennent
ductflity capacity of the critical elem en:isnotexceeded.H owever.in the
consideration of the displacenAents of
elel
ments of a systenn.torsionaleffect
needalso to be accounted forScience and Technologx'

A llocation ofStrength

ln the first case the colum n shear


forces and hence rele:ant strenghts.
are assigned in accordance with conN'entional l,
/hS
jproportionality. In the
other case shear forces are m ade
inN'ersell' proportional to colum n
heights.Forthiscasethe bi-linearsim ulation of the fivc colum ns is also

Itw as posttllated that.irrespective of


the way strengths were assigned to
colnponents.the sequence ofyielding
w i11depend only on theyield displacel'
nentsofthecom ponents.A
'spreviously stated.forgiven m aterialproperties
(6.,).these are controlled by the geo- showninFiz.6.Sviththeuseofeq.(7).
the bi-linearsim ulation allowsthe elennetry rather than the strength of
m entyield displacem ents.A)u,.foreach
case.i.e..1.364 and 1.667 displacem ent
units. respectively- to be estim ated.
=
X 1 VE EED EZZ
These are inverselyproportionalto elennentstift
-ness.
h1

The criticalconApollentis colunln (l)


u'ith the snnallest relatiN'e (unit)yield

i U/ #,

(&

TheadN'antagesofarbitrarystrength
allocation to components outNveigh
thedisadq'antage associated Nviththe
somewhat reduced energl'dissipation capacity'of the elelnent.Nole
that post-l'ield stiffness. ineN'itably
present-hasnotbeen considered in
thisconlparison.
'
Fhe exanlples shou'
' t1)at an allo'
wance for a global displaceluent
ductility capacity of 6. Uhich lnaj'
haq'
e been adopted for ductile
fralnes.ur
ould grosslytlnderestinlate
the ductility delnand on the criticltl
col
nponent-i.e.colunl11(1).(Forexaluple p-..1ma.= 6 )<:1-667= 1().>> 6).
The exanaple denxonstrateshow the
elennent displacenlent ductiIity delnand sh0uld be lil
mited to approxiI'
nately 4.ifthe criticalctlnlponentis
to be protected agai!'
1st excessiN'e
displacenlentdennands.A s a corollary.the studl'ofthe exanlple strtlcture revealsthatitisnotptlssible to
tltilize the ductility capacit#'ol'con-lponents Uit11 larger l'icld displacenents. For exanlple l
a--j = (n?'4 =
1.5% <6.Such coltlnlns Izlay thus be
designed and detailed according to
recomm endCttions1
-01-linlitcd ductilit(
h.
'd
-en'
lands.
Existing design procedures- based
on traditionaldefinitions01-conlponentstiffness-do notallow displacen-lent ductility demands on constituent conlponents B'
ith different
geonAetric proprties to be related
to thatinnposed on theelenltlnt.

displacenlent.It is assulned that the


displacem entductility deluand on appropriately detailed colum ns should
be linlited to 6. Therefore. the dis13lacenAentdem and on the elem entat
thecom ponents.ltwas-therefore.con- theultinlate lim itstatenaustbe lilzlited
cluded that. within rati011al lil
'
nits. to the displacenAent iapacity of colstrength to com ponents Inay be as- ulun (1).i.e...
lu< 6 )<t l.()= 6.0.This
signed in anq'arbitrary m ttnner.The inlpliesthen thattheelelnentdisplacecriterion to besatisfied isthatthe total mentductility'del
uands.tlsing the trastrength oftheelenlent.i.e..thesum of ditional and arbitrary assignm ent of
the strengths of ist colnponents.m ust colum n strengths.shoulclbe lim ited to
be n-laintained.Tw o exalnplesare pre- p-ac !
G 6.0/1.364 = 4.4t
'
) and rtsa s;
sented here to sl
now how arbitrary yet 6.().1.667= 3.6().respectiN'
ely'
.
astute.choicesofstrengthallocation to
1'
he n'
lajor pointshighlighted by this
colnponents l'
nay lead to appealing ex
a
m
pl
eare:
structuralsolutions.
The arbitrary distribtltion of the
base shear force. V s. am ong
.4 M r
uctureSupported on C'tp/lzzna.
&
coluluns. relying on equaltlexural
with J/hr/f/:/cH eights
strengths.enablesthesam edetailing
,As Fi
e6'shows. a tw'
o storey rigid
of the verticalreinforcem entin a11
block structtlre is supported on five
five colulnns.
., St
ructureSupported /?J'Columns
coIuluns ofN'ariable heights.The relaThe extrenlely displ-oportionate 4
wi
t
h
L'
nequalD epth
tiN'e heigths ofthe colunnns (l)to (5)shear de1u?1nd.associated u'
ith the
hi.are l.00.1.25.1.50.1.7jand 2.00-reconventional allocation of com po- This exam ple elem ent. presented in
spectively.'
l-1le inertia forctt.Vu.needs
nentstrengths-iselim inated.
Fig.7.issim i1arto thatseen in Ftv.5.
to be transl
uitted by the 5 colunAns
The reduction of the elenlentstiff- H ow ever.in this sym m etricalA'ariant
with identicalcross sections.Because
nessby afactorof1.364/l.667 = 0.82 the colum ns are ofequalheights but
ofidenticalsectionaldim ctnsions. the
is not likell'to affectadversely d)'- haN'e differentdepth.hci.The conNv
enyield curN'
attlres at both ends of a1l
nalnicresponse.
tionalstiffnessproportionalallocation
colulnns.f
-t
)y.B'illbe practically identical.Smalldifferences in the N'alues of
A az
zz1.667
yield displaceluentsresultfronadifferA -1.364
Total
ent axial colupression loads on the
(
e
1
)
1.
O
.
z
z
colunans.Consequentlq'
.the yield displacen-lentsN5'
i1Ibeapproxiluately -.
#i=
.2 0
?Z
Conventional
!
A c=4.40
-

(i)t.
h/'6.i.e..proportionalto thesquare

* 8

otthe colulmn heights.Therelore-irrespectiA'e 01-tlexuralstrength.theyield

Rrzl 0. 6

displacementofcolum IA(5)isapproximatelq'4 tim esthatofcolumn (l).


Com ponent strengths.yield displacem ents and hence stiffness. allow the
non-linear shear force - lateral displacementrelationshipsfortu'o differentelem tlnts-considered aboN'
e.to be
plotted.These are presented in Fi:.6.
Science and Technology

N bitra

VM =& C0

r%

#
lsJ= .
Ijw=a.8g
,

.
.Q
4
(M 0.
y 02
Q

ly#

as
4-.y

67
#/
I.
s3
y=
=2
1'
.:6

k 0
.0

Ayl A
1

..Ay3

u.-t.sn
5 .,
6

Norm alized Iateraldisplacements

StructuralEngineering International 1/2001

A =0.572

E
1E1 L DT (
D Q E1
ED D E1 E1 El Q E1 ED
1.0

'

w-

(D

1.5
.-

Vi

2.0

.*-

o
i 1.0

1'0
v .

q--e-*=

:
?
so 0
> .8

,
,.

cm

=o

0.6

Q 04
'
o
Q
8

of shear strength w ould follow the


principles described in connection
w ith the m odel elem ent slzow n in

ph=5.2

Total

y = /:2 (arbitrarv

yni=j,c
'
a(conventional
)
t
2 & 4

k o'2
0.0

i
1
j

ni .c

.z

I
0.5

1 d 5

J
?

I
(
I
1

; Jta2=6O
i
.

k 3
!

1.5

/faa=4.5

k/ta:=3.0
.
q

2.5

Normalized Iateraldisplacem ents

Fig..
?(a).D etails ofthisare notgiven
here.

The relativedepthsofcolumnsf'l)-(2)
and (3)are 1.()-2.
()and l.5.respective- having q'ield displacements larger that R eferences
13.
7
.Correspondingly the relative yield 0.500 displacem ent units. cannot be
(11RAL-LA'
L'-'
F..
b'ei.
%lllil'Desiz??flll
'77'
)F'
$
.
;/f1lltII
displacenlents.being inversely propor- utilized.
'

Jt(opt),'!.
5'(7t?./'I.
)ltL'ff'/t,Bltiltlillgs.Bullt2ti1'1()!'the

tionalto the depth ofthe sections(eq.


(4))-are 1).1= l.()00
.-.
l).a= ().51
.
)()and
s.
yA= 0.667.respectively.The conven- C onclusions

N Z Stlciet: 1'tlr EClrtl'lquake Engineering.N'ol.


''9. N. t
3.3.Sk?plcn'llaer.pp.17S-l9Fl.1t?t?6.

tionalassignmentof(shear)strengths.

g'
3IPkL-LZ
AN'T -tRcl'/t
als't?f('
f.
'rJt,P?
x)'
b'
l
$i(,lls

in proportion to h?
:. would result in 1.The recognition of the eriteria 0f
these 3colum nshaving to resist4.7%'.
perform ance-based seism ic design
37.48'
?
', and l5.8%' of the total base
necessitates m ore attention to be
shear. V s. respectively. '
The lateral
given torealisticestim atesoflateral
t'
orce-displacem entresponse ofthe 5force-induced structural deform acom p.onent elem ent so designed. is
tions.
show n by the dashed line in Fig 8.
2. lm proN'ed techniques for the estiNlaking use ofthe designer'
s freedom
m ation of yield displacem ents are
in choosing any arbitrary assignm ent
postulated.The latter depend only
of com ponent strengths.it is decided
on m atcrialproperties.such aslim to m ake colum n base shear strengths
iting strains. and the geometry of
proportionalto the square ofthe seccom ponents of elem ents of the
tion depths.This technique leads to a
structure.Fordesign purposesgenw ellbalanced design-whereby the raerally yield displacements may be
tio ofthe verticalreinforcem entisapconsidered to be independent of
proximatell'the same in allcolumns.
the strength assigned to com poAs the full lines in Fig.8 shoNv.the
nentsorelem ents.
strengthsofthethreetypesofcolum ns
would thenco1'respondto8.2% .32.69,() 3. Because the sequcnce ofthe onset
ofyielding ofcom ponentsofaplasand 18.4g'
o respectively. of the total
tic m echanism is independent of
base shear.V n.The bi-linearidealizatheirstrength.wilhnrationallimits.
tion foreachco1um n com ponentsodestrengths m al'be assigned to com signed isalso shown in Fig.8.ltisseen
ponents in an),
'way that suits the
thatthe diffttrence between theoqeerall
responses.based on the conventional
designer'sintentitlns.
and arbitrary assignment of componentstrengths.is negligible.ltshould 4. Re-defined stiffness.relating freell'
chosk?n strengths to strength-indebt
tnoted thatthe yield displacem ents
pendent yield displacem ents. en01'the colum ns are the sam e in b0th
ables2
4m ore realistieassessm entof
cases. A ssum ing again that the disthestiffnessofelem entsorofasysplacem entcapacityofthecolum nsis6.
t11e e1em ent disp.lacem ent capacity is
tem to be m ade.

It
2ti11oft1)u?NZ S()c1et),
'for Earthquake Engi-

!'
t))''.'
-/crA'illBdiltl-illt'
rh.BuiJ??7'.$'@t).?lt?/.9(-1*.b'1??iL'EJ'

nt.ltlring.Nbl.3().N
'k).3-Septen'lbe1*.1717.252-2(?3lt)97
'

'(.
FjP.
AL'Ln
.
4'
N'
.T.'
Jt't
.
'
/ztl
/?J
'
.
5
'
,??()f.
I)lfz'
IilfnB1/jJt
'
J/?7tr
% ('(f'
r
.
'L':,ti/')'
''
I'I.
.
tr.%i()I:.ProcvwcklngsNZ
Sb'st'Q,?l;s a.
.L
St1c1et).'forEa2*thquake Engineering.Tcchnical
(.'t312ference.5N'airakei.1.3p.l1-l11hl.ltl9s.
'

l41PRIES-I'LEN.5,
1.J.N..
.
fj'
lll5tl
?lt/I.
tlIILl(.
'
i(z.
%fFJ
IL
'('?rr/?(jI/(f/..
't' ILtlv//3f.,t.
z?'j/?ft - (.-t??3''liL'lff ll(.,l'b.
b'ecll
flcif/tv?(l!(I/?c/?D'r'
k'..
Anlerican Cilncrtttc In>titute (SP-l577).RtlcentD evelopmcntin Latklrl
ForceTransferi11Bui1J1
'ngs.pp.2-71---'5.1t)95.

(51PARK.R.alpdPAL'LAS.T.Rel
bl/'
f?7'
('
('
f/Cf1ll('r(,:(,.
h'rF-l/t'!I/?'t4.y.Jtlhn'
sN'ile).
'andSons.Neu''York.
pp.76C
.).1975.

(6!PRIESTLE5'
.5l..
l.N..andKIIB'
.
ALSKA'
.$.1.
J.z.l%j36Q-1.
%f)4*/)l'iti/F2/dqllfL'ililb'U-tlJ?fJt'/J)'()/'Jt(>c'?(.111trdIft'f?'.$'i'?'/d(..i'lf2'tlI '$1-t111.%.(Bu11r
.
.'
.Lti.
1'1(Jf tllt2'Z'
N;I
Z
Societ)'for Earthquakrz Engineering.N'01..
31.
No.2.pp.73-85.lgt?7.
'

'
71P/%.t,
-LA5'
.T..
,
4.'
f?
.
)z?/J/t4f.
'
:?
'
.

'
/?.
//k
!T'
t,???t>
??/f3.
3
114.
3:.
)tlscd JltzA'jtrp? .$'lrt/itzfk''
$' tk)?' J?gJ'??t'
tl?'c'tat'
/
f?/')j/Ff'
- l'
C-ol.rkol'e f'
l,?j,J/i'F2zs. Proceedings t)1'tl)e 12th
5A''t)rld Confere1:ce o1)L(arlhquakk2Engin,Jcring.
PaperN0.()()(n2.2t)()().

controlled b),
'thatofcolumns(2)and
(4)-i.e..p-aia
lyi= 6 X 0.5 = 3.0displace- 5.Clearly defined yield displacennents
l
'
nent units. Therefore. the displacem entductility dem and on the 15ve-colum n elem entshotlld be lim ited to p-ac
:
G 6.0/0.572 = 5.2.'
The fulldisplacen'
lentductilitl'capacity ofthe t'
olum ns
structuralEngineering lnternational 122001

ofconzponentsenabledisplacenlent
andtlisplacelnentductility denAands
on the systelm to related to the displacem entductilitl'capacity ofthe
crilicalconlponents.
Science aT
'
ld '
IQc'hu()?('gr

Das könnte Ihnen auch gefallen