Selsmlc Analysls of 8raced lan lrregular

SLrucLures uslng Pybrld 1esung

uanlel McCrum

Lunchume lecLure aL Lhe ueparLmenL of Clvll, SLrucLural &
LnvlronmenLal Lnglneerlng, 1rlnlLy College uublln on 30
Lh
november
2011.

CuLllne of resenLauon
! Biief desciipiion of plan iiiegulai siiuciuies and Euiocode S.
! Numeiical modelling of conceniiic biacing.
! Biief desciipiion of lisioiy of lybiid iesiing.
! Ouiline ile ley concepis of ile lybiid iesi meilod.
! Hybiid iesi iesulis of biaced plan iiiegulai siiuciuie.
! Paiameiiic siudy iesulis.

! Conclusions.



2 of 25
lan lrregular SLrucLures
! Plan iiiegulai siiuciuies aie ilose in wlicl seismic iesponse is noi only
iianslaiional bui also ioisional, and is a iesuli of siiness and]oi mass ecceniiiciiy in
ile siiuciuie.

! In-plan siaiic ecceniiiciiy, !
"
is defned as ile disiance beiween ile Ceniie of Mass
(CM) and Ceniie of Resisiance (CR) of ile siiuciuie ai eacl ooi.
! Tle iieaimeni of siiuciuial iiiegulaiiiy in seismic design is laigely based on siudies
of ile iesponse of simplifed single-sioiey models, wlicl weie piefeiied as ileii
iesponse could be easily paiameieiised e.g. (Coienza ei al., (iqq)).

! Pievious ieseaicl inio mulii-sioiey plan iiiegulai siiuciuies las focused mainly on
ile iesponse of ieinfoiced concieie momeni iesisiing fiames (}eong and Elnaslai,
(:oo)), bui some ieseaicl las been undeiialen inio ile seismic iesponse of
iiiegulai mulii-sioiey sieel siiuciuies analysing veiiical mass iiiegulaiiiy (Tiemblay
and Poncei, (:oo)) and modelling in-plan iiiegulaiiiy foi a sieel momeni iesisiing
fiame (Fajfai ei al., (:oo)).
! Liiile expeiimenial validaiion exisis foi plan iiiegulai siiuciuies.
3 of 25
8raced lan lrregular SLrucLures
4 of 25
"#$% &% '()* +#,- ./ 0()* #11,$2()1 31)4,5 67,,(
/1)8, 67124721,
(1)
(2) &
! Siiuciuial paiameieis:
! Siaiic ecceniiiciiy, !
"
.
! Laieial ioisional fiequency iaiio, #
$
.
! Plan aspeci iaiio |lengil of building]
widil of buildingj.
! Biace non-dimensional slendeiness, .
! Level of peal giound acceleiaiion, PGA.
! Bi-diieciion oi uni-diieciional exciiaiion.
! Designed foi ECS accidenial ecceniiiciiy.
!
wleie; %
&
is ile siiness iadius of gyiaiion
and %
'
is ile mass iadius of gyiaiion
wleie; N
y
is ile squasl load, N
(%
is ile elasiic
buclling load, ) is ile Young's Modulus, and f
y
is ile
yield siiess
lan lrregularlLy ln LC8
e
L
x
6 , 0 1+ = !
! Euiocode S laieial puslovei meilod iiansfoims peal giound acceleiaiion inio
laieial puslovei foice.

! Tle Euiocode S accidenial ecceniiciiy piovision iliougl ile ioisional eecis
facioi, , applied io ile seismic aciion aecis is:

wleie; * is ile disiance fiom ile elemeni undei consideiaiion io ile CM of ile
siiuciuie measuied peipendiculaily io ile seismic aciion; and +
!
is ile
disiance beiween ile iwo ouieimosi laieial load iesisiing elemenis.

! Tle ioisional eecis facioi, is ilen muliiplied by ile ielevani design aciion
foi ile given biace seciion iesuliing in a new design aciion foi ile biace
membeis io iesisi.

5 of 25
(3)
numerlcal Modelllng l
! OpenSees used foi numeiical
modelling.
! y No. iniegiaiion poinis along ile
membei.
! io No. sub-divisions along ilin
wall and No. sub-divisions acioss
eacl ilin wall.
! Iniiial oui-of-plane impeifeciion of
o.i% of + (Uiiz 8 Malin (:ooS)).
! Uniaxial sieel maieiial model is
based on ile Giuie-Menegoiio-
Pinio (Menegoiio and Pinio
(iqy)) model wiil isoiiopic siiain
laidening.
! Lineai beam-column elemenis foi
beams, nonlineai beam-column
elemenis foi columns.


"#$% 9% :1)4, 8,83,1 623;5#+#6#.*
)(.*$ #76 (,*$7<


"#$% =% "#3,1 ,(,8,*7 41.66 6,47#.* 623;5#+#6#.*
6 of 25
! Numeiical model validaied againsi :D cyclic iesis foi iioomm long sieel
specimens
numerlcal Modelllng ll
"#$% >% ?.)5 +,1626 5#60()4,8,*7 /.1 9@ 8.5,( +)(#5)7#.* ./ 4A4(#4 7,676 .*
B)C 9DE9DE9%F88 GHGI )*5 B3C >DE>DEF88 GHG 60,4#8,*6BJ.$$#*6 B9DD>CC
7 of 25
(a
)
(b)
! In Fig. ((a) ile iensile loading plase and peal iensile capaciiy is modelled
accuiaiely wiil a :% dieience in ile peal iensile load. Tleie is a iy%
undeiesiimaiion of ile iniiial buclling load. Tle Bauslingei eeci is modelled
accuiaiely.
! In Fig. ((b) ileie is an oveiesiimaiion of ile peal iensile capaciiy and iniiial
buclling load in ile numeiical model of q% and % iespeciively. Tle iesi specimen
failed by oveiall laieial buclling and local buclling.
! Numeiical model was
validaied againsi D
dynamic slale iable
iesis.
! Peal displacemeni eiioi
o f : q % a n d p e a l
acceleiaiion eiioi of
:(%.
numerlcal Modelllng lll
"#$% K% B)C @#60()4,8,*7 +,1626 7#8,I )*5 B3C )44,(,1)7#.* +,1626
7#8, /.1 =@ *28,1#4)( 8.5,( +)(#5)7#.* ./ 6<)L, 7)3(, 7,676
8 of 25
(a
)
(b)
"#$% F% B)C M(,+)7#.*I )*5 B3C 0()* +#,- ./ 7,67 67124721,
BJ.$$#*6 B9DD>CC
(a
)
(b)
8raced lan lrregular SLrucLures



9 of 25
! Exploie dieieni levels of mass disiiibuiion
on ile seismic iesponse.
! Laieial ioisional fiequency iaiio piovides a
means of assessing if ile siiuciuie is
ioisionally sii wlen #
!
> i oi ioisionally
exible wlen #
!
< i.
! Evaluaie ile iniuiiive design pioceduie foi
ioisional eecis using ile ECS accidenial
ecceniiiciiy piovision. Fig. 7. Schematic of braced plan irregular structure
with uneven distribution of lumped masses shown
! Esseniially wani io impiove ile undeisianding of ile seismic iesponse of plan
iiiegulai siiuciuies.
! Expand ieseaicl io mulii-sioiey siiuciuies moie iealisiic siiuciuies.
! Use lybiid iesiing io piovide mucl needed expeiimenial validaiion of ile
iesponse of plan iiiegulai siiuciuies.
WhaL ls Pybrld 1esung
! Tle lybiid iesi meilod is a iecenily developed iesi pioceduie ilai plysically
iesis a ley componeni wlilsi simulianeously numeiically modelling ile
suiiounding siiuciuie.
! Tle concepi was fisi developed in }apan by Haluno ei al. (iq6q) and ilen
advanced by Talanasli ei al. (iqy) in ile foim of ile pseudodynamic meilod.
Sling 8 Malin in ile U.S. lelped io develop ile meilod in ile eaily So's.


"#$% N% G4<,8)7#4 ./ 7<, <A31#5 7,67 8,7<.5 B'<#((#06 B9DDNCC
10 of 25
Why uo 8eal-ume Pybrld 1esung?
Advaniages:
! Hybiid iesiing sysiem is compaiaiively cleap compaied io full-scale dynamic
iesiing.
! Do noi gei scaling issues associaied wiil scaled models in slale iable iesis.
! Can iesi sysiems wiil iaie-dependeni belavioui. Any quasi-siaiic iesi using
slow speed aciuaiois does noi incoipoiaie iaie-dependeni belavioui.
! Abiliiy io peifoim geogiaplically disiiibuied expeiimenis.

Disadvaniages:
! Delay can cause insiabiliiy and ile pioceduie io fail so pioceduie las io iun
veiy ecienily in oidei io woil (ioial iime siep io-:oms).
! Accumulaiion of small eiiois can cause numeiical insiabiliiy.
! Incieased diculiy incoipoiaiing MDOF exciiaiions.

11 of 25
seudodynamlc 1esL MeLhod
! Tle basic piinciple of ile pseudodynamic (PSD) meilod is ile solving of ile
equaiion of moiion using plysically measuied foices 8 displacemenis and
numeiically calculaied displacemenis.

! Equaiion of moiion:
wleie; is ile nodal acceleiaiion vecioi, is ile nodal velociiy vecioi,
is ile numeiical iesioiing foice vecioi, expeiimenial iesioiing foice
vecioi and is ile exieinal exciiaiion foice applied io ile sysiem.

1 1 1 1 1 + + + + +
= + + +
i i i i i
F R R x C x M E N N N ! ! !
1 + i
x! !
1 + i
x!
1 + i
N R
1 + i
F
1 + i
E R
12 of 25
SubsLrucLurlng
! Subsiiuciuiing is a ley io ile
simpliciiy and eciency of ile PSD
iesi meilod.
! Tle siiuciuie is biolen up inio a
p l y s i c a l a n d n u m e i i c a l
subsiiuciuie.
! Typically ile plysical subsiiuciuie
las liglly nonlineai oi iaie
dependeni iesponse and ile iesi of
ile siiuciuie is ilen modelled
numeiically.
! Pievenis iesiing of laige, cosily
siiuciuies and ile need foi scaling.


"#$% O% P.*4,07 ./ 62367124721#*$ )00(#,5 7.
31#5$, B)C ,*7#1, 67124721, B3C 41#7#4)( ,(,8,*7
7,67 )*5 B4C *28,1#4)( 8.5,(%
B3C
B)C
B4C
critical element
bridge deck
reaction frame
actuator
bridge pier
critical element
numerically modelled
13 of 25
Soware Communlcauon - M1S
"#$% &D% G4<,8)7#4 ./ 5)7) 4.882*#4)7#.* #* 6./7 1,)(;
7#8, <A31#5 7,67
14 of 25
osslble Lrrors
! An eveni diiven siiaiegy is implemenied
! Applicable io geogiaplically disiiibuied
lybiid iesis, lowevei if ileie is a delay in
ile sending of ile nexi command iaigei
displacemeni io ile seivo-coniiol due io
excessive iieiaiion in ile numeiical
iniegiaiion ilen ilis communicaiion delay
is deali wiil in ile same mannei as a
communicaiion delay ovei ile inieinei.
"#$% &&% Q472)7.1 5,()A BH.1#24<# ,7 )(% B&OOOCC
15 of 25
"#$% &9% G7)7, 71)*6#7#.* 5#)$1)8
BG7.R)5#*.+#4 ,7 )(% B9DDKCC
! Sysiemaiic eiiois:
! Aciuaioi delay.
! Aciuaioi oveislooi]undeislooi.
! Random eiiois:
! Noise.
! Round-o.
Pybrld 1esL


16 of 25
S)3(, &% ME0,1#8,*7)( 01.$1)88,
! Subjecied io scaled Taiwan (iqS6)
eaiilquale iecoid.
! No biaces iesi, cyclic iesi wiil biaces,
:D scaled elasiic, D scaled elasiic, D
unscaled inelasiic and D scaled
inelasiic iesis.
! :( No. iesis in ioial.
! Siiain gauges ai biace mid-span and
LVDTs ai biace ciossing and aciuaioi
lead plaie.


!"#$
%"&'"#
()*+"&
,&-."
%'/"
0**1
(2&*-3'#"4
%3"54"&5"##6
!
%$-8.
9.."5$&'.'$:6
"
#

;-$"&-3
!2&#'25-3
<&"=> ?-826 @
A

B3-5
C#D".$
?-82
1
2
3
4
30x30x3 SPS
30x30x3 8PS
30x30x3 SPS
30x30x3 8PS
1.68
1.37
1.68
1.37
0.13
0.13
0.13
0.13
0.73
0.73
1.23
1.23
2.0
2.0
2.0
2.0
"#$% &=% S,67 6,7;20
Pybrld 1esL 8esulLs


17 of 25
"#$% &>% @#60()4,8,*7 +,1626 7#8, B)C /2(( 7#8, <#67.1AI
)*5 B3C 4(.6,;20
! D scaled (facioi of :.o) inelasiic iesi foi TEST wiil D

= i.:.
! Aveiage displacemeni eiioi of y.y% ims and aveiage acceleiaiion eiioi of
S.i% ims.

(a
)
(b)
arameLrlc SLudy


18 of 25
"#$% &F% G4<,8)7#4 ./ T*)(
*28,1#4)( 8.5,(
! Li mi i ed compai i son of ley
siiuciuial paiameieis duiing iesi
piogiamme.
! Paiameiiic siudy undeiialen io
exiend ile undeisianding ile
inuence of ley paiameieis on ile
seismic iesponse of biaced plan
iiiegulai siiuciuies.


! Base model siiuciuie wiil a PA
iaiio of :.o, D

= i.o, e
s
= o.o+ 8 z
= i.6S.
! Siiuciuie subjecied io No.
eaiilquales wiil vaiying A]V
iaiios.


Parameter Range Considered
\
e
s

U
0

A rauo
A/v rauo
1.3, 1.37, 1.68, 2.0
0.0#, 0.023#, 0.03#, 0.073#, 0.1#, 0.123#, 0.13#
0.73, 0.873, 1.0, 1.123, 1.23
1.0, 1.3, 2.0, 2.3, 3.0
0.33, 0.69, 0.74, 0.89, 1.0, 1.33, 1.77
S)3(, 9% U,A 0)1)8,7,16 )*5 1)*$, 4.*6#5,1,5
8esulLs - SLauc LccenLrlclLy
19 of 25
"#$% &K% @247#(#7A 5,8)*5 +,1626 67)7#4 ,44,*71#4#7A /.1 GG )*5
"G ./ B)C V
W
XD%YFI )*5 B3C V
W
X&%9F 67124721,6
"#$% &Y% "#167;67.1,A #*7,167.1,A 51#/7 +,1626 67)7#4 ,44,*71#4#7A
/.1 GG )*5 "G ./ B)C V
W
XD%YFI )*5 B3C V
W
X&%9F 67124721,6
(a
)
(b)
(a
)
(b)
! Duciiliiy demand incieases
on SS and ieduces on FS of
siiuciuie wiil incieasing
level of !
"
.
! SS of siiuciuie subjecied io
gieaiei duciiliiy demand.


! Fiisi-sioiey inieisioiey
diifis aie ile same foi SS
and FS of siiuciuie.


uucullLy uemand =
u
/
y

(4)
8esulLs - LaLeral 1orslonal lrequency
20 of 25
"#$% &N% Z.18)(#6,5 5247#(#7A 5,8)*5 +,1626 67)7#4 ,44,*71#4#7A
-#7< +)1A#*$ V
W
/.1 B)C GGI )*5 B3C "G ./ 7<, 67124721,6
"#$% &O% Z.18)(#6,5 T167;67.1,A #*7,167.1,A 51#/7 +,1626 67)7#4
,44,*71#4#7A -#7< +)1A#*$ V
W
/.1 B)C GGI )*5 B3C "G ./ 7<, 67124721,6
(a
)
(b)
(a
)
(b)
! Toisionally exible siiuciuies
lave incieased duciiliiy
d e ma n d c o mp a i e d i o
ioisionally sii siiuciuies.
! SS of siiuciuie subjecied io
gieaiei duciiliiy demand.


! Fiisi-sioiey inieisioiey
diifis aie similai foi SS
and FS of siiuciuie.


8esulLs - LC8 rovlslon
21 of 25
"#$% 9D% Z.18)(#6,5 5247#(#7A 5,8)*5 +,1626 67)7#4 ,44,*71#4#7A
-#7< +)1A#*$ V
W
/.1 GG -#7< B)C MS'I )*5 B3C [S' 67124721,6
"#$% 9&% Z.18)(#6,5 5247#(#7A 5,8)*5 +,1626 67)7#4 ,44,*71#4#7A
-#7< +)1A#*$ V
W
/.1 "G -#7< B)C MS'I )*5 B3C [S' 67124721,6
(a
)
(b)
! ITP (Including Toisional
Piovision) las a duciiliiy
demand on aveiage ieduced
by 6% on ile SS of ile
siiuciuie compaied io ETP
( I n c l u d i n g T o i s i o n a l
Piovision) siiuciuie.


! Duciiliiy demand on FS
of ITP siiuciuie ieduced
by q% o n a v e i a g e
c o mp a i e d i o E T P
siiuciuie.


(a
)
(b)
8esulLs - 8l-dlrecuonal LxclLauon
22 of 25
(a
)
(b)
"#$% 99% Z.18)(#6,5 5247#(#7A 5,8)*5 +,1626 ()7,1)( 7.16#.*)( /1,\2,*4A 1)7#.
/.1 2*#; )*5 3#;5#1,47#.*)( ,E4#7,5 67124721,6 .* B)C GGI )*5 B3C "G ./ 67124721,6
! Compaiison of uni- and bi-diieciional exciied siiuciuies indicaied ilai ile
iesponse was similai foi uni- and bi-diieciional exciied siiuciuie.
! Analysis limiied io PA iaiio of i.o siiuciuie.



8esulLs - CA LecL
23 of 25
"#$% 9=% @247#(#7A 5,8)*5 +,1626 67)7#4 ,44,*71#4#7A /.1 'JQ ./ D%>$] D%F$ ^ D%K$
.* B)C GGI )*5 B3C "G ./ 67124721,6
"#$% 9>% [*7,167.1,A 51#/7 +,1626 67)7#4 ,44,*71#4#7A /.1 'JQ ./ D%>$] D%F$ ^ D%K$
.* B)C GGI )*5 B3C "G ./ 67124721,6
(a
)
(b)
(a
)
(b)
8esulLs - lan AspecL 8auo
24 of 25
"#$% 9F% J1)0< ./ *.18)(#6,5 5247#(#7A 5,8)*5 +,1626 0()* )60,47 1)7#.
/.1 B)C GGI )*5 B3C "G ./ 67124721,6
! Tle PA iaiio was vaiied fiom i.o io .o.
! Tle level of duciiliiy demand is noi signifcanily eecied by ile level of PA
iaiio.


(a
)
(b)
Concluslons
! Hybiid iesiing las been slown io be an eeciive meilod of dynamic iesiing.
! Tle value of siaiic ecceniiiciiy las a signifcani eeci on ile seismic iesponse
of biaced plan iiiegulai siiuciuies.
! Tle laieial ioisional fiequency iaiio piovides an indicaioi foi ile level of
ioisional exibiliiy]siiness of ile siiuciuie and ile value of laieial ioisional
fiequency iaiio las a signifcani impaci on ile seismic iesponse of biaced
plan iiiegulai siiuciuies.
! Tle iesulis indicaie ilai ile Euiocode S piovision foi accidenial ioision eeci
is adequaie even ai liglei levels of accidenial ecceniiiciiy exceeding ile
piovision limiis.
! Tle eeci of bi-diieciional exciiaiion is noi signifcani compaied io uni-
diieciional exciied siiuciuies.
! Tle eeci of ile plan aspeci iaiio on ile seismic iesponse is noi signifcani.
! Recommended foi a clecl io be piovided, e.g. #
!
> i, sucl ilai biaced plan
iiiegulai siiuciuies can noi be ioisionally exible. Paiiiculaily if ile level of
PGA exceeds ile design PGA.



25 of 25