highrise building response comparison—use of time domain...

Proceedings of The 2018 IAJC International Conference ISBN 978-1-60643-379-9

Highrise Building Response Comparison—Use of Time Domain Scaling vs. Spectral Matching Input Ground Motion

Mohammad T. Bhuiyan

West Virginia State University [email protected]

Abstract While performing non-linear response history analyses of highrise buildings, designers and researchers discover crucial modeling questions, including use of appropriately selected and scaled ground motion. Once appropriate ground motions are selected, the question becomes how these ground motion time histories can be modified to be compatible with the design target acceleration response spectrum. Modification can be performed in two ways: (a) direct time domain scaling of the acceleration time-histories of the ground motions, and (b) transforming the time-acceleration data into the frequency domain, making adjustments to be compatible with the target spectrum, and transforming back into the time domain. Both the methods are mentioned in Guidelines for Performance-based Seismic Design of Tall Buildings (PEER 2010/05) and FEMA P-1050-1, 2105 edition. ASCE 7-10 mentions the direct scaling approach but does not explicitly mention the other method. The objective of this paper is to determine the extent of differences in response of highrise buildings using both time domain scaled and frequency domain adjusted ground motions. For this purpose, several example structures were selected to be analyzed: (a) 42-story concrete dual core wall-frame structure, (b) 40-story steel space frame structure, and (c) 40-story buckling-restraint-braced frame structure. Detailed non-linear models of these structures were developed in PERFORM-3D, and seven sets of appropriate ground motions were selected for the non-linear time history analyses. Results from analyses show differences in the response of these buildings using time domain scaled and spectral matching input ground motions. 1. Introduction Recent decades have seen a surge in highrise building construction around the world in high seismic areas. While designing those buildings, it is of paramount importance to the designer to select appropriate ground motions and scale those motions for the numerical analysis and evaluation of the design. How to scale appropriately selected ground motion to be compatible with a target spectrum is an important decision for the designer. Two ways to do that are direct time domain scaling of ground motions and transforming the time-acceleration data into the frequency domain, making adjustments (to be compatible with the target spectrum), and transforming back into the time domain. Both methods are mentioned in building guidelines and codes. The objective of this paper is to determine the extent of differences in response of highrise buildings using both time domain scaled and frequency domain adjusted ground motions. 2. Case Study Buildings

ThreestructframeResea2011)build 2.1 42 The bshowresistlinearshow

Figu

2.2 4

e structures wture, (b) 40-e structure. Tarch Center ). The third

dings are prov

2-Story Core

buildings havwn in Figure ting frames ar model was

wn in Figure

ure 1. (a) Ty

di

0-Story Buck

Proceedi

were selectestory bucklinThe first two(PEER) “Tastructure wavided below

e Wall-Spec

ve 42 stories1b. The dualat the perime developed i1b.

(a) ypical plan vmensional re

kling-Restra

ings of The 20ISBN 9

ed for this stung-restraint-

o structures wall Buildingsas used by H

w.

ial Moment

s above groul systems haeter of the buin PERFORM

view at grouendering of

Reprint

aint-Braced (

018 IAJC Inte978-1-60643-

udy: (a) 42-s-braced framwere used bys Initiative” i

Hutt (2013) fo

Frame

und, 4 storiesave a core wauilding as shM-3D (2011

und floor andstructure froed with perm

(BRB) Fram

ernational Co-379-9

story concretme structure, y Pacific Earin their case or his case st

s below grouall and four-hown in Figu1) and a 3D r

d below (Moom PERFORmission.

me Structure

onference

te dual core and (c) 40-srthquake Enstudies (Motudy. Descri

und, and a pe-bay special ure 1a and b.rendering of

(b) oehle et al., 2RM-3D mode

wall-frame story steel spgineering

oehle et al., iptions of tho

enthouse as moment . A detailed f the structur

2011); (b) Thel.

pace

ose

non-re is

hree


The footprint of the above ground structure is 170 ft by 107 ft as shown in Figure 2. It also shows the location of buckling-restrained chevron braces. The building consists of four basement levels as shown in Figure 3a. The footprint of the basement level is 227 ft by 220 ft. Lateral forces were entirely resisted by buckling-restraint braces. PERFORM 3D was used to develop a detailed non-linear model for the numerical analyses in this study.

Figure 2. Typical plan view of the BRB building, above ground (Moehle et al., 2011).

Reprinted with permission. 2.3 40-Story Steel Space Frame Structure The steel space frame structure consists of special moment-resisting frames in both directions. This particular structure has three basement levels and a 120 ft by 80 ft footprint as shown in Figure 3b. Like the other two structures, detailed non-linear model was developed in PERFORM 3D (Figure 3b) for further study.

Figu

(b) 3. Gr 3.1 Sp All thset ofreseaBozoused recordomamotio

(aure 3. (a) Thr) three-dimen

round Motio

Spectral Matc

he case studyf seven pairs

arch team froorgnia, 2008)by the PEER

rded earthquaain to match ons will be c

Proceedi

) ree dimensionsional rend

ons Used in

ching Groun

y buildings as of responseom the PEER). The sameR tall buildinake time histhe target sp

called “frequ


onal renderindering of the

This Study

nd Motions

are located ine spectrum cR Center at Ue spectra andng initiative tories listed pectrum as suency modifi

018 IAJC Inte978-1-60643-

ng of the BRspace frame

y

n Log Angelompatible g

University ofd spectrum co

for analyzinin Table 1 whown in Figied” motions

ernational Co-379-9

(b) RB structure e structure fr

les. Site-speround motiof California,ompatible gr

ng tall concrewere used angure 4. Theses in this pape

onference

from PERFOrom PERFO

cific responsons were pro Berkeley (Mround motioete buildingsnd modified e seven pairser.

ORM-3D moRM-3D mod

se spectra anovided by a Mahin, Yangons have beens. The actualin frequencys of ground

odel; del

nd a

g, & n l y


Figure 4. Site-specific response spectrum and seven pairs of spectrum compatible ground motions.

3.2 Time Domain Scaled Ground Motions Time domain scaling requirements for 3D dynamic analysis are provided in Section 16.1.3.2 of ASCE 7-10:

For each pair of horizontal ground motion components a square root of the sum of the squares (SRSS) spectrum shall be constructed by taking the SRSS of the 5-percent damped response spectra for the scaled components (where an identical scale factor is applied to both components of a pair). Each pair of motions shall be scaled such that for each period in the range from 0.2T to 1.5T, the average of the SRSS spectra from all horizontal component pairs does not fall below the corresponding ordinate of the design response spectrum, determined in accordance with Section 11.4.5 or 11.4.7. (ASCE, 2011)

The problem with these requirements is that no guidance is provided on how to deal with different fundamental periods in the two orthogonal directions. Because an infinite number of sets of scale factors will satisfy the criteria, different engineers are likely to obtain different sets of scale factors for the same ground motions (Soules, 2013). This study uses the two-step scaling method followed in FEMA P-751 (National Institute of Building Sciences, 2012): i) Scale each SRSS’d pair to the average period (Tavg) as shown in Figure 5. This factor will be different for each of SRSS spectra. This scale factor is denoted by S1 in Table 1. Here Tavg is the average of the fundamental periods in each principal direction. ii) As shown in Figure 6, the average of the scaled spectra will match the target spectrum at Tavg. Now a second factor (S2 in Table 1) is applied equally to each motion (already scaled once) such that the scaled average spectrum lies above the target spectrum from 0.2Tavg to 1.5Tavg.

0.000.20

0.400.60

0.801.00

1.201.40

1.601.80

2.00

0 2 4 6 8 10Period (Sec)

Sa

(g

)

Target MCE Sa

Set1_H1

Set2_H1

Set3_H1

Set4_H1

Set5_H1

Set6_H1

Set7_H1


The final scale factor for each motion is the product of the two-scale factors. Detailed calculation steps are provided in Table 1 for the 40-story buckling restraint braced frame structure. Figure 7 provides a comparison of target spectrum and the average SRSS spectrum of 7-pairs of motions after the scaling factor in Table 1 applied to the motions. These seven pairs of ground motions will be called “amplitude scaling” motions in this paper.

Figure 5. Step-1 of time-domain scaling (Soules, 2013).

Figure 6. Step-2 of time-domain scaling (Soules, 2013).

Table 1. 40-story BRB scaling factor.

Record number

Earthquake Name

SRSS Ordinate at

T=Tavg (g)

Target Ordinate

at T=Tavg (g)

S1 S2 SS = S1*S2

Set 1 Superstition Hills-02 0.159 0.143 0.90 1.3 1.166 Set 2 Denali, Alaska 0.063 0.143 2.26 1.3 2.942 Set 3 Northridge-01 (Converter Sta) 0.155 0.143 0.92 1.3 1.193 Set 4 Loma Prieta 0.099 0.143 1.44 1.3 1.877

Set 5 Northridge-01 (Olive View

Med FF) 0.102 0.143 1.39 1.3 1.812

Set 6 Landers 0.116 0.143 1.23 1.3 1.598 Set 7 Kocaeli, Turkey 0.078 0.143 1.81 1.3 2.357

Fig

4. Re As mstudy14 pamodiamplimodimost comp 4.1 42 Figurand f Figurcore wfor stSigniaccelaccelampli

gure 7. Avera

esults

mentioned in y buildings wairs of grounification. Allitude scalingification” mopart, compa

paring the av

2-Story Core

res 8 to 13 sufrequency mo

res 8 and 9 cwall, respecttory shear anificantly highleration for fleration-sensitude scaling

Proceedi

age of SRSSfor the

Section 2, dwere developnd motions, sl of the figurg individual otion responarison betweverage respon

e Wall-Spec

ummarize thodification f

compare the tively. High

nd story momher levels offrequency mositive equipmg motion to e


S spectrum afe buckling re

detailed, threped in PERFseven from ares in this secearthquake mse, and bolden amplitudnse lines.

ial Moment

he analyses rfor the 42-sto

results of stoer demands

ment comparf responses wodification m

ment were plevaluate the

018 IAJC Inte978-1-60643-

fter the applestraint brace

e-dimensionORM-3D so

amplitude scction followmotion respo

d lines represe scaling vs.

Frame

results and coory core wal

ory shear anwere exhibitred to amplitwere observemotion, as calaced in the bbuilding per

ernational Co-379-9

ication of sced frame stru

nal, non-lineoftware. Eaccaling and se

w the same styonse, dottedsent the aver. frequency m

omparison bll-special mo

nd story momted by frequtude scaling,ed for inter-san be seen inbuilding, it wrformance.

onference

cale factors lucture.

ar models foh building w

even from freyle: solid lin

d lines represrage of the mmodification

between ampoment frame

ment for momuency modifi, seen in Figstory drift ann Figures 10will be wiser

listed in Tab

or all three cwas subjectedequency nes representsent “frequen

motions. For n will be don

plitude scaline building.

ment frame aication motiogure 9. nd floor and 11. So,

r to use

le 1

ase d to

t ncy the

ne by

ng

and on

if

Figurrespemodihighe

Fig

am

res 12 and 13ectively. As cification shower responses.

gure 8. Mommplitude sca

Proceedi

3 show the rcan be seen, wed higher r.

(a) ment frame; c

ling; dotted


results of axidifferences

responses, an

comparison o

lines for fre

018 IAJC Inte978-1-60643-

ial strain in cwere not signd at other s

of (a) Story squency mod

response.

ernational Co-379-9

core wall andgnificant; at story levels,

shear; (b) St

dification and

onference

d coupling bsome story lamplitude sc

(b) ory momentd bold lines f

beam rotationlevels, frequcaling showe

t. Solid lines for average

n, uency ed

for of

F

am

F

am

Figure 9. Complitude sca

Figure 10. C

mplitude sca

Proceedi

(a) ore wall; comling; dotted

(a) Comparison ling; dotted


mparison of (

lines for fre

of inter-storlines for fre

018 IAJC Inte978-1-60643-

(a) Story shequency mod

response.

ry drift: (a) Equency mod

response. ernational Co-379-9

ear; (b) Storydification and

EW drift; (b)dification and

onference

(b) y moment. Sd bold lines f

(b) ) NS drift. Sod bold lines f

Solid lines fofor average

olid lines forfor average

r of

r of

F

acce

Fig

Soli

Figure 11. Celeration. So

gure 13. Comid lines for a

Proceedi

(a) Comparison oolid lines for

(a) mparison of amplitude sca


of floor accer amplitude s

bold lines fo

axial strain ialing; dotted

aver

018 IAJC Inte978-1-60643-

eleration: (a)scaling; dottfor average o

in core wall:d lines for frrage of respo

ernational Co-379-9

EW story a

ted lines for fof response.

: (a) at P2 loequency mod

onse.

onference

(b)

acceleration; frequency m

(b) ocation; (b) adification an

(b) NS storymodification a

at P10 locationd bold lines

y and

on. s for

Fi

betw

4.2 4 Figurvs. fr FigurexhibInter-frequshow Signimodiwere build

igure 13. Coween P11-P1

0-Story Buck

res 14 to 16 requency mo

re 14 comparbited by freq-story drift, i

uency modifiwed higher re

ificantly highification motplaced in th

ding perform

Proceedi

(a) omparison of10. Solid line

an

kling-Restra

summarize todification fo

res the resulquency modifin Figure 15ication showesponses.

her level of rtion, as can b

he building, imance.


f coupling bees for amplitnd bold lines

aint-Braced (

the analysis or the 40-sto

lts of story shfication mot, do not show

wed higher re

responses wbe seen in Fiit will be wis

018 IAJC Inte978-1-60643-

eam (CB) rotude scaling;s for average

(BRB) Fram

results and cory (BRB) fra

hear and stortion for storyw a significaesponses, and

were observedigure 16. Agser to use am

ernational Co-379-9

otation: (a) C; dotted linese of respons

me Structure

comparisonsame structur

ry moment. y moment, asant differencd at other sto

d for floor acgain, if accelmplitude scal

onference

(b) CB between Ps for frequene.

s between amre.

Higher dems can be seen

ces; at some ory, levels am

cceleration fleration-sensling motion

P3-P2; (b) Cncy modifica

mplitude scal

mands were n in Figure 1story levels,mplitude sca

for frequencysitive equipmto evaluate t

CB ation

ling

14b. , aling

y ment the

Fsc

F

am

Figure 14. Cocaling; dotted

Figure 15. C

mplitude sca

Proceedi

(a) omparison od lines for fr



f (a) Story srequency mo


018 IAJC Inte978-1-60643-

hear; (b) Sto

odification an


response.

ernational Co-379-9

ory moment.nd bold lines


onference

(b) Solid lines fs for averag


for amplitudge of respons


de se.

r of

F

acce

4.3 4 Figurfigure But mmodiequipto eva


0-Story Stee

res 17 and 18es do not sho

much higher ification motpment were taluate the bu

Proceedi


el Space Fram

8 compare thow a signific

level of resption, as can bto be placed uilding perfo



bold lines fo

me Structure

he results of cant differen

ponses werebe seen in Fiin the build

ormance.

018 IAJC Inte978-1-60643-


e

f story shear,nce.

observed foigure 19. Oning, it will b

ernational Co-379-9

EW story a


, story mome

or floor accelnce again, if be wiser to u

onference

(b) acceleration; frequency m

ent, and inte

leration for faccelerationse amplitude


r-story drift.

frequency n-sensitive e scaling mo

y and

. The

otion

F

sc

F

am

Figure 17. Ccaling; dotted

Figure 18. C

mplitude sca

Proceedi

(a) omparison od lines for fr



of (a) story srequency mo


018 IAJC Inte978-1-60643-

hear; (b) sto

odification an


response.

ernational Co-379-9

ory moment. nd bold lines


onference

(b) Solid lines fs for averag


for amplitudge of respons


de se.

r of

F

acce

5. Co Threeof spematchmotio Whenaccelin a hbuild Refer ASCE


onclusion

e selected buectral matchhing groundons will prov

n spectral mleration werehigh-rise buiding perform

rences

E. (2011). AReston, V

Proceedi


uildings werhing ground md motions yievide compar

atching groue observed. Tlding, it is a

mance.

SCE 7-10, MVA: America



bold lines fo

e analyzed wmotions. Coelds higher drable structur

und motions Therefore, ifdvisable to u

Minimum desan Society of

018 IAJC Inte978-1-60643-


with seven pomparison shdemand. But ral design of

were used, f accelerationuse amplitud

sign loads fof Civil Engin

ernational Co-379-9

EW story a


airs of amplhows that, in

use of time f a high-rise

significantlyn sensitive ede scaled gro

or buildings neers.

onference

(b) acceleration; frequency m

itude scaled n general, use

domain scalbuilding.

y higher leveequipment wound motion

and other st


and seven pe of spectral led ground

els of floor were to be plans to evaluate

tructures.

y and

pairs

aced e the


Hutt, C. M. (2013). Non-linear time history analysis of tall steel moment frame buildings in LS-DYNA. Proceedings of the 9th European LS-DYNA Conference. Salt Lake City, UT: ARUP.

Mahin, S., Yang, T., & Bozorgnia, Y. (2008). Personal communication. Moehle, J., Bozorgnia, Y., Jayaram, N., Jones, P., Rahnama, M., Shome, N., Tuna, Z. …, &

Zareian, F. (2011, July). Case studies of the seismic performance of tall buildings designed by alternative means. Task 12: Report for the tall buildings initiative. Berkeley, CA: Pacific Earthquake Engineering Research Center.

National Institute of Building Sciences. (2012, September). 2009 NEHRP recommended seismic provisions: Design examples. Retrieved from https://www.fema.gov/media-library-data/1393877415270-d563663961c9f40e88ce3ad673377362/FEMA_P-751.pdf

PERFORM 3D. (2011). Nonlinear analysis and performance assessment of 3D structures. Walnut, CA: Computers and Structures, Inc.

Soules, J. G. (2013, June). Instructional material complementing FEMA P-751, Design examples. Retrieved from https://www.fema.gov/media-library-data/1393890432586-c071850c3050c34da135376baa478a1b/P-752_Unit13.pdf

Biography MOHAMMAD T. BHUIYAN is currently an assistant professor of Civil Engineering at West Virginia State University. He earned his BSc in Civil Engineering from Bangladesh University of Engineering & Technology, Dhaka; an MSc in Earthquake Engineering jointly from Universite Joseph Fourier, France, and ROSE School, Italy; and a PhD in Earthquake Engineering from ROSE School with a joint program at Georgia Tech, Atlanta. His research interests include tall buildings, earthquake engineering, and soil-structure interaction. Dr. Bhuiyan may be reached at [email protected].

highrise building response comparison—use of time domain...

Documents