Professional Documents
Culture Documents
1
Dept. of Civil Engineering, Univ. of British Columbia, 2002-6250 Applied Science
Lane, Vancouver, BC V6T 1Z4. E-mail: ventura@civil.ubc.ca
2
Dept. of Civil Engineering, Univ. of British Columbia, 2002-6250 Applied Science
Lane, Vancouver, BC V6T 1Z4. E-mail: armin@civil.ubc.ca
3
Dept. of Civil Engineering, Univ. of British Columbia, 2002-6250 Applied Science
Lane, Vancouver, BC V6T 1Z4. E-mail: fairhurstmike@gmail.com
4
TGB Seismic Consultants, 1945 Llewellyn Pl, North Saanich, BC V8L 1G4. E-mail:
gwt@tbgsc.bc.ca
5
Dept. of Civil Engineering, Univ. of British Columbia, 2002-6250 Applied Science
Lane, Vancouver, BC V6T 1Z4. E-mail: finn@civil.ubc.ca
Abstract
This paper summarizes the current state of the province-wide retrofit program
and introduces the performance based methodology that has been used to assess and
retrofit school blocks. Some of the methodology changes that will be implemented in
SRG3 are also introduced. These changes include revisions to the seismic hazard, a
new ground motion selection and scaling procedure, and new retrofit prototypes and
modeling approaches.
© ASCE
ATC & SEI 2015 742
INTRODUCTION
In total the BC MOE has around 1600 provincial public schools, of which
approximately 1000 are in regions with a high seismic risk. Currently 339 of these
buildings have a high seismic risk and are part of the seismic mitigation program
(SMP). Of these 339 schools, 50% have started or finished seismic mitigation (see
Table 1). Of the remaining high-risk blocks, 47% are concrete (shearwalls or non-
ductile frames), 24% are wood, 23% are masonry, 3% are steel construction, and 3%
are partition walls.
The 3rd edition of the Seismic Retrofit Guidelines (SRG3) will incorporate
several modifications based on recent relevant research. First, the seismic hazard will
be revised to match the seismic hazard proposed for the 2015 National Building Code
of Canada (NBCC), which includes major revisions to the seismic demand along the
West Coast of Canada. Demand will be based on a tri-hazard probabilistic approach,
in which the contribution of all three B.C. seismic sources (crustal, subcrustal, and
subduction sources) is considered. In order to facilitate improved selection and
© ASCE
ATC & SEI 2015 743
scaling of ground motion records, conditional spectra (CS) will replace uniform
hazard spectra (UHS) as a target for record selection and scaling in several high-
hazard locations.
In SRG the five main principal building elements are 1) vertical load-bearing
supports (VLS); 2) lateral deformation resisting systems (LDRSs); 3) partition walls
rocking out-of-plane; 4) diaphragms; and 5) connections.
Many older low-rise buildings have heavy non load-bearing partition walls
that are constructed of unreinforced masonry. The life safety concern posed by these
© ASCE
ATC & SEI 2015 744
One of the key concepts of the SRG methodology is that deformations are
used to estimate the damage and corresponding risk level of a structure. While lateral
strength certainly affects the dynamic response of a structure, it is the inelastic
deformation levels that govern the damage induced in a structure and are used to set
decision limits. This is quite different to force-based methods, typical of design
codes, in which pseudo-static lateral forces are applied to the structure in order to
design members. SRG utilizes interstory drift levels to quantify performance of
structures.
Nonlinear Models
© ASCE
ATC & SEI 2015 745
knoowledge of thet elastic an nd inelastic response off the structurre, which alllows for the
modeling of th he post-yiellding behav vior of the structure. T This allows the entire
respponse of th he structure to be cap ptured whenn subjected to significant ground
exccitations. To expedite thee INDA pro ocess, concenntrated plastticity elemennts are used
in the modelin ng of low-rrise school buildings. These elem ments are caalibrated to
phyysical test reesults, wheree available, or other to oother standaards such as ASCE/SEI
41--13 (ASCE/S SEI, 2013).
New
w Prototypee Models for SRG3
© ASCE
ATC & SEI 2015 746
(a) (b)
Fig
gure 2. Wood Shear Waall Modeled
d with Deterrioration Coompared to Test
Ressults: (a) CU
UREE Load ding (Gatto and Uang,
ding and (b)) Near Faultt (NF) Load
200
02)
Seismic Hazarrd
© ASCE
ATC & SEI 2015 747
Gro
ound Motio
on Selection and Scaling
It is bein
ng proposedd to use condditional spec tra (CS) as ttarget spectra for record
seleection and scaling
s in SRG3.
S The three distinnct seismic hhazard sourrces in BC:
crustal, subcru ustal, and subduction,
s have drasttically diffeerent characcteristics in
geoophysical properties (deepths, magniitudes, etc.) and spectraal ordinates and shape.
Beccause of thiss, it was deemmed over-coonservative tto scale recorrds from eacch source to
the same unifo orm hazard spectrum
s (U
UHS). Lowerr scaling facctors and eaasier record
seleection can be introdu uced by developing
d individual CS for eaach source
inddependently, and selectin ng and scalinng records too the properr CS. Additiionally, it is
extrremely unlik kely that a grround motio
on record prooduces specttral accelerattions with a
uniform probab a periods ((say, 2% proobability of exceedance
bility of exceedance at all
at all
a periods),, which makes scaling to a UHS inherently conservativee. Figure 3
illu
ustrates the 2015
2 Victorria, Site Claass C, UHS S compared to a condittional mean
speectrum (CMS S), condition
ned at a periood of 1.0 secconds, for eaach BC seismmic source.
Fig
gure 3. Victo
oria 2% in 50
5 Year UH HS and CMSS for Crustaal, Subcrusttal, and
Sub
bduction Soources Condditioned at 1.0
1 second
From thhis figure, it can be seen that crustal and subducttion records which have
a 1.0 second sp pectral acceeleration witth a low proobability of exceedancee (2% in 50
yeaars) tend to have simillarly high spectral acceelerations att lower perriods, while
speectral accelerrations at hig
gher periods tend to decrrease compaared to the UUHS. This is
typical of lower magnitudee (Mw < 8.0), short distan ance, crustal and subcrusstal records,
whiich tend to have
h high en
nergy in shorrt periods annd less energgy at longer pperiods. On
the other hand d, subductio on records, which typiccally come from large magnitude
eveents (Mw > 8.0)
8 at far disstances, havee more energgy in their loonger periodds. This can
be clearly seen n in the su ubduction CMS,
C whichh has lowerr spectral acccelerations
commpared to th he UHS at shorts period
ds and simillar accelerattions at longger periods.
Thiis means a subduction ev vent that pro
oduces a 2%% in 50 year spectral accceleration at
1.0 second willl likely pro oduce spectrral accelerattions with a similar proobability off
excceedance at longer
l period
ds.
© ASCE
ATC & SEI 2015 748
Selectin
ng and scalinng records to a CS invoolves matchiing the meann spectrum,
but also match hing the vaariance abou ut that meaan. The variiance comes from the
dev
viations of th
he ground motion
m predicction equatioons as well aas the uncerttainty in the
epssilon correlation coefficiients used too develop thhe spectrumm. Because thhe variance
abo
out the mean n spectrum isi accounted d for in the rrecord selecttion, the usee of a CS is
recoommended forf probabiliistic-based methods,
m succh as the SR
RG methodollogy, where
both the mean and
a standard d deviation ofo the structuural responsee are requireed (NEHRP,
20111). Figure 4 illustrates an examplee 10 subducction recordss selected too match the
meaan and variaance of the Victoria,
V Sitee Class C subbduction CS from Figuree 3.
For more informatiion about CMS C and CSS the reader is referred to NEHRP
(20 a b), Baker and Jayraam (2008), and Baker aand Cornell
011), Lin et al. (2013a and
(20
006). For moore details ab
bout the imp plementationn of CS for S
SRG3 the reader should
see the paper by
y Bebamzad deh et al. (20
015).
gure 4. Set of
Fig o 10 Subducction Recorrds Selected d and Scaled
d to Match tthe Mean
and
d Variance ofo the Victo
oria, BC, Sitte Class C S
Subduction CCS
SR
RG3 Tri-hazard Probab
bilistic Dema
and Approaach
Previouus GSC hazaard models (u used in the 22010 NBCC C) combined crustal and
sub
bcrustal hazards prob babilistically
y; the subbduction hhazard wass analyzed
deteerministicallly and checcked separaately. SRG22 applied thhis hazard model and
deteermined riskk by consideering crustall and subcruustal hazardss with a 2% probability
of exceedance
e in
i 50 years.
© ASCE
ATC & SEI 2015 749
The annual rate of drift exceedance for each seismic source is calculated by
multiplying the individual CPDE for each level of shaking by its probability of
occurrence and then summing the contributions from all levels of shaking as follows:
Where dλSa is the annual frequency of ground motions with intensity Sa, which is
directly calculated from the Probabilistic Seismic Hazard Analyses. CPDE is the
conditional probability the drift, dr, exceeds a certain drift limit, Dr (typically the
DDL), at the given intensity, Sa. The total annual rate of drift exceedance is then
calculated by summing up the rates over all three sources of hazards: crustal,
subcrustal, and subduction. The total PDE is estimated using the temporal Poisson
probability model at given time interval, T, as shown below:
n
PDE(dr > Dr ) = 1 − exp − T λi
i =1
Figure 5a illustrates the CPDE for CDL = 4% vs. level of shaking curve for
the W-1 - blocked OSB/plywood shearwall - prototype with factored resistance of
20% of the weight of the structure (%W) and a height of 3m. Fig. 5b shows the
hazard curves (annual rate of exceedance vs. level of shaking) for different seismic
sources for Victoria, on Site Class C. In Figure 5, the 100% level of shaking
corresponds to the ground motion with a 2% probability of exceedance in 50 years at
period of one second.
© ASCE
ATC & SEI 2015 750
(a) (b)
Fig
gure 5. (a) CPDE
C vs. Leevel of Shak
king for thee W-1 Protootype with a Height of
3m
m and Factored Resista ance of 20%%W and (b)) Annual R Rate of Exceeedance vs.
Levvel of Shakiing for eachh Earthquaake Source ffor Victoriaa, Site Classs C (100%
leveel of shakin
ng = 2% in 50
5 year haza ard at periood 1.0 sec).
Fig
gure 6. PDEE of CDL = 4%
4 vs. Facttored Resisttance for th
he W-1 Prottotype with
a Height
H of 3m
m.
© ASCE
ATC & SEI 2015 751
SEIISMIC PER
RFORMAN
NCE ANALY
YZER
The Seiismic Perforrmance Ana alyzer, or A nalyzer for short, is thhe principal
anaalytical tool of
o this methodology. Th he tool proviides the engiineer to instaantly access
a highly
h advannced, peer-reeviewed anaalytical dataabase. This aallows the eexperienced
eng
gineer to co ombine his practical
p kn
nowledge annd judgemennt with overr 9 million
INDDA results to o demine thee risk of his//her particulaar building bblock and deevelop cost-
effeective retrofi
fit solutions.
The Annalyzer perm mits the eng gineer to quuickly analyyze the threee principal
building elemeents that hav ve analyticaally complexx behaviour.. These threee principal
building elemen nts are LDRRSs, walls rocking out-off-plane and ddiaphragms For each off
these three building elemeents, the Ana alyzer perforrms a risk aassessment oor a retrofit
dessign. After making
m the basic
b parameetric selectioons (input daata), the engineer clicks
t Analysiss button and the analysiss results are iinstantly displayed (Figuure 7).
on the
Fig
gure 7. Screeenshot of th
he Seismic Performance
P e Analyzer
© ASCE
ATC & SEI 2015 752
CONCLUSION
This paper introduced the SRG methodology including several of the major
changes that will be adopted by the Seismic Retrofit Guidelines, 3rd Edition (SRG3),
for use in the performance-based seismic assessment and retrofit of BC school
buildings. These changes are aimed to allow SRG3 to continue to provide cost-
effective retrofit solutions and user-friendly guidelines while evolving to incorporate
state-of-the-art knowledge of the seismic hazard in BC.
Three of the main components that will help to reach this goal are the
redefinition of target demands from UHS to CS; the adoption of more cost-efficient,
better performing retrofit prototypes with modern modeling techniques; and the
change to a tri-hazard probabilistic approach to replace the previous deterministic
approach in classifying prototype performance. The use of CS will facilitate ground
motion selection and scaling while being consistent with the hazard demands for each
earthquake source. New prototypes will allow engineers to design and benefit from
better retrofit solutions. The more advanced numerical models adopted for these
prototypes will justify the use of larger deformation limits which will increase their
performance.
Finally, the change to a tri-hazard probabilistic CPDE check will make the
guidelines more probabilistically robust and similar to the new GSC hazard model,
which also includes all hazard sources in its probabilistic seismic hazard analysis. The
updated guidelines will continue to provide safe and cost and time efficient retrofit
solutions for BC’s at-risk school buildings.
ACKNOWLEDGEMENTS
REFERENCES
© ASCE
ATC & SEI 2015 753
Chin, D., Ventura, C. V., Bebamzadeh, A., and Fairhurst, M. “Effects of subduction
ground motions on the probability of collapse of low-rise buildings.”
Proceedings of the 11th Canadian Conference on Earthquake Engineering.
Gatto, K. and Uang, C. M. (2002) “Cyclic response of woodframe shearwalls: loading
protocol and rate of load effects: CUREE W-13 Report.” Consortium of
Universities for Research in Earthquake Engineering, Richmond, Ca.
Ghobarah, A. (2001). “Performance-based design in earthquake engineering: state of
development.” Engineering Structures, 23(8), 878-884.
Halchuk, S., Allen, S., Adams, J., and Rogers, G. (2014) “Fifth Generation Seismic
Hazard Model Input Files as Proposed to Produce Values for the 2015 National
Building Code of Canada.” Geological Survey of Canada Open File 7576.
Lin, T., Haselton, C. B., Baker, J. W. (2013a) “Conditional spectrum-based ground
motion selection. Part I: Hazard consistency for risk-based
assessments.” Earthquake Engineering & Structural Dynamics, 42(12) 1847-
1865.
Lin, T., Haselton, C. B., Baker, J. W. (2013b) “Conditional spectrum-based ground
motion selection. Part II: Intensity-based assessments and evaluation of
alternative target spectra.” Earthquake Engineering & Structural Dynamics,
42(12) 1847-1865.
NEHRP Consultants Joint Venture. (2011). “Selecting and scaling earthquake ground
motions for performing response-history analyses.” NIST GCR, 11-917.
Priestley, M. J. N. (2000). “Performance based seismic design.” Bulletin of the New
Zealand Society for Earthquake Engineering, 33(3) 325-346.
Raghunandan, M., and Liel, A. B. (2013). “Effect of ground motion duration on
earthquake-induced structural collapse.” Structural Safety, 41(1), 119-133.
Raghunandan, M., Liel, A. B., and Luco, N. (2014). “Collapse risk of buildings in the
Pacific Northwest region due to subduction earthquakes.” Earthquake Spectra.
Vamvatsikos, D. and Cornell, A. C. (2002). “Incremental dynamic analysis.”
Earthquake Engineering & Structural Dynamics, 31(3) 491-514.
Wen, Y. K. (2001). “Reliability and performance-based design.” Structural Safety,
23(4), 407-428.
© ASCE