You are on page 1of 8

Stessa 2012 Mazzolani & Herrera (eds)

2012 Taylor & Francis Group, London, ISBN 978-0-415-62105-2

Performance evaluation of three pre-qualified steel systems in Canada


T.Y. Yang & M. Murphy
Department of Civil Engineering, University of British Columbia, Vancouver, Canada

ABSTRACT: Performance-based earthquake engineering aims to describe the seismic performance of


structures using metrics that are of immediate use to both engineers and stakeholders. A rigorous yet
practical implementation of performance-based earthquake engineering methodology is presented. The
methodology consistently accounts for the uncertainties in the hazards exposed by the structure, structural response and structural damage to quantify the performance of the entire system. A prototype
office building designed according to the National Building Code of Canada was used to compare the
seismic performance of three pre-qualified steel seismic force resisting systems. Both the material use and
expected repair cost at different levels earthquake shaking intensities are presented in this paper. The outlined example demonstrates a clear and transparent procedure to compare the performance of different
seismic force resisting systems and allows stakeholders to make an informed risk-based decision to select
the best structural system.
1

INTRODUCTION

Steel structures are prevalent structural systems


for seismic applications around the world. Among
different Seismic Force Resisting Systems (SFRS),
Moment Resisting Frame (MRF), Concentrically
Braced Frame (CBF) and Buckling Restrained
Braced Frames (BRBF) are found throughout
North America. Selection of the structural system is
usually based on engineering judgment. The typical
approach is to select a structural system which satisfies the minimum standard specified by the local
building code(s) and which carries the minimum
initial construction cost. However, the relative
seismic performance of the system throughout its
life cycle is not usually considered. To compare the
seismic performance of the MRF, CBF and BRBF
systems at different levels of earthquake shaking,
a five-story office building located in Vancouver,
British Columbia is designed and analyzed. The prototype building is designed with the assistance of a
renowned Vancouver-based structural engineering
firm according to the requirements specified in the
National Building Code of Canada (NBCC 2011)
and Canadian Steel Code CSA S16-09 (CSA 2010).
A detailed performance assessment of the prototype building using each of these three systems was
analyzed using the performance-based assessment
methodology presented by Yang et al. (2009a).
This methodology uses a Monte-Carlo simulation
procedure in which the building is analyzed under
numerous earthquake ground motions, with repair
costs aggregated to determine rates at which different repair costs occur. To carry out the procedure,

major structural and nonstructural components


of the buildings are identified and grouped into
performance groups. Damage fragility relations,
corresponding repair methods, repair material
quantities and repair cost functions are defined for
each performance group. Finite element models of
the buildings, one for each of the MRF, CBF and
BRBF systems were developed. Nonlinear dynamic
analyses for individual earthquake ground motion
records are conducted to establish peak response
quantities. Given peak response quantities, a random number generator is used to assign the specific
damage states based on the probabilities defined
by the fragility relations for each performance
group. Repair cost is then calculated based on the
building damage state. To generate cost statistics,
the process is repeated a large number of times for
different earthquake ground motion records. The
results of the performance assessment are used to
compare the relative performance merits of these
three structural systems.
2

DESCRIPTION OF THE PROTOTYPE


BUILDINGS

A prototype five-story (seven bays by five bays)


office building without basement level was selected
for this study. The building has a fixed bay width
of 9 meters, a first-floor height of 4.25 meters
and a floor height of 3.65 meters at other floors.
The building is symmetric in both the NorthSouth and East-West directions and has 6 SFRS
bays in each direction. Figure 1 shows a typical

815

3
2

Plan and perspective view of prototype

W31067

W1000 483

4th Floor
3rd Floor
2nd Floor

A1

0
585

Roof
W310 179

W690 125
00
= 30
A5 W690 125
00
= 30
A4 W690 125
50
= 43
A3 W690 125
50
= 58
A2 W690 125

5th Floor
4th Floor
3rd Floor
2nd Floor
1st Floor

PERFORMANCE GROUPS
AND DAMAGE STATES

Major components of the prototype building were


assigned to 26 performance groups (PGs). These
include: one structural PG at each floor level (15),

W760 257
W760 284
W760 284

W360 421

3.65
3.65
3.65
4.25

W690 125

Roof
5th Floor
4th Floor
3rd Floor

W360 509

3.65

b) BRBF
W690 125

floor plan and perspective view of the structure.


The SFRS bays are shown in bold in Figure 1a.
All structural members were designed according
to the National Building Code of Canada 2010
(NBCC 2010) and the Canadian Institute of Steel
Construction design standard (CSA 2010). Three
code pre-qualified SFRS are included in this
study. These include: a) Type DMoment Resisting Frame (MRF); b) Type MDConcentrically
Braced Frame (CBF); and c) Type DBuckling
Restrained Braced Frame (BRBF). Figure 2 shows
the member sizes and configurations for the SFRS
bays and the typical gravity bays. Note that gravity
columns for the MRF structure are different
because the columns were designed with a column
effective length factor of K = 1.2, whereas the gravity columns of the braced frames were designed
with K = 1.0. This reflects the ability of the gravity
frame to sway laterally in the MRF. Overall, the
building equipped with the BRBF system uses 21%
less total structural steel than the MRF, while the
CBF system uses 7% less total structural steel than
the MRF.
3

W1000 483

W310 342

4.25

b) Perspective view

2nd Floor
1st Floor

4.25

3.65

3.65

W690 125
W690 125
W690 125
W690 125

Roof
W250 67

3.65

3.65

c) MRF
W690 125

5th Floor
4th Floor

W250 101

Figure 1.
building.

W310179

3.65
3.65

eters

7@9m

W310342

rs

W360 421

ete

W1000 350

5th Floor

1st Floor

3.65

9m

W200 100
braces

3.65

5@

W200 86
braces

W1000 x 350

a) CBF

4.25 meters

a) Typical floor plan


N

4 @ 3.65 meters

1
7 @ 9 meters

W200 71
braces

Roof

W310x143

W200 59
braces

W310143

W360 509

5 @ 9 meters

W200 42
braces

W310x67

W920 223

W250 67

W250 101

3.65

3.65

3.65

3.65

4.25

3rd Floor
2nd Floor

1st Floor
d) Gravity Frame

Figure 2. Sections for CBF, BRBF, MRF & Gravity


Frame.
Note: 1) All length units in meters. 2) Brace core area (A1
to A5) for BRBF is shown in mm^2. 3) The gravity column for the MRF uses a W310 179 at the lower floors
and a W310 79 at the upper floors. 4) Column splices
are located at the 3rd floor mid-height. 5) All connections are pin connected unless fixity indicated by

816

one exterior (610) and one interior (1115)


drift-sensitive non-structural component PG at
each floor, one interior acceleration-sensitive nonstructural component PG at each floor (1620),
one acceleration-sensitive content PG at each floor
(2125) and one HVAC PG at the roof (26). The
PGs are selected based on a collection of components whose performance is similarly affected by a
particular Engineering Demand Parameter (EDP).
For example, the structural components were
assigned to PGs whose performance is associated with interstory drift in the story where the
components are located. The nonstructural components were divided into displacement and
acceleration groups. The displacement groups use
interstory drift ratios to define the performance,
while the acceleration groups use absolute floor
level accelerations. Multiple damage states were
defined for each PG. These states correspond to
different levels of damage and the resulting repair
actions. For example, the interior drift-sensitive
non-structural component performance group at
the first floor (PG 11) has three states. States range
from none (DS1) to minor (DS2) to severe damage
(DS3). For each state, a model (fragility relation)
defines the probability of damage being less than
or equal to the threshold damage given the value
of the EDP associated with the PG. Figure 3 shows
the fragility curves defined for PG 11. On this
figure, if the interstory drift ratio is 0.8%, the PG
has a 60% probability in DS2 and 40% probability
in DS3. Table 1 shows a summary of the performance groups included in this study. Symbols dui and
ai represent the interstory drift ratio at the ith story
and the total floor acceleration at the ith floor,
respectively. The performance groups presented in
this study are obtained from the ATC 58 research
team and the values used in this study are summarized in Yang et al. (2009b).

Table 1.

Performance
group

Name

Location

SH12

SH23

SH34

SH45

SH5R

EXTD12

EXTD23

EXTD34

EXTD45

10

EXTD5R

11

INTD12

12

INTD23

13

INTD34

14

INTD45

15

INTD5R

16

INTA2

17

INTA3

18

INTA4

19

INTA5

20

INTAR

21
22
23
24
25
26

CONT1
CONT2
CONT3
CONT4
CONT5
EQUIPR

between
levels 1 & 2
between
levels 2 & 3
between
levels 3 & 4
between
levels 4 & 5
between
levels 5 & R
between
levels 1 & 2
between
levels 2 & 3
between
levels 3 & 4
between
levels 4 & 5
between
levels 5 & R
between
levels 1 & 2
between
levels 2 & 3
between
levels 3 & 4
between
levels 4 & 5
between
levels 5 & R
below
level 2
below
level 3
below
level 4
below
level 5
below
level R
at level 1
at level 2
at level 3
at level 4
at level 5
at level R

Figure 3.
2009a).

Example of fragility curves (Yang et al.

Summary of performance groups.


Engineering
demand
parameters
du1
du2
du3
du4
du5
du1
du2
du3
du4
du5
du1
du2
du3
du4
du5
a2
a3
a4
a5
aR
ag
a2
a3
a4
a5
aR

FINITE ELEMENT MODELS

Three finite element models of the building were created for each of the CBF, MRF and BRBF systems.
The models were developed using OpenSees (UCB
1997). Due to symmetric nature of the building,

817

only half was modeled. For simplicity, only the


response in the East-West direction is presented in
this paper. A 2-D model consisting of the frames in
line 1, 2 and 3 (Figure 1a) is placed in series and tied
together using multi-point constraints at each floor.
To model the stiffness and strength of the beam to
column connections and brace end connections, all
the pin connections are modeled using the semirigid connection proposed by Astaneh-Asl (2005).
The ends of the BRBF braces were pin connected,
while the moment connections are fixed. The BRB
braces were modeled using the nonlinear truss element in OpenSees, while the all other elements were
modeled using the flexibility-formulation nonlinear
fiber-cross-section beam-column elements in OpenSees. Masses were lumped at the nodes according
to the tributary area. The P- effect was accounted
using the corotational transformation in OpenSees (de Souza 2000, Filippou and Fenves 2004).
Rayleigh damping of 2% was assigned to the first
and third vibration modes of building. Table 2
shows the first five periods of the buildings.
5

GROUND MOTION SELECTION

A detailed seismic hazard analysis is conducted for


the building. Three hazard levels representing the
2% probability of exceedance in 50 years (2/50),
10% probability of exceedance in 50 years (10/50)
and 40% probability of exceedance in 50 years
(40/50) are included in this study. A total of 16,
18 and 17 ground motions were selected for the
2/50, 10/50 and 40/50 hazard levels, respectively.
Ground motions were taken from the PEER
NGA database (PEER, 2011) for each of the
hazard levels. The ground motions were amplitude scaled such that the mean spectrum of the
set of ground motions over the period range from
0.2T1 to 1.5T1 (with T1 being the first vibrational
period of the structure) is within 10% of the target
spectrum. Figure 4 shows an illustrative example
of the scaled spectra for the 2/50 hazard level for
the MRF. Tables 3, 4 and 5 list the selected ground
motions for this study.
Table 2.

Periods for the 3 structural systems.

Mode

Moment
resisting
frame
[seconds]

Buckling
restrained
braced frame
[seconds]

Concentrically
braced frame
[seconds]

T1
T2
T3
T4
T5

1.14
0.41
0.22
0.14
0.10

0.86
0.32
0.19
0.15
0.11

0.47
0.18
0.12
0.09
0.07

Figure 4. Example of ground motion scaling procedure


for 2% in 50 years hazard for the MRF system.
Table 3.

Selected records 2% in 50 years hazard.

Event

Year

Station

San Fernando
Tabas- Iran
Coyote Lake
Imperial Valley
Imperial Valley
Irpinia- Italy
Coalinga
Morgan Hill
N. Palm Springs
Chalfant Valley
Whittier Narrows
Loma Prieta
Northridge-01
Kobe- Japan
Chi-Chi- Taiwan
Manjil- Iran

1971
1978
1979
1979
1979
1980
1983
1984
1986
1986
1987
1989
1994
1995
1999
1990

CastaicOld Ridge Rt
Dayhook
Gilroy Array #6
El Centro Array #8
SAHOP Casa Flores
Sturno
Pleasant Valley P.P.y
Gilroy Array #4
Desert Hot Springs
BishopLADWP S. St
Santa Fe Spring
Gilroy Array #3
Pacoima Dam
Kakogawa
TCU089
Abhar

Table 4.

Selected records 10% in 50 years hazard.

Event

Year

Station

San Fernando
Imperial Valley
Imperial Valley
Imperial Valley
Mammoth Lakes
Coalinga
Morgan Hill
N. Palm Springs
Chalfant Valley
Whittier Narrows
Loma Prieta
Loma Prieta
Landers
Northridge
Kobe- Japan
Manjil- Iran
Hector Mine
Chi-Chi- Taiwan

1971
1979
1979
1979
1980
1983
1984
1986
1986
1987
1989
1989
1992
1994
1995
1990
1999
1999

LAHollywood Stor FF
Calexico Fire Station
El Centro Array #3
Holtville Post Office
Convict Creek
Pleasant Valley
Gilroy Array #4
Desert Hot Springs
BishopLADWP S. St
GlendaleLas Palmas
Agnews State Hospital
FremontEmerson Ct.
Amboy
LABaldwin Hills
Kakogawa
Abhar
Amboy
TCU122

818

Table 5.

Selected records 40% in 50 years hazard.

Event
Imperial Valley
Imperial Valley
Victoria- Mexico
Irpinia- Italy
Coalinga
Hollister
Chalfant Valley
Chalfant Valley
Whittier Narrows
Landers
Big Bear
Northridge
Northridge
Hector Mine
Denali- Alaska
Chi-Chi- Taiwan
Chi-Chi- Taiwan

Year
1979
1979
1980
1980
1983
1986
1986
1986
1987
1992
1992
1994
1994
1999
2002
1999
1999

Median peak interstory drift ratio.

Hazard

Building

du2
[%]

du3
[%]

du4
[%]

du5
[%]

duR
[%]

2/50

MRF
CBF
BRBF

0.93
0.46
0.59

1.23
0.46
0.69

1.1
0.5
0.8

1.2
0.39
0.82

1.05
0.27
0.43

10/50

MRF
CBF
BRBF

0.5
0.26
0.4

0.66
0.26
0.44

0.6
0.3
0.5

0.64
0.24
0.52

0.55
0.16
0.29

40/50

MRF
CBF
BRBF

0.23
0.11
0.19

0.28
0.11
0.2

0.3
0.1
0.2

0.31
0.1
0.21

0.27
0.07
0.13

Table 7.

Median peak floor acceleration.

Station
Plaster City
El Centro Array #3
SAHOP Casa Flores
Rionero In Vulture
ParkfieldStone Corral
Hollister Diff Array #3
Convict Creek
Long Valley Dam
SylmarSayre St
LAN Westmoreland
Featherly ParkMaint
ComptonCastlegate
MalibuP. Dume Sch
Whitewater Trout Farm
TAPS Pump Station #09
CHY111
TCU076

ag
Hazard Building [g]

a3
[g]

a4
[g]

a5
[g]

aR
[g]

MRF
CBF
BRBF

0.4 0.64
0.49 0.8
0.49 0.65

0.7
1
0.7

0.67
1.13
0.71

0.6
1.25
0.62

1.11
1.73
0.94

10/50

MRF
CBF
BRBF

0.21 0.31
0.22 0.43
0.22 0.35

0.3
0.6
0.4

0.36
0.65
0.4

0.33
0.75
0.45

0.56
0.99
0.64

40/50

MRF
CBF
BRBF

0.1
0.1
0.1

0.2
0.3
0.2

0.18
0.29
0.17

0.17
0.32
0.19

0.27
0.45
0.29

0.14
0.18
0.15

Table 8.
ratio.

Standard deviation of peak interstory drift

Hazard

Building

du2
[%]

du3
[%]

du4
[%]

du5
[%]

duR
[%]

2/50

MRF
CBF
BRBF

0.21
0.13
0.2

0.29
0.12
0.26

0.3
0.1
0.3

0.23
0.07
0.22

0.22
0.05
0.09

10/50

MRF
CBF
BRBF

0.16
0.06
0.1

0.22
0.06
0.15

0.2
0.1
0.2

0.21
0.05
0.16

0.16
0.03
0.05

40/50

MRF
CBF
BRBF

0.05
0.03
0.06

0.07
0.03
0.07

0.1
0
0.1

0.07
0.03
0.06

0.06
0.02
0.03

COMPUTE THE REPAIR COSTS

The computed response presented in Tables 6 to 9


were used in a mathematical model to systematically generate a large numbers of additional simulated response maxima having the same statistical
properties as the original set. Detailed procedure
to synthetically generate the large array of EDP
matrix is presented in Yang et al. (2009a). The
generated EDP matrix is then used to identify the
damage states of each performance groups. Once
the damage state of each performance group is
identified, the repair action and associate repair
cost to each performance groups shown in Table 1
is then calculated. Finally the total repair cost for

a2
[g]

2/50

SEISMIC RESPONSE QUANTIFICATION

Nonlinear dynamic analyses were conducted to


determine the seismic response of the buildings
to each of the scaled ground motions. Tables 6
and 7 show the median peak drift and floor
acceleration recorded from the time history
analyses, respectively. Tables 8 and 9 show corresponding standard deviation of the peak drift
and floor acceleration recorded from the time
history analyses, respectively. The result shows
the MRF has the highest median story drift followed by the BRBF than the CBF for all hazard
levels considered. On the other hand, the median
peak floor accelerations are highest in the CBF,
followed by BRBF and MRF for all the hazard
levels considered.

Table 6.

the entire building is then summed over all performance groups. The process is repeated a large
number of times to quantify the distribution of the
repair costs at different levels of earthquake shaking intensities. Figure 5 shows the discrete cumulative distribution functions of the total repair costs
for the three systems at each of the three hazard

819

Table 9.

Standard deviation of peak floor acceleration.

ag
Hazard Building [g]

a2
[g]

a3
[g]

a4
[g]

a5
[g]

aR
[g]

2/50

MRF
CBF
BRBF

0.11 0.18 0.2 0.14


0.11 0.16 0.2 0.21
0.11 0.15 0.2 0.13

0.11
0.24
0.15

0.21
0.32
0.18

10/50

MRF
CBF
BRBF

0.04 0.11 0.1 0.08


0.06 0.08 0.1 0.14
0.06 0.08 0.1 0.11

0.1
0.17
0.09

0.15
0.2
0.12

40/50

MRF
CBF
BRBF

0.02 0.03 0
0.03
0.02 0.03 0.1 0.07
0.02 0.03 0
0.04

0.04
0.08
0.04

0.05
0.1
0.07

earthquake engineering methodology presented


here offers important selection criteria in which
the seismic loss of the facility at different levels
of earthquake shaking intensities is compared.
Under the circumstances presented in this study,
the buckling restrained braced frame offers the
lowest material usage (a 21% less structural steel
than the moment resisting frame) and lowest
expected median repair costs for all hazard levels
considered. Conclusions regarding the relative
merits of these three systems are somewhat anecdotal and may vary when the configurations of the
structure are changed. However, the overall methodology presented is widely applicable and offers
quantitative measure which can be used to make
informed decision in selecting the best structural
system for a given project.

ACKNOWLEDGEMENTS

Figure 5.

Discrete CDF of the repair cost distribution.

levels considered. The result shows that at the lowest shaking intensity (40/50) hazard level, the MRF
and BRBF have the lowest median repair cost. As
the shaking intensity increases to 10/50 hazard levels, all three systems have roughly the same median
repair cost. As the shaking intensity increases to
2/50 hazard level, the BRBF and CBF have roughly
the same median repair costs while the MRF has
the highest repair cost.
8

This work was funded in part by the Natural


Sciences and Engineering Research Council of
Canada (NSERC) jointly with the Steel Structures Education Foundation (SSEF), award
reference number 150656511 (application ID
411465). The authors would like to acknowledge
David MacKinnon of the SSEF for making this
project possible. The authors would also like to
thank: Dr. Didier Pettinga, and Rob Simpson, of
Glotman Simpson Consulting Engineers for their
help in designing the prototype models; Members
of the ATC-58 project team for providing fragility curves, repair method and repair cost information for the performance groups used in this study.
Any opinions, findings and conclusion or recommendations expressed in this material are those of
the authors and do not necessarily reflect those of
the Natural Sciences and Engineering Research
Council of Canada or the Steel Structures Education Foundation.

SUMMARY AND CONCLUSIONS

This paper presents a detailed performance assessment of a five-story office building designed
using three options of steel seismic force resisting
systems. This includes the concentrically braced
frame, buckling restrained braced frame and
moment resisting frame. The purpose of the investigation was to understand the relative seismic performance of these systems by comparing the initial
material use and the post-earthquake repair costs
at different levels of shaking intensities. The results
provide a quantitative measure for engineers and
other stakeholders to make an informed decision
to select the best system to achieve the design
objective. The advent of performance-based

REFERENCES
Astaneh-Asl, A. 2005. Design of shear tab connections
for gravity and seismic loads, SteelTips, Structural
Steel Educational Council.
ATC. 2007. Development of next-generation
performance-based seismic design procedures for new
and existing buildings, Applied Technology Council,
Redwood city, CA. http://www.atcouncil.org/atc-58.
shtml.
Cochran, Michael & Honeck, William, 2004. Design of
Special Concentric Braced Frames, SteelTips, Structural Steel Education Council.
CSA S16-09. 2010. Handbook of Steel Construction,
Canadian Institute of Steel Construction, Markham,
Ontario.

820

de Souza, R.M. 2000. Force-based finite element for


large displacement inelastic analysis of frames,
Doctoral dissertation, Graduate Division of the
University of California, Berkeley.
Filippou, F.C. & Fenves, G.L. 2004. Methods of analysis for earthquake-resistant structures, Chapter 6 in
Earthquake Engineering From Engineering Seismology
to Performance-based Engineering, RCR Press, Boca
Raton, Florida.
Lpez, Walterio & Sabelli, Rafael 2004. Seismic Design
of Buckling-Restrained Braced Frames, SteelTips,
Structural Steel Educational Council.
NBCC 2010, National Building Code of Canada,
National Research Council Canada, Ottawa, ON.
Pacific Earthquake Engineering Research Center
(PEER) Strong Motion Database 2011. University
of California at Berkeley http://peer.berkeley.edu/
peer_ground_motion_database.

UCB 1997. Open system for earthquake engineering


simulation (OpenSees) framework, Pacific Earthquake Engineering Research Center, University of
California, Berkeley, http://opensees.berkeley.edu/.
Yang, T.Y., Bozidar Stojadinovic & Jack Moehle.
Demonstration of a practical method for seismic
performance assessment of structural systems, Earthquake Spectra, EERI. (accepted for publication).
Yang, T.Y., Moehle, J., Stojadinovic, B. & Der
Kiureghian, A. 2009a. Performance evaluation of
structural systems: theory and implementation, Journal of Structural Engineering, ASCE. Vol. 135, No. 10,
pg. 1146pg 1154.
Yang, T.Y., Moehle. J. & Stojadinovic, B. 2009b.
Performance evaluation of innovative structural
framing systems, PEER report, Pacific Earthquake
Engineering Research Center, University of California,
Berkeley. Report number 2009/103.

821

You might also like