832

Seismic performance of a full-scale, reinforced

high-performance concrete building. Part I:
Experimental study
Sébastien Mousseau and Patrick Paultre

Abstract: Full-scale tests provide valuable information on the characteristics of building structures that can be used to
evaluate design methods, to calibrate modelling techniques, and to determine damage corresponding to loading levels.
These tests are scarce due to the enormous requirements in testing space and specialized testing equipment. The seismic
behaviour of a full-scale, two-storey, reinforced high-performance concrete building designed with moderate ductility de-
tailing is evaluated by pseudo-dynamic testing, during which increasing seismic loads are applied, resulting in increasing
levels of permanent damage to the structure. To monitor the level of damage, a series of successive forced-vibration tests
are also carried out at each step of the process and are used to track changes in the key dynamic properties of the building.
The paper presents the design of the test structure according to the new edition of the CSA A23.3-04 Design of concrete
structures standard, the series of pseudo-dynamic tests simulating different levels of earthquake excitation consistent with
the 2005 edition of the National building code of Canada, and the evaluation of the performance of the building. It is
shown that the detailing requirements of CSA A23.3-04 are more than adequate to provide the ductility and overstrength
expected.
Key words: high-performance concrete, reinforced concrete, moment-resisting frames, earthquake engineering, force
modification factors, overstrength, ductility, pseudo-dynamic testing, push-over.
Résumé : Les essais à grande échelle fournissent des informations importantes sur les caractéristiques des structures pou-
vant être utilisées pour évaluer les méthodes de dimensionnement, pour calibrer des techniques de modélisation, et pour
déterminer l’endommagement correspondant à des niveaux de chargement. La rareté de ces essais est due principalement à
la nécessité de disposer de grands espaces de laboratoire et d’équipements d’essais spécialisés. Dans cet article, le compor-
tement d’un bâtiment en vraie grandeur de deux étages est évalué par essais pseudo-dynamiques dans lesquels le bâtiment
est soumis à des niveaux d’excitation sismique croissants causant des dommages de plus en plus importants à la structure.
Le bâtiment est construit en béton à hautes performances pour posséder une ductilité modérée. Des essais en vibration for-
cée sont utilisées pour évaluer les changements dans les propriétés dynamiques de la structure. L’article présente le dimen-
sionnement de la structure d’après la nouvelle édition de la norme CSA A23.3-04, Dimensionnement des structures en
béton, la suite des essais pseudo-dynamiques simulant les différents niveaux d’excitation des tremblements de terre consis-
tants avec le Code national du bâtiment du Canada 2005, et l’évaluation de la performance du bâtiment. On montre que
les règlements de la norme CSA A23.3-04 sont tout à fait adéquats pour assurer le niveau de ductilité et de surcapacité dé-
siré.
Mots-clés : béton à hautes performances, béton armé, ossature résistant aux moments, génie parasismique, facteurs de
modification de force, surcapacité, ductilité, essai pseudo-dynamique, poussée progressive.

Introduction the development of new numerical models and design meth-

Experimental research is the most reliable means of as-
sessing seismic performance of structures and is crucial to
Received 31 May 2007. Revision accepted 9 January 2008.
Published on the NRC Research Press Web site at cjce.nrc.ca on
30 July 2008.
S. Mousseau. SNC-Lavalin inc, 455 René-Lévesque Boulevard
West, Montreal, QC H2Z 1Z3, Canada.
P. Paultre.1 Department of Civil Engineering, Université de
Sherbrooke, Sherbrooke, QC J1K 2R1, Canada.
Written discussion of this article is welcomed and will be
received by the Editor until 31 December 2008.
1Corresponding author (e-mail: patrick.paultre@usherbrooke.ca).
ods. Ultimately, the performance of buildings during actual
earthquakes is the only measure of the adequacy of new de-
sign procedures and detailing practice. Numerical predic-
tions with calibrated models using full-scale test results
provide the only choice of systematically evaluating the in-
fluence certain parameters have on the likely response of
building structures. This paper reports on a comprehensive
test series performed on a full-scale, two-storey reinforced
high-performance concrete (HPC) building. The building
was subjected to repeated pseudo-dynamic tests as part of a
research project on the seismic behaviour of HPC structures
and damage detection. The main objectives of the experi-
mental investigation were to carry out pseudo-dynamic tests
on a two-storey reinforced HPC building to: (i) determine
the damage and performance under increasing seismic loads,

Can. J. Civ. Eng. 35: 832–848 (2008) doi:10.1139/L08-017 # 2008 NRC Canada
Mousseau and Paultre 833

(ii) investigate the force modification and overstrength fac- Fig. 1. Full-scale two-storey reinforced high-performance concrete
tors suitable for the structure that is designed and detailed building with reaction wall.
to exhibit moderate ductility (nominal ductility in past co-
des), (iii) investigate whether the concrete compressive
strength limit of 55 MPa in Canadian standard CSA-A23.3-
94 (CSA 1994) is too conservative, (iv) investigate the ad-
equacy of the design and detailing requirements for moder-
ate ductility, (v) provide experimental evidence for
predicting damage under increasing seismic loads, and
(vi) evaluate the performance of different inelastic time-
history dynamic analyses. This paper focuses on the design
of the test structure, the experimental procedures, the
pseudo-dynamic and push-over tests, and the test results.
A companion paper (Mousseau et al. 2008) compares the
experimental results obtained during the seismic tests with
the predictions of inelastic time-history dynamic analyses.

High-performance concrete in the Canadian standard

Recent advances in material technology facilitate the
production of high-performance concrete. The term high-
performance concrete encompasses improved mechanical
characteristics, increased durability, enhanced workability,
better behaviour under loading, etc. In this research project,
only the improved mechanical characteristics of the high- Fig. 2. Plan and elevation views of the building.
performance concrete are considered, especially high-
compressive strength and its effects. In this paper, the
terms HPC and high-strength concrete (HSC) are used in-
terchangeably. High-strength concrete is more fragile than
normal-strength concrete (NSC) and columns made of
HSC need to be properly confined to display the same
level of ductility as NSC columns. In view of limited ex-
perimental results at the time of its publication, the 1994
Canadian standard CSA-A23.3 (CSA 1994) limited the
specified concrete compression strength to 55 MPa in seis-
mic zones. A large number of experimental works have
shown that HSC columns and beams, when properly de-
tailed, can exhibit sufficient ductility. Most, if not all,
tested specimens in these experimental works were single
cantilever columns or beams or beam–column–slab sub as-
semblages. The present research program, therefore, was
developed to study the behaviour of HSC members in a
full-scale reinforced concrete structure designed according
to the new Canadian standard CSA-A23.3-04 (CSA 2004)
under realistic earthquake excitation.

Moment-resisting frame with moderate ductility

The 1984 version of Canadian standard A23.3 (CSA 1984)
introduced new requirements for the design and detailing of
concrete structures in seismic zones, particularly for frames
with what was then called nominal ductility. This new duc-
tility level is between the ductile structures built according
to capacity design procedures to dissipate energy in specially
detailed members and all the other structures.
The suitability of the new requirements for nominal duc-
tility was confirmed by results from reversed cycling load-
ing tests on full-scale beam–slab–column subassemblages
and by inelastic time-history dynamic analyses (Paultre et value of R = 2.0 was assigned to a frame with nominal duc-
al. 1989; Paultre and Mitchell 1987). Ductility level was ac- tility, which did not change up to the 1995 version of the
counted for by using a force-modification factor, R, varying NBCC (NRCC 1995). One of the objectives of this research
from 1 to 4 for concrete structures in the 1985 version of the program was to study the possibility of increasing the force-
National building code of Canada (NBCC) (NRCC 1985). A modification factor for moment-resisting frames with nomi-
# 2008 NRC Canada
834 Can. J. Civ. Eng. Vol. 35, 2008

Table 1. Specified loads used in design.

Dead loads, D
Weight of structural members 24 kN/m3
Mechanical services, floor finishing, and partition loading for the first floor 1.70 kN/m2
Mechanical services and insulation weight loading for the roof 2.10 kN/m2
Walls, cladding, and insulation 2.40–3.60 kN/m2
Live load, L
Office building (first floor) 2.40 kN/m2
Snow load, S
Building located in Montréal 2.32 kN/m2
Earthquake load, E
Base shear: V = 0.1260 W (NRCC 1995) 95.69 kN
Wind load, W
Total pressure on building 0.866 kN/m2

nal ductility from 2.0 to 2.5 to do justice to these type of Table 2. Seismic response coefficient.
structures, which, as was judged by many designers and
members of the CSA A23.3 committee, possess more ductil- Period, T (s) Za/Zv S
ity than the factor 2.0 implied. It is important to note that, T £ 0.25 <1.0 2.1
when this research program began, the change of the force = 1.0 3.0
modification factor from 2.0 to 2.5 had not been made. >1.0 4.2
Therefore, this research was a part of the process of making 0.25 < T < 0.50 <1.0 2.1
that change. = 1.0 3.0 – 3.6(T – 0.25)
>1.0 4.2 – 8.4(T – 0.25)
pffiffiffiffi
Design and description of the building T ‡ 0.50 All values 1:5= T
The two-storey reinforced concrete building used for this
research project has a 5 m bay in the east–west (E–W) di- acceleration-related seismic zone (Za), and the velocity-
rection and a 4 m bay in the north–south (N–S) direction. related zone (Zv). For the building located in Montréal,
The storey height from top of slab to top of slab was 3 m. v = 0.10, I = 1.0, F = 1.0, Za = 4.0, Zv = 2, and W =
The columns were all 300 mm
300 mm. The two-way 759.4 kN.
slab floor system consisted of a 150 mm thick slab sup- Significant changes were introduced in the 2005 edition
ported by beams 300 mm
300 mm on all four sides. The of the NBCC (NRCC 2005) to determine the seismic base
specified concrete strength was 70 MPa and the specified shear, as given by:
steel yield strength was 400 MPa. Plan and elevation views
SðTa ÞMv IE W Sð2:0ÞMv IE W
of the building are shown in Figs. 1 and 2. ½3 V¼ 
The design of the building is summarized below; more Rd Ro Rd Ro
complete information can be found in Mousseau and Paultre where S(Ta) is the design spectral response acceleration for
(2005). The design forces used are in accordance with the the fundamental lateral period of vibration of the building
1995 edition of the NBCC (NRCC 1995) for a site located Ta, Mv is a factor to account for higher mode effects on
in Montréal. The moment-resisting frame concrete structure base shear, IE is an earthquake importance factor for the
was designed for nominal ductility with a seismic force structure, W is the seismic weight (100% of the dead load
modification factor R = 2.0. Table 1 presents the loads used plus 25% of the design snow load), Rd is a ductility-related
in the design process. Load combinations, as prescribed by force-modification factor, and Ro is an overstrength-related
the 1995 NBCC, were used to calculate the design forces. force-modification factor. The NBCC 2005 also specifies
The seismic base shear, V, was calculated from the fol- that, for structures with Rd ‡ 1.5, V need not be taken as
lowing equation: being greater than two thirds of the value calculated for
Ve Ta = 0.2 s, i.e.,
½1 V¼ U
R 2 Sð0:2ÞIE W
½4 V
where U = 0.6 is a calibration factor and Ve is the elastic 3 Rd Ro
base shear given by:
Figure 3 compares the base shear coefficients for
½2 Ve ¼ vSIFW moment-resisting frames with moderate (called nominal in
previous codes and standards) ductility built in Montréal ac-
where v is the zonal velocity ratio, I is the seismic impor- cording to the NBCC 1995 and 2005 in the form of base
tance factor, F is a foundation factor, and W is the seismic shear coefficients (V/W) as a function of the fundamental
weight (100% of the dead load plus 25% of the design snow lateral period of vibration (T). The base shear was calculated
load). S is the seismic response factor (Table 2), which is a for a structure located on site classification C, which is the
function of the period of vibration of the building and the reference site classification in the NBCC 2005, hence Fa =
# 2008 NRC Canada
Mousseau and Paultre 835

Fig. 3. Base shear coefficient (V/W) for reinforced concrete moder-
ately ductile moment-resisting frames in Montréal.
Fig. 4. Summary of detailing requirements for reinforced concrete
moderately ductile moment-resisting frames. ‘n, clear height of col-
umn; c, column dimension; d, effective height; dbh, diameter of the
hoop bars; db‘, diameter of the smallest longitudinal bars; h, mem-
ber depth.

Fv = 1.0. The other parameters were Rd = 2.5, Ro = 1.4, IE =

1.0, and Mv = 1.0. The 5% damped spectral response accel-
eration values for the Montréal site are Sa(0.2) = 0.69g,
Sa(0.5) = 0.34g, Sa(1.0) = 0.14g, Sa(2.0) = 0.048g. The de-
sign spectral acceleration values, S(T), are equal to the cor-
responding Sa(T) values as Fa = Fv = 1.0, with the additional
values S(T ‡ 4.0) = 0.024g.
The equations given in the NBCC 2005 to determine the
fundamental lateral period of vibration of moment-resisting
frames are:

½5 Ta ¼ 0:075h3=4
n

for concrete frames, and

½6 Ta ¼ 0:1N
for all other moment frames; where hn is the total height of
the building above the base and N is the total number of
storeys above exterior grade. The same expressions were
also specified in the NBCC 1995, except that eq. [6] can
also be used for any moment frames.
Figure 3 shows that the seismic forces calculated with the
NBCC 2005 are slightly higher than those calculated with
the NBCC 1995 for structures having a fundamental period
lower than 1.0 s located on site classification C. For the
building being studied, the base shear coefficient is 4.3%
higher according to the NBCC 2005.
Design and detailing of the members of the structure were
in agreement with the special provisions for seismic design
of CSA A23.3-94 Design of concrete structures (CSA
1994), but accounted for the new requirements of CSA
A23.3-04 (CSA 2004), which were then in a preliminary
stage. Hence, the strong column – weak beam concept and
new requirements for the design of confinement reinforce-
ment were all incorporated into the design. These new con-
finement requirements state that the total effective area in
each of the principal directions of the cross section within a
spacing s of rectangular hoop reinforcement, Ash, shall not
be less than the larger of:
Ag fc0
½7 Ash ¼ 0:15kn kp shc
Ach fyh
fc0
½8 Ash ¼ 0:09 shc
fyh
where kn = n‘/(n‘ – 2) is a factor accounting for the effec-
tiveness of the transverse reinforcement arrangement (where
n‘ is the total number of longitudinal bars in the column
cross section that are laterally supported by the corner of
hoops or by hooks of seismic crossties), kp = Pf /Po is a fac-
tor accounting for the level of axial load (where Pf is the
maximum factored axial load for earthquake loading cases
and Po is the nominal axial load resistance at zero eccentri-
city), Ag is the gross area of section, Ach is the cross-
sectional area of core, fc0 is the specified compressive con-
crete strength, fyh is the specified yield strength of hoop
reinforcement, s is the spacing of transverse reinforcement,
and hc is the dimension of concrete core of rectangular
section measured perpendicular to the direction of the
hoop bars to outside of peripheral hoop. The total effective
area of rectangular hoops reinforcement is based on the
# 2008 NRC Canada
836
Fig. 5. Reinforcement details of beams and columns. E–W, east–
west; N–S, north–south. All dimensions in millimetres.
curvature ductility demand and is, therefore, a function of
the axial load, effectiveness of the transverse reinforce-
ment, concrete strength, and the yield strength of hoop re-
inforcement. General detailing requirements for moderately
ductile moment-resisting frames are summarized in Fig. 4.
Figures 5 and 6 show the reinforcement details of the
beams, columns, and slabs, and Fig. 7 shows a picture of a
beam–slab–column connection of the two-storey frame
structure being studied.
Casting of the building took place in seven stages. The
pedestals were first installed on the strong floor and casting
proceeded for the columns of the first storey, two columns
at a time, followed by casting of the first-floor slab. The
procedure was then repeated for the second storey. The con-
Can. J. Civ. Eng. Vol. 35, 2008
Fig. 6. Reinforcement details of slabs: (a) roof reinforcement and
(b) first floor reinforcement. E–W, east–west; N–S, north–south.
All dimensions in millimetres.
crete formulation is shown in Table 3. A large number of
100 mm
200 mm and 150 mm
300 mm cylinders and
100 mm
100 mm
400 mm prisms were made at the
time of casting to determine concrete compressive and ten-
sile strengths at different ages. The average compressive
and tensile (modulus of rupture) concrete strengths at 28 d
were 63 and 5.5 MPa, respectively. At the time of testing,
the average compressive strength had increased to 66 MPa
for the low seismic excitation tests (April 2001) and was
73 MPa during very severe seismic excitation tests (Sep-
tember 2004).
Tension tests were performed on at least three steel sam-
ples of each bar diameter for each batch of steel bars. For
the 10M bars, the average values of yield strength, tensile
strength, and Young’s modulus were fy = 436 MPa, fu =
555 MPa, and Es = 205 300 MPa, respectively. For the 15M
bars, the average values were fy = 401 MPa, fu = 631 MPa,
and Es = 200 200 MPa.
# 2008 NRC Canada
Mousseau and Paultre 837

Fig. 7. First floor joint reinforcement.
forced concrete buildings. The rapid development of
computer-assisted computations over the last 30 years as
well as the possibility of realising a complete on-line
computer-controlled acquisition system and data processing
allowed the development of new hybrid experimental techni-
ques, such as the pseudo-dynamic testing method (PSD)
(Takanashi 1975).

Pseudo-dynamic testing methodology

Given a discrete parameter model, the equations of mo-
tion for a structural system can be written in matrix form as
½9 M€u ðtÞ þ C_u ðtÞ þ fr ðtÞ ¼ fðtÞ
where M and C are the mass and viscous damping matrices
of the structure, u_ ðtÞ and u€ ðtÞ are the vectors of nodal rela-
tive velocity and acceleration, t is time, fr(t) is the nodal re-
storing force vector, and f(t) is the external excitation vector
applied to the system. If a shear-type structure is subjected
to a horizontal ground acceleration u€ g ðtÞ, the force vector
f(t) = –Mr€u g ðtÞ, where r is an influence vector with each
element equal to one in the case of the tested building.
In an elastic system, the restoring force is fr(t) = Ku(t),
where K is the stiffness matrix and u(t) is the displacement
vector. For inelastic systems, more complex models are re-
quired to represent the relation between restoring forces and
displacements. The solution of eq. [9] can be approximated
using direct step-by-step integration techniques where the
time is divided into n equal steps
t = td/n, where td is the
total time of the external excitation, giving
Table 3. Concrete formulation.
½10 M€u i þ C_u i þ fri ¼ fi
Mixing water 155 kg/m3
Cement, type 10-SF* 470 kg/m3 in which u€ i = u€ ði
tÞ, u_ i ¼ u_ ði
tÞ, and fi = f(i
t).
Natural sand 830 kg/m3 Pseudo-dynamic testing involves an interactive closed-
Coarse aggregate, 2.5–10 mm 980 kg/m3 loop testing procedure that simulates dynamic loading. A
Superplasticizers (EUCON 37) 5 L/m3 computer program, which predicts the incremental displace-
Water reducer (EUCON DX) 250 mL/100 kg ments to be imposed on the structure after receiving nonlin-
ear force response feedback ( fri) from the actual structure,
Mass density 2435 kg/m3 models the interactive dynamic response. This process is
Water to cement ratio 0.33 carried out for the duration (td) of a particular input accel-
erogram ( fi). Therefore, measurements of the actual stiffness
*Approximately 7.5% of silica-fume content.
properties of the structure provide a realistic simulation of
Pseudo-dynamic and push-over testing the dynamic response histories. Whereas inertial (M€u i ) and
program viscous (C_u i ) damping forces are modeled analytically with
known and measured quantities, respectively, the nonlinear
The analysis of buildings subjected to seismic loadings structural restoring forces are measured experimentally be-
has always proved to be a delicate task. The inelastic behav- cause of the difficulty in modeling them accurately (Fig. 8).
iour of these structures is often complex to model despite In this way, the process automatically accounts for the hys-
the availability of different kinds of structural analysis soft- teretic damping due to inelastic deformation of the rein-
ware. Therefore, experimental investigation remains the forcement and irreversible damage of the concrete material,
most reliable method to predict seismic performance of which are the two major sources of energy dissipation.
civil-engineering structures and to calibrate numerical mod- To simulate the earthquake response of a structure, a re-
els. cord of an actual or artificially generated earthquake ground
Shake-table tests make it possible to obtain the response acceleration history is used as input data for the computer
of a structure subjected to seismic loadings. The structures running the pseudo-dynamic algorithm. The horizontal dis-
are usually small-scale specimens due to limitations of the placements of the control degrees-of-freedom (DOF) where
shaking tables. Therefore, size effects cannot be reproduced the mass of the structure is considered concentrated are
and detailing of reinforced concrete structures becomes calculated for a small time-step using a suitable time-
problematic. Consequently, shake-table studies are generally integration algorithm. These displacements are then applied
limited to very simplified models or scaled structures requir- to the test structure by servo-controlled hydraulic actuators
ing scaling laws that poorly represent the behaviour of rein- reacting against a stiff reaction wall. Load cells on the ac-
# 2008 NRC Canada
838
Fig. 8. Simplified representation of the pseudo-dynamic method. DOF, degrees of freedom; LVDT, linear voltage displacement transducer.
tuators return the forces necessary to reach the required
displacements and these structural restoring forces are used
by the computer program to predict the next time-step dis-
placement increment. A typical flowchart for the pseudo-
dynamic test method is illustrated in Fig. 9.
Instrumentation
Pseudo-dynamic tests require essentially the same equip-
ment as conventional quasi-static tests, in which prescribed
histories of load or displacement are imposed on specimen
structures by means of displacement-controlled hydraulic ac-
tuators. As shown in Figs. 1 and 10, the lateral seismic loads
were applied to the building by four double-acting,
dynamic-rated servo-hydraulic actuators with a stroke
of ±400 mm and 500 kN capacity reacting on a large-
capacity vertical reaction wall. Two actuators were placed
at each floor level and were attached at mid-span of the
slabs and spandrel beams running in the E–W direction.
The imposed displacements were measured with respect to
two independent steel frames (Fig. 10) using displacement
transducers. The two-storey building was fully instru-
mented with strain gauges and displacement transducers to
measure the deformations in the longitudinal reinforcement
in the beams, columns, and slabs, and in the transverse re-
inforcement in the columns and beams.
Undamaged dynamic properties
As the pseudo-dynamic testing method requires the devel-
opment of a precise numerical model for the structure,
forced-vibration tests (FVT) were carried out to calibrate
this model and to evaluate its performance (Paultre et al.
2003). The FVT consisted of subjecting the structure to a
sinusoidal horizontal harmonic load generated by an
eccentric-mass shaker. The recorded acceleration responses
were then used to extract various vibration properties (reso-
Can. J. Civ. Eng. Vol. 35, 2008
nant frequencies, mode shapes, and modal damping). Con-
crete blocks were used on each floor to simulate the added
mass representing the combined dead and live loads. The
E–W axis was chosen for the pseudo-dynamic earthquake
loading as well as the FVT. The shaker was placed on the
roof in a position chosen to excite both flexural and tor-
sional modes of vibration. Low-frequency force-balanced
accelerometers were used on each floor to record horizontal
accelerations in two orthogonal directions (parallel and per-
pendicular to the simulated earthquake loads). The FVT
were carried out with the shaker operating from 1 to 16 Hz,
with increments of 0.01 Hz.
Figure 11 illustrates typical frequency responses for the
building in the undamaged state, i.e., before any seismic ex-
citations were applied. Accelerations are plotted with respect
to frequency. The building’s resonant frequencies (flexure
and torsion) can be readily identified from the peaks of
these frequency response curves. The amplitude recorded at
position TNEW (top floor, northeast corner, measuring axis
parallel to the shaker force in the E–W direction) is plotted
in Fig. 11a. The building’s two resonant flexural frequencies
in the E–W direction (2.47 and 7.56 Hz) and the two tor-
sional frequencies (4.14 and 12.59 Hz) can be identified.
The corresponding modal damping ratios are then obtained
using the half-power bandwidth method and range from
0.76% of critical damping for the second flexural frequency
to 1.57% for the first flexural frequency. Response obtained
at position TNEN in the orthogonal axis (same position, but
axis perpendicular to the shaker force) is also illustrated in
Fig. 11b. The torsional modes identified in the top graph
(Fig. 11a) have a relatively stronger perpendicular compo-
nent in the N–S direction and exhibit larger resonant ampli-
tudes in that direction (4.14 and 12.59 Hz). The two
resonant flexural frequencies in the N–S directions that
were barely visible in graph (Fig. 11a) are now clearly appa-
# 2008 NRC Canada
Mousseau and Paultre
Fig. 9. Pseudo-dynamic flowchart.
rent (2.80 and 8.34 Hz). Table 4 presents the complete set of
undamaged dynamic properties for building.
Test procedure
Repeated pseudo-dynamic tests were carried out to ob-
serve the behaviour of the structure under increasing simu-
lated earthquake loading. At the end, a push-over test was
performed to evaluate the structure’s overstrength and struc-
tural ductility. After each level of seismic load applied to
the structure, the dynamic characteristics of the building
were obtained by forced-vibration tests. The structural signa-
ture was, therefore, obtained for each loading step and was
related to the level of damage in the structure (Paultre et al.
2007). Table 5 presents the chronology of the tests per-
formed on the building.
The lumped mass matrix for the pseudo-dynamic test is a
2
2 diagonal matrix with the masses concentrated at the
floor levels. The mass of the structure was provided by eight
1300 mm
1300 mm concrete blocks, four on each floor.
The thickness of the blocks was determined to represent
probable loading on the floor, i.e., 100% of design dead
loads and 50% of design live loads. Hence, the mass matrix
839
consisted of a 20 050 kg mass at DOF 1 and a 18 730 kg
mass at DOF 2. An initial damping matrix and stiffness ma-
trix need to be specified at the beginning of the pseudo-dy-
namic test. The damping matrix was based on results from
the FVT obtained from the structure in the undamaged state.
This initial damping matrix was computed to give damping
ratios of 1.57% and 0.76% in the two flexural modes in the
E–W direction (Table 4). The initial stiffness matrix for the
structure was determined by carrying out a static displace-
ment test on the building and the results were confirmed by
parameter identification after the first dynamic test.
The pseudo-dynamic tests were carried out in quasi-static
condition to prevent the generation of inertial forces
(accounted for in the computer model) and a maximum
velocity was specified for each actuator to control force re-
laxation during the test (Thewalt and Mahin 1991). Hence,
a maximum speed of 0.2 mm/s was used to move the
building. The time-integration scheme adopted is based on
the
-method (Hilber et al. 1977), which is an uncondi-
tionally stable implicit algorithm. A time step of 0.02 s
was used with the algorithm.
Ground motion time histories
Two different accelerograms were used for the pseudo-
dynamic test. The first ground motion was the S00E
component of the accelerogram recorded in El Centro,
California, during the 18 May 1940, Imperial Valley earth-
quake. The first 30 s time history for the input motion is
shown in Figs. 12b and 13e. This ground motion was
scaled to different intensities to meet the objectives of the
investigation.
The second time history used to carry out the PSD tests
was the M7R70A1 accelerogram generated for Montréal
and having a probability of exceedance of 2% in 50 years
(Atkinson and Beresnev 1998). This ground motion is com-
patible with the uniform hazard spectra (UHS) used in the
NBCC 2005 (NRCC 2005, Adams and Halchuk 2003). This
time history (Figs. 12a and 14c) was generated for a mo-
ment magnitude 7.0 earthquake at a hypocentral distance of
70 km. The peak ground acceleration (PGA) is 0.271g. Fig-
ure 12 presents 5% damped pseudo-acceleration response
spectra (PSA) for the ground motion used in the tests. This
last figure also illustrates the uniform hazard spectra (UHS)
for a very dense soil and soft rock (class C) site located in
Montréal according to the NBCC 2005.
Pseudo-dynamic and push-over test results
The earthquake responses of the structure, in terms of
storey displacements and base shear forces, are presented in
Figs. 13 and 14, respectively, for different seismic excitation
levels of the Imperial Valley earthquake (El Centro accel-
erogram). Figure 15 presents the response to the M7R70A1
accelerogram. The force–displacement hysteretic curves for
all seismic excitations are shown in Figs. 16 and 17 for
both floors (interstorey drift versus storey shear) and in
Figs. 18 and 19, where the global hysteretic behaviour of
the building is depicted in terms of base shear versus roof
displacement. Table 6 presents the maximum response ob-
tained during the pseudo-dynamic tests carried out on the
# 2008 NRC Canada
840 Can. J. Civ. Eng. Vol. 35, 2008

Fig. 10. Experimental setup for pseudo-dynamic tests.

Fig. 11. Measured frequency response functions, horizontal accel- Table 4. Dynamic characterization results for undamaged building.
eration: (a) east–west direction, north frame and (b) north–south
direction, east frame. Frequency, Period, Damping,
Mode f (Hz) T (s)  (%)
First flexural (E–W) 2.47 0.405 1.57
First flexural (N–S) 2.80 0.358 1.56
First torsional 4.14 0.242 1.21
Second flexural (E–W) 7.56 0.132 0.76
Second flexural (N–S) 8.34 0.120 0.86
Second torsional 12.59 0.079 1.14

versus displacement responses and the base shear versus

roof displacement are plotted in Figs. 16a and 18a, respec-
tively.
The maximum interstorey drifts were 6.5 and 5.2 mm for
the first and second storeys, which correspond to drift ratios
of 0.22% and 0.20%, respectively. According to the NBCC
2005 (NRCC 2005), drift ratios shall be limited to 1% for
post-disaster buildings, 2% for schools, and 2.5% for all other
buildings. The maximum base shear, V0max, was 68.1 kN dur-
ing the test.
The global behaviour of the building during the low seis-
mic excitation test was excellent. The inspection of the
structure after testing did not reveal any major cracks other
than hairline cracks due to the preliminary forced-vibration
tests, which had been performed before the first PSD test to
obtain initial dynamic properties (frequencies and mode
shapes).

building. The results are presented in chronological order of
the tests.
Response to low seismic excitation
The responses of the building to the El Centro ground
motion scaled to 0.078g PGA were basically within the line-
arly elastic range. According to the NBCC 1995 (NRCC
1995), this level of seismic loading corresponds to a 1% in
1 year probability level for Montréal. The storey displace-
ment time history is shown in Fig. 13a and the storey shear
Response to moderate seismic excitation
Moderate seismic excitation was achieved by scaling the
El Centro ground motion to 0.180g PGA. The structure’s
earthquake responses are shown in Figs. 13b, 15b, 16b, and
18b. Some period elongation can be observed from the re-
sponse history. All floor-displacement time histories are in
phase, pointing to a dominant first-mode participation.
Under this seismic input, the structure exhibited signifi-
cant cracking in the beams, columns, and slabs, but per-
formed with no measured yielding of the reinforcement and
# 2008 NRC Canada
Mousseau and Paultre 841

Table 5. Chronology of the tests carried out on the building.

Date Type of test Description of test

September 2000 Forced vibration Preliminary dynamic characterization of the building
September 2000 Forced vibration Dynamic characterization without the actuator brackets
January 2001 Forced vibration Dynamic characterization with the actuator brackets
April 2001 Pseudo dynamic Test with the El Centro* recording (ügmax = 0.078g)
April 2001 Pseudo dynamic Test with the El Centro recording (ügmax = 0.129g)
April 2001 Forced vibration Dynamic characterization after PSD (ügmax = 0.129g)
July 2002 Pseudo dynamic Test with the El Centro recording (ügmax = 0.180g)
August 2002 Forced vibration Dynamic characterization after PSD (ügmax = 0.180g)
October 2003 Forced vibration Dynamic characterization after PSD (ügmax = 0.180g){
April 2004 Forced vibration Dynamic characterization after PSD (ügmax = 0.180g){
July 2004 Pseudo dynamic Test with the El Centro recording (ügmax = 0.270g)
July 2004 Forced vibration Dynamic characterization after PSD (ügmax = 0.270g)
September 2004 Pseudo dynamic Test with the M7R70A1 accelerogram (ügmax = 0.271g)
September 2004 Pseudo dynamic Test with the El Centro recording (ügmax = 0.430g)
September 2004 Forced vibration Dynamic characterization after PSD (ügmax = 0.430g)
November 2004 Push over Push-over test until 
= 3.62
November 2004 Forced vibration Dynamic characterization after push-over test
Note: ügmax, maximum ground acceleration; PSD, pseudo-dynamic testing method; 
, maximum structural
ductility ratio.
*El Centro, Calif., during the 18 May 1940, Imperial Valley earthquake
{
New equipment.
{
New accelerometer positions.
Fig. 12. Spectral accelerations of (a) M7R70A1, and (b) El Centro drifts were 24.5 and 28.3 mm for the first and second stor-
earthquake north–south component scaled at different levels. PSA, eys, respectively, which are about 82% and 94% of the max-
pseudo-acceleration response spectra. imum 1% interstorey drift ratio allowed by the NBCC 2005
(NRCC 2005) for post-disaster buildings. Note that the max-
imum base shear measured just reached the design base
shear (99.8 kN) in the NBCC 2005 (NRCC 2005). The hys-
teretic curves confirm that no evident yielding had occurred
at this level of excitation. This PSD test indicates that HPC
performs well up to this moderate seismic excitation with
very little damage.

Response to severe seismic excitation

no spalling of the concrete covers. The building displayed a
maximum first-storey displacement of 24.5 mm and a maxi-
mum top-storey displacement of 52.4 mm. Maximum storey
Two different PSD tests were carried out at severe seismic
excitations. The first test was done with the El Centro ground
motion scaled to 0.270g PGA. As can be seen in Fig. 12b, the
pseudo-acceleration response spectra for this level of excita-
tion at the measured periods of vibration of the building is
clearly higher than the design spectrum for Montréal.
The responses of the structure for the first test are shown
in Figs. 13c, 15c, 16c, and 18c. A vibration period of ap-
proximately 0.68 s can be observed from the storey dis-
placement time history. Maximum interstorey drifts were
34.6 and 41.9 mm for the first and second storeys, which
correspond to drift ratios of 1.15% and 1.40%, respectively.
These values exceed the drift ratio allowed by the NBCC
2005 (NRCC 2005) for post-disaster buildings, but are lower
than the drift ratios for schools and for all other buildings.
Maximum base shear developed during the test was
165.1 kN, which is significantly higher than the design base
shear (99.8 kN).
The severe seismic excitation test caused the first meas-
ured yielding of reinforcement steel. Measured strains in the
longitudinal reinforcement at the bottom of the first storey
column reached 1.18y, where y is the yield strain of rein-
forcing bars. At the time of the first yielding, the base shear
was 144.4 kN.
# 2008 NRC Canada
842
Fig. 13. Time histories of first floor and roof displacements for El
Centro earthquake, N–S component scaled to (a) 0.078g peak
ground acceleration (PGA), (b) 0.180g PGA, (c) 0.270g PGA,
(d) 0.430g PGA, and (e) N–S component of horizontal ground ac-
celeration of El Centro earthquake.
The structure performed very well during the test. Few
new cracks developed, but existing cracks widened. Neither
spalling of the cover nor local instabilities of reinforcement
were observed.
The second test was carried out with the M7R70A1 accel-
erogram generated for Montréal (see Fig. 12a). The peak
ground acceleration of the accelerogram is 0.271g. The PSD
test results obtained with the accelerogram are shown in
Can. J. Civ. Eng. Vol. 35, 2008
Fig. 14. Time histories of first floor, roof displacements, and base
shear for the M7R70A1 time history scaled to 0.271g peak ground
acceleration.
Figs. 14, 17, and 19. The maximum interstorey drifts were
17.7 and 21.6 mm for the first and second storeys, respec-
tively. These values are less than the limits imposed by the
NBCC 2005 (NRCC 2005) for all types of structures. Fur-
thermore, as shown in Fig. 14, the maximum base shear de-
veloped was only 76.3 kN, significantly lower than the
design base shear (99.8 kN) in the NBCC 2005.
No new cracks appeared during the severe seismic excita-
tion test. Existing cracks opened (and closed) in the critical
regions of the beams and columns according to the displace-
ments imposed on the structure by the seismic forces. Meas-
ured displacements and storey shears were clearly lower
than those measured during the PSD test with the El Centro
ground motion scaled to 0.270g PGA, even though the max-
imum ground accelerations of these two seismic functions
are almost identical. This result highlights that the high fre-
quency content of Eastern Canada earthquakes are not crit-
ical for flexible structures.
Response to very severe seismic excitation
The last pseudo-dynamic test was performed with the El
Centro ground motion scaled to 0.43g PGA. The PGA level
corresponds to a 2% in 50 year probability level for Mon-
tréal (Adams and Halchuk 2003). Even though the PGA
value is in accordance with the prescribed acceleration for
Montréal, the pseudo-acceleration response spectra for the
very severe seismic excitation is clearly higher than the de-
# 2008 NRC Canada
Mousseau and Paultre
Fig. 15. Time histories of base shear for El Centro earthquake
north–south (N–S) component scaled to (a) 0.078g peak ground ac-
celeration (PGA), (b) 0.180g PGA, (c) 0.270g PGA, (d) 0.430g
PGA, and (e) N–S component of horizontal ground acceleration of
El Centro earthquake.
sign spectra of Montréal for all given frequencies (Fig. 12b).
This excitation level was used to study the behaviour of the
structure under very severe seismic excitation.
The earthquake responses of the structure are shown in
Figs. 13d, 15d, 16d, and 18d. An important period elonga-
tion can be observed from the response history. The meas-
ured vibration period is about 0.85 s. The maximum
interstorey drifts were 50.9 and 67.5 mm for the first and
843
second storeys, which corresponds to drift ratios of 1.70%
and 2.25%, respectively. The values exceed the drift ratio al-
lowed by the NBCC 2005 (NRCC 2005) for post-disaster
buildings. Furthermore, the interstorey drift of the second
floor also exceeds the limit of 2% imposed on school build-
ings. On the other hand, the limit concerning all other build-
ing categories was not exceeded. Maximum base shear was
204.3 kN during the test, which is more than twice the de-
sign base shear.
The hysteretic curves in Fig. 16d show a light pinching of
the hysteresis. Furthermore, analysis of the curves reveals
the structure has significant inelastic behaviour. The longitu-
dinal reinforcement in the beams, near the column faces,
and at the base of the first-storey columns suffered inelastic
tensile strains. The column longitudinal bars, below the first-
floor beam–column joints, and all the longitudinal bars in
the second-storey columns remained in the elastic range of
the steel reinforcement.
This intensity level caused some damage to the structure.
Several new cracks in the beams, columns, slabs, and joints
appeared and most of the existing cracks widened signifi-
cantly. On the other hand, no spalling of the cover was ob-
served and, hence, it can be implied that no local
instabilities of reinforcement took place. Despite the very
high level of excitation, the building preserved its structural
integrity and its capacity to sustain gravity loads.
Response to push-over
After the last very severe seismic excitation PSD test, a
push-over test was carried out on the building. The objective
of the test was to verify the ultimate capacity of the struc-
ture, thereby measuring its overstrength and ductility levels.
The structure was tested for incrementally increasing lateral
loads. The shape of the lateral load distribution applied to
the structure was calculated as combinations (square root of
sum of squares) of the lateral load distributions obtained
from modal analyses of the structure with the NBCC 2005
(NRCC 2005) design spectra.
Figure 20 illustrates the building’s response during the
push-over test. It shows the overall hysteretic behaviour of
the building in terms of lateral load versus roof displace-
ment. The maximum positive displacement of the roof was
183.1 mm, with a corresponding base shear of 237.6 kN. In
contrast, the maximum negative roof displacement was
203.3 mm, with a corresponding base shear of 233.7 kN.
Obviously, these displacements exceeded the drift ratio al-
lowed by the NBCC 2005 for all buildings. The recorded
base shear was 2.38 times greater than the design base shear
and surpassed the base shear recorded during the pseudo-
dynamic test with the El Centro ground motion scaled to
0.43g PGA.
Spalling of the cover was observed at the base of the first-
floor column during the push-over test. Furthermore, severe
spalling of the concrete cover occurred in the spandrel
beams of the first floor perpendicular to the loading (N–S)
under significant diagonal torsional compression stresses.
This is indicative of torsional yielding, which limits the ef-
fective width of the slab contributing to the negative bend-
ing resistance of the main beam in the seismic loading
direction as was explained by Di Franco et al. (1995). Even
at this level of loading and after several pseudo-dynamic
# 2008 NRC Canada
844 Can. J. Civ. Eng. Vol. 35, 2008

Fig. 16. Shear-displacement response for first storey and second storey for El Centro earthquake north–south component scaled to
(a) 0.078g peak ground acceleration (PGA), (b) 0.180g PGA, (c) 0.270g PGA, and (d) 0.430g PGA.

# 2008 NRC Canada

Mousseau and Paultre
Fig. 17. Shear–displacement response for (a) first storey and
(b) second storey for M7R70A1 time history scaled to 0.271g peak
ground acceleration.
tests, the building maintained its structural integrity and
gravity-load-carrying capacity. In addition, according to the
shape of the hysteretic curve, the structure had reserve
strength and ductility at the end of the push-over test.
Analysis of test results
According to Table 6, the base shear increased from
68.1 kN (9.0%W) to 204.3 kN (26.9%W) during the pseudo-
dynamic tests carried out on the building, representing a
300% increase. For the same tests, roof displacement in-
creased by approximately 1000%. Structural displacement
ductility demand for the PSD tests progressed from 0.21 to
2.11. Even with the very severe seismic excitation test, the
building’s overall behaviour was still very good.
Table 7 presents the maximum interstorey drift ratios re-
corded during the pseudo-dynamic tests. The drift limit of
1% for post-disaster buildings was exceeded during the
0.180g PGA test with the El Centro ground motion, and the
drift ratio for schools (2%) was reached at the second storey
during the PSD test with 0.43g PGA. As for the 2.5% limit
prescribed for the category of all other buildings, it was
never exceeded during the pseudo-dynamic tests carried out
on the building.
Overstrength
The experimental estimation of overstrength for a com-
plete building with several degrees of freedom is linked to
845
the definition of the concept itself. A very conservative ap-
proach is to consider the base shear recorded during first
yielding of a member. In this experimental investigation,
first yielding occured in a ground-storey column at a base
shear of 144.4 kN during the PSD test with the El Centro
ground motion scaled to 0.270g PGA. Therefore, a conser-
vative overstrength factor is equal to 144.4/99.8 = 1.45. On
the other hand, this approach does not take into account the
fact that, even after yielding of some structural elements,
buildings have important reserve strength and the capacity
to sustain additional loads. Mitchell and Paultre (1994) esti-
mated the overstrength of reinforced concrete structures
when the drift ratios reached the limit permitted by the
NBCC. In the 2005 edition of the code (NRCC 2005), the
limit for all other buildings is 0.025 times the storey height.
This interstorey drift was obtained during the push-over test
(Fig. 20). The corresponding base shear was 225.6 kN. Us-
ing this value of base shear, the overstrength factor is 225.6/
99.8 = 2.26. For a moderately ductile moment-resisting
frame, the overstrength-related force-modification factor,
Ro, is 1.4 in the NBCC 2005. Experimental results ob-
tained in this investigation seem to confirm the validity
and safety of this factor.
Displacement ductility
Displacement ductility, as used in this paper, is equal to
the ratio of maximum displacement reached by the roof to
the displacement at general yielding as used in Mitchell and
Paultre (1994). The general yielding level was determined
by using a bilinear idealization of the building response
(Fig. 21). The initial stiffness was determined as the average
of the initial force–displacement slope of the first three low-
level pseudo-dynamic test results. The second segment of
the bilinear representation was determined as the tangent to
the post-yield force–displacement response obtained from
the push-over test. With the bilinear idealization, the roof
displacement corresponding to general yielding,
y, was
56.1 mm. Using this displacement, the maximum structural
ductility ratio, 
can be obtained from

max
½11 
¼

y
where Dmax is the maximum measured displacement at roof
level.
During the push-over test, the structural displacement
ductility reached 3.62, which is more than the ductility-re-
lated force-modification factor, Rd. This factor is 2.5 in the
NBCC 2005 (NRCC 2005) for a moment-resisting frame
with moderate ductility.
During the pseudo-dynamic tests, the structural displace-
ment ductility demand progressed from 0.21 to 2.11
(Table 6). After these tests, a push-over test was carried out
on the structure. The structural ductility reached 
= 3.62.
At this level, the interstorey drift limit for all building usage
was exceeded. When considering this limit, the structural
ductility was 2.67. All these values exceed the ductility-
related force-modification factor, Rd, in the NBCC 2005
for this seismic-force-resisting system chosen. Furthermore,
these results demonstrate that the appropriate use of high-
strength concrete in seismic zones seems safe and justified
# 2008 NRC Canada
846 Can. J. Civ. Eng. Vol. 35, 2008

Fig. 18. Base shear–roof displacement response for El Centro earthquake north–south component scaled to (a) 0.078g peak ground accel-
eration (PGA), (b) 0.180g PGA, (c) 0.270g PGA, and (d) 0.430g PGA. VNBCC-05, design base shear according to the 2005 NBCC (NRCC
2005).

Fig. 19. Base shear-roof displacement response for M7R70A1 time Fig. 20. Base shear–roof displacement response for push-over test.
history scaled to 0.271g peak ground acceleration.

if the applicable standards and particularly the detailing re- crete, the building has shown excellent ductility without
quirements are respected. During the last PSD test (El showing signs of major failure.
Centro ground motion scaled at 0.43g PGA) and push-over
test, the average compressive strength of the concrete used
in the building was 73 MPa. Despite the very high level of
Conclusions
excitation and the high compressive strength of the con- Forced vibrations were used to extract the key dynamic
# 2008 NRC Canada
Mousseau and Paultre 847

Table 6. Maximum response parameters recorded during pseudo-dynamic testing method tests.

u1max u2max (u2–u1)max f1max f2max V0max M0max

Accelerogram (mm) (mm) (mm) (kN) (kN) (kN) (kNm) 

El Centro (ügmax = 0.078g) 6.5 11.8 5.2 39.2 42.5 68.1 306.4 0.21
El Centro (ügmax = 0.180g) 24.5 52.4 28.3 68.0 104.1 123.8 661.5 0.93
El Centro (ügmax = 0.270g) 34.6 75.9 41.9 93.3 130.5 165.1 837.1 1.35
M7R70A1 (ügmax = 0.271g) 17.7 39.2 21.6 68.7 66.9 76.3 384.8 0.70
El Centro (ügmax = 0.430g) 50.9 118.4 67.5 110.5 169.0 204.3 1031.4 2.11
Note: u1max, maximum displacement of DOF 1 (first floor); u2max, maximum displacement of DOF 2 (roof); f1max, maximum first
storey shear; f2max, maximum second storey shear; M0max, maximum base overturning moment.

Table 7. Maximum interstorey drift ratios recorded. as gauges for expected damage levels corresponding to
sectional behaviour.
Interstorey drift ratios (%)
Accelerogram Storey 1 Storey 2 Acknowledgements
El Centro (ügmax = 0.078g) 0.22 0.20 The authors gratefully acknowledge the financial support
El Centro (ügmax = 0.180g) 0.82 0.94 of the Natural Sciences and Engineering Research Council
El Centro (ügmax = 0.270g) 1.15 1.40 of Canada. They would also like to thank Claude Aubé, Sé-
M7R70A1 (ügmax = 0.271g) 0.59 0.72 bastien Gauthier, and Laurent Thibodeau, technicians in the
El Centro (ügmax = 0.430g) 1.70 2.25 structural laboratory of the Department of Civil Engineering
at the University of Sherbrooke, for their role in the test
Fig. 21. Bilinear idealization of building behavior for push-over preparation. The help of Martin Beauvilliers, Najib Bouaa-
test. PSD, pseudo-dynamic testing method. nani, Jean-Christophe Kombila, and Olivier Lefebvre is also
acknowledged. Charles Savard and Tommy Prevost partici-
pated in the design and construction of the structure. This
research was carried out as part of the work by the second
author on the CSA A23.3-04 seismic design subcommittee.

properties of the structure (vibration frequencies and mode
shapes as well as modal damping ratios) to be used in PSD
tests and to set the initial state of the structure. A suite of
PSD tests with increasing seismic excitations demonstrated
the excellent performance of the high-strength concrete
structure designed using the new CSA A23.3-04 Design of
concrete structures standard (CSA 2004) under earthquake
excitations. In particular, displacement ductility and over-
strength are higher than the prescribed values in the new
NBCC 2005 (NRCC 2005). The HSC columns designed
with the new requirements of the A23.3-04 standard per-
formed very well under PSD tests and push-over testing,
even at drift levels significantly larger than the limits al-
lowed by the NBCC 2005. These results confirm the appro-
priateness of the new CSA A23.3-04 standard, in particular
(i) the new strong column – weak beam requirements and
(ii) the new confinement requirements for columns. The test
results also confirm the safe values of the ductility-re-
lated force-modification factor and the overstrength-re-
lated force-modification factor. These unique test results
can be used for the calibration of numerical models and
References
Adams, J., and Halchuk, S. 2003. Fourth-generation seismic hazard
maps of Canada: Values for over 650 Canadian localities in-
tended for the 2005 National Building Code of Canada. Open
File 4459, Structural engineering series. Earthquakes Canada,
Geological Survey of Canada, Ottawa, Ont.
Atkinson, G.M., and Beresnev, I.A. 1998. Compatible ground-
motion time histories for new national seismic hazard maps.
Canadian Journal of Civil Engineering, 25(2): 305–318.
doi:10.1139/cjce-25-2-305.
CSA. 1984. Design of concrete structures for buildings. CSA stan-
dard CSA-A23.3-84. Canadian Standards Association, Rexdale,
Ont.
CSA. 1994. Design of concrete structures. CSA standard CSA-
A23.3-94. Canadian Standards Association, Rexdale, Ont.
CSA. 2004. Design of concrete structures. CSA standard CSA-
A23.3-04. Canadian Standards Association, Mississauga, Ont.
Di Franco, M.A., Mitchell, D., and Paultre, P. 1995. Role of span-
drel beams in slab-beam-column connections subjected to seis-
mic loading. Journal of Structural Engineering, ASCE, 121(3):
408–419. doi:10.1061/(ASCE)0733-9445(1995)121:3(408).
Hilber, H.M., Hughes, T.J.R., and Taylor, R.L. 1977. Improved nu-
merical dissipation for time integration algorithms in structural
dynamics. Earthquake Engineering & Structural Dynamics,
5(3): 283–292. doi:10.1002/eqe.4290050306.
Mitchell, D., and Paultre, P. 1994. Ductility and overstrength in
seismic design of reinforced concrete structures. Canadian Jour-
nal of Civil Engineering, 21(6): 1049–1060. doi:10.1139/l94-
109.
Mousseau, S., and Paultre, P. 2005. Esssais pseudo-dynamiques et
dynamiques d’un bâtiment de deux étages en béton à hautes per-
formances. CRGP-2005/01. Centre de recherche en génie para-
sismique et en dynamique des structures, Departement of Civil
# 2008 NRC Canada
848
Engineering, Université de Sherbrooke, Sherbrooke, Que. [In
French.]
Mousseau, S., Paultre, P., and Mazars, J. 2008. Seismic perfor-
mance of a full-scale, reinforced high-performance concrete
building. Part II: Analytical study. Canadian Journal of Civil En-
gineering, 35(8): 849–862. doi:10.1139/L08-019.
NRCC. 1985. National building code of Canada. National Research
Council of Canada, Ottawa, Ont.
NRCC. 1995. National building code of Canada. National Research
Council of Canada, Ottawa, Ont.
NRCC. 2005. National building code of Canada. National Research
Council of Canada, Ottawa, Ont.
Paultre, P., and Mitchell, D. 1987. Evaluation of seismic perfor-
mance of concrete frame structures in Canada. Structural engi-
neering series, Report 87–4. McGill University, Montréal, Que.
Paultre, P., Castele, D., Rattray, S., and Mitchell, D. 1989. Seismic
response of reinforced concrete frame subassemblages — a Ca-
nadian code perspective. Canadian Journal of Civil Engineering,
16(5): 627–649. doi:10.1139/l89-097.
Paultre, P., Proulx, J., Mousseau, S., Prévost, T., and Savard, C.
Can. J. Civ. Eng. Vol. 35, 2008
2003. Pseudo-dynamic and forced vibration tests of a full-size
two-story reinforced high-performance concrete building. In
Large-scale structural testing, SP-211. Edited by M.A. Issa and
Y.L. Mo. American Concrete Institute, Farmington Hills, Mich.
Paper SP-211-7, pp. 135–160.
Paultre, P., Weber, B., Mousseau, S., and Proulx, J. 2007. Measur-
ing earthquake damages in a high-strength concrete structure.
Report CRGP-2007/05. Centre de recherche en génie parasismi-
que et en dynamique des structures, Université de Sherbrooke,
Que.
Takanashi, K. 1975. Nonlinear earthquake response analysis of
structures by a computer-actuator on line system. Bulletin of
the Earthquake Resistant Structure Research Center. No. 8. Insti-
tute of Industrial Science, University of Tokyo, Tokyo, Japan.
pp. 205–219.
Thewalt, C.R., and Mahin, S.A. 1991. The pseudo-dynamic test
method: Verification and extensions. In Experimental and nu-
merical methods in earthquake engineering. Edited by J. Donea
and P. M. Jones. Kluwer Academic Publishers, the Netherlands.
pp. 101–118.
# 2008 NRC Canada