You are on page 1of 14

Displacement-Based Method of Analysis for Regular

Reinforced-Concrete Wall Buildings: Application


to a Full-Scale 7-Story Building Slice Tested
at UC–San Diego
Marios Panagiotou, M.ASCE1; and José I. Restrepo, M.ASCE2
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

Abstract: This paper describes a displacement-based method of analysis for the performance-based seismic design of regular buildings
with reinforced-concrete bearing walls acting as the lateral-force resisting system. The method considers two performance levels:
immediate occupancy (IO) and life safety (LS), each anchored at a specific seismic hazard level. It explicitly accounts for the
combined effects of inelastic first mode of response, kinematic system overstrength, and higher modes of response. Quantification of
these effects is required to capacity-protect the structure and to ensure the intended performance. As an example, the method is applied
to a full-scale 7-story reinforced-concrete building slice, built and tested on the George E. Brown Jr. Network for Earthquake Engineering
Simulation Large Outdoor High-Performance Shake Table at the University of California–San Diego. The response of the test
building largely verified the method discussed in this paper. DOI: 10.1061/(ASCE)ST.1943-541X.0000333. © 2011 American Society
of Civil Engineers.
CE Database subject headings: Displacement; Load-bearing capacity; Walls; Lateral forces; Reinforced concrete; Seismic design;
Multistory buildings; Concrete structures; California.
Author keywords: Displacement-based design; Performance-based design; Capacity design; Load-bearing walls; Lateral forces;
Reinforced concrete; Seismic design; Three-dimensional effects; Higher-mode effects.

Introduction Displacement-based methods differ in how they determine


the fundamental period of the building, use of elastic design
Reinforced-concrete (RC) walls have an excellent reputation for spectrum, and an equivalent viscous damping ratio or inelastic
resisting earthquake-induced lateral forces in low- to medium-rise design spectrum. Among the different methods, only Priestley et al.
buildings in high-seismic regions (Wood 1991; Fintel 1997) and (2007) includes specific recommendations to estimate the
are a preferred lateral-force resisting system in many parts of
higher-mode effects. A review of the literature indicated that
the world. Since the advent of ductile seismic design, the methods
none of these methods explicitly quantifies the effect of the
used in the design of RC walls recognize the inherent ductility in
three-dimensional (3D) kinematic-interaction between walls and
such structural elements (Paulay et al. 1990; Paulay and Priestley
1992). Derived lateral forces required for elastic response are the elements framing to them, defined herein as kinematic system
divided by a response modification coefficient. Recently, a number overstrength.
of displacement-based (DB) methods of analysis have been This paper describes a DB method of analysis for performance-
proposed for determining design lateral forces for RC wall build- based seismic design of regular RC wall buildings. The method
ings. Many of these methods incorporate the principles of concrete considers two performance levels, immediate occupancy (IO) and
mechanics and structural dynamics, which make them quite life safety (LS). In comparison with previous DB approaches, this
suitable for performance-based seismic design. A discussion of method explicitly accounts for the combined effects of the
the different methods can be found in Moehle (1992), Fédération inelastic first mode of response, kinematic system overstrength, and
Internationale du Beton (FIB) (FIB 2003), Sullivan et al. (2003), higher modes of response. This allows for a better estimation of
and Englekirk (2007). different response quantities required to capacity-design the walls
and to ensure a satisfactorily performance at LS. The first part of the
method, which accounts for the inelastic first-mode response,
1 builds on existing substitute structure methods that use inelastic
Assistant Professor, Dept. of Civil and Environmental Engineering,
Univ. of California–Berkeley, Berkeley, CA (corresponding author). design spectra (e.g., Fajfar 2000; Chopra and Goel 2001). The
E-mail: panagiotou@berkeley.edu method is specifically employed in the paper to determine the
2
Professor, Dept. of Structural Engineering, Univ. of California–San design lateral forces for a full-scale slice of a 7-story reinforced-
Diego, La Jolla, CA. concrete residential building, tested under input ground-motion
Note. This manuscript was submitted on April 14, 2009; approved on excitation in the George E. Brown Jr. Network for Earthquake
September 26, 2010; published online on October 21, 2010. Discussion
Engineering Simulation Large Outdoor High-Performance Shake
period open until November 1, 2011; separate discussions must be sub-
mitted for individual papers. This paper is part of the Journal of Structural Table (LHPOST) at the University of California–San Diego. Key
Engineering, Vol. 137, No. 6, June 1, 2011. ©ASCE, ISSN 0733-9445/ results of the test program are presented in an accompanying paper
2011/6-677–690/$25.00. (Panagiotou et al. 2011).

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 677

J. Struct. Eng. 2011.137:677-690.


Description of Method of the regions above the potential plastic hinges and adequate shear
strength and anchorage of the reinforcement along the entire length
Performance Levels of the wall.
In the proposed method, the spread of plasticity in the walls of
The proposed method considers two performance levels, IO and the building is deliberately precluded from extending beyond the
LS, each of which is anchored to a specific seismic-hazard level. first floor by special detailing of the longitudinal reinforcement.
Specific limit states are associated with each performance level, The spread of plasticity in the wall is constrained for two reasons:
as described in the following. Apparently, the limit states can be (1) to optimize construction, and (2) to reduce the uncertainty in
modified according to the design performance objectives. estimating the plastic hinge length. Constraining the plastic hinge
IO is anchored to a seismic hazard level associated with fre- length to within the base of the wall allows for special reinforcing
quently occurring earthquakes of low intensity, with 50% probabil- detailing only in the bottom part of the walls. Building on the Hines
ity in exceedance in 50 years (return period of 72 years). The limit et al. (2004) observation that there are large variations between
states selected for this performance at IO are such that no visual measured equivalent plastic hinge lengths in walls and those esti-
damage occurs in the building. This is translated into the following mated for using existing approaches, it is expected that for this type
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

strains and interstory drift limits: (1) maximum tensile strain of 1% of plastic hinge, the shape of plastic curvatures in the wall tends
in the wall longitudinal reinforcement; (2) maximum compressive toward being rectangular rather than linear, as is usually the case
strain of 0.4% in the concrete; and (3) maximum interstory drift when plasticity spreads unconstrained. The design procedure is
ratio of 1%. The interstory drift-ratio limit is set to control damage performed in the four steps described as follows.
to nonstructural elements (Lang and Restrepo 2006).
LS is anchored to a seismic hazard level associated with the Step 1: First-Mode Lateral Forces
design basis earthquake (DBE) (ASCE-7 2006). Extensive yielding
This part of the method determines the design-bending moments
and nonlinear inelastic response are anticipated for the building
at the base of the walls by using lateral forces derived from the
walls responding to a rare and strong-intensity earthquake. There-
first mode of response. Using code-prescribed load combinations
fore, structural damage is accepted in critical regions in the build-
and strength reduction factors, the longitudinal reinforcement
ing, in this case, at the base of the critical walls. Limit states
in the walls at their bases is established. The proposed method
associated with expected building performance at LS are (1) maxi-
employs an inelastic single degree-of-freedom (SDOF) approxi-
mum tensile strain of 5% in the longitudinal reinforcement, and
mation at LS and uses the effective stiffness at yield. This
(2) maximum roof drift ratio of 3%. In this procedure, the concrete
approach differs from previously proposed methodologies for
compressive strain is not considered a limiting strain for the design
RC wall buildings (e.g., Priestley et al. 2007), which use an equiv-
of walls. Instead, this strain is calculated from the curvature de-
alent linear-elastic SDOF oscillator with secant stiffness properties
mand required by the governing limit state. Transverse reinforce-
at maximum response and an equivalent viscous damping ratio.
ment is provided in the potential plastic hinge regions of the walls
This part of the method is more aligned with the substitute-
to ensure that the compressive strain demand is met. The maximum
structure methods that use inelastic design spectra (e.g., Fajfar
tensile strain limit is set to avoid premature fracture of the longi-
2000; Chopra and Goel 2001).
tudinal reinforcement, and the maximum roof drift ratio limit is
Because of space limitations, not all the variables in the
used to restrict large deformation (P-delta) effects.
equations that follow are defined. (For details, see the “Notation”
Assumptions and Limitations of Method section.) For cantilever wall buildings, the shape of the first mode is
approximated by the following polynomial expression:
The method of analysis discussed in this section makes the follow-      
ing four assumptions: (1) the bending moment at the critical section 1 hi 5 10 hi 3 20 hi 2
Φ1;i ¼  þ ð1Þ
at the base of the cantilever walls, where plastic hinges will ulti- 11 H 11 H 11 H
mately develop, is attributable to the first mode of response only;
(2) all walls are cracked, and no tension stiffening exists in the This mode shape is obtained from the deformed shape of a pris-
reinforced-concrete walls; (3) the effect of kinematic system over- matic cantilever wall subjected to distributed lateral forces whose
strength, as defined in the following, is not accounted for to deter- magnitude is directly proportional to the height. This mode shape
mine the required base bending moment strength of the walls; does not differ significantly from that of a cantilever wall of uni-
(4) the lateral deformations in the building are solely caused by form mass and reduced stiffness at its base (Panagiotou 2008). The
the first mode of response. The second assumption results in a mode shape allows for the computation of the first-mode modal
conservative design, particularly in regions of a low seismic hazard, weight, W e;1 , participation factor, Γ1 , and contribution factor, Γ1 .
in which the likelihood of significant cracking caused by ground P
ð n W i Φ1;i Þ2
shaking at IO is low. The fourth assumption limits the number W e;1 ¼ Pn i¼1 ;
i¼1 W i ðΦ1;i Þ
2
of stories in a building to which the method proposed is suitable:
Pn ð2Þ
as the number of floors increases, the participation of the higher
i¼1 W i Φ1;i 
modes of response increases to the extent that their contribution Γ1 ¼ Pn and Γ1 ¼ Γ1 Φ1;n
i¼1 W i ðΦ1;i Þ
2
to lateral displacements and more importantly, to interstory drifts,
becomes nonnegligible. Target roof displacements are calculated for the limit states
governing the IO and LS performance objectives. For IO, the
Mechanism of Inelastic Deformation
two strain-limit states are met by conservatively assuming that
The preferred mechanism of inelastic deformation in cantilever the roof displacement in the critical wall, the longest in a regular
walls is through the development of plastic hinges at their bases. building (Paulay and Restrepo 1998), does not exceed the yield
Capacity design (Park and Paulay 1975) is used to ensure that this roof displacement, δ y;r , that is, δ IO;r ≤ δy;r . This assumption has
mechanism develops and is maintained throughout. Other plastic the advantage of not requiring the choice of a value for the plastic
mechanisms are deliberately precluded from developing through hinge length. For such moderate plastic strains, the plastic hinge is
a design-strength hierarchy that requires adequate flexural strength beginning to spread and has not necessarily reached its maximum

678 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.


length (Hines et al. 2004). The yield displacement for the prismatic of 5%, δLS;r is computed from Eq. (5b), and the corresponding
cantilever examined to obtain the mode shape is displacement ductility is μΔ ¼ δLS;r =δy;r .
The following step in the method requires converting the limit-
11 11 H 2 ing IO and LS roof displacements into equivalent displacements,
δ y;r ¼ φy H 2 ¼ λy ð3Þ
40 40 Lw δ IO;e and δ LS;e , respectively, of linear single-degree-of-freedom
oscillators [see Fig. 1(e)].
where φy ¼ λy =Lw = idealized yield curvature (Priestley et al.
2007). When the yield displacement is reached and the building’s
lateral displacement is compatible with the first-mode shape previ- δy;r δ LS;r
δ IO;e ¼ and δLS;e ¼ ð6Þ
ously presented, the maximum interstory drift ratio is Γ1 Γ1 C μ
15 δ y;r 3 H
θy;r ¼ ¼ λy ð4Þ
11 H 8 Lw The definition of δIO;e is taken from Chopra (2001). The defi-
nition for δ LS;e is based on the empirical relationships between the
The first definition given in Eq. (4) relates the interstory drift displacement responses of inelastic and elastic oscillators of equal
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

ratio to the roof drift ratio, while the second definition relates initial periods (Chopra and Chintanapakdee 2001; Ruiz-García and
the interstory drift ratio and the geometrical aspect ratio of the wall. Miranda 2003, 2006). The maximum periods T IO and T LS corre-
As pointed out by Priestley et al. (2007), coefficient λy depends spond to the critical limit states at IO and LS, respectively, and are
essentially on the cross section and the yield strain of the longitu- computed using the equivalent displacements, as shown in Fig. 2.
dinal reinforcement, although it is independent of the longitudinal The smallest of the periods obtained indicates the performance
reinforcement ratio and axial load ratio. For λy ¼ 0:0034, and level that controls the design of the building. In the final design,
checking that θy;r is equal to or less than the IO interstory drift limit the fundamental period of the building, calculated ignoring tension
state of 1%, the geometric aspect ratio of the web wall calculated by stiffening, must be T 1 ≤ T D , where T D ¼ minðT IO ; T LS Þ. When
solving Eq. (4) is H=Lw ≤ 7:8. Given the wall geometry, the roof T LS > T IO , the LS strain limit may be revised iteratively until
displacement is estimated using Eq. (3). The roof displacement, T LS ≈ T IO . This allows relaxation of the reinforcing detailing
δ LS;r , calculated for LS considers the elastic and plastic contribu- in the potential plastic hinge regions. Once the period T D is
tions (see Fig. 1). This displacement is given by established, the first-mode design base shear, V b;1 , is calculated
M LS as follows:
δ LS;r ¼ δ þ δ p;r
M n y;r
     
11 M LS H ℓp ℓp δ y;r 2π 2 W e;1
¼ þ ðμϕ;LS  1Þ 1  λH ð5a Þ V b;1 ¼ ð7Þ
40 M n Lw 2H Lw y Γ1 T D g

or in a simpler form
  The first-mode base shear is distributed in lateral forces F 1;i in
11 H ℓp proportion to W i Φ1;i (Chopra 2001). The design-bending moment
δ LS;r ≈ þ ðμϕ;LS  1Þ λH ð5b Þ
40 Lw Lw y at the wall base, M u , is computed by taking the moments of the
lateral forces about the wall base. The design of a building with
where the LS curvature ductility μϕ;LS ¼ φLS =φy in the critical equal length walls is rather simple. In buildings with walls of differ-
section of the wall is calculated using the curvature φLS , which cor- ent geometries, the design engineer apportions the lateral forces
responds to the governing LS-limit state. Ignoring the influence among the different walls. Paulay and Restrepo (1998) commented
of ratios M LS =M n and ℓp =H, the simplified Eq. (5b) is obtained. that a distribution method based on elastic properties of the gross
The roof drift ratio at LS is calculated based on Eq. (5a) or (5b) section of a wall is a possible way to achieve this, but it is not nec-
as θy;r ¼ δ LS;r =H For a target μϕ;LS , satisfying the strain limit state essarily the only or the most effective method.

Fig. 1. First-mode bending-moment diagrams, curvature diagram, target displacements, and SDOF target displacements

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 679

J. Struct. Eng. 2011.137:677-690.


IO
Sa,2
LS

δ
LS,e
Sd (m)

Sa (g)
Sa aLS
1
Sd
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

aIO 1
δ
IO,e

0
T2 TIO TLS
T (s)

Fig. 2. Design acceleration and displacement-response spectrum

Step 2: System Static Overstrength Lateral Forces factor found for the load combination, resulting in the minimum
axial force in the wall.
The system static overstrength lateral forces (SSOLF) are those
lateral forces greater than the design lateral forces that are required System Mechanism
to push the critical building wall (with its actual boundary condi- Three-dimensional interaction effects caused by deformation com-
tions and longitudinal reinforcement specified at the critical sec- patibility between walls and elements framing into them have been
tions) to the lateral displacement corresponding to LS. This studied both experimentally (Paulay and Taylor 1981; Bertero et al.
analysis uses expected rather than specified material properties. 1985), and numerically (Ozcelik et al. 2008; Balkaya and Kalkan
In practice, the SSOLF could be determined using an adaptive 2003, 2004). This interaction increases the lateral system resistance
pushover analysis of a representative mathematical model built of reinforced-concrete-wall buildings. It also causes a large in-
with nonlinear analysis software (Reinhorn 1997; Satyarno et al. crease in the shear-force demand in walls. Because such forces
1998). It is useful to estimate these forces as the addition of arise from deformation compatibility, this source of additional
two independent and additive mechanisms: (1) an element mecha- strength is referred to here as kinematic system overstrength.
nism that assesses the effects of flexural overstrength at the critical This section, based on plastic analysis, presents a way to explic-
section of the wall, and (2) a system mechanism that assesses the itly quantify the forces that arise in RC wall buildings because of
effects caused by kinematics on the elements framing onto the kinematic overstrength. Two structural configurations are consid-
laterally deformed wall. These two mechanisms are examined in ered. Fig. 3 depicts the plan view and the deformed side view
the next section. of two structural configurations. In both cases, structural walls,
which provide all the lateral force resistance in the direction of
Element Mechanism excitation, and gravity columns exist. The walls are designed to
The excess of flexural strength at the critical wall section at LS develop a single plastic hinge at their bases. The walls are coupled
in the plastic hinge of a cantilever wall is caused by excess longi- with each other and with the columns through the floor slab. The
tudinal reinforcement, expected rather than specified material prop- side-view sketches refer to the maximum lateral deformation of the
erties, hardening in the longitudinal reinforcement, larger axial walls. The building of Fig. 3(a) has external walls. Fig. 3(b) has
compression in the wall, and the use of a strength-reduction factor interior walls in the direction of excitation. The corresponding
in design. When these factors are considered, the flexural capacity elevations of these configurations are shown in Figs. 3(c) and 3(d),
at the critical section at LS is M LS > M u . Therefore, the flexural respectively.
factor is defined as ΩLS;j ¼ M LS;j =M u . To attain the LS limit The migration of the neutral-axis depth in a reinforced-concrete
state on the bare cantilever wall, the first-mode lateral forces must member, such as walls, and the large curvatures expected in plastic
be increased to ΩLS;j F 1;i . Factor ΩLS;j may be computed from hinge regions cause significant elongation of their tension chord.
moment-curvature analyses of the critical section of the wall, as Additional but much smaller chord elongation also occurs in the
detailed, for each load combination, j. These analyses are per- elastic portion of the walls. Shortening of the compression chord
formed using expected material properties. Because of the rela- will also occur, although in comparison with the magnitude of the
tively low axial-compression force levels that nearly all building expected tension chord lengthening, the shortening will be small.
cantilever walls experience, demands for combined flexure-axial For the building shown in Fig. 3(c), these deformations are consid-
load design most often are governed by the flexural overstrength ered sufficiently large to mobilize a number of positive and

680 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.


Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 3. Plan and elevation of system-mechanism mobilizing yield lines in slabs framing into wall for two structural configurations

negative yield lines. Swaying of the walls also induces warping in X


n
the slabs because the gravity columns restrain the slab from rotat- V K;W;C ¼ ½M W þ QK ða þ Lw =2 þ eC Þ 1=hi ;
ing, thus enabling the development of moment M W at each level i¼1
ð10Þ
[see Fig. 3(c)]. In the undeformed state, the lateral forces F K;i X
n

needed to form the mechanism shown in Fig. 3(c) are given by V K;W;T ¼ ½M W þ QK ða þ Lw =2  eT Þ 1=hi
i¼1

2M W þ QK ð2a þ Lw þ eC  eT Þ In Eqs. (9) and (10), the subscripts T and C denote the wall that
F K;i ¼ ð8Þ resists tension and compression axial forces, respectively, caused
hi
by the plastic framing actions QK . The lateral forces of Eqs. (8)
These forces are distributed to the two walls as and (9) take the form of a harmonic series, with the force in the
lowest level being the greatest. In multistory buildings, the height
M W þ QK ða þ Lw =2 þ eC Þ of these resultant lateral forces is generally low. To quantify the
F K;i;C ¼ ; relative magnitude of these forces in relation to the first-mode lat-
hi
ð9Þ eral forces, the following example is considered. Consider a 6-story
M W þ QK ða þ Lw =2  eT Þ structure with H ¼ 6h and the plastic moment capacity of the
F K;i;T ¼
hi framing elements to be only aQK ¼ 0:03M u, M W ¼ 0, Lw ¼ a,
and eC ¼ 0:4Lw . For this case, the shear force at the base of the
Eq. (10) gives the shear forces at the base of the walls derived left wall of Fig. 3(c) is V K;W;C ¼ 0:84M u =H. Considering a
from the forces of Eq. (9). first-mode effective height equal to 2=3H, the first-mode base shear

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 681

J. Struct. Eng. 2011.137:677-690.


force of the wall is V b;1 =2 ¼ 1:5M u =H. It is remarkable that despite form of Eq. (2). If the first-mode period and the structural system
the sum of framing moments acting on the right wall on all six type are known, the second-mode period can be approximated.
floors is only 6QK ða þ Lw =2Þ ¼ 0:27M u, the base-shear ratio is For prismatic Euler-Bernoulli cantilever beams of uniform distrib-
V K;W;C =0:5V b;1 ¼ 0:56. Such a large ratio suggests that kinematic uted mass, the second mode is T 2 ¼ T 1 =6:3. The DBE spectral
effects induced by slabs and/or gravity load beams framing into acceleration, corresponding to T 2 , is available from the design
cantilevered walls should be considered in the design, even if only spectra (see Fig. 2). Thus, the second-mode base shear is
in an approximated way, or accounted for with a pushover analysis.
For the building shown in Fig. 3(d), the lateral forces F K;i Sa;2
V b;2 ¼ W e;2 ð15Þ
required to form the plastic framing mechanism shown are g
M W þ QK ð4a þ Lw Þ
F K;i ¼ ð11Þ For most medium-rise buildings, there is no gain by computing
hi
T 2 accurately. This is because T 2 most often falls in the region
For simplicity, Eq. (11) ignores the axial-force eccentricity at of constant spectral acceleration (Eibl and Keintzel 1988). The
the base of the columns. The forces resisted by the wall and each second-mode base shear is distributed in lateral forces F 2;i in
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

gravity column are given by proportion to W i Φ2;i . No base shear-force reduction is made in
Eq. (15) for the inelastic response of the wall. It is assumed that
M W þ QK ð2a þ Lw Þ aQK
F K;W;i ¼ ; F K;G;i ¼ ð12Þ the second mode of response of the building is not significantly
hi hi reduced by nonlinear response at the base of the wall (Rodriguez
The shear forces at the base of the wall and the gravity columns et al. 2002; Panagiotou 2008).
from the forces of Eq. (12) are
Xn Step 4: Design Envelopes
V K;W ¼ ½M W þ QK ð2a þ Lw Þ 1=hi ; At every story level, system overturning, shear forces, and axial
i¼1
ð13Þ forces are computed from the two mechanisms previously evalu-
X
n
ated and the second mode, and then combined using the square root
V K;G ¼ aQK 1=hi
i¼1
sum of the squares (SRSS) rule to obtain design actions as follows:
Initially, it seems counterintuitive that there is a single or qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
no axial force-resisting moment at the base of the walls of U i ¼ ð1:0U D;i þ 1:0U Lr;i Þ  1:0 ðΩU 1;i þ U K;i Þ2 þ ðU 2;i Þ2 LS
Figs. 3(c) and 3(d), respectively. This is because the tensile force ð16Þ
in the reinforcement has already been considered to resist bending
in the flexural overstrength mechanism, and these two mechanisms
For the walls’ bending-moment, Eq. (16) is used, where i ¼ 2 to
are considered independent and additive. Furthermore, for compat-
n. This is because at the base of the web wall the bending-moment,
ibility, the eccentricity between the axial forces in the critical
including flexural overstrength, has been previously established.
section in the two mechanisms must be the same. The actions attrib-
In Eq. (16), the forces of the element and system mechanisms
utable to kinematic overstrength, apart from increasing the shear
are added before squaring them because they develop concurrently.
force of the walls, cause significant variation of the walls’ axial
force and thus of their bending-moment resistance. This can be im-
portant when plasticity has to be constrained to develop over a spe-
cific region. A specific example demonstrating this is presented in
the next section, in which the method is applied to the design of a
7-story structure.

Step 3: Higher Mode Effects


Higher modes are well known for having a significant effect on the
response of structural walls (Blakeley et al. 1975; Derecho et al.
1981; Paulay and Priestley 1992; Rutemberg and Nsieri 2006;
Panagiotou and Restrepo 2007, 2009). Codes, such as New Zealand
Standards (NZS 4203), Canada (CSA 2005), Eurocode 8 (CEN
2004); or design guidelines (SEAOC 1999) incorporate procedures
that account for the effects of higher modes in design. Typically,
these procedures multiply the design lateral forces by a dynamic
magnification factor greater than unity.
In the method proposed, a different approach is implemented to
consider the higher modes of response. Parametric analyses per-
formed by Panagiotou (2008) suggest that in most wall buildings,
including regular tall buildings, the effects of the higher modes of
response are largely dominated by the second translational mode.
For the design of the building, the second mode is approximated by
the following cubic polynomial:
 3  2
h h h
Φ2;i ¼ 2:4 i  8:6 i þ 5:2 i ð14Þ
H H H
The second-mode modal weight, W e;2 , participation factor, Γ2 ,
Fig. 4. Southwest view of building
and contribution factor, Γ2 , are calculated using the generalized

682 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.


Application of Method to Full-Scale 7-Story Test for the limitations of experimental costs and capabilities. The lateral
Structure force resistance in the building was provided by a 3.66-m-long
load-bearing reinforced-concrete rectangular wall, referred to as
Description of Test Structure the web wall. The web wall, which was 0.20 m thick at the first
and seventh levels and 0.15 m elsewhere, provided lateral force
The test structure represented a slice of a 7-story multistory resi- resistance in the east–west direction and supported seven 0.20-m-
dential load-bearing wall building prototype located in Los Angeles thick slabs spaced at 2.74 m intervals. Two transverse walls built
(Englekirk 2007) (see Figs. 4 and 5). The building slice is referred east and west of the web wall provided lateral and torsional stability
to as the building hereafter in this paper. Fig. 4 shows an overall to the building. The precast segmental west wall was jointed using
view of the building and its main components. Fig. 5 presents the mortar bed joints, and then prestressed. This wall had a footing that
main dimensions and a definition of a level number. The test struc- allowed rocking in the east-west direction. The east wall, referred to
ture was selected to represent the structural configuration of some as the flange wall hereafter, was 4.87 m long and 0.20 m thick at the
real RC wall buildings with a structural configuration similar to that first story, and 0.15 m thick elsewhere. The web and the flange
shown in Fig. 3. A main concern in the selection of the structural walls were cast into a T-shaped footing that was tied to the moving
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

system was to represent the kinematic interaction between walls, platen of the shake table. The building height, measured from the
slabs, and gravity systems in this type of building, while accounting top of the foundation to the top of the roof slab, was 19.2 m. The

Fig. 5. Dimensions, building geometry, and lumped seismic weights

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 683

J. Struct. Eng. 2011.137:677-690.


building total weight, excluding the foundation, was 2,046 kN. H=Lw ≤ 7:8. The aspect ratio of the web wall is H=Lw ¼ 5:25,
Table 1 lists the tributary seismic weights in the building. which meets the IO design requirement. With the given web
The slab between the web and flange walls was 635 mm wide × wall geometry, the roof displacement estimated using Eq. (3) is
203 mm deep, and at the east and west ends, it had 140-mm-deep × δ y;r ¼ 94 mm.
51-mm-wide slots. With this geometry, the slab acted like a near- The spread of plasticity in the web wall of the building is,
pinned link, enabling the transfer of in-plane forces (shear, bend- theoretically, deliberately precluded from extending a distance
ing-moment, and axial force) and reduced out-of-plane bending beyond 1.4 m from the wall base by special detailing of the lon-
moments and shear forces. A 635-mm-wide vertical gap between gitudinal reinforcement. When the strain penetration of the
the web and the flange was left to avoid shear transfer between the longitudinal reinforcement in the wall is considered, the plastic
web and flange walls. The precast segmental wall was connected to hinge length in the web wall is ℓp ¼ 1:56 m ¼ 0:43Lw . As exam-
each of the slabs with a horizontal steel truss with proprietary low- ined in the following and discussed in the companion paper
friction ball bearing connections. The north and south ends of (Panagiotou et al. 2011), the amount of longitudinal reinforcement
the slabs were supported on four gravity columns consisting of at the second level in the web wall was less than that needed to fully
102-mm-diameter extra-strong steel pipes, which were filled with constrain the spreading of plasticity to within the desired length.
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

high-performance grout to bond a concentric high-strength steel In this building, the governing LS limit state was the roof drift
threaded bar that was used to form the column’s end connections. ratio, set at θy;LS ¼ 3% and thus δ LS;r ¼ 567 mm, corresponding to
The gravity columns were pinned at the ends and were able to carry displacement ductility μΔ ¼ δ LS;r =δ y;r, equal to 6. For δLS;r ¼
axial tension and compression. The walls and slabs were built using 567 mm from the simplified Eq. (5b) μφ;LS ¼ 18. The calculated
tunnel-form construction. Horizontal construction joints in the wall moment-curvature analyses of the critical section of the web wall
were located at the top of the slabs and again at 102 mm above the for various ratios of longitudinal reinforcement indicate that the
top of the slabs. Concrete with specified compressive strength of tensile strain in the reinforcement at μφ;LS ¼ 18 is slightly smaller
27.6 MPa and ASTM A615 Grade 60 steel reinforcement were than the LS strain limit of 5%.
used throughout. A specific study of the 360° component Sylmar ground-motion
obtained during the 1994 Northridge, California, earthquake gives
Performance Levels and Design Spectra
values of C μ between 0.7 and 1.2 in the period range of 1 to 2.2 s
The building was designed for the IO and LS performance levels for μΔ ¼ 6. This motion was used in the experimental program to
related to seismic hazard levels for a site in Los Angeles. The limit represent the design-basis earthquake (Panagiotou et al. 2011). The
states considered are the same described in the “Performance maximum value of C μ ¼ 1:2 was adopted for the design. The rel-
Levels” section. The displacement and acceleration design spectra atively small value of C μ for the specific ground motion is because
for the two hazard levels are shown in Fig. 2. The displacement the predominant periods of the distinct pulses contained in this
spectra are linear in the period range exceeding 0.48 s, where motion are smaller than the period of the building (Panagiotou
the spectral velocity is constant; the linear portion of the spectra 2008). The equivalent displacements calculated from Eq. (6) are
has a slope aIO ¼ 0:06 m=s and aLS ¼ 0:22 m=s for the IO and δ IO;e ¼ 64 mm and δ LS;e ¼ 323 mm. The maximum periods T IO
LS seismic hazards, respectively. and T LS corresponding to the critical limit states at IO and LS,
respectively, are T IO ¼ 1:06 s and T LS ¼ 1:53 s (see Fig. 2).
Step 1: First-Mode Design Lateral Forces In the design of the building, IO governs. When the longitudinal
Using the values of hi and W i listed on Table 1 and making use of reinforcement LS tensile-strain limit is revised to 3%, then
Eqs. (1) and (2), W e;1 ¼ 0:66W t and Γ1 ¼ 1:46. The target roof μφ;LS ¼ 10, δ LS;r ¼ 344 mm, δ LS;e ¼ 236 mm, and T LS ¼ 1:12 s,
displacements are calculated for the limit states governing the and the first-mode design base-shear, V b;1 , is calculated as
IO and LS performance objectives, respectively. Coefficient λy ¼ V b;1 ¼ 292 kN, corresponding to a design base-shear coefficient
0:0034 was calculated from a moment-curvature analysis of the C s ¼ 0:15. The first-mode base shear is distributed in lateral forces
web wall. With the value of λy and checking that θy;r is equal F 1;i in proportion to W i Φ1;i (see Table 2). The design bending-
to or less than the IO interstory drift limit state of 1, the geometrical moment at the wall base is M u ¼ 4243 kN-m.
aspect ratio of the web wall, calculated by solving Eq. (4), is
Step 2: System Static Overstrength Lateral Forces

Table 1. Seismic Weights and Corresponding Heights For the revised LS tensile strain limit of 3%, ΩLS ¼ 1:44. The sys-
tem mechanism was ignored in the initial design of the building;
Weight (kN)
however, since this mechanism had a large effect on the building’s
Height hi Web Flange Segmental
(m) Referencea wall Slab wall wall Other Total
Table 2. Design Lateral Forces
0.0 W0 22 0 31 20 0 73
2.6 W1 39 169 55 39 8 310 Lateral force (kN)
5.4 W2 33 156 48 39 8 285 Level i F 1;i F k;i;C F k;i;T F 2;i
8.1 W3 33 156 48 39 8 284
7 78 33 14 263
10.9 W4 33 156 48 39 8 284
6 77 38 16 113
13.6 W5 33 156 48 39 8 284
5 57 46 19 66
16.4 W6 39 156 49 40 8 291
4 40 58 24 196
19.1 W7 22 156 30 20 6 235
3 24 77 32 268
2,046
2 12 116 48 267
Note: Does not include foundation’s weight; height measured from 1 3 237 98 195
wall base.
a Sum 292 605 250 615
Refer to Fig. 2.

684 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.


response (Panagiotou et al. 2011), it was included as part of the Step 3: Higher-Mode Effects
design in this paper.
For the revised LS tensile strain limit of 3% in the extreme lon- Based on Eq. (2), W e;2 ¼ 0:17W t , Γ2 ¼ 0:62, and T 2 ¼ T 1 =6:3 ¼
gitudinal bar in tension, the expected tensile chord elongation at the 0:18 s. The DBE spectral acceleration, corresponding to T 2 , is
first level of the web wall is 47 mm. This elongation is large enough available from the design spectra, Sa ¼ 1:81g (see Fig. 2), and
to mobilize a local slab mechanism in the slotted slab. The wall V b;2 ¼ 615 kN, which is equal to 2:1V b;1 . The second-mode base
tensile-chord lengthening causes slab end rotations of at least shear is distributed in lateral forces F 2;i in proportion to W i Φ2;i , as
0.077 rad. At such a large rotation, the slab develops its small flexu- listed in Table 2.
ral capacity at the slotted ends over the entire 4.88 m length of the
Step 4: Design Envelopes
slots at each floor. Plastic analysis of the local mechanism gives a
shear force of QK ¼ 131 kN and a shear span of a ¼ 379 mm (see System and Web Wall Overturning and Bending-Moment
Fig. 3). Here, 80% of the slab shear-force QK flows toward the wall Envelopes
and resisted at the base by a compressive force, acting at an eccen- Fig. 6(a) plots the system-overturning moment and the web-wall
tricity ec from the wall’s centerline. The remaining 20% is carried bending-moment envelopes for the eastward and westward re-
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

by the gravity columns. The additional web wall base sponse. The system-overturning moment demand varies approxi-
compressive force adds to the force needed for equilibrium in mately linearly with height as a result of the combined first and
the flexural overstrength mechanism. second modes. This figure shows asymmetry among the eastward
Swaying of the wall also induces combined 3D bending in the and westward system overturning-moment envelopes, which is
slabs. This is because the gravity columns that are located at the caused by the slotted slabs and flange wall that act together as
slab north and south edges restrain the slab from rotating, thus an outrigger system. The system overturning moment is partially
enabling the development of a warping moment M W at each level resisted by the wall in bending, while the remainder is resisted
(see Fig. 3). A nonlinear finite-element analysis of the slab gave a by a tension-compression couple in the web wall-outrigger system.
moment capacity of M W ¼ 204 kN-m. This moment is resisted by The base web wall’s share of the system overturning moment is
a tension-compression pair by the gravity columns. The lateral 70% in the west direction and 60% in the east direction.
forces F k;i;C needed to form the mechanism shown in Fig. 3 are
given by Eq. (9). The additional but small lateral forces carried System Shear Force Envelopes
by the flange wall are ignored. Table 2 lists forces F K;i;C calculated Fig. 6(b) plots the system shear force design envelopes obtained for
for the building displaced westward. Compared with the first-mode the eastward and westward demands. Shear forces are significantly
lateral forces F 1;i (see Table 2), the forces F K;W;i are quite remark- influenced by the second mode and kinematic effects, particularly
able, and although the sum of slab moments on all seven levels in the lower levels of the building. The maximum base shear force
is 7½M W þ QK ða þ ðLW =2ÞÞ ¼ 0:73M u , the base shear ratio is of 1.2 MN is 4:1V b;1 , which is equivalent to a base shear coefficient
V K;W;C =V b;1 ¼ 2:07. The resultant force V K;W;C is located at of C s ¼ 0:59. The asymmetry among the east and westward shear
0.39% of the roof height. Kinematic effects caused by the swaying force envelopes is attributable to the kinematic effects previously
of the building eastward can be used to estimate in a similar way to discussed.
the westward direction previously examined. The lateral forces
F K;i;T required to develop the eastward mechanism computed with Web Wall Axial-Force Envelopes
Eq. (9) are listed in Table 2. The lateral forces in this direction are The axial forces in the web wall are plotted in Fig. 6(c). Under
smaller than those determined for the westward direction, yet they gravity loading only, the web wall experiences a compressive axial
sum up to a base shear that is comparable to that calculated for the force of 809 kN at the base. The outrigging action of the slab causes
first mode. This is because the base moment caused by the reacting axial-force changes in the web wall, raising it to N u ¼ 1534 kN ¼
vertical force (tension) in the wall has the same sign of the slab 0:06Ag f 0c during westward response and reducing it to only N u ¼
moments, while the opposite occurs in the westward direction. 75 kN ¼ 0:003Ag f 0c during eastward response.

Fig. 6. System and web-wall design envelopes: (a) moment; (b) shear force; (c) axial force

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 685

J. Struct. Eng. 2011.137:677-690.


Web Wall Design
Potential Plastic Hinge Region
As previously discussed, the reinforcement in the web wall is
determined from the single-load combination presented in Eq. (16),
using U Lr ¼ 0 and strength reduction factors for bending and shear
equal to 1. For practical applications, the load combination with the
lowest axial force would govern the design for combined bending
and axial force; strength reduction factors would be those recom-
mended by a design code and would be less than 1.
The longitudinal reinforcement in the potential plastic hinge
region of the web wall is calculated from M u ¼ 4243 kN-m and
N u ¼ 809 kN. With specified material properties f y ¼ 414 and
f 0c ¼ 30 MPa, the required longitudinal reinforcement ratio for
the ðM u ; N u Þ pair is ρl ¼ 0:59%. Figs. 7 and 8 show the relevant
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

reinforcing details for the web wall. The longitudinal reinforcement


in this region of the web wall consisted of 8 No. 5 bars in the
boundary elements. In between the boundary elements, 13 No. 4
longitudinal bars were detailed in a single curtain (see Fig. 7).
The reinforcement ratio in the first level of the web wall was
ρl ¼ 0:66%. Moment-curvature analyses were performed on the
section of the wall, using the expected wall section axial forces con-
sidering the kinematical effects, to calculate of the expected east- Fig. 8. Web wall elevation Levels 1 and 2 showing longitudinal and
ward and westward neutral axis depths. The boundary elements transverse reinforcement
extend from the edge of the wall toward the wall centerline, a length
beyond which the longitudinal bars will experience tensile and
compressive strain cycles. ACI 318-05 (ACI 2005) requires minimum shear reinforcement.
The transverse reinforcement in the boundary elements con- A level just above the minimum shear reinforcement is required
sisted of No. 3 weld-resistance grids spaced at sh ¼ 102 mm. for the second phase of the experimental program. A single
Distance sh provided adequate lateral stability to the longitudinal reinforcement curtain No. 4 bars at 203 mm on center is prescribed
Number. 5 bars (sh =d b ¼ 6:4) for the expected plastic strain range for the potential plastic hinge region of the web wall [see Figs. 7(a)
of about 3.5%. The critical strain for the confinement of the con- and 8]. This results in a transverse reinforcement ratio ρt ¼ 0:31%.
crete core in the boundary elements occurred in a second phase of
the test program, when the system resisted lateral forces with a Other Regions
monolithic T-wall and the flange of this wall was in tension (where The previous section described the design of the plastic-hinge
the maximum expected concrete compressive strain occurs in Phase region in the web wall. Capacity design was used to design this
II and is only εc ¼ 0:5%). At this low compressive strain level, region for shear. This required assessing the probable shear-force
only minimum confinement of the concrete core is needed. demand and included (1) increased shear attributable to flexural
In the design for shear, the entire shear force of V u ¼ 1:2 MN overstrength; (2) increased shear caused by kinematic system
was allocated entirely to the web wall, thus ignoring the small par- overstrength effects; and (3) increased shear attributable to the
ticipation of the flange wall. Even with this shear-force demand, second mode.

Fig. 7. Building—Plan view of reinforcement

686 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.


The design of the longitudinal reinforcement above the plastic The expected moment strengths M E determined for the web wall
hinge region must consider the axial-flexure-shear interaction. for Levels 2 through 7 for the eastward and westward directions are
Codes generally consider the effect of diagonal cracking indirectly shown in Fig. 9. Expected moment strengths are determined using
caused by shear through their prescriptions for the development the ACI 318-05 (ACI 2005) simplified-flexure method, and the
of the flexural reinforcement. The design of the longitudinal longitudinal reinforcement provided the corresponding axial force
reinforcement in the web wall was carried out using the fictitious pair and expected material properties of f y ¼ 456 MPa and
bending-moment approach proposed by Presland et al. (2001). This f 0c ¼ 35 MPa. The extent of the flexural strength deficiency in
approach recognizes explicitly that in a section of a diagonally the web wall becomes evident when the expected moment strengths
cracked beam-column element, the presence of shear increases M E are compared with the moments M V in this figure. When
the tensile force demand in the section, while it ignores the reduc- M V ≥ M E , yielding of the wall is likely to occur. The flexural-
tion in the compressive force demand. Using the Mörsch parallel- strength deficit is pronounced when the wall is deflected eastward,
angle truss analogy (Mörsch 1909; Collins and Mitchell 1991), the because yielding of the wall is likely throughout Levels 2 and 3 and
additional tensile force needed for equilibrium in the section is at the base of Level 4. In the westward direction, yielding of the
ΔT ¼ V u =ð2 tan θv Þ. A similar additional tensile-force demand is web wall is likely to spread through the bottom half of Level 2 and
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

obtained from flexure theory if the bending-moment demand in at the base of Level 3. Because of the final detailing of the longi-
the section is fictitiously increased from M u to M v where tudinal reinforcement in the web wall, yielding of the wall is no
longer expected only at the base of the wall, as initially intended.
Vu As discussed in the companion paper (Panagiotou et al. 2011) dur-
M V ¼ M u þ ΔTjd ¼ M u þ jd ð17Þ ing the experimental response, yielding of the reinforcing steel was
2 tan θv
measured up to Level 4, verifying to a large extent the analysis
method.
The values of M v are calculated for the web wall, assuming Shear-force demand in Level 2 also led to minimum reinforce-
jd ¼ 0:8ℓw . Because minimum shear reinforcement is required ment. The transverse reinforcement was also controlled by the
for the first phase of testing, it is assumed that θv ¼ 35° in second-phase demands. A single reinforcement curtain of No. 4
Eq. (17). The values of M u and M V are plotted in Fig. 9 for Levels 2 bars at 203 mm on center was provided at Levels 2 through 6
through 7 and for the eastward and westward directions, respec- [see Figs. 7(b) and 8], resulting in a transverse reinforcement ratio
tively. The values of M V computed for the eastward and westward of ρt ¼ 0:41%. The detailing of the reinforcement at Level 7 was
directions are paired with the corresponding axial forces N u to cal- identical to that at Level 1; at some point in the test program, it was
culate the required longitudinal reinforcement ratios in these levels. envisaged that the wall would be coupled at the top by a stiff beam
The longitudinal reinforcement ratios required at Levels 2, 3, and 4 that would force a plastic hinge there.
were 1.28, 1.12, and 1.01%, respectively. The reinforcement ratios
provided in Levels 2, 3, and 4 of the web are smaller than those Comparison of Measured and Analytically Predicted
required by the proposed analysis and equal to 0.81%. When Response Envelopes
the building was designed, kinematical effects caused by the fram-
ing slabs onto the web wall were ignored, and the demands are Fig. 10 compares the experimentally measured (Panagiotou et al.
more stringent when these effects are included. 2011) and analytically predicted, with the method described in
Figs. 7(b) and 8 show the longitudinal reinforcement arrange- this paper, envelopes of system-bending moment and shear force,
ment of Level 2. All longitudinal bars start in a single curtain, respectively, for the eastward and westward response in the DBE
104 mm above the first-floor slab. These bars were spliced with test. Regarding the system bending-moment, the method of analy-
bars fully anchored in the upper half of Level 1 (see Fig. 8). To sis estimated to a fair degree of accuracy the system bending-
enhance the splicing of the longitudinal reinforcement at the ends moment envelope for the eastward and westward response. A small
of the web wall, U-shape No. 3 reinforcement was used in the lower underestimation is observed in the lower half of the building in the
half of Levels 2 through 6. westward response. The method of analysis did not account for the
small bending moment carried by the flange and segmental walls.
The underestimation was larger for the eastward response. The
analysis did not account for the strain aging effects that played
a role in the eastward response of the building (Panagiotou et al.
2011). Regarding the system shear forces, the predictions obtained
through the method of analysis were found in very good agreement
with the measured results throughout the height of the building for
both the eastward and westward responses. The very good match
between measured and estimated shear forces may be a coinci-
dence, as some prediction errors were expected for the following
reasons: (1) of the way that lateral forces attributable to the first
mode of response, kinematic system overstrength and second mode
of response were combined; (2) the second-mode spectral acceler-
ation was that of the design spectrum; (3) the contribution of modes
higher than the second one were ignored; (4) strain rate and strain-
aging effects were not accounted for in the analysis procedure;
(5) the shear forces carried by the segmental and flange walls were
ignored in the analysis method; and (6) the analysis procedure
assumed a damping ratio of 5%, which is significantly higher than
Fig. 9. Web-wall bending-moment envelopes: designed, expected, and
the damping ratio identified for the building in the highly inelastic
provided strength
range (Panagiotou 2008).

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 687

J. Struct. Eng. 2011.137:677-690.


7 7
Eastward Westward Eastward Westward
6 6

5 5
Measured
4 4
Story

Story
Measured
3 3

2 2
Analysis Analysis
1 1

0 0
-15 -10 -5 0 5 10 15 -1.5 -1 -0.5 0 0.5 1 1.5
Bending Moment (MN-m) Shear Force (MN)
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 10. Comparison of analytically computed and experimentally measured (Panagiotou et al. 2011) system-response envelopes

As also discussed in Panagiotou et al. (2011), the experimental overstrength and higher modes of response for this test
response of the 7-story building verified the displacement-based structure.
method of analysis. The method of analysis estimated very satis- 5. The combination of these effects also resulted in a significant
factorily the interstory drift at IO, as well as the maximum tensile variation of the axial load and bending-moment resistance
strain of the wall longitudinal reinforcement (Panagiotou et al. along the height of the web wall. It was analytically computed
2011). Similarly, the method estimated with fair accuracy the roof that yielding will extend up to Level 4 of the web wall. This
displacement, as well as the maximum tensile strain of the wall was verified by the experimentally measured response, as dis-
longitudinal reinforcement at LS. Finally, the experimental cussed in the companion paper Panagiotou et al. (2011).
response additionally verified the method regarding the effect 6. The proposed method of analysis was largely verified from the
of kinematic overstrength to the yielding of the web wall at test response of the 7-story structure, including the ways in
Levels 2, 3, and 4. which the effects of kinematic system overstrength and second
mode of response were accounted for.

Conclusions
Acknowledgments
This paper described a displacement-based method of analysis for
regular reinforced-concrete-wall buildings. The method explicitly The authors sincerely thank the Englekirk Board of Advisors, a
accounts for the effects of system overstrength and higher modes University of California–San Diego industry group supporting
of response. It also relies on capacity design to ensure the intended research in structural engineering. In particular, they thank
seismic performance at LS. The method is specifically applied to Dr. Robert Englekirk for his support and advice.
the full-scale 7-story reinforced-concrete building built and tested
on the George E. Brown Jr. Network for Earthquake Engineering
Simulation Large Outdoor High-Performance Shake Table of the Notation
University of California–San Diego. The following conclusions
are drawn: The following symbols are used in this paper:
1. The proposed method of analysis considers the combined Ag = wall cross-sectional area;
effects of inelastic first mode of response, kinematic system a = shear span of framing element;
overstrength, and higher modes of response. This allows for aIO = slope of the linear portion of the IO displacement
a fair estimation of the response quantities required to capacity spectrum;
design and control the inelastic response of reinforced- aLS = slope of the linear portion of the LS displacement
concrete-wall buildings. spectrum;
2. The kinematic interaction between walls and elements framing C S = base shear coefficient;
into them, defined as kinematic system overstrength, is expli- C μ = constant ductility inelastic to elastic displacement ratio;
citly evaluated in the proposed method. It can have a signifi- d b = bar diameter;
cant effect on reinforced-concrete-wall buildings, increasing eC = eccentricity of compressive force at base of wall
the shear-force demand in the walls and significantly varying attributable to kinematic overstrength;
their axial force and thus their bending-moment resistance. eT = eccentricity of tensile force at base of wall attributable to
3. The effect of higher modes of response is assessed by explicitly kinematic overstrength;
considering the second-mode response. Higher modes are F K;G;i = gravity column lateral force attributable to kinematic
shown to increase the shear force and the bending-moment overstrength at level i;
demand in the walls. F K;W;i = wall lateral force attributable to kinematic overstrength
4. The application of the method to a 7-story full-scale building at level i;
slice/shake table test, described in the accompanying paper F K;i = lateral force attributable to kinematic overstrength at
by Panagiotou et al. (2011) analytically computed a base level i;
shear-force demand 4.1 times larger than the design base shear F K;i;C = lateral force attributable to kinematic overstrength at
force considering only the first-mode response. This was attri- level i for wall resisting compression attributable to
butable to the significant combined effects of kinematic system kinematic overstrength;

688 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.


F K;i;T = lateral force attributable to kinematic overstrength at V K;W;T = base shear force attributable to kinematic system
level i for wall resisting tension attributable to kinematic overstrength of the wall resisting tension caused by
overstrength; kinematic system overstrength;
F 1;i = first-mode lateral force at level i; V u = design shear force;
F 2;i = second-mode lateral forces at level i; W e;1 = first-mode effective seismic weight;
f y = steel yield strength; W e;2 = second-mode effective seismic weight;
f 0c = concrete compressive strength; W i = seismic weight at level i;
g = acceleration of gravity; W t = total seismic weight excluding foundation weight;
H = total building height; Γ1 = first-mode participation factor;
hi = height at level i; Γ2 = second-mode participation factor;
h = floor height; Γ1 = first-mode contribution factor;
i = level number; Γ2 = second-mode contribution factor;
j = load combination; ΔT = additional tensile force needed for equilibrium in the
jd = internal lever arm between the section resultant tensile wall section attributable to flexure-shear interaction;
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

and compressive forces; δ IO;e = equivalent SDOF target displacement at IO;


Lw = length of wall; δ IO;r = roof target displacement at IO;
ℓp = wall plastic hinge length; δ LS;e = equivalent SDOF target displacement at LS;
M E = expected moment strength; δ LS;r = roof target displacement at LS;
M LS = wall base moment at LS; δ p;r = plastic roof displacement;
M LS;j = moment at LS for load combination j; δ y;r = yield roof displacement;
M n = wall nominal base moment; εc = concrete compressive strain;
M o = calculated base moment at 3% steel tensile strain for εs = reinforcing steel strain;
eastward or westward response; εy = reinforcing steel yield strain;
M u = design base moment; θv = angle of inclination, measured with respect to the line at
M u;i = bending moment demand; 90 degrees from the longitudinal axis of the element, of
M v = fictitious bending moment demand to consider for the the diagonal compression stress field carrying the shear
flexure shear interaction in a section; force;
M w = slab-warping moment; θy;r = maximum interstory drift ratio at δ y;r ;
n = number of floors; λy = geometrical coefficient of wall;
Pu = design compressive axial force; μΔ = displacement ductility;
Pu;j = wall axial load for load combination j; μφ = wall curvature ductility;
QK = framing element plastic shear force; μφ;LS = wall base curvature ductility at LS;
R = response modification coefficient; ρl = web wall longitudinal reinforcement ratio;
Sa = spectral acceleration; ρt = web wall transverse reinforcement ratio;
Sd = spectral displacement; Φ1;i = first-mode shape at level i;
sh = distance between confinement grids; Φ2;i = second-mode shape at level i;
T = period of free vibration; φLS = wall base curvature at LS;
T D = design fundamental period; φy = reference wall yield curvature;
T IO = maximum fundamental period at IO limit state; ΩLS;E = web wall flexural-overstrength factor for eastward
T LS = maximum fundamental period at LS limit state; response;
T 2 = second-mode period of building; ΩLS;j = base-moment flexural-overstrength factor at LS limit
U D;i = design quantity (bending moment, shear force, or axial state for load combination j; and
force) attributable to dead load at level i; ΩLS;W = web wall flexural-overstrength factor for westward
U K;i = design quantity (bending moment, shear force, or response.
axial force) at level i attributable to kinematical
effects;
U Lr;i = design quantity (bending moment, shear force, or axial
force) attributable to reduced live load at level i in the
building U Lr;I ¼ 0; References
U 1;i = design quantity (bending moment, shear force) at level i
attributable to first-mode lateral forces; American Concrete Institute (ACI). (2005). Building Code Requirements
U 2;i = design quantity (bending moment, shear force) at level i for Structural Concrete and Commentary, ACI Committee 318-05,
Farmington Hills, MI, 430.
attributable to second-mode lateral forces;
ASCE. (2006). “Minimum design loads for buildings and other structures.”
V b;1 = first-mode design base shear; SEI 7-2005, Reston, VA.
V b;2 = second-mode design base shear; Balkaya, C., and Kalkan, E. (2003). “Nonlinear seismic response evalu-
V K = wall–base shear force attributable to kinematic system ation of tunnel-form building structures.” Comput. Struct., 81(3),
overstrength; 153–165.
V K;G = gravity column base shear force attributable to Balkaya, C., and Kalkan, E. (2004). “Three-dimensional effects on
kinematic system overstrength; openings of laterally loaded pierced shear walls.” J. Struct. Eng.,
V K;W = wall–base shear force attributable to kinematic system 130(10), 1506–1514.
overstrength; Bertero, V. V., Atkan, A. E., Charney, F., and Sause, R. (1985). “Earthquake
V K;W;C = base shear force attributable to kinematic system simulator tests and associated experiments, analytical and correlation
studies of one-fifth scale model.” Earthquake effects on reinforced
overstrength of wall resisting compression caused by
concrete structures, U.S.–Japan Research, ACI Publication SP-84,
kinematic system overstrength; American Concrete Institute, Detroit, 375–424.

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011 / 689

J. Struct. Eng. 2011.137:677-690.


Blakeley, R. W. G., Cooney, R. C., and Megget, L. M. (1975). “Seismic Panagiotou, M., and Restrepo, J. I. (2007). “Lessons learned from the
shear loading at flexural capacity in cantilever wall structures.” Bull. N. UCSD full-scale shake table testing on a 7-story residential building
Z. Natl. Soc. Earthquake Eng., 8, 278–290. slice.” Proc., SEAOC Convention, Reno, NV.
Canadian Standard Association (CSA). (2005). Design of Concrete Struc- Panagiotou, M., and Restrepo, J. I. (2009). “Dual plastic hinge concept for
tures, Standard A23.3-04, Rexdale, Canada, 214. arresting higher mode effect on high-rise cantilever wall buildings.”
Chopra, A. K. (2001). Dynamics of structures: Theory and applications to Earthquake Eng. Struct. Dyn., 38(12).
earthquake engineering, 2nd Ed., Prentice Hall, Upper Saddle Panagiotou, M., Restrepo, J. I., and Conte, J. P. (2007). “Shake table test of
River, NJ. a 7-story full scale reinforced concrete structural wall building slice
Chopra, A. K., and Chintanapakdee, C. (2001). “Comparing response phase II: t-wall.” SSRP 07-08, Dept. of Structural Engineering, Univ.
of SDOF systems to near-fault and far-fault earthquake motions in of California–San Diego.
the context of spectral regions,” Earthquake Eng. Struct. Dyn., 30 Panagiotou, M., Restrepo, J. I., and Conte, J. P. (2011). “Shake-table test of
(3), 417–442. a full-scale 7-story building slice. Phase I: Rectangular wall.” J. Struct.
Chopra, A. K., and Goel, R. K. (2001). “Direct displacement-based design: Eng., 137(6), 691–704.
Use of inelastic versus elastic design spectra.” Earthquake Spectra, Park, R., and Paulay, T. (1975). Reinforced concrete structures, Wiley,
17(1), 47–64. Hoboken, NJ.
Downloaded from ascelibrary.org by University of Leeds on 08/21/13. Copyright ASCE. For personal use only; all rights reserved.

Collins, M. P., and Mitchell, D. (1991). Prestressed concrete structures, Paulay, T., Bachmann, H., and Moser, K. (1990). Erdbebenbemessung von
Prentice Hall, Englewood Hills, NJ, 766. Stahlbetonhochbauten, Birkhäuser, Berlin, 1990 (in German).
Derecho, A. T., Iqbal, M., Ghosh, S. K., Fintel, M., Corley, W. G., and Paulay, T., and Priestley, M. J. N. (1992). Seismic design of reinforced
Scanlon, A. (1981). Structural walls in earthquake-resistant buildings concrete and masonry buildings, Wiley, Hoboken, NJ.
dynamic analysis of isolated structural walls development of Paulay, T., and Restrepo, J. I. (1998). “Displacement and ductility compat-
design procedure—Design force levels, Portland Cement Association, ibility in buildings with mixed structural systems.” J. New Zealand
Skokie, IL. Struct. Eng. Soc., 2(1), 7–12.
Eibl, J., and Keintzel, F. (1988). “Seismic shear forces in RC cantilever Paulay, T., and Taylor, R. G. (1981). “Slab coupling of earthquake-resisting
shear walls.” Proc., 9th World Conf. on Earthquake Engineering, shear walls.” ACI Struct. J., 78(2).
Tokyo, Paper 9-1-1. Presland, R. A., Restrepo, J. I., and Park, R. (2001). “Seismic performance
Englekirk, R. E. (2007). “Rational seismic design procedures for shear wall of retrofitted reinforced concrete bridge piers: Laboratory testing of
braced buildings.” Charles Pankow Foundation, Claremont, CA, 〈http:// the proposed Thorndon Overbridge retrofit scheme.” Research Rep.
www.spur.org/documents/1106englekirk%20final%20report042107 2001-03, Dept. of Civil Eng., Univ. of Canterbury, Christchurch,
.pdf〉, (Jan. 13, 2009). New Zealand, 513.
European Committee for Standardization (CEN). (2004). EC8: Design of Priestley, M. J. N., Calvi, G. M., and Kowalsky, M. J. (2007). Displace-
structures for earthquake resistance, Brussels, Belgium. ment-based seismic design of structures, IUSS, Pavia, Italy.
Fajfar, P. (2000). “A nonlinear analysis method for performance based Reinhorn, A. (1997). “Inelastic analysis techniques in seismic evaluations.”
seismic design.” Earthquake Spectra, 16(3), 573–592. Seismic design methodologies for the next generation of codes,
Fédération Internationale du Beton (FIB). (2003). Displacement-based P. Krawinkler and H. Fajfar, eds., Balkema, Rotterdam, Netherlands,
seismic design of reinforced-concrete buildings, State-of-the-art Rep., 277–287.
Lausanne, Switzerland, 196. Rodriguez, M. E., Restrepo, J. I., and Carr, A. J. (2002). “Earthquake-
Fintel, M. (1995). “Performance of buildings with shear walls in earth- induced floor horizontal accelerations in buildings.” Earthquake Eng.
quakes of the last thirty years.” Precast Prestressed Concr., 40(33), Struct. Dyn., 31, 693–718.
62–80. Ruiz-García, J., and Miranda, E. (2003). “Inelastic displacement ratios for
Hines, E. M., Restrepo, J. I., and Seible, F. (2004). “Force-displacement evaluation of existing structures.” Earthquake Eng. Struct. Dyn., 32(8),
characterization of well-confined bridge piers.” ACI Struct. J., 101(4), 1237–1258.
537–548. Ruiz-García, J., and Miranda, E. (2006). “Inelastic displacement ratios for
Lang, A. F., and Restrepo, J. I. (2006). “Seismic performance evaluation evaluation of structures built on soft soil sites.” Earthquake Eng. Struct.
of gypsum wallboard partitions.” Proc., 8th U.S. National Conf. on Dyn., (35), 679–694.
Earthquake Eng. (EERI), San Francisco, Paper No. 1353. Rutenberg, A., and Nsieri, E. (2006). “The seismic shear demand in ductile
Moehle, J. P. (1992). “Displacement-based design of RC structures cantilever wall systems and the EC8 provisions.” Bull. Earthquake
subjected to earthquakes.” Earthquake Spectra, 8(3), 403–428. Eng., 4, 1–21.
Mörsch, E. (1909). “Concrete-steel construction,” Engineering News, New Satyarno, I., Carr, A. J., and Restrepo, J. (1998). “Refined pushover analy-
York, 368. 〈http://books.google.com/books?id=RO8OAAAAYAAJ sis for the assessment of older reinforced concrete buildings.” Proc.,
&printsec=frontcover&dq=concrete-steel+construction#PPP4,M1〉 New Zealand Society of Earthquake Engineering Technology Conf.,
(Jan. 28, 2009). Wairakei, New Zealand, 75–82.
New Zealand Standards (NZS). (1992). 4203, Vols. 1 and 2: Code of prac- Structural Engineers Association of California (SEAOC). (1999). SEAOC
tice for general structural design and design loadings for buildings, blue book: Seismic design recommendations of the SEAOC seismology
Wellington, New Zealand. committee, Sacramento, CA.
Oztoruna, N. K., Citipitioglu, E., and Akkas, N. (1998). “Three- Sullivan, T. J., Calvi, G. M., Priestley, M. J. N., and Kowalsky, M. J. (2003).
dimensional finite-element analysis of shear wall buildings.” Comput. “The limitations and performances of different displacement based
Struct., 68. design methods.” J. Earthquake Eng., 7, 201–241.
Panagiotou, M. (2008). “Seismic design, testing and analysis of reinforced Wood, S. (1991). “Performance of reinforced-concrete buildings during the
concrete wall buildings.” Ph.D. Thesis, Dept. of Structural Eng., Univ. 1985 Chile earthquake: Implications for the design of structural walls.”
of California–San Diego. Earthquake Spectra, 7(4), 607–638.

690 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / JUNE 2011

J. Struct. Eng. 2011.137:677-690.

You might also like