Professional Documents
Culture Documents
Abstract: The reported experimental study documents the performance of six fully grouted reinforced concrete block structural walls tested
under quasistatic cyclic loading. The walls fall under the ductile shear walls and the special reinforced masonry walls seismic force resisting
system (SFRS) classification of the Canadian and U.S. standards, respectively. The test matrix consisted of one rectangular, one flanged, and
two slab-coupled walls, all with an overall aspect ratio of 1.4. In addition, two rectangular walls, representing the individual components of
the slab-coupled wall systems, were tested to quantify the wall slab coupling effects. In addition to discussing the experimental results, the
study also presents key force-based seismic design (FBSD) parameters, such as the wall lateral load capacity, plastic hinge length, wall failure
modes, and displacement ductility capacities. Moreover, the effects of wall cross-sectional configuration and slab coupling on the cyclic
response and deformation capabilities of the walls are discussed. In general, the yield and ultimate loads were found to be accurately predicted
using the Canadian Standards Association (CSA) S304-14 and Masonry Standards Joint Committee (MSJC) formulations. The wall exper-
imental displacement ductility values (calculated at 20% strength degradation) ranged between 5.4 and 7.6, whereas the idealized displace-
ment ductility values at the same strength degradation level ranged between 3.4 and 5.4. The analysis results reported in the paper highlight
the fact that walls designed and detailed within the same SFRS classification possess significantly different FBSD parameters. The results also
indicate that slab coupling, although not recognized as a wall coupling mechanism in the current editions of the CSA and MSJC, can have
significant influences on the seismic response of ductile/special reinforced masonry wall systems. DOI: 10.1061/(ASCE)CF.1943-5509
.0000794. © 2015 American Society of Civil Engineers.
Author keywords: Cyclic loading; Ductility; Plastic hinge length; Reinforced masonry; Structural wall; Seismic response; Slab-coupling.
slab located at height of 1.04 m on center from the foundation/wall the scaled version of the conventional full-scale M20 steel bars.
interface, and a roof slab. In this respect, in their text, Harris and For the horizontal (shear) reinforcement, W1.7 bars (3.8 mm diam-
Sabnis (1999) discuss several experimental studies that focused eter), which represented a scaled version of the full-scale M10 bars,
on the performance of scaled reinforced masonry at the material, were used and were hooked to the end of the walls’ outermost
assemblage, component, and system levels. In addition, earlier vertical bars. The mechanical properties of the wall constituent
research studies by Hamid et al. (1985) have indicated a good materials (blocks, mortar, grout, and reinforcement) were obtained
correlation between a full-scale masonry prototype and the corre- through a series of standardized tests ASTM CI09-08 (ASTM
sponding scaled models. More recently, there has been a consider- 2008a), ASTM CI019-08 (ASTM 2008b), CSA A165 (CSA
able number of research studies that focused on utilizing scaled 2014b), CSA A179-14 (CSA 2014d), and the average yield
reinforced masonry shear wall models to predict the response of strengths, f y of the D7 and W1.7 reinforcements, were 495 and
their full-scale counterparts (Shedid et al. 2010a, c; Shedid and 670 MPa, respectively, based on the tensile strength tests according
El-Dakhakhni 2014; Banting and El-Dakhakhni 2012, 2014a, b; to CSA G30.18-09 (CSA 2014a).
Hereema et al. 2015a, b). The average compressive strength of the masonry blocks was
As shown in Fig. 1, the test matrix consisted of one rectangular 25.2 MPa according to ASTM C140-08 (ASTM 2008c) and CSA
wall (W1), one flanged wall (W2) and two slab-coupled walls A165-14 (CSA 2014b). Type S mortar was used in wall construc-
systems, (W3 and W4), with an overall aspect ratio of 1.4. In order tion with weight proportions corresponding to 1.0∶0.2∶3.5∶0.85
to facilitate quantifying the slab coupling influence on the wall re- (portland cement: lime: dry sand: water), and having an average
sponse, two individual rectangular walls (W5 and W6) were also flow of 127%. Forty-two mortar cubes were tested in compression
constructed and tested as they presented individual components of according to the CSA A179-14 (CSA 2014d) and resulted in an
206 mm 70 mm
133 mm
194 mm 133 mm
133 mm
2160 mm
1533 mm 1533 mm
2160 mm
337 mm
1533 mm
200 mm
W1 W2 W3
1932 mm
600 mm 133 mm
133 mm
133 mm
80 mm
2160 mm
2160
2160 mm
mm
1533 mm
598 mm 465 mm
W4 W5 W6
Fig. 1. Wall dimensions
V15 V16
Bracing
beam H5
H6 V13 V14
Out-of-plane 4E 1E
Strain gauge
supports H7 V11 V12
4D 3B 2B 1D
Concrete H8 V7 V8 V9 V10
Foundation
V5 4C 1C V6
V3
Downloaded from ascelibrary.org by Universidad de los Andes on 08/29/22. Copyright ASCE. For personal use only; all rights reserved.
V2 V4
V1 4B 1B
H9 Displacement
3A H10 H11 2A
4A 1A Potentiometer
63 mm
Structural Post-tensioned Rigid steel
Horizontal 1533 mm Vertical
Floor steel rods Foundation
reinforcement reinforcement
(a) (b)
Fig. 2. Test setup: (a) layout; (b) external and internal instrumentation
100 outermost bars yielded. Diagonal shear cracks were first observed
Force - control Displacement - control
80 at the onset of yield (1Δy ) and kept increasing in number, length,
60
and width. The shear cracks were mainly observed over two-thirds
of the walls’ first story heights and were concentrated around the
Displacement (mm)
40
wall mid-lengths. At higher load cycles, spalling of masonry oc-
20
curred at wall toes and resulted in exposing the reinforcement. With
0 increased loading, buckling of the reinforcement bars developed
-20 and the outermost bars eventually fractured at both wall ends.
-40 At such point the test was terminated, typically with the walls
Fy
-60 experiencing more than 50% strength degradation.
80% Fy
-80 60% Fy
The slab-coupled Walls W3 and W4 showed different failure
40% Fy mechanisms compared with Walls W1 and W2, where, as expected,
-100
0 5 10 15 20 25 30 these walls failed by forming plastic hinges at the bottom of the two
Number of cycles walls in addition to hinges at the wall/slab interface regions (Fig. 5).
Following a weak beam (slab)/strong column (wall) mechanism,
Fig. 3. Sample loading history for Wall W1 the damage started at the slab, followed by damage at the wall
bases. Due to continued load reversal at high displacement de-
mands, the walls outermost reinforcement bars eventually frac-
tured, and subsequently the test was terminated.
load, while monitoring the onset of yielding of the outermost bars With almost no shear cracks observed, the crack pattern for
using the strain gauges installed at the wall-foundation interface Walls W5 and W6 was dominated by bed joint (flexural) cracks
level. Once the yield load was reached, the walls were cycled twice during initial stages of loading and then head joint cracks were ob-
at multiples of their respective yield displacements until the wall served at the first yield load cycle. At higher displacement levels,
outermost reinforcement bars fractured; at which point the test
the walls exhibited uplift on the tension side accompanied by ver-
was terminated. The following sections discuss the test observa-
tical splitting cracks on the compression toe. Under increased wall
tions followed by analyses of the experimental results from the
top displacement demands, face shell and grout spalling occurred
FBSD perspective.
and was followed by buckling of the outermost reinforcement bars
and their eventual fracture. Fig. 6 shows the extent of cracking in all
Test Results six walls at 20% peak strength degradation.
(d)
Bed joint cracking
(a)
Downloaded from ascelibrary.org by Universidad de los Andes on 08/29/22. Copyright ASCE. For personal use only; all rights reserved.
Masonry spalling
at wall toes
Initiation of shear
cracks at yield load
(e)
63 mm Bar Buckling
(b)
(f)
Horizontal
Wall uplift 1-2 mm reinforcement
Bar fracture
Grout
(c) (g)
Fig. 4. Damage sequence of walls W1 and W2 at % top wall drift: (a) 0.07%; (b) 0.3%; (c) 0.6%; (d) 0.6–1.9%; (e) 1.2–1.5%; (f) 1.5%; (g) 1.8–2%
the displacement recorded experimentally at first yield of the outer- of þ118.5 and −116 kN for the positive and negative loading di-
most vertical reinforcement in each wall is also presented on the rection, respectively.
graphs in Fig. 7. The effect of altering the wall configuration on the lateral
The response of all the walls was approximately linear elastic strength and top drift is illustrated in Fig. 8(b). The peak load
up to the wall yield strength level, corresponding to a wall top of the flanged Wall W2 is 1.4, 2.8, and 4.3 times that of rectangular
drift value that ranged between 0.2 and 0.5%, accompanied by Wall W1 and the slab-coupled Walls W3 and W4, respectively. The
minimal energy dissipation. At higher displacement levels, the flanges in Wall W2 resulted in a reduced compression zone depth,
wider hysteresis loops were characterized by a reduced capacity which in turn resulted in increasing the moment arm and ultimately
at the same displacement level due to stiffness and strength led to higher wall cross section moment capacity. Wall W3 has 45%
degradations. The ultimate load for all the walls was reached higher ultimate load than Wall W4. Compared with Wall W1, the
approximately between 2Δy and 3Δy top displacement, with ultimate capacity of Wall W2 (with the same reinforcement ratio)
all the walls displaying symmetrical responses in both loading increased by 37%.
directions. The responses of the slab-coupled walls were compared with
that of two individual wall components linked together (i.e., through
link members). Fig. 8(c) shows the predicted load-displacement re-
Force-Based Seismic Design Parameters sponse of doubling that of the individual Wall W5 (noted in the
figure as 2W5) and doubling that of Wall W6 (noted in the figure
The load-displacement envelopes for the walls presented in as 2W6) as well as the load-displacement relationship of the cor-
Fig. 8(a) show that Wall W6, with the smallest cross-sectional area, responding slab-coupled Walls W3 and W4, respectively. The in-
had the lowest capacity of þ9.9 and −8.8 kN, whereas Wall W2, creased strength acquired by the slab-coupled walls is attributed to
with the largest cross-sectional area, yielded the highest capacity the coupling moment, M c ¼ Flc , generated by the slabs within the
Slab Slab
rotation rotation
(b) (c)
z F
y
F F
lc
Potential plastic
hinge regions
Mi Mj
Fig. 5. Weak beam/strong pier failure mechanism: (a) slab rotation; (b) west slab interface; (c) east slab interface
coupled wall system where F denotes the axial force either com- The capacities of the slab-coupled wall systems (Walls W3 and
pression or tension developed in the individual walls and lc repre- W4) listed in Table 2 are predicted using the procedure proposed by
sents the net span of the coupling slab (Fig. 5) as will be discussed Harries et al. (2004) and Hoenderkamp (2012). These two ap-
in the next section. proaches are originally based on the continuous medium method
(Chitty 1947), which requires calculation of the degree of coupling
(DOC) parameter between the walls. The DOC indicates the ratio
Wall Strength Predictions between the coupling moment, F · lc , and the overall overturning
moment, M, resisted individually by the two coupled walls’ cross
The flexural capacity of RM shear walls can be accurately predicted sections, M i and M j , in addition to that facilitated through wall
using cross-sectional analysis (Priestley and Elder 1982; Priestley coupling as given Eq. (1). Eq. (2) presents the same relationship
1986). The predicted and the experimental yield and ultimate loads in a format that enables evaluating the coupling moment, Mc ¼
of the walls are presented in Table 2. The value of 0.003 was used F · lc , directly
for the ultimate masonry strain as specified by the CSA S304-14
(CSA 2014c) to calculate the wall ultimate flexural capacity. For F · lc
DOC ¼ ð1aÞ
the yield strength, a linear strain profile, with a yield strain of M i þ M j þ F · lc
the outermost steel reinforcement set to 0.0025, was used. For all
walls, the ratio between the theoretical and the experimental DOC
strength V exp =V pred ranged between 0.9 and 1.2, and between 0.8 F · lc ¼ ðMi þ M j Þ ð1bÞ
1 − DOC
and 1.1, for the yield and ultimate loads calculations, respectively.
The equations used for the calculations summarized in Table 2
are listed in the appendix at the end of the paper for ease of In order to evaluate the coupling moment given by Eq. (2), the
reference. proposed DOC formulation by Hoenderkamp (2012), given in
W4 W5 W6
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Eq. (2), will be used, where C is a constant that depends on the load ðAi þ Aj ÞðI i þ I j Þ
configuration (Hoenderkamp 2012) k¼ 1þ ð5Þ
Ai Aj l2c
2 1 þ C=ðk2 − 1Þ −1
DOC ¼ 1 þ ðk − 1Þ ð2Þ In the Eqs. (4) and (5), lc , hf , Ai , Aj , I i , and I j represent the net
1−C slab span between the walls, the center-to-center floor height, the
cross-sectional area of each of the individual walls, and the indi-
In this respect, Eq. (2) was derived by solving the governing differ- vidual wall’s moment of inertias, respectively. Finally, I c represents
ential equation for coupled walls [Eq. (3)], and enforcing the ap- the moment of inertia of the coupling slab accounting for shear
propriate boundary conditions where dy=dz, Em , M e , and I define deformations and is calculated by Eq. (6)
the slope of the walls centerlines at level z above the base, the elas-
Ib
tic modulus of the masonry walls, the externally applied bending Ic ¼ ð6Þ
12Ec I b
moment and the summation of the two individual walls’ moments 1þ l2c Gc Ac
λ
of inertias, respectively
40
10
0 0
| | | | | |
= µΔ ρh 0.26% 6 5 4 3 2 1 = µΔ ρh 0.26%
6 5 4 3 2 1 -40
-120
-50 -40 -30 -20 -10 0 10 20 30 40 50 -100 -80 -60 -40 -20 0 20 40 60 80 100
(a) Top Displacement (mm) (d) Top Displacement (mm)
60
5
20
0
-20
lw 1533 mm -5 lw 598 mm
-60
Vy hw 2160 mm Vy hw 2160 mm
-100 -10
0.8Vu 0.8Vu
ρv 0.6% Vu ρv 0.6%
Vu -15
-140 | | | | | | | | | | | | |
= µΔ ρh 0.26% = µΔ ρh 0.26%
6 5 4 3 2 1 7 6 5 4 3 2 1
-180 -20
-50 -40 -30 -20 -10 0 10 20 30 40 50 -90 -60 -30 0 30 60 90
(b) Top Displacement (mm) (e) Top Displacement (mm)
Vy
Lateral Load, V (kN)
20 5
0 0
lw 1533 mm lw 465 mm
-20 -5
Vy hw 2160 mm
hw 2160 mm
Vy 0.8Vu
-40 0.8Vu ρv 0.6% -10 Vu ρv 0.6%
| | | | | | |
Vu | | | | |
7 6 5 3 2 = µΔ ρh 0.26% 5 4 3 2 1 = µΔ ρh 0.26%
4 1
-60 -15
-80 -60 -40 -20 0 20 40 60 80 -100 -80 -60 -40 -20 0 20 40 60 80 100
(c) Top Displacement (mm) (f) Top Displacement (mm)
Fig. 7. Load-displacement relationships: (a) W1; (b) W2; (c) W3; (d) W4; (e) W5; (f) W6
2×V individual values increased by 59 and 34%, respectively. As such, the results
V coupled ¼ ð8Þ
1-DOC show that, although no special reinforcement detailing of the slabs
was adopted in the current study, the slabs provided significant wall
A simple comparison to illustrate the strength and stiffness en- coupling that altered the latter’s strength and stiffness characteris-
hancements realized by the slab-coupled walls (compared with the tics. This in turn indicates that ignoring slab-coupling in the
corresponding linked ones) is shown in Fig. 8(c). The numbers in design of RM SFRS may lead to inaccurate estimation of the
the figure show that, by considering the slab coupling effects, design lateral shear capacity, stiffness, and period; an issue that
the yield, V y , and ultimate strength, V u of Wall 5 increased by 83.2 might need to be addressed by the MSJC and CSA S304 seismic
and 44.9%, respectively, and the corresponding lateral stiffness design subcommittees. However, as only two slab-coupled walls
W5 3.6 Rectangular ductility as well as to define the area within the wall where special
20
W6 4.6 Rectangular detailing might be necessary to facilitate plastic curvature develop-
0
ment with minimal (controlled) wall damage. For the tested rectan-
-20 gular and flanged walls (e.g., Walls W1, W2, W5, and W6), plastic
-40 W1 hinging would occur exclusively at the base of the walls. On the
-60 W2 other hand, for the slab-coupled walls, the hinging mechanism de-
W3 pends on the DOC between the walls. For such walls, plastic hing-
Downloaded from ascelibrary.org by Universidad de los Andes on 08/29/22. Copyright ASCE. For personal use only; all rights reserved.
-80 W4
W5 ing can develop at the base of the two walls and at the wall/slab
-100
W6 interfaces.
-120
-100 -80 -60 -40 -20 0 20 40 60 80 100
To date there has been no consensus on quantifying lp for
(a) Top Displacement (mm) rectangular RM walls (Shedid and El-Dakhakhni 2014). In addi-
tion, there are only a few analytical studies that have focused on
Drift (%) evaluating the influence of connecting walls, with different cross-
-4.6 -4 -3.4 -2.8 -2.2 -1.6 -1 -0.4 0.2 0.8 1.4 2 2.6 3.2 3.8 4.4 sectional characteristics, on their plastic hinge lengths (Bohl and
120
Wall A r Configuration V y (kN) V u (kN)
Adebar 2011).
100 Because of their nonlinear behavior, idealizing the load-
W1 1.4 Rectangular 66.9 85.9
80 W2 1.4 Flanged 94.3 117.3 displacement envelopes of RM walls is necessary to quantify
W3 1.4 Coupled I 33.9 42.3
60 W4 1.4 Coupled II 21.3 27.4 their idealized plastic hinge length, lp;ideal based on mechanistic
40 models. According to Tomaževič (1999), the experimental load-
Lateral Load (kN)
20
displacement envelopes can be idealized using three key points:
0
(1) the point where there is a major change in the slope of the
load-displacement envelope, (typically at Δym ); (2) the point
-20
representing the maximum capacity, V u , attained during testing;
-40
and (3) the displacement at 20% peak load degradation, Δ0.8 Vu .
-60 W1 Tomaževič (1999) also suggests that, if the load-displacement
-80 W2 envelope is idealized with a elastic-plastic relationship, then
W3 the idealized elastic-plastic resistance Fep can be evaluated by
-100
W4
-120 equating the energy under the experimental load-displacement
-100 -80 -60 -40 -20 0 20 40 60 80 100 envelope to that under the idealized elastic-plastic relationship [rec-
(b) Top Lateral Displacement (mm)
ognizing that the yield (effective) stiffness K e ¼ Fy =Δy ] as shown
Drift (%)
in Eq. (8)
-4.6 -4 -3.4 -2.8 -2.2 -1.6 -1 -0.4 0.2 0.8 1.4 2 2.6 3.2 3.8 4.4
50 0 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1
Wall Ar Configuration V y (kN) V u (kN)
2 × Aenv A
40 W3 1.4 Coupled I 33.9 42.3 Fep ¼ K e @Δ0.8V u − Δ0.8V 2u − ð9Þ
2 W5
W4
1.4
1.4
Linked
Coupled II
18.8
21.3
29.6
27.4
Ke
30
2 W6 1.4 Linked 14.6 18.8
20
where Aenv = area under the experimental load-displacement
Lateral Load (kN)
10
envelope up to Δ0.8 Vu . Subsequently, the idealized plastic hinge,
0
lp;ideal values (Table 4) can be calculated by solving for lp;ideal in
-10 Eq. (10) (Paulay and Priestley 1992), utilizing the values of μep
Δ0.8 Vu
and μϕ ¼ ϕu =ϕep y , presented in Table 4
-20
-30
W3
2 W5 lp;ideal lp;ideal
-40 W4 μep
Δ0.8Vu ¼ 1 þ 3ðμϕ − 1Þ 1 − 0.5 × ð10Þ
2 W6
hw hw
-50
-100 -80 -60 -40 -20 0 20 40 60 80 100
(c) Lateral Displacement (mm) In the above equation, μep Δ0.8 Vu is the idealized displacement
ductility at 20% peak load degradation, ϕu , ϕep y , and μϕ denotes
Fig. 8. Load-displacement envelopes: (a) all the walls; (b) walls with the ultimate curvature, the idealized yield curvature, and the cur-
same overall aspect ratio; (c) slab-coupled and linked walls vature ductility of the walls, respectively. The curvature ductility
was determined using the idealized yield curvature corresponding
to the idealized yield displacement from the idealized elastic-
plastic relationships depicted in Fig. 9. Walls W1, W2, W3, and
were considered in this study, further testing of slab-coupled W4, with the same overall aspect ratios, showed different idealized
RM walls with different DOC is necessary to gain a better under- equivalent plastic hinge lengths ranging between 0.05lw and 0.32lw
standing of slab-coupled wall behavior under different levels of (Table 4). Walls W5 and W6 had lp;ideal values of approximately
seismic demands. 0.26lw and 0.36lw , respectively. These values will be compared
fu
r ¼ 0.2 × − 1 ≤ 0.08 ð14Þ
fy
with available equivalent plastic hinge length model predictions in
the next section. where lsp and f u denote the strain penetration length and the ulti-
mate strength of the reinforcement, respectively. The prediction
Plastic Hinge Length Predictions results of the three different approaches presented in Table 5 over-
estimate the idealized plastic hinge length for Walls W1 and W2.
Table 5 shows the predicted equivalent plastic hinge length, lp The high μϕ values for Walls W1 and W2 result in lower values for
values from different plastic hinge quantification approaches in- the idealized plastic hinge lengths. For the remaining walls how-
cluding those proposed by Paulay and Priestley (1992), Priestley ever, there is no general trend relating the lp;ideal values and the
and Kowalsky (1998) and Priestley et al. (2007). Approach I, pro- corresponding theoretical predictions. Nevertheless, it can be con-
posed by Paulay and Priestley (1992), is given by Eq. (11), where cluded that Approach I (Paulay and Priestley 1992) yielded better
lp is a function of the wall’s length and the strain penetration length predictions of lp for Walls W1 and W2, while Approach III devel-
(function of the vertical reinforcement diameter, db and its yield oped by Priestley et al. (2007) yields the closest agreement with
strength, fy ) lp;ideal for Walls W4, W5, and W6, respectively. On the other hand,
Approach II (Priestley and Kowalsky 1998) provided the closest
lp ¼ 0.08lw þ 0.022db fy ð11Þ approximation for Wall W3. However, because of the limited
number of walls reported in the current study, analyses of a more
Approach II was proposed by Priestley and Kowalsky (1998) comprehensive experimental RM wall database are necessary to
and is given by Eq. (12), in which the plastic hinge length, lp is generalize such conclusions for RM walls detailed within the
proportional to both lw and hw ductile/special RM SFRS classification.
Table 4. Walls’ Idealized Ultimate Load, Experimental Displacements, and Idealized Plastic Hinge Lengths
Idealized
Theoretical Experimental displacement
displacement (mm) displacement (mm) ductility Curvature ductility
Loading Idealized
Wall direction Fep (kN) Fep =V u (%) Δyth Δuth Δym Δ0.8 Vu μep
Δ0.8 Vu μϕ (ϕu =ϕep
y ) lp;ideal (mm) lp;ideal =lw (%)
W1 þve 83.2 91.9 5.7 9.7 6.0 35.2 4.6 4.6 191.1 12.5
−ve 71.8 88.4 6.4 29.1 4.3 4.3
W2 þve 112 94.5 5.7 37.5 6.8 40.5 4.7 4.7 83.0 5.4
−ve 107.3 92.5 7.1 30.1 3.9 3.9
W3 þve 34.8 84.6 11.5 26.7 10.8 55.6 5.0 5.0 484.3 31.6
−ve 39.3 90.5 9.5 62.0 5.7 5.7
W4 þve 25.1 88.6 11.7 25.8 13.8 61.2 3.9 3.9 278.4 18.2
−ve 25.1 94.5 10.9 66.2 4.9 4.9
W5 þve 13.4 86.5 14.7 21.3 9.5 37.6 2.9 2.9 153.1 25.6
−ve 12.4 90.4 10.1 56.4 3.9 3.9
W6 þve 9.1 92.1 18.7 28.3 13.8 59.7 3.2 3.2 167.6 36.0
−ve 7.5 84.9 15.7 78.6 5.2 5.2
40 25
capacities, V u , averaged from both directions.
Downloaded from ascelibrary.org by Universidad de los Andes on 08/29/22. Copyright ASCE. For personal use only; all rights reserved.
20
30 W3 W4 The ductility values evaluated at maximum load, μΔVu , listed in
15 Table 6 ranged between 1.9 and 4.3, with the highest values
20
10 corresponding to the slab-coupled Walls W3 and W4. The displace-
10 5 ment ductility values at 1%, μΔ1% range between 1.4 and 3.6,
0 0 within which Walls W1 and W2 (the most stiff walls in the test
0 20 40 60 0 20 40 60 matrix) had the highest values as opposed to the slab-coupled Walls
Top Displacement (mm) Top Displacement (mm) W3 and W4, which reached values of only 2.2 and 1.8, respectively.
(c) (d) Such relatively higher ductility values for Walls W1 and W2 are
16
attained due to the walls’ lower yield displacements compared
10 with those of the more flexible walls. At 20% load degradation,
14
Lateral Load (kN)
Lateral Load (kN)
12 8 μΔ0.8 Vu values ranged between 4.0 and 6.5, while the μΔfr values
10 W5 W6 (corresponding to the fracture of the walls’ outermost bars) ranged
6
8
6
between 5.4 and 7.6, with the highest ductility values again corre-
4
4 sponding to the slab-coupled Walls W3 and W4.
2 2 Fig. 8(b) presented earlier shows the influence of alternating
0 0 wall cross-section configuration on its stiffness and displacement
0 20 40 60 0 20 40 60 80 capacities. In this respect, the wall top drift capacities are compared
Top Displacement (mm) Top Displacement (mm)
using the corresponding Δ0.8 Vu and Δfr levels. At Δ0.8 Vu , the
(e) (f)
slab-coupled Walls W3 and W4 have similar percentage drifts
Fig. 9. Average load-displacement envelopes and bilinear idealization (for the (+) and (−) loading directions) of approximately 2.8
curves: (a) W1; (b) W2; (c) W3; (d) W4; (e) W5; (f) W6 and 3.0, respectively. At the Δ0.8 Vu level, the corresponding aver-
age percentage drift from Walls W3 and W4 is higher by 87 and
76% than of Walls W1 and W2, respectively. In addition, the top
wall drifts corresponding to the Δfr are very high for the slab-
coupled walls reaching 3.4%. On the other hand, the rectangular
Wall Ductility Quantification
Wall W1 and the flanged Wall W2 reached only 1.8 and 2.0%
Displacement ductility quantification is key to evaluate the RM drifts, respectively, at the Δfr level.
walls’ inelastic deformation capacities and to facilitate predicting Moment-curvature analysis was used along with Approach III
the drift and damage levels under different level of seismic de- (Priestley et al. 2007) to obtain the theoretical yield and ultimate
mands (Park and Paulay 1975; Paulay and Priestley 1992; Priestley displacements, presented in Table 6. Such values are then used to
et al. 1996; Tomaževič 1999). To date, there has been no consensus calculate the theoretical displacement ductility μth Δ , shown in the
amongst researchers in terms of identifying RM and RC walls’ table, which are compared with the μep Δ0.8 Vu values. The idealized
yield displacement point (Shedid et al. 2010b). As can be observed displacement ductility values at 20% ultimate load degradation,
from Fig. 8(a) presented earlier, none of the walls had a well- μep
Δ0.8 Vu are higher than μΔ by at least 200% for most of the walls,
th
ep
defined yield plateau that can be used to calculate the displacement where the μΔ0.8 Vu values range between 3.4 and 5.4. The exper-
ductility. As such, the equivalent elastic-perfectly plastic RM wall imental ductility values corresponding to μΔ1% , μΔ0.8 Vu and μΔfr
W1 þve 5.7 9.7 12.8 35.2 38.5 2.1 3.6 5.9 1.6 4.6 6.4 1.8 5.1 1.7 2.6
−ve 12.9 29.1 38.6 2.0 3.4 4.5 1.3 4.3 6 1.8 5.7
W2 þve 5.7 37.5 13.9 40.5 42.3 2.0 3.2 6 1.9 4.7 6.2 2.0 5.0 6.6 0.7
−ve 13.9 30.1 43.2 2.0 3.0 4.2 1.4 3.9 6.1 2.0 5.6
W3 þve 11.5 26.7 31.1 55.6 72.6 2.9 2.0 5.1 2.6 5.0 6.7 3.4 6.5 2.3 2.3
−ve 41.2 62 72.6 4.3 2.3 6.5 2.9 5.7 7.6 3.4 6.6
W4 þve 11.7 25.8 31.1 61.2 74.1 2.3 1.6 4.4 2.8 3.9 5.4 3.4 4.8 2.2 2.0
−ve 41.2 66.2 74 3.8 2.0 6.1 3.1 4.9 6.8 3.4 5.5
Downloaded from ascelibrary.org by Universidad de los Andes on 08/29/22. Copyright ASCE. For personal use only; all rights reserved.
W5 þve 14.7 21.3 19.6 37.6 65.7 2.1 2.3 4.0 1.7 2.9 6.9 3.0 5.1 1.4 2.4
−ve 29.4 56.4 65.3 2.9 2.1 5.6 2.6 3.9 6.5 3.0 4.5
W6 þve 18.7 28.3 29.5 59.7 74.1 2.1 1.6 4.3 2.8 3.2 5.4 3.4 4.0 1.4 3.0
−ve 29.6 78.6 102.8 1.9 1.4 5.0 3.6 5.2 6.5 3.4 6.7
a
Theoretical displacement ductility calculated using Approach III plastic hinge length prediction.
b
Average values from positive and negative cycles.
demonstrate that RM walls detailed following the same prescriptive relating the lp;ideal values and the corresponding theoretical predic-
requirements for the ductile/special RM SFRS classification can tions. Nevertheless, it can be concluded that Approach I (Paulay
experience significantly different ductile capabilities based on their and Priestley 1992) yielded better predictions of lp for Walls
configurations. W1 and W2, whereas Approach III developed by Priestley et al.
(2007) yields the closest agreement with lp;ideal for Walls W4,
W5, and W6, respectively, whereas Approach II (Priestley and
Conclusions Kowalsky 1998) provided the closest approximation for Wall W3.
In conclusion, the experimental results presented in this
The paper presented an experimental study on ductile shear walls or study showed that RM walls designed within the same SFRS clas-
special reinforced masonry walls SFRS classification as per the sification level could show a significant difference in their
CSA S304-14 (CSA 2014c) and MSJC-13 (2013) standards, re- load-displacement relationship characteristics and ultimate drift
spectively. The objective of the study was to evaluate key FBSD capacities. The seismic design process has to incorporate the dif-
of flexural dominant individual and slab-coupled RM shear walls. ference in configuration even for walls within the same SFRS
Quantifying such FBSD parameter would facilitate assessing classification level to accurately predict the behavior of walls de-
whether walls with different cross-sectional configurations, but signed and detailed for such level. Moreover, North American
with the same overall aspect ratio and detailed following the same masonry code seismic design subcommittees might find it benefi-
prescriptive SFRS classification requirements, would develop cial to consider incorporating RM slab-coupled walls, as a separate
similar response under seismic events. In general, all walls showed SFRS classification, as they showed better performance than sim-
a ductile behavior and failed in flexure. However, due to the varia- ilar walls with the same overall aspect ratio. This however would
tion in their cross section configurations, crack patterns were differ- require further testing than what is presented in the current study.
ent, with the rectangular and flanged Walls W1 and W2 showing a The companion paper will further investigate displacement- and
combination of flexure and shear cracks. The slab-coupled Walls performance-based seismic design parameters in order to gain more
W3 and W4 failed by forming plastic hinges at the bottom of insight into the behavior of ductile/special RM walls.
the two walls in addition to hinges at the wall/slab interface of
the walls. Walls W5 and W6 exhibited a dominating flexural failure
characterized by bed and head joint cracking, toe crushing, and bar
Appendix. Moment Capacity Calculations for the
fracture at the end of the test. Based on the FBSD parameter quan-
Walls
tification of the test walls, the following conclusions can be made:
Displacement ductility capacities corresponding to different
wall response levels varied when considering different wall con- Formulation Used to Predict the Walls’ Yield Strengths
figurations. Using elastoplastic idealization, on average the ideal-
ized displacement ductility capacities, μep Δ0.8 Vu are higher than
theoretical, μth
Δ by at least 200% for most of the walls. The values P ¼ Cm þ Cs þ T
ranged from 3.4 to 5.4, indicating the variability in the ductility
capacity within the walls. Utilizing the idealized load-displacement
Cm ¼ 0.5 εm Em tc
relationships, Walls W1, W2, W3, and W4 showed different ideal-
ized equivalent plastic hinge length ranging between 0.05lw and
0.32lw . Walls W5 and W6 had an lp;ideal value ranging between Cs ¼ ΣAs0 fs
0.26lw and 0.36lw . The prediction results of the three different ap-
proaches presented in Table 5 overestimate the idealized plastic T ¼ ΣAs f s
hinge length for Walls W1 and W2. The high μϕ values for Walls
W1 and W2 result in lower values for the idealized plastic hinge εy × c
εm ¼
lengths. For the remaining walls however, there is no general trend d−c