Sei sulla pagina 1di 12

ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title No. 116-S20

Experimental Investigation: New Ductility-Based Force


Modification Factor Recommended for Concrete Shear
Walls Reinforced with Glass Fiber-Reinforced Polymer Bars
by Ahmed Hassanein, Nayera Mohamed, Ahmed Sabry Farghaly, and Brahim Benmokrane

Reinforced concrete shear walls are the most common lateral-force- different aspect ratios and one steel-reinforced shear wall
resisting system in reinforced concrete structures. Recent exper- under combined reversed lateral and axial loading. The
imental studies have proven the applicability of using glass GFRP-reinforced shear walls evidenced good performance
fiber-reinforced polymer (GFRP) reinforcement in lateral-resisting in resisting lateral loads associated with adequate strength
concrete structural systems (shear walls and columns). In this
and deformation capacity compared to the steel-reinforced
study, five concrete shear walls reinforced with GFRP bars and
shear wall. The elastic behavior of the GFRP reinforcement
spirals were tested under reversed cyclic quasi-static loading and
constant axial load. The main difference between the walls was kept the hysteresis-envelope tips ascending until failure
the GFRP reinforcement configuration in the boundary elements. with no strength degradation and with an acceptable level
Two shear walls included boundaries reinforced with square GFRP of energy dissipation. Mohamed et al. (2014b) proposed a
spirals, while the third shear wall had boundaries reinforced with deformation capacity model based on the test specimens
circular GFRP spirals. The remaining two shear walls had higher with a minimum level of confinement.
confinement of boundary elements consisting of square GFRP spiral Confinement level is a key parameter that affects defor-
embedded inside rectangular GFRP spiral in one and rectangular mation capacity, as confining the concrete at the boundary
GFRP spiral with two GFRP ties in the other. The main objectives of a shear wall with transverse reinforcement would signifi-
were to assess the impact of increasing the confinement level in the cantly increase the compressive strain at which the concrete
boundaries and the effect on inelastic deformation capacity. The
fails (Wallace and Moehle 1992). In the case of GFRP-
walls with higher confinement clearly achieved higher drift ratios
reinforced shear walls, concrete crushing at the base of the
and strength. The envelope curves were bilinearly idealized. The
elastic-plastic transition point and the maximum deformation limit wall should control the strength, inelastic deformation, and
were identified based on the seismic performance of the test spec- energy dissipation due to the elastic behavior of the GFRP
imens. The recorded inelastic rotation capacity of the test walls bars. In other words, the concrete is the source of deforma-
achieved the required level for lateral-resisting systems. Moreover, bility (Mohamed et al. 2014b) and, therefore, increasing the
the ductility-based force modification factor was assessed and a new level of confined concrete at the boundary of the shear wall
value of 2.4 was suggested for implementation in FRP design codes. would delay the concrete’s compression failure and thus
enhance the shear wall’s deformability. Therefore, this study
Keywords: deformation; force modification factor; glass fiber-reinforced
addressed the effect of the confinement level. In this study,
polymer; seismic; shear wall.
five full-scale GFRP-reinforced shear walls were tested
under quasi-static cyclic loading with different levels of
INTRODUCTION
confinement to assess the effect of confinement on the calcu-
Glass fiber-reinforced polymer (GFRP) reinforcing bars
lated strength and drift capacity. Increasing the confinement
are finding greater acceptance as a construction material in
level of GFRP-reinforced shear walls might lead to modi-
reinforced concrete (RC) structural elements (ACI 440.1R-
fications to the suggested design equations of deformation
15). Numerous research studies have proven that lateral-cy-
capacity and the deformation-demand model proposed by
clic-resisting concrete elements reinforced with GFRP bars
Mohamed et al. (2014b).
exhibit acceptable levels of strength, stiffness, and energy
Moreover, in its provisions for designing shear wall lateral-
dissipation in addition to corrosion resistance (Sharbatdar
resisting elements, the National Building Code of Canada
and Saatcioglu 2009; Mady et al. 2011; Mohamed et al.
(NBCC 2015) adopts a force modification factor for each
2014a,b; Tavassoli et al. 2015).
seismic-force-resisting system (SFRS) to determine the
Observations from earthquakes have indicated that
design lateral force. This force modification factor is deter-
well-designed RC shear walls can be used effectively as a
mined by the equal-displacement approach for the ductili-
primary lateral-load-resisting system in response to both
ty-related force modification factor (Rd) and a statistical
wind and earthquake loading (Fintel 1995). In compar-
approach to evaluate the overstrength-related force modi-
ison to other lateral-resisting systems, RC shear walls have
proven to provide excellent, cost-effective, lateral resis- ACI Structural Journal, V. 116, No. 1, January 2019.
MS No. S-2017-483, doi: 10.14359/51710867, was received December 24, 2017,
tance (Cardenas et al. 1973; Fintel 1995). A recent experi- and reviewed under Institute publication policies. Copyright ©  2019, American
mental study conducted by Mohamed et al. (2014a) involved Concrete Institute. All rights reserved, including the making of copies unless
permission is obtained from the copyright proprietors. Pertinent discussion including
three concrete shear walls reinforced with GFRP bars with author’s closure, if any, will be published ten months from this journal’s date if the
discussion is received within four months of the paper’s print publication.

ACI Structural Journal/January 2019 213


Fig. 1—Cross section and reinforcement details of tested walls.
fication factor (Ro) as reported by Mitchell et al. (2003). 1. Estimate the elastic-plastic and maximum allowable
According to CSA S806 (2012), due to the lack of research deformation;
on the seismic response of FRP-reinforced elements, the 2. Bilinearly idealize the envelope curve;
force modification factor (Rd Ro) is suggested to be equal 3. Reevaluate the values for ductility-related and over-
to unity. Mohamed et al. (2014b, 2015) proposed a value strength force modification factors; and
for the force-reduction factor based on the limited experi- 4. Modify the existing design equations.
mental program with a suggestion of limiting the deforma-
tion capacity to 2.5%. They suggested taking the ductility EXPERIMENTAL PROGRAM
and overstrength force modification factors (Rd and Ro) as Five full-scale concrete shear walls reinforced with GFRP
1.5 and 1.3, respectively, for GFRP-reinforced shear walls. bars were fabricated and tested under the combined action
In this study, based on the previous experimental data in of axial and lateral cyclic loads. All the walls had the same
addition to the recently tested GFRP-reinforced shear wall dimensions with an aspect ratio of 2.33 to be classified as
specimens with higher confinement levels, the force modifi- midrise shear walls according to CSA and ACI codes. The
cation factor and equations to calculate the rotation capacity following sections describe the design and details of the tested
and demand were reevaluated. shear walls, materials, testing procedure, and instrumentation.

RESEARCH SIGNIFICANCE Specimen dimensions and design


FRP design codes and guidelines (ACI 440.1R-15 and The tested GFRP-reinforced shear-wall specimens were
CSA S806 [2012]) do not have seismic provisions due to 200 mm (7.87 in.) in thickness, 1500 mm (59.1 in.) in length,
the lack of seismic experiments on FRP-reinforced concrete and 3500 mm (137.8 in.) in height, resulting in a model of
elements. Hence, this study aimed at filling the knowledge a single midrise shear wall. The five fabricated and tested
gap by assessing the force modification factor and equations GFRP-reinforced shear walls were GX, GnoX, GCi, GDC1,
to calculate the rotation capacity and demand of concrete and GDC2. In these labels, “G” stands for GFRP reinforce-
shear walls reinforced with GFRP bars. For this purpose, ment; “X” and “noX” indicate the presence or absence of
five full-scale GFRP-reinforced concrete shear walls with diagonal bars as sliding-shear reinforcement between the
different levels of confinement were constructed and tested shear-wall specimen and footing, respectively; “Ci” for
under reversed cyclic quasi-static loading and constant axial a circular spiral in the boundary element; and “DC” for
load at the Canadian Foundation for Innovation (CFI) Struc- double-confined boundary elements.
tural Laboratory in the Department of Civil Engineering at Figure 1 shows the cross section and the detailed reinforce-
the University of Sherbrooke, Sherbrooke, QC, Canada. The ment and dimensions of the tested shear walls. Table 1 pres-
objectives of this study were to: ents the dimensions and reinforcement ratios. The design

214 ACI Structural Journal/January 2019


Table 1—Details of test specimens
Wall Axial ratio fc′, MPa b, mm lb, mm lweb, mm ρb, % ρweb, % ρh, % ρv, %
GX 0.15 26.1 200 165 1170 1.73 0.55 1.58 4.0
GnoX 0.15 29.5 200 165 1170 1.73 0.55 1.58 4.0
GCi 0.15 29.1 200 190 1120 1.50 0.57 1.58 3.8
GDC1 0.15 28.9 200 275 950 1.81 0.53 1.58 5.8
GDC2 0.15 24.3 200 275 950 1.56 0.53 1.58 5.0
Notes: ρb is boundary longitudinal ratio; ρweb is web longitudinal ratio; ρh is horizontal reinforcement ratio; and ρv is boundary volumetric ratio. 1 mm = 0.0394 in.; 1 MPa = 145 psi.

of reinforcement details and wall thickness complied with Table 2—Mechanical properties of reinforcement
CSA S806-12, ACI 440.1R-15, CSA A23.3-14, and ACI
Bar db A Ef ffu εf
318-14, where applicable. The shear-wall specimens had
two boundary elements at both sides with different lengths GFRP #3 9.5 71.3 62.5 1346 2.30
and reinforcement ratios. Plane sectional analysis was used GFRP #4 12.7 126.7 61.3 1303 2.35
to calculate the flexural strength. Shear-wall specimens were
Notes: db is bar diameter (mm); A is area (mm2); Ef is modulus of elasticity (GPa);
designed with an adequate amount of distributed and concen- ffu is tensile strength (MPa); and εf is tensile strain (%). 1 mm = 0.0394 in.; 1 mm2 =
trated reinforcement to ensure flexural domination and to 0.00155 in.2; 1 MPa = 145 psi.
prevent shear and sliding-shear failures. Sufficient shear rein-
forcement was provided according to CSA S806-12 and CSA
A23.3-14. Sliding-shear resistance was calculated according
to CSA A23.3-14. Detailed design equations and procedures
can be found in Mohamed et al. (2014a). Ignoring the dowel
action of the longitudinal FRP reinforcement (ACI 440.1R-
15) resulted in one additional layer of diagonal No. 3 GFRP
bars in two directions across the potential sliding planes at
an angle of 45 degrees and spaced at 100 mm (4 in.) in GX.
This additional sliding-shear reinforcement was not used in
GnoX. The footings were heavily reinforced with steel bars
to avoid any influence on wall deformations during the tests.
The footings were used to attach the specimens to the rigid
lab floor and served as an anchorage length for all vertical
bars, as no lap splices were used. Fig. 2—Boundary, horizontal, and vertical reinforcement.
Generally, GX, GnoX, and GCi had the minimum confine-
ment specified in design codes (CSA S806-12, ACI 440.1R-
15) of one spiral in each boundary. GDC1 and GDC2 had
higher confinement of boundary elements, consisting of
square spiral embedded inside rectangular spiral in GDC1
and rectangular spiral with two ties in GDC2 (Fig. 1).

Materials
The specimens were divided into three areas: two boundary
elements on each end and web in the middle. The vertical web
reinforcement consisted of two vertical layers of No. 3 sand-
coated GFRP bars spaced at 120 mm (4.72 in.) extending
with no lap splice into the footing to the full depth of
700 mm (27.6 in.). The horizontal reinforcement consisted of
two layers of No. 4 GFRP bars spaced at 80 mm (3.15 in.).
The clear concrete cover was 25 mm (1 in.) in all cases. The
tensile properties of the GFRP reinforcing bars, presented in
Table 2, were obtained from tensile tests according to ASTM
D7205. Figure 2 shows the vertical, horizontal, tie, and spiral
reinforcement and a typical assembled cage. All the shear
walls were cast with ready mixed normalweight concrete with
a target compressive strength (fc′) of 30 MPa (4.35 ksi).

Test procedure and instrumentation


Figure 3 shows the test setup. A specially fabricated Fig. 3—Test setup.
steel-beam assembly was used to transfer both axial and

ACI Structural Journal/January 2019 215


Fig. 4—Applied-load history.
lateral loads to the shear-wall specimens. An axial load of
0.15Agfc′ was applied and maintained constant by manually
controlling the two hydraulic jacks mounted on both sides
of the steel beam. Moreover, strain gauges were mounted on
the axial-load-applying bars to monitor the axial load during
the test. Cyclic lateral displacements were applied through a
1000 kN (224.8 kip) MTS actuator mounted horizontally to
a reaction wall. The displacement history applied was iden-
tical to that used by Mohamed et al. (2014a) and is shown in
Fig. 4. A lateral bracing system was mounted at the level of
the steel-beam assembly and was fixed to a reaction wall and
rested on the steel-beam through bearings to prevent out-of- Fig. 5—Specimen instrumentation.
plane displacement with no friction resistance to the in-plane the lateral strength kept increasing. Vertical splitting cracks
displacement. typically appeared at the boundaries when the concrete
The shear-wall specimens were instrumented with a compression strain ranged between 3000 and 3500 με.
combination of potentiometers and linear variable differen- After this level, gradual cover spalling was observed with
tial transducers (LVDTs) to measure displacements. Strain the formation of a plastic hinge; Fig. 6(a) presents a typical
gauges were used to measure critical response quantities in crack pattern at failure. Despite the absence of GFRP bar
the vertical and horizontal web reinforcement and the longi- yielding, a plastic-hinge region formed at the lower portion
tudinal reinforcement and spirals in the boundary elements. of the wall due to the plastic deformation of the concrete
Concrete strain in the boundary elements was measured with in the boundaries. The height of the plastic hinge zone was
two LVDTs mounted close to the base of the shear walls. approximately at half of the wall length.
Shear deformations were measured with two sets of four The failure was mainly controlled by concrete crushing as
potentiometers in an X-configuration at two segment heights flexural behavior dominated the response; Fig. 6(b) shows
at 1 and 2 m (39.4 and 78.7 in.) from the base. Four potenti- a typical failure mode of the tested specimens. Failure
ometers were used to measure the flexural deformation in the occurred at the maximum lateral strength of the specimens
boundary elements at 1 and 2 m (39.4 and 78.7 in.) from the in the following sequence: 1) vertical-bar fracture occurred
base. Three potentiometers were used to measure the lateral on the compression side in the boundary element (Fig. 6(c)),
deformation at heights of 0.5, 1.5, and 3.5 m (19.7, 59.1, and leading to a significant increase in the boundary-core internal
137.8 in.) from the base. Figure 5 shows the instrumentation pressure on the confining spirals accompanied with a slight
of the tested shear walls. degradation in axial and lateral strengths; 2) displacement
increased with almost constant lateral load, forming a small
TEST RESULTS AND DISCUSSION horizontal plateau in the load-displacement curve; and 3)
Cracking progress and failure mode another vertical-bar fracture was recorded, followed by rupture
Cracking progressed similarly in all specimens with of the spirals and concrete crushing in the boundary-element
flexural-dominated behavior. The first crack was observed core.
on the tensile side at an approximate height of 300 mm
(11.81 in.) from the base of the specimens and was asso- Lateral load capacity
ciated with significant stiffness degradation. As the applied Table 3 provides the predicted and experimental ultimate
horizontal displacement increased, horizontal cracks gradu- flexural strength. Plane sectional analysis was used to predict
ally formed along the wall height and extended diagonally the flexural behavior through force equilibrium based on the
toward the center of the wall. The length of the horizontal following assumptions: 1) plane sections before bending
cracks originating from the wall edge decreased as the remain plane after bending, which implies that the strains
distance from the wall base increased. The stiffness kept across the depth of a section are proportional to distances
decreasing throughout the test due to crack spreading and

216 ACI Structural Journal/January 2019


Fig. 6—Typical failure pattern.

Table 3—Experimental results


Wall εcu Vexp Vper Vexp/Vper, % Δu Drift V1st crk Vsplit Vspalling
GX 0.013 475 469 101.2 113.2 3.23 176 318 385
GnoX 0.013 498 496 100.5 116.1 3.32 196 315 361
GCi 0.012 492 507 97.0 100.0 2.86 181 281 399
GDC1 0.016 693 652 106.2 156.0 4.46 187 344 436
GDC2 0.015 594 571 104.0 146.8 4.19 161 309 422

Notes: εcu is concrete compression strain at failure; Vexp is experimental ultimate load (kN); Vper is predicted ultimate load based on plan sectional analysis (kN); Δu is displacement
corresponding to ultimate load (mm); “Drift” is drift values corresponding to ultimate load (Δu/lw); V1st crk is load at first crack (kN); Vsplit is load at concrete cover splitting (kN); and
Vspalling is load at concrete cover spalling (kN). 1 mm = 0.0394 in.; 1 kN = 0.225 kip.

from the neutral axis; 2) perfect bond was assumed between to a lateral load of 176, 196, 181, 187, and 161 kN (39.6,
the GFRP bars and concrete; and 3) the experimentally 44.1, 40.7, 42, and 36.2 kip) for GX, GnoX, GCi, GDC1,
measured concrete compression strain at ultimate strength and GDC2, respectively. Cracks were propagated both in the
was used for the calculations. The agreement between web and along the wall height, nearly reaching half the wall
predicted and experimental values confirms the applica- height at a concrete compression strain ranging between
bility of using plane section analysis to predict the flexural 3000 and 3500 με. Cover splitting started to visually appear
strength of FRP-reinforced shear walls. on the boundary element side at cycles +35 to +40 mm (1.4
to 1.6 in.), corresponding to a lateral load of 304, 345, 315,
Hysteretic response 327, and 317 kN (68.3, 77.6, 70.8, 73.5, and 71.3 kip) for
Figure 7 shows the load-displacement hysteretic responses GX, GnoX, GCi, GDC1, and GDC2, respectively. Spalling
for the test specimens. In general, all specimens reached their of the concrete cover was not observed at that point, but
ultimate predicted strength with pinched hysteretic loops wall stiffness was significantly reduced due to the initiation
with no strength decay up to failure. The reloading branches of inelastic compression deformation in the concrete in the
followed a loading path like the previous cycle path, but at a boundaries. Gradual spalling of concrete cover was associ-
lower loading stiffness, resulting in lower peak strength. An ated with the progress of flexural-shear cracks along the wall
initial linear branch corresponding to the uncracked condi- height, which continue to propagate near failure. GDC1 and
tion of the wall was evident. The first crack was formed at the GDC2 attained a higher displacement level of 4.46% and
lateral displacement of +8.00 mm (0.31 in.), corresponding 4.19% drift, respectively, due to the higher level of confine-

ACI Structural Journal/January 2019 217


Fig. 7—Hysteresis response.
ment provided by the rectangular spiral with square spiral compressive strain at which the concrete fails and, thus,
or ties compared to GX, GnoX, and GCi, which achieved delay compression failure of the concrete there. It can help
3.23%, 3.32%, and 2.86% drift. to form a plastic hinge at the base of the wall before the
The diagonal X-bars had no effect, as no sliding was observed collapse and ensure that the wall fails in a less brittle manner.
between the wall and base of GnoX; no differences were That is why ACI 318-14, CSA A23.3-14, and ASCE/SEI
observed in the strength and deformation of GX and GnoX. 41-13 require boundary element confinement for shear walls
subject to earthquake loading.
Effect of confinement configuration With increasing applied loads, the concrete core in the
Confining the concrete at the base of a shear wall with boundary elements compresses and starts to expand. GFRP
transverse reinforcement would significantly increase the spirals and ties restrain such expansion and lead to a substan-

218 ACI Structural Journal/January 2019


Fig. 8—Concrete compression strain.
tial improvement in strength and ultimate strain. The test
specimen results clearly show that increased confinement
level at the boundary element significantly enhanced both
the lateral strength level and drift ratio (Table 3). Figure 8
shows the developed concrete compression strain for the
tested walls throughout the loading cycle. The developed
compression strain was similar for the five shear walls up to
1% drift ratio, which corresponds to a concrete compressive
strain of 3000 με. GCi with circular spirals at the boundary
Fig. 9—Determination of lateral design force V.
developed the lowest strain rate compared to the other spec-
imens with square and rectangular spirals due to the lower where Rsize accounts for overstrength in code-restricted
strength capacity of the bent portion of the square and rect- dimensions for design; R accounts for overstrength of
angular spirals. The concrete compressive strain achieved material resistance factors defined in CSA standards; Ryield
at failure in the specimens with higher volumetric ratios is the ratio of “actual” yield strength to minimum speci-
(GDC1 and GDC2) was 40% higher than that of the walls fied yield strength; Rsh accounts for the overstrength due
with lower volumetric ratios (GX, GnoX, and GCi) due to to the development of strain hardening; and Rmech is the
the higher confinement at the boundary. overstrength of mobilizing the full capacity of the structure
such that a collapse mechanism is formed. All the parameters
FORCE MODIFICATION FACTORS in Eq. (2) can be derived statistically, exhibiting the difference
Force modification factors are essential seismic design between design solutions and actual behavior. The rationale for
tools that define the level of inelasticity expected in struc- calculating each parameter was given by Mitchell et al. (2003).
tural systems during an earthquake event. They also reflect The ductility-related modification factor (Rd) is a measure
the structure’s capacity to dissipate energy through inelastic of the global nonlinear response of a structure and depen-
behavior. They are used to reduce the design forces in dent on building period and ductility. Therefore, to estimate
earthquake-resistant designs, which accordingly would the ductility modification factor for the GFRP-reinforced
reduce construction costs. The force modification factors concrete walls, the actual strength-displacement response
account for damping, energy-dissipation capacity, and should be idealized with the linearly elastic-perfectly plastic
structure overstrength. According to the NBCC (2015), the curve. To generate the bilinear idealization, a well-defined
seismic design load (V) is determined by reducing the elastic explicit transition between the elastic and inelastic deformation
seismic load (Ve) by ductility-related modification factor (Rd) must be clear to identify the elastic-plastic transition point. In
and overstrength-related modification factor (Ro), as shown addition, the maximum deformation limit should be identified.
in Fig. 9 and expressed by the following equation
Maximum deformation limit
S (Ta ) M v I eW Various assumptions are available in the literature for
V= (1)
Rd Ro steel-reinforced structural elements to identify the maximum
deformation capacity (Δu) of shear walls. Park (1989)
where S(Ta) is the design spectral acceleration with a proba- suggested four possible assumptions: 1) maximum displace-
bility of exceedance of 2% in 50 years (2475-year return) at ment corresponding to a limit value for compression strain
the selected period; Mv is the higher mode effect factor; W is (εc); 2) displacement corresponds to ultimate load-carrying
the seismic weight; and Ie is the structure importance factor. capacity (Pu); 3) post-peak displacement at 20% reduction
To account for the various components contributing to of the load-carrying capacity; and 4) displacement at which
the overstrength-related force modification factor Ro, the the longitudinal or transverse reinforced material fractures
following formulation is used under tension or buckles under compression. The latter three
assumptions are not applicable to GFRP-reinforced shear
Ro = RsizeRRyieldRshRmech (2) walls, as there was no strength degradation up to the failure
point (refer to Fig. 7). Accordingly, and as recommended by

ACI Structural Journal/January 2019 219


Mohamed et al. (2014b), the ultimate displacement ended
at the ultimate lateral strength, where the longitudinal rein-
forcement fractured under compression, followed by spiral
rupture and concrete crushing. Using a limit value for
compression strain in FRP-reinforced shear walls would be
more conservative to prevent collapse. The measured strains
in the GFRP reinforcement and concrete through the failure
progression were as follows:
1. The initial stiffness slope for all walls was similar until
the first crack initiated at nearly 0.2% drift and at different
lateral strengths. The measured strains were minimal near
the base in both the concrete and longitudinal bars (less than
500 µε). As the confinement was not activated at that load
level, the measured stains at the spirals were not significant.
2. At a lateral drift ranging from 0.7% to 0.8%, the Fig. 10—Equal-energy idealized bilinear elastic-perfectly
concrete compressive strain was between 3000 and 3500 µε. plastic response.
Concrete cover splitting was observed, indicating the start of capacity to maintain a conservative value and avoid collapse.
inelastic deformations due to the progression of microcracks in This maximum allowable limit of Δu was 2.5% drift, which is
the boundary element core. At this point, strains in the spirals the maximum drift under seismic loads allowed under ASCE
were initiated and measured between 800 and 1000 µε. The 7 (2010) and NBCC (2015). Confinement details should be
corresponding strains in the longitudinal bars on the tension considered in the design procedure based on the concrete
side were 4000 to 6000 µε. compressive strain achieved at the suggested value of 2.5%
3. The concrete cover gradually spalled until the lateral drift (Wallace and Moehle 1992; Paulay and Priestley 1995;
drift reached 2%. The concrete compressive strain was in a Saatcioglu et al. 2013; Mohamed et al. 2014a,b).
range of 6500 to 7000 µε. Hysteresis loops were stable and
lateral strength increased with no decay. Strains in the longi- Elastic-plastic transition point
tudinal bars were redistributed along the height away from The elastic-plastic transition point of a structural element
the base, causing new cracks to propagate along the height is defined as the point at which it begins to produce plastic
and existing cracks to extend through the web. Strains in deformations, while ductility is a measure of its ability to
the spirals varied between 1500 and 2000 µε. The maximum deform plastically without substantial reduction in strength.
recorded strains in the outermost longitudinal bars on the Plastic deformation starts in steel-reinforced structural
tension side were in the range of 9000 to 12,000 µε. elements as soon as the steel bars reach their yield strength.
4. As the lateral drift increased, the specimens continued There are several methods for obtaining yield displacement
to carry the load without any decay or instability. The cracks in steel-reinforced shear walls through the bilinear idealiza-
extended, however, all over the web and a sign of residual tion of the load-displacement hysteretic response.
strains in longitudinal bars corresponding to zero displace- The literature contains several useful discussions about
ments was observed. The average drift ratio for this level of the appropriate definition of elastic-plastic transition point
damage was 3% drift. Concrete compressive strains measured (Δy) for steel-reinforced shear walls (Park 1989; Paulay and
from 9800 to 10,000 µε. The corresponding strains in the Priestley 1995). The best procedure based on the experi-
longitudinal bars on the tension side were 14,000 to 15,000 µε mental results was used. Park (1989) discussed the defini-
and, in the spirals, measured from 3000 to 3400 µε. tions of Δy. The yield point (Δy) was defined as one of: 1)
5. Failure of the tested walls occurred when a longitudinal yield displacement of the equivalent elastic-plastic system
bar fracture occurred on the compression side, accompanied with the same elastic stiffness and ultimate load as the real
with axial strength degradation and a significant increase in system; 2) yield displacement of the equivalent elastic-plastic
the confining pressure leading to spiral rupture followed by system with the same energy absorption as the real system;
crushing of the boundary concrete core. The recorded strains and 3) yield displacement of the equivalent elastic-plastic
at failure were concrete compressive strains of 12,000 to system with reduced stiffness, found as the secant stiffness
17,000 µε, longitudinal bar strains on the tension side of at either first yield or at 0.75 of ultimate load (Pu), whichever
20,000 to 22,000 µε, and less than 7000 µε on the spirals. is less, where 0.75Pu is due to cracking.
The ultimate displacement (Δu) of steel-reinforced shear Other common procedures used in Europe, Australia,
walls is generally defined as a 20% loss of ultimate lateral North America, and Japan for estimating the idealized curve
strength (NBCC 2015). In contrast, the actual ultimate from the actual load-displacement curves generated from
displacement in the ends of GFRP-reinforced shear walls laboratory tests are explained and reviewed in Munoz et al.
occurs at the ultimate lateral strength at which the spiral (2008) and Mohamed et al. (2014b).
ruptures and longitudinal reinforcement fractures under Based on these reviews of existing idealization methods
compression. In this study, however, the authors defined for concrete shear walls as well as data interpretation proce-
the maximum allowable displacement (Δu) of the GFRP- dures for nonlinear testing (Park 1989), the authors decided
reinforced concrete shear walls corresponding to an earth- to incorporate the equivalent energy elastic-plastic (EEEP)
quake to be less than the actual ultimate displacement method for bilinear idealization (Fig. 10). It should be noted,

220 ACI Structural Journal/January 2019


Table 4—Ductility force modification factor
Px/Pu
Wall Δe idz, mm (Fig. 12) P1, kN P2, kN Rd = P2/P1
GX 22.4 56% 388 1014 2.61
GnoX 18.9 57% 411 1183 2.88
GCi 20.6 54% 404 1107 2.74
GDC1 27.3 40% 439 1068 2.43
GDC2 23.0 43% 397 1020 2.57

Notes: 1 mm = 0.0394 in; 1 kN = 0.225 kip.

The ductility-based force modification factor (Rd) is


defined as the ratio between elastic lateral load (P2) and
the idealized wall capacity (P1). As listed in Table 4, the
GFRP-reinforced concrete shear walls had Rd values ranging
from 2.4 to 2.9. Based on the estimated values, the lower-
bound value of 2.4 for (Rd) is recommended for the tested
walls, despite the pinched behavior of the GFRP-reinforced
walls. Recent research (Sharbatdar and Saatcioglu 2009;
Mohamed et al. 2014a) indicates that hysteretic pinching
without strength degradation does not produce undesirable
responses and can result in less structural deformation than
elastic-perfectly plastic behavior. Therefore, the behavior of
Fig. 11—Determination of equivalent linear elastic response. the tested GFRP-reinforced shear walls can be accepted in
resisting lateral forces. Moreover, the lack of early strength
however, that the idealized curve of the GFRP-reinforced
decay indicates no lack in shear capacity, concrete confine-
concrete shear walls using 0.4Pu did not capture the actual
ment, or joint and bond failure, thereby justifying that the
initiation of inelastic deformation, which was experimen-
pinched hysteretic response has no effect on the proposed
tally observed for walls with lower confinement levels (GX,
Rd = 2.4.
GnoX, and GCi) and which led to actual stiffness degrada-
tion higher than the idealized curve.
DESIGN OF GFRP-REINFORCED CONCRETE
SHEAR WALLS
Ductility-based force modification factor (Rd)
To ensure adequate seismic performance, CSA A23.3-14
To estimate the ductility-modification factor (Rd), the
and ACI 318-14 require the inelastic rotation demand (θid)
actual strength-displacement response should be idealized
and inelastic rotation capacity (θic) to be calculated and
with the linearly elastic-perfectly plastic curve. As stated,
compared when designing steel-reinforced concrete shear
however, using 0.4Pu to estimate the idealized curve would
walls. Based on the same hypothesis, the rotation demand
not produce accurate values of the elastic-plastic transition
(θd) for GFRP-reinforced concrete shear walls should be less
point. The concrete is the source of plasticity in the FRP-
than the rotation capacity (θc)
reinforced concrete shear walls and, therefore, the confine-
ment level could affect plasticity level and, in return, affect
θc ≥ θd (3)
the percentage at which the elastic-plastic point is defined.
Therefore, having a constant value (0.4Pu according
The total rotation demand (θd) consists of two components:
to EEEP) is not applicable. Based on that, the actual
the elastic rotation demand (θed) and the inelastic rotation
elastic-plastic transition point was used with the maximum
demand (θid), which are illustrated in Fig. 12. The demanded
deformation point to obtain the corresponding percentage
elastic and inelastic rotations (θed and θid, respectively) can
of ultimate load intersection (Fig. 10). The best percentage
be expressed as follows
of the maximum load was approximately 0.4Pu for highly
confined walls (GDC1 and GDC2) and 0.55Pu for lower
θd = θed + θid (4)
confinement levels (GX, GnoX, and GCi).
According to current design codes, the elastic-plastic tran-
θed = ∆ed/hw (5)
sition point (Δe) represents the design capacity (strength
capacity) equal to or exceeding the required factored
∆ id
code-specified seismic force (P1), which is the maximum θid = (6)
load, as shown in Fig. 11. P2 is the design seismic force due (hw − 0.5λ p )
to an earthquake of intensity as specified in the given seismic
map area but corresponding to full elastic structural response; Combining Eq. (4), (5), and (6), the rotation demand can
P2 was obtained using equal-energy principle (Fig. 11). be expressed as

ACI Structural Journal/January 2019 221


Fig. 12—Elastic and inelastic deformations of shear walls.

∆ ed ∆ id
θd = + (7)
hw (hw − 0.5λ p )

The elastic displacement (Δed) was estimated as the total


displacement (Δtotal) divided by the force modification factor
(R), resulting in the following expressions for calculating the
total rotation demand (θd), as proposed by Mohamed et al.
(2014b)
Fig. 13—Plastic-hinge length analysis.
∆  ∆ (1 − 1/R ) 
θ d = total +  total  (8)
Rhw  (hw − 0.5λ p )  h 
θc = φec  w − l p  + φc l p (12)
 3 
∆ ed ( Rhw − 0.5λ w )
θd = (9) where the elastic curvature capacity (ec) was experimen-
hw (hw − 0.5λ w )
tally calculated based on the longitudinal-bar strain gauges
by idealizing the curvature profile over height (Fig. 13),
where the elastic displacement demand (Δed) can be calcu-
corresponding to the maximum deformation limit at 2.5%
lated from the nominal strength of the shear wall (P) as
drift. The curvature profile for each wall specimen was
follows
idealized into a portion of the linear curvature (ec) over the
height and a portion of the equivalent inelastic curvature
 h3 h  (ic) over the plastic hinge length, as depicted in Fig. 13.
∆ ed = P  w + w  (10)
3 E
 c eI Gc Ae  Figure 14 provides the resultant idealized curvature profiles.
The elastic curvature values were 0.014, 0.013, 0.012, 0.022,
where P is the lateral design force; hw is the wall height; and 0.021 rad/m (0.00036, 0.00033, 0.0003, 0.00056 and
lw is the wall length; EcIe is the effective stiffness taken as 0.00053 rad/in.) for GX, GnoX, GCi, GCD1, and GCD2,
a suggested percent of the gross stiffness of the uncracked respectively. Accordingly, the elastic curvature capacity was
section (EcIg); Gc is the concrete shear modulus assumed as set to 0.018/lw, equal to the lowest measured elastic curva-
0.4 of the elastic modulus (Ec); and Ae is the effective shear ture (0.012 rad/m [0.0003 rad/in.]).
area (area between the extreme compression fiber and the The plastic-hinge length (lp) was calculated from the
center of the tension reinforcement). The effective shear area curvature profile along the height (Fig. 14) and ranged from
of an elastic homogenous uncracked member with a slender 0.5lw to 0.7lw. The lower limit of 0.5lw was set for calculating
section is equal to the rotation capacity as a conservative value. The ultimate
curvature capacity (c) is related to the ultimate concrete
Ae = Ag(5 + 5ν)/(6 + 5ν) (11) compressive strain (εcu) corresponding to the maximum
specified deformation limit (2.5% drift) and the compression
where Ag = bwlw and ν = 0.2, which is a typical Poisson’s zone length (c), as follows
ratio for concrete.
On the other hand, the rotation capacity (θc) was estimated c = εcu/c (13)
based on experimental strain profiles as the average curva-
tures over the wall height by using the idealized curvature where the compression-zone length c was calculated based
profile illustrated in Fig. 13. Based on the figure, the rotation on the plane-sectional analysis at the same deformation limit
capacity consists of elastic and inelastic portions and can be (2.5% drift). Using these specified limits of curvature and
expressed based in Eq. (12) (Mohamed et al. 2014b) plastic hinge length in Eq. (12), the rotation capacity equa-

222 ACI Structural Journal/January 2019


Fig. 14—Curvature profiles at 2.5% drift.

Table 5—Rotation capacity and demand


θc, % (Eq. (14))
Wall εcu at 2.5% drift Δed, mm c, mm 2.5% drift At failure θd, % (Eq. (9))
GX 0.010 55.0 436 2.22 2.98 1.79
GnoX 0.010 55.1 429 2.25 3.03 1.80
GCi 0.009 57.1 348 2.44 3.30 1.94
GDC1 0.012 61.5 431 2.59 3.85 2.26
GDC2 0.011 59.7 397 2.58 3.41 2.13
G10 *
0.007 54.2 288 2.57 3.13 1.85
G12* 0.007 51.9 337 2.31 2.85 1.88
G15 *
0.007 55.9 393 2.34 2.94 1.75
*
GFRP-reinforced walls tested by Mohamed et al. (2014a).

tion proposed by Mohamed et al. (2014b) can be modified (Eq. (9)) at the 2.5% drift limit and at failure. As listed in
and expressed as follows Table 5, the calculated rotational capacities of all the shear
walls exceeded the rotational demand values. Moreover,
ε cu lw 0.006 hw they also achieved the desired rotation capacity of 2.5% set
θc = − 0.009 + (14)
2c lw by NBCC (2015) and ASCE/SEI 41 (2013).

Equation (14) was proposed based on the modification CONCLUSIONS


of the equation presented by Mohamed et al. (2014b) to An experimental study comprising five full-scale
account for the confinement effect at the boundary elements. GFRP-reinforced concrete shear walls with different levels
Table 5 presents the calculated rotation capacity (Eq. (14)) of confinement was conducted to estimate the ductility-based
of the tested walls in the current study and the three walls force modification factor and evaluate the calculated rota-
tested by Mohamed et al. (2014a). The calculated rotation tion demand and capacity. The concluding remarks can be
capacity was compared to the calculated rotation demand summarized as follows:

ACI Structural Journal/January 2019 223


• The tested concrete walls achieved an acceptable level the Canada Research Chair in Advanced FRP Composite Materials for Civil
Structures, the NSERC Research Chair in FRP Reinforcement for Concrete
of deformation according to NBCC (2015) and ASCE/ Infrastructure, the Fonds de la recherche du Québec en nature et technol-
SEI 41 (2013) as well as a stable hysteretic response ogies (FRQ-NT), the Canadian Foundation for Innovation (CFI), and the
with no sign of premature failure. technical staff of the structural lab in the Department of Civil Engineering
at the University of Sherbrooke.
• The diagonal X-bars had no effect as no sliding between
the wall and base was observed in GnoX.
• The presence of more confinement in the boundary REFERENCES
ACI Committee 318, 2014, “Building Code Requirements for Struc-
element (GDC1 and GDC2) clearly enhanced the tural Concrete (ACI 318-14) and Commentary (ACI 318R-14),” American
strength and deformation capacity of the walls. Concrete Institute, Farmington Hills, MI, 520 pp.
ASCE/SEI 41-13, 2013, “Seismic Evaluation and Retrofit of Existing
• The plane sectional analysis was confirmed as being Buildings,” American Society of Civil Engineers, Reston, VA, 518 pp.
applicable in predicting the ultimate flexural strength of ASTM D7205/D7205M-06(2011), 2011, “Standard Test Method for
the GFRP-reinforced shear walls. Tensile Properties of Fiber Reinforced Polymer Matrix Composite Bars,”
ASTM International, West Conshohocken, PA, 13 pp.
• The elastic-plastic transition point could be identified Canadian Commission on Building and Fire Codes, 2015, “National
based on the seismic performance of the test specimens Building Code of Canada (NBCC),” National Research Council of Canada,
and was found to correspond to a concrete compressive Montreal, QC, Canada.
Canadian Standards Association, 2014, “Design of Concrete Structures
strain ranging from 3000 to 3500 µε. The maximum Standard (CSA A23.3-14),” Mississauga, ON, Canada, 240 pp.
deformation limit was set to 2.5% drift. Canadian Standards Association, 2012, “Design and Construction of
• The ductility force modification factor was evaluated Building Components with Fiber-Reinforced Polymers (CSA S806-12),”
Mississauga, ON, Canada, 208 pp.
based on the idealized curve and found to range from 2.4 Cardenas, A. E.; Hanson, J. M.; Corley, W. G.; and Hognestad, E., 1973,
to 2.9. A conservative value of Rd = 2.4 is recommended. “Design Provisions for Shear Walls,” ACI Journal Proceedings, V. 70,
• For the test specimens, the achieved rotation capacity No. 3, pp. 221-230.
Fintel, M., 1995, “Performance of Buildings with Shear Walls in Earth-
exceeded the evaluated rotation demand as required by quakes of the Last Thirty Years,” PCI Journal, V. 40, No. 3, pp. 62-80. doi:
seismic design guidelines. 10.15554/pcij.05011995.62.80
This study is expected to provide the impetus for building Mady, M.; El-Ragaby, A.; and El-Salakawy, E., 2011, “Seismic Behavior
of Beam-Column Joints Reinforced with GFRP Bars and Stirrups,” Journal
concrete shear walls reinforced with GFRP bars and consti- of Composites for Construction, ASCE, V. 15, No. 6, pp. 875-886. doi:
tute a step toward using GFRP bars in such lateral-resisting 10.1061/(ASCE)CC.1943-5614.0000220
systems. Analytical studies that incorporate the load- Mitchell, D.; Tremblay, R.; Karacabeyli, E.; Paultre, P.; Saatcioglu, M.;
and Anderson, D. L., 2003, “Seismic Force Modification Factors for the
displacement relationships of shear walls are, however, Proposed 2005 Edition of the National Building Code of Canada,” Cana-
needed before final provisions for seismic design can be dian Journal of Civil Engineering, V. 30, No. 2, pp. 308-327. doi: 10.1139/
proposed. This includes nonlinear time-history dynamic l02-111
Mohamed, N.; Farghaly, A. S.; Benmokrane, B.; and Neale, K. W.,
analysis of GFRP-reinforced structures, designed according 2014a, “Experimental Investigation of Concrete Shear Walls Reinforced
to the recommended procedure, under selected earthquake with Glass Fiber-Reinforced Bars under Lateral Cyclic Loading,” Journal
records, as well as shake-table testing. These studies would of Composites for Construction, ASCE, V. 18, No. 3, p. A04014001 doi:
10.1061/(ASCE)CC.1943-5614.0000393
allow for better assessment of the demand for this type of Mohamed, N.; Farghaly, A. S.; Benmokrane, B.; and Neale, K. W.,
shear wall and building. 2014b, “Drift Capacity Design of Shear Walls Reinforced with Glass
Fiber-Reinforced Polymer Bars,” ACI Structural Journal, V. 111, No. 6,
Nov.-Dec., pp. 1397-1406. doi: 10.14359/51687099
AUTHOR BIOS Mohamed, N.; Farghaly, A. S.; and Benmokrane, B., 2015, “Seismic
Ahmed Hassanein is a Doctoral Candidate in the Department of Civil Engi-
Response Modification Factors for GFRP-Reinforced Concrete Shear
neering at the University of Sherbrooke, Sherbrooke, QC, Canada. His research
Walls,” The 11th Canadian Conference on Earthquake Engineering, Cana-
interests include testing of concrete structures reinforced with fiber-reinforced
dian Association for Earthquake Engineering (11CCEE), Victoria, BC,
polymers and seismic analysis of reinforced concrete structures.
Canada, 8 pp.
Munoz, W.; Salenikovich, A.; Mohammad, M.; and Quenneville, P.,
Nayera Mohamed is an Assistant Professor at Assiut University, Assuit
2008, “Determination of Yield Point and Ductility of Timber Assemblies:
City, Egypt. She received her BSc and MEng from Assiut University and
in Search for a Harmonised Approach,” Proceedings of Meeting 41 of
her PhD from University of Sherbrooke. Her research interests include
CIB-W18, St. Andrews, NB, Canada.
seismic analysis of reinforced concrete structures and behavior of struc-
Park, R., 1989, “Evaluation of Ductility of Structures and Structural
tural concrete reinforced with fiber-reinforced polymers.
Assemblages from Laboratory Testing,” Bulletin of the New Zealand
National Society of Earthquake Engineering, V. 2, No. 3, pp. 155-166.
Ahmed Sabry Farghaly is a Research Associate in the Department of Civil
Paulay, T., and Priestley, M. J. N., 1995, Seismic Design of Reinforced
Engineering at the University of Sherbrooke, Sherbrooke, QC, Canada. He
Concrete and Masonry Buildings, John Wiley and Sons, New York, 735 pp.
received his MEng and PhD from Hokkaido University, Sapporo, Japan.
Saatcioglu, M.; Palermo, D.; Ghobarah, A.; Mitchell, D.; Simpson, R.;
His research interests include nonlinear analysis of reinforced concrete
Adebar, P.; Tremblay, R.; Ventura, C.; and Hong, H., 2013, “Performance
structures and seismic behavior of structural concrete reinforced with
of Reinforced Concrete Buildings during the 27 February 2010 Maule
fiber-reinforced polymers.
(Chile) Earthquake,” Canadian Journal of Civil Engineering, V. 40, No. 8,
pp. 693-710. doi: 10.1139/cjce-2012-0243
Brahim Benmokrane, FACI, is a Professor of civil engineering and Tier-1
Sharbatdar, M. K., and Saatcioglu, M., 2009, “Seismic Design of FRP
Canada Research Chair in Advanced Composite Materials for Civil Struc-
Reinforced Concrete Structures,” Asian Journal of Applied Sciences, V. 2,
tures, and NSERC Research Chair in FRP Reinforcement for Concrete
No. 3, pp. 211-222. doi: 10.3923/ajaps.2009.211.222
Infrastructure in the Department of Civil Engineering at the University of
Tavassoli, A.; Liu, J.; and Sheikh, S., 2015, “Glass Fiber-Reinforced
Sherbrooke, Sherbrooke. He is a member of ACI Committee 440, Fiber-
Polymer-Reinforced Circular Columns under Simulated Seismic Loads,”
Reinforced Polymer Reinforcement.
ACI Structural Journal, V. 112, No. 1, Jan.-Feb., pp. 103-114.
Wallace, J. W., and Moehle, J. P., 1992, “Ductility and Detailing
ACKNOWLEDGMENTS Requirements of Bearing Wall Buildings,” Journal of Structural
The authors would like to express their special thanks and gratitude to the Engineering, ASCE, V. 118, No. 6, pp. 1625-1644. doi: 10.1061/
Natural Science and Engineering Research Council of Canada (NSERC), (ASCE)0733-9445(1992)118:6(1625)

224 ACI Structural Journal/January 2019

Potrebbero piacerti anche