h —c—c—c—c— behavior of precast, prestressed concrete ... · pdf...
TRANSCRIPT
Spr ing 2011 | PCI Journal42
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
Editor’s quick points
n This paper describes the structural behavior of precast, prestressed concrete sandwich wall panels reinforced with a carbon-fiber-reinforced polymer (CFRP) shear grid to achieve composite action.
n Use of CFRP as a shear transfer mechanism was intended to increase thermal insulation efficiency, enhance service life, and increase overall structural capacity
n Test results of the experimental program were compared with theoretical predictions of fully composite and noncomposite actions to evaluate the percent composite action and to assess the optimum panel configuration.
Behavior of precast, prestressed concrete sandwich wall panels reinforced with CFRP shear gridBernard A. Frankl, Gregory W. Lucier, Tarek K. Hassan, and Sami H. Rizkalla
Precast, prestressed concrete sandwich wall panels are typically used for building envelopes. Such panels con-sist of two outer layers of precast, prestressed concrete separated by an inner layer of insulation. The panels can support gravity loads from floors or roofs, resist normal or transverse lateral wind loads, insulate a structure, and provide interior and exterior finished wall surfaces. Typical panels are fabricated with heights up to 45 ft (14 m) and with widths up to 12 ft (3.7 m). Concrete wythe thick-nesses range from 2 in. to 6 in. (50 mm to 150 mm) with overall panel thicknesses ranging from 5 in. to 12 in. (130 mm to 300 mm).
Precast, prestressed concrete sandwich wall panels may be designed as noncomposite, partially composite, or fully composite. Defining and designing for a partial degree of composite action can significantly increase the structural efficiency and reduce both initial and life-cycle costs of a panel, compared with the fully noncomposite case. The degree of composite action depends on the nature of the connections between the two concrete wythes. Commonly used shear transfer connectors include wire trusses, bent wires, and solid zones of concrete penetrating the insula-tion wythe (Fig. 1). Increasing the degree of composite ac-tion between wythes increases the structural capacity of a given panel, making it more structurally efficient. However,
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
43PCI Journal | Spr ing 2011
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
Lee and Pessiki found that a three-wythe panel with stag-gered longitudinal solid concrete zones exhibits behavior similar to that of a fully composite panel.6 They also observed that transfer of the prestressing force induced cracks in the concrete wythes parallel to the prestress-ing strands. A finite-element analysis was conducted to investigate the prestressing forces during release. Results of the analysis showed that modeling the concrete and the insulation with solid block elements provided results close to the measured values.
Salmon et al.7 introduced the use of fiber-reinforced polymer (FRP) bars formed in a truss orientation in place of metal wire trusses. Test results showed that the use of FRP achieved a high level of composite action and provided thermal benefits similar to noncomposite precast, prestressed concrete sandwich wall panels. Following the same concept, a carbon-fiber-reinforced polymer (CFRP) shear connection grid was used in the construction of pre-cast, prestressed concrete sandwich wall panels in 2003.8
Because carbon fibers have a thermal conductivity that is about 14% that of steel, connecting concrete wythes with carbon grid allows a panel to develop composite structural action without thermal bridges. Therefore, the insulating value of the panel is maintained.1 The grid was oriented di-agonally between the concrete wythes, normal to the wall surface, allowing a truss mechanism to develop.
This paper describes the behavior of six full-scale precast, prestressed concrete sandwich wall panels. The panels were composed of two outer precast, prestressed concrete wythes and an internal layer of insulation with shear grid reinforcement placed through the core into each concrete wythe. The various parameters considered in the current study included the type of insulation, presence of solid
traditional composite shear connections have the nega-tive consequence of thermally bridging the two concrete wythes, thus decreasing the thermal efficiency.
Wall panels were first introduced during the 1960s as double-tee sandwich panels.1 Solid concrete zones were used between the double-tee and the inner concrete wythe to develop composite action. Double-tee sandwich panels provided a robust structural wall but sacrificed the poten-tial thermal savings. Flat concrete slabs were soon used in place of double-tees to reduce the thickness of the building envelope and to improve the aesthetics of a structure. As in double-tee sandwich panels, composite action between the wythes of flat slab sandwich wall panels was often achieved through solid concrete zones.
More recently, steel ties and wire trusses were introduced to replace solid concrete zones. Steel wythe connections improved the thermal performance of sandwich wall panels compared with solid concrete zones, but such ties still act as thermal bridges.1 Noncomposite panels were introduced in the 1980s and aimed at addressing the thermal deficien-cies created by steel ties. Noncomposite panels contain minimal shear connectors to substantially reduce the potential for thermal bridging but sacrifice the structural efficiency of a composite structure.
Despite their lower structural capacity, noncomposite panels became popular due to their thermal savings and architectural characteristics. The typical design method for precast, prestressed concrete sandwich wall panels often assumes noncomposite action.2 In practice, however, panels generally exhibit partially composite behavior. Test results by several researchers have shown that significant shear transfer occurs between the wythes.3–5
Figure 1. Commonly used shear transfer connectors include wire trusses, bent wires, and solid zones of concrete penetrating the insulation core. Note: CFRP = carbon-fiber-reinforced polymer.
Wire truss connector Bent wire connectors Solid concrete zone CFRP grid material sample CFRP grid shear transfer mechanism in section cut from a tested pane (insulation removed)
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
Spr ing 2011 | PCI Journal44
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
(Fig. 2). All panels were 8 in. (200 mm) thick and consist-ed of three layers. Table 1 summarizes the configurations of the tested panels.
Panels EPS1, EPS2, XPS1, XPS3, and XPS4 consisted of two 2-in.-thick (50 mm) concrete wythes with a 4-in.-thick (100 mm) insulation layer in between. This arrangement was designated as a 2-4-2 panel configuration. One wythe of a 2-4-2 panel included two 2-in.-thick (50 mm) and 24-in.-wide (610 mm) internal pilasters along the full height of each panel at 1/4 and 3/4 widths (Fig. 3). The two pilasters were provided to carry axial loads from the two corbels located at the top of the inner panel face. Lifting anchors on the inner face were centered on the pilasters in 2-4-2 panels, so these anchors did not bridge the concrete wythes. Two lifting anchors were also included on the top edge of each panel. These anchors spanned between concrete wythes. Panel XPS2 consisted of a 4-in.-thick (100 mm) concrete wythe, a 2-in. (50 mm) layer of insula-tion, and an outer 2-in. (50 mm) concrete wythe. Figure 4 shows this configuration, which was designated 4-2-2 with two corbels located at the top of the 4-in.-thick wythe. The 4-2-2 panel was designed to carry the axial load through its thicker concrete wythe and therefore did not have internal pilasters.
concrete zones, panel configuration, and shear grid rein-forcement ratio.
The loading sequence for each panel was selected to simu-late the effect of service gravity and wind loads for a 50-year lifespan. Load and support conditions were designed to mimic field conditions. Test results from the experimen-tal program were compared with theoretical predictions to evaluate the percent composite action achieved by each tested panel.
Experimental program
Six precast, prestressed concrete sandwich wall panels were designed and tested to evaluate their flexural response under combined vertical and lateral loads. The study included panels fabricated with two different insulation types: expanded polystyrene (EPS) insulation and extruded polystyrene (XPS) insulation. According to the manufac-turer, the selected EPS insulation had a nominal density of 1 lb/ft3 (16 kg/m3) and a nominal compressive strength of 13 psi (90 kPa). The selected XPS insulation had a nominal density of 1.8 lb/ft3 (29 kg/m3) and a nominal compressive strength of 25 psi (170 kPa).
The panels were 20 ft tall × 12 ft wide (6.1 m × 3.7 m)
Inner panel view during testing Outer panel view during testing
Figure 2. The panels were 20 ft tall × 12 ft wide (6.1 m × 3.7 m).
45PCI Journal | Spr ing 2011
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
CFRP shear grid was provided between the two concrete wythes to transfer the shear stresses across the insulation and to develop a composite action between the wythes. The CFRP grid was placed in strips running parallel to the longitudinal axis of each panel (Fig. 5). The CFRP strips were embedded 3/4 in. (19 mm) into each concrete wythe. All panels except XPS4 contained the same grid layout. Panel XPS4 contained an additional 30 ft (9.1 m) of shear grid. In addition to the CFRP grid, panel XPS1 contained
Each concrete wythe was reinforced with a sheet of welded-wire reinforcement in the plane of the wythe and prestressed in the longitudinal direction by five 270 ksi (1860 MPa), low-relaxation prestressing strands. The diameters of the prestressing strands in the 2-in.-thick (50 mm) and 4-in.-thick (100 mm) concrete wythes were 3/8 in. (10 mm) and 1/2 in. (13 mm), respectively. Figures 3 and 4 show the strands used and their initial tension levels.
Figure 3. The configuration and dimensions of 2-4-2 panels. Panel in photo cut to show cross section. Note: CFRP = carbon-fiber-reinforced polymer; WWR = welded-wire reinforcement.
Inner face
Outer face
Foam core
CFRP shear grid (Typical)
Panel width = 12 ft
2 in
.
2 ft 4 ft
16x10 W2.1xW3.0WWR Continuous (Typical)
2 ft 2 ft 2 ft
2 in
.
8 in
.Five 3/8 in. low relaxationstrands at 17,200 lbeach wythe (not shown)
Table 1. Summary of experimental tests and results
Panel identification
InsulationWythe
thicknesses, in.Solid zones
Shear grid layout
Concrete strength, psi
Failure load (1.2D + 0.5Lr +…)
Service load deflection D + Lr + W
EPS1 EPS 2-4-2 No Layout 1 7620 2.8W120 h/460
EPS2 EPS 2-4-2 No Layout 1 7670 1.8W150 (2.8W120) h/500
XPS1 XPS 2-4-2 Yes Layout 2 10,080 1.6W120 h/1480
XPS2 XPS 4-2-2 No Layout 1 8790 3.2W120 h/755
XPS3 XPS 2-4-2 No Layout 1 7670 0.7W120 n.a.
XPS4 XPS 2-4-2 No Layout 3 7340 1.8W120 h/700
Note: D = dead load; h = height = 240 in.; Lr = roof live load; n.a. = not applicable; W = wind load at a selected design wind speed; W120 = 6.96 kip; W150 = 10.56 kip. 1 in. = 25.4 mm; 1 kip = 4.448kN; 1 psi = 6.895 kPa.
Spr ing 2011 | PCI Journal46
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
10 discretely located solid concrete zones throughout the height and width of the panel (Fig. 5).
Test setup
All panels were tested in the laboratory using a steel testing frame that allowed for simultaneous application of gravity- and lateral loads (Fig. 2). Reverse-cyclic lateral loads were applied at a rate of 1 cycle per 10 sec (0.1 Hz) to simulate wind pressures. The testing frame consisted of one braced frame on each side of the panel to support an upper cross beam. This cross beam in turn provided the upper lateral support to the panel. The entire setup was anchored to the laboratory strong floor. A closed-loop hydraulic actuator, supported by a strong reaction wall, applied the lateral load.
Each panel was simply supported in the testing frames at the top and bottom edges. The bottom of the panel was supported by a hinge, which restrained horizontal and vertical movements while allowing for rotation. The center of the hinge was located 1 in. (25 mm) below the bottom of each panel. The top of each panel was supported using a sliding pin connection that restrained horizontal motion but
allowed for vertical movement and rotation. The center of the sliding pin was located 4 in. (100 mm) above the top of the panel.
Vertical loads were applied to the top of each corbel by a hydraulic jack and cable (Fig. 2) to simulate the effects of a 60-ft-span (18.3 m) double-tee roof system. The jack was connected to an accumulator to maintain a constant axial load as the panel deformed. Lateral loads were applied by the actuator connected to a spreader beam system, which was used to push and pull the panel to simulate wind pres-sure and suction. Two loading tubes were provided at each ¼-height of each panel, one on each wythe, to distribute the lateral load across the width of the panel. The lateral-loading mechanism included a vertical spreader beam that could shorten and elongate as the panel deformed to prevent the transfer of any unintended forces to the panel. Each panel was subjected to reverse-cyclic loading begin-ning at a level equivalent to 70% of the service load. The loading regime was selected using a Weibull distribution to simulate wind loads over a 50-year service life.9
All panels were instrumented to measure lateral deflection, relative displacement between the two concrete wythes,
Figure 4. The configuration in this drawing was designated as 4-2-2 with two corbels located at the top of the 4-in.-thick wythe. Panel cut to show cross section. Note: CFRP = carbon-fiber-reinforced polymer; WWR = welded-wire reinforcement. 1 in. = 25.4 mm; 1 ft = 0.305 m; 1 lb = 4.448 N.
Inner face
Outer face
CFRP shear grid (Typical)
Panel width = 12 ft
2 in
.
2 ft 4 ft
16x10 W2.1xW3.0WWR Continuous (Typical)
2 ft 2 ft 2 ft
4 in
.
8 in
.
Foam core
Five 3/8 in. low relaxationstrands at 16,500 lbouter wythe (not shown)
Five 1/2 in. low relaxationstrands at 31,000 lbinner wythe (not shown)
47PCI Journal | Spr ing 2011
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
surface strain of the concrete, and applied axial and lateral loads. The strain profile across the thickness of each panel was measured with four electrical-resistance strain gauges across the panel section at three locations along the height.
Each panel was subjected to 3710 fully reversed lateral-load cycles at 45% of the factored lateral wind load, equiv-alent to 0.7W, with a factored axial load of 1.2D + 0.5Lr in place, where W is wind load, D is dead load, and Lr is roof live load. The initial cycles were followed by 177 cycles at 50% of the factored lateral wind load (0.8W) with the fac-tored axial load applied. Subsequent individual cycles were applied at 60%, 80%, and 100% of the factored lateral wind load (1.0W, 1.3W, 1.6W), all with the factored axial load applied. After completion of all lateral cycle loads in the presence of gravity load, the lateral load was increased in one direction only until failure.
Results and discussion
Lateral displacements
Figure 6 depicts the measured lateral displacement at midheight for the different panels. In general, the measured lateral deflections due to the applied axial load only were found to depend on the thickness of the panel configuration (2-4-2 or 4-2-2) and also on the type and configuration of shear transfer mechanism used. Lateral deflection due to axial load alone was shown as the offset deflection at zero lateral-load level.
The allowable displacement of h/360 (where h is the height of the panel) at service load level (as per the American Concrete Institute’s ACI 533R-93, Guide for Precast Concrete Wall Panels10) was compared with the measured values for the tested panels. The two EPS-insulation panels EPS1 and EPS2 behaved almost identically throughout the loading cycles. The panels’ stiffnesses remained constant to a lateral load of 15 kip (67 kN), or 2.2W120 (where W120 is the wind load at a wind speed of 120 mph), beyond which concrete cracking occurred, considerably reducing the stiffness (Fig. 6). Similar behavior was observed for XPS2. The behavior of panels XPS1 and XPS4 remained linear to a lateral load level of about 10 kip (44.5kN), or 1.4W120. Panel XPS3 failed prematurely at a lateral load level of 5 kip (22 kN).
The maximum measured displacements at service-load lev-el for EPS1 and EPS2 were equivalent to h/460 and h/500, respectively. These service-level displacements were well within the ACI 533R limit. Among the XPS-insulation panels, XPS1 experienced the least stiffness degradation with increased load cycles because of the presence of solid concrete zones connecting the inner and outer wythes. The maximum lateral displacement at service-load level for XPS1 was equivalent to h/1480.
Panels XPS2 and XPS4 exhibited minimal lateral-load degradation. The measured lateral displacements for both panels did not increase noticeably throughout the fatigue cycles. The maximum lateral displacements at service-load level for XPS2 and XPS4 were equivalent to h/755
Figure 5. Layout of CFRP shear grids. Note: CFRP = carbon-fiber-reinforced polymer. 1 in. = 25.4 mm; 1 ft = 0.305 m.
CL
CL
CLCL
CLCL
CLCL
CL
CL
CL
CL
CL
CLCL
Layout 1Panels EPS1, EPS2, XPS1, XPS2
Layout 2Panel XPS3
Layout 3Panel XPS4
Spr ing 2011 | PCI Journal48
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
Figure 6. Measured lateral displacement at mid-height for the different panels.
-25,000
-15,000
-5,000
5,000
15,000
25,000
-0.5 0 0.5 1 1.5 2 2.5Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
-111
-67
-22
22
67
111-13 0 13 25 38 51 64
Lateral displacement at mid-height, mmLa
tera
l loa
d, k
N
-25,000
-15,000
-5,000
5,000
15,000
25,000
-0.5 0 0.5 1 1.5 2 2.5Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
-111
-67
-22
22
67
111-13 0 13 25 38 51 64
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
-25,000
-15,000
-5,000
5,000
15,000
25,000
-0.5 0.0 0.5 1.0 1.5 2.0 2.5Lateral displacement at mid-height, in.Load-displacement behavior of EPS1 Load-displacement behavior of EPS2
Load-displacement behavior of XPS1 Load-displacement behavior of XPS2
Load-displacement behavior of XPS3 Load-displacement behavior of XPS4
Late
ral l
oad,
lb
-111
-67
-22
22
67
111-13 0 13 25 38 51 64
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
-25,000
-15,000
-5,000
5,000
15,000
25,000
-0.5 0.0 0.5 1.0 1.5 2.0 2.5Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
-111
-67
-22
22
67
111-13 0 13 25 38 51 64
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
-25,000
-15,000
-5,000
5,000
15,000
25,000
-0.5 0.0 0.5 1.0 1.5 2.0 2.5Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
-111
-67
-22
22
67
111-13 0 13 25 38 51 64
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
-25,000
-15,000
-5,000
5,000
15,000
25,000
-0.5 0.0 0.5 1.0 1.5 2.0 2.5Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
-111
-67
-22
22
67
111-13 0 13 25 38 51 64
Lateral displacement at mid-height, mmLa
tera
l loa
d, k
N
49PCI Journal | Spr ing 2011
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
XPS4 showed a clear reduction in composite action with increasing lateral load. A significant discontinuity in the strain at ultimate load was observed for these panels, indi-cating a partial composite behavior at ultimate.
Failure modes
The observed failure modes for EPS1, XPS1, and XPS2 were localized at the tops of the panels in the corbel zones. Failure was characterized by a shear failure around the cor-bels extending down about 2 ft (600 mm). Figure 8 shows an example of these failures, which were accompanied by separation of the top of the panel. Panel EPS2 exhibited a flexural-shear failure across the width of the panel at about 7/8 the panel’s height. Panels XPS3 and XPS4 exhibited a flexural-shear failure across the width of the panel at about 7/8 the panel’s height along with a simultaneous top-of-panel separation.
All panels with sufficient shear transfer mechanisms exhibited deflections well below the limiting value speci-fied by ACI 533R and sustained loads prior to failure in excess of their factored design loads (Table 1). However, panel XPS3, which had a gap between the outer wythe and foam, failed prematurely prior to the service load with high deflections.
Uniform design pressures for panels EPS1, XPS1, XPS2, XPS3, and XPS4 were assumed to be 29 lb/ft2 (1.4 kPa), corresponding to a design wind speed of 120 mph (193 km/hr). Although designed for 29 lb/ft2 (1.4 kPa), EPS2 was tested for an equivalent pressure of 44 lb/ft2 (2.1 kPa), corresponding to a design wind speed of 150 mph (241 km/hr). Thus, the lateral fatigue loading on EPS2 was
and h/700, respectively. Failure of panel XPS3 occurred before reaching the design service-load level. Test results suggested that the accumulated degradation for XPS3 was substantial compared with other panels. A gap between the outer concrete wythe and the foam, observed before testing, likely contributed to the premature failure. This gap measured about 1/4 in. (6 mm) and was visible along the majority of both 20 ft (6.1 m) panel sides. The gap was identified by the precast concrete producer after stripping the forms, but because the panel was intended for testing, it was decided not to reject the piece.
Strain profiles
The strain profiles across the thickness of the panels were measured to determine the degree of composite action between the two concrete wythes. Figure 7 shows typical results recorded during the testing of the panels at ultimate load. For all panels, the inner wythe experienced com-pressive strains while the outer wythe experienced tensile strains under the effect of applied factored gravity loads. Test results showed that EPS-insulation panels EPS1 and EPS2 as well as XPS1 with solid concrete zones exhib-ited and maintained a high level of composite action until failure.
The strain profile at ultimate load indicates that the neutral axes of these panels were located closer to the elastic cen-troid of the composite cross section rather than the elastic centroid of each individual wythe. The measured strains for XPS2 (4-2-2 configuration) indicated that each wythe acted independently in carrying the applied loads and the neutral axis was located within the thickness of each wythe. This behavior indicated noncomposite action. Panels XPS3 and
Figure 7. Strain profile distribution for different panels at ultimate loads. Note: The negative (-) sign indicates compression. Note: 1 kip = 4.448 kN.
-178
-40
94
19
Outer wythe
Inner wythe
EPS1, EPS2, XPS1, XPS3, XPS4
-230
-12
113
8
-142
-6
38
36
-160
10
15
-70
-200
5
40
-72
Outer wythe
Inner wythe
XPS2
-322
148
85
-109
EPS1 EPS2 XPS1 XPS3 XPS4
XPS2
19.4 kip Ultimate lateral load =
20 kip 11.2 kip 5.0 kip 12.6 kip
22.5 kip Ultimate lateral load =
Spr ing 2011 | PCI Journal50
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
Mcr = cracking moment
Ma = acting moment
Ig = gross moment of inertia of the section
Icr = cracked transformed moment of inertia of the section
Recent research findings by Bischoff and Scanlon have demonstrated that Eq. (2) is only applicable for flexural members with an Ig/Icr ratio less than 3, which corresponds to beams or slabs with a steel reinforcing ratio greater than 1%.12 It has been demonstrated that applying Eq. (2) to cross sections with an Ig/Icr ratio greater than 3 consider-ably overestimates the member stiffness.11 Equation (3) is an alternative formulation of the effective moment of inertia that was proposed by Bischoff and Scanlon.11
1
I
M
M
I
I
II
1
ef f
a
cr
g
cr
cr
g2#=
- -f p > H (3)
The ratios of Ig/Icr for the inner and outer wythes of the tested panels ranged from 20 to 183, which is significantly higher than the limiting value of 3 proposed by Bischoff and Scanlon.11 Figure 9 shows comparisons between the measured and predicted displacements for different panels using both approaches for the effective moment of inertia.
For the theoretical fully composite case, applied loads and moments were assumed to act on the fully composite sec-tion. For the theoretical noncomposite case, it was assumed that the applied axial load was resisted by the inner wythe (with corbels) alone. Applied moments were assumed to be distributed to the inner and outer wythes depending on
higher than the load used for the other panels.
Analysis of sandwich wall panels
To evaluate the degree of composite action for the panels, the measured lateral displacements were compared with the predicted values assuming both fully composite and fully noncomposite behaviors. The percentage of composite ac-tion k was evaluated for all tested panels using Eq. (1).
k 100exp
noncomposite
noncomposite erimental
compositeD D
D D=
-
-a k (1)
where
∆experimental = measured displacement at a selected load level
∆composite = corresponding theoretical displacement as-suming fully composite behavior
∆noncomposite = corresponding theoretical displacement as-suming fully noncomposite behavior
To determine theoretical panel displacements beyond cracking, the effective moment of inertia Ieff was calcu-lated by Eq. (2) in accordance with ACI’s Building Code Requirements for Structural Concrete (ACI 318-05) and Commentary (ACI 318R-05).11
1IM
M
M
MI I I
ef f
a
cr
a
cr
g cr g
33
#= + -f fp pR
T
SSSS
V
X
WWWW
(2)
where
Figure 8. Typical observed failure modes for EPS1, XPS1, and XPS2.
Corbel-zone shear failure on inner wythe Outer wythe cracking following top-of-panel separation at failure
51PCI Journal | Spr ing 2011
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
for XPS2 was significantly lower than for other panels, the maximum displacement at service load level was still within the ACI 533R limitations. Panel XPS3 failed prior to reaching the combined axial and lateral service-load condition.
Conclusion
The flexural behaviors of six full-scale insulated precast, prestressed concrete sandwich wall panels were investi-gated. The panels were subjected to monotonic axial and reverse-cyclic lateral loading to simulate gravity and wind pressure loads, respectively. Based on the findings of this
their individual stiffnesses.
The distribution ratio was calculated using the average of their gross and cracked moments of inertia. The percentage of composite action was calculated for each panel under the combined axial and lateral service load, and Table 2 summarizes the results. In all calculations, a 20 ft (6.1 m) nominal span was assumed.
The estimated composite action for EPS1, EPS2, XPS1, and XPS4 was nearly 100%. Panel XPS2 exhibited 18% composite action under the combined axial and lateral service load. Although the percent of composite action
Figure 9. Load-lateral displacement behavior for the panels.
0
5,000
10,000
15,000
20,000
25,000
0 2 4 6 8 10
Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
0
22
44
67
89
1110 51 102 152 203 254
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
Experimental EPS1
Experimental EPS2
Composite behavior
Non-composite behavior using Bischoff's formula
Non-composite behavior using ACI's formula
Service load level
0
5,000
10,000
15,000
20,000
25,000
0 2 4 6 8 10
Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
0
22
44
67
89
1110 51 102 152 203 254
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
Experimental
Composite behavior
Non-composite behavior using Bischoff's formula
Non-composite behavior using ACI's formula
Service load level
0
5,000
10,000
15,000
20,000
25,000
0 2 4 6 8 10
Lateral displacement at mid-height, in
Late
ral l
oad,
lb
0
22
44
67
89
1110 51 102 152 203 254
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
Experimental Composite behaviorNon-composite behavior using Bischoff's formulaNon-composite behavior using ACI's formula
Service load level
0
5,000
10,000
15,000
20,000
25,000
0 2 4 6 8 10
Lateral displacement at mid-height, in.
Late
ral l
oad,
lb
0
22
44
67
89
1110 51 102 152 203 254
Lateral displacement at mid-height, mm
Late
ral l
oad,
kN
Experimental XPS3
Experimental XPS4
Composite behavior
Non-composite behavior using Bischoff's formula
Non-composite behavior using ACI's formula
Service load level
Panels EPS1 and EPS2 Panel XPS1
Panel XPS2 Panels XPS3 and XPS4
Spr ing 2011 | PCI Journal52
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
study, several conclusions were made:
• Panel stiffness and deflections are significantly af-fected by the type and configuration of the shear transfer mechanism. Panel stiffness is also affected by the type of foam.
• Values of percent composite action near 100% can be achieved with CFRP grid shear connections or with solid concrete zones.
• Appropriate use of CFRP shear grid can provide an effective shear transfer mechanism in precast, pre-stressed concrete sandwich wall panels, as evidenced by the behavior of panels EPS1, EPS2, XPS2, and XPS4. All panels sustained loads in excess of their factored design loads and exhibited large deformations before failure. CFRP grid can provide the required composite action between wythes using either EPS or XPS foam.
• Appropriate selection of the CFRP shear-grid quantity and configuration is critical to achieve optimal struc-tural performance of a panel. Proper quality control in production is especially important for composite wall panels.
• For a given shear transfer mechanism, a higher percent composite action can be achieved using EPS insula-tion rather than XPS insulation. Use of XPS insulation requires an increase of the shear reinforcement ratio compared with EPS insulation.
Acknowledgments
The authors are grateful for the support of AltusGroup and the assistance provided by Harry Gleich of Metromont Corp. and Steve Brock of Gate Precast.
Table 2. Percentage of composite action for different panels at service load level
Panel identification
Experimental displacement, in.
Composite displacement, in.
Noncomposite displacement, in. k, % k, %
Bischoff formula for Ieff
ACI 318 formula for Ieff
Bischoff formula for Ieff
ACI 318 formula for Ieff
EPS1 0.21 0.200 39.6 6.70 100.0 99.8
EPS2 0.24 0.200 39.6 6.70 99.9 99.4
XPS1 0.14 0.087 28.8 2.20 99.8 97.6
XPS2 0.46 0.089 0.5 0.50 17.7 17.7
XPS3 n.a.* 0.087 n.a. n.a. n.a. n.a.
XPS4 0.31 0.087 20.1 3.40 98.9 93.5
*Specimen XPS3 failed at a lateral load level less than the design service load.
Note: Composite displacements differ between EPS and XPS 2-4-2 panels due to different values for concrete elastic modulus. Ieff = effective moment of inertia; k = percentage of composite action; n.a. = not applicable. 1 in. = 25.4 mm.
53PCI Journal | Spr ing 2011
—C—C—C—C—
–H –H –H
–H –H –H
–H
–H
References
• Gleich, H. 2007. New Carbon Fiber Reinforcement Advances Sandwich Wall Panels. Structure Magazine (April): pp. 61–63.
• PCI Committee on Precast Sandwich Wall Panels. 1997. State-of-the-Art of Precast/Prestressed Sand-wich Wall Panels. PCI Journal, V. 42, No. 2 (March–April): pp. 92–134.
• Pessiki, S., and A. Mlynarczyk. 2003. Experimental Evaluation of Composite Behavior of Precast Con-crete Sandwich Wall Panels. PCI Journal, V. 48, No. 2 (March–April): pp. 54–71.
• Bush, T. D., and G. L. Stine. 1994. Flexural Behavior of Composite Prestressed Sandwich Panels. PCI Jour-nal, V. 39, No. 2 (March–April): pp. 112–121.
• Lee, B., and S. Pessiki. 2007. Design and Analysis of Precast, Prestressed Concrete Three-Wythe Sandwich Wall Panels. PCI Journal, V. 52, No. 4 (July–August): pp. 70–83.
• Lee, B., and S. Pessiki. 2008. Experimental Evaluation of Precast, Prestressed Concrete, Three-Wythe Sand-wich Wall Panels. PCI Journal, V. 53, No. 2 (March–April): pp. 95–115.
• Salmon, D. C., A. Einea, M. K. Tadros, and T. D. Culp. 1997. Full Scale Testing of Precast Concrete Sandwich Panels. ACI Structural Journal, V. 94, No. 4 (July–August): pp. 354–362.
• Frankl, B. 2008. Structural Behavior of Insulated Pre-cast Prestressed Concrete Sandwich Panels Reinforced with CFRP Grid. M.Sc. thesis. Department of Civil, Construction and Environmental Engineering, North Carolina State University, Raleigh, NC.
• Xu, Y. L. 1995. Determination of Wind-Induced Fa-tigue Loading on Roof Cladding. Journal of Engineer-ing Mechanics, ASCE, V. 121, No. 9 (September): pp. 956–963.
• American Concrete Institute (ACI) Committee 533R. 2004. Guide for Precast Concrete Wall Panels. ACI 533R-93. Farmington Hills, MI: ACI.
• ACI Committee 318. 2005. Building Code Require-ments for Structural Concrete (ACI 318-05) and Commentary (ACI 318R-05). Farmington Hills, MI: American Concrete Institute (ACI).
• Bischoff, P. H., and A. Scanlon. 2007. Effective Mo-ment of Inertia for Calculating Deflections of Concrete
Members Containing Steel Reinforcement and Fiber-Reinforced Polymer Reinforcement. ACI Structural Journal, V. 104, No. 1 (January): pp. 68–75.
Spr ing 2011 | PCI Journal54
—C—C—C—C—–H –H –H
–H –H –H
–H
–H
Notation
D = dead load
h = height of the panel
Icr = cracked transformed moment of inertia of the section
Ieff = effective moment of inertia
Ig = gross moment of inertia of the section
k = percentage of composite action
Lr = roof live load
Ma = acting moment
Mcr = cracking moment
W = wind load
W120 = wind load at a wind speed of 120 mph
W150 = wind load at a wind speed of 150 mph
∆composite = corresponding theoretical displacement as-suming fully composite behavior
∆experimental = measured displacement at a selected load level
∆noncomposite = corresponding theoretical displacement as-suming fully noncomposite behavior
About the authors
Bernard A. Frankl graduated with his MS in civil engineering from North Carolina State University in Raleigh, N.C., and now works for the South Dakota Department of Transportation.
Gregory W. Lucier is the manager of the Constructed Facilities Laboratory at North Carolina State University.
Tarek K. Hassan, PhD, is an associate professor for the Structural Engineering Depart-ment at the Faculty of Engineer-ing at Ain Shams University, and a senior structural engineer at Dar Al Handasa, Cairo, Egypt.
Sami H. Rizkalla, PhD, P.Eng., is a Distinguished Professor of Civil, Construction and Environmental Engineering, director of the Constructed Facilities Laboratory, and director of the National Science Foundation Industry/
University Cooperative Research Center at North Carolina State University.
Synopsis
This paper describes the structural behavior of precast, prestressed concrete sandwich wall panels reinforced with carbon-fiber-reinforced polymer (CFRP) shear grid to achieve composite action. Use of CFRP as a shear transfer mechanism was intended to increase the thermal insulation efficiency, enhance the service
life, and increase the overall structural capacities of the panels.
This study included testing of six full-scale sandwich wall panels, each measuring 20 ft × 12 ft (6.1 m × 3.7 m). The panels consisted of two outer prestressed con-crete wythes and an inner insulation wythe. The study included two types of insulation and several shear transfer mechanisms with different CFRP reinforce-ment ratios to examine the degree of composite action developed between the two concrete wythes.
All panels were simultaneously subjected to applied gravity and lateral loads. Reverse-cyclic lateral loads simulated the effects of wind pressure and suction. All panels were subjected to approximately 4000 cycles of lateral loading with the presence of factored gravity load. Following each fatigue regime, the lateral loads were increased until failure was achieved. Test results of the experimental program were compared with theo-retical predictions of fully composite and noncompos-ite actions to evaluate the percent composite action and to assess the optimum panel configuration.
Keywords
Carbon-fiber-reinforced polymer, CFRP, composite, insulated wall panel, sandwich wall panel, shear grid.
Review policy
This paper was reviewed in accordance with the Precast/Prestressed Concrete Institute’s peer-review process.
Reader comments
Please address any reader comments to [email protected] or Precast/Prestressed Concrete Institute, c/o PCI Journal, 200 W. Adams St., Suite 2100, Chicago, IL 60606. J