Construction and Building Materials 98 (2015) 722–734

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Durability and long term behavior of FRP/foam shear transfermechanism for concrete sandwich panels

http://dx.doi.org/10.1016/j.conbuildmat.2015.08.1050950-0618/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: sami_rizkalla@ncsu.edu (S.H. Rizkalla).

Hamid Kazem a, William G. Bunn b, Hatem M. Seliem c, Sami H. Rizkalla a,⇑, Harry Gleich b

aNorth Carolina State University, USAbMetromont Corporation, USAcHelwan University, Egypt

h i g h l i g h t s

� Precast concrete sandwich panels can provide the shear transfer mechanism.� Quality of fabrication can affect the shear strength of the panels.� Design strength should be 30% of the sustained loads ultimate shear capacity.� The EPS panels are more susceptible to aging effect compared to the XPS panels.

a r t i c l e i n f o

Article history:Received 11 February 2015Received in revised form 5 August 2015Accepted 10 August 2015

Keywords:Wall panelsCFRP gridGFRP gridShear strengthSustained loadingOutdoor exposure

a b s t r a c t

This paper presents an experimental program to evaluate the effect of sustained loading and outdoorexposure on the shear strength of precast concrete sandwich panels connected with FRP grid/rigid foaminsulation as shear transfer mechanism. CFRP and GFRP grids were considered in this study along withEPS and XPS foam. The experimental program is comprised of three different studies with a total of 26test panels using different combinations of FRP grid and foam insulation.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Precast concrete sandwich panels have been widely used forbuilding envelopes of wide range of structures. Precast concretesandwich panels have been produced in North America for morethan 50 years [1]. The panels generally consist of two discrete con-crete wythes separated by a rigid foam layer. Load-bearing con-crete sandwich panels are typically subjected also to sustainedgravity roof loads in addition to lateral wind pressure loads. Theprecast concrete panels can be designed to resist the applied loadsas full-composite, partial-composite, or non-composite action. Thedegree of composite action depends on the ability of the FRP grid totransfer the shear forces between the two concrete wythes.

Early panels were designed as non-composite, where the innerthick wythe was designed to carry the load and the outer wythe

was non-structural to protect the insulation layer between theconcrete wythes [1]. Use of non-composite panels showed goodthermal efficiency, while sacrificing the structural performance.Later, solid concrete zones, metal trusses and steel ties were usedto achieve the composite action between the two concrete wythes.However, both types of shear connectors introduce thermal bridgeand result in reduction of the R-values. Recently the use of Fiber-Reinforced Polymer (FRP) grid has been introduced as a shear con-nector due to their structural appealing benefits in addition to theirsuperior thermal performance [2]. This approach was adopted inresponse to the growing demand for energy-efficient buildingsand recognition of Leadership in Energy and Environmental Design(LEED) certification. Several studies have been reported on thebehavior of short-term strength of precast concrete sandwich pan-els utilizing FRP grid/rigid foam as a shear transfer mechanism.

Frankl et al. [3] investigated the flexural behavior of six full-scaleinsulated precast/prestressed concrete sandwich wall panels. Thepanels were subjected to monotonic axial load and reverse-cycle

Front Elevation Cross - Sectional Elevation

Fig. 1. Typical sectional-elevation of test panels.

H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734 723

lateral load, simulating gravity and wind pressure loads, respec-tively. The study concluded that a high degree of composite actioncan be achieved by using Carbon FRP (CFRP) grid as shear connec-tors. Further, the lateral stiffness of the test panels were greatlyaffected by the type and configuration of the shear connectors usedas well as the type of foam.

Kim et al. [4] used CFRP grid as a shear transfer mechanism toevaluate effect of different parameters on shear flow capacity ofconcrete sandwich panels. The parameters included grid embed-ment length, insulation type and thickness, shear grid density(spacing), and the effect of repeated loading. Results indicated thatthe shear grid density, insulation type, and thickness have majoreffects on the shear strength of the system.

Bunn [5] tested 66 three-wythe precast concrete panels withCFRP grid/rigid foam insulation as shear connectors. The studyinvestigated different parameters including, type and thickness offoam insulation, spacing between CFRP gridlines, and discontinuityof the CFRP grid. Test results showed that panels utilizingExpanded PolyStyrene (EPS) foam have higher shear strength com-pared to those using Extruded PolyStyrene (XPS) foam using thesame CFRP grids. Bond between the rigid foam and the concretewas confirmed and evaluated by testing panel specimens withplastic sheets between the foam and the concrete to break thebond. Furthermore, the performance of CFRP grid only was investi-gated by removing the foam insulation. Test panels without FRPgrid confirmed that the bond between foam insulation and con-crete increased the shear strength of the panels. The research indi-cated also that increasing the thickness of EPS insulation foamdecreased the shear strength of the panels.

Sopal [6] and Hodicky et al. [7] conducted a comprehensiveresearch program to study CFRP grid/rigid foam insulation as ashear transfer mechanism for precast concrete sandwich wall pan-els. Research findings concluded that increasing the spacingbetween the CFRP grids resulted in increase of the overall shearstrength due to the increase of the concrete/foam interface area.However, increasing the thickness of the insulating foam resultedin decreasing of the shear strength of the panels. Based on theresearch findings, an equation was proposed to estimate the shearstrength of precast concrete sandwich wall panels utilizing CFRPgrid and rigid foam insulation as shear transfer mechanism. Sopal[6] also performed tests on panel specimens subjected to a freeze–thaw aging regimen. Test results indicated that the freeze–thawcycle tests, specified by the Acceptance Criteria of InternationalCode Council Evaluation Service (ICC-ES AC422) [8] had no notice-able effect on the shear strength of the test panels.

Soriano [9] tested three-wythe precast concrete panels withGFRP grid/rigid foam. Test results showed that EPS foam providedhigher shear strength in comparison to XPS foam without surfacetreatment. However, it was found that sandblasting of the XPSincreased the shear strength of the panels due to enhancementof the bond between the foam and the concrete.

Naito et al. [10] examined the shear-deformation relationship of14 different types of shear connectors, including CFRP and GFRPgrid. Test results indicated that the shear performance of ties var-ied considerably and the use of EPS foam increased the shearstrength in comparison to XPS foam.

Insel [11] tested sandwich wall panels utilizing CFRP grid sub-jected to monotonic and cyclic loadings to investigate fatigue per-formance. Monotonic load and various levels of cyclic load wereapplied to small- and medium-scale wall panels. Results indicatedthat for cyclic loads below the 30% of the ultimate monotonicloads, number of cycles to failure increases considerably. Further-more, results showed a reasonable agreement between analyticalmodel and experimental results.

Despite the extensive work performed to evaluate the behaviorand strength of precast concrete sandwich panels with FRP grid as

shear connectors, very limited research has been reported relatedto the durability and long-term behavior of these types of panels.This paper presents an experimental program undertaken to eval-uate the durability of precast concrete sandwich panels connectedby FRP grid/foam insulation as a shear transfer mechanism. Thestudy investigated the effect of sustained load and outdoor expo-sure conditions on the shear strength of the panels.

2. Experimental program

The experimental program reported in this paper comprises three studies. Theobjective of the first study was to determine the short-term ultimate shear strengthof the panels. The second study investigated the effect of sustained loads on thebehavior of FRP gird/rigid foam as a shear transfer mechanism. The third study eval-uated the effect of outdoor exposure conditions on the ultimate strength of thepanels.

The three studies included a total of 26 precast concrete sandwich panel seg-ments (referred to as panels hereafter) constructed using different FRP grid and dif-ferent rigid foam types. CFRP and GFRP grids were used along with the twocommercially-available foam types, EPS and XPS. EPS and XPS foams are polystyr-ene (transparent thermoplastic) insulations, which include polyurethane, polyiso-cyanurate, and phenolic [8]. Two different densities of the EPS foam were used,namely Normal Density (ND) and High Density (HD). Some panels were constructedwith Sand-Blasted (SB) XPS foam, while other panels had XPS foam without surfacetreatment.

All test panels consisted of three-wythe concrete panels separated by two layersof rigid foam insulation and bridged by FRP grid. The panels were 44 in. (1118 mm)tall and 20 in. (508 mm) wide as shown in Fig. 1. The FRP grids were cut at a 45-degree angle of an orthogonal grid and were placed in truss orientation. Segmentsof three-wythe panels were used instead of two-wythe panels to allow testing theFRP grid/rigid foam mechanism in direct shear and to minimize the bending effects.

The panels reinforced with CFRP grid had a total thickness of 12 in. (305 mm),two outer concrete wythes of 2 in. (51 mm) each, inner concrete wythe of 4 in.(102 mm), and two foam layers of 2 in. (51 mm) each as shown in Fig. 2. The panelswith GFRP grid had a total thickness of 16 in. (406 mm), two outer concrete wythesof 3 in. (76 mm) each, inner concrete wythe of 6 in. (152 mm), and two foam layersof 2 in. (51 mm) each as shown in Fig. 2. The width of the CFRP and GFRP grids usedin this study were 3.5 in. (89 mm) and 4.0 in. (102 mm), respectively as shown inFig. 3. The selected widths are used to simulate the typical dimensions used forthe sandwich panels connected by CFRP and GFRP constructed in USA and SouthKorea, respectively. The selected widths also provided an embedment length intoeach concrete wythe of ¾ in. (19 mm) and 1.0 in. (25 mm), for the CFRP and GFRPgrids respectively, indicating superior bond characteristics of CFRP grids. In addi-tion, Fig. 3 shows the spacing between the strands of each grid.

Panels with GFRP grid Panels with CFRP grid

Fig. 2. Cross-section of test panels.

Fig. 3. FRP grids cut at 45-degree, (a) CFRP, (b) GFRP.

Table 2Test matrix of the sustained load effect.

Panel designation Foam FRP grid Applied sustained load

% kip (kN) Duration

E(ND).C.0 EPS(ND) CFRP 0 0 12 monthsE(ND).C.20 20 6.0 (26.7)E(ND).C.35 35 10.5 (46.7)E(ND).C.45 45 14.0 (62.3)X.C.0 XPS 0 0X.C.20 20 3.5 (15.6)X.C.35 35 6.0 (26.7)X.C.45 45 8.0 (35.6)

E(HD).G.0 EPS(HD) GFRP 0 0 6 monthsE(HD).G.20 20 11.1 (49.4)E(HD).G.30 30 16.4 (73.0)E(HD).G.50 50 27.0 (120.1)X(SB).G.0 XPS(SB) 0 0X(SB).G.20 20 10.5 (46.7)X(SB).G.30 30 15.5 (68.9)X(SB).G.50 50 25.5 (113.4)

724 H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734

2.1. Short-term strength

A total of six panels were tested at age of 28 days in direct shear up to failure todetermine the short-term strength of precast concrete sandwich panels with FRPgrid/rigid foam insulation to establish a reference for the long-term strength. Themeasured ultimate short-term strength was used to determine the sustained loadlevels to be used to study the long-term behavior.

Four panels utilizing CFRP grid were included in the study, where two duplicatepanels used EPS(ND) foam and the other two duplicate panels used XPS foam with-out surface treatment. The purpose of testing duplicate panels was to establish con-fident in the test results. The two panels connected with GFRP grid, used EPS(HD)foam for the first panel and Sand-Blasted (SB) XPS foam for the second panel. Thetest matrix of the first study is given in Table 1. Use of different types of foam fol-lowed typical foams used in practice in USA and South Korea for CFRP and GFRPsandwich panels, respectively.

Table 1Test matrix of the short-term strength.

Panel designation Foam FRP grid

E(ND).C.a EPS(ND) CFRPE(ND).C.bX.C.a XPSX.C.b

E(HD).G EPS(HD) GFRPX(SB).G XPS(SB)

2.2. Effect of sustained load

A total of 16 concrete panels were subjected to selected sustained load levelsbefore testing in direct shear up to failure. Eight panels utilized CFRP grid and eightpanels utilized GFRP grid. Throughout the application of the sustained load, the rel-ative vertical displacement between the inner concrete wythe and the two outerwythes was measured using digital calipers with an accuracy of 0.0008 in.(0.02 mm) and recorded daily for the first four weeks and then bi-weekly for theremaining duration of the test. All the panels were kept in controlled environmentwith a temperature of 20 �C ± 2 (68 �F ± 4) and relative humidity of 70% ± 5.

Six panels with CFRP grid were subjected to a sustained load for 12 monthsbefore testing in direct shear up to failure. The three selected sustained load levelswere 20%, 35%, and 45% of the measured ultimate short-term strength determinedfrom the first study. The remaining two panels with CFRP grid were not subjected tosustained load, except their own weight until they were tested in direct shear up tofailure after 12 months. The two foam types, EPS(ND) and XPS without surfacetreatment were used in the panels with CFRP grid as given in Table 2.

Six panels with GFRP grid were subjected to a sustained load for six monthsbefore testing in direct shear up to failure. Since, glass fibers are more susceptibleto creep in comparison to carbon fibers, the duration of sustained load used forGFRP panels was six months while CFRP panels were tested for 12 months to collectsufficient data. The three selected sustained loads were 20%, 30%, and 50% of themeasured ultimate short-term strength determined from the first study. In addi-

Table 3Test matrix of the exposure condition effect.

Panel designation Foam FRP grid

E(HD).G.EXP.A EPS(HD) GFRPE(HD).G.EXP.BX(SB).G.EXP. A XPS(SB)X(SB).G.EXP. B

Fig. 4. Outdoor exposure of test panels.

Table 4Material properties of FRP grid.

FRP % fiber Cross-sectional area in.2 (mm2) Tensile strength ksi (N/mm2) Elongation at break Elastic modulus ksi (N/mm2) Grid spacing in. (mm)

Glass 55 0.032 (20.70) 48 (330) 1.78% 2700 (18,600) 2.0 � 2.0 (53 � 53)Carbon 55 0.002860 (1.845) 290 (2000) 0.76% 34,000 (234,400) 1.8 � 1.6 (46 � 41)

Table 5Material properties of rigid foam.

Foam Densitypcf (N/m3)

Compressive strength (10%deformation), psi (kN/m2)

Elastic moduluspsi (N/mm2)

EPS (ND) 1.00 157 12 (83) 200 (1.38)EPS (HD) 1.80 283 29 (200) 480 (3.31)

XPS 1.55 243 25 (172) 675 (4.65)

H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734 725

tion, two panels with GFRP grid were not subjected to externally applied load untiltested in direct shear up to failure after six months. The test matrix of the sustainedload study is given in Table 2.

2.3. Effect of outdoor exposure

Since, the GFRP panels perceived to be more susceptible to environmental con-ditions [12], four panels with GFRP grid, EPS(HD) and XPS(SB), given in Table 3,were subjected to outdoor weathering conditions. The panels were exposed forseven months, as shown in Fig. 4, before they were tested in direct shear up to fail-

Fig. 5. Test setup of d

ure. The exposure period extended between the months of April and October. Theclimate in Raleigh, North Carolina, like much of the southeastern United States, isa humid subtropical climate. During this period of the year the average temperatureranges between a minimum of 10 �C (50 �F) and a maximum of 33 �C (90 �F), andthe average relative humidity ranges between 40% and 95%. April is the driestmonth, while summer is hot and humid, followed by the rainy fall season. There-fore, it shall be noted that, the selected condition is not a representative of coldand polar zones. After testing those four panels up to failure, the obtained ultimateshear strength was compared to those were not subjected to sustained loads in thesustained load program.

3. Material properties

The material properties of the CFRP and GFRP grids used in thisstudy are summarized in Table 4. The two FRP grids are manufac-tured and provided by two different producers. The material char-acteristics of the different foam types are given in Table 5. Materialproperties of the FRP grids and rigid insulations are provided bymanufacturers.

irect shear test.

(a) (b)

Fig. 6. Test setup for sustained loading.

Fig. 8. Shear flow vs. relative vertical displacement of specimens with GFRP gridsunder short-term loading.

Table 6Short-term shear flow strength of panels with CFRP and GFRP grid.

Panel designation Concretestrength

Maximumshear flow

Relativeverticaldisplacement

psi N/mm2 lb/in. N/mm in. mm

E(ND).C.a 5800 40.0 368.4 64.5 0.228 5.75

726 H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734

The measured concrete compressive strength at day of testingwas determined by testing 4 � 8 in. (100 � 200 mm) cylindersaccording to ASTM C 39 [13]. After 24 h from casting, the concretecylinders were removed from the molds and were kept next to thecorresponding panels to be subjected to the same curingconditions.

E(ND).C.b 6400 44.1 278.5 48.8 0.148 3.77AVG 6100 42.1 323.5 56.7 0.188 4.76

X.C.a 6100 42.1 209.3 36.7 0.168 4.26X.C.b 6550 45.2 162.4 28.4 0.157 3.99AVG 6330 43.6 185.9 32.5 0.162 4.12

E(HD).G 5030 34.7 601.9 105.4 0.157 3.98X(SB).G 4840 33.4 577.1 101.1 0.249 6.32

4. Test setup

Two different test setups were used in this experimental pro-gram. The first test setup was designed to test the panels in directshear up to failure, while the second setup was designed to subjectthe test panels to sustained load. Details of the two test setups aregiven in the following sub-sections.

4.1. Direct shear test setup

The test setup used for direct shear testing is shown in Fig. 5.The test panels were supported vertically at the bottom edges of

Fig. 7. Shear flow vs. relative vertical displacement of sp

the two outer concrete wythes using Hollow Structural Sec-tion (HSS) steel members. Load was applied using a 120 kip(534 kN) hydraulic jack and a stiff HSS steel member to ensure uni-form load distribution. The top HSS steel member was centered onthe inner wythe to spread load evenly across the entire panel

ecimens with CFRP grids under short-term loading.

Fig. 9. Relative vertical displacement vs time of panels with CFRP grid and EPS(ND) foam under sustained load.

H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734 727

width. Relative vertical displacements between the outer and innerconcrete wythes were measured at different locations. Appliedload was measured using a 100 kip (445 kN) load cell. The relativevertical displacement between the inner and outer concretewythes was measured throughout the loading history using linearpotentiometers as shown in Fig. 5. The relative vertical displace-ment was measured at three different locations and the reportedvalues are average of the three different locations. All instrumentswere connected to an electronic data acquisition system for contin-uous data collection.

4.2. Sustained load setup

The setup was designed to subject the test panels to sustainedshear loads. The test panels were supported vertically at the bot-tom edge of the two outer concrete wythes by steel HSS members.

Fig. 10. Relative vertical displacement vs. time of panels

A calibrated hydraulic jack and a constant pressure accumulatorwere used to apply the selected load to the top surface of the innerconcrete wythe through two HSS steel member as shown in Fig. 6.The top HSS members were centered on the inner wythe and wereseparated by a steel roller, to ensure uniform distribution of theload. The relative vertical displacement between the inner wytheand the outer wythes was measured using digital caliper with anaccuracy of 0.02 mm at four different locations throughout theapplication of the sustained load.

5. Results and discussion

5.1. Short-term shear strength

Figs. 7 and 8 show the measured shear flow versus the relativevertical displacement of the panels with CFRP grid and GFRP grid

with CFRP grid and XPS foam under sustained load.

Fig. 11. Relative vertical displacement vs. time of panels with CFRP grid and EPS(ND) foam under sustained load using logarithmic scale.

728 H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734

tested at 28 days, respectively. The shear flow on the proposedshear mechanism was computed as the applied load acting onthe total length of the FRP grid used, i.e. 88 in. (2235 mm). Thereported relative vertical displacement is an average of three mea-sured values at three different locations of the test panels. Themaximum measured shear flow and the corresponding relativevertical displacement for the six test panels are summarized inTable 6 and are also identified by the dots on the correspondingcurve in Figs. 7 and 8. The measured concrete compressive strengthat day of testing is also given in Table 6. The maximum measuredshear flow is considered to be the short-term shear strength of thetest panels.

Test results indicate that both types of FRP grid can be used as ashear transfer mechanism for precast sandwich wall panels. Shearstrength of the panels utilizing GFRP grid was slightly higher than

Fig. 12. Relative vertical displacement vs. time of panels with CFRP g

that with CFRP grid due to the use of larger cross section of thestrands and smaller diameters of the grid. The use of EPS foam typeresults in higher shear strength in comparison to XPS foamwithoutsand blasting. However, sand blasting the XPS foam enhances theperformance of the panels as reported by Sopal [6] by comparingthe behavior of XPS panels with and without Sand-Blasting. Thebehavior emphasize that increasing the surface roughness of thefoam insulation provides better bond between the concrete wythesand the foam, ultimately leading to higher shear flow per the sameline of grid. The noticeable difference in the strength of the dupli-cate segments highlights the impact of the manufacturing processon the strength. During casting, there is a possibility that air maybe entrapped between the insulating foam and concrete interfacecausing reduction of the bond and consequently the overall shearflow strength.

rid and XPS foam under sustained load using logarithmic scale.

Fig. 13. Relative vertical displacement vs. time of panels with GFRP grid and EPS(HD) foam under sustained load.

Fig. 14. Relative vertical displacement vs. time of panels with GFRP grid and XPS(SB) foam under sustained load.

H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734 729

5.2. Effect of sustained loading

5.2.1. Creep of panels with CFRP gridThe measured relative vertical displacement throughout the

duration of sustained load of 12 months for the panels with CFRPgrid and EPS(ND) or XPS foam, respectively are shown in Figs. 9and 10. The reported displacement is an average of four measure-ments, two on each side of the test panels. In order to extrapolatethe behavior of the panels under sustained load, the same data ispresented in Figs. 11 and 12 for the EPS(ND) and XPS foam usinglog–log scale, respectively.

Test results of the panels with CFRP grid indicate that creepeffect due to sustained load occurred approximately in the firstmonth, after which the effect was significantly reduced. For thepanels with EPS(ND) foam and subjected to sustained load levelof 20% and 35%, the relative vertical displacement was virtuallyconstant after the first month. For the same panels subjected tosustained load level of 45%, the vertical displacement continuously

increased with time; however the rate of increase was remarkablyreduced after the first month. The panels with XPS foam withoutsurface treatment exhibited the same trend as those with EPS(ND) panels, except that the panels subjected to sustained loadlevel of 35%. The panels subjected to sustained load of 35% hadsame behavior as those subjected to 45% sustained load. Thisbehavior highlights that the panels with XPS foam without surfacetreatment are more vulnerable to creep effect due to sustainedload in comparison with panels with EPS(ND) foam. In light ofthe test results, it can be justified that for panels with CFRP grid,to minimize the creep, sustained load level should be limited to aload equivalent to 35% of the ultimate load. This sustained loadlimits the long-term differential movement between the two con-crete wythes to 1/16 in. (2 mm) as shown in Figs. 11 and 12.

5.2.2. Creep of panels with GFRP gridThe measured relative vertical displacement throughout the

duration of application of the sustained load of six months for

Fig. 15. Relative vertical displacement vs time of panels with GFRP grid and EPS(HD) foam under sustained load using logarithmic scale.

Fig. 16. Relative vertical displacement vs time of panels with GFRP grid and XPS(SB) foam under sustained load using logarithmic scale.

730 H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734

the panels with GFRP grid and EPS(HD) or XPS(SB) foam, respec-tively are shown in Figs. 13 and 14. The reported displacement is

Table 7Results of direct shear test for long-term behavior of panels with CFRP grid.

Panel designation Concrete strength Maximum shear flow

psi N/mm2 lb/in. N/mm Ratio

E(ND).C.0 6090 42.0 260.8 45.7 1.00E(ND).C.20 7300 50.3 288.0 50.4 1.10E(ND).C.35 7170 49.4 298.5 52.3 1.14E(ND).C.45 6830 47.1 261.0 45.7 1.00

X.C.0 7850 54.1 184.8 32.4 1.00X.C.20 7430 51.2 188.9 33.1 1.02X.C.35 7160 49.4 145.2 32.1 0.79X.C.45 6680 46.1 183.3 32.1 0.99

an average of four measurements taken, with two readings on eachside of the test panels.

Table 8Results of direct shear test for long-term behavior of panels with GFRP grid.

Panel designation Concrete strength Maximum shear flow

psi N/mm2 lb/in. N/mm Ratio

E(HD).G.0 6530 45.0 526.9 92.3 1.00E(HD).G.20 6530 45.0 524.9 91.9 0.99E(HD).G.30 6950 47.9 507.8 88.9 0.96E(HD).G.50 6950 47.9 525.3 92.0 0.99

X(SB).G.0 6360 43.9 546.9 95.8 1.00X(SB).G.20 6360 43.9 544.4 95.3 0.99X(SB).G.30 6570 45.3 536.4 93.9 0.98X(SB).G.50 6570 45.3 512.9 89.8 0.94

(a) (b)

Fig. 17. Typical shear deformation and cracking of foam during testing (a) EPS(HD), (b) XPS(SB).

Fig. 18. Shear flow vs. relative vertical displacement of CFRP grid panels tested at 28 days and panels not subjected to the sustained load and tested at 12 months.

Table 9Aging effect of panels with CRFP grid.

Panel designation Concrete strength Maximum shear flow

psi N/mm2 lb/in. N/mm Ratio

E(ND).C.AVG 6100 42.1 323.5 56.7 1.00E(ND).C.0 6090 42.0 260.8 45.7 0.81

X.C.AVG 6330 43.6 185.9 32.5 1.00X.C.0 7850 54.1 184.8 32.4 0.99

H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734 731

Similar to the test panels with CFRP grid, the panels with GFRPgrid exhibited the same trend. Test results of the panels with GFRPgrid indicate that most of the creep effect due to sustained loadoccurs approximately in the first two weeks, after which the effect

was significantly reduced. For panels with EPS(HD) the sustainedload level of 20% and 30% had almost the same effect and the rel-ative vertical displacement was approximately constant after thefirst two weeks. For the same panels subjected to sustained loadlevel of 50%, the vertical displacement slightly increased with timeafter the first two weeks. Similar behavior was exhibited by thepanels with XPS(SB) throughout the duration of the sustained load.Similar to the panels with CFRP grid, it can be justified that for pan-els with GFRP grid creep effect can be minimized if the sustainedload level be limited to a load equivalent to 30% of the ultimatestrength of the GFRP grid/rigid foam connection. This also limitsthe long-term differential movements to 1/16 in. (2 mm) as shownin Figs. 15 and 16, respectively.

Fig. 19. Shear flow vs. relative vertical displacement of GFRP grid panels tested at 28 days and panels not subjected to the sustained load and tested at seven months.

Table 10Aging effect of panels with GRFP grid.

Panel designation Concrete strength Maximum shear flow

psi N/mm2 lb/in. N/mm Ratio

E(HD).G 5030 34.7 601.9 105.4 1.00E(HD).G.0 6530 45.0 526.9 92.3 0.88

X(SB).G 4840 33.4 577.1 101.1 1.00X(SB).G.0 6360 43.9 546.9 95.8 0.95

732 H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734

5.2.3. Sustained loading effect on shear strength of panels with CFRPgrid

After the application and removal of the sustained load, the testpanels were tested in direct shear up to failure. Table 7 gives the

Fig. 20. Shear flow vs relative vertical displacement of panels with GFRP

maximum measured shear flow and the measured concrete com-pressive strength at day of testing. It should be noted that two pan-els that were kept unloaded for 12 months [E(ND)C.0 and X.C.0]were also tested and were used as control panels to assess theeffect of the sustained load on the ultimate shear strength.

Test results show that the ultimate shear strengths of the panelswith CFRP grid and EPS(ND) foam subjected to different sustainedload levels were not affected by the sustained loading. Similarlythe shear strengths of the panels with CFRP grid and XPS foamwithout surface treatment were not impacted by the sustainedload. The shear strength of the panel with XPS foam and subjectedto load level of 35% of the ultimate strength (X.C.35) only wasreduced. Inspection of this panel prior to ultimate testing, indi-cated loss of bond between the XPS foam and the concrete wythe

grid and EPS(HD) foam subjected to seven month outdoor exposure.

Fig. 21. Shear flow vs relative vertical displacement of panels with GFRP grid and XPS(SB) foam subjected to seven month outdoor exposure.

Table 11Summary and comparison of results measured for environmental panels.

Panel designation Concretestrength

Maximum shear flow

psi N/mm2 lb/in. N/mm Ratio

E(HD).G.0 6530 45.0 526.9 92.3 1.00 AVGE(HD).G.EXP.A 6470 44.6 537.4 94.1 1.02 1.00E(HD).G.EXP.B 6470 44.6 515.5 90.3 0.98

X(SB).G.0 6360 43.9 546.9 95.8 1.00 AVGX(SB).G.EXP.A 6470 44.6 484.9 84.9 0.89 0.94X(SB).G.EXP.B 6470 44.6 536.2 93.9 0.98

H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734 733

at one interface, hence the shear transfer mechanism for that panelwas mainly provided by the CFRP grid and possibly some frictionbetween the foam and the concrete surface. The reduction of theshear strength of this panel is attributed to the loss of bond of thesepanels. Generally, the sustained loading has no effect on the ulti-mate shear strength. Accordingly, based on these limited data, itcan be claimed that long-term strength of the sandwich concretepanels utilizing CFRP grid is equal to their short-term strength.

5.2.4. Sustained loading effect on shear strength of panels with GFRPgrid

After the application and removal of the sustained load, the testpanels were ultimately tested in direct shear to failure. The mea-sured shear strengths of the loaded panels were compared to thoseof the unloaded panels [E(HD).G.0 and X(SB).G.0] for comparisonpurposes. Table 8 gives the maximum measured shear flow andthe measured concrete compressive strength at day of testing.

Test results show that the shear strengths of the panels withGFRP grid regardless of the foam type were virtually unaffectedby the sustained load. The bond between the EPS(HD) rigid foamand the concrete wythes remained intact throughout the durationof subjection to the sustained loading. This was evident by theobserved shear deformation and cracks in the EPS(HD) rigid insu-lation during testing, as shown in Fig. 17(a). Similarly, the bondbetween the XPS(SB) foam and the concrete wythes remainedintact during the sustained loading phase of the program for allpanels. However, it was observed that the failure was caused by

shear sliding of the foam at the concrete interface, as shown inFig. 17(b).

5.2.5. Aging effect of panels with CFRP gridIn order to evaluate the aging effect, the shear strengths of the

panels with CFRP grid that are not subjected to sustained load andtested at age of 12 months, were compared to those tested at28 days reported in the short term study as shown in Fig. 18.Table 9 gives the maximum measured shear flow of the aged andcontrol panels as well as the ratios of their strengths along withthe measured concrete compressive strength at day of testing. Itshould be noted that for each foam type, two identical panels weretested at 28 days; therefore the average strength of the two panelsgiven in Table 6 was used as the control value.

Test results show that the shear of strength was reduced by 19%and 1% for panels with EPS(ND) and XPS foam, respectively. Thereduction in the strength could be related to the aging of the foamwhich could lead also to deterioration of the bond. The reduction ismore pronounced for EPS(HD) foam in comparison to the XPS(SB)foamwhich has more closed cells; and is attributed to obtain lowerstiffness for panels with EPS(ND). Since only one panel was testedfor each foam type and due to the lack of statistical data, no definedconclusion can be recommended on the effect of aging. However,the results highlights that EPS(ND) foam might be more vulnerableto aging effect in comparison to XPS foam without surfacetreatment.

5.2.6. Aging effect of panels with GFRP gridThe measured shear strengths of the panels with GFRP grid, EPS

(HD) and XPS(SB) foams that are not subjected to sustained loadand tested at age of seven months were compared to those testedat 28 days of the first study [E(HD).G and X(SB).G in Table 6] asshown in Fig. 19. Table 10 gives the measured shear flow the mea-sured concrete compressive strength at day of testing of the fourpanels with GFRP grid and EPS(HD) or XPS(SB) foam.

Test results show that the shear of strength was reduced by 12–5% for EPS(HD) and XPS(SB) panels, respectively. In addition, it isworth noting that specimens tested at 28 days of curing exhibitedsignificantly greater ductility compared to the specimens testedafter seven months of indoor exposure. It is believed that the

734 H. Kazem et al. / Construction and Building Materials 98 (2015) 722–734

reduction in both strength and ductility could be attributed to theaging of the foam which could lead to deterioration of the bond.The reduction is more pronounced for EPS(HD) foam in comparisonto the XPS(SB) foam which has more closed cells.

Similar to the panels with CFRP grid, since only one panel wastested for each foam type and due to the lack of statistical data,no defined conclusion can be recommended on the effect of aging.Test results indicates that EPS(HD) foammight be more susceptibleto aging effect in comparison to XPS foam with Sand-Blasted sur-face similar to EPS(ND) and XPS without surface treatment usedfor panel with GFRP grid. This behavior confirms that, generallyEPS foam is more prone to aging effect compared to XPS foam.

5.3. Outdoor exposure

The four test panels with GFRP grid and subjected to outdoorexposure, were tested in direct shear up to failure after sevenmonths of exposure. Effect of outdoor weather conditions wasdetermined by comparing the behavior and the measured shearstrength of the exposed panels to the panels kept inside the labo-ratory (controlled environment with a temperature of 20 �C ± 2(68 �F ± 4) and relative humidity of 70% ± 5) and not subjected tosustained load in the long-term study [E(HD).G.0 and X(SB).G.0in Table 8]. It is worth noting that the strengths of the exposedpanels were not compared to those tested at 28 days to avoid tothe aging effect. Figs. 20 and 21 show the measured shear flow ver-sus relative vertical displacement of the GFRP panels with EPS(HD)and XPS(SB) foam, respectively. In addition, Table 11 gives themaximum measured shear flow along with the concrete compres-sive strength.

It is evident from test results that limited-period outdoor expo-sure for the reported weathering conditions did not have a notice-able effect on the shear strength of panels with GFRP grid. Asshown in Table 11, the shear strength of the exposed panels withEPS(HD) was virtually equal, while that of the panels with XPS(SB) was reduced by 6% on average. This highlights that the panelswith XPS(SB) foam are more prone to outdoor exposure effect com-pared to panels with EPS(HD) foam.

6. Conclusions

Based on the test results, the following conclusions and recom-mendations can be drawn:

(1) Carbon and glass FRP grid with EPS and XPS foam insulationcan provide the shear transfer mechanism for precast con-crete sandwich panels.

(2) Shear strength of panels with EPS foam outperformed thosewith XPS foam; however, bond of XPS foam and conse-quently shear strength of the panels can be greatly enhancedby sandblasting the surface of XPS foam.

(3) Quality of fabrication of precast concrete sandwich panelscan affect the shear strength of the panels due to possibletrapping of air between the concrete and the foam. Trappedair at the interface can reduce the bond between the con-crete and the foam, and consequently the shear strength ofthe panel.

(4) It is recommended that the design strength of the panelshould be limited to 30% of the ultimate shear strengthcapacity under the sustained loads to hinder the creep effect.

Design of FRP grid (grid spacing and foam thickness) of load-bearing sandwich panels should limit the induced service(un-factored) shear stress to 30% of its ultimate short-termshear strength. Accordingly, the concrete sandwich panelsshould be designed for the sustained load factor of 3.33(1/0.3).

(5) The panels with EPS foam are more susceptible to agingeffect in comparison to the panels with XPS foam regardlessof the type of FRP grid used.

(6) The limited-period outdoor exposure for the reportedweathering conditions did not have a noticeable effect onthe ultimate shear strength of the panels with GFRP grid.However, test results highlights that the XPS foam may bemore sensible to weather condition. Accordingly, it may beadvisable to protect the foam edges from direct exposureto the environment.

Acknowledgments

The authors are grateful to the Altus Group and the Korea Insti-tute of Construction Technology (KICT) for their financial supportand providing of the materials. The authors are also grateful tothe staff of the Constructed Facilities Laboratory at North CarolinaState University for their sincere help throughout the experimentalprogram.

References

[1] PCI Committee on Precast Sandwich Wall Panels, State of the art of precast/prestressed concrete sandwich wall panels, PCI J. 52 (2 (spring)) (2011) 131–176.

[2] H. Gleich, New carbon fiber reinforcement advances sandwich wall panels,Struct. Mag. (2007) 61–63.

[3] B.A. Frankl, G.W. Lucier, T.K. Hassan, Sami H. Rizkalla, Behavior of precast,prestressed concrete sandwich wall panels reinforced with CFRP shear grid,PCI J. 52 (2(spring)) (2011) 42–54.

[4] Y.J. Kim, H. Messenger, T. Harmon. Composite insulated precast wall panelswith shear transfer provided by carbon fiber grid, in: Proceedings of 3rd FibInternational Congress, Washington, D.C. (May–June), 2010.

[5] Bunn, G. William, CFRP Grid/Rigid Foam Shear Transfer Mechanism for Precast,Prestressed Concrete Sandwich Wall Panels (MSc thesis), Department of Civil,Construction, and Environmental Engineering, North Carolina State University,Raleigh, NC, 2011, pp. 190.

[6] G.J. Sopal, Use of CFRP Grid as Shear Transfer Mechanism for Precast ConcreteSandwich Wall Panels (PhD dissertation), Department of Civil, Constructionand Environmental Engineering, North Carolina State University, Raleigh, NC,2013, pp. 174.

[7] K. Hodicky, G. Sopal, S. Rizkalla, T. Hulin, H. Stang, Experimental and numericalinvestigation of the FRP shear mechanism for concrete sandwich panels, ASCEJ. Comp. Constr. (2014). 10.1061/ (ASCE) CC.1943-5614.0000554, 04014083.

[8] ICC-ES AC422, Acceptance criteria for semi-continuous fiber-reinforced gridconnectors used in combination with rigid insulation in concrete sandwichpanel construction, International Code Council (ICC), 2010.

[9] J.G. Soriano, GFRP Shear Grid for Precast, Prestressed Concrete Sandwich WallPanels (MSc thesis), Department of Civil, Construction and EnvironmentalEngineering, North Carolina State University, Raleigh, NC, 2013, pp. 239.

[10] C. Naito, J. Hoemann, M. Beacraft, B. Bewich, Performance and characterizationof shear ties for use in insulated precast concrete sandwich wall panels, ASCE J.Struct. Eng. 138 (1) (2012) 52–61.

[11] E. Insel, Shear testing of concrete sandwich panels with carbon fiber gridreinforcement (MSc thesis), Faculty of graduate school, University ofWyoming, 2007, pp. 106.

[12] ASTM C39/C39M-12a, Standard Test Method for Compressive Strength ofCylindrical Concrete Specimens, ASTM International, West Conshohocken, PA,USA, 2012 (pp. 7).

[13] ASTM C39/C39M-12a, Standard Test Method for Compressive Strength ofCylindrical Concrete Specimens, ASTM International, West Conshohocken, PA,USA, 2012 (pp. 7).