ACI

STRUCTURAL J O U R N A L

A JOURNAL OF THE AMERICAN CONCRETE INSTITUTE

Prepublished Paper
This is a prepublished manuscript. The
final manuscript is tentatively scheduled
for V. 120, No. 4 and is subject to change.

The DOI for this paper is 10.14359/51738717

and will not change, but won’t be activated
until the issue has been published.
 1 Design and Testing of LWSCC Bridge-Deck Slabs Reinforced with GFRP Bars

2 Mahdi Aflakisamani, Salaheldin Mousa, Hamdy M. Mohamed, Ehab A. Ahmed, and Brahim

3 Benmokrane

4 Biography:

5 Mahdi Aflakisamani is a doctoral candidate in the Department of Civil and Building Engineering at

6 the University of Sherbrooke (Sherbrooke, QC, Canada). He received his BSc from the Department of

7 Civil Engineering, Shahrekord University (Iran) and MSc from the Department of Civil Engineering,

8 Islamic Azad University, Najafabad branch (IAUN, Iran). His research interests include the use of

9 fiber-reinforced polymers (FRPs) in reinforced concrete structures.

10 Salaheldin Mousa is an FQRNT postdoctoral fellow in the Department of Civil and Building

11 Engineering at the University of Sherbrooke (Sherbrooke, QC, Canada), where he also received his

12 PhD. He received his BSc and MSc from the Faculty of Engineering at Shoubra, Benha University

13 (Cairo, Egypt). His research interests include the use of fiber-reinforced polymers (FRPs) in reinforced

14 concrete structures.

15 Hamdy M. Mohamed is a lecturer and research associate in the Department of Civil and Building

16 Engineering at the University of Sherbrooke (Sherbrooke, QC, Canada). He received his BSc and MSc

17 from the Faculty of Engineering at Helwan University (Cairo, Egypt) and his PhD from the University

18 of Sherbrooke (Sherbrooke, QC, Canada). His research interests include the use and field applications

19 of fiber-reinforced polymers (FRPs) in reinforced-concrete structures.

20 Ehab A. Ahmed is a former research associate in the Department of Civil and Building Engineering at

21 the University of Sherbrooke (Sherbrooke, QC, Canada). He is currently an instructor at Concordia

22 University (Montreal, QC, Canada). He received his BSc and MSc from the Faculty of Engineering,

23 Menoufia University (Menoufia, Egypt), and his PhD from the University of Sherbrooke. His research

24 interests include the use and field applications of FRPs in reinforced concrete structures.

1
 1 Brahim Benmokrane, FACI, is Professor in the Department of Civil and Building Engineering at the

2 University of Sherbrooke (Sherbrooke, QC, Canada) and Tier-1 Canada Research Chair Professor in

3 Advanced Composite Materials for Civil Structures and Industrial Research Chair Professor in FRP

4 Reinforcement for Concrete Infrastructure and Director of the University of Sherbrooke Research

5 Center on Structural FRP Composite Materials for Concrete Structures (CRUSMAC). He is a current

6 member of ACI Committees, 440 on Fiber-Reinforced Polymer Reinforcement and 435 on Deflection

7 of Building Concrete Structures, and a member of ACI Subcommittees 440-H, FRP-Reinforced

8 Concrete and Building Code Requirements for Structural Concrete Reinforced Internally with Fiber

9 Reinforced Polymer (FRP) Bars; ACI 440-L, FRP-Durability; ACI 440-I, FRP-Prestressed Concrete;

10 ACI 440-F, FRP-Repair-Strengthening, and ACI 440-E, Professional Education. He is current member

11 and past Co-Chair of ACI Subcommittee 440-K, FRP-Material Characteristics. He serves as co-chair of

12 CSA S806 and CSA S6. He is the founding chair of CSA S807 and CSA S808. His research interests

13 include the development of FRP reinforcement for concrete structures and their durability, structural

14 performance, and field applications.

15

16

17

18

19

20

2
 1 ABSTRACT
2 Advances in new lightweight self-consolidating concrete (LWSCC) mix designs have led to the

3 construction of new concrete structures with much lower weight and higher strengths. The integration

4 of glass fiber-reinforced polymer (GFRP) bars with LWSCC can be used effectively to build durable

5 bridges with smaller cross sections and extended service lives. This study aimed at evaluating the

6 effectiveness of this type of concrete for building concrete bridge-deck slabs with GFRP reinforcement.

7 Five full-scale edge-restrained concrete bridge-deck slabs were fabricated, simulating a slab-on-girder

8 bridge deck commonly used in North America. The bridge-deck slabs were 3000 mm (118.1 in.) in

9 length, 2500 mm (98.4 in.) in width, and 200 mm (7.9 in.) in thickness. The test parameters included

10 reinforcement type (Sand coated GFRP or helically wrapped GFRP and steel) and reinforcement ratio

11 (ranging from 0.44% to 1.15%). The bridge-deck slabs were designed according to the Canadian

12 Highway Bridge Design Code. The specimens were exposed to a concentrated load over a contact area

13 of 250 × 600 mm (9.8 × 23.6 in.), which simulates the footprint of a sustained truck wheel load (87.5

14 kN CL-625 truck), as specified in Canadian standards. The test results indicate that the failure mode of

15 all deck slabs was punching shear. The recorded ultimate load capacities for all specimens exceeded the

16 design factored load, which validates the use of GFRP-reinforced LWSCC for the construction of

17 bridge-deck slabs. It was also concluded that the surface conditions of the GFRP bars (sand coated or

18 helically wrapped) had a minor effect on the cracking, deflection, and behavior of the tested LWSCC

19 deck slabs. In addition, increasing the axial-reinforcement stiffness in the GFRP-reinforced slabs

20 significantly increased the ultimate capacity and reduced maximum crack width, reinforcement strains,

21 and mid-span deflection at ultimate load.

22

23 Keywords: Lightweight self-consolidating concrete; GFRP rebars, bridge deck slab; wheel-load;

24 ultimate capacity; cracking patterns and strains; load-deflection; punching shear; design codes.

25
3
 1 INTRODUCTION

2 Concrete made with lightweight aggregates can have many significant applications, including the

3 construction of bridge elements. This type of concrete reduces structure mass, consequently reducing

4 the cross section and the reinforcement used in concrete structures. The weight of long-span bridges is

5 considerably more than imposed traffic loads, so that minimizing the structure's dead load becomes an

6 essential design parameter. Replacing normal aggregates with lightweight (LW) aggregates reduces the

7 weight of the concrete by 25% to 35% while maintaining structural capacity (Harmon 2007).

8 Concrete bridge-deck slabs are more likely to deteriorate than any other bridge element because they

9 experience harsh environmental conditions, including the routine use of deicing salts as well as traffic

10 loads, freeze–thaw cycles, and wet–dry cycles, all of which corrode steel reinforcing bars (Benmokrane

11 et al. 2020; Gooranorimi et al. 2019; Elgabbas et al. 2016; Arafa et al. 2016). As steel reinforcement

12 corrodes when exposed to chlorides and deicing chemicals, fiber-reinforced polymer (FRP) bars are

13 increasingly used in the construction industry, particularly for marine structures, concrete bridge-deck

14 slabs, and parking garages (Benmokrane et al. 2020; ACI 440.1R-15 2015; Benmokrane et al. 2007).

15 FRP bars have many advantages compared to steel, such as high tensile strength and light weight

16 (Mehany et al. 2021; Benmokrane et al. 2021; Gooranorimi et al. 2019; Mousa et al. 2018; ACI

17 440.1R-15 2015). Besides, FRP bars help reduce the cost of maintenance, repair, and rehabilitation

18 because of their noncorrodible nature (Benmokrane et al. 2020). Using lightweight concrete (LWC)

19 reinforced with FRP bars would be an effective solution to improve the strength-to-weight ratio and

20 reduce the lifetime cost of the bridge-deck slabs (Shirley-Smith et al. 2020). In addition, the use of

21 LWC reinforced with FRP bars in accelerated bridge construction (ABC) can minimize traffic

22 disruptions, improve work-zone safety, minimize environmental impacts, improve constructability,

23 enhance quality, and lower life-cycle costs (Youssef et al. 2019).

4
 1 Limited experimental programs have been conducted to evaluate the behavior of GFRP-reinforced

2 lightweight concrete members (ACI 440.1R-15 2015). Comparing the results of the shear behavior of

3 one-way glass FRP-reinforced lightweight concrete slabs with the ACI 440.1R-06 (2006) design

4 provisions showed that these predictions are less conservative for lightweight concrete (LW) than for

5 normal-weight (NW) concrete (Pantelides et al. 2012a&b). Their experimental results, however,

6 showed that the GFRP-reinforced lightweight panels behaved similar to the NW concrete panels

7 (Pantelides et al. 2012b). Canadian (CSA S806-02 2002) and Japanese (JSCE 1997) standards provide

8 predictions of the shear strength of LW and NW concrete panels reinforced with GFRP bars with a

9 lower degree of conservativism. Consequently, a reduction factor was determined to be unnecessary

10 (Liu and Pantelides 2013). The AASHTO LRFD Bridge Design Guide Specifications for GFRP-

11 Reinforced Concrete (AASHTO 2018) has no provisions for the use of lightweight concrete reinforced

12 with GFRP bars because of the lack of experimental studies. This guide’s specifications allow the use

13 of LW to improve the predicted values based on the experimental results obtained from tested LW

14 slabs reinforced with GFRP bars.

15 The results from past studies on FRP-reinforced bridge-deck slabs indicate that punching shear was the

16 failure mode of the tested bridge-deck slabs (El-Gamal et al. 2005). They also show that reinforcement

17 type (GFRP for glass FRP, CFRP for carbon FRP, and steel) had a negligible effect on the deflection

18 behavior of these deck slabs (Ahmad et al. 1993; El-Gamal et al. 2005; 2007). Moreover, FRP

19 reinforcement can improve the cracking performance of and induced strains in FRP-reinforced concrete

20 slabs by enhancing the bond between the concrete and reinforcing bars (Hussein and El–Salakawy

21 2018; Elgabbas et al. 2016).

22 Laboratory Tests on Lightweight Concrete Panels

23 Pantelides et al. (2012a) implemented an experimental study focusing on the shear strength of GFRP-

24 reinforced lightweight concrete panels. The specimens were subjected to a load simulating the wheel-

5
 1 load area of the AASHTO (2009) design truck with simply supported edges and center-to-center

2 spacing of 2440 mm (96 in.). The results show that the LWC specimens obtained 80% of the ultimate

3 shear strength of the NW concrete specimens and 89% of their maximum deflection. A reduction factor

4 of 0.85 was proposed to modify the ACI 440.1R-06 (2006) shear equations for GFRP-reinforced LWC.

5 Liu and Pantelides (2013) extended the study mentioned above by testing 12 LWC and 8 NWC panels.

6 All specimens were reinforced with GFRP bars and all obtained 1.3 times the predicted ultimate shear

7 capacity provided by the ACI 440.1R-06 (2006) equation. In contrast, the Canadian (CSA S806-02

8 2002) and Japanese (JSCE 1997) standards predicted the shear capacity of GFRP-reinforced NWC and

9 LWC panels with similar degrees of conservatism that were less than that calculated according to

10 ACI 440.1R-06 (2006) guidelines. Hence a reduction factor of 0.8 was proposed to calculate the shear

11 capacity of GFRP-reinforced LWC panels defined in ACI 440.1R-06 (2006) to acquire a similar degree

12 of conservatism as that of NWC panels.

13 Experimental Projects of FRP-Reinforced Bridge-Deck Slabs

14 Elgabbas et al. (2016) experimentally investigated the behavior of edge-restrained concrete bridge-deck

15 slabs reinforced with basalt FRP (BFRP) bars. Seven full-scale slabs were tested under concentrated

16 load over a contact area of 250 × 600 mm (9.8 × 23.6 in.), simulating the footprint of a sustained truck

17 wheel load (87.5 kN CL-625 truck). The results show that ACI 440.1R-15 (2015) produced

18 conservative predictions in terms of the punching-shear strength of BFRP-reinforced concrete bridge-

19 deck slabs, while CSA S806-12 (R2017) yielded predictions with a lower degree of conservatism.

20 Moreover, the reinforcement ratio of the bottom assembly in the transverse direction was recognized as

21 the main parameter affecting the structural behavior (deflection, strains, and crack width). It is worth

22 mentioning that El-Gamal et al. (2005) used a similar test procedure and specimen dimensions but with

23 different reinforcing bars (GFRP and CFRP bars). In their study, punching shear was the failure mode

24 for all the tested specimens. The recorded ultimate load capacities were more than three times the

6
 1 design factored load specified in CSA S6-10 (2010). Generally similar results were obtained from the

2 investigation conducted by Bouguerra et al. (2011).

3 RESEARCH SIGNIFICANCE

4 No research has yet been reported on the behavior of LWSCC bridge-deck slabs reinforced with FRP

5 bars. This study experimentally investigated the punching-shear behavior of a series of full-scale

6 LWSCC bridge-deck slabs reinforced with different types of GFRP bars subjected to a concentrated

7 load, simulating truck wheel load. The experimental results were used to evaluate the FRP punching-

8 shear design provisions in CSA S806-12 (R2017), ACI 440.1R-15 (2015), and AASHTO (2018). It

9 also establishes a step towards developing design provisions and recommendations for engineers in

10 designing LWSCC bridge-deck slabs with GFRP reinforcement. Moreover, the findings of this

11 pioneering study will support the work of North American technical committees engaged in developing

12 standards and design provisions for GFRP-RC deck slabs made with lightweight-aggregate self-

13 consolidating concrete (LWSCC).

14 EXPERIMENTAL PROGRAM

15 Material Properties

16 Concrete

17 Lightweight self-consolidating concrete (LWSCC) was used in this study (see Table 1 for the mix

18 design). The mix was made with lightweight coarse and fine aggregates according to the specifications

19 in ASTM C330/C330M (2017) and natural normal-weight sand. The coarse and fine aggregates had a

20 maximum size of 14 mm (0.55 in.) and 5 mm (0.2 in.), respectively. The lightweight aggregates were

21 expanded shale from Northeast Solite (Solite 307, Solite 343), as shown in Figure 1-(a). As the

22 water/cement (w/c) ratio of 0.33 was chosen, the aggregates were submerged in water before mixing to

23 maintain the desired ratio by saturation. The LWSCC was prepared using TerC3 cement containing

24 75% general-use cement (GU), 20% fly ash (FA), and 5% silica fume (SF) complying with ASTM

7
 1 ASTM C150/C150M (2018). The equilibrium density of the LWSCC was 1906 kg/m3 (119 lb/ft3) as

2 measured according to ASTM C567/C567M (2019). The LWSCC was designed with a target 28-day

3 compressive strength of 40 MPa (5.8 ksi). The curing process was started after casting by covering

4 specimens with polyethylene sheets. The slabs were unmolded one day after casting and the water-

5 curing process was performed for seven days.

6 Reinforcing bars

7 The GFRP bars used in this study were Type I (No. 5) and Type II (No. 5). Type I bars had a sand-

8 coated surface, while Type II bars had a helically wrapped surface, as shown in Figure 1-(b). The

9 GFRP bars were made with vinyl-ester resin and EC-R glass fibers, classified as Grade III according to

10 CSA S807-19 (2019). The tensile strength and the modulus of elasticity of the GFRP reinforcement

11 were determined according to ASTM D7205/D7205M (2016). Deformed steel bars (15M) were used in

12 the control specimen. Table 2 presents the mechanical properties of the reinforcing bars.

13 Test Specimens

14 Five full-scale concrete bridge deck-slabs measuring 3000 × 2500 × 200 mm (118.1 × 98.4 × 7.9 in.)

15 were fabricated. Two parallel steel girders were used to support the deck slabs with 2000 mm center-to-

16 center spacing, as shown in Figure 2. The specimens had two rows of holes (27 mm (1.06 in.) in

17 diameter) at each edge, 160 mm (6.3 in.) and 250 mm (9.8 in.) from the restrained edges. These holes

18 were fitted with steel bolts to provide edge restraint. The top and bottom concrete cover for all

19 specimens was kept constant at 30 mm (1.2 in.) and 25 mm (1 in.), respectively, as specified in Clause

20 16.4.5 of CSA S6-19 (2019). The ratio of supporting girder spacing to slab thickness was less than 18,

21 as specified in Clause 8.18.3.1 of CSA S6-19 (2019). Slab length was defined to prevent one-way shear

22 before punching failure and to consider the slab area affected by the wheel load. This area was assumed

23 to be based on the outer diameter of the wedge formed during punching failure (El-Gamal et al. 2005).

24 The chosen test parameters were: (1) reinforcement ratio in the transverse direction of the bottom

8
 1 reinforcement (0.44%, 0.83%, and 1.15); (2) surface conditioning of the GFRP bars (sand coated or

2 helically wrapped); (3) and reinforcement type (steel or GFRP bars).

3 Table 3 summarizes the reinforcement details of the tested slabs. Number 5 sand-coated GFRP bars

4 were used in G1-150, G1-100, and G1-300. Number 5 helically wrapped GFRP bars were used in G2-

5 300. Deformed M15 steel bars were used in S1-300 to serve as a control specimen. The amount of

6 bottom transverse reinforcement used in the tested slabs satisfies the empirical and flexural design

7 methods in CSA S6-19 (2019) and AASHTO (2018), with a minimum reinforcement ratio of 0.44%

8 and a maximum bar spacing of 300 mm. Specimens G1-150 and G1-100 had a bar spacing of 200 mm

9 (7.9 in.) in the longitudinal and transverse directions of the top assembly. In the bottom assembly, the

10 spacing between bars was 150 mm (5.9 in.) and 100 mm (3.9 in.) in the main direction (transverse

11 direction) in G1-150 and G1-100, respectively, and 200 mm (7.9 in.) in the longitudinal direction in

12 both slabs. Specimens G1-300, G2-300, and S1-300 had a bar spacing of 300 mm (11.8 in.) in the

13 longitudinal and transverse directions of the top and bottom assembly. Figure 2 shows the geometry

14 and the reinforcement details of the tested deck slabs. The location of some reinforcing bars has been

15 modified to prevent conflicts between prefabricated holes and bars. Figure 3 shows the slab

16 construction.

17 Instrumentation

18 External and internal instrumentation was used in this study, as shown in Figure 4. Fifteen strain

19 gauges measuring 6 mm (0.24 in.) in length were installed on the surface of the top and bottom

20 reinforcing bars, as shown in Figure 4-(a). Four electrical strain gauges measuring 60 mm (2.4 in.) in

21 length were used around the loading plate to measure the concrete compressive strains. Seven

22 potentiometers (POTs) (P1 to P7) were installed at different locations on the specimens to measure slab

23 deflection, as shown in Figure 4-(b). Potentiometers P8 and P9 were installed to measure the side

24 movements of specimens while testing. When the first three cracks appeared, three LVDTs, with a

9
 1 precision of 0.001 mm, were installed to measure the crack width. All the deflections, reinforcement

2 and concrete strains, and loads were recorded by a data-acquisition system connected to a computer.

3 Test Setup and Procedure

4 The specimens were tested up to failure under a concentrated monotonic load applied to the center of

5 the slab with a servo-controlled, 1000 kN (224.8 kip) MTS hydraulic actuator. The contact area of the

6 applied load measured 250 × 600 mm (9.8 × 23.6 in.), which is specified in the Canadian Highway

7 Bridge Design Code (CSA S6-19 2019) to simulate the footprint of a sustained truck wheel load (87.5

8 kN CL-625 truck). The load was applied at a displacement-controlled rate of 0.3 mm/min (0.01

9 in./min). A 10 mm (3.9 in.) thick sheet of neoprene was used under the loading plate to ensure uniform

10 stresses on the concrete surface. The edges of the specimens were supported on two steel girders with a

11 span of 2000 mm (78.7 in.), which is similar to the test setup of Elgabbas et al. (2016). Two neoprene

12 strips 100 mm (3.9 in.) in width between the concrete surface and steel supports and two neoprene

13 strips 100 mm (3.9 in.) in width between the concrete surface and steel channels were used. The strips

14 were 3 mm (0.12 in.) and 10 mm (0.39 in.) thick, respectively. Figure 5 shows the test setup and data-

15 acquisition system. The corresponding loads were noted adjacent to each crack. The longitudinal edges

16 of the slabs were partially restrained 22 mm (0.87 in.) diameter steel bolts that tied the deck slabs to the

17 steel parts of the test setup. As shown in Figure 2, four rows of holes 27 mm (1.06 in.) in diameter were

18 prepared in the slabs during casting to fit the bolts. The steel bolts play the role of shear connectors

19 between the specimens and girders in actual slab-on-girder bridges. The slabs were restrained by

20 tightening the steel anchors with a torque moment of 160 N.m (116 lb-ft) to create uniform restraint

21 along the slab edges. The torque moment was calculated to generate horizontal shear friction between

22 the steel girder and the slab bottom surface above the horizontal component of the arching stress and to

23 prevent lateral movement between the slab and girder.

24 TEST RESULTS AND DISCUSSION

10
 1 The experimental results for the GFRP-reinforced LWSCC slabs consist of cracking patterns and

2 failure mode, cracking and ultimate load capacity, concrete and reinforcement strains, and deflection

3 behavior. Table 4 presents a summary of the test results.

4 Cracking Patterns and Loads

5 All the LWSCC specimens had similar cracking patterns despite their differences in reinforcement

6 details. Figure 6 shows the cracking patterns of all the deck slabs. The formation and propagation of

7 cracks were observed and identified from the beginning of loading. The first cracks appeared

8 underneath the loading plate, mostly in the longitudinal direction. The cracks then propagated radially

9 to the slab edges. The recorded cracking loads of the slabs were between 99.2 and 115.2 kN (22.3 and

10 25.9 kips). The average cracking load was 107.2 kN (24.1 kips), which is similar to the predicted

11 service load (110.25 kN [24.8 kips]) according to the CSA S6-19 (2019) design provisions. Specimen

12 G1-100 had the highest cracking load (115.2 kN [25.9 kips]) of the slabs. A comparison of the recorded

13 cracking loads for G1-100 (115.2 kN [25.9 kips]), G1-150 (106.9 kN [24 kips]), and G1-300 (100.0 kN

14 [22.5 kips]) confirms that reducing reinforcement spacing improved the cracking performance of the

15 deck slabs. Cracking loads recorded for G1-300 and G2-300 were 100.0 and 114.5 kN (22.5 and 25.7

16 kips), respectively. This could be related to the effect of the GFRP-bar surface treatment.

17 Approximately similar cracking load values were recorded for G1-300 (100.0 kN [22.5 kips]) and S1-

18 300 (99.2 kN [22.3 kips]). This can be attributed to the fact that the sand-coated GFRP bars

19 significantly affected the cracking performance of the deck slabs.

20 Figure 7 shows the load to average crack-width relationship for all the LWSCC specimens. Table 4

21 provides the values of the crack widths at the service, factored, and ultimate loads. The maximum

22 measured crack widths for G1-150, G1-300, and S1-300 at the service designed load were 0.12, 0.2,

23 and 0.1 mm (0.005, 0.008, and 0.004 in.), respectively. These values are less than the permissible value

24 defined in ACI 440.1R-15 (2015) and CSA S6-19 (2019) of 0.5 mm (0.02 in.) for exterior exposure.

11
 1 The cracking loads for G1-100 and G1-300 were higher than the service load, which means that the

2 first cracks appeared after the service load had been reached. Specimens G1-100 and S1-300 had

3 similar levels of axial-reinforced stiffness. Given the same load level, G1-100 had narrower average

4 crack widths than S1-300. At the factored design load level, G1-100 and G1-150 behaved similarly to

5 S1-300, while G1-300 and G2-300 had twice the crack width as S1-300. The crack-width values

6 recorded for all specimens at the factored design load level did not exceed the limit (0.5 mm) defined in

7 CSA S6-19 (2019).

8 At the ultimate load level, G1-100 showed the smallest crack width, followed by G1-150. Moreover,

9 S1-300 and G1-100, which had similar axial-reinforcement stiffness (ρ × E), exhibited different load-

10 to-crack-width relationships. Spacing the GFRP bars at 100 mm (3.9 in.) in G1-100 enhanced the slab’s

11 cracking performance compared to the specimen with steel bars (S1-300) spaced at 300 mm (11.8 in.).

12 In addition, G1-150 had lower axial-reinforcement stiffness and bar spacing than S1-300, but had a

13 better load–crack-width relationship. This confirms that using lower bar spacing reduces crack width,

14 even at lower levels of axial-reinforcement (Elgabbas et al. 2016). Furthermore, it can be posited that

15 having the same bottom reinforcement ratio and bar spacing might have led to S1-300, G1-300, and

16 G2-300 having a similar load-to-crack-width relationship, as shown in Figure 7. On the other hand, at

17 the factored design load and ultimate load level, comparable crack-width values were recorded for G1-

18 300 and G2-300, which indicates that using different bar surface treatments had a negligible effect on

19 the cracking behavior of the tested specimens.

20 Ultimate Load Capacity and Failure Mode

21 Figure 6 shows that a similar cracking pattern appeared on the bottom surfaces of all LWSCC slabs

22 after testing. All the specimens experienced punching-shear failure around the loading plate. No signs

23 of concrete crushing in the top surface or bar rupture were observed, which confirms punching shear as

24 the failure mode. The maximum loads recorded were 541, 525, 442, 421, and 538 kN (121.6, 118, 99.4,

12
 1 94.6, and 121 kips) for G1-150, G1-100, G1-300, G2-300, and S1-300, respectively. As expected, the

2 slabs with wider GFRP bar spacing in the transverse direction of the bottom mesh showed lower

3 punching-shear strengths. The recorded punching-shear capacities of the test specimens were 2.02 to

4 2.60 times the factored design load (208.25 kN [46.8 kips]) as estimated with the CSA S6-19 (2019)

5 design provisions. This proves that the slabs experienced punching-shear failure despite being designed

6 based on flexural failure. Similarly, to the findings of past studies on GFRP-, CFRP-, and BFRP-

7 reinforced concrete bridge-deck slabs, the ultimate capacity of the deck slabs exhibited a high level of

8 conservativeness (Hewitt and Batchelor 1975; Perdikaris and Beim 1988; El-Gamal et al. 2005; 2007;

9 Bouguerra et al. 2011; Elgabbas et al. 2016).

10 Table 4 presents the punching-shear strengths of all the LWSCC specimens. G1-100 and S1-300

11 showed similar punching-shear capacities (525 and 538 kN [118 and 121 kips], respectively) with close

12 axial-reinforcement stiffness (711 and 880 MPa [103.1 and 127.6 ksi], respectively). The findings were

13 similar for G1-300 and G2-300 with respect to axial-reinforcement stiffness (272 and 282 MPa [39.5

14 and 40.9 ksi], respectively) and ultimate loads (442 and 421 kN [99.4 and 94.6 kips]), respectively.

15 This confirms that the slabs with similar degrees of axial-reinforcement stiffness could achieve similar

16 levels of ultimate capacity.

17 Reinforcement and Concrete Strains

18 Table 4 presents the strains recorded by S3B (strain gauge located at the center of the middle bar in the

19 bottom reinforcement of the LWSCC slabs) and C2 (strain gauge located on the top concrete surface

20 near the loading plate), while Figure 8 shows the load-to-strain graphs. The behavior of the strains in

21 the graphs was linear up to the failure of the slabs. Generally, Figure 8 confirms that increasing the

22 axial-reinforcement stiffness reduced tensile strains in the bottom reinforcement. The effect of axial-

23 reinforcement stiffness can be seen in G1-100, G1-150, and G1-300, which had reinforcement ratios of

24 1.15%, 0.83%, and 0.44%, respectively. At the maximum load level of each slab, the bottom

13
 1 reinforcement strains were 4,600, 6,700, and 7,050 µɛ for G1-100, and G1-150, G1-300, respectively.

2 Reducing the reinforcement ratio from 1.15% to 0.83% and 0.44% increased the maximum

3 reinforcement strains by 145.8% and 153.0%, respectively. Moreover, the axial-reinforcement stiffness

4 significantly affected the maximum recorded values of bottom reinforcement strains in the tested slabs.

5 Approximately the same level of reinforcement strains was recorded for G1-300 and G2-300 (7,050

6 and 7,100 µɛ, respectively); the reinforcement had almost the same axial-reinforcement stiffness in the

7 transverse direction of bottom reinforcement (272 and 282 MPa [39.5 and 40.9 ksi], respectively). This

8 confirms that the GFRP-bars surface treatment had a negligible effect on the bottom transverse

9 reinforcement.

10 The ranges of recorded strains at the service load level (110.25 kN [24.8 kips]) for the top concrete

11 surface and middle reinforcing bars varied from -180 µɛ to -70 µɛ and from 90 µɛ to 1,250 µɛ,

12 respectively. At the factored design load level (208.25 kN [46.8 kips]), these values were from -700 to -

13 150 µɛ and from 580 µɛ to 3,260 µɛ, respectively. The large difference in the axial-reinforcement

14 stiffness of the specimens produced a large difference in the recorded strains in the top concrete surface

15 and middle reinforcing bars. The maximum strains recorded at service load were 15.8% and 18.4% of

16 that recorded at ultimate load for the top concrete surface and middle reinforcing bar, respectively.

17 These values for the factored design load were 62.4% and 48.5% of the ultimate load, respectively. The

18 safety factor values ranged from 1.6 to 5.5 for the GFRP-reinforced slabs. The recorded tensile strain

19 values at the ultimate load for the Type I GFRP bars ranged from 20% to 32% of the ultimate tensile

20 strain, and for GFRP bars type II, the recorded tensile strain value was 30% of the ultimate tensile

21 strain. Furthermore, the recorded compressive strains ranged from 27% to 37% of the maximum

22 allowable concrete compressive strain (3,500 µɛ) (CSA S806-12 [R2017]). This proves that the mode

23 of failure for all specimens was punching shear, where no concrete crushing or bar rapture was

24 observed.

14
 1 Figure 9 gives the distribution of the reinforcement strains from strain gauges installed on the middle

2 bar in the bottom reinforcement of the slab. Higher values of the strains were recorded in the middle of

3 the specimens and, conversely, the lowest values were recorded towards the supports. Eventually, these

4 values dropped to zero due to slab edge restraint. The fact that a similar trend can be seen in the graph

5 for all the tested specimens proves that no debonding of reinforcement occurred during the test.

6 Load–Deflection Behavior and Effect of Test Parameters

7 This section presents the load–deflection curves in three groups to show the effect of the test

8 parameters on the load–deflection behavior of the LWSCC deck slabs, as shown in Figure 10. It should

9 be noted that the self-weight of the deck slabs was not included in the values of the deflection and the

10 applied load. All the specimens exhibited similar linear load–deflection behavior from initial loading

11 up to the initiation of the first crack. The uncracked response for all the deck slabs showed insignificant

12 deflection, reflecting gross section stiffness. The post-cracking stiffness was considerably reduced. This

13 change in stiffness represents the transition from gross-to-effective section properties. After cracking,

14 the LWSCC deck slabs reinforced with GFRP bars behaved nearly linearly with reduced stiffness up to

15 failure. This can be attributed to the linear elastic characteristics of the GFRP reinforcement. After

16 cracking occurred, however, the load–deflection curve of the slab reinforced with steel was initially

17 linear and then exhibited reduced stiffness after yielding.

18 The influence of the reinforcement ratio on the punching-shear strength of the LWSCC deck slabs

19 reinforced with GFRP was assessed. Number 5 (16 mm) GFRP bars spaced at 300 mm were used to

20 reinforce G1-300 in the transverse direction of the bottom assembly at a reinforcement ratio of 0.44%.

21 The reinforcement ratio in G1-150 and G1-100 was 0.83% and 1.15%, respectively. Figure 10-(a)

22 provides a comparison of the moment–deflection curves for the three LWSCC slabs reinforced with

23 GFRP bars (Type I, sand coated). The figure indicates that increasing the reinforcement ratio enhanced

24 the post-cracking stiffness of the LWSCC slabs.

15
 1 Specimens G1-150 and G1-100 experienced lower deflection and higher punching-shear strength than

2 G1-300. Increasing the reinforcement ratio from 0.44% to 0.83% and 1.15% decreased the measured

3 deflections by 8.0% and 22.0% at the service-load level (110.25 kN [24.8 kips]), respectively, and at

4 the same time, the punching-shear strength enhanced by 22.3% and 18.7%. Indeed, the punching-shear

5 strength and behavior of the LWSCC slabs depended significantly on the GFRP reinforcement ratio.

6 Specimens G1-300 and G2-300 were reinforced with GFRP bars with different conditioning (sand

7 coated and helically wrapped, respectively). Figure 10-(b) shows the load–deflection behavior of G1-

8 300 and G2-300, as well as the comparable punching-shear strength of 442 and 421 kN (99.4 and 94.6

9 kips), respectively. It can be concluded that the LWSCC slabs with similar reinforcement ratios

10 behaved similarly with respect to deflection, stiffness, and punching-shear strength, regardless of the

11 reinforcing-bar surface conditioning.

12 The GFRP- and steel-reinforced LWSCC slabs (G-100 and S-300) were designed to have similar levels

13 of axial-reinforcement stiffness. As depicted in Figure 10-(c), G1-100 and S1-300 had comparable

14 load–deflection responses up to the yielding of the steel reinforcement in S1-300. After the

15 reinforcement yielded, S1-300 experienced severe deformation, resulting in higher deflection at

16 ultimate load compared to G1-100. The two slabs exhibited almost similar punching-shear strength:

17 525 and 538 kN (118 and 121 kips), respectively. GFRP-reinforced slab needed three times as much

18 reinforcement to achieve the same behavior of steel-reinforced slab, however, the use of LWSCC

19 compensated the constructability and congestion issues. This can prove that the combination of the

20 LWSCC and GFRP reinforcement is the great alternative to the conventional steel-reinforced concrete

21 slab.

22 COMPARISON OF THE EXPERIMENTAL PUNCHING-SHEAR CAPACITIES AND

23 THEORETICAL PREDICTIONS

16
 1 The equations in CSA S806-12 (R2017), ACI 440.1R-15 (2015), and AASHTO (2018) for predicting

2 the punching-shear strength of GFRP-reinforced deck slabs can be compared with the experimental

3 results of this study. CSA S806-12 (R2017) provides three design equations (Eqns. 1, 2, and 3) for

4 estimating the punching-shear capacity; the smallest estimated value should be considered as the design

5 value.

 2  ' 3
1
6 Vc  1   0.028c ( E f  f f c )  b0 d (1)
 c   

 d  1

7 Vc   s  0.19  0.147c ( E f  f f c' ) 3  b0 d (2)
 b0  

 1

8 Vc  0.056c  E f  f f c'  3  b0 d (3)
 

9 where βc is the ratio of the long side to the short side of the loading plate; λ is a factor to consider

10 concrete density; f'c is the concrete compressive strength; bo is the perimeter of the critical area

11 underneath the applied load at a distance of d/2 from the loading plate edges; d is the distance from the

12 top surface of the concrete to the center of the bottom reinforcing bars; and αs is a factor to consider

13 column location. CSA S806-12 (R2017) considers the reduction factor λ for normal-density concrete,

14 structural semilow density concrete in which all the fine aggregate is natural sand, and structural low

15 density concrete in which none of the fine aggregate is natural sand to be values of 1.0, 0.85, and 0.75,

16 respectively. Considering the weight of the lightweight sand and normal-weight sand used in the

17 LWSCC mix, the value of λ for calculating the punching-shear capacity of the tested slabs, using

18 interpolation, was considered to be 0.794.

19 According to ACI 440.1R (2015), the punching-shear capacity of two-way slabs reinforced with GFRP

20 bars is calculated with Eqns. (4) and (5), which considers the stiffness of reinforcement as an influential

21 factor in the punching-shear strength of concrete slabs.

17
 4
1 Vc  f c' b0c (4)
5

2 k  2  f n f  (  f n f )2   f n f (5)

3 where c is the neutral-axis depth (mm) of the cracked transformed section; c is kd; and nf is the modular

4 ratio, nf = Ef /Ec. ACI 440.1R-15 (2015) does not provide a specific factor for calculating the punching-

5 shear capacity of different types of LWC. ACI 318-19 (2019), however, provides a modification factor

6 λ based on equilibrium density in order to compensate for the reduction of mechanical properties of

7 LWC compared to NWC with the same compressive strength. Given a value of 1906 kg/m3 (119 lb/ft3)

8 for the equilibrium density of LWC, the value of modification factor λ for calculating the punching-

9 shear capacity of the deck slabs was considered to be 0.9.

10 According to AASHTO (2018), the shear resistance of two-way slabs can be computed Eqn. 6, where

11 Vc is the nominal shear resistance of the concrete.

12 Vc  0.316k f c' bo d v (6)

13 where k is the ratio of the depth of the neutral axis to the depth of the flexural reinforcement; f’c is the

14 specified compressive strength of concrete (ksi); bo is the perimeter of critical section calculated at a

15 distance of d/2 from the concentrated load (in.); and dv is the effective shear depth (in.). Although

16 AASHTO (2018) does not cover the use of LWC, with regards to the AASHTO (2020) design

17 specifications, when the splitting tensile strength (fct) is provided, a modification factor (λ) for concrete

18 density is determined with Eq. (7):

f ct
19   4.7  1.0 (7)
f c'

20 Table 4 presents the experimental-to-predicted ratio of the punching-shear strength. According to the

21 equations in CSA S806-12 (R2017), the experimental-to-predicted strength ratio was 1.18, 1.01, 1.22,

22 and 1.11 for G1-100, G1-150, G1-300, and G2-300, respectively. The corresponding values using the
18
 1 ACI 440.1R-15 (2015) design equation were 1.37, 1.14, 1.48, and 1.37, respectively. Moreover, the

2 calculated ratios obtained with the equations in AASHTO (2018) were 1.40, 1.15, 1.59, and 1.41,

3 respectively. According to the concrete density reduction factor in AASHTO (2020), these values

4 would be 1.40, 1.15, 1.78, and 1.47, respectively. It can be concluded that CSA S806-12 (R2017)

5 provided accurate predications compared to the experimental results. ACI 440.1R-15 (2015) and

6 AASHTO (2018) conservatively underestimated the punching-shear capacity of the tested GFRP-

7 reinforced LWSCC slabs. Moreover, all the predicted concrete punching-shear capacities have been

8 estimated without considering the reduction factor (strength reduction factor, concrete and GFRP

9 material reduction factors). Taking into account the reduction factor in the design will provide a

10 conservative prediction to the shear capacity to avoid the brittle failure. Furthermore, according to the

11 equations provided in CSA S6-19 (2019) and AASHTO LRFD Bridge Design Specifications (2020),

12 the experimental-to-predicted strength ratios for steel-reinforced slab (S1-300) were 1.22 and 1.33,

13 respectively. This showed that Canadian bridge design code S6-19 (2019) predicted more accurate

14 shear strength than AASHTO (2020). Past findings revealed that ACI 440.1R-15 (2015) predicted the

15 punching-shear strength of GFRP-reinforced LWC panels more accurately than that of GFRP-

16 reinforced NW concrete panels (Pantelides et al. 2012a; Liu and Pantelides 2013). ACI 440.1R-15

17 (2015) and AASHTO (2009), however, conservatively predicted the punching-shear capacity of GFRP-

18 reinforced NW concrete deck slabs (Bouguerra et al. 2011), BFRP-reinforced NW concrete deck slabs

19 (Elgabbas et al. 2016), and two-way slabs (Dulude et al. 2013), as well as the shear resistance of

20 GFRP-reinforced UHPFRC closure joints between one-way bridge-deck slabs (Youssef et al. 2019).

21 SUMMARY AND CONCLUSION

22 In this study, a new mix design for lightweight self-consolidating concrete (LWSCC) was developed. A

23 comprehensive experimental program was designed and conducted to investigate the behavior of edge-

19
 1 restrained GFRP-reinforced LWSCC bridge-deck slabs. The following conclusions were drawn based

2 on the results of this study.

3 1. All the specimens experienced punching-shear failure; the recorded ultimate capacities were

4 higher than the predictions and factored design load specified in CSA S6-19 (2019).

5 2. The behavior of the LWSCC deck slabs was significantly affected by the reinforcement ratio as

6 the main parameter. The cracking performance, reinforcement strains, and deflection of the

7 GFRP-reinforced tested specimens were reduced by increasing the reinforcement ratio of the

8 bottom assembly.

9 3. Comparing LWSCC deck slabs with the same axial-reinforcement stiffness (ρ × E) leads to the

10 conclusion that these slabs have similar behavior and punching-shear strength. Increasing the

11 axial-reinforcement stiffness in the GFRP-reinforced slabs significantly increased the ultimate

12 capacity and reduced maximum crack width, reinforcement strains, and mid-span deflection at

13 ultimate load.

14 4. The surface conditions of the GFRP bars (sand coated or helically wrapped) had a minor effect

15 on the cracking, deflection, and behavior of the tested LWSCC deck slabs.

16 5. A comparison of the concrete punching-shear capacities of the LWSCC deck slabs to their

17 capacities predicted with the CSA S806-12 (R2017) equations based on a concrete density

18 reduction factor revealed that these equations yielded more accurate predictions of GFRP-

19 reinforced LWSCC deck slabs than did the ACI 440.1R-15 (2015) and AASHTO (2018)

20 equations.

21 6. The experimental evidence from this investigation provides some experimental backbone for

22 including design provisions in bridge-design specifications for the use of GFRP bars as internal

23 reinforcement in LWSCC bridge-deck slabs.

20
 1 Based on the results and conclusions of this study, further investigations into the behavior of LWSCC

2 bridge deck slabs reinforced with GFRP bars under fatigue loading should be conducted to generate

3 more confidence and encourage wider acceptance of this new material. In addition, experimental tests

4 could be extended to study the use of different types of microfibers in LWSCC bridge deck slabs

5 reinforced with GFRP bars.

6 ACKNOWLEDGMENTS

7 This research was conducted with funding from the Tier 1 Canada Research Chair in Advanced

8 Composite Materials for Civil Structures, the Natural Sciences and Engineering Research Council of

9 Canada (NSERC), the Fonds de la recherche du Québec en nature et technologies (FQR-NT), and the

10 Canadian Foundation for Innovation (FCI), and for the technical help provided by the staff of the

11 structural lab of the Department of Civil Engineering at the University of Sherbrooke. The authors

12 would like to express their special thanks and gratitude to Northeast Solite Corporation for their

13 generosity. Their donation of Solite® aggregate was instrumental to the success of this research

14 project.

15 REFERENCES

16 American Association of State Highway and Transportation Officials (AASHTO), (2009). "AASHTO

17 LRFD Bridge Design Guide Specifications for GFRP-Reinforced Concrete Decks and Traffic

18 Railings," first edition, Washington, DC, 68 pp.

19 American Association of State Highway and Transportation Officials (AASHTO), (2018). “AASHTO

20 LRFD Bridge Design Guide Specifications for GFRP-Reinforced Concrete.” 2nd Edition, American

21 Association of State Highway and Transportation Officials, Washington, D.C.

22 American Association of State Highway and Transportation Officials (AASHTO), (2020). “AASHTO

23 LRFD Bridge Design Specifications.” 9th Edition, American Association of State Highway and

24 Transportation Officials, Washington, D.C.

21
 1 ACI Committee 318, (2019). "Building Code Requirements for Structural Concrete and Commentary

2 (ACI 318R-19)." American Concrete Institute, Farmington Hills, Michigan, 623 pp.

3 ACI Committee 440, (2006). "Guide for The Design and Construction of Structural Concrete

4 Reinforced with FRP Bars (ACI 440.1R-06)," American Concrete Institute, Farmington Hills, MI,

5 44 pp.

6 ACI Committee 440, (2015). "Guide for The Design and Construction of Structural Concrete

7 Reinforced with FRP Bars (ACI 440.1R-15)," American Concrete Institute, Farmington Hills, MI.

8 Ahmad, S.H., Zia, P., Yu, T., and Xie, Y. (1993). "Punching Shear Tests of Slabs Reinforced with 3-D

9 Carbon Fiber Fabric." ACI Concrete International, 16 (6), 36–41.

10 Arafa, A., Farghaly, A., Ahmed, E., Benmokrane, B. (2016). "Testing of GFRP-RC Panels with

11 UHPRFC Joints of Nipigon River Cable Stayed Bridge in Northwest Ontario, Canada." ASCE

12 Journal of Bridge Engineering, Vol.21, No. 11, p. 05016006-1-05016006-14.

13 ASTM C150/C150M-18. (2018). "Specification for Portland Cement." ASTM International, West

14 Conshohocken, PA.

15 ASTM C330/C330M-17a (2017). "Standard Specification for Lightweight Aggregates for Structural

16 Concrete." ASTM International, West Conshohocken, PA.

17 ASTM C567/C567M-19. (2019). "Standard Test Method for Determining Density of Structural

18 Lightweight Concrete." ASTM International, West Conshohocken, PA.

19 ASTM D7205/D7205M-16. (2016). "Standard Test Method for Tensile Properties of Fiber-Reinforced

20 Polymer Matrix Composite Bars." American Society for Testing and Materials, Conshohocken,

21 USA, 13 pp.

22 Benmokrane, B., El-Salakawy, E., El-Gamal, S., and Goulet, S. (2007). "Construction and Testing of

23 an Innovative Concrete Bridge Deck Totally Reinforced with Glass FRP Bars: Val-Alain Bridge on

24 Highway 20 East." ASCE Journal of Bridge Engineering, 12 (5), 632-645.

22
 1 Benmokrane, B., Mousa, S., Mohamed, H.M., El-Safty, A., and Nolan, S. (2021). "Design,

2 Construction, Testing, and Behavior of Driven Precast Concrete Piles Reinforced with GFRP Bars

3 and Spirals." ASCE Journal of Bridge Engineering, Vol. 26, No. 8, pp.

4 Benmokrane, B., Mohamed, H.M., Mohamed, K., and Mousa, S. (2020). "Recent Canadian

5 Developments Related to Unconventional Reinforcing for Concrete Structures, Design Codes, and

6 Applications in Buildings and Bridges. " ACI Special Publication, SP-346, American Concrete

7 Institute, pp. 93-112.

8 Bouguerra, K., Ahmed, E., El-Gamal, S., and Benmokrane, B. (2011). "Testing of Full-Scale Concrete

9 Bridge Deck Slabs Reinforced with Fiber-Reinforced Polymer (FRP) Bars." Construction and

10 Building Materials, 25 (10), 3956–3965.

11 Canadian Standards Association (2010). "Canadian Highway Bridge Design Code (CAN/CSA S6–

12 10)." Rexdale, ON, Canada.

13 Canadian Standards Association (2019). "Canadian Highway Bridge Design Code (CAN/CSA S6–

14 19)." Rexdale, ON, Canada.

15 Canadian Standards Association. (2002). "Design and construction of building components with fiber-

16 reinforced polymers." (CAN/CSA S806-02), Rexdale, Canada.

17 Canadian Standards Association. (2012). "Design and construction of building components with fiber-

18 reinforced polymers." (CAN/CSA S806-12), Rexdale, Canada.

19 Canadian Standards Association. (2017). "Design and construction of building components with fiber-

20 reinforced polymers." (CAN/CSA S806-12 R2017), Rexdale, Canada.

21 Canadian Standards Association. (2019). "Specification for fiber-reinforced polymers." (CAN/CSA

22 S807-19), Toronto, Canada.

23
 1 Elgabbas, F., Ahmed, E. A., and Benmokrane, B. (2016). " Experimental Testing of Concrete Bridge-

2 Deck Slabs Reinforced with Basalt-FRP Reinforcing Bars under Concentrated Loads." ASCE

3 Journal of Bridge Engineering, 21(7): 04016029.

4 El-Gamal, S., El-Salakawy, E., and Benmokrane, B. (2007). "Influence of Reinforcement on the

5 Behavior of Concrete Bridge Deck Slabs Reinforced with FRP Bars." ASCE Journal of Composites

6 of Construction, 11 (5), 449–458.

7 El-Gamal, S., El-Salakawy, E.F., and Benmokrane, B. (2005). "Behavior of Concrete Bridge Deck

8 Slabs Reinforced with FRP Bars under Concentrated Loads." ACI Structural Journal, 102 (5), 727–

9 735.

10 Gooranorimi, O., Gremel, D., Myers, J.J., Nanni, A. (2019). “Long-term Durability of GFRP Internal

11 Reinforcement in Concrete Structures,” American Concrete Institute (ACI) Special Publication 331,

12 Symposium Volume-Durability of Concrete Structures Incorporating Conventional and Advanced

13 Materials, Farmington Hills, MI, SP-331-5, March 2019, pp. 69-79.

14 Harmon, K. S. (2007). "Engineering Properties of Structural Lightweight Concrete." Carolina Stalite

15 Company.

16 Hewitt, B.E., and Batchelor, B. (1975). "Punching Shear Strength of Restrained Slabs." ASCE Journal

17 of the Structural Division, 101 (9), 1837–1853.

18 Hussein, A., and El–Salakawy, E., (2018). "Punching Shear Behavior of Glass Fiber–Reinforced

19 Polymer–Reinforced Concrete Slab-Column Interior Connections." ACI Structural Journal, 115 (4),

20 1075-1088.

21 Japan Society of Civil Engineering (JSCE). (1997). “Recommendation for Design and Construction of

22 Concrete Structures using Continuous Fiber Reinforcing Materials.” Concrete Engineering Series.

23 23, Tokyo, Japan.

24
 1 Liu, R., and Pantelides, C. P. (2013). "Shear Strength of GFRP Reinforced Precast Lightweight

2 Concrete Panels." Construction and Building Materials, 48, 51-58.

3 Mehany, S., Mohamed, H. M., and Benmokrane, B., (2021). “Contribution of Lightweight Self-

4 Consolidated Concrete (LWSCC) to Shear Strength of Beams Reinforced with Basalt FRP Bars.”

5 Engineering Structures, 231, 111758.

6 Mousa, S., Mohamed, H. M., and Benmokrane, B., (2018). “Flexural Strength and Design Analysis of

7 Circular reinforced concrete Members with Glass Fiber-Reinforced Polymer Bars and Spirals.” ACI

8 Structural Journal, 115 (5), 1353-1364.

9 Pantelides, C. P., Besser, B. T., and Liu, R. (2012a). "One-Way Shear Behavior of Lightweight

10 Concrete Panels Reinforced with GFRP Bars." Composites for Construction, 16 (1), 2-9.

11 Pantelides, C. P., Liu, R., and Reaveley, L. D. (2012b). "Lightweight Concrete Precast Bridge Deck

12 Panels Reinforced with Glass Fiber-Reinforced Polymer Bars." ACI Structural Journal, 109 (6),

13 879-888.

14 Perdikaris, P. C., and Beim, S. (1988). "RC Bridge Decks under Pulsating and Moving Loads." ASCE

15 Journal of Structural Engineering, 114 (3), 591–607.

16 Shirley-Smith, H., Billington, P. N., and Billington, D. P., (2020). “Bridge.” Encyclopedia Britannica.

17 https://www.britannica.com/technology/bridge-engineering. Accessed 21 April 2021.

18 Youssef, M. H., Ahmed, E. A., Benmokrane, B., (2019). “Structural Behavior of GFRP-RC Bridge

19 Deck Slabs Connected with UHPFRC Joints Under Flexure and Shear.” ASCE Journal of Bridge

20 Engineering, 24 (9), 04019092.

21 Appendix. Design of Slab G1-150 (CSA S6-19 2019)

22 This section presents how specimen G1-150 was designed according to CSA S6-19 (2019) provisions,

23 mentioned here where applicable. Figure A1 shows the cross section of G1-150.

25
 1

2 Figure A1 - Cross section of specimen G1-150 (Note: 1 mm = 0.0394 in.).

3 According to the information presented in Table 2, GFRP bars with an ultimate capacity (fFRP.u) of

4 1,316 MPa and tensile modulus of elasticity (EFRP) of 61.8 GPa were used as reinforcement in the

5 specimen.

6 Deck-slab thickness, h = 200 mm

7 Deck-slab effective span, Se = 2,000 mm

8 Bottom concrete cover, CC = 25 mm

bh3 1000  2003

9 Ig    666.67 106 mm4
12 12

10 The Empirical Design Method (Clause 16.8.8.1)

11 According to the empirical design method, the amount of reinforcement required for the bottom

12 assembly in the transverse direction is calculated as follows:

500d s 500 167

13 AFRP min   1000  1351mm2 / m
EFRP 61800

14 where

16
15 ds = h – bottom concrete cover – bar diameter/2 = 200 – 25 – =167 mm
2

16 Using 16 mm GFRP bars @ 150 mm for the bottom transverse direction has been determined.

17 The Flexural Design Method

26
 1 Designing the bridge-deck slabs based on the flexural method considers the service and ultimate limit

2 states. The truck wheel load (87.5 kN CL-625 truck) and the self-weight of the slab and pavement are

3 used to calculate the design moments. The slab self-weight was not considered in the experimental

4 results.

5 Concrete

6 f c'  40 MPa

7  
Ec  3000 f c'  6900   c / 2300 
1.5
 19.52 MPa (Clause 8.4.1.5.3)

8 c  0.75 (Table 8.1)

9 Reinforcement

10 The 16 mm GFRP bars @ 150 mm were used for the bottom transverse direction.

11 AFRP  1393 mm2/m

16
12 d = h – bottom concrete cover – bar diameter/2 = 200 – 25 – =167 mm
2

AFRP 1393
13 f  100  100  0.834%
1000  d 1000 167

14 Specific tensile strength, f FRP.u = 1316 MPa

15 Bond-dependent coefficient, k b = 0.8 (Clause 16.8.2.3, sand-coated bars)

16 FRP  0.65 (Table 16.2, GFRP reinforcement in concrete)

17 Stress at service-load level, f SLS  0.25 f u' (Clause 16.8.3, GFRP bars)

18 Cracking Moment
27
 1 The cracking strength, fcr, is specified as follows (Clause 8.4.1.8.1):

2 a) 0.4 f c' for normal-density concrete

3 b) 0.34 f c' for concrete with semilow density

4 c) 0.30 f c' for low-density concrete

5 Linear interpolation based on the fraction of natural sand in the mix:

6  Lightweight sand in the mix design: 488 kg

7  Normal sand in the mix design: 381 kg

8 → f cr  0.32 f c'  0.32 40  2.02 MPa

2 f cr I g 2  2.02  666.67 106

9 M cr   106  13.49 kN ‧ m/m
h 200

10 Load and Transverse Moment attributable to Uniform Load

11 Slab self-weight = 0.2 × 19.06 = 3.81 kN/m2

12 Pavement self-weight = 0.065 × 24.0 = 1.56 kN/m2

13 Service dead load, Wd  SLS = 3.81 + 1.56 = 5.37 kN.m2

14 Factored dead load, Wd ULS = 1.2 × 3.81 + 1.5 × 1.56 = 6.91 kN.m2

15 Service moment attributable to dead load, M d  SLS  0.071Wd  SLS Se2

16 = 0.071 × 5.37 × 2.02 = 1.525 kN. m/m

17 Factored moment attributable to dead load, M d ULS  0.071Wd ULS Se2

18 = 0.071 × 6.91 × 2.02 = 1.962 kN m/m

28
 1 Load and Transverse Moment corresponding to Wheel Load

2 The maximum transverse moment, M y 

 Se  0.6  P  22.75 kN ‧ m/m (Clause 5.7.1.2)
10

3 Wheel load service moment, M y  SLS = 1.4 × 22.75 × 0.9 = 28.665 kN m/m

4 Wheel load Factored moment, M y ULS = 1.4 × 22.75 × 1.7 = 54.145 kN m/m

5 where 1.4 is the impact factor, and 0.9 and 1.7 are the service and ultimate load combination factors,

6 respectively (Table 3.1).

7 Transverse Design Moment (Total Loads)

8 Total service moments, M SLS = 1.525 + 28.665 = 30.19 kN m/m

9 Total factored moment, M ULS = 1.962 + 54.145 = 56.107 kN m/m

10 Check for Crack Width and Stress Limit under Service Load

Ef 61800
11 nf    3.17
Ec 19519

12 k  2  f n f  (  f n f ) 2   f n f  0.22

M SLS
13 fs   140 MPa 0.25 f u*  0.25 1316  329 MPa (OK)
AFRP d (1  k / 3)

14 h2  h  kd  200  0.22 167  163.26 mm

15 h1  d  kd  167  0.22 167  130.26 mm

16 dc  h  d  200  167  33 mm

29
 f s h2 140 163.26
1 wcr  2 kb dc2  ( s / 2) 2  2    0.8 332  (150 / 2) 2  0.3
EFRP h1 61800 130.26

2  0.30 mm  0.50 mm (OK)

3 Check for Stress under Ultimate Load

4 1  0.85  0.0015 f c'  0.79  0.67 (Clause 8.8.3)

5 1  0.97  0.0025 f c'  0.87  0.67 (Clause 8.8.3)

6 FRP AFRP f FRP  1c f c' ab

7 0.65 1393  (61800   FRP )  0.79  0.75  40  (0.87  c) 1000

8 c  2713.8 FRP

 c   d   167 
9    
 0.0035   0.0035   FRP   0.0035   FRP 

10 Solving equations (45) and (46)

11  FRP  0.01303

12 c  35.36 mm

13 a  1c  30.76 mm

14 f f  EFRP FRP  805.25 MPa fu*  1316 MPa (Overreinforced)

15 M r  1c f c' ab(d  a / 2)  110.5 kN.m

16  1.5M cr  1.5 13.49  20.24 kN.m (OK) (Clause 16.8.2.2)

30
1  1.5M u  1.5  56.107  84.16 kN.m (OK) (Clause 16.8.2.2)

2 Use 16-mm GFRP @ 150 mm.

31
1 List of Tables

2 Table 1 - Mixture proportions for the LWSCC.

3 Table 2 - Mechanical properties of the reinforcing bars.

4 Table 3 - Reinforcement details of the tested bridge-deck slabs.

5 Table 4 - Test results and comparison to theoretical predictions of the punching-shear capacities.

32
1 Table 1 - Mixture proportions for the LWSCC

Cement type TerC3 (kg/m3) 543

Water/cement ratio 0.33

Lightweight coarse aggregate (kg/m3) 369

Lightweight sand (kg/m3) 488

Natural sand (kg/m3) 381

Air entrainment (mL/100 kg) 70

Superplasticizer (L/m3) 3.51

3 Table 2 - Mechanical properties of the reinforcing bars

Bar db Af a Ef ffu εfu

Reinforcement type Bar type Surface configuration
size (mm) (mm2) (GPa) (MPa) (%)

Type I #5 15.9 199 61.8 1316 2.13 Sand-coated

GFRP
Type II #5 15.9 199 64.1 1528 2.4 Helically Wrapped

Steel b - M15 16 200 200 fyc = 460 εf c = 0.23 Ribbed

4 a
Nominal cross-sectional area.
5 b
The manufacturer provided the tensile properties of the steel bars.
6 c
fy and 𝛆y are yield strength and strain of steel bars, respectively.
7 Note: Properties calculated based on the nominal cross-sectional area, (1 mm = 0.0394 in; 1 mm2 = 0.00155 inch2;
8 1 MPa = 0.145 ksi).

33
 9 Table 3 - Reinforcement details of the tested bridge-deck slabs

Bottom reinforcement Top reinforcement

Slab IDa Transverse direction Longitudinal direction Transverse direction Longitudinal direction

RFTb ρ (%) ρ×E (MPa) RFTb ρ (%) ρ×E (MPa) RFTb ρ (%) ρ×E (MPa) RFTb ρ (%) ρ×E (MPa)

G1-150 16 @150 0.83 513 16 @200 0.62 383 16 @200 0.6 371 16 @200 0.62 383

G1-100 16 @100 1.15 711 16 @200 0.62 383 16 @200 0.6 371 16 @200 0.62 383

G1-300 16 @300 0.44 272 16 @300 0.43 266 16 @300 0.44 272 16 @300 0.43 266

G2-300c 16 @300 0.44 282 16 @300 0.43 276 16 @300 0.44 282 16 @300 0.43 276

S1-300 16 @300 0.44 880 16 @300 0.43 860 16 @300 0.44 880 16 @300 0.43 860

10 a Reinforcement type (GFRP Type I and Type II and Steel) followed by reinforcement spacing (100, 150, or 300 mm).
11 b Bar diameter and spacing in mm (diameter in mm @ spacing in mm).
12 c Helically wrapped GFRP bars (GFRP Type II).
13 Note: 1 MPa = 0.145 ksi, 1 mm = 0.0394 in.
14

34
15 Table 4 - Test results and comparison to theoretical predictions of the punching-shear capacities.

Deflection (mm) FRP strain (µɛ) Concrete strain (µɛ) Crack width (mm) Pmax. exp/Pmax. pred
f'c σt Pcr Pmax CSA ACI ACI AASHTO AASHTO
ID P P Pmax Pserv Pfact Pmax Pserv Pfact Pmax Pserv Pfact Pmax
(MPa) (MPa) (kN) (kN) serv fact (2012) (2015)1 (2015)2 (2018)3 (2018)4
G1-150 47.3 3.96 106.9 541.4 2.06 6.34 24.12 1,239 3,262 6,719 -121 -236 -1,305 0.12 0.27 0.82 1.18 1.53 1.37 1.40 1.40
G1-100 47.3 3.96 115.2 525.6 1.75 4.51 19.01 422 1,600 4,608 -181 -503 -1,147 --- 0.22 0.61 1.01 1.26 1.14 1.15 1.15
G1-300 42.4 3.25 100.0 442.8 2.24 8.2 25.81 90 2,939 7,051 -174 -698 -1,119 0.20 0.60 1.68 1.22 1.64 1.48 1.78 1.59
G2-300 46.1 3.64 114.5 421.1 1.13 6.88 24.47 453 2,159 7,077 -123 -273 -944 --- 0.49 1.89 1.11 1.53 1.37 1.47 1.41
S1-300 46.1 3.64 99.2 538.7 1.27 4.18 25.72 107 581 3,892 -71 -151 -1,206 0.10 0.30 2.01 ---- ---- ---- ---- ----

Average 45.8 3.69 107.2 493.9 1.69 6.02 23.8 462 2108 5869 -134 -372 -1144 0.14 0.38 1.40 1.13 1.49 1.34 1.45 1.39

SD 2.0 0.3 7.6 57.4 0.5 1.7 2.8 466.3 1074 1506 44.9 224 132.7 0.1 0.2 0.6 0.1 0.2 0.1 0.3 0.2

CV (%) 4.4 8.0 7.1 11.6 28.6 28 11.7 101 51 26 -34 -60 -11.6 37.8 43.0 45.8 8.1 10.9 10.7 17.9 13.0
16 Notes: σt, Pcr and Pmax are the spitting tensile strength, cracking, and failure loads, respectively; service (Pserv) and factored (Pfact) design loads are calculated as follows:
17 Pserv  1.4  0.9  87.5  110.25 kN (24.8 kips) and Pfact  1.4 1.7  87.5  208.25 kN (46.8 kips), respectively (CL625 truck) (CSA S6-19 2019).
18 1
Considering the concrete density reduction factor provided in AASHTO (2020).
19 2
Without considering the concrete density reduction factor provided in AASHTO (2020).
20 Note: 1 kN = 0.225 kips, 1 MPa = 0.145 ksi, 1 mm = 0.0394 in.
21

35
22 List of Figures

23 Figure 1 – (a) Lightweight aggregates from Northeast Solite; (b) GFRP and reinforcing bars.

24 Figure 2 – Geometry of the tested deck slabs and reinforcing details (Note: 1 mm = 0.0394

25 in.).

26 Figure 3 – Construction of the bridge-deck slabs.

27 Figure 4 – Typical instrumentation of the test slabs: (a) strain gauges on the top (T) and

28 bottom (B) reinforcing bars; and (b) LVDTs and strain gauges on the top concrete surface

29 (Note: 1 mm = 0.0394 in.).

30 Figure 5 – Overview of the test setup.

31 Figure 6 – Crack patterns at failure: (a) G1-100; (b) G1-150; (c) G1-300; (d) G2-300; and

32 (e) S1-300.

33 Figure 7 – Load-to-average crack-width curves (Note: 1 kN = 0.225 kips, 1 mm = 0.0394

34 in.).

35 Figure 8 – Strains in the bottom reinforcement and surface of the concrete (Note: 1 kN =

36 0.225 kips).

37 Figure 9 – Distribution of the reinforcement strains at the middle section of the slabs (m)

38 (Note: 1 m = 3.28 ft.).

39 Figure 10 – Effect of test parameters on the load–maximum deflection curves: (a)

40 reinforcement ratio; (b) surface conditioning; and (c) reinforcement type (Note: 1 kN = 0.225

41 kips, 1 mm = 0.0394 in.).

42

36
43

44 Figure 1 – (a) Lightweight aggregates from Northeast Solite; (b) GFRP reinforcing bars.

45

46 Figure 2 - Geometry of the tested deck slabs and reinforcing details (Note: 1 mm = 0.0394
47 in.).

37
48 .

49 Figure 3 - Construction of the bridge-deck slabs: (a) formwork; (b) meshes; (c) prepared
50 formwork for casting; (d) mixing LW aggregates with water (pre-wetting) 3 days before
51 casting; (e) casting; and (f) demolding.

52

53

54

55 Figure 4 - Typical instrumentation of the test slabs: (a) strain gauges on the top (T) and
56 bottom (B) reinforcing bars; and (b) LVDTs and strain gauges on the top concrete surface
57 (Note: 1 mm = 0.0394 in.).

58

38
59
60

61

62

63

64
65
66 Figure 5 - Overview of the test setup.

39
67

68 Figure 6 - Crack patterns at failure: (a) G1-100; (b) G1-150; (c) G1-300; (d) G2-300; and
69 (e) S1-300

40
70

71 Figure 7 - Load–average crack-width curves (Note: 1 kN = 0.225 kips, 1 mm = 0.0394 in.).

72

73 Figure 8 - Strains on the bottom reinforcement and surface of the concrete (Note: 1 kN =
74 0.225 kips).

75

76 Figure 9 - Distribution of the reinforcement strains at the middle section of the slabs (m)
77 (Note: 1 m = 3.28 ft).

41
78
79 (a)

80
81 (b)

82
83 (c)

84 Figure 10 - The effect of test parameters on the load–maximum deflection curves: (a)
85 reinforcement ratio; (b) surface conditioning; and (c) reinforcement type (Note: 1 kN =
86 0.225 kips, 1 mm = 0.0394 in.).

42