The Response of Reinforced Concrete Composite Beams Reinforced with Pultruded GFRP to Repeated Loads

T his paper investigates the experimental response of composite reinforced concrete with GFRP and steel I-sections under limited cycles of repeated load. The practical work included testing four beams. A reference beam, two composite beams with pultruded GFRP I-sections, and a composite beam with a steel I-beam were subjected to repeated loading. The repeated loading test started by loading gradually up to a maximum of 75% of the ultimate static failure load for five loading and unloading cycles. After that, the specimens were reloaded gradually until failure. All test specimens were tested under a three-point load. Experimental results showed that the ductility index increased for the composite beams relative to the reference specimen by 156.2% for a composite beam with GFRP with shear connectors, 148.6% for composite beams with GFRP without connectors, and 96% for the composite beam with a steel I-section.


INTRODUCTION
For recent research, the Glass Fiber Reinforced Polymer (GFRP) profiles were frequently utilized in composite beams because of their higher tensile performance, and many loading types have been done on this material. (Ibrahim et al. 2022) tested eight composite beams with pultruded GFRP under static and impact loading. Also, (Ali and Allawi, 2021) and (Allawi and Ali 2020) tested hybrid beams with GFRP under static and impact loading.
In recent decades, special attention has been paid to structures under repeated loads, such as railways and highway bridges. Limited research is related to composite concrete structures' behavior under repeated loads. A particular case of cyclic loading is repeated or cyclic -half loading, in which the loading is applied in unidirectional cycles. (Sivagamasundari and Kumaran, 2008) evaluated the behavior of one-way slabs reinforced with GFRP bars and traditional reinforcement under cyclic loading with variable and constant amplitude fatigue loads. Eleven specimens have steel reinforcement, and 28 have GFRP. Seventeen of the 39 specimens were examined under variable-amplitude fatigue loading. Sixteen were tested under static load. The specimen is 2400mm long, 600mm wide, and 100mm to 120mm thick. Two concrete grades (20 MPa and 30 MPa) and three reinforcement percentages (0.65%, 0.82%, and 1.15%) were used. All slabs failed as flexural. At failure load, GFRP-reinforced slabs crushed concrete and fractured GFRP. (10-17%) of the slabs' flexural strength increased. by increasing compressive strength by 50% for the same slabs and slab thickness by 20% for the static load specimen. Increasing the compressive strength of concrete by 50% increased fatigue performance by 33% for the same slabs; GFRP reinforcement caused less damage than steel reinforcement. Also, slabs subjected to constant amplitude fatigue loading had a larger residual and final deflection and fracture width than slabs subjected to variable amplitude fatigue loading.
(Allawi and Jabir, 2016) tested nine RC one-way slabs with and without lacing reinforcement. The tests were designed to study the effect of the lacing reinforcement on the flexural response of one-way slabs under repeated load. The loading was applied as (5 cycles) loading-unloading to 80% of the ultimate load of the control specimen, then loaded up to the failure. Also, (Mohammed and Fawzi, 2016) tested nine burned RC beams subjected to the effect of repeated loading (loading-unloading) for five cycles and then up to failure. Furthermore, (Allawi, 2017) studied composite prestressed concrete girders with an external post-tensioned technique under static and repeated loading. In this research, the beams were subjected to five loading and unloading cycles up to 75% of the ultimate load. After that, the girders were reloaded gradually until failure. Also (Hasan and Allawi, 2019) tested eighteen simply supported reinforced concrete beams under static and fatigue loads with displacement control technique, which were exposed to high frequency (10 Hz) by fixing the fatigue load in each cycle.
(Fathuldeen and Qissab, 2019) studied the repeated loading of RC beams strengthened with NSM CFRP strips. Fifteen NSM-CFRP beams were loaded monotonically and repeatedly. Three beams were left unreinforced as references, and the others were strengthened with NSM-CFRP strips. Each group has two beams tested under monotonic loads as a control for those tested under repeated loads. For all specimens, the loading cycles were applied until failure. The test results showed that NSM-CFRP strips boosted beams' flexural strength and stiffness. The load-carrying capacity was increased from 1.47 to 4.49 times. After repeated loading, the overall area of CFRPs increased by 1.02 times the control value. Increasing the total area of CFRP strips reduced the ductility factor to 0.71, while cumulative energy absorption increased by 1.22 times for the stronger reference specimens tested under repeated loads. (Khalaf and Al-Ahmed, 2021) used repeated loading to investigate the behavior of the existence of large openings in reinforced concrete continuous deep beams. The range of the repeated loading varied between 30% and 70% of the ultimate load of the beam subjected to static load.
(Zhu et al., 2022) studied high-strength concrete beams reinforced with BFRP bars and steel fibers under four-point cyclic loading. Five 150×300×2100mm concrete beams were built and tested. Four BFRP-reinforced concrete beams with different reinforcement ratios (ρs), 0.56%, 0.77%, 1.15%, and 1.65%, and one conventional steel-reinforced concrete beam were tested. Cracking, failure mechanisms, load deflection, residual deformation, and stiffness degradation were examined. An increase in ρs restrained fracture widths, deflections, and residual deformation while increasing beams' flexural bearing capacity. The bearing capacity was reduced by 10% in the third cycle compared to the first displacement cycle. Stiffness degraded quickly before failing. Higher ρs beams have higher residual stiffnesses.
The repeated load was applied by incremental loads gradually up to (75%) of the ultimate load level of the control specimen and then released the load gradually to zero with (5 cycles) of loading-unloading.

Specimens Configuration
Four composite reinforced concrete specimens which were cast in different configurations. The overall length of the specimens was 3000mm with a support location of 125mm from each end of the beam giving a clear span of 2750mm. All specimens have the same cross-sectional dimensions, the beam width is (200 mm), and the total height of the beam is (300 mm) (see Figure 1). The arrangement of reinforcements consists of 2 ф16 mm rebars as longitudinal bottom reinforcement and 2 ф10 mm rebars as top reinforcement. Stirrups of ф10 mm spaced every 125 mm c/c were used as shear reinforcement.
The specimen NR-R was used as a reference specimen without additional reinforcement. The specimen CG-R was reinforced with pultruded GFRP I-beams positioned at the centroid of the cross-section (see Figure 1-b). CGC-R was a composite specimen of GFRP with shear connectors provided in the top flange of the GFRP I-section in the specimen. The last specimen CS-R, was reinforced with a steel I-section positioned at the center of the cross-section (see Figure 1-d). The diameter of these connectors was 12 mm, with a height of 70 mm and a spacing of 375 mm. Shear connectors were stiffened with washers and nuts after being inserted through drilled holes on both sides of the top flange of the GFRP beam (see

Materials Preparation
Normal-weight concrete with a cylindrical compressive strength of about 25 MPa was produced for casting the test specimens. The yield stress and ultimate strength of steel reinforcements for a bar diameter of 16 mm were 520.73 MPa and 687.07 MPa, respectively; for a bar diameter of 10 mm, were 407.7 MPa and 465.63 MPa, and 375.9 MPa and 479.63 MPa for steel plate that was used to fabricate the steel I-section. The GFRP compressive and tensile strengths were 326.14 MPa and 347.5 MPa, respectively.

Test Setup and Instrumentations
The experimental program consisted of four composite reinforced concrete beams subjected to a non-reversed repeated loading regime depending on the ultimate load of specimens subjected to static load. The applied load was performed using an electric hydraulic jack with a 1000 kN capacity controlled using a 1000 kN load cell( see Figure 2). The repeated loading test sequence was started from zero value up to a certain cracking load, and then specimens were unloaded. After that, they were reloaded gradually again with a 5 kN load increment, up to a maximum of 75% of the ultimate static failure load. Then the load was gradually released to zero for five loading and unloading cycles. After that, the specimens were reloaded gradually until failure. In all specimens, the test was terminated when defection increased dramatically under an approximately constant load.

Test Results
The simply supported four composite beams were loaded under a three-point load. The repeated test program included testing reinforced specimens loaded with a concentrated repeated load of five cycles. At the beginning of the test, each specimen was loaded with a monotonic concentrated load till a certain cracking load was reached. The specimens were then unloaded; after that, they reloaded gradually again with a 5 kN load increment up to a maximum of (75%) of the ultimate static failure load which was tested by (Ibrahim et al,  2022). Then loading was released gradually to zero with five loading and unloading cycles, as shown in Figure 03. After that, the specimens were reloaded gradually until failure. In all specimens, the test was terminated when deflection increased dramatically under an approximately constant load. The test results were divided into five parts:

Load -Deformation Behavior
The load-deflection curves for the beam specimens tested under the effect of repeated loads are shown in Error! Reference source not found. and Table 1 Summary of the repeated loading test results.

Cracks Propagation and Failure Modes
The first flexural crack appears at the middle third of the beams whenever the tensile stresses exceed the modulus of rupture of concrete, this crack occurred at the load range of (15.86 % to 27.22 %) from the ultimate load capacity of the repeated-tested specimens, and this crack develops slowly across the width of the beam as shown in Figure 5-8. The crack pattern of the samples applying to the repeatedly loaded was almost the same as in the samples tested under static load. However, more cracks appeared when the number of loading cycles increased, and their width grew. It was shown in Table 2 that the ultimate load of the composite specimens CG-R, CGC-R, and CS-R increases by about (77.82 %, 81.07 %, and 148.15 %) relative to the NR-R specimen. Table 2 shows cracking and ultimate load for repeated load specimens. As expected and explained before, the ultimate load capacity increases for the composite specimens CG-R, CGC-R, and CS-R, respectively, concerning the specimen NR-R.
For the NR-R specimen, after loading and unloading five cycles up to the ultimate loads, specimens showed the flexural failure mode by yielding steel reinforcement and compression concrete crushing and propagation of flexural cracks. After that, a sudden fracture in tensile steel reinforcement happened, leading to the collapse of the specimen, as shown in Figure 5.
For composite specimens, CG-R and CGC-R flexural cracks gradually spread in the midspan region in the first cycle after reaching the cracking load. No more cracks appeared for the last four cycles, but the previous cracks began to elongate. After the five cycles, concrete crushing started at a yield load; the number of cracks increased with loading increment; wider cracks were developed. The deflection began to overgrow when the applied load reached 127.78 kN and 130.12 kN for specimens CG-R and CGC-R, respectively. After applying the ultimate loads, the compression steel rebars were bent over. The concrete cover was spalled, accompanied by a loud noise produced by the initial interlaminar failure and rupture of the GFRP profile, as illustrated in Fig. 6 and Fig. 7. Finally, the testing was stopped due to slipping and crushing in the GFRP profile. The applied loads began to drop gradually. The web of the GFRP beam crashed, resulting in longitudinal shear failure.
At the end of the first cycle for the composite specimen with the steel I-section CS-R, flexural cracks grew in the middle of the span. For the four cycles, shear cracks appeared: no more flexural cracks appeared, but the previous cracks began to elongate. After the five cycles, concrete crushing started at a yield load of 135 kN. At the maximum load of 178.32 kN, the number of cracks increased with loading increment; wider cracks were developed. The test was terminated due to steel yielding followed by buckling and twisting in the steel I-section inside the concrete. This caused radial cracks at the end flange on the beam sides and crushing at the end supports' position, as shown in Figure 8.

Residual Deflection Response
In the loading and unloading process, the load-deflection curve's ascending and descending portions take different patterns; the difference between the two is commonly referred to as residual deflection. The results of the tests showed that, for all the specimens, the amount of deflection at the same point and load increment increased as the number of loading cycles increased. This means the specimen did not return to its original position when the load was released, and a residual deflection was observed. As shown in Table 3, the test results for the composite specimens CG-R, CGC-R, and CS-R showed that the residual deflection was 12.52 %, 15.29 %, and 21.2 % less than that of the reference beam NR-R. This could be because of the benefits of high stiffness of specimens, which led to increased ductility of concrete and decreased permanent deformations. It was noticed that the most residual deflection happened in the first cycle, while Figure 9 shows that the differences in residual deflection between the last four cycles were not as significant as the difference between the first and second cycles.

Load-strain relation
The same locations of the strain gauges for the specimens tested under static load are adopted to measure and represent the load-strain relations of the specimens tested under repeated load. Figure 10-Figure 13 illustrates the maximum compressive strain of the concrete was 0.005 mm/mm, while the GFRP within the elastic range has a tensile strain of 0.006 at the ultimate load level of the specimens CG-R and CGC-R. And for the CS-R specimen, the I-steel profile's maximum compressive and tensile strain was about 0.025 mm/mm.
The failure load of the composite specimens with GFRP CG-R and CGC-R was determined from the maximum strain value recorded in the web of the GFRP profile.
Regarding the load-strain relation Figure 10-Figure 13, it was noticed that the most significant effect of the repeated load was in the first cycle. In contrast, it is clear from Figure  10 that the differences in the strains between the last four cycles were relatively small in comparison with the difference between the first and the second cycles.

2.4.5.Ductility
Ductility is a requirement for structural design in most design codes. It is defined in RC structures as the ultimate deformation relative to the yield point deformation, which is usually caused by steel reinforcement. Typical ductility definitions do not apply to GFRPreinforced structures because of the linear strain-stress relationship of GFRP. Several approaches have been presented for determining the ductility index of GFRP-reinforced structures. The most common module was proposed by (Naaman AE, 1995), as illustrated in Figure 14. It has been used in some previous studies ( Where: Etot : the total energy, which was calculated by the area under the load-midspan deflection curve up to the failure load.
Eel : the elastic energy, which was calculated as the triangle area produced at the failure load Pfail by the line with the weighted average slope of the two starting straight lines S1 and S2 of the load-deflection relationship.
P1 and P2 : are the loads at the end of the initial two lines, respectively, as illustrated in Figure  14.
Energy ductility is also defined as the ability to absorb inelastic energy without compromising load capacity. Higher inelastic energy absorption equates to higher ductility of the same system. The GFRP and the steel I-section improve the system's ductility significantly compared to the reference specimens. The ductility index depends on specimens' elastic and total energy amounts are shown in Table 4. It was observed from Table 4 that for all the composite beams, the ductility factor increased as compared to the reference beam. The greatest increases were recorded at the composite beam CGC-R, which was approximately 156%, while specimens CG-R and CS-R were increased by 148.6% and 96%, respectively.