Water absorption
The water absorption of GFRP rebar can vary depending on several factors, including the specific type and composition of the GFRP material, the manufacturing process, and the environmental conditions it is exposed to. In general, GFRP rebar absorbs water through capillary action and diffusion. In this study, Fig. 1 shows that the water absorption of 10-Yrs specimens was continued to increase gradually over the time and demonstrated a non-Fickian diffusion process [18]. The non-Fickian behaviour was due to gradual capillary transport of water into the polymer matrix and fibre/matrix interfaces during the immersion period at elevated temperature [9]. It is worth noting that the water absorption of GFRP composites does not always reach equilibrium even after a long immersion period. In this work, the average water absorption of 10-Yrs specimens was determined to be 0.215% after being immersed for 13,008 h at 60 oC. This water absorption value was lower compared to the other reported results in the literature regarding GFRP rebar [3, 19, 20]. The lower water absorption observed in the 10-Yrs rebar can be attributed due to its lower porosity content, as mentioned in another report [21]. The reduced porosity likely limited the amount of water that could penetrated and be absorbed by the material, resulting in the lower absorption observed in this study.
Glass transition temperature
Glass transition is an important property of the polymer matrix, as it affects the durability and performance of the GFRP rebar in different hostile environment. The polymer matrix serves to protect the load-carrying glass fibre component from degradation, particularly in corrosive conditions. In the DSC traces shown in Fig. 2(a), the heat flow curves exhibit a phase change step, which is referred to as the glass transition. This glass transition represented the shift of the polymer matrix from an elastic behaviour to a viscoelastic behaviour within a temperature range of 90 to 115°C. Furthermore, the DSC heat flow curves also provided information about the curing process of the rebars. The absence of any residual exothermic cure reaction during the first heat scans indicates that the rebars were well cured. This means that the polymer matrix had its desired level of cross-linking and bonding, contributing to the overall structural integrity of the GFRP rebar.
The purpose of evaluating the glass transition temperature of the first heating scan (Tg1) was to assess the thermal stability changes in the polymer matrix caused by conditioning. On the other hand, the glass transition temperature of the second heating scan (Tg2) was measured to determine the degradation of the polymer matrix resulting from conditioning. Figure 2 (b) presents the average Tg1 and Tg2 values of both the control and conditioned rebars. For the control rebar, Tg1 was found to be 102.6°C, which increased to 108.1°C after the second heating scan (Tg2). This increase in Tg of the control rebar was attributed to anti-plasticization, caused by the evaporation of inherent moisture and volatile processing additives during the elevated temperature thermal scanning.
After 10 years of conditioning in an alkaline solution, Tg1 of the rebar decreased to 97.18°C. The decrease of approximately 5.4°C in Tg was attributed to absorbed water that plasticized the polymer matrix during the conditioning period. In the plasticization process, the absorbed water molecules caused swelling of the polymer matrix, resulting in increased distance between polymer segments and enhanced polymer chain mobility or free volume. Consequently, the Tg of the polymer matrix decreased. Following the second heating scan, Tg of the conditioned rebar increased to 107.1°C, which approached the Tg2 value of the control rebar. This increase in Tg was again attributed to anti-plasticization due to the evaporation of absorbed water during the thermal scan. Therefore, conditioning the GFRP rebar in the alkaline solution caused a reversible chemical change in the polymer matrix.
Statistical analysis (Paired Sample t Test) confirmed that the difference between Tg2 of the control and Tg2 of the conditioned rebars was not statistically significant. Thus, based on the thermal analysis, it can be concluded that the polymer matrix of the GFRP rebar did not chemically degrade after being conditioned in the alkaline solution for 10 years. Other research studies [4, 22, 23] have also reported that the epoxy vinyl ester matrix of the rebar is less susceptible to hydrolysis or degradation reactions at ambient temperatures.
Transverse shear strength
GFRP rebars have various applications, including their use as dowels at joints in concrete pavements and for reinforcing concrete in direct shear [24]. The behaviour of the control rebar and 10-Yrs conditioned rebar under transverse shear load is discussed, the results are presented in Fig. 3 (a). It can be observed that the transverse shear load of the control rebar increased linearly up to 43% of the maximum shear breaking load, afterwards the shear phase changed slightly due to buckling of the glass fibres under compression. Once the maximum breaking load was reached, the recorded transverse shear load was then diminished irregularly and caused to split the test specimen into three pieces as shown in the photos in Fig. 3 (a). In contrast, the conditioned rebar exhibited changes in the shear phase at relatively low shear loads, specifically at 11 and 39% of the maximum breaking load. These observed differences were attributed to softening of the polymer matrix (indicated by a decrease in Tg) due to water absorption during the conditioning process. Figure 3 (b) presents the average transverse shear strength (TSS), which was found to be 178.5 MPa for the control rebar. After 10 years of conditioning, the TSS decreased by 2.75%. This slight reduction in TSS was also attributed to the softening of the matrix in the 10-year rebar. However, it is noteworthy that the TSS retention of the 10-Yrs rebar outperformed the reference works [22, 25, 26], indicating its favourable performance.
SEM fractography was utilised to investigate the causes and mechanisms of rebar failure following conditioning. In Fig. 4 (a), the fractured surface displayed fragmented fibres and matrix debris, which resulted from compressive shear load. The fibres were observed to be split into small pieces instead of being pulled out from the matrix while bearing the load; this was ascribed to strong bonding between the fibre and matrix. Figure 4 (b) provides evidence of buckling of fibre and shear fracture of the rebar due to transverse shear load. Moving on to Fig. 5 (a), the presence of single fractured fibre ends provided proof of "chop marks" caused by compression failure. Additionally, smooth, featureless brittle fracture surfaces on individual fibres indicated tension failure. In Fig. 5 (b), a mixed mode of tension and compression failures was observed on the fractured fibre end, resulting from micro-buckling. The well-defined anchoring region of the matrix onto the fibre once again suggested a strong adhesion between the fibre and the matrix. This strong interfacial bonding played a significant role in the superior TSS retention of the 10-Yrs rebar.
Short beam shear strength
The short beam shear strength (SBSS) results are presented in Fig. 6 (a). The average SBSS of the control rebar was determined to be 66.5 MPa, which showed a retention of 101.3% after 10 years of conditioning. However, statistical analysis using the paired sample t-test indicated that the difference in SBSS between the control and 10-Yrs rebars was insignificant at a 95% confidence level. The rebar specimens were failed in shear horizontally, initiating at the mid-span and propagating along the neutral axis, as depicted in Fig. 6 (b).
In a study by Ramanathan et al. [27], it was reported that the SBSS of GFRP rebars decreased by up to 30% after 18 years of service life due to the presence of voids in the fibre/matrix interface, fibre/matrix delamination and damage of matrix. In contrast, another research [28] found that SBSS remained unaffected when GFRP rebars were exposed to concrete under high level of sustained load for 10 years in natural weathering conditions. Additionally, in a separate report [29], SBSS of GFRP rebars extracted from seven bridges after 15 to 20 years of service was found to be 5 to 16% higher than the original rebar. This increase in SBSS was attributed to post-curing of the matrix over the time.
However, it is commonly reported that the SBSS of GFRP rebars decreases when conditioned in alkaline environment at elevated temperatures. For example, in a previous study [14], SBSS decreased by 8 to 20% after conditioning in an alkaline solution for 44 days at 60 oC. Another work [9] found a decrease of approximately 8.5% in SBSS after conditioning in an alkaline solution for 24 months at 60 oC. In the present study, the SBSS remained unaffected because the core of rebar was minimally affected by water diffusion, which was a result of low porosity content of the rebar as reported in [21].
Flexural strength and flexural modulus
While reports on the flexural analysis and design of GFRP rebar reinforced concrete beams exist [28, 30–33], there is limited documentation on the effect of conditioning on the flexural properties of GFRP rebar sample alone. This work focuses on the residual flexural properties and failure analysis of the 10 years conditioned rebars.
Figure 7 (a) illustrates the flexural load versus displacement curves of the control and conditioned rebars. The curves exhibited a sharp increase with increasing load until reaching the maximum load, at which point the fracture process was initiated. Subsequently, irregularities and staggered decreases in load occurred due to tensile shear failures. The maximum axial fibre stresses were concentrated in the outer fibres, and the shear failure was primarily influenced by the interfacial strength between the outer fibres and the matrix, as evident in the photo in Fig. 7 (a). It is also noticeable that the areas under the curves decreased after conditioning. This decrease was attributed to the failure of the softened matrix and the weakening of the interfacial bonding between the fibres and matrix on the outermost surface, which has been in direct contact with the alkaline solution for an extended period. However, it is important to note that the core of the rebar remained unaffected, as confirmed by the earlier SBSS findings.
Figure 7 (b) presents the flexural strength and flexural modulus retentions of the rebars. The average flexural strength and flexural modulus of the control rebar were measured to be 1035 MPa and 52.5 GPa, respectively. Following conditioning, the flexural strength and flexural modulus were found to retain up to 77.2% and 91.8%, respectively. The reductions in flexural strength and flexural modulus were again attributed to the softening of the matrix and the deterioration of the fibre/matrix interfacial strength on the outer surface, as mentioned earlier. SEM fractographic analysis was conducted to gain insights into the flexural failure process.
Figure 8 reveals that the fibres were embedded within the matrix, demonstrating good wetting of the fibres and strong fibre/matrix interfacial bonding. The matrix exhibited excellent adherence to the fibres, indicating strong fibre/matrix interfacial bonding once again. During the failure process, the polymer matrix fractured first due to its lower modulus, upon the application of stress to the rebar. As a result of the differing elasticities between the matrix and the fibre, interfacial stresses developed at the fibre/matrix interface during the flexural stress transfer process, led to the generation of micro-cracks (Fig. 9). The coalescence of these micro-cracks under increasing stress ultimately caused the rebar to fail. The "riverlines" observed in the matrix originated from the fibre/matrix interface (Fig. 10). These lines, perpendicular to the direction of crack propagation, indicated the gradual development of the crack. Furthermore, the characteristic "cusps" in the matrix indicated shear fracture and exhibited greater deformation due to the increased plasticity of the softened matrix in the 10-Yrs rebar.
Comparisons
The comparative residual properties of the 10-Yrs rebar, including the previously reported tensile properties [21], are depicted in Fig. 11. It is evident that the glass transition temperature of the matrix decreased as a result of plasticization and softening caused by water absorption during conditioning. Among the various mechanical properties, the flexural strength and tensile strength were significantly affected due to the softening of the matrix and the deterioration of the fibre/matrix interfaces in the circumferential areas of the rebar. The flexural modulus was also affected by conditioning, whereas the tensile modulus remained unaffected. While the modulus of the rebar was primarily influenced by factors such as fibre content, fibre type, and manufacturing process, the observations suggested that the properties of the matrix also impacted the flexural modulus, as the softened matrix deformed at relatively low stresses under flexural loading. Although the transverse shear strength (TSS) experienced a slight decrease, the short beam shear strength (SBSS) remains almost unchanged. These findings indicate that the softened matrix had a limited effect on the short beam shear and transverse shear properties, primarily because the fibre/matrix interfaces in the core of the rebar are minimally affected due to the slow and minimal diffusion of water at ambient temperatures during the conditioning period.