Abstract
Alkali-activated materials (AAMs) are considered an eco-friendly alternative to ordinary Portland cement (OPC) for mitigating greenhouse-gas emissions and enabling efficient waste recycling. In this paper, an innovative seawater sea-sand concrete (SWSSC), that is, seawater sea-sand alkali-activated concrete (SWSSAAC), was developed using AAMs instead of OPC to explore the application of marine resources and to improve the durability of conventional SWSSC structures. Then, three types of fiber-reinforced polymer (FRP) bars, that is, basalt-FRP, glass-FRP, and carbon-FRP bars, were selected to investigate their bond behavior with SWSSAAC at different alkaline dosages (3%, 4%, and 6% Na2O contents). The experimental results manifested that the utilization of the alkali-activated binders can increase the splitting tensile strength (ft) of the concrete due to the denser microstructures of AAMs than OPC pastes. This improved characteristic was helpful in enhancing the bond performance of FRP bars, especially the slope of bond-slip curves in the ascending section (i.e., bond stiffness). Approximately three times enhancement in terms of the initial bond rigidity was achieved with SWSSAAC compared to SWSSC at the same concrete strength. Furthermore, compared with the BFRP and GFRP bars, the specimens reinforced with the CFRP bars experienced higher bond strength and bond rigidity due to their relatively high tensile strength and elastic modulus. Additionally, significant improvements in initial bond stiffness and bond strength were also observed as the alkaline contents (i.e., concrete strength) of the SWSSAAC were aggrandized, demonstrating the integration of the FRP bars and SWSSAAC is achievable, which contributes to an innovative channel for the development of SWSSC pavements or structures.
Keywords
Introduction
The demand for marine engineering construction materials has gradually increased with the emerging marine industry and increasing infrastructure construction (e.g., pavements, ports, and dams) on islands and reefs. The vast consumption of traditional building materials has led to reduced freshwater and river sand resources, further resulting in significant failure to the natural ecological environment (Ahmed et al., 2020; Zhou et al., 2021a; Zhang et al., 2020a). In addition, using ships to transport these traditional building materials to offshore islands and reef areas will exhaust a large amount of human and material resources because of the scarcity of conventional materials in these areas (Bazli et al., 2020a; Guo et al., 2018; Wang et al., 2018a). Consequently, the effective use of local marine resources is of great significance to alleviate the consumption of traditional resources.
In recent years, using seawater and sea-sand instead of freshwater and river sand to prepare seawater sea-sand concrete (SWSSC) has captured the attention of a wide range of researchers (Dong et al., 2018; Li et al., 2021; Zhang et al., 2020b). The utilizations of local marine resources (i.e., sea-sand and seawater) not only effectively protect the river ecosystem, but also avoid detrimental impacts on project schedules caused by inevitable weather changes or transportation (Dong et al., 2020; Wang et al., 2021). Thus, the use of seawater, sea-sand or coral sand can be considered instead of freshwater and river sand for concrete preparation applied in port works, dams, and road pavements on these coastal or island areas. Meanwhile, to improve the shrinkage and cracking resistance of these pavements, a new kind of pavement structure, continuously reinforced concrete pavement, emerged in marine engineering construction (Kim et al., 2020; Zhang et al., 2017).
However, the high chloride content of seawater and sea-sand can induce the corrosion of steel bars inside the concrete. These corrosion products develop an expansion effect that can accelerate the propagation of corrosion-induced cracks, seriously affecting the load-bearing capacity or serviceability degradation of SWSSC pavements or structures (Li et al., 2021; Zhang et al., 2019a). Fiber-reinforced polymer (FRP) bars as a new type of rebars can avoid the aforementioned corrosion issue related to steel rebars, originating from their outstanding properties such as superior corrosion resistance, lightweight, and high tensile strength (Li et al., 2018a; Rolland et al., 2018; Zhang et al., 2020c). Currently, FRP bars have gained wide acceptance as an ideal alternative to traditional reinforcement for pavements or structures subjected to harsh environments (Liu et al., 2020; Zhang et al., 2017). However, previous studies (Bazli et al., 2020b; Wang et al., 2018b; Zhou et al., 2021b) have indicated that FRP bars are not entirely immune to marine environments with high temperature, high humidity, and high salinity, and still exhibit a trend of performance deterioration after exposure to these harsh environments. Wang et al. (2018b) found that the ultimate bond stress of GFRP and CFRP bars embedded in coral concrete decreased by approximately 54% and 17%, respectively, when the specimens were soaked in seawater at 60°C for 120 days. This degraded behavior will greatly affect the durability or serviceability of FRP-reinforced SWSSC structures (Dong et al., 2018). Therefore, it is necessary and desirable to exploit a new approach to improve the durability and serviceability of FRP reinforced concrete pavements or structures subjected to marine conditions.
Alkali-activated materials (AAMs) or geopolymers, which are prepared by chemical polymerization of aluminosilicate materials (e.g., slag, metakaolin, coal gangue, and fly ash) with alkaline activators (e.g., alkali silicates and/or alkali hydroxide), are deemed to be innovative substitutes for OPC in concrete preparation on account of their superior characteristics, containing dense microstructures, rapid strength development, good resistance to chemical attack, excellent fire, and chemical resistance (El-Hassan et al., 2021; Huseien et al., 2019; Toragall et al., 2021; Zhang et al., 2021a). The main reaction products of AAMs are N-A-S-H and C-(A)-S-H gels, which not only form a dense paste matrix but also have the ability to absorb or assimilate chloride ions (Ismail et al., 2013; Zhang et al., 2019b, 2021b), thereby obtaining better durability performance than OPC (Wang et al., 2020). This dense microstructure and good durability of AAMs can retard the intrusion of corrosive ions from seawater into the structures to corrode the reinforcement, and thus achieve the goal of improving the durability of FRP structures. Furthermore, AAMs use industrial by-products as precursor materials, which significantly lessen greenhouse-gas emissions and energy consumption (Luukkonen et al., 2018; Zhang et al., 2020d; Zhang et al., 2022a), thereby achieving the double advantages of eco-friendliness and cost-effectiveness. Previous studies have concentrated on using AAMs or geopolymers in ordinary reinforced concrete pavements or structures (Deepti et al., 2016; Maranan et al., 2015). Relative less investigation, nevertheless, was conducted on the utilization of AAMs in SWSSC pavements or structures in the current study (Zhang et al., 2021c). Given that the reliable bond between FRP bars and surrounding concrete directly affects the load-bearing capacity of the structure, this paper will examine the feasibility of preparing SWSSC with AAMs instead of OPCs, and investigate the coordinated deformation ability between FRP bars and SWSSC from the bond characteristics.
In this investigation, three types of FRP bars, namely BFRP, GFRP, and CFRP bars, were selected to determine their bond behavior with seawater sea-sand alkali-activated concrete (SWSSAAC) for various alkaline concentrations (3%, 4%, and 6% Na2O contents), and the cement-based SWSSC was also used for the reference. The differences in failure modes and bond characteristics, including bond strength and bond stiffness, were compared between the SWSSC and SWSSAAC.
Experimental program
Materials
FRP bars
The ribbed BFRP, GFRP, and CFRP bars manufactured by a pultrusion method were adopted for the pull-out test samples, as depicted in Figure 1. The basic mechanical characteristics of these FRP bars are summarized in Table 1. The CFRP bars contained a higher tensile strength (ftu) and an elastic modulus (Ef) than those for the BFRP and GFRP bars, while the difference in ftu and Ef between GFRP bars and BFRP bars was very tiny. The detailed testing procedure for ftu and Ef of the FRP bars is illustrated as follows. First, two epoxy-filled steel tubes were anchored at both ends of the FRP bars and then clamped by the lower and upper clucks of a 1000 kN servo-hydraulic universal testing machine. Then, a 50 mm long extensometer was mounted to the middle height of the FRP bars during the initial loading period, and it was removed after the load value increased to about half of ftu. Finally, the loading rate was set to 2 mm/min until the final breakage of the FRP bars occurred. Schematic diagram of ribbed FRP bars. Mechanical characteristics of ribbed FRP bars.
Concrete
Chemical constituents of GBFS, FA, SF, and OPC.
Chemical composition used for preparing seawater (g/L).
Mix proportions and mechanical properties of SWSSC and SWSSAAC.
The 28 days compressive strength (fcu) and splitting tensile strength (ft) of the SWSSAAC and SWSSC were tested by adopting cube samples (150 mm×150 mm×150 mm), according to the Chinese Code GB/T 50152-2012 (2012). The mechanical properties of the SWSSAAC and SWSSC are provided in Table 4. It can be seen that the SWSSAAC had a higher slump than the SWSSC. This may be because the OPC contained a relatively high water absorption compared to the GBFS. Additionally, a higher ft/fcu ratio for the SWSSAAC was also observed than that of the SWSSC under the same fcu. This may be related to the different reaction products between AAMs and OPCs (Li et al., 2018b; Zhang et al., 2021e). The dense hydration products of alkali-activated binders, that is, C(N)-A-S-H and C-S-H gels, improved the interfacial microstructures and mechanical interlocking between the paste matrix and aggregates and further resulted in an increased ft (Zhang et al., 2021f, 2022b). Moreover, the ft/fcu ratio of the SWSSAAC gradually declined with increasing fcu, and this finding agrees with the natural aggregate concrete (NAC).
Specimen preparation
A 150 mm × 150 mm × 150 mm concrete cube and a 600 mm length FRP bar were prepared for the pull-out test specimens. The FRP bars were placed at the concrete block center and had an embedment length of 70 mm. To prevent the concrete splitting failure, steel spiral stirrups having a 6 mm diameter and a 40 mm spacing were positioned into the bond length region to confine the inner concrete. As shown in Figure 2, the unbonded section was achieved by wrapping PVC tubes around the bars at both ends of the specimen. The internal gap between the PVC tubes and the FRP bars was sealed with a silicone sealant to avert the slurry from soaking into it while the concrete is being poured. Geometrical size of the pull-out specimen (unit: mm).
Loading procedure
After curing for 28 days, the specimens were subjected to a pull-out test on a specific loading device based on ASTM D7913/D7913M-14 specification (2020). As plotted in Figure 3, a linear variable differential transformer (LVDT) located at the free-end of the bars was employed to collect the relative slip value between the FRP bars and the concrete. In addition, a 20 t load sensor installed on the loaded end of the bars was adopted to capture the pull-out force. The loading procedure was then applied to the FRP bars by a displacement control procedure at a 1.2 mm/min rate. The applied load was terminated when the slip value recorded at the free-end increased to about 3 mm or when the FRP bars were fractured. Schematic diagram of the pull-out test setup.
Results and discussion
Tested results
The average bond stress (τ) can be defined by equation (1), which supposed that the bond stress is uniformly distributed within the bond length region (Zhang et al., 2019a).
Pull-out test results of the SWSSC and SWSSAAC specimens.
Failure modes
After the tests were finished, the pull-out specimens were split to observe the interfacial failure between the FRP bars and surrounding concrete, as presented in Figure 4. Two characteristic failure patterns, that is, bar being pulled out, and bar rupture or cracking, were detected in the SWSSC and SWSSAAC specimens during the pull-out process. The detailed discussion is elaborated on below. Typical failure modes for the SWSSC and AWSSAAC specimens with different bar types and alkaline contents. (a) Bar being pulled out; (b) Bar being pulled out; (c) Bar being pulled out; (d) Bar being pulled out; (e) Bar being pulled out; and (f) Bar rupture.
Pull-out failure
For the SWSSAAC specimens with a lower alkaline content (3%) and the SWSSC specimens, the pull-out failure of the FRP bars was their primary failure mode, as illustrated in Figure 4(a)–(e). During the testing process, it was observed that the FRP bars were tardily pulled out from the concrete blocks, and no detectable cracks emerged on the concrete surface. However, continuous FRP bar rib tracks were imprinted on the concrete bonding section. Additionally, the peeling and wear of the surface fibers or ribs of the FRP bars can be observed within the bond length range, which implied that the shear capacity of the FRP bars is somewhat weaker than the concrete strength. As the loading procedure continued, the concrete ribs at the bonded interface were partially abraded and a small amount of crushed concrete was pulled out together with the FRP bars, which lessened the mechanical interlocking and friction resistance between the FRP bars and surrounding concrete. Therefore, a gradual decline in the bond stress after peak value occurred as the loading procedure proceeded. Moreover, the exfoliation of the surface ribs or fibers of the FRP bars and the wear of the concrete at the bond length region for those specimens reinforced by CFRP bars were severer than those for specimens with the GFRP and BFRP bars. The main reason may be that the specimens reinforced with the CFRP bars exerted a higher pull-out load, and then this improved load capacity increased the failure of the FRP-concrete interface.
Rupture of the bars
The FRP bars underwent rupture or cracking failure for the SWSSAAC specimens with concrete strength over 50 MPa. As plotted in Figure 4(f), sudden cracking or rupture of the FRP bars was observed as a result of the stress concentration during the pull-out process. This cracking or rupture of the FRP bars primarily occurred in the load-end within the embedment length region, and no visible cracks were observed at the free-end. The main reason may be attributed to the fact that the bond stress is unevenly distributed within the bond length region, resulting in the bond stress at the load-end being greater than that of the free-end when the specimens failed. Meanwhile, the ultimate pull-out capacity of these specimens was only about 60 kN, which was much smaller than the ultimate load (about 100 kN) of the FRP bars. This phenomenon occurrence is related to the differences in a stress state and the weak shear capacity of the FRP bars (Islam et al., 2015). The bars are uniformly stressed at both ends during the tensile strength testing, while the bars are stressed at one end during the pull-out testing. The stress distribution of the bars along the embedment length is uneven and the transverse shearing stress is generated due to the constraint force produced by the surrounding concrete. In addition, the worn fibers or ribs on the surface of FRP bars also decreased the area of the FRP bars. Therefore, the fractured load under this failure mode was lower than that of the ultimate tensile strength of the FRP bars. Furthermore, this premature fracture of the FRP bars in concrete also resulted in a small slip value collected at the free-end.
Additionally, the bar ribs stayed firmly embedded into the concrete block and no visible concrete failure was detected, while the surface ribs and fibers of the BFRP bars were exfoliated and worn. This may be due to the fact that the concrete strength of the pull-out specimens was greater than the shear capacity of the BFRP bars. Moreover, due to the premature failure of the BFRP bars, the slippage value was relatively small and an intact rib mark could be observed in the bond region of the specimens.
Bond stress-slip curves
Figure 5 plots the effects of bar types and alkaline contents on the bond stress-slip responses of the SWSSC and SWSSAAC specimens. It can be found that the slope of the curves in the upward phase and the ultimate bond stress were higher for specimens using CFRP bars than those using BFRP and GFRP bars. This is because the CFRP bars achieve higher elastic modulus and tensile strength than the GFRP and BFRP bars (see Table 1). Furthermore, the SWSSAAC specimens exhibited a higher initial bond stiffness than the SWSSC specimens, which may be related to the dense interfacial transition zone between the aggregate and alkali-activated binders. As the alkaline dosage (i.e., concrete strength) was aggrandized, the initial bond stiffness and bond strength incrementally increased (see Figure 5(b)). This can be explained by the fact that the higher concrete strength improved the chemical adhesion and mechanical interlocking between the FRP bars and the concrete, further upgrading the binding force of surrounding concrete to the FRP bars. It is noteworthy that those specimens with concrete strengths above 50 MPa exhibited a sharp reduction in the load value due to the sudden failure in FRP bars, when the load applied reached its maximum value, as described in Figure 5. This abrupt failure resulted in the specimens with fcu>50 MPa having a lower peak slip (Su) than that of the specimens with fcu<50 MPa. Bond-slip curves of the specimens at various bar types and alkaline contents. (a) Influence of bar type; and (b) Influence of alkaline content.
Ultimate pull-out stress
Figure 6(a) displays the variation tendency of the ultimate pull-out stress (i.e., bond strength) of the specimens with the BFRP, GFRP, and CFRP bars. It can be observed that the bond strength of the CFRP bars in SWSSAAC was better than that of the GFRP and BFRP bars, while the bond strength between the BFRP bars and the GFRP bars exerted a slight difference due to their similar mechanical properties. Compared with the BFRP and GFRP bars, there were approximately 8.0% and 8.6% improvements in the bond strength for SWSSC and SWSSAAC specimens reinforced with the CFRP bars, respectively. The main reason is that the CFRP bars had higher ftu and Ef than those of BFRP and GFRP bars (see Table 1). Higher ftu and Ef can limit the deformation of the bars and retard the wear of the surface fibers. It should be noted that although the concrete strength of SWSSC was approximately 9% greater than that of SWSSAAC-3%, there was little difference in bond strength between the SWSSC and SWSSAAC, which implied that SWSSAAC specimens contained better bond performance than that of SWSSC specimens at the same concrete strength. Effects of bar type and concrete strength on the bond strength for SWSSC and SWSSAAC. (a) Effect of bar type; and (b) Effect of alkaline content.
As demonstrated in Figure 6(b), the bond strength for the SWSSC and SWSSAAC specimens was gradually improved with the increase of the alkaline content or concrete strength. When the concrete strength increased from 42.3 MPa to 56.9 MPa, an enhancement of approximately 12.2% in the bond strength was detected for the BFRP bars. However, this improved trend regarding the bond strength was not apparent when the concrete strength reached 64.6 MPa, which is related to the failure patterns of the specimens. The exfoliation of the surface fibers or ribs of the BFRP bars dominated the bond failure of the specimens originating from insufficient shear capacity and weak elastic modulus of the BFRP bars. These weak characteristics rendered the shear capacity of the BFRP bars lower than the strength of concrete. Furthermore, this premature failure that occurs in FRP bars also imposed the development of the bond stress, as plotted in Figure 6.
Compared with the SWSSC specimens, the SWSSAAC specimens exerted a higher bond strength, which is associated with the higher ft of the SWSSAAC than SWSSC, as demonstrated in Table 4. A higher ft can delay the propagation of internal cracks inside the concrete and increase the constraint force of surrounding concrete to the FRP bars, which is conducive to enhancing the bond strength of the FRP bars.
Initial bond stiffness
The slope of the stress-slip curves at the upward branch (i.e., initial bond stiffness) was changed accordingly with the variation of the initial slip values. This slope was adopted as the tangential bond stiffness (k), as defined by equation (2), which mirrored the energy consumed by the specimens to resist pull-out failure. To analyze the effects of different slip values (S=0.01, 0.05, 0.10, and 0.20 mm) on the initial bond stiffness of the pull-out specimens with various bar types and alkaline contents, Figure 7 draws the variation pattern of the initial bond stiffness at various slip values. Effects of various bar types and alkaline contents on the initial bond rigidity of the samples for various slip values. (a) Effect of bar type; and (b) Effect of alkaline content.
As described in Figure 7, an apparent decline in the initial bond rigidity of all specimens was observed with increasing initial slip value. The reduction in initial bond stiffness mainly occurred before the slip value reached 0.05 mm. The main reason may be that the friction resistance, chemical adhesion, and mechanical interaction were gradually lowered as the initial slip value of the FRP bars raised. When the slip value was between 0.05 mm and 0.20 mm, the degradation rate of bond stiffness became slower. At this stage, the bond stress was primarily supported by mechanical interlocking and friction resistance. As a result, it was found that the SWSSAAC obtained a greater initial bond spiffiness than that found in the SWSSC. There was an almost three times increase regarding the initial bond stiffness of the SWSSAAC than the SWSSC at the same fcu. Such remarkable result was associated with the dense interfacial microstructures, which has been justified by a higher ft of the SWSSAAC than that of the SWSSC (see Table 4). A higher ft can hinder the development of internal cracks inside the sample during the pull-out procedure and further upgrade the binding force of surrounding concrete to FRP bars. Therefore, a higher initial bond stiffness was expected to occur with increasing fcu.
Additionally, the SWSSC specimens with the CFRP bars exerted a higher initial bond stiffness, followed by GFRP and CFRP bars, respectively, while the SWSSAAC specimens with the GFRP bars exhibited the largest initial bond stiffness, and its order was observed to be GFRP bars>CFRP bars>BFRP bars. This phenomenon is linked with the material properties of the FRP bars. A higher ftu and Ef can reduce the exfoliation of the surface fibers of the FRP bars and produce a higher mechanical interlocking at the FRP-concrete interface. As the compressive strength of the concrete (i.e., alkaline content) enhanced, the bond rigidity of the SWSSAAC at various slip stages, except for S=0.01 mm, was expected to rise gradually. The main reason may be that this increased concrete strength develops a comparatively high chemical adhesion and mechanical interaction.
Conclusions
In this study, bond characteristics of FRP bars in seawater sea-sand alkali-activated concrete (SWSSAAC) were investigated, and the effects of various bar types and alkaline contents were also considered. Some key findings obtained were as follows. (1) The ft/fcu ratio of the SWSSAAC exhibited a degradation tendency with increasing fcu. Compared with the SWSSC, the SWSSAAC exerted a higher ft/fcu ratio at the same fcu due to the improved interfacial microstructures between aggregates and the paste matrix with the use of AAMs. (2) The cracking or rupture in FRP bars was the dominant failure pattern for the SWSSAAC specimens with fcu over 50 MPa. This is related to the uneven distribution of bond stress and the weak shear strength of the FRP bars, which results in premature failure of the FRP bars. In this failure mode, the fracture load of FRP bars reached only 60% of the ultimate load of the FRP bars. (3) The SWSSAAC was expected to obtain higher bond strength and bond rigidity than those for the SWSSC. Compared with the SWSSC, the SWSSAAC archived approximately three times higher initial bond stiffness at the same fcu, which may be attributed to the denser microstructure of AAMs than with OPCs. Furthermore, compared with the BFRP and GFRP bars, the specimens reinforced with the CFRP bars achieved a higher bond strength and bond rigidity due to the relatively high elastic modulus and tensile strength of the CFRP bars. Additionally, the initial bond stiffness and bond strength of the SWSSAAC increased with the increase of fcu. However, since the investigation in this paper is in an exploratory experiment, only the effects of the type of the FRP bars and the concrete strength are considered. Many factors are affecting the bonding performance, including bond length, bar diameter, and restraint state. Thus, more in-depth studies are needed to be carried out in the subsequent research. (4) Using AAMs to replace OPC was an efficacious way to improve the mechanical interlocking at the paste-aggregate interface and to promote the anchorage behavior of FRP bars in concrete. This improved behavior provides support for applying AAMs in FRP-reinforced SWSSC pavements or structures. Thus, it is appropriate and promising to develop innovative FRP reinforced SWSSC structures for marine environment application by combining corrosion-resistant FRP with well-durable AAMs. However, the durability of AAMs used in FRP reinforced SWSSC structures still needs to be further proven by subsequent studies.
Footnotes
Acknowledgments
The authors gratefully acknowledge the funding support from the National Natural Science Foundation of China (Grant No. 51778136).
Author Contributions
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was supported by National Natural Science Foundation of China (Grant No. 51778136); Scientific Research Foundation of Graduate School of Southeast University (Grant No. YBPY2020).
