Abstract
In this paper, static performances of a new precast concrete arch system with carbon fiber reinforced polymer (CFRP) reinforcement were investigated. Varied concrete strengths and segment types of precast concrete arches were tested. The results show that the increase of concrete strength can improve the bearing capacity of the precast arches with CFRP reinforcement. The ultimate bearing capacity of the precast arches segmented at the arch vault with CFRP reinforcement can reach or even exceed that of complete arches, while the precast arch segmented at the shoulder had lower ultimate load. A theoretical analysis method was proposed to predict the failure and bearing capacity of the precast concrete arches with CFRP reinforcement, which were found to be in close agreement with the experimental result.
Introduction
Arch structures have been widely used for thousands of years to build bridges, buildings, culverts and cross drains, because they can withstand high compression force and are suitable to act as compressive structures (Lai et al., 2023; Tan et al., 2015; Xia et al., 2023; Yang et al., 2023). Precast concrete arches are preferable solution for new construction and replacement due to their low initial cost and rapid installation (Tan et al., 2015; Zoghi and Farhey, 2006). Lim and Plantier reported that increasing precast concrete arch tunnels have been designed and constructed in Korea (Lim and Plantier, 2020). Strasky et al. (Strasky et al., 2017) presented that it is applicable to use precast members to build economical and architecturally interesting membrane-arch roof structures. Ong et al. (Chong Yong et al., 2015) reviewed 9 kinds of available precast concrete closed spandrel arch bridge systems in the market and their advantages. The precast concrete closed arches are generally assembled from single-leaf, double-leaf or multi-leaf precast elements. The single spans of single-leaf precast concrete arches are usually smaller than their double-leaf or multi-leaf counterparts with the limitation of transportation and lifting. The structural behavior and jointing system are the further research focuses of precast concrete systems (Chong Yong et al., 2015).
Fiber reinforced polymer (FRP) strengthened concrete structures have become a popular technology in the past decade. Fiber reinforced polymers have also been used to reinforce and strengthen arches. Numerous experimental and numerical studies were performed on the FRP strengthening masonry or concrete arches, shells and vaults and confirmed the efficiency of the FRP strengthening systems (Ali and Hamza, 2015; Avorio and Borri, 2001; Bati et al., 2013; Cao et al., 2010; Chen et al., 2011, 2015; Fagone et al., 2018; Flaga and Kwiecien, 2010; Jiang et al., 2012; Liu et al., 2020; Meier et al., 2011; Oliveira et al., 2010; Wang et al., 2011, 2017, 2018, 2019; Xie et al., 2014; Zhang et al., 2015; Zhao et al., 2021). Firat and Eren (Firat and Eren, 2015) found that the capacity of damaged masonry arch strengthened with FRP increases at least 58% in comparison with unstrengthen masonry arches and strengthening of joints with epoxy resin is an effective alternative to increase the damaged masonry arch capacity. Hamed et al. (Hamed et al., 2015) found that applying the FRP strips leads to an increase of about 40% in the failure load of concrete arch, changes to the cracking pattern, and a significant increase in deflection capacity. Abdulhameed and Said (Abdulhameed and Said, 2019) investigated self-form segmental concrete masonry arches which employ carbon fiber reinforced polymer (CFRP) fabrics to bond the voussoirs and form the masonry arches and proved that CFRPs were efficient for strengthening the extrados of the arch rings under service loadings. The precast concrete arch systems with CFRP reinforcement are promising structures with high-speed construction and good loading capacity.
In this paper, a new precast concrete arch system with CFRP reinforcement was introduced. CFRP sheets are longitudinal bonded on the extrados and intrados to reinforce and connect the arch segments, which results in high flexural strength and low requirement for connection accuracy of the system. To improve the bonding behavior, another CFRP sheets are fully wrapped on the concrete arches with longitudinal CFRP reinforcement. The whole system is only consisting of concrete and CFRP materials, which is corrosion resistant and suitable for using in varied environment. Small-scale precast concrete arches with CFRP reinforcement were tested to investigate their static performances. The plain concrete arches were divided into 1∼4 segments to precast. Their strengthening mechanism and effects were revealed through analyses.
Experimental protocol
Material
Physical and mechanical parameters of carbon fiber sheets and their adhesive.
According to GB/T50081-2002 standard for test methods of mechanical properties of ordinary concrete, the average measured compressive strength of standard cube of C30 and C60 concretes was 44.45 MPa and 61.39 MPa, respectively.
Test setup
Constants of the strengthened arches.
The plain concrete arch structure without segments is shown in Figure 1. The test arch is cast by C30 and C60 concrete at one time, in which the segments are separated by baffles, and then bonded and reinforced with CFRP after 27 days through natural curing. CFRP sheets were longitudinally bonded on the extrados and intrados to reinforce and connect the arch segments and then fully wrapped on the longitudinal CFRP reinforcement. For the specimens with more than two segments, splitting planes were bonded using impregnated resin and reinforced with 40 mm CFRP sheets before overall longitudinal reinforcement. It worth noting that the small-scale arches in the tests were made by hand and 2500 kg hydraulic carrier without any cranes. The realistically-sized precast arches would be constructable with large cranes and careful construction organization design. Scheme of the concrete arch and strain gauge (units: mm).
In order to test the strain change of concrete during test, six strain gauges (Figure 1) are arranged on the plain arch surface. The strain gauges are attached under circumferential CFRP sheets which can affect the strain data to varying degrees. The strain measured in this group of tests is not necessarily an accurate value, but it still has a certain reference value for analyzing failure characteristics and evaluating reinforcement effect.
The test is carried out by applying concentrated load at the center of the arch vault. The tension steel plate is set at the two arch feet for lateral restraint, which can be regarded as hinged restraint, as shown in Figure 2. 1000 kN jack is used for manual loading. Two LVDTs were placed at the extrados and intrados of the vault of the arches, respectively. Data acquisition and crack observation are carried out every 10 kN before 100 kN and every 5kn after 100 kN. During the loading process, observe the structural damage and take photos to record the typical damage characteristics. Static loading set-up.
Results and analysis
Failure behavior
During the compression of G30-1 arch, the arch feet were compacted first, and then the whole arch were stressed together. This step can be removed by preloading. When the load on the vault reaches 40 kN, the CFRP sheets on the lower side of the vault is stripped, and small cracks are found in the transverse distribution of the vault, as shown in Figure 3(a). With further loading, the peeling sound continued. Due to the comprehensive wrapping, no obvious test phenomenon was found outside. At 160 kN, a series of fracture sounds were heard, and it was found that there was peeling in front of the front and left arch foot of the arch vault, and the crack penetrated in front of the arch vault. At 170 kN, corresponding through cracks were also found at the rear side of the arch vault. Finally, during the loading process of 220 kN to 225 kN, the right arch foot was shear damaged, as shown in Figure 3(b). Due to the restraint of CFRP on both sides and sides of the arch, the upper part of the failure area did not slide to the left. Under the combined action of CFRP sheets around, a section failure was formed at the arch vault and right arch foot, and the forming mechanism could not bear the load and failed. The overall failure results of arch G30-1 are shown in Figure 3(c). It can be seen from the figure that the outer side of the side CFRP sheets is tensioned and the inner side is wrinkled, which shows that the side CFRP sheets is very helpful to improve the mechanical performance of the structure. Failure behavior of G30-1: (a) Cracked at vault; (b) Shear failure of right arch foot; (c) Overall failure behavior.
When the load of G30-2 vault reaches 80 KN, the CFRP sheets on the lower side of the vault is partially stripped, and the vault is subject to sliding deformation, as shown in Figure 4(b). G30-2 arch is divided into segments at the arch vault, and two layers of carbon fiber reinforcement are adopted at the segments. Although the carbon fiber sheets is partially stripped when the arch vault is deformed, it can still bear the load. With further loading, the peeling sound continued. Due to the comprehensive wrapping, no obvious test phenomenon was found outside. Until 210 kN, the left arch waist concrete is cracked and the carbon fiber sheets is broken, as shown in Figure 4(a). The failure behaviors of arch G30-2 are shown in Figure 4. Failure behavior of G30-2: (a) Fracture at shoulder; (b) Crack at vault; (c) Overall failure.
When the load of G30-3 vault reaches 70 kN, it is found that the front side of the vault is circularly cracked. At 135 KN, the right side makes a loud slip, and the mid span crack penetrates longitudinally, as shown in Figure 5(b). At the end of 145 kn, the joint at the left arch waist burst and the carbon fiber sheets was broken, as shown in Figure 5(a). Compared with other arches, the failure is relatively sudden. Failure behavior of G30-3: (a) Fracture at shoulder; (b) Crack at vault; (c) Overall failure.
When the load of G30-4 vault reaches 80 KN, it is found that the front side of the vault is circularly cracked. At the end of 270 kN, the joint at the left arch waist burst and the carbon fiber sheets was broken, as shown in Figure 6(a). It can be seen that the outer side of the joint concrete is broken, indicating that the carbon fiber sheets and concrete are subject to bending together, and the neutral axis is basically at the center line of the section. Failure behavior of G30-4: (a) Fracture at shoulder; (b) Overall failure behavior.
The failure of arch G60-1 is shown in Figure 7. At 160 kN, the concrete crack is found on the arch vault. At 235 kN, the crack runs through the whole section. With the further increase of load, the crack appears at the arch waist and continues to develop. Until 290 kN, the crack runs through and destroys at the arch vault and right arch waist, so the arch cannot be further loaded. Compared with G30 series arch, G60 series arch has less deformation when damaged. Failure behavior of G60-1: (a) Crack at vault; (b) Fracture at shoulder; (c) Overall failure.
The failure of arch G60-2 is shown in Figure 8. Cracks are found at the arch waist at 125 kN, and shear failure occurs at the arch feet on both sides at 310 kN. At that time, the composite arch cannot continue to bear the failure. Compared with G60-1 arch, the mid span of arch G60-2 is segmented, but the segmented part is bonded and reinforced with two layers of carbon fiber. Obvious damage is found during loading. Failure behavior of G60-2: (a) crack of left arch foot; (b) Shear failure of right arch foot; (c) Overall failure behavior.
Load-displacement curves and ultimate loads
The load and displacement curves of G30 arches are shown in Figure 9, which is not compared with those of G60 as the LVDTs of G60-2 could not obtain correct signal. The stiffness change of different segment type can be obtained. The initial slopes of all curves are similar as the initial cross aera of different arches were almost at the same level. After that, cracks formed on the extrados surface of the vault and shoulder. The G30-2 and G30-4 arches were reinforced at the vault using two layers of CFRP sheet. The stiffness of those arches was still high after initial cracking, which reveals that the crack propagation of G30-2 and G30-4 was slower than that of other arches. For G30-2, large displacement was observed without much increase of load around the peak load. It could be a result of delamination and slip at the vault which was a joint of the two segments. The peak loads of the arches were reached at similar displacement around 18 mm. The ultimate loads were varied for different segmentation with different failure mode. Load-displacement curves.
Comparison of test results.
aNote: Failure mode of G30-1 was observed as shear failure. The predicted load for flexural failure was closer to the ultimate load.
Strain curves
Figure 10 shows the strain curves of arches in the test. It should be noted that some strain gauges cannot collect data due to concrete cracking and other reasons during loading. This is the reason that some curves did not reach to ultimate load of arches and data of some strain gauges were not plotted in the Figure 10. As shown in Figure 10, there existed turning points in the growth curves of the tensile strain. The turning points of strain curves of the arch vaults are lower than those of the middle part of the arches, which reveals that concrete cracks started early at the arch vaults. The compressive strains were almost linear when the strains were low. Some bi-linear curves were also be obtained and resulted in large ultimate strain which is more than 3500 με. Strain curves of arches. (a) G30-1. (b) G30-2. (c) G30-3. (d) G30-4. (e) G60-1. (f) G60-2.
Analysis
Utilization rate of CFRP
Utilization rate of CFRP.
Effect of number and location of segments
The number of segments of the precast concrete arch with CFRP reinforcement was not an important effect for the static bearing capacity. The arches with 1 segment, 2 segments and 4 segments had similar ultimate bearing capacities as shown in Table 3. It reveals that the reinforcement of CFRP is effective to connect different segments of the arches. However, the location of the segments is an important factor for both the static capacities and failure modes. The precast concrete arch with CFRP reinforcement to be segmented and reinforced at the top of the arch would prevent the vault failure. The arches without vault deviation first cracked and failed at the vault, and collapsed with another hinge and failure development at the shoulder or foot. The section failure at vault was naturally divided the arches into segments separated at the vault, which can be equal to the vault deviation during the precast. That is the reason that the ultimate bearing capacity of the precast arches with vault deviation can reach or even exceed that of the single leaf arches. For the arches with 3 segments, the early shoulder hinge development due to the shoulder deviation during precast resulted in early failure and lower loading capacities.
Theoretical analysis and failure mechanism
The CFRP reinforced concrete arches can be simplified and analyzed as two-hinge arches using the force method as shown in Figure 11(a) and also used in (Zhang et al., 2015). The internal axial force FN, shear force FV and bending moment M of any section in the arch can be expressed as: Theoretical analysis model. (a) Two-hinge arches. (b) Five/four-hinge arches.

The failure modes found in the tests were flexural failure and shear failure. The strain and stress distribution at ultimate condition of flexural failure are shown in Figure 12, where the tensile strength of concrete and compressive strength of FRP strips are neglected. Based on the equilibrium of forces and strain compatibility based on plane section assumption, the nominal flexural strength can be derived Strain and stress distribution at ultimate condition of flexural failure.

Shear failure depends on the shear strength of the cross section, Vu, given by
Conclusion
Performances of new precast concrete arch system with CFRP reinforcement were studied. The effect of concrete strength, CFRP reinforcement and number/location of segments has been investigated, and a theoretical analysis method was proposed. The results and discussions presented in the present paper allow the following conclusions to be made. (1) The increase of concrete strength can improve the bearing capacity of precast concrete arch system with the same amount of CFRP reinforcement. The ultimate load of 2 segments precast arch with concrete strength of 44.45 MPa is 1.58 times of that of arch with concrete strength of 61.39 MPa. Both the flexural failure and shear failure of the arches depends on the strength of the concrete. (2) The ultimate bearing capacity of reinforced arch is related to the reinforcement form of CFRP. The denser the circumferential CFRP sheets, the more the amount of CFRP on both sides of the arch, and the greater the failure strength. (3) The precast concrete arch with CFRP reinforcement is better to be segmented and reinforced at the vault of the arch. The ultimate bearing capacity of the precast arches with vault deviation can reach or even exceed that of single leaf arch with CFRP reinforcement. The enhanced nominal flexural strength at the vault delayed the form of vault hinge of the five/four-hinge collapse.
Footnotes
Acknowledgments
The authors gratefully acknowledge the financial support provided by the National Natural Science Foundation of China (52008209 and 51778622) and the Fundamental Research Funds for the Central Universities (NS2020048).
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 work was supported by the National Natural Science Foundation of China (52008209, 51778622), and Fundamental Research Funds for the Central Universities (NS2020048).
