Experimental and numerical study on the axial compression behavior of circular concrete columns confined by BFRP spirals and ties

2.1 Specimen design

A total of 90 specimens were prepared which included 9 cubes of size 150 × 150 × 150 (mm), and 81 RC cylinders, measuring 150 mm in diameter and 300 mm in height. The RC cylinders were divided into five reinforcement groups, i.e., spirals made of 6 mm BFRP bars, ties made of 6 mm BFRP bars, spirals made of 8 mm BFRP bars, spirals made of 6 mm BFRP bars along with four steel longitudinal bars of 6 mm diameter, and steel ties with steel longitudinal rods of 3 mm diameter. The stirrups were arranged in three spacings, i.e., 45, 60, and 90 (mm). All the RC cylinders had two extra steel stirrups of 6 mm diameter at the top and bottom sections. Steel rods of 3 mm diameter were utilized for the specimens that didn’t have the steel longitudinal bars, to support and keep the proper spacing between the stirrups. Figure 1(a) depicts the specifics of BFRP spirals/ties in the transverse direction and steel bars/rods in the longitudinal direction respectively. Figure 1(b) illustrates the cross-sectional details of the reinforcement. Longitudinal steel bars were arranged in specimens with an average interval of 90°. The spirals/ties were fixed to longitudinal steel bars/rods at specified spacings with steel wires. The overview of reinforcement setup is shown in Figure 1(c). The variables were the spiral spacing (s), transverse bars diameter, and type.

Figure 1:

Design details of the column specimens.

Steel bars/steel rods with a constant diameter of 6 and 3 (mm) were used for longitudinal reinforcement while BFRP/steel stirrups of diameters 6 and 8 (mm) bars were used for transverse reinforcement. Due to the commercial availability and compatibility with the common transverse reinforcement employed in practice, the BFRP bar diameters of 6 and 8 (mm) were chosen. These diameters are frequently employed in spirals and ties to confine circular RC columns, particularly those that are intended for moderate loading. The chosen range offers realistic reinforcing choices while preserving constructability and spacing compatibility in a 150 mm diameter column. The test matrix is presented in Table 1. Groups RC-1 and RC-2 specimens were prepared following three distinct CMD codes, having 45, 60, and 90 (mm) BFRP stirrup spacings. Each code contained a set of three RC cylinders for specific stirrup spacing to ensure the consistency of the test results. Groups RC-3, RC-4, and RC-5 samples were fabricated by following the ACI CMD code for specific stirrup spacings, and these samples had different reinforcement configurations as mentioned in Table 1. To avoid the end extrusion damage during axial loading, a unidirectional BFRP sheet was used for the external confinement of RC cylinders at the top and bottom 50 mm sections, as shown in Figure 1(d). An overlap of 120 mm in length was specified at the end of the BFRP external layer.

2.2 Material properties

In this experimental program, ordinary Portland cement of grade strength 42.5 MPa with a specific gravity of 3.15 was used as a binder. The aggregates utilized were river sand and crushed basalt CA with a maximum size of 19 mm. Tap water was used for the concrete batching. BFRP bars with 6 and 8 (mm) diameters were used to create stirrups. HRB400 steel bars of 6 mm diameter were utilized for the steel longitudinal and transverse (ties) reinforcements. A unidirectional BFRP sheet of a density of 380 g/m³ was used for the external confinement of cylinders at the top and bottom sections. The Sanyou epoxy resin adhesive was used to confine the BFRP sheet on the concrete cylinders. The epoxy resin and the hardener were mixed with a ratio 1:2 to make a proper mix composition for the BFRP external confinement. The material properties are listed in Table 2.

2.3 Concrete casting procedure

Concrete samples were prepared under standard conditions, according to three CMD codes, i.e., ACI [49], BS [50], 51], and Chinese code (JGJ) [52]. The various CMD codes were chosen because of their different concrete proportioning approaches and worldwide applicability. These codes exhibit variations in critical parameters, including aggregate content, w/c ratio, and slump control for any specific strength concrete. Their consideration gives a fair evaluation of the effects of regionally varied CMD guidelines on the mechanical properties of RC concrete columns. A 28-day design CS of 35 MPa was anticipated for the concrete. Table 3 shows the calculated quantities of concrete ingredients. Before concrete batching, the necessary tests were conducted to know the aggregates’ nature by determining the specific gravity and fineness modulus according to the American Society of Testing and Materials (ASTM) C128 [53] and ASTM C127 [54]. A revolving concrete mixer was used to prepare the concrete mixtures. Before mixing, the mixing pan was cleaned to ensure the complete utilization of all mixing water for hydrating the cementitious material. Subsequently, all dry components were added to the pan and water mixed thoroughly until uniform consistency was achieved. Workability tests were conducted according to the ASTM C143 [55] before pouring the concrete into the molds. All the molds, were placed on a vibrator for the proper compaction of concrete. After 24 h, the specimens were demolded and put in water tanks to cure for 28 days. The specimens were prepared and cured by following the guidelines of ASTM C192 [56].

2.4 Test setup

For all the plain concrete cubes and RC specimens testing, a servo-hydraulic machine of capacity 2000 kN was used. All the cylindrical specimens were tested under the displacement control mode at a fixed loading rate of 0.3 mm/min. The cube specimens were tested under compression mode at the standard loading rate of 0.5 MPa/s. Four strain gauges, i.e., two vertical and two horizontals, were installed at the outer surface of the RC cylinders at the mid-height section, and for the inner reinforcements, two strain gauges were installed on the BFRP stirrups. All the strain gauges were connected to the data acquisition system during the testing to record the strain data. The axial CS properties of the specimens were determined as per the ASTM C39 [57]. To ensure the uniform distribution of load, the specimens’ surfaces were made smooth. The testing arrangement is given in Figure 2.

Figure 2:

Experimental arrangements.

3.1 Concrete mix design codes comparison

Many countries have standardized CMD codes because of the noteworthy development of concrete technology in the construction field. The CMD codes of grade strength 35 MPa, i.e., ACI, BS, and JGJ were studied in terms of methodology, ingredient proportioning, and test results. The ACI, BS, and JGJ codes are based on the empirical relations developed through the experimental process, concerning the material properties and construction specifications. The concrete mixtures were calculated according to equations (1)–(3) specified by CMD codes. The standard deviation (σ) for the ACI code is 8.33, but the BS and JGJ codes have the same σ of 5.0. In this particular case, the BS code had the largest cement content, followed by the JGJ and ACI codes. Compared to the JGJ and BS codes, the ACI code has the highest fine aggregate content. In terms of CA, the ACI code has the lowest content and the JGJ code has the highest. The JGJ code recommended the highest W/C ratio compared to the ACI and BS codes. Due to the difference in concrete ingredients, the designed concrete exhibited variations in workability and compressive properties.

The workability tests were carried out in compliance with the ASTM standard. The slump values for ACI, BS, and JGJ concrete mixtures adhered to the ranges outlined by the respective codes. The results of the slump test are presented in Table 4. The BS concrete mixture showed 12.5 % and 27.5 % higher workability than ACI and JGJ concrete mixtures, respectively. The concrete mixtures are symbolized as follows: the first letter refers to the CMD code, followed by the type of concrete, and then followed by the CA type and size. For instance, ACI-PC-B19 indicates the ACI CMD code plain concrete made with basalt CA of size 19 mm.

3.2 Failure modes

3.2.1 Failure modes of plain concrete

During compression testing of the PC cubes, the failure patterns were observed. In the initial stage of the loading, no visible cracks appeared on the specimen’s surface but as the load increased, micro-cracks appeared and increased in length, width, and number. The failure occurred in the vertical pattern, from top to bottom, at both edges. The matrix, along with CA, chipped off from the sides. The crack paths mainly propagated through the matrix, and it was observed that basalt CA was not mainly fractured upon the failure of the cubes. The failure modes of cube specimens, prepared by ACI, BS, and JGJ codes, are presented in Figure 3. The Scanning Electron Microscopic (SEM) study was carried out to examine the micro-structural behaviors of tested specimens, which identified the interfacial transition zone (ITZ) as the main location for developing microcracks under compressive loading which likely caused fractures to gradually spread and eventually result in the specimens’ failure. Specimens manufactured using the BS and JGJ codes exhibited more extensive microcracking branching and greater densities than the ACI specimen, with the majority of the cracks forming radially around the basalt CA. It raises the possibility that the various CMD codes have caused variations in the microstructural characteristics. See Figure 4.

Figure 3:

Failure modes of cube specimens.

Figure 4:

Overview of micro-structural failure: SEM analysis.

3.2.2 Failure modes of reinforced concrete

The failure pattern of RC cylinders was examined. The vertical macro cracks were observed in the RC cylinders, which extended from top to bottom with lateral expansion, and the concrete cover was separated from the main core of the concrete. The BFRP strips were ruptured at the bottom sections of some specimens. After the ultimate failure, the internal reinforcement was checked, and it was found that the BFRP transverse reinforcement (spiral/ties) ruptured in the mid-height region. The vertical supporting steel rods were found buckled significantly in the mid-height region where the BFRP stirrup was ruptured. The degree of buckling in the longitudinal steel rods varied based on the spacing of spirals/ties. RC cylinders with close spirals/tie spacing, i.e., 45 mm exhibited less buckling compared to those with larger spacings, i.e., 60 mm and 90 mm (Figure 5a–g). The failure patterns of RC cylinders, reinforced with 6 mm BFRP spirals and 6 mm longitudinal steel bars, included basalt CA crushing, vertical macrocracks, and chipping of concrete. The steel bar buckled as a result of the BFRP spirals breaking at mid-height. Closer BFRP spiral spacing (45 mm) resulted in less buckling, which indicates a greater confinement effect. Concrete failure and steel bar buckling were more severe in the cylinders with a larger spacing of 90 mm (Figure 5h). The RC cylinders strengthened with 6 mm steel ties and 3 mm longitudinal steel rods, exhibited substantial vertical macro cracks and concrete cover spalling. The degree of concrete degradation was more evident in specimens with broader steel tie spacings, i.e., 60 mm and 90 mm. The spacing of the steel ties influenced crack propagation, with increased spacing resulting in more significant cracking. The longitudinal steel rods demonstrated significant buckling, with larger tie spacing (Figure 5i). The internal reinforcement sectional views are illustrated in Figure 6.

Figure 5:

Failure modes of RC cylinders.

Figure 6:

Internal view of the reinforcement sections after ultimate failure.

The RC cylinders failed, generating a loud cracking noise. The concrete cover spalling was marked by separating large pieces of concrete from sides. Once the surface spalled, substantial cracking in the concrete core caused the lateral expansion, starting the passive confining pressure of the BFRP transverse reinforcements. At this moment, the confining restraint provided by the BFRP transverse reinforcement was activated, and the cylinder became able to carry the compressive loading further. As a result, the concrete was crushed, and the BFRP spirals/ties ruptured with the buckling of longitudinal steel bars/steel rods. The observations at the maximum load stage indicate that the BFRP spirals/ties activated to confine the concrete core after concrete spalling started from the specimen’s sides and macro cracks developed. In the end, the cylinders repeatedly exhibited the development of buckling of the steel bars/steel rods, yielding of the BFRP spirals/ties, fracture of BFRP spirals/ties, and crushing of concrete. The major vertical cracks were spread between the ends of the BFRP-RC cylinders.

The tested RC cylinders’ visual examination showed that the transverse reinforcement spacing and arrangement had a substantial impact on the degree and pattern of cracking, concrete spalling, and crushing. Specimens with 90 mm BFRP stirrup spacing had extensive vertical cracks and localized spalling in the mid-height area, signifying inadequate confinement. Conversely, specimens with 45 mm spacing demonstrated a distribution of fine cracks accompanied by controlled spalling, indicating efficient lateral confinement pressure and energy accumulation. The observed failure was qualitatively compatible with the estimated G_F and ductility index, despite the fact that exact crack widths and spalled concrete volume were not quantified. In general, specimens that displayed significant spalling and extensive longitudinal cracks demonstrated reduced G_F and G_Acc. In contrast, dense transversely spaced specimens retained core integrity and showed superior post-peak behavior. The function of BFRP transverse confinement in controlling crack development and postponing brittle failure is further supported by these failure characteristics.

3.3 Compressive strength

3.3.3 Comparison to contemporary work

To assess the reliability and technical validity of the current study, the strength results were compared with relevant studies that investigated the axial compressive behavior of concrete columns reinforced with BFRP stirrups. The studies were chosen for comparison due to their utilization of comparable transverse reinforcement configurations, materials, and stirrup spacings.

A study [19] investigated concrete columns reinforced with BFRP bars and transversely confined using BFRP spirals of varying diameters and spacings. Specimens f-4 and f-8 were chosen in a nearly matched arrangement, using concrete of comparable strength and longitudinal epoxy resin rods with 8 mm BFRP spirals spaced at 46 and 92 (mm), respectively. The specimens f-4 and f-8 obtained CS of 42 MPa and 34 MPa, demonstrating the reduction of strength with the increased BFRP spiral spacing. In the current study, the specimens ACI-B(S)8–45 and ACI-B(S)8–90 of group RC-4 achieved 38.8 MPa and 34.2 MPa CS. These specimens were reinforced with 8 mm BFRP spirals at 45 and 90 (mm) spacings, respectively. The observed strength trends are notably consistent, as both studies validate that closer BFRP spiral spacing considerably improves strength. The observed strength reduction in the current ACI-B(S)8–45 specimen can be linked to variations in the type of longitudinal rods and the CMD code utilized.

In another investigation [20], square columns measuring 250 × 250 (mm) and 800 mm in height were reinforced with steel longitudinally, and with BFRP/steel ties of 8 and 12 (mm) diameter spaced at 60, 95, and 140 (mm) intervals. A CS of 38.2 MPa was attained by the reference specimen 8S–95C–S reinforced with 8 mm steel ties spaced 95 mm apart. However, despite the lower tie diameter, the group RC-5 specimen ACI-S(T)6–90 reinforced with 6 mm steel ties at 90 mm spacing achieved 36.3 MPa in the current investigation. The similar strength indicates that the diameter difference was compensated by more consistent confinement provided by the circular section geometry in this study as opposed to the square section used in the reference study.

In a specific case, reference specimens 8-95C–S, which were reinforced using steel longitudinal bars and 8 mm BFRP ties at 95 mm spacing, achieved 27.4 MPa. Conversely, the specimens of RC-2 group of current study (ACI-B(T)6–90, BS-B(T)6–90, and JGJ-B(T)6–90), using 6 mm BFRP ties at 90 mm spacing, attained 37.3 MPa, 36.7 MPa, and 31.6 MPa, respectively. This research demonstrates that the notably superior CS (13 %–26 %) underscore the benefits of circular cross-sections and potentially more efficient tie details, especially with smaller diameter BFRP ties. The findings emphasize the possible limitations of square-section columns when reinforced with different BFRP ties, as stress concentrations at the corners lead to less effective confinement. The reference specimen 8-60 A-S achieved 39.2 MPa by using 8 mm BFRP ties spaced at 60 mm. The corresponding specimens (ACI-B(T)6–60, BS-B(T)6–60, and JGJ-B(T)6–60) in the current study achieved 38.6 MPa, 40.4 MPa, and 33.2 MPa, respectively. These findings further illustrate that CMD code has a meaningful influence on confined strength, with BS code performing better in this configuration. Additionally, it supports the usefulness of closely spaced BFRP connections, even at lower diameters. The graphical presentation is illustrated in Figure 7.

Figure 7:

Comparison to contemprary studies.

3.4 Load-displacement behavior

The compression behaviors of RC cylinders are graphically illustrated in terms of load-displacement curves. Figure 8 represents the overall compression behaviors of group RC-1 (ACI, BS, and JGJ) specimens. The load-displacement curves begin with a linear pattern, which illustrates the concrete’s and the BFRP reinforcement’s elastic nature. This linearity indicates that the concrete and reinforcement are functioning together without considerable cracking. In Figure 8A, the ACI RC cylinders showed linear behavior in ascending regions until their respective P_max but displayed variation in post-peak behavior. The BS-B(S)6–45 and BS-B(S)6–60 curves exhibited linear behavior in the elastic region up to their P_max, as shown in Figure 8B, whereas the BS-B(S)6–90 curve displayed greater plasticity and increased gradually. The stiffness is reduced as microcracks develop into macrocracks in the non-linear section of BS-B(S)6–90. In Figure 8J, the curves JGJ-B(S)6–60 and JGJ-B(S)6–90 start to deviate from their initial linearity as the load approaches the peak as compared to the JGJ-B(S)6–45 curve. It indicates the development of micro-cracks in the elastic region. After the P_max, the curves showed ductile behavior and failed gradually. Figure 9 presents the compressive behaviors of group RC-2 specimens. In Figure 9A, curves ACI-B(T)6–45 and ACI-B(T)6–60 showed identical pre-peak behaviors up to the P_max, with a noticeable variation in the post-peak region. The curve ACI-B(T)6–90 showed deviation in the transition and post-peak regions. The curves BS-B(T)6–45 and BS-B(T)6–90 ascended identically to the maximum load point but showed noticeable variation in the post-peak region while BS-B(T)6–60 showed deviation in the elastic region (Figure 9B). In Figure 9J, all the curves of JGJ-RC cylinders followed almost the same ascending pattern till their respective P_max but displayed a noticeable variation in post-peak region. In Figure 10, the load-displacement curves of group RC-3 specimens are displayed. In the elastic region, all the curves upsurged identically while the curve ACI-B(S)6-90-SB6 showed a slight variation in the transition zone and a noticeable drop in the post-peak region. This decline indicates the cracks enlargement, resulting in less load-bearing capability. The ACI-B(S)6-45-SB6 and ACI-B(S)6-60-SB6 showed more resistance against the loading. In Figure 11, the specimens of group RC-4 displayed the same compressive behavior and mounted identically in the elastic and transition regions but showed different post-peak behaviors. At the P_max stage, internal damage in the concrete reached a critical level, though the BFRP confinement delayed full collapse by containing the lateral expansion (ACI-B(S)8–45). Figure 12 displays the load-displacement curves of group RC-5 specimens. All the curves ascended almost in the same manner with slight deviation but showed different post-peak behaviors.

Figure 8:

Group RC-1: Load-displacement behavior of RC cylinders confined with 6 mm BFRP spirals at various spacings.

Figure 9:

Group RC-2: Load-displacement behavior of RC cylinders confined with 6 mm BFRP ties at various spacings.

Figure 10:

Group RC-3: Load-displacement behavior of ACI-RC cylinders confined with 6 mm BFRP spirals & 6 mm steel bars.

Figure 11:

Group RC-4: Load-displacement behavior of ACI-RC cylinders confined with 8 mm BFRP spirals at various spacings.

Figure 12:

Group RC-5: Load-displacement behavior of ACI-RC cylinders confined with 6 mm steel ties at various spacings.

The post-peak behaviors of RC cylinders demonstrate that BFRP reinforcements continue to support some load despite extensive concrete damage. This behavior shows that the RC cylinders retained some residual capacity due to the transverse confining effect of BFRP. The analysis of failure modes and the assessment of load-displacement behavior indicate that differences predominantly occur in the post-peak region, attributable to variations in the rib spacings of BFRP spirals/ties, the diameter of BFRP spirals, the type of transverse reinforcement, and the substitution of longitudinal steel rods with steel rebars. The CMD codes exhibited minimal influence on the failure modes of the RC cylinders due to the dominant behavior of reinforcements.

3.5 Energy accumulation

The total energy accumulation during the testing, from the first loading point to the ultimate failure point, is termed the accumulated energy (G_Acc). Table 6 presents the results of the G_Acc of RC cylinders (groups RC 1–5). In the group RC-1, ACI-B(S)6–45 achieved 20.6 % and 16.7 % higher G_Acc than those with larger spacings, i.e., 60 and 90 (mm), respectively. In the case of BS RC cylinders, BS-B(S)6–45 gained about 15 % higher G_Acc than BS-B(S)6–60 and BS-B(S)6–90, respectively. The JGJ RC cylinders (JGJ-B(S)6–45) exhibited 20.2 % and 7.4 % higher G_Acc than the JGJ-B(S)6–60 and JGJ-B(S)6–90. In this case, the RC cylinders with condensed BFRP spiral spacings (45 mm) exhibited higher G_Acc than those with larger BFRP rib spacings. The reason could be the higher confining pressure provided to the concrete core by the concentrated BFRP spirals, resulting in higher CS and G_Acc. In the case of group RC-2 specimens, the ACI-B(T)6–60 offered 7.2 % and 12.5 % higher G_Acc than ACI-B(T)6–45 and ACI-B(T)6–90. Similarly, the BS-B(T)6–60 offered 13.5 % and 11.5 % higher G_Acc than BS-B(T)6–45 and BS-B(T)6–90, respectively. The same trend was found in the JGJ RC specimens, i.e., JGJ-B(T)6–60 achieved 14.8 % and 40 % higher G_Acc than JGJ-B(T)6–45 and JGJ-B(T)6–90, respectively. The group RC-2 specimens of moderate BFRP tie spacings (60 mm) offered the highest G_Acc than those having tie spacings of 45 and 90 (mm). The group RC-3 specimens ACI-B(S)6-45-SB6 achieved 4.1 % and 30.2 % higher G_Acc than those with larger spiral spacings, i.e., ACI-B(S)6-60-SB6, and ACI-B(S)6-90-SB6, respectively. In the case of group RC-4, the specimens ACI-B(S)8–45 achieved 49.7 % and 58.9 % higher G_Acc than the ACI-B(S)8–60 and ACI-B(S)8–90, respectively. In the case of group RC-5, the specimen ACI-S(T)6–45 exhibited 8.8 %, and 31.4 % higher G_Acc than those with the larger ties spacings, i.e., ACI-S(T)6–60 and ACI-S(T)6–90, respectively. In groups RC-3, RC-4, and RC-5, the close rib spacings (45 mm) was more effective in terms of G_Acc. The G_Acc graphical presentation of groups RC 1–5 is illustrated in Figures 13, 14, and 15.

Figure 13:

Accumulated energy of RC cylinders: BFRP spirals of 6 mm at various spacings (group RC-1).

Figure 14:

Accumulated energy of RC cylinders: BFRP ties of 6 mm at various spacings (group RC-2).

Figure 15:

Accumulated energy results of RC cylinders: Group RC-3, group RC-4 and group RC-5.

The comparison of G_Acc between group RC-1 and group RC-2 was made. In this case, ACI-B(S)6–45 specimens achieved 11.5 % higher G_Acc than ACI-B(T)6–45 while ACI-B(T)6–60 and ACI-B(T)6–90 offered 28.7 % and 14.5 % higher G_Acc than ACI-B(S)6–60 and ACI-B(S)6–90. It was found that RC specimens reinforced with BFRP spiral having small spacings of 45 mm are more effective than the RC specimens reinforced with BFRP ties having same spacings. In contrast, in the case of larger spacings, i.e., 60 and 90 (mm), the RC specimens with BFRP ties performed well and resisted longer against the applied load, hence gaining a higher G_Acc. The BS-B(S)6–45 and BS-B(T)6–45 having 45 mm BFRP spirals and ties spacing offered almost the same G_Acc with a 2.4 % variation while BS-B(T)6–60 and BS-B(T)6–90 exhibited 24.5 % and 10.2 % higher G_Acc than specimens having BFRP spirals, i.e., BS-B(S)6–60 and BS-B(S)6–90, respectively. The results of the G_Acc of JGJ RC specimens showed that JGJ-B(S)6–45 and JGJ-B(S)6–90 achieved 10.5 % and 32 % higher G_Acc than JGJ-B(T)6–45 and JGJ-B(T)6–90, respectively. However, in the case of 60 mm spacings, the JGJ-B(T)6–60 gained 24 % higher G_Acc than JGJ-B(S)6–60. In comparison of group RC-1 and group RC-3, the ACI-B(S)6–45 and ACI-B(S)6-45-SB6 offered almost the same G_Acc. In the case of 60 mm spiral spacings, ACI-B(S)6-60-SB6 gained 18 % higher G_Acc than ACI-B(S)6–60 while in the case of 90 mm spiral spacings, ACI-B(S)6–90 achieved 15.2 % higher G_Acc than ACI-B(S)6-90-SB6. The G_Acc of group RC-1 and group RC-4, the ACI-B(S)8–45 and ACI-B(S)8–60 achieved 42.8 % and 15.8 % higher G_Acc than ACI-B(S)6–45 and ACI-B(S)6–60. However, in the case of 90 mm spacings, the ACI-B(S)6–90 offered 7.5 % higher G_Acc than ACI-B(S)8–90. The G_Acc of group RC-2 and group RC-5 was compared and it was found that ACI-S(T)6–45, ACI-S(T)6–60 exhibited 27 % and 13.8 % higher G_Acc than ACI-B(T)6–45, ACI-B(T)6–60, while ACI-S(T)6–90 and ACI-B(T)6–90, offered nearly the same G_Acc, respectively.

The G_Acc results indicate a significant influence of reinforcing type, rebar diameter, stirrup spacing, and CMD codes on the energy absorption capabilities of RC cylinders. RC groups using larger BFRP bar size (group RC-4) or steel ties (group RC-5) continuously showed greater G_Acc. In BFRP-reinforced configurations, BFRP ties (group RC-2) showed superior performance compared to BFRP spirals (group RC-1) at larger spacings, suggesting that BFRP ties are more efficient for specimens requiring enhanced crack control at larger stirrup spacing. The results indicate that designs with steel ties and larger diameter BFRP bars with closer stirrup spacing are most effective for applications requiring significant energy absorption under compressive loads. Conversely, BFRP ties may provide a feasible alternative for comparable advantages, particularly in configurations with low confinement demands. This knowledge can direct the choice of reinforcement configurations and types to maximize the durability and compressive resilience of RC structures in a range of structural applications.

3.6 Post-peak fracture energy

The fracture energy (G_F) quantifies the energy dissipation during the fracture growth in concrete components and the failure procedure after the P_max. The G_F was calculated from the point P_max to the ultimate failure point under the load-displacement curve. It’s an essential metric for comprehending concrete’s post-peak behavior, especially in loaded structural components. This crucial fracture mechanics parameter should be explored comprehensively and employed in the structural design process to guarantee that structures can resist anticipated loading conditions without failing catastrophically. This section analyzes the post-peak behavior results of RC cylinders with various reinforcement arrangements in terms of G_F. The G_F results of groups RC-1 - RC-5 are given in Table 6.

The group RC-1 specimens experienced improved confinement and resistance to crack propagation following the P_max due to the increased G_F with closer BFRP spiral spacing. In this case, ACI-B(S)6–45 offered 15.5 % and 14.7 % higher G_F than ACI-B(S)6–60 and ACI-B(S)6–90, respectively. The specimens BS-B(S)6–45 gained 15.8 % and 19.3 % higher G_F than BS-B(S)6–60 and BS-B(S)6–90, respectively. Similarly, the specimens JGJ-B(S)6–45 exhibited 28.6 % and 14.3 % higher G_F than JGJ-B(S)6–60 and JGJ-B(S)6–90, respectively. The G_F results of group RC-2 demonstrated that BFRP ties offer increased G_F at intermediate tie spacing (60 mm) in contrast to the closer tie spacing of 45 mm, which achieved lower G_F and G_Acc due to the unexpected post-peak failure despite achieving the highest CS. The RC cylinders with 45 mm BFRP tie spacing demonstrated decreased G_F compared to the 60 mm arrangement, contrary to usual confinement patterns. In this case, the dense ties reduced post-peak toughness by causing localized stiffness zones that either caused premature microcracking or stress concentrations that limited the development of distributed microcracks. It was found that the ACI-B(T)6–60 achieved 11.6 % and 20.2 % higher G_F than ACI-B(T)6–45 and ACI-B(T)6–90, respectively. The same trend was found in BS RC specimens, i.e., BS-B(T)6–60 offered 33 % and 22.3 % higher G_F than BS-B(T)6–45 and BS-B(T)6–90, respectively. The specimens JGJ-B(T)6–60 exhibited 27.9 % and 50.6 % higher G_F than JGJ-B(T)6–45 and JGJ-B(T)6–90, respectively.

In the group RC-3 specimens, maximum G_F was attained with moderate (60 mm) and close (45 mm) BFRP spiral spacings while G_F decreased significantly at larger spacings (90 mm). Based on the results, it appeared that closer BFRP confinement is needed to maintain optimal fracture resistance, as the BFRP spiral efficacy decreases with wider distances. In this case, the specimens ACI-B(S)6-45-SB6 and ACI-B(S)6-60-SB6 achieved almost the same G_F but ACI-B(S)6-90-SB6 showed about 50 % decline in the post-peak G_F. The G_F results of group RC-4 demonstrated that larger diameter BFRP bars enhance stiffness; however, the findings imply that wider BFRP spiral spacing can decline the G_F, indicating the necessity for a balance between BFRP bar size and spacing to achieve effective confinement. In this configuration, the ACI-B(S)8–60 showed the highest G_F, i.e., 10.3 % and 27.8 % higher than ACI-B(S)8–45 and ACI-B(S)8–90, respectively. In the group RC-5, the use of steel ties offered superior confinement, enhancing G_F by effectively limiting lateral expansion, hence enabling the concrete to withstand greater fracture stresses. In this case, the ACI-S(T)6–45 and ACI-S(T)6–60 offered nearly the same G_F with a variation of 3.4 % but ACI-S(T)6–90 showed 34.1 % decreased G_F than ACI-S(T)6–60. The graphical results of G_F of group RC-1 to group RC-5 are illustrated in Figures 16, 17, 18, 19, 20, and 21.

Figure 16:

G_F results of ACI-RC: 6 mm BFRP spirals/ties.

Figure 17:

G_F results of BS-RC: 6 mm BFRP spirals/ties.

Figure 18:

G_F results of JGJ-RC: 6 mm BFRP spirals/ties.

Figure 19:

G_F results of ACI-RC: 6 mm BFRP spirals with steel longitudinal rods/bars.

Figure 20:

G_F results of ACI-RC: 6 & 8 mm BFRP spirals.

Figure 21:

G_F results of ACI-RC: 6 mm BFRP/steel ties.

The results of G_F of group RC-1 with 6 mm BFRP spirals and group RC-2 with 6 mm BFRP ties were compared. The results indicated that specimens of RC-2 generally yielded higher G_F than specimens of RC-1, particularly the specimens with the 60 mm ties spacing. The ACI-B(T)6–45, ACI-B(T)6–60, and ACI-B(T)6–90 achieved 6.6 %, 30.3 %, and 11.7 % higher G_F than ACI-B(S)6–45, ACI-B(S)6–60, and ACI-B(S)6–90, respectively. In the case of BS RC specimens, the BS-B(S)6–45 exhibited 19.4 % higher G_F than BS-B(T)6–45 wihile BS-B(T)6–60 and BS-B(T)6–90 offered 30.1 %, 12.7 % higher post-peak G_F than BS-B(S)6–60 and BS-B(S)6–90, respectively. In the case of JGJ RC specimens, the JGJ-B(S)6–45 and JGJ-B(S)6–90 dissipated 24.6 %, 39.7 % higher G_F than JGJ-B(T)6–45 and JGJ-B(T)6–90 respectively while the specimens JGJ-B(T)6–60 acieved 31.8 % higher G_F than JGJ-B(S)6–60. The specimens of group RC-3 achieved higher G_F in closer and intermediate BFRP spiral spacings (45 mm & 60 mm) but dropped sharply at wider spacing (90 mm). The G_F results showed that ACI-B(S)6-45-SB6 and ACI-B(S)6-60-SB6 exhibited 3.7 % and 20.8 % higher G_F than ACI-B(S)6–45, ACI-B(S)6–60, respectively but in the case of larger spiral spacing, i.e., ACI-B(S)6–90 offered 39.2 % higher G_F than ACI-B(S)6-90-SB6. The larger bar diameter in RC-3 enhanced G_F under more confined conditions, but it became less effective at wider spacings where the loss of confinement allowed more crack propagation. The G_F results of groups RC-1 and RC-4 showed that ACI-B(S)6–45 and ACI-B(S)8–45 achieved nearly the same G_F. The ACI-B(S)8–60 offered 23.6 % higher G_F than ACI-B(S)6–60 while ACI-B(S)6–90 offered 6.5 % higher G_F than ACI-B(S)8–90. The group RC-4 specimens with 8 mm BFRP bars, demonstrated relatively consistent G_F but did not outperform RC-1 specimens by a large margin except in intermediate configurations. The larger BFRP bar diameter in RC-4 added stiffness, but without closer confinement, it did not significantly increase G_F beyond RC-1’s performance, except at optimal spacings. Group RC-5 specimens generally achieved higher G_F than group RC-2, especially in the case of 45 and 60 (mm) tie spacing configurations. The G_F results showed that ACI-S(T)6–45 and ACI-S(T)6–60 achieved 19.1 % and 11.6 % higher G_F than ACI-B(T)6–45 and ACI-B(T)6–60, but in the case of 90 mm tie spacings, the specimens ACI-B(T)-6-90 offered 6.6 % higher G_F than ACI-S(T)6–90. Steel ties in RC-5 outperformed BFRP ties in RC-2 in the case of close and medium tie spacings, by providing more rigid confinement, preventing lateral expansion and enhancing energy absorption during post-peak fractures.

The G_F results underscore the critical influence of reinforcement type, BFRP bar diameter, and configuration of transverse reinforcements on the fracture resistance of RC cylinders. Among the evaluated programs, group RC-2 and RC-5 stand out for their superior G_F results. Group RC-4 also demonstrated enhanced G_F, especially in configurations with intermediate stirrup spacing, where the balance of bar stiffness and confinement was optimal. The utilization of BFRP stirrups can be a good substitute of steel stirrups in the RC columns, as the specimens transversally reinforced with BFRP bars exhibited comparable compressive performance across the different CMD codes. In essence, for applications requiring high G_F and resistance to cracking under compressive loading, configurations with steel ties (RC-5) or larger diameter BFRP bars with optimal spacing (RC-2 and RC-4) are recommended and should be explored further on large scale. These setups not only improve fracture resistance but also contribute to the durability and resilience of concrete structures, particularly in demanding structural environments where cracking and failure resistance are paramount.

3.7 Ductility characteristics

From the load-displacement curves, the ductility index of RC cylinders was calculated using the formula β = Δ 85 % Δ   P max where β is the ductility index coefficient, Δ85 % is the axial displacement at 85 % of P_max, and ΔP_max is the axial displacement at P_max. The ductility findings are presented in Table 6. The ductility results provided insights into the deformability of RC cylinders with varying reinforcement configurations. The ductility of the RC cylinders (groups RC-1 to RC-4) ranged from 1.09 to 1.37, indicating that variations in reinforcement materials and configurations affect ductility. The ductility of RC cylinders was found to be influenced by the reinforcement type, stirrup bar diameter, and rib spacing. The RC cylinders with concentrated and medium stirrup spacings, i.e., 45 and 60 (mm) demonstrated elevated ductility values. It indicates that concentrated transverse reinforcement enhances the material’s capacity to deform following the P_max. The RC cylinders having BFRP transverse reinforcement (spirals/ties) exhibited moderate ductility and controlled deformability due to better confinement. RC cylinders of (group RC-5) showed a little higher ductility, particularly in 45 and 60 (mm) spacings, suggesting better post-peak deformation than BFRP ties (group RC-2) under specified circumstances. The ductility also showed a notable increase with medium BFRP spiral spacing of larger diameter BFRP bar (group RC-4). Figure 22 shows that close and medium spiral/tie spacings have higher ductility than larger stirrup spacings. RC cylinders with higher ductility (i.e., close to 1.37), which showed a higher ability to absorb energy and deformation before complete collapse, are preferable in low seismic zones where deformability is crucial. The reinforcement type, arrangement, and stirrup spacing are essential in determining the ductility of RC cylinders. The observed differences in ductility assessments provided a foundation for optimizing reinforcement systems to satisfy particular structural requirements for improved ductility.

Figure 22:

Ductility index of RC cylinders (group RC-1 – group RC-5).

4.1 Model geometry and configuration

The finite element models were developed to match the tested specimen dimensions and reinforcement configurations described in Section 2. Each RC column model had a diameter of 150 mm and a height of 300 mm. The transverse reinforcement was modeled as continuous BFRP spirals, in accordance with the experimental arrangements. Longitudinal reinforcement was modeled as steel rods (∅3 mm) or steel bars (∅6 mm) depending on the group type. The model incorporated an explicit representation of the concrete core, longitudinal reinforcement, transverse reinforcement, and the thin BFRP sheet wraps at the exterior top and bottom 50 mm sections of the columns. To ensure correct interaction between components, the longitudinal and transverse reinforcements were embedded within the concrete elements. The 3D models for the cylindrical specimens and internal reinforcements arrangement of groups RC-1 and RC-3 are illustrated in Figure 23 (a & b).

Figure 23:

Details of the FE model: (a) concrete solid part and (b) reinforcement truss elements.

4.2 Material models

Concrete behavior was simulated using the Concrete Damaged Plasticity (CDP) model in ABAQUS, which accounts for nonlinear compressive and tensile behavior, stiffness degradation due to damage, and different failure mechanisms under varying stress states. The key CDP parameters, derived from experimental cube compression tests (Table 5) and literature values for similar concrete grades, were as follows: dilation angle (ψ) of 30°, eccentricity (ε) of 0.1, fb0/fc0 ratio of 1.16, K (ratio of second stress invariant) of 0.667, and a viscosity parameter of 0.0001 [58], 59]. The compressive stress–strain curve was defined from the experimental results of ACI CMD, with post-peak softening calibrated to match the observed ductility and G_F values.

Longitudinal steel bars were modeled as elastic–plastic materials with isotropic hardening, using experimentally measured yield strengths (Table 2) and an elastic modulus of 200 GPa. Poisson’s ratio was set to 0.3. BFRP spirals and sheets were modeled as linearly elastic up to rupture, consistent with their brittle tensile failure observed experimentally. Elastic modulus and tensile strength were taken from Table 2. The rupture strain was computed as tensile strength divided by elastic modulus. No compression stiffness contribution was included for BFRP bars in the CDP model.

4.3 Meshing strategy

The concrete domain was discretized using C3D8R (8-node linear brick, reduced integration) solid elements to improve computational efficiency while maintaining accuracy. Reinforcement bars were modeled with T3D2 (2-node linear truss) elements. Mesh sensitivity analysis indicated that an element size of approximately 15 mm for the concrete and 10 mm for the reinforcement provided stable results without excessive computational demand.

4.4 Interaction and bonding assumptions

A perfect bond between reinforcement and surrounding concrete was assumed using the embedded region constraint in ABAQUS. The BFRP sheet wraps were modeled as surface layers tied to the concrete in the wrapped regions, assuming full composite action with the underlying core [60], 61].

4.5 Boundary conditions and loading setup

The base of each model was fixed in all translational and rotational degrees of freedom to replicate the rigid platen in the test setup. The load was applied to the top face of the specimen as a displacement-controlled vertical movement, matching the experimental displacement control method. To ensure uniform load transfer, a rigid analytical surface was coupled to the top face of the concrete model.

4.6 Model validation

The FEM models were validated against the experimental results by comparing the predicted CS values and the observed damage modes. Table 7 presents the comparison between experimental and FEM-predicted CS values. Across all specimens, the difference between the simulated and experimental CS values was within 7 %, demonstrating strong agreement and confirming that the modeling approach accurately captured the axial capacity of BFRP-confined RC columns. The highest deviation occurred for the specimen ACI-B(S)6–45 with a difference of 6.87 %, while the smallest deviation (2.23 %) was observed for ACI-B(S)6-45-SB6.

In terms of damage patterns, the FEM simulations reproduced the key failure characteristics observed experimentally. In both RC-1 and RC-3 models, vertical cracking initiated near the mid-height region, followed by progressive widening of cracks and localized concrete crushing. The simulations also captured the spalling of the concrete cover and the concentration of damage in the central region of the column, corresponding to the experimentally observed rupture locations of BFRP spirals and the buckling zones of longitudinal steel bars. For specimens with closer spiral spacing (45 mm), the numerical damage contours indicated a more uniform distribution of cracking and delayed crushing, while specimens with larger spacing (90 mm) exhibited concentrated damage zones, consistent with reduced confinement efficiency seen in the tests (Figure 24).

The close agreement in CS values and the accurate replication of the failure modes confirm the reliability of the developed FEM framework in predicting the compressive behavior and damage progression of BFRP-confined RC columns. This validated model provides a sound basis for conducting further parametric studies and design optimizations (Figure 24).

Figure 24:

Failure modes of FE model of: (a) group RC-1 cylinders and (b) group RC-3 cylinders.