Static and Flexural Fatigue Behavior of GFRP Pultruded Rebars

This paper presents the experimental results of composite rebars based on GFRP manufactured by a pultrusion system. The bending and radial compression strength of rods was determined. The elastic modulus of GFRP rebars is significantly lower than for steel rebars, while the static flexural properties are higher. The microstructure of the selected rebars was studied and discussed in light of the obtained results—failure processes such as the delamination and fibers fracture can be observed. The bending fatigue test was performed under a constant load amplitude sinusoidal waveform. All rebars were subjected to fatigue tests under the R = 0.1 condition. As a result, the S-N curve was obtained, and basic fatigue characteristics were determined. The fatigue mechanism of bar failure under bending was further analyzed using SEM microscopy. It is worth noting that the failure and fracture mechanism plays a crucial role as a material quality indicator in the manufacturing process. The main mechanism of failure under static and cyclic loading during the bending test is widely discussed in this paper. The results obtained from fatigue tests encourage further analysis. The diametral compression test reflects the weakest nature of the composite materials based on the interlaminar compressive strength. The proposed methodology allows us to invariantly describe the experimental transversal strength of the composite materials. Considering the expected durability of the structure, the failure mechanism is likely to significantly improve their fatigue behavior under the influence of cyclic bending. The reasonable direction of searching for reinforcements of composite structures should be the improvement of the bearing capacity of the outer layers. In comparison with steel rebars (fatigue tensile test), the obtained results for GFRP are comparable in the HCF regime. It is worth noting that in the near fatigue endurance regime (2–5 × 106 cycles) both rebars exhibit similar behavior.


Introduction
Rapid growth in the civil engineering field has enforced the development of new technologies. Composite bars FRP (Fiber Reinforced Polymers) are used in the process of concrete structure reinforcement, and they are a good example of dynamically developed material in recent years. Węglowski [1] gives examples of the application of composite rebars in modern constructions, mostly as reinforcement in bridge decks, acoustic barriers, and underground reinforcement. Moreover, this reinforcement is widely used in constructions such as swimming pools, airports, parking lots, and soft-eye (cavity) walls in underground tunnels [2][3][4][5].
The composite rebars provide advantages including high specific strength; by applying various resins and fibers they may fit specific requirements. The composite rebars prices plummet due to the expansion of the manufacturing technologies, used more often in civil engineering. The expansion of composite rebars application significantly increased in the early 1990s. Initially, they were applied in Western Europe, the USA, Japan, and Canada [6]. The external constructions and projects based on this composite reinforcement showed profitable results in terms of many experiments and proved that composite rebars might be an alternative for steel bars, and a breakthrough should be expected in their applications [7]. concrete structures-flexion from the global deformation of the structural element, and transverse compression from local pressure transferred through the concrete. Additionally, the analysis of the failure and fracture mechanism of these rebars was investigated in terms of the base material used in the manufacturing process.

FRP Reinforcement
Composite rebars (FRP) are marked by fair mechanical properties such as great tensile strength, and physical properties such as lower density in comparison to steel reinforcement. They have instanced the deployment in the objects particularly exposed to aggressive environments, and such where the proper working enables electromagnetic indifference, including construction. To reach a high corrosion resistance and electromagnetic indifference, a suitable material should be used for manufacturing not only the longitudinal reinforcement (to carry the bending load) but also the reinforcement to carry the transverse forces [30]. The basic properties are presented in Table 1. According to the ACI 440.1R standard [31], FRP reinforcement is characterized by over 4-times lower density compared to steel, which significantly influences reinforcement works and decreases transportation costs. The thermal expansion is taken into account and the steel is an isotropic material, whereas the behavior of FRP rods is dependent on the fiber direction. The type of fiber decides on longitudinal thermal expansion of reinforcement, and the transverse thermal expansion depends on the resin type [17].
Composite rebars show elastic-brittle tensile strength characteristics. In contrast to the classic steel reinforcement, composite rebars are characterized by greater tensile strength oscillating between 600 and 3700 MPa, and relatively low Young Modulus between 35 and 125 GPa, excluding carbon fiber reinforcement polymer (CFRP). Table 2 presents selected typical mechanical properties in a parallel direction to fibers [17]. For the experimental campaign, steel rods were devoted, as well as composite rebars. The rods with a 10 mm diameter were used for experimental testing (steel and GFRP), as presented in Figure 1. For the experimental campaign, steel rods were devoted, as well as composite rebars. The rods with a 10 mm diameter were used for experimental testing (steel and GFRP), as presented in Figure 1. Composite rebars subjected to the experimental campaign are made from glass fibers in an epoxy resin matrix with the fibers content of 65-70%. Moreover, an over-wrap is applied over the plastic rods to increase the buckling and flexural strength [32,33]. The parameters of this rebar are shown in Table 3. For each conducted experiment, the samples needed to be cut to the proper length. An accurate cut was required to obtain the accuracy of the results as best as possible. The specimens were cut for specific length using metallographic cutters METACUT 302 (METKON, Bursa, Turkey). It cuts with the manual feed of the grinding wheel.
Moreover, the structural research using a scanning electron microscope required appropriately prepared samples including mounting and burnishing using grinder-polisher FORCIPOL 2V (METKON, Bursa, Turkey). The objective of mounting was to stabilize the sample in the neutral material (ELECTROMIX resin) allowing the controlled machining.
The microstructure of the composite reinforcement was investigated before and after the fatigue tests. For microscopic analysis specimens were selected from outer and central parts of rebars, as is shown schematically in Figure 2. The longitudinal and transverse microsections were prepared to investigate the variations in microstructures. Composite rebars subjected to the experimental campaign are made from glass fibers in an epoxy resin matrix with the fibers content of 65-70%. Moreover, an over-wrap is applied over the plastic rods to increase the buckling and flexural strength [32,33]. The parameters of this rebar are shown in Table 3. For each conducted experiment, the samples needed to be cut to the proper length. An accurate cut was required to obtain the accuracy of the results as best as possible. The specimens were cut for specific length using metallographic cutters METACUT 302 (METKON, Bursa, Turkey). It cuts with the manual feed of the grinding wheel.
Moreover, the structural research using a scanning electron microscope required appropriately prepared samples including mounting and burnishing using grinder-polisher FORCIPOL 2V (METKON, Bursa, Turkey). The objective of mounting was to stabilize the sample in the neutral material (ELECTROMIX resin) allowing the controlled machining.
The microstructure of the composite reinforcement was investigated before and after the fatigue tests. For microscopic analysis specimens were selected from outer and central parts of rebars, as is shown schematically in Figure 2. The longitudinal and transverse microsections were prepared to investigate the variations in microstructures.
For the composite material, quality evaluation is important to identify the microstructural features of composites like fiber distribution, evidence of voids and bubbles, fiber aggregation rate, as well as the measurement of the fibers diameters and its variability. Figure 3 shows the part of the rebar's cross-section, and Figure 4 gives the longitudinal section of the specimen. In the observed microstructure it can be stated the uniformity of fibers' distribution is relatively high, and the fibers are dominant without any inclusions. Comparable diameters of fibers warrant a good quality of fibers used by manufacturers. For the longitudinal cut sample, the SEM image shows 2 different parts, the outer (A, B), and the central (C, D) part of the rebar (Figure 4). Part A and B allow us to observe an overwrapped layer on the GFRP rod, in which fibers were cut transversally, and longitudinally in the major part of the rod. Other parts (C, D) show the distribution of the fibers in the central part of the rod. Also based on the measurements from Figure 5 it can be concluded that the glass fiber diameters vary from 15-17 µm. FORCIPOL 2V (METKON, Bursa, Turkey). The objective of mounting was to stabilize the sample in the neutral material (ELECTROMIX resin) allowing the controlled machining.
The microstructure of the composite reinforcement was investigated before and after the fatigue tests. For microscopic analysis specimens were selected from outer and central parts of rebars, as is shown schematically in Figure 2. The longitudinal and transverse microsections were prepared to investigate the variations in microstructures.  For the composite material, quality evaluation is important to identify the microstructural features of composites like fiber distribution, evidence of voids and bubbles, fiber aggregation rate, as well as the measurement of the fibers diameters and its variability. Figure 3 shows the part of the rebar's cross-section, and Figure 4 gives the longitudinal section of the specimen. In the observed microstructure it can be stated the uniformity of fibers' distribution is relatively high, and the fibers are dominant without any inclusions. Comparable diameters of fibers warrant a good quality of fibers used by manufacturers. For the longitudinal cut sample, the SEM image shows 2 different parts, the outer (A, B), and the central (C, D) part of the rebar (Figure 4). Part A and B allow us to observe an overwrapped layer on the GFRP rod, in which fibers were cut transversally, and longitudinally in the major part of the rod. Other parts (C, D) show the distribution of the fibers in the central part of the rod. Also based on the measurements from Figure 5 it can be concluded that the glass fiber diameters vary from 15-17 µ m.  For the composite material, quality evaluation is important to identify the microstructural features of composites like fiber distribution, evidence of voids and bubbles, fiber aggregation rate, as well as the measurement of the fibers diameters and its variability. Figure 3 shows the part of the rebar's cross-section, and Figure 4 gives the longitudinal section of the specimen. In the observed microstructure it can be stated the uniformity of fibers' distribution is relatively high, and the fibers are dominant without any inclusions. Comparable diameters of fibers warrant a good quality of fibers used by manufacturers. For the longitudinal cut sample, the SEM image shows 2 different parts, the outer (A, B), and the central (C, D) part of the rebar (Figure 4). Part A and B allow us to observe an overwrapped layer on the GFRP rod, in which fibers were cut transversally, and longitudinally in the major part of the rod. Other parts (C, D) show the distribution of the fibers in the central part of the rod. Also based on the measurements from Figure 5 it can be concluded that the glass fiber diameters vary from 15-17 µ m.     For comparative analysis, we selected B500SP steel. The number 500 in the steel designation indicates the value of the characteristic yield strength in MPa, and SP marks its weldability. This steel meets, among others, the requirements of the PN-B-03264:2002 standard and is classified as the highest strength class A-IIIN. Also, for steel rebars, the homogeneous steel structure typical for the B500SP type of heat-treated steel (outer surface) was exhibited after etching 3%HNO3, see Figure 6.

Bending Test
This research aimed to compare the flexural strength of composite (ANKRA LLC) and steel (B500SP) rebars. The calculations of flexural strength factors and Young modulus for both rods were carried out for comparison in accordance with the ASTM D790 [34] and ASTM D7957 [17]. Considering that three-point bending can cause damage to the composite material at the point of load application, it was decided to use four-point bending. In four-point bending, there are lower contact stresses at the point of load and thus constant bending moment without shear forces at the pure bending section, which will allow to verify the usefulness of both methods in the analysis of composite rods. Regarding the mentioned standards, the four-point bending, and three-point bending tests were conducted using the universal testing machine MTS 810 (MTS Systems, Corporation, Eden Prairie, MN, USA). This device is controlled by a special software MTS Flex Test For comparative analysis, we selected B500SP steel. The number 500 in the steel designation indicates the value of the characteristic yield strength in MPa, and SP marks its weldability. This steel meets, among others, the requirements of the PN-B-03264:2002 standard and is classified as the highest strength class A-IIIN. Also, for steel rebars, the homogeneous steel structure typical for the B500SP type of heat-treated steel (outer surface) was exhibited after etching 3%HNO 3 , see Figure 6.   For comparative analysis, we selected B500SP steel. The number 500 in the steel designation indicates the value of the characteristic yield strength in MPa, and SP marks its weldability. This steel meets, among others, the requirements of the PN-B-03264:2002 standard and is classified as the highest strength class A-IIIN. Also, for steel rebars, the homogeneous steel structure typical for the B500SP type of heat-treated steel (outer surface) was exhibited after etching 3%HNO3, see Figure 6.

Bending Test
This research aimed to compare the flexural strength of composite (ANKRA LLC) and steel (B500SP) rebars. The calculations of flexural strength factors and Young modulus for both rods were carried out for comparison in accordance with the ASTM D790 [34] and ASTM D7957 [17]. Considering that three-point bending can cause damage to the composite material at the point of load application, it was decided to use four-point bending. In four-point bending, there are lower contact stresses at the point of load and thus constant bending moment without shear forces at the pure bending section, which will allow to verify the usefulness of both methods in the analysis of composite rods. Regarding the mentioned standards, the four-point bending, and three-point bending tests were conducted using the universal testing machine MTS 810 (MTS Systems, Corporation, Eden Prairie, MN, USA). This device is controlled by a special software MTS Flex Test

Bending Test
This research aimed to compare the flexural strength of composite (ANKRA LLC) and steel (B500SP) rebars. The calculations of flexural strength factors and Young modulus for both rods were carried out for comparison in accordance with the ASTM D790 [34] and ASTM D7957 [17]. Considering that three-point bending can cause damage to the composite material at the point of load application, it was decided to use four-point bending. In four-point bending, there are lower contact stresses at the point of load and thus constant bending moment without shear forces at the pure bending section, which will allow to verify the usefulness of both methods in the analysis of composite rods. Regarding the mentioned standards, the four-point bending, and three-point bending tests were conducted using the universal testing machine MTS 810 (MTS Systems, Corporation, Eden Prairie, MN, USA). This device is controlled by a special software MTS Flex Test Console (Series 793, MTS Systems, Corporation, Eden Prairie, MN, USA). The working range of the hydraulic pulsator is equally ±100 kN with an additional load cell 5 kN. The specimen length in this test is equally 200 mm. For strain/deflection measurements, LVDT and MTS displacement gauges (clip-on gage with base 2.5 mm (+3/−1) mm travel arms-for elastic range) were used for registration of the deflection in load application points. Also, displacement signals were registered from the MTS 810 displacement transducer (cross head) of the tensile machine.

Radial Compression Test
This test was focused on the longitudinal tensile strength of FRP rebars. The length is a crucial parameter that influenced the value of stress, so the Weibull distribution was applied to normalize the obtained results. During the diametral compression test, the composite rebar was loaded perpendicular to its long axis. Also, in this experiment, the MTS 810 testing machine was equipped with compression platen as is shown in Figure 7. The preparation of the samples included their cut for a specific length. In this research, three various lengths (10 mm, 20 mm, and 30 mm) were investigated with three repetitions.
Console (Series 793, MTS Systems, Corporation, Eden Prairie, MN, USA). The working range of the hydraulic pulsator is equally ±100 kN with an additional load cell 5kN. The specimen length in this test is equally 200 mm. For strain/deflection measurements, LVDT and MTS displacement gauges (clip-on gage with base 2.5 mm (+3/−1) mm travel armsfor elastic range) were used for registration of the deflection in load application points. Also, displacement signals were registered from the MTS 810 displacement transducer (cross head) of the tensile machine.

Radial Compression Test
This test was focused on the longitudinal tensile strength of FRP rebars. The length is a crucial parameter that influenced the value of stress, so the Weibull distribution was applied to normalize the obtained results. During the diametral compression test, the composite rebar was loaded perpendicular to its long axis. Also, in this experiment, the MTS 810 testing machine was equipped with compression platen as is shown in Figure 7. The preparation of the samples included their cut for a specific length. In this research, three various lengths (10 mm, 20 mm, and 30 mm) were investigated with three repetitions.

Flexural Fatigue Test
The materials subject to static loads show greater strength than materials subjected to cyclic loads. This phenomenon is known as fatigue, and it is measurable by mechanical tests, which include repeating various stresses altering in a regular cycle from minimum to maximum value. Most fatigue testing machines use eccentric rotational weight for generating cyclic altering loads. To conduct this experiment, MTS 810 equipped with support for four-point bending was used. It accomplished the array measurement for the various value of the maximum load. The load used for the maximum load of composite rebars includes from 20-60% of force drives to failure of composite rebar obtain from bending tests. The frequency oscillates in the range of 1-1.5 Hz. In practical terms, fatigue testing of polymers, including composites, provides some difficulties, mainly related to the influence of frequency and thus generated heat. In the case of steel, the 1-10 Hz HCF (High Cycle Fatigue) range does not bring any changes. In the case of composites, thermal stability is no longer the same. Thus, the experimental observations with the use of a thermographic camera consisted of checking whether there is a significant difference in temperature during fatigue tests due to cyclic loading. In the presented experimental campaign, it was established that for low frequency f, no changes greater than 1-2 degrees C

Flexural Fatigue Test
The materials subject to static loads show greater strength than materials subjected to cyclic loads. This phenomenon is known as fatigue, and it is measurable by mechanical tests, which include repeating various stresses altering in a regular cycle from minimum to maximum value. Most fatigue testing machines use eccentric rotational weight for generating cyclic altering loads. To conduct this experiment, MTS 810 equipped with support for four-point bending was used. It accomplished the array measurement for the various value of the maximum load. The load used for the maximum load of composite rebars includes from 20-60% of force drives to failure of composite rebar obtain from bending tests. The frequency oscillates in the range of 1-1.5 Hz. In practical terms, fatigue testing of polymers, including composites, provides some difficulties, mainly related to the influence of frequency and thus generated heat. In the case of steel, the 1-10 Hz HCF (High Cycle Fatigue) range does not bring any changes. In the case of composites, thermal stability is no longer the same. Thus, the experimental observations with the use of a thermographic camera consisted of checking whether there is a significant difference in temperature during fatigue tests due to cyclic loading. In the presented experimental campaign, it was established that for low frequency f, no changes greater than 1-2 degrees C over a wide range of fatigue lifetimes were observed. For this purpose, the thermographic camera FLIR E6-XT (Wilsonville, OR, USA) was used. The exemplary thermogram obtained during cyclic loading is shown in Figure 8. over a wide range of fatigue lifetimes were observed. For this purpose, the thermographic camera FLIR E6-XT (Wilsonville, OR, USA) was used. The exemplary thermogram obtained during cyclic loading is shown in Figure 8.

Bending Results
During the experiment, the force, time, and displacement signals were registered. Representative force-displacement curves were plotted in  According to this date, the graphs were compiled for three-point bending (Figures 9 and 10), and four-pointbending (Figures 11 and 12) for steel, and composite rebar.

Bending Results
During the experiment, the force, time, and displacement signals were registered. Representative force-displacement curves were plotted in  According to this date, the graphs were compiled for three-point bending (Figures 9 and 10), and four-pointbending (Figures 11 and 12) for steel, and composite rebar.
All results from three-point bending and four-point bending tests are summarized in Table 4. over a wide range of fatigue lifetimes were observed. For this purpose, the thermographic camera FLIR E6-XT (Wilsonville, OR, USA) was used. The exemplary thermogram obtained during cyclic loading is shown in Figure 8.

Bending Results
During the experiment, the force, time, and displacement signals were registered. Representative force-displacement curves were plotted in Figures 9-12. According to this date, the graphs were compiled for three-point bending (Figures 9 and 10), and four-pointbending (Figures 11 and 12) for steel, and composite rebar.    All results from three-point bending and four-point bending tests are summarized in Table 4.     All results from three-point bending and four-point bending tests are summarized in Table 4.   In both 3PB and 4PB loading cases, the ductile behavior of B500SP material with a distinct plastic deformation and modulus of elasticity typical for structural steel is noted. The GFRP bar, on the other hand, has a higher brittleness with a local maximum force and a lower (about 3 times the modulus of elasticity). However, the final bending strength value results in higher results which is a good parameter in the application context.

Diametral Compression Test
All specimens during the diametral compression test failed in the central part of the rebar, fragmenting into 2 or 3 parts. Also, a flattening of the sample is noticeable, which did not return to the original shape. Figure 13 presents specimens used in the test before and after it. During the experiment, force-displacement curves were plotted and collected in Figures 14-16.

Diametral Compression Test
All specimens during the diametral compression test failed in the central part of the rebar, fragmenting into 2 or 3 parts. Also, a flattening of the sample is noticeable, which did not return to the original shape. Figure 13 presents specimens used in the test before and after it. During the experiment, force-displacement curves were plotted and collected in Figures 14-16.

Diametral Compression Test
All specimens during the diametral compression test failed in the central part of the rebar, fragmenting into 2 or 3 parts. Also, a flattening of the sample is noticeable, which did not return to the original shape. Figure 13 presents specimens used in the test before and after it. During the experiment, force-displacement curves were plotted and collected in Figures 14-16.       While the circular rod is subjected to diametral compression, the central zone of the cylinder is subjected to biaxial stress. The normal stress in the perpendicular direction to applied load is tension stress, while in the parallel direction to the applied force is compression stress. The tension and compression stress in the central zone of the rod can be calculated using Equations (1) and (2) [35]. where: -diametral compression strength, F-maximum force, D-diameter of the rod, L-length of the rod.
x-distance from damage surface. While the circular rod is subjected to diametral compression, the central zone of the cylinder is subjected to biaxial stress. The normal stress in the perpendicular direction to applied load is tension stress, while in the parallel direction to the applied force is compression stress. The tension σ x and compression σ y stress in the central zone of the rod can be calculated using Equations (1) and (2) [35].
where: σ x -diametral compression strength, F-maximum force, D-diameter of the rod, L-length of the rod.
x-distance from damage surface. The analysis of Equation (1) shows that the maximum stress appears perpendicular to the failure surface (x = 0). The virtual transverse failure stress during elongation represents the load at break was calculated using Equation (1) assuming x = 0. The results obtained from the calculations are shown in Table 5. The transverse tensile strength against specimen length is plotted in Figure 17 using linear regression and a 95% confidence level. The transverse tensile strength against specimen length is plotted in Figure 17 using linear regression and a 95% confidence level. According to the Weibull distribution, the probability F that the rod of L length undergoes failure influenced by stress σ is described by Equation (3): where 0 and are called scale and shape parameters. These two parameters may be determined by estimation of the highest probability. Especially, the parameters and 0 are connected with strength data, according to Equations (4) and (5): According to the Weibull distribution, the probability F that the rod of L length undergoes failure influenced by stress σ is described by Equation (3): where σ 0 and α are called scale and shape parameters. These two parameters may be determined by estimation of the highest probability. Especially, the parameters α and σ 0 are connected with strength data, according to Equations (4) and (5): For the investigated composite material, the scale parameter equals σ 0 = 14.262 and the shape parameter α = 15.715. The normalized strength was calculated using Equation (6): Transverse tensile strength was calculated by Equation (7): According to these formulas, Table 6 was composed. It shows the values of stress σ, normalized strength σ, and the probability of failure F(σ). The observed length effect on the probability of failure is noticeable-in Table 6 with increasing L, the probability of failure increases too. This seems to be consistent with the Weibull theory.

Fatigue Bending Test Results
During the fatigue test, the specimens were loaded by cyclic force ∆F, in which maximum values ranged from 20 to 60% of the critical bending stress calculated from the bending test 3PB/4PB. The experiment was conducted with a frequency of 1-1.5 Hz keeping a constant stress ratio R = σ min / σ max = 0.1. During the experiment, the number of cycles to failure (criterion-stiffness drop at 40%) was registered. For fatigue characterization, the S-N curves were drawn. The object of interest was the high cycle fatigue regime-which can be described by the power-law relationship: where: A, n-experimentally determined parameters in HCF regime (high cycle fatigue). According to the obtained results, the S-N curve for composite rebar was compiled ( Figure 18). The slope parameter m was calculated as: For the comparison of the fatigue properties, Szlachetka et al. [36] discussed the fatigue phenomenon of the ribbed steel rebars under tensile mode fatigue (using R = 0). Based on the data available in [36] the S-N curve with 95% confidence intervals is plotted in Figure 19. Also, statistical outputs are computed in Table 7. For the comparison of the fatigue properties, Szlachetka et al. [36] discussed the fatigue phenomenon of the ribbed steel rebars under tensile mode fatigue (using R = 0). Based on the data available in [36] the S-N curve with 95% confidence intervals is plotted in Figure 19. Also, statistical outputs are computed in Table 7. For the comparison of the fatigue properties, Szlachetka et al. [36] discussed the fatigue phenomenon of the ribbed steel rebars under tensile mode fatigue (using R = 0). Based on the data available in [36] the S-N curve with 95% confidence intervals is plotted in Figure 19. Also, statistical outputs are computed in Table 7. Based on Equations (8) and (9), statistical outputs are calculated in Table 7. It is noticeable that in a high cycle fatigue regime, there is a higher slope of the S-N curve for the steel material. Also noticeable is the slightly better fatigue resistance for higher stress levels, which is confirmed by the better ductility of the material. However, for fatigue lifetimes close to 10 6 cycles, GFRP exhibit similar fatigue lifetimes in comparison with steel. However, it is important to be particularly careful in drawing final comparative conclusions since both materials were tested with different conditions-the steel in tension and Based on Equations (8) and (9), statistical outputs are calculated in Table 7. It is noticeable that in a high cycle fatigue regime, there is a higher slope of the S-N curve for the steel material. Also noticeable is the slightly better fatigue resistance for higher stress levels, which is confirmed by the better ductility of the material. However, for fatigue lifetimes close to 10 6 cycles, GFRP exhibit similar fatigue lifetimes in comparison with steel. However, it is important to be particularly careful in drawing final comparative conclusions since both materials were tested with different conditions-the steel in tension and the composite in bending. However, the observed fatigue properties encourage further studies on the programmable structure of the composites-in pultrusion, it is possible to shape the geometry and cross-section in any way to ensure optimal fiber arrangement. Fatigue damage analysis using SEM will be an important part of the study. The fatigue fracture of B500SP material is a typical and well-known fatigue damagedas it is evidenced in [36]. Due to the higher interest in the preliminary fatigue study of GFRP rebars, it was extremely important to perform, after the fatigue tests, the microscopic analysis of damaged specimens. The selected samples were prepared for SEM (HITACHI S-3400N) investigations. The two samples were selected for examination of the damage mechanism of the composite rebars. The first one was subjected to the force of 1500 N, with frequency 1 Hz withstood 14 649 cycles. The force equal 1000 N, and frequency 1 Hz acted on the second sample withstood 76 636 cycles. Once more, the transverse, and longitudinal microsection of investigated rebars were taken. Figures 20 and 21 show the cross-section of the rebars after the experiment. Further, Figures 22 and 23 present the longitudinal microsection.
Cross-sectional images (Figures 20 and 21) show noticeable delaminations and fiber fracture in both specimens, however, it is more severe for the specimen subjected to the greater maximum load (Figure 20). The larger force likely drives this, inducing higher stress in fibers. The longitudinal microsections (Figures 22 and 23) give information about the zones where the fibers were elongated (A, B), and compressed (C, D). In the longitudinal section of both specimens, the fiber fracture can be observed even with the lowest magnitude, most evidently in the compressed part of the specimens. It is much more prominent in the specimen subjected to the greater maximum load (Figure 22). This kind of damage may be ascribed to fiber microbuckling, which is the typical behavior of reinforcing fibers after exceeding the critical compressive load (fibers behaving like slender columns embedded in a relatively soft matrix). This fiber buckling subsequently causes delamination. The fatigue fracture of B500SP material is a typical and well-known fatigue damaged-as it is evidenced in [36]. Due to the higher interest in the preliminary fatigue study of GFRP rebars, it was extremely important to perform, after the fatigue tests, the microscopic analysis of damaged specimens. The selected samples were prepared for SEM (HI-TACHI S-3400N) investigations. The two samples were selected for examination of the damage mechanism of the composite rebars. The first one was subjected to the force of 1500 N, with frequency 1 Hz withstood 14 649 cycles. The force equal 1000 N, and frequency 1 Hz acted on the second sample withstood 76 636 cycles. Once more, the transverse, and longitudinal microsection of investigated rebars were taken. Figures 20 and 21 show the cross-section of the rebars after the experiment. Further, Figures 22 and 23 present the longitudinal microsection.      Cross-sectional images (Figures 20 and 21) show noticeable delaminations and fiber fracture in both specimens, however, it is more severe for the specimen subjected to the greater maximum load (Figure 20). The larger force likely drives this, inducing higher stress in fibers. The longitudinal microsections (Figures 22 and 23) give information about the zones where the fibers were elongated (A, B), and compressed (C, D). In the longitudinal section of both specimens, the fiber fracture can be observed even with the lowest magnitude, most evidently in the compressed part of the specimens. It is much more prominent in the specimen subjected to the greater maximum load (Figure 22). This kind of damage may be ascribed to fiber microbuckling, which is the typical behavior of reinforcing fibers after exceeding the critical compressive load (fibers behaving like slender columns embedded in a relatively soft matrix). This fiber buckling subsequently causes delamination.

Conclusions
Based on the experimental research, the following conclusions can be drawn:

•
The elastic modulus of GFRP rebars is significantly lower than for steel rebars (approx. 3x). It is worth underlining the higher static flexural properties (similar for 3PB and 4PB test) of GFRP materials than high-quality B500SP steel (see Table 4).

•
Looking at the specimens with a higher scale (σ 0 ) loaded with a higher stress level, the failure processes such as delamination and fiber fracture can be observed. The higher stress level in fibers is influenced by the greater force, driving this phenomenon. In the longitudinal sections of investigated samples, even the lowest magnitude provides the locations of the fractures. The inferred failure mode is fiber buckling under compressive load exceeding the critical load for slender fibers.

•
The results obtained from fatigue tests (m = 4.33, A = 4692 MPa) encourage further analysis. Considering previous conclusions, the diametral compression test reflects the weaker nature of the composite materials based on the interlaminar compressive strength. The proposed methodology allows us to describe invariantly the experimental transversal strength of the composite materials. For tested GFRP rebars (epoxy resin) σ med varies from 17.38-18.60 MPa independent of the scale length.

•
Considering the expected durability of the structure, the failure mechanism is likely to significantly improve their fatigue behavior under the influence of cyclic bending. It was expected to obtain greater resistance in extreme fibers as a result of single line wrapping. It seems that the reasonable direction of searching for reinforcements of composite structures should be the improvement of the bearing capacity of outer layers.

•
In comparison with steel rebars (fatigue tensile test), the obtained results for GFRP are comparable in the HCF regime. It is worth noting that in near fatigue endurance regime (2-5 × 10 6 cycles) both rebars exhibit similar behavior.

•
Further studies considering the environmental influence on fatigue endurance both GFRP and steel rods are necessary regarding the possible anti-fatigue design of the geometry of the rebars using pultrusion process in order to redistribute damage stress in the outer part of the composite rods. Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.