Bond behaviour between H-shaped steel and concrete after freeze-thaw cycles

Specimen failure mode

Xie et al.⁴³ conducted pull-out tests on steel-reinforced concrete and identified three main failure modes: (1) Bond-splitting failure. Upon reaching approximately 80% of the ultimate load, vertical cracks begin to appear on the inner side of the protective layer at the edge of the steel flange. As the load further increases, these cracks gradually expand outward at approximately a 45° angle or parallel to the direction of the profiled steel, and extend along its length. (2) Bond-anchorage failure, also known as push-out or pull-out failure. Upon reaching approximately 80% of the ultimate load, microcracks perpendicular to the flange’s edge begin to appear in the marginal protective layer of the steel flange. Shear cracks on the inner side of the steel’s protective layer predominantly extend along the entire length of the steel within a narrow thickness range, while also slightly expanding towards the outer side of the protective layer. (3) Profiled steel yielding failure, which occurs after the profiled steel itself yields.

According to the test results, it can be seen that the failure mode did not undergo a significant change after a lower number of F–T cycles. The specimens consistently exhibited a failure mode of push-out for steel sections with an anchorage length of 400 mm embedded in concrete after undergoing various F–T cycles. Figure 8 presents the typical push-out failure cross-sections of the profiled steel embedded in concrete after 0, 25, and 50 F–T cycles. From Fig. 8, it can be seen that before the F–T cycle, the crack at the top of the specimen is relatively small, primarily oriented perpendicular to the flange, indicating that the structural integrity of the concrete is maintained in the initial state, and its failure mode is primarily push-out failure. After 25 F–T cycles, vertical splitting cracks appeared at the top of the specimen, caused by F–T damage, and extended from the top toward the sides, forming through cracks. This finding suggests that the specimen’s failure pattern progresses from crack development to initial degradation in bonding performance as the number of F–T cycles increases. After 50 F–T cycles, small initial cracks were observed at the end of the top flange. The crack development of the specimen was still relatively limited, but the damage to the bottom concrete was more severe. Although the bond performance further decreased, crack development was not pronounced, and the failure mode remained push-out failure. Therefore, F–T cycles have a particular impact on the failure mode at the interface between H-shaped steel and concrete. As the number of F–T cycles increases within a specific range, the interfacial bond performance gradually decreases, and the crack expansion pattern exhibits different characteristics, but the failure mode does not change.

Guo⁹ conducted 0, 50, 100, and 200 F–T cycle push-out tests on concrete specimens with profiled steel anchorage lengths of 380 mm and 560 mm and volumetric stirrup ratios of 0.4%, 0.8%, and 1.2%. For specimens with a volumetric stirrup ratio of 0.4%, the failure mode was push-out failure after 50 F–T cycles. For specimens with a volumetric stirrup ratio of 0.8% and a profiled steel anchorage length of 380 mm, the failure mode was splitting failure before F–T cycling, which changed to push-out failure after 100 and 200 F–T cycles. However, after 50 F–T cycles, one of the three specimens exhibited splitting failure, while the remaining two specimens exhibited push-out failure. For specimens with a volumetric stirrup ratio of 0.8% and a profiled steel anchorage length of 560 mm, splitting failure was observed before F–T cycles, and it changed to push-out failure after 50, 100, and 200 F–T cycles.

Bond strength

In this study, the average bonding stress at the interface between steel and concrete is defined as:

where τ represents the average bond stress in units of MPa; P is the load applied at the loading end in units of kN; C is the circumference of the profiled steel cross-section in units of mm; and l is the bond length of the profiled steel in units of mm.

The bond strength test results between the profiled steel and the concrete after F–T cycling are presented in Table 2. The specimen numbering follows the format “F200V08L560-3”, where “F200” indicates 200 freeze–thaw cycles, “V08” denotes a volumetric stirrup ratio of 0.8%, “L560” specifies an anchorage length of 560 mm, and “-3” represents the third parallel specimen in the group. An asterisk (“*”) indicates data obtained from the present experiments, while the remaining values were adopted from reference⁹; however, all the failure modes were re-examined and consistently reclassified in this study.

Each configuration was tested with 1 to 3 specimens, depending on material availability and experimental feasibility. For groups with multiple specimens, the bond strength is reported as the mean ± standard deviation; for single specimens, the measured value is presented directly without deviation.

The influence of F–T cycles on bonding stress

The impact of F–T cycling on the bond strength between profiled steel with a volumetric stirrup ratio of 0.8% and concrete with anchorage lengths of 380 mm and 560 mm is illustrated in Fig. 9. Table 3 shows the percentage decrease in the bond strength of profiled steel embedded in concrete with anchorage lengths of 380 mm and 560 mm as the number of F–T cycles increases. From Fig. 9 and Table 3, it can be seen that as the number of F–T cycles increases, the bond strength decreases. The specimens with a profiled steel anchorage length of 380 mm show a greater reduction in bond strength than those with an anchorage length of 560 mm. After 50 F–T cycles, the bond strength between the steel shapes with anchorage lengths of 380mm and 560mm and the concrete decreased to 82.2% and 84.7% of their initial strength, respectively. After 100 F–T cycles, the bond strength between the steel shapes with anchorage lengths of 380mm and 560mm and the concrete decreased to 75.3% and 78.6% of its initial strength, respectively. After 200 F–T cycles, the bond strength decreased to 67.2% and 71.7% of the precycle strength for anchorage lengths of 380 mm and 560 mm, respectively.

The influence of F–T cycles on the bond strength of steel-reinforced concrete

According to the experimental results of this study, after 25 and 50 F–T cycles, the bond strengths of the profiled steel-concrete with an anchorage length of 400 mm were 98.8% and 64.3%, respectively, of their initial value before the F–T cycling. Yao et al.³⁹ used the same experimental method to investigate the bond strength after 0, 50, and 100 F–T cycles for profiled steel with anchorage lengths of 380 mm and 560 mm. The conclusion remains the same: with an increase in the number of F–T cycles, the decline in the specimen’s bond strength intensifies. As the number of F–T cycles increases, the accumulation of damage leads to the expansion and penetration of cracks in the concrete⁴⁴ which reduces the chemical bonding force, mechanical interlock, and friction of the contact surface between the steel and the concrete to varying degrees⁹thus reducing the bonding stress between the steel and the concrete.

Potential sources of error include variations in specimen preparation, such as mixing and reinforcement placement, which could affect bond strength measurements. Inconsistent material distribution or specimen geometry could also introduce variability. Fluctuations in environmental conditions, particularly temperature and humidity, may affect hydration and strength development, thereby influencing bonding performance. Meanwhile, the brief temperature fluctuation (± 2 °C) of the F–T test device may prolong a single F–T cycle, thereby affecting the degradation of the bond stress between the steel and concrete after F–T.

Influence of stirrup volumetric stirrup ratio on bond strength

The impact of the volumetric stirrup ratio on the bond strength between the profiled steel with an anchorage length of 560 mm and the concrete after 0 and 100 F–T cycles is shown in Fig. 10. As illustrated in Fig. 10, the bond strength tends to increase with the volumetric stirrup ratio. At the same volumetric stirrup ratio, the bond strength decreases as the number of F–T cycles increases. Table 2 shows that after the same number of F–T cycles, the push-out force and bond strength between the profiled steel and concrete with a higher volumetric stirrup ratio are greater than those with a lower volumetric stirrup ratio. For example, before undergoing F–T cycles, the bond strength of a specimen with a volumetric stirrup ratio of 1.2% is 1.08 (1.21) times that of a specimen with a volumetric stirrup ratio of 0.8% (0.4%). This is because the presence of stirrups provides lateral restraint for the interface between steel and concrete. The confining effect of stirrups increases the lateral pressure on the concrete, which helps resist crack formation and reduces the likelihood of debonding. This confinement effect indirectly improves the slip resistance between steel and concrete, leading to increased bond strength. After 100 F–T cycles, the bond strength of the specimen with a 1.2% volumetric stirrup ratio is 1.03 (1.06) times that of the specimen with a 0.8% (0.4%) volumetric stirrup ratio.

Influence of volumetric stirrup ratio on bond strength.

According to the research findings of Guo⁹ the bond strength between profiled steel and concrete with a volumetric stirrup ratio of 0.4% was 82.7% of that with a 1.2% before undergoing F–T cycling. However, after 100 F–T cycles, the bond strength between the profiled steel and concrete with a volumetric stirrup ratio of 0.4% reached 94.3% of that with a volumetric stirrup ratio of 1.2%. This discovery reveals a significant difference in the bond strength between steel and concrete with different volumetric stirrup ratios before and after undergoing F–T cycles. After F–T cycling, specimens with a higher volumetric stirrup ratio exhibited a greater reduction in bond strength than did those with a lower volumetric stirrup ratio. Yao et al.³⁹ used a similar experimental method to explore the relationship between the thickness of the steel protective layer (60 mm, 90 mm, and 120 mm) and the bond strength of concrete after undergoing 100 F–T cycles. The research results show that as the number of F–T cycles increases, the bond strength of specimens with a protective layer thickness of 60 mm, 90 mm, and 120 mm decreases to 80.1%, 80.1%, and 87.3% of their original values, respectively. This phenomenon indicates that an increase in the thickness of the steel protective layer can effectively mitigate the damage caused to concrete specimens by F–T cycling.

Bond stress-slip curve

The measured bond stress-slip curves of steel embedded in concrete at different F–T cycles are shown in Fig. 11. Specifically, Fig. 11a shows the bond stress-slip curves obtained from this study for profiled steel embedded in concrete, while Fig. 11b shows the curves obtained from the literature⁹corresponding to the specimen marked with “#” in Table 2. A curve in Fig. 11b is obtained by averaging three specimens with the same number of F–T cycles. The bond stress-slip curves in Fig. 11 indicate that, after F–T cycles, both steel-reinforced concrete and recycled concrete exhibit similar behavior to untreated steel-reinforced concrete. Similar results have also been reported in the literature⁴⁵. The impact of F–T cycling on the bond performance between profiled steel and concrete is evident in the bond stress-slip curves. These changes occur across different phases: microslip, internal cracking, push-out, decline, and residual.

Bond stress-slip curve of specimens under different F–T cycles

(1) Before F–T cycling, the microslip phase of the bond stress-slip curve shows that when a small thrust is applied at the loading end of the profiled steel, the slip is not significant, and the curve maintains an approximately linear relationship. After the F–T cycles, the slope of this linear segment changes slightly, and the length of the bond stress-slip curve becomes shorter. (2) During the internal cracking stage, as the applied thrust increases, the free end of the steel begins to slip, indicating that the adhesion between the steel and concrete is approaching exhaustion. With further F–T cycles, slip can be observed even with the application of a smaller force. (3) Subsequently, the slip increases significantly, and the rising part of the bond stress slip curve exhibits nonlinear characteristics.4) With F–T cycling, during the bond stress-slip stage of profiled steel, when the applied thrust reaches the peak load, multiple longitudinal splitting cracks appear in the thinnest area of the concrete cover. With the F–T cycles, the peak load decreases, while the peak slip amount increases. 5) In the stage of bond stress-slip reduction before F–T cycles, the thrust applied to the specimen rapidly decreases, and the slip amount of the steel increases until the steel is completely pushed out and the loading stops. Under F–T cycling, the stress decreases more gradually. 6) During the residual phase of the bond stress‒slip, although the applied thrust decreases compared to the peak load, the slip amount significantly increases.

Under F–T cycles, the volume of water in steel-reinforced concrete specimens changes during freezing and thawing, leading to internal damage within the steel-reinforced concrete structures⁴⁶. As the number of F–T cycles increases, this internal damage expands and progresses. Consequently, a shorter length of the micro-slip section in the bond stress-slip curve indicates that the internal structure of the steel-reinforced concrete specimen becomes relatively loose. Both the freezing of water and the melting of ice crystals cause changes in the internal stress of the steel-reinforced concrete specimen, which can lead to cracking and eventual failure. An increase in F–T cycles decreases the peak load in the bond stress-slip curve corresponding to the slip stage. Jin et al.⁴⁷ developed a physical model to simulate the internal microstructure of concrete and used an equivalent circuit model to analyze electrochemical impedance spectroscopy results. The findings demonstrated that F–T cycles promote crack propagation within the concrete, leading to a decrease in continuous path resistance, as confirmed by microscopic observations using environmental scanning electron microscopy.