Experimental study on the axial behavior of steel slag fine aggregate concrete stub columns

Abstract Axial compression tests were conducted on reinforced concrete stub columns to evaluate the effects of replacing natural sand with steel slag fine aggregate (SSFA) on axial behavior and axial compressive capacity prediction. Four design parameters were considered, namely the SSFA volumetric replacement ratio, concrete strength, longitudinal reinforcement ratio and transverse reinforcement ratio. The SSFA volumetric replacement ratios were 0%, 40%, and 100%. The results showed that increasing the SSFA replacement ratio generally improved the peak load and initial stiffness, except in the higher‐strength concrete group. For the control group, the peak load at replacement ratios of 40% and 100% increased by 32.39% and 41.42%, respectively, compared with that at 0% replacement, while the initial stiffness increased by 11.37% and 19.67%, respectively. However, increasing the SSFA replacement ratio generally led to more pronounced post‐peak softening. Some specimens with 100% SSFA exhibited a steeper post‐peak decline, indicating a more brittle post‐peak response. The ductility coefficients ranged from 1.03 to 1.41, indicating limited ductility. Comparisons with the predictions of the Chinese, American and European design codes showed that the Chinese code provided the closest agreement with the test results, although it remained conservative overall, mainly because the contribution of transverse confinement was not explicitly considered. Based on the Park confined concrete model, a modified equation for the axial compressive capacity of steel slag fine aggregate concrete stub columns was proposed by incorporating the strength enhancement of the core concrete. Its predictions showed better agreement with the test results and lower scatter than those of the current design codes. Finally, the test results confirmed that the three‐dimensional solid finite element model developed in MSC. Marc could reasonably reproduce the failure modes of the specimens. The mean absolute deviations of the peak load and peak displacement were 4.93% and 3.54%, respectively, confirming the validity of the numerical model.