Phenomenology of moderate interface debonding in concrete-filled steel tubular columns: Experimental and numerical assessment

Concrete-filled steel tubular (CFST) columns are recognized for their high load-bearing capacity, ductility, and energy dissipation, primarily due to the confinement effect provided by the steel tube on the core concrete. However, interface debonding, possibly induced by temperature gradients, shrinkage, creep, or axial loads, can significantly compromise this performance. Debonding can occur at different severity levels. While severe debonding (gap thicknesses > 3 mm) has been widely studied, the mechanical effects of moderate debonding remain less explored. This study investigates the role of moderate interface debonding through both experimental testing and advanced numerical modeling. Five CFST short columns were subjected to axial compression tests. Debonding was introduced using stainless steel inserts to precisely simulate various gap thicknesses (0.5 mm and 1.0 mm) and arc-length ratios (0.25 and 0.5). The results show that even moderate debonding can reduce axial capacity due to early loss of confinement in the concrete core. The study identifies three distinct response types associated with moderate debonding, depending on the possibility and extent of steel–concrete re-engagement. These were further investigated through a validated finite element model developed in Abaqus and benchmarked against existing literature. A response chart was established to characterize CFST performance as a function of debonding thickness and arc-length. The findings highlight: (i) a response comparable to fully bonded specimens when debonding is negligible, (ii) a transitional regime with two distinct load-bearing peaks resulting from delayed confinement recovery, and (iii) a degraded response with a single, early peak and reduced ductility under larger debonding configurations. The paper provides practical recommendations for evaluating the structural performance of CFST under moderate debonding scenarios. • Simulated small debonding gaps (0.5–1.0 mm) using removable steel inserts. • Identified three CFST response types based on debonding severity. • Dual-peak behavior observed in moderate debonding with delayed confinement. • Validated FEM captured stress redistribution and interface re-engagement. • Proposed a CFST response map based on debonding gap and arc-length ratio.