Abstract Recycled aggregate concrete‐filled steel tubes consist of outer steel tubes filled with recycled aggregate concrete, produced from aggregates obtained by crushing construction and demolition waste. These structural members are designed to enhance the performance of recycled aggregate concrete, as its mechanical properties improve when cast into circular steel tubes, creating a 3D stress state. Under axial compression, these columns exhibit increased structural capacity due to the confinement effect of the outer steel tube and improved ductility in post‐peak behaviour. This paper presents a numerical modelling scheme to simulate the structural response of these still‐pioneering composite elements. It details the computational procedure used to calibrate the constitutive model of recycled aggregate concrete, essential for accurate numerical analysis of these columns. The proposed concrete model modifies the stress‐strain relationships defined in Eurocode 2. The paper also discusses selection of a non‐linear model for steel based on the material behavior defined in Eurocode 3. The procedure involves developing a 3D finite element model in ABAQUS via a Python script, combined with Bayesian optimisation in MATLAB. Finally, the efficiency of the proposed approach is demonstrated through the application of machine learning models to predict the axial capacity of these columns, enabling rapid and accurate estimation for any given specimen.
Nikolić et al. (Mon,) studied this question.