<p>This study focuses on stabilizing the calculations involved in topology optimization for elastoplastic material models, specifically the optimization process which often faces challenges due to the discontinuity of the slope in the stress–strain curve at the yield point. Traditional approaches to topology optimization for elastoplastic material models sometimes suffer from stagnation at mechanically irrational solutions and fluctuations in design variables, caused by the non-smooth characteristic of the stress–strain relationship. Since gradient-based methods used in topology optimization require the evaluated functions to be differentiable, it is crucial to ensure that the stress–strain curve which effects the functions transitions smoothly between the elastic and plastic regions. To address this issue, this study proposes the use of the subloading surface model, which enables a smooth transition from the elastic to the plastic region, thereby stabilizing the optimization process. The objective function is formulated to maximize the energy absorption capacity of the structure, and sensitivity formulas are derived analytically based on the subloading surface model. The numerical examples show improved convergence performance, indicating the potential applicability of structural designs under other materially nonlinear constitutive models and geometrically nonlinear conditions.</p>