Numerical analysis of thermo-mechanical response in wheel–brake shoe systems: friction power vs energy conversion methods under extreme downhill braking

Purpose Accurate prediction of wheel tread and brake shoe surface temperatures under constant-speed braking on ultra-long downhill slopes is critical for railway safety. There is currently no numerical solution that balances both computational accuracy and efficiency and is suitable for ultra-long downhill ramps. This study aims to innovatively propose a reduced-order modeling strategy that balances computational accuracy and efficiency, developing a novel numerical model in ANSYS APDL with parameterized modeling and establishing analytical heating and cooling models in MATLAB to describe frictional heat generation and convective dissipation. Design/methodology/approach A coupled MATLAB–ANSYS APDL simulation framework is implemented to efficiently simulate thermal responses under varying braking speeds (20–120 km/h), axle loads (20–25 t), slope gradients (20‰–30‰) and slope lengths (10–20 km). A systematic comparison between the Friction Power Method (FPM) and Energy Conversion Method (ECM) reveals differences in temperature predictions and sensitivities to operational parameters. Findings Results show that ECM predicts consistently higher temperatures than FPM, with percentage differences in wheel tread temperature increasing from 3.7% to 16.3% as speed rises, while differences decrease under high axle load and steep gradient conditions, indicating convergence behavior. The analysis demonstrates that FPM is suitable for high-precision temperature field prediction and thermal damage risk assessment, whereas ECM is advantageous for rapid estimation and preliminary scheme evaluation, reflecting a complementary compensatory relationship between accuracy and efficiency. Originality/value Overall, this study develops a novel ANSYS APDL–MATLAB co-simulation framework for thermo-mechanical analysis of wheel–brake shoe systems under ultra-long downhill braking. Unlike conventional axisymmetric and harmonic reduced-order models, it enables flexible heat source modeling and efficient parametric studies across multiple operational conditions via a 1/6 circumferential strategy, thereby balancing accuracy and computational efficiency for design and thermal safety assessment.