Abstract Understanding and predicting wear evolution in composite piston rings for Stirling engines is challenging due to multiscale rough surface contact and dynamic transfer film formation. This study develops a unified wear prediction framework integrating fractal contact theory with visualized transfer film characterization. Fractal parameters (D, G) capture initial asperity-dominated contact, while a transfer film index (TFI) is synthesized from coverage, connectivity, and porosity with weights ωC = 0.573, ωk = 0.334, ωH = 0.093 to quantify film structure and load-bearing capacity. Experimental results show the transfer film evolves from discrete to dense-continuous, with TFI increasing from 0.0427 to 0.8083. TFI correlates negatively with the friction coefficient (R2 ≈ 0.91) and governs wear rate attenuation, where TFI ≥ 0.7 indicates a stable low-wear regime. Incorporating TFI improves wear rate prediction accuracy from R2 = 0.9149 to 0.9545 (relative error: 4.85~31.48%), which captures the full wear evolution. The framework elucidates the transition from asperity contact to transfer film-dominated lubrication, and its forward-validation against extended-time experimental data demonstrates its potential for reliable lifetime assessment of composite piston rings in Stirling engines.
Yang et al. (Mon,) studied this question.