Abstract: To address the wear prediction problem of piston ring sealing surfaces under dry friction conditions in Stirling engines, a microstructural characterization of the piston ring sealing surface was performed based on three-dimensional fractal theory. A multi-scale contact model of the friction interface was developed, incorporating the wear mechanism at the interface. Additionally, a wear prediction model was constructed based on a modified Archard wear law. The accuracy of the model was subsequently validated through experimental testing. Using the validated model, an interaction analysis was conducted to examine the factors influencing the wear rate, revealing the specific influence of different operating conditions on the wear rate of sealing surfaces with varying roughness. Based on these findings, a genetic algorithm was applied to optimize the morphology of the piston ring sealing surface and its precompression amount to achieve optimal sealing performance. The results showed that, for given contact load, running speed, and precompression, there exists an optimal surface roughness that matches these parameters. When the contact pressure P₁ was 3.5 MPa, the running speed V was 2 m/s, and the precompression s was 0.021 4 mm, the optimal surface roughness Ra of the modified Polytetrafluoroethylene piston ring for the minimum wear was found to be 0.37 μm. The proposed wear model could provide decision-making support for selecting sealing components and for the operation and maintenance of Stirling engines. Moreover, the optimization results offered theoretical guidance for improving the efficiency, extending the lifespan, and guiding the material processing of sealing components in stirling engine applications.