Abstract: Wear is commonly observed in the interaction between various objects, and accurately and efficiently simulating the contact wear behavior of these objects is crucial for studying the hazards of wear and its prevention. Based on a penalty function contact algorithm and the Archard wear model, combined with the Coulomb friction model that considers constant, linear and exponential pressure-dependent friction coefficients, incorporating complex nonlinear factors such as geometric nonlinearity, material nonlinearity, and boundary nonlinearity into the numerical simulation method for contact wear. To improve the efficiency and accuracy of the variable friction coefficient model, this paper proposes an adaptive penalty factor (VFAPF) algorithm and applies it to the simulation and analysis of contact wear behavior in linear elastic materials and Mooney-Rivlin hyperelastic materials. The results show that compared with the APF algorithm, the VFAPF algorithm possesses better accuracy stability in the variable friction large slip problem, and when the elastic slip is dominant, although there is a small decrease in the efficiency (about 12 %), the improvement of the accuracy is more obvious (about 62 %). For the small-sliding contact wear problem of linear elastic material, despite differences in wear gap and stress distribution among the three friction coefficient models, their effects on the wear location and contact pressure are minimal. In contrast, in the large-sliding contact wear problem for Mooney-Rivlin hyperelastic material, the three friction coefficient models show significant differences in wear gap, wear location, contact pressure, and stress distribution, with the linear friction coefficient model exhibiting the most pronounced variation.