Design and evaluation method of erosion-resistant and wear-resistant CoCrFeMnNi high-entropy alloy coating based on molecular dynamics simulation and machine learning

The CoCrFeMnNi high-entropy alloy is widely regarded as one of the most promising materials for wear resistance, particularly in mitigating erosion failure in critical components of oil and gas equipment. However, its performance is significantly influenced by factors such as elemental composition and proportion. Therefore, determining the optimal parameters for preparing erosion-resistant alloys has become a significant challenge. In this study, for the first time, we establish an erosion resistance design evaluation system that integrates molecular dynamics and machine learning. This system incorporates parameters such as pressure, friction coefficient, and the number of wear atoms, and a high-performance machine learning model is trained to predict the optimal erosion-resistant coating and verify. Furthermore, to enhance the applicability of the design evaluation method, the trained machine learning model can be directly employed to guide the alloy component design under different crystal orientations and structures post-verification. The results of this study offer significant theoretical insights for the development of the optimal erosion-resistant CoCrFeMnNi high-entropy alloy for oil and gas equipment, while also substantially reducing both experimental and temporal costs.