Artificial intelligence and molecular dynamics assisted analysis of wear behavior of high entropy alloys

Understanding the underlying wear mechanisms of high entropy alloys (HEAs) offers opportunities for the development of advanced materials for tribological conditions. In this work, we proposed an artificial intelligence and molecular dynamics (MD) based computational framework to explore the wear behavior of Al-Cr-Co-Fe-Ni based HEAs. The influence of alloying elements on coefficient of friction (COF) has been further discovered. To estimate the artificial neural network (ANN) model, the microstructure, hardness and tribological property of the (FeNi)xCoCrAl (x=2, 3) HEAs were studied at dry sliding condition. The results showed that Fe3Ni3CoCrAl HEAs exhibited a coarser dendritic structure than Fe2Ni2CrCoAl HEAs, with the microhardness of 309.2HV and 283.5HV, respectively. The tribological results showed that the wear rate of Fe2Ni2CoCrAl HEAs and Fe3Ni3CoCrAl HEAs was 5.2 × 10 − 5 mm 3 / ( Nm ) and 5.4 × 10 − 5 mm 3 / ( Nm ) , the average friction coefficient in the steady state denoted ∼0.55 and ∼0.56, respectively. The good agreement between the predicted values and the measured values demonstrated that the model exhibited superior prediction accuracy. The dominant wear mechanism was abrasive wear for Fe2Ni2CoCrAl HEAs, whereas it turned to adhesive wear and delamination wear for Fe3Ni3CoCrAl HEAs. Both alloy surfaces experienced oxidation wear due to the frictional heat. Atomic insight into friction and wear behavior was investigated at nano-scratch condition by using MD. The dislocations and defects in the material subsurface layer were captured to understand the wear mechanism. Our findings revealed the deformation mechanisms and microstructure evolution during the friction and wear process, offering a new perspective for designing alloys with excellent wear resistance in the future. • An artificial intelligence and molecular dynamics based computational framework was proposed to explore the wear behavior. • Al showed a significant negative effect on the coefficient of friction. • Fe3Ni3CoCrAl HEAs exhibited coarser dendritic structure and lower hardness than Fe2Ni2CoCrAl HEAs. • The wear mechanism turned from abrasive wear to adhesive and delamination wear with increasing Fe and Ni. • The point defects and vacancies, atomic clusters and SFs were revealed at the subsurface layer under nano-scratch.