EXTENDED ABSTRACT: Wear-resistant materials, as key foundational materials in industry, possess excellent characteristics such as high strength, high wear resistance, high impact toughness, and thermal stability. These materials ensure the development of fields such as aerospace, energy transportation, equipment manufacturing, mining, electronics, and resource processing. Coupling computational simulation with key experiments is an efficient approach for designing and developing new types of wear-resistant materials. In this work, the integration of multiscale theoretical design, material preparation, performance testing, and machine learning methods is used to study the wear-resistant materials. We employ phase diagram thermodynamic calculation methods to investigate the composition and process windows of multicomponent wear-resistant material systems. Key experiments are de signed to measure the influence of composition and processes on the microstructure. Based on the microstructure, finite element models of cemented carbides are established, mapping the morphology and distribution of hard phases and binder phases into the finite element models to make the simulation process align with actual conditions. This allows for simulation and analysis of their mechanical states and service performance. Machine learning is used to analyze the correlations between computational and experimental data, establishing a mapping relationship between the composition, process, structure, and performance of wear-resistant materials. This design strategy leverages multiscale computational simulation combined with various experimental data to achieve precise control over microstructure and performance throughout the preparation process of wear-resistant materials, ultimately accelerating the research and development efficiency of these materials.
Keywords:Wear-Resistant Materials; Data-driven; Microstructure Design; Performance Regulation