A digital twin framework for predicting the equivalent elastic modulus of tungsten carbide coating

• Published in: Ceramics international Vol. 51; no. 18; pp. 26266 - 26279
• Main Authors: Zeng, Xin, Ping, Xuecheng, Zhao, Huan, Zhao, Qian
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.07.2025
• Subjects: Digital twin; Equivalent elastic modulus; PSO-BPNN; WC coating; WC coating; Digital twin; PSO-BPNN; Equivalent elastic modulus
• ISSN: 0272-8842
• DOI: 10.1016/j.ceramint.2025.03.309

In order to accurately predict the macroscopic equivalent mechanical properties of the Tungsten carbide (WC) coating, a digital twin (DT) framework for predicting equivalent mechanical properties of WC coatings was established in this paper. Firstly, the mesostructure and elastic modulus of WC coating materials were obtained by experimental methods. Then, a 2D representative volume element (RVE) statistical finite element (FE) model was established by introducing the homogenized multi-scale method and statistical model to mirror the physical entity obtained from the experiment. Next, a surrogate model of the FE modal was established based on improved particle swarm optimization (PSO) and back propagation neural network (BPNN) to improve computing efficiency. The results show that the simplified FE model can accurately predict the equivalent elastic modulus of WC coatings after calibration, the surrogate model established using the improved PSO-BPNN algorithm has accurate prediction results and significantly improved computing efficiency, and the average particle size and proportion of WC particles are the key factors affecting the equivalent elastic modulus of WC coatings.