Machine Learning-Assisted Thermo-Elastic Analysis of Fiber-Reinforced Composite Rotating Disks

Abstract

This study presents a numerical investigation of the thermo-mechanical behaviour of Glass Fiber Reinforced Polymer (GFRP) and Basalt/Epoxy composite rotating disks, focusing exclusively on the computation of radial displacements under coupled thermal and mechanical fields. The elastic modulus of the materials was assumed to remain constant with temperature, and the analysis was conducted over the temperature range 40 °C to 120 °C in 20 °C increments. Results indicate that temperature rise leads to noticeable changes in radial displacement distributions, exhibiting distinct deformation characteristics for each composite type. Basalt/Epoxy disks showed comparatively higher displacement responses than their GFRP counterparts, reflecting their different thermo-elastic sensitivities. In addition to the numerical analysis, the obtained displacement data were compared with machine learning predictions, demonstrating close agreement and indicating that the generated dataset is suitable for training data-driven models. These findings suggest that the numerical results can be effectively utilized within machine learning frameworks for predictive modelling and optimization tasks. Overall, the study provides theoretical guidance for safe design, performance evaluation, and material selection of composite rotating disks in aerospace, automotive, and energy applications.