Research on a PINN-Based Predictive Model for the Dynamic Ablation Process of Insulation Layer

Thermal protection is critical for solid rocket motor safety, yet ablation prediction of insulation layers remains a core challenge. To achieve rapid and accurate prediction, this study develops a Physics-Informed Neural Network (PINN) model for EPDM-based insulation materials. The model is constructed based on three submodels (flow, heat transfer, and mass transfer in the surface boundary layer of thermal insulation layers; flow, heat transfer, and diffusion in porous media; thermal decomposition and thermochemical ablation of insulation layers) along with two external modules (gas and particle erosion). Results show a maximum prediction error of 10.76%, in good agreement with experiments. Compared with traditional ablation models, the proposed prediction model achieves faster prediction speed while maintaining good accuracy, which is conducive to providing suggestions for thermal protection design.