Prediction of Low‐temperature Dynamic Mechanical Behavior of HTPB Propellant Based on Artificial Neural Networks

Le résumé fourni par la source

ABSTRACT To accurately describe the dynamic mechanical response of HTPB propellants under extreme environments, this paper employs artificial neural networks to conduct modeling and prediction research on their dynamic mechanical behaviors, and proposes and verifies a complete data‐driven solution. This solution adopts a hierarchical modeling and dual‐network architecture, which is different from the conventional idea of a single network. It achieves high‐precision fitting and multi‐condition generalization by constructing different targeted networks, and establishes a dual‐dimensional generalization verification framework. The research results show that compared with the traditional damage‐coupled thermoviscoelastic constitutive model, the proposed data‐driven method has better performance and can accurately capture various complex nonlinear behaviors and low‐temperature mechanical characteristics of HTPB propellants. Through rigorous verification, the model has good interpolation and extrapolation capabilities; combined with relevant optimization and verification methods, the computational efficiency, robustness and stability of the model are further improved.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.