Answers are grounded in this paper and cite their sources.
High-performance components made from high-temperature alloys often operate under prolonged exposure to elevated temperatures and multiaxial stresses, where creep fracture is a dominant failure mechanism. Accurate prediction of long-term creep life is therefore crucial for ensuring structural reliability. In this study, we propose a data-driven framework for rapid multiaxial creep life prediction, combining the multiaxial stress rupture criterion (MSRC) with the skeletal point concept. A recurrent neural network (RNN) is trained on finite element data from 1278 notched specimens to predict skeletal point stress, while a fully connected neural network (FCNN) is used to estimate rupture time based on material properties and stress states. Testing results show that the proposed method maintains prediction errors within a two-fold tolerance range, demonstrating high accuracy, generalisation, and computational efficiency. This approach offers a practical tool for evaluating the long-term creep behaviour of high-temperature components under complex multiaxial loading.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: