This is the authors' abstract. We don't add key points for this paper.
Optimizing hot deformation conditions is critical for achieving efficient thermo-mechanical processing of advanced alloy systems. In this study, a multicomponent Ni 48 Cu 10 Co 2 Ti 38 Ta 2 alloy was developed, exhibiting a refined eutectic microstructure composed of NiTi and Ni 3 Ti phases, along with coarse Ti 2 Ni and NiTi dendritic phases. High-temperature compression tests were performed using a Gleeble® thermo-mechanical simulator over a temperature range of 973–1273 K and strain rates from 10 -2 to 10 s -1 to investigate the alloy flow behavior. To reduce experimental efforts and enhance prediction accuracy, five machine learning (ML) models random Forest (RF), XGBoost (XGB), decision tree (DT), K-Nearest neighbor (KNN), and gradient boosting (GB) were applied to predict the flow stress–strain response and construct processing maps. Among these, the RF model demonstrated superior predictive performance, particularly at a strain rate of 0.1 s -1 , with R 2 = 0.97, RMSE = 10.1%, and MAE = 8.9%. The flow curves predicted by the RF model were used to develop precise processing maps, identifying optimal and safe deformation conditions. The resulting processing maps were validated through experiments, confirming that the alloy can be safely deformed within the temperature range of 1173–1273 K and strain rates between 10 -0.8 and 10 -2 s -1 . This integrated experimental–computational approach offers a reliable and efficient strategy for determining hot working conditions, reducing material and energy consumption. It also presents a robust framework for advancing the development of high-temperature alloy systems through the combination of machine learning techniques and experimental validation.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Jain et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: