Synergistic Machine Learning and Microstructural Analysis of Hot Deformation Behavior in a Novel 22Cr–25Ni Austenitic Heat‐Resistant Alloy

The hot deformation behavior of a novel 22Cr–25Ni austenitic heat‐resistant alloy was investigated through hot compression tests in the temperature range of 950–1150 °C and strain rate range of 0.01–10 s −1 . An artificial neural network model was developed, achieving exceptional accuracy in flow‐stress prediction ( R = 0.9998, AARE = 0.711%). Microstructural analysis revealed a complex dynamic recrystallization (DRX) behavior, the DRX degree increased with temperature but exhibited a distinctive nonmonotonic dependence on strain rate. Both continuous DRX (CDRX) and discontinuous DRX (DDRX) recrystallization mechanisms were found to coexist, with their dominance shifting under varying thermomechanical conditions. Texture evolution was strongly dependent on temperature and strain rate, showing a transition from brass‐/goss‐dominated components to a strong cube texture at elevated temperatures, while increasing strain rate reintroduced brass components. Notably, at 950 °C/0.01 s −1 , grain‐boundary M 23 C 6 carbides promoted dislocation pile‐ups and substructure development, leading to the identification of particle‐induced CDRX acting synergistically with DDRX as the primary softening mechanism. The results demonstrate that the hot deformation response of this multicomponent alloy is governed by the interplay between high stacking fault energy, solute drag, and precipitate pinning, which collectively explain its high activation energy, narrow processing window, and competitive DRX mechanisms.