Evaluating constitutive model applicability and controlling steady flow stress via multi-path loading during hot deformation for 7B50 aluminum alloy

The prediction of flow stress and the realization of steady flow stress under varying deformation conditions have always been critical challenges in the hot working process. In this work, hot compression experiments of as-hot rolled 7B50 aluminum alloys were conducted at the strain rate interval between 0.001 s⁻¹ and 1 s⁻¹ and with hot deformation temperatures ranging from 440 °C to 500 °C. The lattice diffusion model, Arrhenius model, and artificial neural network (ANN) model were employed to reveal the hot deformation behavior of 7B50 aluminum alloy. Among them, the Arrhenius model exhibited the best predictive ability with an R ² of 0.99743. The lattice diffusion model was suitable for describing internal state variations of the material. The ANN model demonstrated efficient dataset analysis capabilities. In addition, multi-path loading experiments were conducted to fit complex deformation conditions. The cyclic loading tests showed that, by unloading and reloading during the hot compression process, the flow stress could still be restored to the level prior to unloading. However, the flow stress exhibited an abrupt change when the strain rate or temperature varied. The difference in flow stress between the three steady stages did not exceed 10.6 MPa in an optimized multi-path loading test. Therefore, the new approach for achieving steady flow stress was proven to be feasible by simultaneously regulating the strain rate and the temperature.