Strain-rate-sensitive mechanical response, twinning, and texture features of NiCoCrFe high-entropy alloy: Experiments, multi-level crystal plasticity and artificial neural networks modeling

Abstract The present work adopts an extended multi-level crystal plasticity finite element method (CPFEM) framework coupled with an improved artificial neural network (ANN) algorithm to predict the quasi-static and dynamic uniaxial tensile mechanical response, twinning and texture characteristic in NiCoCrFe high-entropy alloy. Firstly, a Split-Hopkinson tensile bar setup is utilized to achieve high strain rates, and the initial microstructure and texture data used in CPFEM is characterized by electron backscattered diffraction. The experimental results show the effect of twins on strain hardening is more obvious with the increase of strain rates. Next, a developed dislocation-density based hardening law taking into account rate and temperature sensitive property is firstly extended by calculating the slip/twin activation stresses through thermal activation mechanism, Hall-Petch resistance, forest dislocation drag and slip-twin interactions. The simulations reveal some underlying defect mechanisms using the temporal evolution of microstructure and macro response over a wide range of strain rates. Then, an advantageous ANN model combined with the Genetic algorithm (GA) optimized is applied to exploit potential law of massive experimental and CPFEM data and serve engineering applications efficiently. The verification results capture well the features of strain rate sensitivity with CPFEM and experimental data sources. Finally, the predictive ability of two models is further demonstrated by validating the uniaxial tension of NiCoCrFe HEA. It is clear that the improved idea (an ANN model through GA optimization) is more efficient than the physical mechanism based CPFEM model.