Can Virtual Experiments Replace Conventional Experiments?

Predicting how metals deform and fail in extreme environments is essential for ensuring the reliability and safety of engineered systems relevant to stockpile stewardship. However, calibrating material models that accurately capture this behavior remains challenging due to the number of traditional experiments needed to identify the complex interactions of different material behaviors. One approach to reducing the cost and time for calibrating these models is to use novel experiments and an inverse identification method; another approach is to use physics informed computational models at smaller length scales to conduct virtual experiments that provide similar information. This work presents a multiscale modeling framework that links microscale deformation physics to macroscale material response by conducting virtual experiments. Crystal plasticity simulations are performed using a computationally generated microstructure of aluminum 6061-T6, incorporating key features such as crystallographic texture and grain morphology, and are simulated in uniaxial and multi-axial loadings. Results of these simulations are compared to traditional experiments and are used to calibrate a component-scale material model that is used in the simulation of a hole-expansion experiment that incorporates multiple strain rates, and multiple multi-axial loadings.