A Metallurgically Informed Multiscale Integrated Computational Framework for Metal Forming Processes

: Predicting the mechanical response of industrial alloys is crucial for optimizing manufacturing processes and improving material performance. Traditional, solely experimental approaches, though effective, are inefficient as they are resource-intensive, requiring extensive laboratory testing and the iterative calibration of processing conditions. These costs can be avoided through computational/virtual experiments based on a multiscale hierarchical framework that integrates macroscopic approaches, mesoscale modelling as well as atomic level and advanced thermodynamical simulations to study and predict the mechanical response of metallic systems. In the context of this work, a framework for studying the effect of forming on metallic materials is proposed, applied, and validated on the hot extrusion of AA6063. Coupling thermodynamic simulations (including Phase Field) results with literature data establishes a microstructurally accurate representative volume element (RVE) design. This way, the phase fraction and the grain size of the RVE are determined by thermodynamic simulations (ThermoCalc, MICRESS), which can be validated via microstructure characterization. It is known that the mechanical properties of the individual phases affect the macroscopical properties of the material. Using atomic level simulations (i.e., molecular dynamics), the dislocation density of the material is calculated and utilized as an input for a Crystal Plasticity Fast Fourier Transformation simulation. This iterative process can be applied to match all stages of manufacturing processes. The hierarchical and systematic integration of these computational methodologies enables a rigorous analysis of the effect that processing parameters have on the microstructure. This work contributes to the broader effort of creating experiment-free workflows for designing materials and processes by leveraging a multiscale modeling approach. Coupled with experimental data, the predictive accuracy of the mechanical behavior can be further enhanced.