Computational prediction of grain features during friction stir processes through a mechanistic discontinuous dynamic recrystallization model

The large amount of strain combined with high temperature during Friction Stir Welding and Processing (FSWP) results in dynamic recrystallization and grain growth. The final properties of the processed material depend on the recrystallized grain structure. The ability to predict recrystallized microstructural features would take the FSWP modeling efforts one step closer to estimating the final weld mechanical properties. Here we present a computational framework for microstructural feature prediction based on the Discontinuous Dynamic Recrystallization (DDRX) principle considering plastic deformation, nucleation, and growth. The computed strains, strain rates and temperatures from an existing Heat Transfer and Material Flow (HTMF) model are utilized as input parameters for the DDRX model. The microstructural features such as average grain size, dislocation density, Taylor's factor, number of new grains formation and grain size distribution are predicted using the DDRX model. The grain size prediction is validated against experimentally measured grain size, demonstrating a remarkable 97% accuracy and the reliability of the DDRX model.