Numerical study of refill friction stir spot welding of dissimilar metallic materials using smoothed particle hydrodynamics (SPH)

Refill friction stir spot welding (RFSSW) has found several industrial applications, especially in the transportation and automotive sectors. However, modeling the RFSSW process has been tackled mainly with empirical approaches. At the same time, the key physical phenomena involved have been explained and predicted by a few numerical studies in the literature. This study uses a fully Lagrangian method, smoothed particle hydrodynamics (SPH), for the simulation of RFSSW. The Lagrangian particle method simulates materials undergoing large deformation, interface dynamic changes, void formations, material temperature, and strain evolution without using complex tracking schemes often required by traditional grid-based methods. As a relevant example, magnesium-to-steel welding simulation is presented by accounting for all the main thermo-mechanical phenomena involved. Temperature, stress, and strain field histories as well as material flow taking place during the process, are determined as characteristic aspects for qualification of RFSSW; the proposed computational approach is validated by comparing the predicted and experimentally measured welding temperature. The results obtained demonstrate that SPH is a reliable tool for welding design and process optimization and provides the information related to the involved physics needed to precisely evaluate the quality, the mechanical characteristics, and the material flow of the joined region.