A multi-objective optimization application in Friction Stir Welding: Considering thermo-mechanical aspects

The objective of this paper is to investigate optimum process parameters in Friction Stir Welding (FSW) to minimize residual stresses in the work piece and maximize production efficiency meanwhile satisfying process specific constraints as well. More specifically, the choices of tool rotational speed and traverse welding speed have been sought in order to achieve the goals mentioned above using an evolutionary multi-objective optimization (MOO) algorithm, i.e. non-dominated sorting genetic algorithm (NSGA-II), integrated with a transient, 2-dimensional sequentially coupled thermo-mechanical model implemented in the FE-code, ANSYS. The thermal model is based on a heat source description which in essence is governed by the rotational speed and the temperature dependent yield stress of the work piece material. This model in turn delivers the temperature field, in order to compute thermal strain field which is the main driver for the mechanical model predicting both transient and finally residual stresses in the work piece. This thermo-mechanical model is then used in the aforementioned constrained MOO case where the two objectives are conflicting. Following this, two reasonable design solutions among those multiple trade-off solutions have been selected based on the cost and the quality preferences.