Multi-Objective Optimization of Friction Stir Welding Process Parameters
CHALLENGE - Friction Stir Welding (FSW) is a relatively new solid state joining technology that has found considerable application in various industries due to its high mechanical performance and low environmental impact. A successful FSW process relies on acquiring the correct thermal conditions that allow for adequate softening and stirring of the workpiece material. As a result it was required to develop a thermal model that would account for the severe change in mechanical properties during the FSW process as a result of the high operating temperature induced by frictional heating.
SOLUTION - The model is based on the FSW of 3mm thick, AA5083-H111 plates. Using temperature dependent data for the aluminum alloy, the model aimed to minimize the rotation speed of the tool, maximize the plunge rate, minimize the dwell time and maximize the feed rate in the aim of predicting as efficient a process as possible. The study employs a combination of a coupled steady-state thermo-mechanical model, a transient thermal model and a steady state thermal model, all of which are implemented in ANSYS 13. The FEM models are then embedded in a modeFRONTIER multi-objective optimization environment that uses a SOBOL DOE approach coupled with MOGA II. Welding trials were performed in order to establish an experimental operating window for the process to verify the framework. A revised optimization workflow was developed that more accurately predicted operating parameters that coincided with the experimentally obtained operating window.
BENEFITS - Discrepancies were found between the FEM models and the experimental data and, as a result, a revised platform was developed that replaced the thermo-mechanical FEM model with an equation relating the rotational speed of the tool to the heat flux input to the system. The parameters of the model were a rotational speed of 507.8 rpm and feed rate of 59.4 mm/min.