Numerical simulation of material flow and defect formation during FSW to predict weld failure location

Predicting the failure location of welded specimens is of importance for various industrial applications. In friction stir welding, material flow and eventually defect have an effect on the failure location. In the current work, a three-dimensional coupled Eulerian Lagrangian (CEL) is developed to study the material flow and predict defects originating during friction stir welding of AA2024 having a thickness of 3 mm. To minimize defects and achieve good weld quality, a square-shaped pin is used. The developed model is validated with experimentally observed axial force and spindle torque. Numerically predicted defects have been validated with experimental fracture locations and strength to test the robustness of the model in quantifying defects. Peak temperature increased by 10.7% when rotational speed was increased from 600 to 1500 rpm. Also, the peak temperature rise of 6.1% is observed when the welding speed is increased from 60 to 150 mm/min. Higher rotational and welding speed led to lower defects. At 1500 rpm and 150 mm/min process conditions, the highest weld strength of 447 MPa is obtained. Material flow analysis is carried out for varying process parameters; an intermixing of material flow with a zig-zag pattern is observed for 1500 rpm, indicating better material flow as compared with 600 rpm.