Thermal optimization of friction stir welding with simultaneous cooling using inverse approach

Abstract

Study utilized an inverse heat transfer approach in order to obtain heat input during friction stir welding with simultaneous cooling process. It specifically applied the conjugate gradient method using limited experimental temperatures in space and time. Experimental dataset was either inputs for convergence of inverse algorithm or outputs for verification of final predicted results. The three-dimensional finite element model was used for direct problem. To obtain better precision and simulate real practical process, slip function has been defined by exponential approximation of matrix rotary speed and calculated Johnson-Cook coefficient as an immeasurable term of analytical solution, which is the main output of inverse determination. Weld line maximum temperature was calculated as second immeasurable parameter. Optimization of normal force and cooling performance for friction stir welding of Al5052 aluminum alloy sheet was sought. Numerical results were consistent with experimental recorded temperatures and the optimization results were acceptably close to the desired maximum temperature, which is practically achievable.