Experimental-based methodology for the double ellipsoidal heat source parameters in welding simulations

The welding numerical simulation involves thermal and mechanical analyses and metallurgical consideration. The thermal analysis determines the transient temperature distribution on the welded part, and its accuracy is relevant for the following analyses on structural integrity, in particular buckling and fatigue failure modes of steel-plated structures. The temperature distribution is strongly influenced by the size and shape of the heat source, which can be represented as a double ellipsoid and defined by Goldak’s parameters. The main difficulty is that the adjustment of these parameters to obtain a suitable temperature distribution is usually determined by trial and error. In this paper, a parametric study is performed to analyze the influence of Goldak’s parameters on the weld bead size for 2D and 3D numerical welding simulations. A method to estimate the parameters of a double ellipsoidal heat source in motion is proposed. An algorithm has been implemented based on the combination of analytical formulation, experimental data, and numerical simulation. As an analytical formulation, the Fachinotti’s solution is used to determine the isotherms. The weld bead dimensions are defined from the melting temperature isotherm of the material. Experimental data, such as the welding process parameters and the weld bead dimensions are input to the algorithm. Numerical simulations of the welding process have been developed to calibrate and validate the method. The numerical simulation results from the obtained Goldak’s parameters by the proposed method are correlated with the experimental results of fillet-welded specimens. © 2020, Sociedade Brasileira de Engenharia Naval.