Development of a Three-Dimensional Heat-Transfer Model for the Gas Tungsten Arc Welding Process Using the Finite Element Method Coupled with a Genetic Algorithm–Based Identification of Uncertain Input Parameters

An accurate estimation of the temperature field in weld pool and its surrounding area is important for a priori determination of the weld-pool dimensions and the weld thermal cycles. A finite element-based three-dimensional (3-D) quasi-steady heat-transfer model is developed in the present work to compute temperature field in gas tungsten arc welding (GTAW) process. The numerical model considers temperature-dependent material properties and latent heat of melting and solidification. A novelty of the numerical model is that the welding heat source is considered in the form of an adaptive volumetric heat source that confirms to the size and the shape of the weld pool. The need to predefine the dimensions of the volumetric heat source is thus overcome. The numerical model is further integrated with a parent-centric recombination (PCX)-operated generalized generation gap (G3) model-based genetic algorithm to identify the magnitudes of process efficiency and arc radius that are usually unknown but required for the accurate estimation of the net heat input into the workpiece. The complete numerical model and the genetic algorithm-based optimization code are developed indigenously using an Intel Fortran Compiler. The integrated model is validated further with a number of experimentally measured weld dimensions in GTA-welded samples in stainless steels.