ABSTRACT This study presents a comprehensive thermal analysis of autogenous tungsten inert gas welding on a 316 LN stainless steel plate, integrating experimental techniques and finite element simulation. The Goldak double ellipsoid heat source model is employed to simulate the heat input accurately. Experimental measurements are conducted using both contact (using K-type thermocouples) and noncontact infrared (IR) thermography methods, comparing their effectiveness in capturing precise thermal gradients and validating the numerical predictions. The peak temperature in the adjacent base metal at a distance of 10 mm from the weld centerline obtained from finite element analysis (FEA) is 527°C, whereas that obtained using the thermocouple and IR thermography are 507°C and 513°C, respectively. The FEA predicted molten weld pool temperature to be 1,970°C, which closely matches with the IR recorded temperature of 1,800°C. The simulation demonstrated high accuracy, with the peak melt pool temperature measured by IR and the peak temperature at the base plate measured using both IR and thermocouples showing a deviation within 10 %. Along with validating computational predictions of weld pool temperature, the IR thermography was also used successfully to estimate the weld-bead width. The fusion zone half-width from the experimental macrograph analysis is 3.2 ± 0.05 mm, which agrees well with the simulation (3.2 mm) and IR thermography (3.6 mm). Additionally, the study explores the effect of varying convection heat transfer coefficient on cooling rate. This integrated approach results in more accurate welding simulations, which can act as the foundation for residual stress analysis and for optimizing the welding parameters. The novelty of this work lies in the combined use of thermocouple, IR thermography, and FE simulation for weld thermal analysis, achieving ≤10 % deviation in thermal history predictions and bead-width estimation within ∼12 % of the measured values.
Pillai et al. (Sun,) studied this question.