Computational Modeling of Thermo-Metallurgical Behavior During the TIG Welding Process

    Welding is a necessary industrial process that persistently needs to be developed. For this aim, numerical prediction of welded parts behaviors is an alternative, which avoids the cost of the experimental analysis. During welding processes, located and moving heat source generates, after cooling, important residual stresses in welded parts. These stresses are the results of the heterogeneity of deformation due to changes in temperature and in metallurgical transformation in the different parts of the weld. This work concerns the modeling and the finite element simulation of these microstructural transformations during the welding operation of a plate in C50 steel.

Several authors have implemented thermomechanical and metallurgical models in numerical calculation codes to study the generation of residual stresses due to metallurgical transformations during welding. Ronda et al. (2000) presented the results of a numerical simulation of welding problem based on electromagnetic, thermal, mechanical and metallurgical modelling. Results of this numerical simulation are shown for two simple welding benchmark problems formulated for two thick plates. Using ABAQUS, Deng (2009) developed a thermo-metallurgical-mechanical computational model to predict the residual stresses in 2.25Cr - 1Mo steel pipes butt-welded in several passes. The simulation results show that yield strength change affect welding residual stresses in 2.25 Cr–1Mo steel pipes. Using SYSWELD software, Li et al. (2017) developed a thermo-metallurgical-mechanical finite element model to predict welding residual stresses in P92 steel joints. The experimental measurements verified the effectiveness of the developed computational approach. Heinze et al. (2012) studied the influence of austenitic grain size on the residual stresses distribution after steel welding. Numerical results have shown the impact of austenite grain size on the residual stress development. Xia et al. (2018) propose a finite element simulation procedure to model the thermo-metallurgical coupling phenomena of welding using the ABAQUS software. The proposed coupled analysis simulation model is confirmed by the excellent agreement between the simulated and experimental results.

      Sun et al. (2019) performed a series of numerical simulations on the SIMUFACT software, to examine the variability of welding residual stresses for different materials. The result shows that the mixed hardening model provides the most accurate prediction of residual stresses. Zain-Ul-Abdein et al. (2011) developed a comparative study between two finite element models established on ABAQUS and SYSWELD. They studied the effect of metallurgical transformations on residual stresses and distortions induced by laser beam welding in a T-joint configuration. The results show that considering metallurgical changes has a negligible effect on the predicted distortions but not on the residual stresses distribution.
    Other researchers have developed numerical computer codes for the welding process simulation. Hamide et al. (2008) contributed to developing an adaptive mesh module for the TRANSWELD software to simulate the thermo-metallurgical coupled behavior of a fusion line produced on a steel plate by a TIG welding station. Hendili et al. (2013) contributed to developing a metallurgical behavior module established in CODE ASTER. They used Leblond's equation to describe the formation of the austenitic phase during heating stage and the Waeckel’s model to describe the decomposition of the austenitic phase during the cooling stage.
    Several other studies present numerical and experimental characterization of welded assemblies. Baskoro et al. (2017) investigated the influence of micro-resistance spot welding parameters on the mechanical properties and failure of an aluminum alloy 1100 nugget. Baskoro et al. (2011) propose a study comparing particle swarm optimization with a genetic algorithm for molten pool detection in fixed aluminum pipe welding.  Rupajati et al. (2021) investigate the characteristics of the lap shear force and microstructure of micro friction stir spot welding joints.
    Our numerical tool developed under MATLAB environment is a new personalized numerical calculation code, simple and flexible. With this tool, we have tried to solve some of the difficulties that exist in using commercial codes to simulate the welding process. For example, the study of metallurgical behavior or the heat source moving simulation during welding cannot be done directly on some commercial software. Indeed, to predict the behavior of parts during welding on ABAQUS for example, we need to program the metallurgical behavior laws on FORTRAN. In our case, the metallurgical changes and source moving peculiarities of the welding processes study are integrated into our numerical calculation code. The computer code will be used to perform thermomechanical and metallurgical calculations during the TIG welding operation. This article is about the development of the thermo-metallurgical model.