Development of a numerical model for microstructural evolution during submerged ARC welding of ASTM A516 GR.70 steel

Abstract

The present study proposes a coupled thermo-metallurgical Finite Element Model (FEM) that has been specifically designed to simulate the Submerged Arc Welding (SAW) process on ASTM A516 Gr.70 hypoeutectoid steel, with the aim of predicting the extent of the Heat-Affected Zone (HAZ), grain size evolution and hardness variation with respect to phase transformations. The thermal field was modelled by considering a three-dimensional Gaussian volumetric heat source implemented through concentric heat-exchange windows in SFTC DEFORM-3D™ environment, ensuring accurate representation of localized energy deposition and heat conduction. A customized user subroutine was developed to couple the computed thermal cycles with metallurgical transformations by considering phase transformation mechanisms based on the Fe–Fe₃C phase diagram, grain refinement kinetics and phase-fraction-dependent hardness models. Temperature-dependent material properties were included to ensure accurate simulation because phase transformations are highly dependent on cooling rates, especially in hypoeutectoid steels. The numerical results were validated with metallographic and microhardness measurements, showing good agreement in terms of HAZ extension (4%), grain size distribution (< 10%) and hardness values (5%). The numerical and experimental validation presented in this study focuses on pass 2.1 within the investigated multipass joint geometry, selected to minimize the influence of reheating from subsequent welding passes. The results show the capability of the model to reproduce grain refinement in the fine-grained HAZ, grain growth in the solid–liquid transition zone along with the hardness gradient from the base material to the weld bead. Overall, the proposed approach provides a calibrated thermo-metallurgical framework for the prediction of microstructural and mechanical response of ASTM A516 Gr.70 steel during the investigated SAW process. The framework may provide a modelling basis for future Digital Twin-oriented developments, although real-time data assimilation, online updating and bidirectional process communication are beyond the scope of the present study.