Influence of welding conditions on microstructure and defect generation in 316 stainless steels: experimentation and modeling

Abstract

Gas Tungsten Arc Welding (GTAW) is a widely used technique for joining stainless steel pipelines, yet it often results in discoloration and oxide layer formation in the heat-affected zone (HAZ). This study investigates the impact of oxygen concentration in the backing gas on the pitting corrosion resistance and oxidation behavior of AISI 316L stainless steel welds. Experimental and numerical approaches were combined to analyze oxidation kinetics and heat distribution during welding. Experimental analysis was conducted on 316L stainless steel pipes welded under different oxygen levels (50, 200, 500, and 5000 ppm) in the purging gas. Additionally, the effects of surface roughness (40- and 60-grit) on discoloration and corrosion resistance were examined. Scanning Electron Microscopy (SEM), X-ray Diffraction (XRD), and electrochemical testing (cyclic polarization and Electrochemical Impedance Spectroscopy) were utilized to characterize oxide layer formation and pitting susceptibility. Results revealed that increasing oxygen concentration in the backing gas initially enhanced corrosion resistance up to a critical threshold, beyond which the formation of porous iron oxide layers and chromium-depleted zones led to a decrease in resistance. Smoother surfaces exhibited improved corrosion resistance, emphasizing the role of surface preparation in mitigating localized corrosion. To further understand the oxidation mechanisms, non-isothermal oxidation kinetics of 316L stainless steel were investigated between 1100 K and 1373 K using Thermogravimetric Analysis (TGA) at heating rates of 5, 10, 15, 20, and 25 K/min. Activation energy (Ea) was determined using model-free isoconversional methods (Friedman, Flynn-Wall-Ozawa, Starink, and Kissinger-Akahira-Sunose) and model-fitting methods (Coats-Redfern and Kennedy-Clark). The determined activation energies ranged from 224.79 to 233.81 kJ/mol, with the F2 (second-order reaction) and F3 (third-order reaction) models providing the best fit to experimental data. The Criado method further confirmed the suitability of these reaction models. FactSage thermodynamic simulations predicted the formation of protective oxide layers, primarily composed of spinel and corundum, at temperatures up to 1373 K. Numerical simulations were conducted using Finite Element (FE) and Finite Volume (FV) methods to predict heat distribution and oxidation layer formation during welding. The Goldak double ellipsoid model was implemented in Abaqusto simulate transient heat input, while oxidation kinetics were analyzed using an Arrhenius-based oxidation rate equation. Activation energy values derived from isoconversional and model-free methods were incorporated into the simulations to enhance accuracy in predicting oxidation behavior. The findings of this research highlight the importance of optimizing oxygen concentration in the backing gas and controlling surface roughness to enhance corrosion resistance and mitigate weld discoloration. By integrating experimental results with numerical modeling, a better understanding was developed on oxidation kinetics and welding process optimization for improved longevity and reliability of stainless steel pipelines in corrosive environments.

Date	19 Dec 2025
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Mohammad Jahazi (Supervisor) & Alireza Khodabandeh (Co-supervisor)