Analysis and Prediction of Welding Distortion in 316L Stainless Steel Using Finite Element Analysis

  • KIM, Jaewoong (Korea Institute of Industrial Technology)
  • KIM, Younghyun (Korea Institute of Industrial Technology)
  • HA, Eulyong (Korea Institute of Industrial Technology)
  • KANG, Sungwook (Changwon National University)
  • SHIN, Kwangyong (Korea Institute of Industrial Technology)
  • LEE, Kiyong (Korea Institute of Industrial Technology)
  • KO, Myeongjin (Korea Institute of Industrial Technology)

Please login to view abstract download link

As regulations for decarbonization in the maritime industry become increasingly stringent, there is growing interest in ships powered by liquefied hydrogen, an eco-friendly fuel. To store liquefied hydrogen, hydrogen must first be liquefied, which occurs at the cryogenic temperature of −253 °C. Structural materials are therefore required to possess high mechanical properties to ensure stable performance under such cryogenic conditions. In the process of manufacturing liquefied hydrogen storage tanks, welding is an essential step, and the welding distortion that occurs during this process has a direct impact on dimensional accuracy and structural safety. Accordingly, the development of technologies capable of predicting and controlling welding distortion in advance is essential. This study aims to measure welding deformation in multi-pass butt welding of 316L stainless steel using the Flux Cored Arc Welding (FCAW) process and to verify the deformation prediction through comparison with thermal and thermo-elasto-plastic finite element analysis. By analyzing deformation characteristics under different welding conditions, this study seeks to provide foundational data for predicting and controlling welding deformation during the fabrication of cryogenic hydrogen storage tanks. Multi-pass welding was performed on 10 mm thick 316L stainless steel using four different welding conditions, varying the number of passes, current, voltage, and welding speed. Bead width and penetration depth were measured through cross-sectional observations, and deformation was quantified. A 3D finite element model was constructed based on actual weld profiles, and heat transfer analysis was carried out using the Goldak double-ellipsoidal heat source model. Thermo-elasto-plastic analysis and the inherent strain method were then applied to predict welding deformation. The results confirmed that the number of passes and heat input significantly affect welding deformation, and the simulation approach could reasonably approximate deformation trends. These findings are expected to serve as a basis for optimizing welding conditions and minimizing deformation in cryogenic structural applications such as liquefied hydrogen storage tanks.