Sensitivity to welding positions and parameters in GTA welding with a 3D multiphysics numerical model - Aix-Marseille Université Access content directly
Journal Articles Numerical Heat Transfer, Part A Applications Year : 2017

Sensitivity to welding positions and parameters in GTA welding with a 3D multiphysics numerical model

Abstract

A three-dimensional numerical model of Gas Tungsten Arc welding has been developed to predict weld a bead shape, fluid flow in the weld pool, as well as thermal field in the workpiece. This model accounts for coupled electromagnetism, heat transfer, and fluid flow with a moving free surface to simulate different welding positions. The solution strategy of the coupled non-linear equations that has been implemented in the CastɜM finite-element code is also discussed. The capabilities of our numerical model are first assessed by comparison to the experimental results. Then, as fluid flows in a weld pool play a prominent role in the weld quality as well as in the final shape of the weld bead seam, the effect of various welding positions on the weld pool shape has been investigated. This constitutes the main novelty of this work. The performed computations point out a strong sensitivity to gravity on the weld pool shape depending on assisting or opposing the weld direction with respect to gravity. This study contributes to assessing the model capabilities that provide a deeper physical insight into a more efficient optimization of welding processes.
Fichier principal
Vignette du fichier
Preprint_revised_25octo2016.pdf (1.98 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01791099 , version 1 (11-06-2021)

Licence

Attribution - NonCommercial - NoDerivatives

Identifiers

Cite

Minh Chien Nguyen, Marc Medale, Olivier Asserin, Stephane Gounand, Philippe Gilles. Sensitivity to welding positions and parameters in GTA welding with a 3D multiphysics numerical model. Numerical Heat Transfer, Part A Applications, 2017, 71 (3), pp.233 - 249. ⟨10.1080/10407782.2016.1264747⟩. ⟨hal-01791099⟩
145 View
121 Download

Altmetric

Share

Gmail Facebook X LinkedIn More