Thermal mechanical modeling of gas tungsten arc welding

Welding problems encountered in development and production activities are typically solved in a ''let's try this'' approach, which can be costly and do not always provide the optimal solution. The Ongoing research at LLNL and Sandia is directed toward improving our basic understanding of what is actually occurring on a phenomenological basis in welding and to apply this knowledge to the development of a general computer-based weld model. Our approach has been to couple a well defined experiment with computer simulation to examine the effects of various moedling assumptions on the computer predictions. The application is confined to the inert gas-tungsten-arc (GTA) process. However, the results give insight into modeling other welding processes (e.g. laser, electron beam).