Abstract: The manual selection of environment parameters during welding simulation will bring a significant error to the simulation result of welding temperature field. By a combination of finite element method (FEM) and infrared thermography, these environment parameters were inversed mathematically in place of previous manual selection. First, FEM model of the welding process was constructed, and the temperature field was computed with initial environment parameters. Then, a real welding process was conducted and the temperature field was measured by infrared thermography. Last, the simulation and measurement results were compared, and the environment parameters were adjusted continuously with the genetic algorithm (GA) until the simulation matched the measurement best. Parameters according to the best-matched simulation results were considered as the most appropriate parameters.

Key words: environment parameters, welding temperature field, inversion