Abstract
It is important to develop a mathematical method for determining the optimum process parameters in welding engineering. An algorithm for optimizing process parameters to achieve the required temperature field on the specimens has been developed. The optimization of process parameters refers to the determination of optimum heat inputs, such as arc current to make the calculated temperature fields correspond with required ones in the present work. That is, the objective function is defined as the optimization problem of minimizing the difference between calculated and required temperatures. The nonlinear programming methods, such as the convex programming and the steepest descent method, are applied to solve the optimization problem in this study. It has been shown that the algorithms based on the optimizing methods are reliable enough to determine the optimum heat inputs in GTA (Gas Tungsten Arc) welding on both the plate and pipe specimens.
