ALGORITHM-BASED ON NONLINEAR-PROGRAMMING METHOD FOR OPTIMUM HEAT INPUT CONTROL IN ARC-WELDING

A simple mathematical method for the estimation of the optimum heat in puts to make the temperature field coincide with a requirement during arc welding has been developed. The optimisation problem in the presen t work means the objective function which is defined as the difference between the calculated and required temperatures. An analytical solut ion of rite heat pow equation is applied for the calculation of the te mperature field in welding, and non-linear programming methods, i.e. t he convex programming method and the steepest descent method, have bee n used to solve the objective function in the problem mathematically. It is concluded that both optimising methods are sufficient to determi ne the optimum heat input pattern in the case of gas tungsten arc (GTA ) welding on thin plate and small sized pipe specimens. Furthermore, t he applicability of the estimated optimum heat inputs is evaluated exp erimentally by bead on plate welding tests.