Advanced Control Methods for Optimization of Arc Welding

Abstract

In this thesis the objective is to optimize the control algorithms for the manual pulsed Gas Metal Arc Welding (GMAW) process, with the aim of enhancing the quality of welded joints. To be able to develop controllers for the GMAW process, and also for enabling process simulation, a mathematical model describing the GMAW process is developed. The mathematical model includes a description of the electrical circuit, the drop dynamics, and the melting rate, and also, the model includes criteria for drop detachments. Basically, the electrical circuit consists of the welding machine, the wires, the electrode, and the electrical arc. The drop is modelled as a mass-spring-damper system influenced by a number of external forces. Of these external forces the electromagnetic force is the most significant. Due to the importance of the electromagnetic force, a thorough derivation of the electromagnetic force is included in the thesis. The melting rate of the consumable electrode is modelled both from a statically (steady-state) point of view, and from a dynamically point of view. The static model is a simple equation describing the melting rate, while the dynamic model is more complex. The structure of GMAW control is discussed, and based on this discussion a general control structure is presented. This structure includes an arc length controller and a metal transfer controller, and also, an inner loop controlling the welding current. The current is assumed to be controlled by a traditional PI-controller, and therefore, this inner control loop is not considered in the thesis. For arc length control, a nonlinear controller based on feedback linearization is proposed, and robustness and performance are considered. For metal transfer control an approach based on obtaining a uniform drop size prior to pulse initiation is proposed. This is carried out by calculating the amount of melted electrode between the current pulses during the welding process. The purpose of the uniform drop size approach is to enhance the robustness of the drop detachment process. In addition to the arc length controller and the metal transfer controller, an arc length minimization algorithm is proposed for enhancing the ability to focus the arc, and also, to minimize the heat input into the workpiece. Simulation programs for testing the controllers and algorithms have been developed, and successful tests have been carried out.