Concept for the calculation of the distribution of heat input in the cathode area by GMA welding

Abstract

Currently used approaches for the modeling of the heat input in gas metal arc (GMA) welding process simulation usually assume an axisymmetrical Gaussian distribution of heat flux in the cathode region. However, as in GMA welding, cathodic electron emission of non-refractory metals is involved; the attachment region consists of multiple highly mobile cathode spots, which have a highly concentrated current density, which cannot be explained solely by thermionic emission, as is present in refractory cathodes. In this work, a novel concept is presented to determine the distribution of the cathode spots, allowing to determine the distribution of heat flux and current density, to serve as boundary conditions in a magneto-hydrodynamic weld pool simulation. While the concept does not yet deliver a fully convergent solution, a model lies at its core, which takes into account the experimentally determined high current density and provides a relationship between the cathode surface temperature, the generated heat flux, and the current density. By applying a stochastic cellular automaton method, the weighted random walk of the movement of the cathode spots is simulated, according to a probability determined by the potentially released heat flux. Averaging over time gives a spatially resolved distribution of heat flux and current density.