Application of pulsed electric current to solidifying conductive materials leads to a rapid and significant reduction of surface roughness. This is validated experimentally in the present work for multi-component oxides. The surface geometry of the casts without and with pulsating treatment was measured using a Leica DCM-3D Profiler. The pulsating treatment reduces the roughness of solidified materials by more than 50%. The unmeasurable points, which account for nearly 2% of total area, were interpolated using an artificial multi-layer neural network. Both the experimentally measured and neural network interpolated surface profile data were implemented to the calculation of electric current free energy. The results show that the electric current free energy can overtake surface energy and provide a significant driving force to the kinetics of surface evolution in solidification. Distortion of the electric field is noticed around the interface.
Keywords: electrical properties; kinetics; machine learning; oxide; solidification.
© 2022 The Authors.