Transverse 2D gliding arc modeling

摘要

Summary form only given. There seems to be growing interest in the numerical simulation of the Gliding Arc (GA) discharge1,2. Results published so far have been 2D simulations of two geometries: (1) axisymmetric stationary discharge channel and (2) “plasma sheet” immersed into 2D laminar gas flow between two diverging electrodes infinite in the third dimension. These approaches however, do not reveal GA interaction with the gas flow, which can be addressed using transverse 2D modeling presented in this paper. GA moves slower than the gas flow and when discharge channel temperature is in the order of 1000 K, transverse gas flow results in strong interaction between the discharge and gas flow, i.e. gas flow partially penetrates and cools the discharge very efficiently, as it accelerates and expands because of decrease in density. On the other hand, most of the gas flows around the discharge and forms so-called Karman vortex street. Thus, presence of discharge inside the gas flow significantly disturbs both of them. Our simulation shows that discharge moves slower than the gas flow even without accounting for interaction with electrodes, because initially a hot stationary discharge channel forms inside cold gas flow. Sudden formation of this stationary channel stops gas in front of it, and therefore acceleration of the discharge channel is not as fast as expected. We used Fluent© with a subroutine that calculates local heat release due to electric current using look-up table where electric conductivity of argon was calculated as function of the reduced electric field and gas temperature. This table was developed using BOLSIG+ and equilibrium electric conductivity of argon dependency on temperature. Initially, slowly increasing electric field was applied to a small conductive spot in a rectangular domain with upward laminar gas flow until the total current value through the spot reached a predetermined value. Power release in each cell of the domain and resulting increase in temperature were computed and used by Fluent© for modified flow calculation. Repeating such iterations revealed peculiarities of the discharge-flow interaction. We hope that demonstrated results will help other researchers involved in development of GA modeling approaches and highlight the importance of gas-discharge interaction.