Speaker
David Trdlicka
Description
See the full Abstract at http://ocs.ciemat.es/EPS2018ABS/pdf/P4.3007.pdf
Modelling of streamer propagation in dielectric liquids
using a dense gas model
D. Trdlička1 , J. Karel1 , P. Bílek2 , J. Fořt1 and Z. Bonaventura2
1 Department of Technical Mathematics, Faculty of Mechanical Engineering, Czech Technical
University in Prague, Prague, Czech Republic
2 Department of Physical Electronics, Faculty of Science, Masaryk University, Brno, Czech
Republic
Ultrafast electrical breakdown in dielectric liquids is of considerable interest for applications
in high-voltage insulation. For liquids with high mobility of charged particles, the breakdown
takes place on a nanosecond time scale, and its mechanism is similar to the streamer breakdown
of gases which is caused by the electron impact ionization of particles, with a distinction that
the electron–ion recombination in the streamer channel plays a significant role compared to
streamers in gases. Also very high-voltage pulses of sub-ns duration provide an extremely high
electrical field in the plasma formation region in the liquid and allow ionization directly in the
condensed phase by direct electron impact. Thanks to the sub-nanosecond time scale, the fluid
lacks time to expand, and so the discharge is formed directly in the liquid phase [1, 2]. In this
contribution we present a study positive streamer dynamics in dielectric liquid using a dense gas
model [3]. The electric discharge propagation in liquid is described by the set of convection-
diffusion equations with source terms for charged particles coupled with the Poisson equation
for the electric potential. The equations are discretized by finite volume method (FVM) on 2D
unstructured triangular grid. The convective terms are computed by upwind scheme and the
accuracy of the scheme is increased by linear reconstruction restricted by the Barth-Jespersen
limiter. The dissipative terms are discretized by the diamond scheme and central approxima-
tion. The second order in time is guaranted by three steps Runge-Kutta method. The Poisson
equation is discretized analogously as dissipative terms in convection-diffusion equations, and
the system of linear equations is afterwards solved by LU decomposition. Implemented multi-
level dynamic grid adaptation algorithm allows to capture sharp peaks and steep gradients of
unknowns occurring in moving region of the streamer head.
This contribution is funded by Czech Science Agency grant no. 18-04676S.
References
[1] Starikovskiy A 2013 Plasma Sources Sci. Technol. 22 012001.
[2] Starikovskiy A, Yang Y, Cho Y and Fridman A 2011 Plasma Sources Sci. Technol. 20 024003.
[3] Babaeva N Y and Naidis G V 1999 Tech. Phys. Lett. 25 91.
