P. Ortega et al., PREDICTIVE MODELING OF POSITIVE LEADER PROPAGATION UNDER STANDARD ANDOSCILLATORY IMPULSE SHAPES, Journal of physics. D, Applied physics, 27(6), 1994, pp. 1233-1241
First we define an optimum discharge corresponding to a minimum energy
consumption (optimum discharge), which may not necessarily correspond
to the shortest discharge path. The definition results from charge an
d voltage measurement in a 2.2 m point-to-plane air gap energized by s
tandard and oscillating impulse shapes. The second step is the creatio
n of a macroscopic model and fitting it, to compare actual discharges
with the optimum one. The variation of observable discharge parameters
such as axial length of the leader channel, velocity of the leader ti
p, apparent injected charge, time-to-breakdown and applied voltage val
ue just before breakdown are readily simulated using this model, whate
ver the waveshape. The third step is to validate the model using previ
ous experimental results on several gap spacings: 5, 10 and 17 m (stan
dard impulse shapes) and 2.2 m (oscillating impulse shapes). Two main
phases are shown to exist in both experiment and model: an unstable on
e resulting from the first corona event and, in some cases, a quasi-st
eady one occurring before the final jump (breakdown cases). For a brea
kdown case, the U50 value can be deduced from a U0 value, the withstan
d level, given by this model. Finally, the well-known U curves for swi
tching impulses are easily reproduced using our modelling work.