Skip to Main content Skip to Navigation

Modélisation de la décharge positive dans les grands intervalles d'air

Abstract : This work presents an autonomous discharge model allowing to predict the dielectric strength of air gaps submitted to a given voltage shape (pure or oscillating). This model is based on an equivalent electrical network, the parameters of which varying with time according to the channel characteristics and the discharge geometry. The leader propagation is based on a criteria linked to the calculation of the field on its tip and where the random path of the discharge is taken into account. The model allows to determine the currents evolution (leader and arc), the corresponding charge, the potential gradient in the channel, a plausible trajectory of the discharge, its instantaneous velocity, the temporal evolution of its thermal channel, the power and energy injected in the gap, breakdown voltage or time to breakdown. Moreover, for a given configuration, this model constitutes a tool for the determination of the U50 voltage (voltage at 50 % breakdown). It also allows simulating the image converter working frame or streak mode. The results issued of the elaborated model are in good agreement with those obtained experimentally. On the basis of great similarities laboratory sparks - atmospheric discharges, a positive lightning model has been elaborated. The electromagnetic environment associated to the leader has been also characterized. The results obtained by the model are also conform to those measured during natural lightning.
Complete list of metadata
Contributor : Publications Ampère <>
Submitted on : Wednesday, May 2, 2007 - 9:31:01 AM
Last modification on : Wednesday, March 31, 2021 - 3:10:05 AM
Long-term archiving on: : Tuesday, September 21, 2010 - 1:32:27 PM


  • HAL Id : tel-00143944, version 2


Issouf Fofana. Modélisation de la décharge positive dans les grands intervalles d'air. Energie électrique. Ecole Centrale de Lyon, 1996. Français. ⟨tel-00143944v2⟩



Record views


Files downloads