Comments on the resistivity of high pressure, nanosecond, spark-gap switches
- Air Force Research Lab., Kirtland AFB, NM (United States)
- Sandia National Labs., Albuquerque, NM (United States)
- Stevens Inst. of Tech., Hoboken, NJ (United States). Physics Dept.
An important class of compact, repetitive ultra-wideband pulsers makes use of the rapid resistance collapse of a spark gap, either self-breaking or triggered, as a closing switch to launch an electromagnetic pulse into a waveguide and ultimately to an antenna. The dynamics of the spark gap during its breakdown phase determine its resistance, inductance, and the risetime of the current, all of which will affect pulser performance. While the early-time streamer phase involving complex kinetic processes on a microscopic scale determines the switch delay and jitter, switching is observed to occur only after the streamer has established a partially ionized channel. The flow of current through this channel provides Joule heating, which not only makes the channel more conducting by increasing the number and mobility of charge carriers, but also causes it to hydrodynamically expand, increasing its cross sectional area. This expansion, which has been shown to be the dominant, though not the only, mechanism affecting channel resistance, at least during the first quarter cycle of a discharge, was modeled first by Braginskii. In the few-hundred-nanosecond timescale, considerable effort has been directed toward matching experimental results with the model of Braginskii for channel expansions while using a variety of assumptions for electrical conductivity, {sigma}. In the nanosecond timescale, however, data is more limited, and a recent effort aimed at explaining this data with the Braginskii model has neglected time dependence of {sigma}. In this paper the authors show that the nanosecond-time-scale data of Sorensen and Ristic is not in fact well modeled by a constant {sigma}, particularly near and after current maximum. The authors show that good agreement with experiment requires that they account both for expansion-related cooling and radiation losses from the channel to obtain a conductivity that agrees well with experiment.
- OSTI ID:
- 346854
- Report Number(s):
- CONF-980601--
- Country of Publication:
- United States
- Language:
- English
Similar Records
Nanosecond spark gap switching measurements
Performance of a 30-kV, 1-kHz, nanosecond source