Dielectric Breakdown
This is the season when our second-year students are picking the projects they want to do in their third year, as are those third-year students intending to carry on for Year 4 of the MPhys programme. I’ve been chatting to quite a few students about this particular project so thought I’d do a quick post here.
One of the third-year projects I’ve got in the current catalogue for next year concerns a computer model of dielectric breakdown based on the idea of diffusion-limited aggregation. This is a neat model that allows the students to simulate pretty patterns like the one shown on the left.
The mathematics of it was first presented in Niemeyer, L., Pietronero, L., Wiesmann, H., “Fractal dimension of dielectric breakdown,” Physical Review Letters 52 (1984), 1033-1036.
Dielectric breakdown happens when a sufficiently large voltage is applied across a material that doesn’t normally conduct electricity, the classic example being a lightning strike. Here’s another example, which I find particularly electrifying…
Classical physics can be fun, you know!
Follow @telescoper
In the Dark 
May 9, 2012 at 10:36 pm
What’s going on at the start? Are some fuses released in response to the initiation of a spontaneous arc, or does the arc form in response to the movements of those bits of the apparatus?
Good old plasma physics!
May 9, 2012 at 11:01 pm
From what I understand it’s pretty complicated to disconnect main power lines from substations – especially when they’re working at 500kV like this one! – so they have complicated mechanical switchers that contain a gas that quenches any sparks formed when they move apart. There are three phases in the supply, so three switches are used. In this video apparently the gas in one leg of the circuit failed to quench the sparks and off it went. It was videotaped as part of an equipment test.
May 9, 2012 at 11:01 pm
See http://www.capturedlightning.com/frames/longarc.htm#500_kV_Switch for an explanation.
May 10, 2012 at 10:56 am
Thanks. The writer is remarkably blase about the “upstream circuit breakers” though, which presumably face the same problem.
May 10, 2012 at 2:15 pm
The problem with this particular breaker/air switch was that the dielectric gas (SF6) in one of the circuit breaker “bottles” became depleted so the breaker couldn’t clear the internal arc when it tried to open. This caused the remaining half of the breaker to be overvolted, and it flashed over. A properly-functioning breakers will easily handle the task of breaking the relatively low reactive current from an unloaded line. BTW, larger versions of SF6 breakers are now being tested for a new 1.2 milllion volt transmission line to be installed in India. A picture of this massive breaker is in the middle of the May, 2012 issue of IEEE Spectrum,