Caution: Research Rant Ahead!I wish I had a time-machine. I would go back in time to 1660, travel to London (maybe the other way round, in fact), burst into the founding of the Royal Society and beat the snot out of them, until they agreed to use the SI and nothing but the SI for all existing and future units, forevermore and until the end of time, but at least until 2009.
For weeks I have intermittently tried to calculate a theoretical breakdown curve, without much success. One of the difficulties (after you have written a program to solve the stupid transcendental equation that laughingly calls itself "breakdown condition") is finding values for the diffusion coefficients, Townsend coefficients and electron mobility in the literature, particulary for the high electrical fields I am using. All those values depend on the electric field over the concentration of your neutral gas, E/N, which is Volts times square metres in good, sensible SI units. But noo, nobody likes using those. Instead, you find Vcm^2, V/(cm torr), V/(m torr), V/(cm mbar), V/(m bar) and my personal favourite, Townsend (Td), which is Vcm^2 x 10^(-17).
Converting all those units is a pain in the ass in and of itself, but the problem is excasperated in that I was using the wonderful little Boltzmann equation solver BOLSIG to calculate my coefficients, and only now found out that BOLSIG expects Townsends, not Vcm^2 as an input. Typical example of garbage in, garbage out - I got diffusion coefficients that were for values 14 orders of magnitude smaller than I was looking for.
Anyway, this is now resolved, so behold the diffusion coefficients (times N) for argon, from 10 to 10000 (grmbl) Townsend.
2 weeks ago