simulating ion/electron injection by way of ionization quest

[happyjack27]

so i'm having trouble with the eguns approach. but now i've read that ionization is the way to go for large systems (which i had presumed, at least for the ions). i know from my "brute force" test that that effectively forms a well, if not as efficiently as one would like. so i'm just going to go ahead and simulate the injection via ionization. but i have questions regarding the spatial distribution of ions and electrons formed by ionization. namely,

1) what would be ideal? presumably electrons in the center, and ions on the edge. but that's impossible. by definition of ionization the electrons and ions have to start off in the same place.
2) what's practical? is there a greater concentration in the center or the outside? is it uneven, like is there really more in the top, when done via gas puffing?
3) how much does it really matter?

my default plan is just to start them all off evenly distributed by volume within a sphere about 75% of the magrid radius. is there anything overly unrealistic about this? will it significantly affect the operation of the machine (e.g. alter the power gain / power loss)?

[D Tibbets]

Two problems. The secondary electrons created from cascading ionization of a neutral gas near the edge are initially low average energy (perhaps ~ 100 eV). You need the continuous injection of hot electrons to heat these cool electrons so that a deep potential well can form.* Though it would not give meaningful electron input energy levels, you could multiply by a constant to magically increase these cool electrons to the designated potential well voltage on the edge, while leaving the created ions alone, so they remain cool on the edge.

The other problem may be the random vectors of these secondary electrons (no radial preference). Actually, I don't know precisely how the injected electrons heat the secondary electrons in a primarily radial direction. Lets see... a slow moving electron on the edge is hit by a hot radial electron. Is the resultant deflections preferentially radial?

* This assumes the electron are mostly radial in the working machine, as opposed to random directions (not speed)., which may actually be the case (at least until the ions start to drag the electrons towards the center (square to parabolic well evolution)).

Dan Tibbets