Possibly. However, the electrons they would be pulling out would effectively be the electrons they left behind when the fuel was ionized.ravingdave wrote:A worrying thought has occurred to me about alpha's streaming out of the cusps. Wouldn't they drag the electrons out with them thereby killing the wiffleball ?
I mean, if they are coming out in a stable beam or some such, then does this not create a sort of static pseudo virtual positive electrode all the way from the core to outside the MagGrid ?
Yeah, the alpha particles may be streaming individually, but collectively they are a large positive attracting force. Eventually, the electrons have to notice this and start moving towards this statically stable stream.
If i'm explaining this idea clearly enough, does anyone have an explanation as to why this won't happen ?
I remember Art Carlson talking about this exact effect with nothing more than just fuel ions, so if it is possible to pull electrons out of the wiffleball with fuel ions, wouldn't it also occur with Alpha particles ?
Just wondering.
David
It Is A Different Machine
Engineering is the art of making what you want from what you can get at a profit.
The alphas won't "pick up" electrons; not at these energies. You won't get a neutral helium beam slamming into the wall.
I estimate 70 A total alpha current for a 100 MWf reactor (~5 A per cusp). Given the exit speed of the alphas, I would expect a line density of about 0.4 uC/m, which for a 5-cm-wide beam results in a potential disruption near the cusp on the order of 10-20 kV IMCAC, assuming no electrons follow them out and flatten the distribution. Completely nulling this effect with a similarly-sized beam would take maybe 800 A of electron current at 50 kV, resulting in a 40 MW outflow (~3 MW per cusp), most of which would be recovered by proper trap design (really, at this power level most of the electrons should just drop back into the cusp, especially considering that they slow down as they travel outwards, producing a negative charge concentration - so maybe this should be 400 A each way, for 20 MW cusp recirculation). This amount would be substantially reduced if low-energy ionization electrons were to move in and plug the cusp, since they can't get very far outside the magrid anyway.
Also remember that the cusp is a lot narrower for the electrons than for the alphas; 5 A at a couple of cm doesn't generate much of a magnetic field. This means the electron beam would have to be much tighter, restricting the amount of actual outflow that could occur.
Actually, that's probably the answer - the high-energy electrons (the ones you don't want to lose) will leave the cusp anyway, whether or not there's an alpha stream. The question is, does the alpha beam cause a sufficient rise in potential that the electrons won't come back? I don't think it does.
This is all BoE at most. What am I missing?
I estimate 70 A total alpha current for a 100 MWf reactor (~5 A per cusp). Given the exit speed of the alphas, I would expect a line density of about 0.4 uC/m, which for a 5-cm-wide beam results in a potential disruption near the cusp on the order of 10-20 kV IMCAC, assuming no electrons follow them out and flatten the distribution. Completely nulling this effect with a similarly-sized beam would take maybe 800 A of electron current at 50 kV, resulting in a 40 MW outflow (~3 MW per cusp), most of which would be recovered by proper trap design (really, at this power level most of the electrons should just drop back into the cusp, especially considering that they slow down as they travel outwards, producing a negative charge concentration - so maybe this should be 400 A each way, for 20 MW cusp recirculation). This amount would be substantially reduced if low-energy ionization electrons were to move in and plug the cusp, since they can't get very far outside the magrid anyway.
Also remember that the cusp is a lot narrower for the electrons than for the alphas; 5 A at a couple of cm doesn't generate much of a magnetic field. This means the electron beam would have to be much tighter, restricting the amount of actual outflow that could occur.
Actually, that's probably the answer - the high-energy electrons (the ones you don't want to lose) will leave the cusp anyway, whether or not there's an alpha stream. The question is, does the alpha beam cause a sufficient rise in potential that the electrons won't come back? I don't think it does.
This is all BoE at most. What am I missing?
Since the alphas are not likely to re-neutralize but rather drag the electrons with them magnetic separation would probably do the trick. And since the device itself generates a magnetic field the separation is intrinsic. Will it be good enough? Hard to say.Aero wrote:Its a bigger problem than that. If the alpha picks up one electron, then the power of direct conversion is halved, and if it picks up two electrons then we're left with a stream of high energy neutral particles, no direct conversion.
There will be a lot to work out once we can prove net power.
Engineering is the art of making what you want from what you can get at a profit.
I know this is offtopic, but in response to what you said, MSimon...MSimon wrote:There will be a lot to work out once we can prove net power.
Imagine it's the moment Rick Nebel and EMC2 are satisfied that the polywell works. So it's reasonably assured that an investment in the engineering questions will result in a net power machine. I say "reasonably assured," but I'm convinced it will be beyond any possible doubt.
There will be two fronts to deal with. One front is the engineering "lot to work out," because of these important design issues.
The other front will be fielding the press and the naysayers (and politically motivated opponents) who will not "get it" yet.
On both those fronts, this website will be a priceless resource for information and archived discussions.
I think the "non-ambipolar plasma diffusion" argument will have something to say here.
Some of these questions are querying effects that are intertwined with multiple competing factors. Back of the enevelope is inadequate as well maybe months of supercomputer CPU time and best guess plasma equations.
Build it and see, if you're feeling lucky.
Some of these questions are querying effects that are intertwined with multiple competing factors. Back of the enevelope is inadequate as well maybe months of supercomputer CPU time and best guess plasma equations.
Build it and see, if you're feeling lucky.