TallDave wrote:Art Carlson wrote:In simple terms, the problem is that an electric field that causes electrons to return will cause ions to be lost with high energy. If you want to fill out the argument, you need to consider quasi-neutrality.
At the edge, I might be able to get away with violating quasineutrality a little bit.
I'm still wondering exactly how strong the force that preferentially expels electrons in a slightly electron-rich plasma would be. Presumably it also sucks on ions. And there's also the Magrid, which makes the edge the "bottom" of the electostatic well for electrons anyway.
Attracting ions could definitely be an issue -- Bussard reported big ion loss problems when he put electron emitters at the cusps. But if the cusps are squeezed small enough and the ions are far enough away, the "pipe" might not carry enough electrons to pull the ions out away from the well in the center.
If we're getting WB confinement as claimed, either the ions aren't getting sucked out at any appreciable rate or WB confinement doesn't work this way.
I understand that quasi neutrality in the cusps is unresolved. I recall that Dr Nebel said they were not quasi neutral based on their 1.5 d simulations, but that experimental measurements were nod adequate to determine the validity of the simulation either way.
Below is a simple model, ignoring magnetic effects except for the cusps being narrow tubes that charged particles can transit. Also, ignoring sheaths that I don't understand anyway.
I see two domains inside the cusps, the region inside the magrid that does not see the positive charge (or alternately the neutral charge) on the magrid and the region outside the magrid that does see it. Assuming the electrons are fast and the ions are slow as they enter the cusps ( I believe this is consistent with the premise of how the potential well needs to operate for the machine to work at all, instead of being merely a bag of thermalized plasma). There is an excess of electrons inside the machine, so the ions would have some force pulling them back, but assume that locally the high speed electrons ahead of them are dominating and accelerating (or at least preventing the ions from decelerating) them outwards. These outward bound electrons would in turn be slowed by the trailing ions.* If this domain is the whole story then I could see the ions escaping to the outside readily. But, in the second domain the outbound electrons suddenly see the positive charge and many of them will reverse and be acellerated back into the machine due to the magrid charge, in addition to the ions trailing, but not yet to the border. These inbound electrons would pass the ions and start pulling them back in, hopefully before the ions inertia and leading outbound electrons(and inbound electrons that have yet to pass the ion) reach the border where they see the positive magrid charge and fly to the walls. Or to state it another way- the outward pressure due to the inertia of the upscattered ions and the escaping electrons in the cusp is counterbalanced by the inward pressure from the excess electrons inside the machine and the inward bound recirculating electrons. Hopefully this would keep the ion confinement time as good or better than the electron confinement time.
Also, keep in mind that escaping ions may not contribute much to the energy balance( they are already at the top of the energy well- again assuming we are not talking about a thermalized plasma), but only to the need to keep the vacuum levels outside the machine low enough to prevent arcing.
* Considering ions that are leading an electron, superfically the effects would be opposite from what I described above. But with the difference in the speeds of the ions and electrons in the two domains- ions slower inside and rapidly accelerated to fast outward speeds once outside the magrid, while the majority of the electrons are fast inside and slow outside (or fast reversed speed) the local densities of the relative charged particles may faver the recirculating of the ions along with the electrons. I suppose this could be called cusp plugging(?).
Dan Tibbets
To error is human... and I'm very human.