Another arguement against Ambipolar cusp flows

[D Tibbets]

Another attempt to force others to my beliefs.

The ambipolar arguments for cusp charged particle flows have been repeatedly given as a non negotiable property of plasmas. While I'll accept that this is necessary for neutral plasmas, the unwillingness of some to even consider that the Polywell plasma is not neutral (and its consequences) seems strange. Non neutral plasmas are readily admitted by several texts, but they state that they are outside the scope of their considerations. While this is probably necessary in order to keep the physics manageable (and because it is not necessary for considerations of mainstream Tokamak research), it in no way negates the possible application of non neutral considerations.

A little different perspective is given by the following definition.

http://www.alcwin.org/Dictionary_Of_Pla ... n-70-A.htm

"Term : Ambipolar plasma potential
Definition : Electric fields that are self-generated by the plasma and act to preserve charge neutrality through ambipolar diffusion."

In that definition the net potential (zero in a neutral plasma) drives the charged particles to ambipolar flows through any exit path- which could be magnetic diffusion, cusp flows, or groundings.

But, if the potential is not Zero, by this definition the charged particles flows will be such that the plasma will leak the charge that is the same as the potential. In the Polywell's case, electrons will have a potential pushing out, while positive ions will be inhibited from leaving. This is the whole idea behind electrostatic containment. I believe no one has any difficulty accepting this idea. Why this idea would be suddenly ignored in a cusp region is beyond me*.

Yes there is a very strong coulomb force that limits the amount of charge separation you can achieve, but inversely, this same force constrains the movement of charged particles that are exposed to that force. IE: if you can maintain a potential against escape mechanisms (consider resistance in an electrical circuit) then the effects of that potential will be felt by any charged particle that is exposed to it. A comparison can be made with an electrical current in a wire. The electrons are mobile, the atomic nuclei (sort of temperary ions) are not. Ambipolarity would dictate that there would not be any current flow, because the electrons and ions need to stay equal in number (neutral charge). But, if excess electrons are introduced on one side of the wire from a battery (electron gun in the Polywell's case) the electrons will flow to ground, while the ions happily stay in place. so long as the resistance is enough that the battery is not drained, this condition can persist indefinitely.

I believe this describes the Polywell situation. The resistance (cusp losses and other loss mechanisms) is good enough that the potential can be maintained at reasonable costs. And, this maintained potential has to limit the positive ion flows to levels proportionate to the potential. Inversely you can say the potential forces more electrons to exit, but the end result is the same. If more ions are added, the potential drops, and the restraint on the ions exiting are relaxed somewhat, but never absent, unless the potential difference inside and outside the magrid is zero (there is no potential well). There are complications that need to be considered, but that is the basic premise.

* In a strongly coupled plasma the approach to ambipolar flows despite the potential would be more tenacious, but it could never be purely ambipolar (in a non neutral plasma) until the coupling between the electrons and positive ions was absolute. IE: the electrons were captured into orbits and result is neutral atoms. Again, fusion plasmas are weakly coupled according to several sources. I assume thy are referring to plasmas with ~ 10^19 to 10^20 particles /M^3 like Tokamaks. The Polywell might have a higher density, but I guessing this would only slue the results a tiny amount.

Dan Tibbets

[chrismb]

A non-neutral plasma may also experience ambi-polar flow. The 'overcharge' will, indeed, tend to pull out of the plasma body more rapidly iinto a region of lower charge, that if it were emanating from a neutral plasm into a charged region, but if the space-charge 'ahead' is the same as the non-neutral plasma 'behind' but the plasma is a lower density or a lower magnetic field, then you will get 'ambipolar' flow from the non-neutral plasma [albeit slightly non-ambipolar in so far as the ratio of the charges will be preserved].

It's all up to Gauss. He's just trying to make sure that neutralising charge gets the priority over any thermal motion.

[TallDave]

I'm not aware that anyone has asserted flows must be ambipolar (though this is a common assumption in neutral plasmas, because they're typically heated neutrally), just that plasmas must be quasineutral.

Joel has simulated a plasma where cusps and exterior are not even quasineutral, let alone ambipolar. Art has claimed this requires huge > MV charges on the Magrid but as I and others suspected at the time, the picture is more complicated than a simple equation can capture and full simulations give a different result (thanks again Joel!).

We don't really know for sure by experiment yet whether cusps/exterior are quasineutral. Rick might have an idea by now, courtesy of those all lovely WB-8 ports.

But I'm not sure how much it really matters, since the whole thing is e-driven. There will always be electrons going places ions don't (such as the Magrid via cross-field transport), and ions will not be able to drag electrons with them to the wall because electron behavior dominates due to the excess of electrons everywhere.

I'd be really interested to see what equations EMC2 is using for cross-field transport...