Polywell In Europe Raising Funds

[TallDave]

Of course, the problem with an electric field that turns back the electrons is that it would pull out any ions that are there.

I haven't seen anyone try to quantify the competing electric forces at the edge of the well (negatively charged plasma ball wants to push electrons out the cusps, electrons want to pull ions with them). I suspect this is difficult to characterize without a detailed simulation. And I'm not sure I would trust a simulation anyway.

Then you haven't been paying attention. If you are not up to solving the homework, you can cheat by looking for the post where I showed that the cusp plasma will be quasi-neutral. It's not as complicated as a simulation. Back-of-the-envelope estimates are perfectly adequate.

IIRC you just calculated a force from an electron fan, without regard to cusp plugging oscillations or WB field geometry. I don't remember for sure, but I also believe you did not account for the force driving the non-ambipolar losses in a Polywell.

And I'll ask again: what do losses from a quasineutral cusp flow look like? I think we agree the answer is: catastrophically bad. But we would have seen that in WB-7 (not to mention all the machines before) and we wouldn't be building WB-8 and contemplating a reactor. So we can infer (always dangerous, yes, but sometimes justified) that quasineutral cusp flows seem unlikely.

Hee's a constructive homework assignment: under what conditions (electron densities, cusp sizes, well depth, ion distribution) does a cusp plugged with oscillating electrons have a weak enough pull (relative to the well) to limit ion losses to a degree that could produce decent confinement? (Basically we're asking: at what conditions do the ravines of the cusps at the edges fail to allow a flood of ions to leave our bowl-shaped Valley of Fusion?)

It seems intuitively obvious that if ions spend nearly all their time far enough from the cusps, then it must be difficult for them to ever see enough of a force to pull them into one. But presumably we could define some limits.

[Art Carlson]

At least as long as alexjrgreen refuses to calculate anything. I'm outta here.

Art, you're sitting on a grid computer. Of all the people here you're the most able to run the numbers.

For these questions, back-of-the-envelope is plenty good enough. Never use a grid computer when a napkin will do.

[Munchausen]

Sorry, but the party line is that the electrons are magnetically confined. At least that would work, albeit poorly. What you describe will not work at all.

Perhabs this is an awfully stupid question, but why is this so? I thought the idea of this contraption was to have a positive grid to confine the electrons?

You claim with considerable emphasis that the potential well cannot be created at any resonable cost of energy.

Besides from losses to the grid, which is the fundamental difference from the Elmore-Watson-Tuck machine? Are there any useful information to be gained from what was learned in 1959?

"On the Inertial-Electrostatic Confinement of a Plasma" William C. Elmore, James L. Tuck, Kenneth M. Watson, "The Physics of Fluids" v. 2, no 3, May-June, 1959

[KitemanSA]

Maybe we are (partly) using the language differently. I can't quote Bussard or Nebel readily, but Wikipedia uses language similar to mine: "the negative charges [are] confined to the inner region of the reactor by magnetic fields". If you want to say the electrons are confined by both magnetic fields (within the magrid) and electric fields (in the cusps), I can live with that.

Folks, Art and I can live with a common statement. This may be a first!

This discussion started with the topological difference between a sphere and a torus when it comes to magnetic confinement. Both ions and electrons can be confined by a magnetic field everywhere parallel to the surface in a torus. In a sphere you have to have cusps somewhere. If you try to plug the cusps with electic fields, you have to decide whether to stop the ions or the electrons. You can't stop both. That is why the topology always plays a role.

Concur. BUT, if you use the fields to stop the electrons, which then efficiently develop a potential well inside the core, and that well contains the ions, then you need not contain said ions with the magnets so the cusps won't leak the IONS, so the electric field that plugs the electron leaks (contains the electrons) don't interact with the ions. Wheww! TaDaa! Polywell!

Having said I can live with the common statement doesn't mean that I really think the containment is from the electric field, and the magnetic field is more of an amplifier. But what the heck, I can live with your statement.

[KitemanSA]

Never use a grid computer when a napkin will do.

I like this. Two agreements in one night!

[D Tibbets]

The electrons leave the central cusps in jets, not fans, because of the magnetic field and they're accelerating away so the jet density is much lower than that in the wiffleball.

I usually talk of fans because I think the line cusps are more important than the point cusps, but the same point applies to the jet or beam of the point cusp.

Here's an extra assignment just for you: Given (a) mono-energetic electrons with 100 keV energy and radial velocity at the edge of the WB, and (b) 100 V potential difference between the edge of the WB and the maximum potential, what is the reduction factor of the electron density between the edge of the WB and the position of maximum potential?

I'm not suer what you are referring to, the WB edge potential compared to the core, or (I suspect) the WB edge potential compared to the magrid potential. If the magrid potential then you are off by several orders of magnitued if one of Bussard's papers is acurate. He stated that thp potential well could be generated up to ~ 80-85% of the drive (magrid) voltage. So if the magrid was at 100,000 V, the potential well would be ~ 85,000 V, a difference of 15,000 V. I'm not sure how that applies to your question. If you are referring to the shape of the potential well- square or elliptical, and the strength og that gradient; then I again don't know where you are going.

Concerning ion flow in the cusps, it is obvous that there is some, or I assume that Bussard's comments about attempts to further contain electrons with repellar plates being a failure because they increased ion losses would not apply. As you said, they are opposing effects between pos. ions and electrons. The question then becomes the realitive numbers and effects by and on the charged particals in the cusp areas- both inside and outside the magrid.

I admit that I am confused by the processes. In a simplistic way I would think that the electrons outside the WB would not tug on the ions (that had not been upscattered above them) because of the symetrical cusps around the machine (Gauss's law). The ions at the nominal top of the potential well would only see the electrons below them. This is contrary to the fact that impellers didn't work, so I assume there is some intermediate result. So, whether the system works, depends on compromises. That and the claimed fact that escaping ions do not cost much energy to the system, only limits the concentration achievable within the magrid compared to the exterior. Picturing the ions and electrons in monoenergetic highly radial paths, fully thermalized paths, or any combination in between boggles my imagination and it seams it boggles the mathmatical explainations as well, or at least the assumptions that the mathmatical explainations are besed on.
Add to that the different cusp hole sizes for the ions and electrons, the differing velocities and opposing acceleration directions at the WB border confounds theoretical reduction. Add to that other poorly understood (by me) effects like Debye leangths at various heights in the potential well and at the borders (between virtual anode (if present) and the electron dominate levels, the WB border, the border between the inside and outside of the magrids, etc. Add to that turbulances, density variations, POPS type effects, etc. and what appears to be a simple sysetem becomes exceeding complex, and I assume difficult to predict with confidence without supporting aviable data.
Admittedly, some well known (?) physisics limits things, under certain conditions, but the conditions seem often vague, and work arounds are claimed to exist. Also, admittedly, there are more than a few of these work arounds and all of them need to work for the overall system to work.

Dan Tibbets

[alexjrgreen]

For these questions, back-of-the-envelope is plenty good enough. Never use a grid computer when a napkin will do.

Only if you have an open mind. As one of the greatest exponents of back-of-the-envelope calculations once said:

I was born not knowing and have had only a little time to change that here and there.

Richard Feynman

[BenTC]

At least as long as alexjrgreen refuses to calculate anything. I'm outta here.

Art, you're sitting on a grid computer. Of all the people here you're the most able to run the numbers.

I think the idea was that YOU calculate it. The human mind is very good at disregarding inconsistencies, to the degree that it prevents itself from noticing this. Its similar to Domestic Blindness. Going through the motions makes inconsistencies pop up and forces the mind to confront them. (Not that I'm in a position to do such calculations myself)

[alexjrgreen]

Doing the right calculation beats doing the calculation right. Every time.

We know that electrons leave the wiffleball - you can see the scorch marks on the older WB models.

We also know that in WB6 the electron losses dropped dramatically (the scorch marks are absent), so the electrons are being successfully recirculated back.

Art wants to show that this results in all the ions flying out the cusps.