Talk-Polywell.org

Well, it has become obvious to me that I am not going to do anything with this idea except ponder it so I now put it into the public domain.

Why not a tokawell?

Imagine a polywell in the shape of a torus. Or, as another way to think of it, imagine a tokamak with an electro-static confinement of ions while the electrons are constrained by the toroidal polywell.

Thoughts?

This came up in the big magrid brainstorming thread years ago.

The stability of plasma in a polywell at high beta derives from the magnetic field profile, weak in the center and strong near the coils, with or without the plasma. A tokawell would appear suffer from the same stability problems as a tokamak.

Ruminating on a toroidal Polywell...
As Hanely pointed out there would be macro instabilities. The outer curving containment is always concave towards the plasma. The inner border may be convex, but this makes up a minority of the confining B field surface. You cannot avoid cusps if you require always convex B fields. In the general discussions much is made of the potential well and the Wiffleball effect. But macro instabilities is also a major concern. With Wiffleball effects the density in a Polywell might be a thousand times greater that the target density in a Tokamak. I believe this means macro instabilities could grow perhaps a ~ million times faster. This means macro instabilities become a very much more significant problem if required Wiffleball confinement conditions/ density is obtained. The only way to avoid this is to control the instabilities like in a Tokamak- but at much higher frequency and corresponding efforts. The always convex B fields of the cusp quasi spherical geometry avoids this problem entirely.

I suspect a Polywell might be stretched out to almost cylindrical proportions , but this has two consequences. First, any confluence towards a common center is lessened- the core is stretched to a long cylinder- line. There is 2 D compression, not 3D. Secondly a long oval magnet might maintain the convex curvature, but I suspect at some point as the radius of curvature approaches a line, local plasma turbulance might push the fleid border outward enough that a small concave pocket forms and the macro instability then grows as with concave fields. It is not an absolute transition between stability and instability with a line separating them, but there is some mild overlap. I have no idea how close you could push to na almost cylindrical shape before macro instabilities became a problem. A torus is just a cyclinder curved onto itself with the added consequence that one side of the magnetic wall is always concave towards the plasma.

One consideration is to stack ~ quasi spherical units together. This is best visualized using the three ring magnet arrangement that I have described and is what the Lockheed design utilizes- I think. Here the two point cusps are not routes of electron escape, but windows into the next segment. Superficially this would seem to have advantages. But, the recently appreciated manifest importance of electron injection efficiency may modify the priorities enough, that this reduced cusp availability may be a net minus. Confinement may already be more than adequate, but getting the electrons/ plasma into the machine without disrupting the containment is the challenge.

Dan Tibbets

hanelyp wrote:This came up in the big magrid brainstorming thread years ago.

The stability of plasma in a polywell at high beta derives from the magnetic field profile, weak in the center and strong near the coils, with or without the plasma. A tokawell would appear suffer from the same stability problems as a tokamak.

Why? The tokamak wouldn't have the electrostatic attraction along the minor axis of the torus, would it?

The instability in a tokamak relates to magnetic field pressure gradients. The electrostatic well doesn't change this. You'd still face instability that increases with beta. I'm also thinking that the electrostatic convergence won't work the same in a low beta configuration like a tokawell would have to be, with magnetic field crosswise to electric.

Why would a tokawell "have to be" low beta?

KitemanSA wrote:Why would a tokawell "have to be" low beta?

Because like a tokamak the magnetic field macro instabilities increase exponentially with increased plasma pressure/ density. I suppose you might have high Beta with a low magnetic strength and corresponding low plasma temperature * density. But that defeats the purpose. The fusion rate could fall below the Bremsstruhlung losses, even if there is perfect particle confinement. In a Polywell the macro instabilities, or MHD or edge instabilities if you prefer is due to the unfavorable orientation of the magnetic fields in at least a substantial part of the machine. The fields are concave towards the plasma. The electrostatic potential well in a Polywell does not negate this concern. If anything it would make edge instabilities for the electrons even worse, if not for the convex shape of the magnetic fields. Even if you only consider the electrons, there would be problamatic and possibly fatal macro instabilities if not for the always convex magnetic fields inherent in the polyhedral cusp design. Electron losses would increase, and higher electron input efforts would be required or soon the Coulomb pressure would overwhelm any containment.

There has been mention of high Beta conditions in a tokamak variant, but my understanding is that this might represent a Beta of perhaps ~ 0.1 instead of the current ~ 0.03 Beta. Better, but still far below Beta of close to 1. In a tokamak I understand there are two major concerns about particle containment. ExB diffusion controlled by large size and limited density. And, macro instabilities which are unavoidable with the closed torus morphology. Various modes and interventions are being pursued to try to control this so that the working density and or temperature does not have to be decreased further. Two essential keys to the Polywell is that ExB issues are addressed by the electrostatic field potential well- for the ions, and convex magnetic fields for the macro stability. Both have to be working for the triple product to reach working levels. The ExB diffusion of electrons would eventually become the dominate loss mechanism if cusp confinement could improved enough. This would limit the smallness of the machine even with excellent electron cusp confinement and Bremsstruhlung control. The macro instabilities remains (I think) a minor player, but let the magnetic fields in even a small part of the machine become parellel or concave and at the envisioned densities the instabilities could be intolorable both from a confinement point of view and from a damage point of view.

Dan Tibbets

I'm not sure the concept is getting thru, and sorry Dan, I've stopped reading you a long time ago. Anyone who still thinks the Ni(p,?)Cu reaction is endothermic... Does anyone else want to weigh in on this?

D Tibbits does tend to ramble, but he is correct that macro instabilities in a tokamak style magnetic field get very bad very fast as plasma pressure rises. In contrast, a Buzzard polywell or skunkworks CFR style field is stable until plasma pressure approaches magnetic pressure at cusp points.

hanelyp wrote:D Tibbits does tend to ramble, but he is correct that macro instabilities in a tokamak style magnetic field get very bad very fast as plasma pressure rises. In contrast, a Buzzard polywell or skunkworks CFR style field is stable until plasma pressure approaches magnetic pressure at cusp points.

Ahh, but that is the point, this is NOT a tokamak style magnetic field, not at all.

KitemanSA wrote:Ahh, but that is the point, this is NOT a tokamak style magnetic field, not at all.

In which case I have NO IDEA what kind of configuration you're talking about, and the 'tokawell' label would not apply.

hanelyp wrote:

KitemanSA wrote:Ahh, but that is the point, this is NOT a tokamak style magnetic field, not at all.

In which case I have NO IDEA what kind of configuration you're talking about, and the 'tokawell' label would not apply.

Imagine a torus tiled with rectangles, each successive rectangle being a magnet with alternating field directions. Stuff in electrons and viola, a toroidal potential well.

KitemanSA wrote:Imagine a torus tiled with rectangles, each successive rectangle being a magnet with alternating field directions. Stuff in electrons and viola, a toroidal potential well.

I can see that working. I propose the name 'Toruwell'. But not practical compared to a spherical or cylindrical version.

KitemanSA wrote:

hanelyp wrote:

KitemanSA wrote:Ahh, but that is the point, this is NOT a tokamak style magnetic field, not at all.

In which case I have NO IDEA what kind of configuration you're talking about, and the 'tokawell' label would not apply.

Imagine a torus tiled with rectangles, each successive rectangle being a magnet with alternating field directions. Stuff in electrons and viola, a toroidal potential well.

Where does the electric field come into it? Can you make a drawing of it?

ohiovr wrote:Where does the electric field come into it? Can you make a drawing of it?

It is like a cubeoctahedron Polywell except the core is not spherical but toroidal. And no, not, really. All I have currently is an iPad with free software.