Non-Conductive Can?

[KitemanSA]

All the pictures I have seen show metalic cans around the coils and metalic bridges between them. I can sort of understand the coil can, it carries the positive charge for the MAGrid. But wouldn't a simple strip on one face do that just as well?

If the can and bridges were non-conductive, wouldn't they pick up a minor negative charge and help prevent further electron losses?

Also, wouldn't it be better to move the bridges back from the cusp to project a smaller cross section?

Just some random thoughts.

[drmike]

An outside insulated can would mean you need to put ground some where. An insulated surface would lead to something like the "plasma ball lamps" where you have arcs that wander all over the place. Only for a polywell device, the arcs would be capable of blowing holes through things. A uniform surrounding ground is a lot easier to work with. It can be far away from the structure though, so you could put a few grids in there to help shape the field distribution.

For connections between coils it can go either way. You might even want to mount each coil separately so the region between them is free to allow currents to flow with no obstructions. It'd be an interesting comparison to study.

[KitemanSA]

drmike,

Thank you for your response.

Obviously I am missing something, which is not surprising being a mechanical vice electrical engineer. But what are we trying to ground with the metallic coil can? I though it was kept at a significant positive voltage.

My idea is to replace the metallic can and bridge with a non-conductive one and add a metallic strip to hold the positive charge of the magrid. Then, under operation, allow the non-conductive can and bridge to build up a slight negative (or perhaps just not so positive) voltage, thereby reducing the electron loss.

True, if the negatives build too high there would be some transfer to the positive grid, but arcing? Wouldn’t adjusting the conductivity of the “insulator” prevent that? And as it is, any electrons that do touch the can and bridge are immediate losses.

BTW, how does an insulator compare to metal with respect to "sputtering" resistance?

[hanelyp]

A buildup of negative charge on the coil shells would mess up the electric fields that help with injection and recirculation of electrons. The positive grid is an important part of how the polywell operates. And I would expect arcing between the positive ring and the edges of the negative charged insulating shell.

[drmike]

hanelyp brings up a good point. The idea of pure metal all around the coils is to ensure a uniform electric field. If you go from metal to insulator which would be necessary for the strips, the jump from metal to insulator will have a huge electric field. That will cause arcing, especially in a dense plasma.

You got it right the first time - the electrons are lighter so they will charge the insulator negative. That builds up a protective sheath, but since the walls are negative they will tend to pull ions towards them. The current that flows into the walls gets neutralized once the ions get slowed down and stick to the wall. So it is a loss mechanism for a plasma.

I think sticking to metal wall around the coil makes the most sense, and using shielded supports at the same potential also makes sense. At some point you have to put in a break with insulator, but it can be close to the outside wall so the plasma density is really low and your loss fraction won't hurt.

[MSimon]

There is also the thermal problem of alpha impingement.

Insulators are not known for high thermal conductivity.

[KitemanSA]

There are a number of excellent electrical insulators that are also very good thermal conductors; most ceramics with diamond like structures, IIRC. Boron nitride comes to mind, though I can't find the data on it just now.

You got it right the first time - the electrons are lighter so they will charge the insulator negative. That builds up a protective sheath, but since the walls are negative they will tend to pull ions towards them. The current that flows into the walls gets neutralized once the ions get slowed down and stick to the wall. So it is a loss mechanism for a plasma.

Again, perhaps I misunderstand the general principles of the MAGrid, but I was under the impression that the quasi-spherical nature of the grid made it so that it cancels out all effect INSIDE the sphere, so the negative portions of the coil can would have no effect on the ions; though there may be a need to pump SOME additional electrons out of the posi-strip to balance the general state of charge. for the external electrons. Hmmm.

[eros]

For connections between coils it can go either way. You might even want to mount each coil separately so the region between them is free to allow currents to flow with no obstructions. It'd be an interesting comparison to study.

I think electrons recirculate back or something. Coil support needs to go through that electron cloud. If support structure go through at thinnest area, it maybe get lessest hits. I quess..

Anyway it is 100% clear that inside support structure should go amperes that shield it from electrons and ions.

[MSimon]

There are a number of excellent electrical insulators that are also very good thermal conductors; most ceramics with diamond like structures, IIRC. Boron nitride comes to mind, though I can't find the data on it just now.

It is brittle. Hard to work. And thin sections are difficult. And as I recall it is not cheap.

[KitemanSA]

There are a number of excellent electrical insulators that are also very good thermal conductors; most ceramics with diamond like structures, IIRC. Boron nitride comes to mind, though I can't find the data on it just now.

It is brittle. Hard to work. And thin sections are difficult. And as I recall it is not cheap.

True, but BN was provided ONLY as an off the top of the head example. SiC may do almost as well and is readily workable. I am pretty sure that if it is in fact desirable, one can be found that is functional. It remains to be seen if it would help.

[MSimon]

There are a number of excellent electrical insulators that are also very good thermal conductors; most ceramics with diamond like structures, IIRC. Boron nitride comes to mind, though I can't find the data on it just now.

It is brittle. Hard to work. And thin sections are difficult. And as I recall it is not cheap.

True, but BN was provided ONLY as an off the top of the head example. SiC may do almost as well and is readily workable. I am pretty sure that if it is in fact desirable, one can be found that is functional. It remains to be seen if it would help.

SiC is used to face metal working tools. It is brittle and not easy to work. That is why tools that use it have steel backing.

[KitemanSA]

SiC is used to face metal working tools. It is brittle and not easy to work. That is why tools that use it have steel backing.

When proposing a non-conductive can, I envisioned a barrier to the immediate loss of electrons. Perhaps a non-conductive coating on a can would suffice?

[MSimon]

SiC is used to face metal working tools. It is brittle and not easy to work. That is why tools that use it have steel backing.

When proposing a non-conductive can, I envisioned a barrier to the immediate loss of electrons. Perhaps a non-conductive coating on a can would suffice?

You then have the problem of alpha sputtering blasting non-fuel elements into the reaction chamber.

If we go to pB11 the thought is to coat the coil cans with B11. The ITER folks have tried that to solve their first wall problem. (elemental Boron rather than refined) It might work.