magrid configuration brainstorming

Discuss how polywell fusion works; share theoretical questions and answers.

Moderators: tonybarry, MSimon

windmill
Posts: 29
Joined: Thu Jul 19, 2007 3:19 pm
Location: Toronto, Canada

Post by windmill »

Being a mechnical designer makes me look at these configurations from the standpoint of "how are you going to support these things?" Inside a vacuum tank. On megavolt standoffs. That support TONS of vectored forces trying to push them apart (or together, depending.) Through which you have to pass cryogenic coolants and high current/voltage power, and fit meter-scale multi-Tesla superconducting field coils. Around which are multi-megavolt grids. All bathed in hundreds of megawatts of alpha or neutron flux, with electrons wanting to jump on anything not magnetically insulated.
Any workable configuration is going to have to satisfy the physics and the engineering.

JohnP
Posts: 296
Joined: Mon Jul 09, 2007 3:29 am
Location: Chicago

Post by JohnP »

multi-Tesla superconducting field coils
I thought the fields would be on the order of kilogauss?

Solo
Posts: 261
Joined: Fri Jun 29, 2007 12:12 pm
Location: Wisconsin

Post by Solo »

@John: I'm guessing that the large machines need bigger B-fields, though I don't know by how much, off-hand. It probably has something to do with Bussard's scaling laws. But I still had the impression it wouldn't make it to the T level.

@Windmill: good to get some input from someone who takes a different perspective. Any way the polywell plays out, it's going to be a big engineering feat. I was hoping we might optimize the topology to get rid of the trade-off that has to be made between electron confinement and losses to the grid. But now that you mention it, probably a bigger issue would be how to reduce the alpha flux intercepted by the coils.

@hanelyp: you've got the right idea: what are the other possible electron containment configurations? That's what I want to know. I think the tokomak is out, though, and probably anything else that is extremely far from spherical, due to the need to keep any scattered ions in the system.

hanelyp
Posts: 2255
Joined: Fri Oct 26, 2007 8:50 pm

Post by hanelyp »

93143 wrote:The trouble with a tokamak configuration is that the centre of charge for the electron distribution is outside the confinement area (ie: in the hole in the middle of the donut). ...
Yes the center of charge would be outside the containment volume, but the electric field generated by the electron ring still points towards the electron ring. To visualize, consider the electrons confined along a line in space, with the resultant electric field. The Ions would be confined to a cylinder coaxial with the line. Then take that line and bend it around into a ring. The ion confinement still surrounds the electron ring, though probably a bit distorted.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

Indrek has done a nice bit of interactive math showing the intercepted areas.

http://www.mare.ee/indrek/ephi/area/

Pick you coil size. Coil Thickness. Coil spacing and Voila. Intercepted area.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

windmill wrote:Being a mechnical designer makes me look at these configurations from the standpoint of "how are you going to support these things?" Inside a vacuum tank. On megavolt standoffs. That support TONS of vectored forces trying to push them apart (or together, depending.) Through which you have to pass cryogenic coolants and high current/voltage power, and fit meter-scale multi-Tesla superconducting field coils. Around which are multi-megavolt grids. All bathed in hundreds of megawatts of alpha or neutron flux, with electrons wanting to jump on anything not magnetically insulated.

Any workable configuration is going to have to satisfy the physics and the engineering.
I love hard problems.

BTW I was discussing the magnet reqmts with a super conductor guy and he said the field around the coils will be 3-4T to get 1 to 2 T at the coil center.

Which is exactly what we see from Indrek's simulations.

windmill
Posts: 29
Joined: Thu Jul 19, 2007 3:19 pm
Location: Toronto, Canada

Post by windmill »

That new MgB2 superconductor should be great for this app. I noticed the company that makes the wire supplies it in round section and in, I believe, kilometer-long continous lengths. I wonder how much wire we would need for a 4T, 3 meter diameter coil? It's ampere-turns that partially determine field strength, but how do you include the coil diameter in the calculation?

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

windmill wrote:That new MgB2 superconductor should be great for this app. I noticed the company that makes the wire supplies it in round section and in, I believe, kilometer-long continous lengths. I wonder how much wire we would need for a 4T, 3 meter diameter coil? It's ampere-turns that partially determine field strength, but how do you include the coil diameter in the calculation?
IIRC a .3 meter coil producing .45T requires 100,000 A turns. Something like that. A 3 m coil to produce the same field would require 1,000,000 A turns. So for 4.5T it would be 10,000,000 A turns.

The current MgB wire is good for 100 A. The .45T 3 m coil would require 10,000 turns. At 10m per turn (roughly) that is 100,000 m, 100 km.

I did talk to the head of an MgB superconductor company. He seemed willing to work with us if we ever got funded. He has other technologies besides MgB so we will probably get steered in the right direction if we ask the right questions.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

In any case I think I would start with MRI coils - modified and then full custom.

93143
Posts: 1131
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

hanelyp wrote:
93143 wrote:The trouble with a tokamak configuration is that the centre of charge for the electron distribution is outside the confinement area (ie: in the hole in the middle of the donut). ...
Yes the center of charge would be outside the containment volume, but the electric field generated by the electron ring still points towards the electron ring. To visualize, consider the electrons confined along a line in space, with the resultant electric field. The Ions would be confined to a cylinder coaxial with the line. Then take that line and bend it around into a ring. The ion confinement still surrounds the electron ring, though probably a bit distorted.
I'm going on intuition here. It seems to me that Gauss' Law has some relevance.

The idea is that once the ions are closer to the centre of the torus than the bulk of the electrons, the electrons on the far side of the torus start to balance out the force from the nearby ones and it becomes very hard to stop the ions from just drifting into the inner wall.

I could probably prove this, but I'm somewhat preoccupied by my doctoral exam committee meeting tomorrow...

EDIT: Now that I think about it, I think Gauss' Law applies strictly. You wouldn't have any way to prevent ions from hitting the inner wall.

hanelyp
Posts: 2255
Joined: Fri Oct 26, 2007 8:50 pm

Post by hanelyp »

Toroidal configuration:
93143 wrote:EDIT: Now that I think about it, I think Gauss' Law applies strictly. You wouldn't have any way to prevent ions from hitting the inner wall.
Ignoring the effect of the chamber walls, the electric field near the inner wall would be weaker due to the effect of electrons on the other side of the ring, but still pointing towards the near side of the ring. Then add in the effect of a conductive chamber wall allowing positive charge to flow towards the inner wall so that the portion of the ring behind the inner wall is masked.

Roger
Posts: 788
Joined: Fri Jul 06, 2007 2:03 am
Location: Metro NY

Post by Roger »

Do I have this right?

IEC Plasma is self organizing, which is a no no in a Tokamak. Didnt the MIT work show this?
I like the p-B11 resonance peak at 50 KV acceleration. In2 years we'll know.

MSimon
Posts: 14331
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

Roger wrote:Do I have this right?

IEC Plasma is self organizing, which is a no no in a Tokamak. Didnt the MIT work show this?
Yes.

93143
Posts: 1131
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

hanelyp wrote:Toroidal configuration:
93143 wrote:EDIT: Now that I think about it, I think Gauss' Law applies strictly. You wouldn't have any way to prevent ions from hitting the inner wall.
Ignoring the effect of the chamber walls, the electric field near the inner wall would be weaker due to the effect of electrons on the other side of the ring, but still pointing towards the near side of the ring. Then add in the effect of a conductive chamber wall allowing positive charge to flow towards the inner wall so that the portion of the ring behind the inner wall is masked.
First: A positive test charge within a ring of electrons would not experience ANY radial confinement, because the potential on the disc of space within the ring is uniform.

Second: The presence of a non-negligible amount of positive charge would improve matters somewhat, but since the confinement of any one ion depends on there being ions closer to the centre than it, the confinement is very poor.

Third: You're quite right about a metal tokamak shell; that hadn't even occurred to me...

I still think you'd run into extra problems with a toroidal system, even before attempting to extract energy directly from the charged fusion products.

jlumartinez
Posts: 143
Joined: Thu Jun 28, 2007 7:29 pm
Location: Spain

Post by jlumartinez »

It is true that in a donut shape there won´t be a radial confinement (undestanding radial toward the center of the toroidal), but Hanelyp is right in thinking that there will be a confinement toward the center-section of the torus: it is like a potential well but having the negative peak with a circular shape.

Maybe it will not be as efficient in confinement as a standar Polywell central potential well (negative peak = spot ) but maybe it will also create a confinement (negative peak = line) and the final design will be a trade-off between a perfect positive ion confinement in Polywell and a possible new design with lower electron losses (if viable) . At the end the most important thing is to control the electron losses because, after getting huge electron lifetimes, the confinement can be forced just by using bigger potentials.

I like a lot Hanelyp way of thinking. We should do the same : be open-minded and try to search for improved configurations. Who knows if someone can guess a better one ...

Post Reply