[DUMB??] Design Question: Concentric polyhedral magnets

Discuss the technical details of an "open source" community-driven design of a polywell reactor.

Moderators: tonybarry, MSimon

laksindiaforfusion
Posts: 22
Joined: Sat Feb 16, 2008 9:48 pm
Contact:

[DUMB??] Design Question: Concentric polyhedral magnets

Post by laksindiaforfusion »

One of the main issues that the magnets would face is going to be structural artifacts due to high energy alpha collision. Which would mean that the magnet has a limited lifetime before the artifacts change the chemical and electrical properties of the magnet.


here is a dumb suggestion:

assuming that all the fusion reactions are happening in a small sphere at the center of the machine, the alpha particles that travel outwards would go hit the outer sphere and thereby generate power. However they might also hit the magnets and get STOPPED(this is a big assumption again) leaving a space that should be alpha free.

Lets call this fusion-shadow region. This shadow region would look exactly like a shadow of the polyhedral magnetic config when a spherical light source is kept in at the center of the machine.

Now lets assume that we have two polyhedral magnets, a smaller one which is made from cheap easily replaceable parts and a very powerful and sophisticated larger one designed so that it can be concentric to the smaller one and lie in its fusion-shadow. However we can make sure that the magnetic fields add up of both the coils causing no change in the potential well or the beta=1 condition. But at the same time increasing the life of the poweful magnet. The smaller one can have just enough magnetic field to support the large one, with a lot more cooling mechanisms, power extraction etc etc.

Would this be feasible?
The believer's burden and a skeptics purpose

scareduck
Posts: 552
Joined: Wed Oct 17, 2007 5:03 am

Post by scareduck »

laksindiaforfusion wrote:assuming that all the fusion reactions are happening in a small sphere at the center of the machine, the alpha particles that travel outwards would go hit the outer sphere and thereby generate power. However they might also hit the magnets and get STOPPED(this is a big assumption again) leaving a space that should be alpha free.
-ish. I imagine there'll be plenty of diffusion past the MaGrid, partly because the MaGrid will bend the trajectories of ions (slightly), both magnetically and electrostatically. But, yes, it's fair to assume that alphas will hit the magrid (possibly damaging the crystalline structure of whatever it is they're impacting).
Now lets assume that we have two polyhedral magnets, a smaller one which is made from cheap easily replaceable parts and a very powerful and sophisticated larger one designed so that it can be concentric to the smaller one and lie in its fusion-shadow. However we can make sure that the magnetic fields add up of both the coils causing no change in the potential well or the beta=1 condition. But at the same time increasing the life of the poweful magnet. The smaller one can have just enough magnetic field to support the large one, with a lot more cooling mechanisms, power extraction etc etc.
1) The inner magnets would have to be held up with something, and would have to be connected to the rest of the magnets with something, too. Bussard wanted interconnects between individual magnets to be a short as possible. Another set of magrids with interconnects just gives the electrons something else to hit.
2) More magnets shortens the Paschen arcing distance between the casings, a major loss mechanism.
3) Similar to (1), there's also cooling issues to be dealt with.

It's not clear to me that this would be any better than Bussard's design, and it might be appreciably worse.

laksindiaforfusion
Posts: 22
Joined: Sat Feb 16, 2008 9:48 pm
Contact:

Post by laksindiaforfusion »

1) The inner magnets would have to be held up with something, and would have to be connected to the rest of the magnets with something, too. Bussard wanted interconnects between individual magnets to be a short as possible. Another set of magrids with interconnects just gives the electrons something else to hit.
Not if they are set up in such a way that the magnetic field dont go INTO any one of the magnets. I assume this must be possible. Ofcourse, electrons can strike anything, but if we make the fields pass THRU one of the coils then they will strike it for SURE. Now whether they strike it or not depends on the distribution of electron energy, which i believe varies widely from inside the machine to outside the machine.

2) More magnets shortens the Paschen arcing distance between the casings, a major loss mechanism.

3) Similar to (1), there's also cooling issues to be dealt with.
If 2 is true then it can be used as a design constraint i.e. distance between the magrid to avoid arcing vs ion/electron density outside the machine vs fusion rate. (It is an engineering optimization problem that way) .

The advantage that I see here is whether we can have a larger magnet that need not worry about alpha bombardment and hence can have minimal cooling systems on that one (which takes care of point 3, because in case of superconducting magnets, heat comes from fusion in forms of alpha or neutrons, both of which are eaten away in the smaller magrid) and can give the larger magrid a longer lifetime. The inner magnet, has to have better protection and should be designed to be replaced easily.
The believer's burden and a skeptics purpose

MSimon
Posts: 14332
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

laksindiaforfusion wrote: I assume this must be possible.
Could you provide a diagram or sketch?
Engineering is the art of making what you want from what you can get at a profit.

laksindiaforfusion
Posts: 22
Joined: Sat Feb 16, 2008 9:48 pm
Contact:

Post by laksindiaforfusion »

Could you provide a diagram or sketch?
Sure, I have a small sketch in my office. I shall scan it and send it accross tommorow when i get there.[/quote]
The believer's burden and a skeptics purpose

scareduck
Posts: 552
Joined: Wed Oct 17, 2007 5:03 am

Post by scareduck »

laksindiaforfusion wrote:
1) The inner magnets would have to be held up with something, and would have to be connected to the rest of the magnets with something, too. Bussard wanted interconnects between individual magnets to be a short as possible. Another set of magrids with interconnects just gives the electrons something else to hit.
Not if they are set up in such a way that the magnetic field dont go INTO any one of the magnets.
No! There will still need to be electrical connectivity, at least, between the inner magnets and the outside. This still increases the places where electrons can intercept the grid.
If 2 is true then it can be used as a design constraint i.e. distance between the magrid to avoid arcing vs ion/electron density outside the machine vs fusion rate. (It is an engineering optimization problem that way) .

The advantage that I see here is whether we can have a larger magnet that need not worry about alpha bombardment and hence can have minimal cooling systems on that one
Minimal cooling? You mean, as in having no superconducting magnets? In which case, why not just reinforce the center-facing portions of the original MaGrid? If you're talking about putting water-cooled copper in the center magnets, that's a huge energy loss.

hanelyp
Posts: 2257
Joined: Fri Oct 26, 2007 8:50 pm

Post by hanelyp »

Sketching up what's described, I see a single set of field lines around both magnets, centered closed to the stronger outer magnet. If I'm seeing laksindiaforfusion's idea correctly, we might do as well with a single magnet of oblong cross section, the cross section major axis tilted towards the core, presenting the minor axis for alpha intercept. A bit more complex fabrication, but it gives a larger magnet cross section for coolant flow without more alpha intercept. Would require a stronger field to shield the magnet from electrons, but we may need that anyway for the proper wiffleball.

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

Post by Roger »

Hopefully not too dumb of a question/s.

Would the resulting MGrid have more nooks and crannies in it ? Would the cusp losses be changed ? For the better ? Does the magnetic field change, and does that change the potential well ? For the better ?

If the Whiffle ball has smaller holes, and leaks less, that might be interesting... On the other hand a more diffuse potential well might suck.
I like the p-B11 resonance peak at 50 KV acceleration. In2 years we'll know.

MSimon
Posts: 14332
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

hanelyp wrote:Sketching up what's described, I see a single set of field lines around both magnets, centered closed to the stronger outer magnet. If I'm seeing laksindiaforfusion's idea correctly, we might do as well with a single magnet of oblong cross section, the cross section major axis tilted towards the core, presenting the minor axis for alpha intercept. A bit more complex fabrication, but it gives a larger magnet cross section for coolant flow without more alpha intercept. Would require a stronger field to shield the magnet from electrons, but we may need that anyway for the proper wiffleball.
There is a limit here (oblong cross section) because you have a limitation of a minimum radius (electrostatic field strength).

In any case I do not see cooling for superconducting WB-100 coils as a serious problem. What is desired is definitely pushing the limits. It is not state of the art. It is not beyond state of the art into TBA territory.

The first working reactor is going to be done with the absolute minimum number of optimizations. I'd use standard (modified) MRI coils. I think they run about 1 m dia. They should get us close to break-even. While we are waiting for custom coils to be designed and built.
Engineering is the art of making what you want from what you can get at a profit.

laksindiaforfusion
Posts: 22
Joined: Sat Feb 16, 2008 9:48 pm
Contact:

Explanation attempt Diagrams

Post by laksindiaforfusion »

First bear with the crude drawings (need to dig out some CAD software soon):
Image

This image shows one cross section of the magrid, in only one quadrant. The virtual anode is the center of the diagram, where the fusion takes place and causes an alpha shadow; which is labeled by region A in the above diagram.

Also shown are two different electron trajectories, one is an electron being captured into the wiffleball and the other an electron with an absurdly high energy electron escaping the well, getting caught in the magnetic field lines and then gets recirculated back into the WB.

This causes a small region around the magrid magnets which have very very low high energy electron density(Region B), however there can be low energy electron(for example a low energy electron moving in a direction perpendicular to the magnetic field can go hit the magrid.) However since the e-gun is placed along the cusp lines there is a large chance that there are no high energy electrons in this region.

Therefore if we keep a second magrid in the intersection of region A and Region B, we can in effect increase the strength of the magnetic field OR, this is what i initially thought it up for, we can make the inner magrid to be made of cheap easily replaceable superconductor, and the outer one to be of high quality superconductor, with lesser protection from Alpha bombardment, thereby giving the largest return on investment.

This is shown in the following figure:
Image

It can be used as a strength boost or a cost mitigating mechanism.

As far as the support structure goes this diagram helps:
Image

Shows that the supporting ceramic struts are aligned in such a way that they lie in the low electron density region, thereby protecting them.

This is one change i want to see on Bussard's design. instead of vertical ceramic supports, make them aligned to the magnetic field so that there is minimal electron damage.
The believer's burden and a skeptics purpose

jtolson
Posts: 3
Joined: Fri Feb 01, 2008 12:22 am
Location: VA

CAD

Post by jtolson »

quote]First bear with the crude drawings (need to dig out some CAD
software soon):
one free and simple CAD program is free from Google
Google SketchUp 6 is a 3D modeling software tool that’s easy to learn, simple to use I'm learning to use this program. 40 years ago I took drafting in school with the pen and paper method

Mikos
Posts: 27
Joined: Wed Jan 16, 2008 3:19 pm
Location: Prague, Czech Republic

Blender

Post by Mikos »

Blender is the way to go for 3D modeling! It is the best 3D modeling software I know, in my opinion much better than 3D Studio MAX (or similar expensive proprietary software) and it has really well-thoughtout user interface. And best of all, it is opensource!

If you are looking for simple 2D CAD software (like AutoCAD LT), then QCad is really great (and it is also opensource).

And if you want really complex and powerful 3D CAD software, you can use BRL-CAD. It is powerful CSG (Constructive Solid Geometry) modeling system. It was developed and used in production by the U.S. military for more than 20 years and now it is opensource. And it is not only used for engineering, but one of primary purposes is also support of ballistic and electromagnetic analyses. I think it would be great CAD for Polywell. There is also really great documentation.
"Those who would give up Essential Liberty to purchase a little Temporary Safety, deserve neither Liberty nor Safety."
-- Benjamin Franklin

laksindiaforfusion
Posts: 22
Joined: Sat Feb 16, 2008 9:48 pm
Contact:

Post by laksindiaforfusion »

I have used belnder before, but it takes a LOT of work to learn and a lot more to master.

Its actually used more by the game dev guys and semiprofessional animators. I guess it should be good enough for a lot of CAD at the amateur level.

BRL-CAD sounds tooo hi-fi for a beginner :)

Yeah.. in the end we all need to take the small step of starting to use any one of them.
The believer's burden and a skeptics purpose

Mikos
Posts: 27
Joined: Wed Jan 16, 2008 3:19 pm
Location: Prague, Czech Republic

Post by Mikos »

laksindiaforfusion wrote:I have used belnder before, but it takes a LOT of work to learn and a lot more to master.
I have used it before too, it took me less than week to learn it and do what I needed (3D model and simple animation), there are great tutorials for Blender. Blender UI is really stunning (you can't learn it without tutorial or documentation, but after you learn it, every other 3D modeling application seems to be horribly cumbersome in comparison with Blender).
"Those who would give up Essential Liberty to purchase a little Temporary Safety, deserve neither Liberty nor Safety."
-- Benjamin Franklin

laksindiaforfusion
Posts: 22
Joined: Sat Feb 16, 2008 9:48 pm
Contact:

Post by laksindiaforfusion »

(you can't learn it without tutorial or documentation, but after you learn it, every other 3D modeling application seems to be horribly cumbersome in comparison with Blender).
Thats the whole issue :)

Personal reminder: Patience is a virtue :)
The believer's burden and a skeptics purpose

Post Reply