Different polyhedra require different strength magnets

[tombo]

The funny cusps are a result of the LOCAL current geometry.
They reside between conductors with oppositely directed currents where those currents come closest together.
Look at the point at the center of the equatorial segment (center of a set of 4 segments actually). Let's call than 45deg West.
The current flows east above the equator, west on the equator and east below the equator.
This forms 3 triangular coils above the equator and 3 below it around that point.
IN-OUT-IN above the equator
OUT-IN-OUT below the equator
The 2 IN-going coils ABOVE the equator are connected by a narrow neck of IN-going field just ABOVE the center of the equatorial coil quadrant.
The 2 OUT-going coils BELOW the equator are connected by a narrow neck of OUT-going field just BELOW the center of the equatorial coil quadrant.
Those 2 necks are the cusps.
No GLOBAL effects force the cusps to be on the global equator.

I agree, that area is not as clean as I would like to see it, but I think it works.
The biggest issue with this geometry is the uninterrupted series of cusps located along the 0, 90, 180 and 270 longitude lines.
It looks too much like the original 2-opposed-coil "cusp machine".
It may be interrupted at the poles by tweaking the feed line locations. But that may not be good enough.

[Heath_h49008]

What would stop us from constructing a hectogon, or megagon for that matter, polywell?

So long as the exterior magnetic field exists, and it's cusps are pushed closed from internal electron population... does it matter what % of the reactor core radius those fields extend?(inward)

If the contained electrons are what creates the well, attracts and heats the ions, incoming electrons, and the exterior magnets have only one job... maintaining the "walls"... what does it matter?

NOTE: Simplicity, obviously I wouldn't want to wire 100 or 1m coils if it would work with 6 or 12. But would it work?

[MSimon]

So long as the exterior magnetic field exists, and it's cusps are pushed closed from internal electron population... does it matter what % of the reactor core radius those fields extend?

I'm not exactly sure about that. But I can tell you that it would be very helpful from a cooling point of view to keep the hole sizes about 10 gyroradii larger than that of a 6 MeV alpha.

IIRC 6 MeV Alpha w 3T field = about 11 cm.

[KitemanSA]

The funny cusps are a result of the LOCAL current geometry.
They reside between conductors with oppositely directed currents where those currents come closest together.

Actually, this is the definition of the line-like cusp that the WB6 generates in lieu of the funny cusp of the patented design.

Look at the point at the center of the equatorial segment (center of a set of 4 segments actually). Let's call than 45deg West.
The current flows east above the equator, west on the equator and east below the equator.
This forms 3 triangular coils above the equator and 3 below it around that point.
IN-OUT-IN above the equator
OUT-IN-OUT below the equator
The 2 IN-going coils ABOVE the equator are connected by a narrow neck of IN-going field just ABOVE the center of the equatorial coil quadrant.
The 2 OUT-going coils BELOW the equator are connected by a narrow neck of OUT-going field just BELOW the center of the equatorial coil quadrant.
Those 2 necks are the cusps.

If so, and I won't argue the technical validity of your statement, they STILL create a north in/north out pair around the equatorial pipe, which would have a NULL field right through its center and would be the location of electron drain. No?
Maybe not. Maybe the transverse field is sufficient to protect it. Can you tweek my spreadsheet to analyse this?

No GLOBAL effects force the cusps to be on the global equator.

I agree, that area is not as clean as I would like to see it, but I think it works.

Would be nice. Can you analyze?

The biggest issue with this geometry is the uninterrupted series of cusps located along the 0, 90, 180 and 270 longitude lines.
It looks too much like the original 2-opposed-coil "cusp machine".
It may be interrupted at the poles by tweaking the feed line locations. But that may not be good enough.

Let me think about those. I may have a new favorite!

[KitemanSA]

What would stop us from constructing a hectogon, or megagon for that matter, polywell?

Sorry for the snootiness, but -ogons are planar figures and as far as i know, we can't have a planar polywell!
_-ahedrons are 3D figures. The hexahedron will only work well if it is truncated, and the more toward fully truncated (rectified) the unit is, the better iaw Dr.B. By the way, the truncated hexahedron is what we have now. All the WBs to date are truncated hexahedra (cubes)! As to "megahedra", yes in theory it should work. See tombo's figures above. the sole requirement, in theory, is that each vertex has an even number of faces around it.

So long as the exterior magnetic field exists, and it's cusps are pushed closed from internal electron population... does it matter what % of the reactor core radius those fields extend?(inward)

Indeed, Dr. B. seemed to think that better sphericity was desirable, and the more faces, the better the sphericity. However, MSimon makes a good point with the gyro-radii thingee.

If the contained electrons are what creates the well, attracts and heats the ions, incoming electrons, and the exterior magnets have only one job... maintaining the "walls"... what does it matter?

In theory, it doesn't...

NOTE: Simplicity, obviously I wouldn't want to wire 100 or 1m coils if it would work with 6 or 12. But would it work?

It should.

[Heath_h49008]

What would stop us from constructing a hectogon, or megagon for that matter, polywell?

Sorry for the snootiness, but -ogons are planar figures and as far as i know, we can't have a planar polywell!
_-ahedrons are 3D figures. The hexahedron will only work well if it is truncated, and the more toward fully truncated (rectified) the unit is, the better iaw Dr.B. By the way, the truncated hexahedron is what we have now. All the WBs to date are truncated hexahedra (cubes)! As to "megahedra", yes in theory it should work. See tombo's figures above. the sole requirement, in theory, is that each vertex has an even number of faces around it.

My apologies. I can be a bit of a savage when it comes to terminology. (It gets worse after 11:00, )

I understand the need to keep the openings large enough to avoid excessive "squeeze" on the alphas trying to get out. So, lets be generous. 15cm openings as a minimum. I believe I read that 2.5 meters was considered ideal for B11, and 1.75m for DD-DT

That was to handle the distance an electron or ion would be deflected by a deflection impact, or was it for another reason?

Was that in reference to the central core span... the same as each coil in a cube... or individual coils? Because if that was only a requirement of the central core, then I understand why they were investigating other polyhedrons.

[KitemanSA]

The ~2m and ~3m cores were the ~ sizes projected to produce 100MW output with DD and pB&J. The measure is of the sphere that can be fit within the magnets. In WB-6, the toroids were about 30cm wide so the core size was ABOUT 30cm which everybody uses as a matter of convenience.

The prime reason I have been looking for other polyhedra is the question of sphericity. MSimon believes that any lacking of sphericity can be made up by additional size; and for terrestrial use, I concur 100%. But trying to squeeze things into ships, submarines, and spacecraft can make efficiency more of an issue. 10% smaller can be a VERY good thing.

[Heath_h49008]

So, a more spherical form, made up of minimally sized coils would be the most efficient from a size to core diameter standpoint. Spheres are the best way to contain any given volume.

We have our min. coil diameter to avoid Alphas heating them up on the way out... 15cm I said, but MSimon said slightly smaller...

We have our minimum core diameter of 2.5m. Again, dictated by the particles, not by choice. But, it can be as much larger as we choose to build it.

So long as round coils are efficient at containing the electrons, we have an erector-set reactor. Build it as large as you like, out of rings of at least 30cm? (thickness of 7.5cm... yes, I pulled that directly out of thin air.) as large in diameter as the magnets are capable of producing a field to cover.

If the connectors can be attached, and the power/coolant connected through them, the magnets are capable of efficiently producing a large enough fields to cover the area, and the internal space is large enough to contain the reaction... we have functional reactor

I'll do some math and look for a graphics program to make some pretty pictures... but I think we have a simpler question to answer.

Edit: I'm saying, the real question is how many polyhedrons have that even number of coils around each cusp, and how many have the ability to contain the 2.5 meter min core with the min number of coils of about 30cm spaced about 4cm apart? It's a simple question, I just have to do it longhand.

[MSimon]

You are up against superconductors. They do not lend themselves well to small coils due to plumbing rqmts.

I estimate a minimum dia of 15 cm for the coil "tube". 20 cm more likely. And maybe as much as 25 cm. Which says 1 m (hole) SC coils are about as small as you can expect to get given the current state of the art.

[mvanwink5]

KitemanSA re:
"The prime reason I have been looking for other polyhedra is the question of sphericity. MSimon believes that any lacking of sphericity can be made up by additional size; and for terrestrial use, I concur 100%. But trying to squeeze things into ships, submarines, and spacecraft can make efficiency more of an issue. 10% smaller can be a VERY good thing."

Maybe I am missing something, but the larger the coil diameter, the harder it is to achieve the B field. So wouldn't it be easier to minimize magnet diameter, maximize the B field for each magnet, and increase the bore diameter by going to a higher order polyhedra, if needed?

Cheers

[Heath_h49008]

KitemanSA re:
"The prime reason I have been looking for other polyhedra is the question of sphericity. MSimon believes that any lacking of sphericity can be made up by additional size; and for terrestrial use, I concur 100%. But trying to squeeze things into ships, submarines, and spacecraft can make efficiency more of an issue. 10% smaller can be a VERY good thing."

Maybe I am missing something, but the larger the coil diameter, the harder it is to achieve the B field. So wouldn't it be easier to minimize magnet diameter, maximize the B field for each magnet, and increase the bore diameter by going to a higher order polyhedra, if needed?

Cheers

What he said.

[mvanwink5]

Also wouldn't smaller solenoid coils be easier to cool, be mechanically stronger, and have less thermal expansion issues? There would be more supports (each coil with wall supports housing plumbing feeds, and electrical conductors), but the supports wouldn't be stressed as much, it would seem.

From the point of view of parts (design, construction, experience), one standardized coil would be nice too.

If some clever person can figure out how to make these coils, pumb them up, and support them, well, that would be something! Isn't the heat flux mainly on the front side?

All this waiting around, when do we start building?

[KitemanSA]

There is one other issue that I don't know how to evaluate "effectwise". All these coils have to be connected together and provide passage of current and maybe even coolant. All those connections are potentially "nubs" across the line-like cusps that uniform circular coils would absolutely generate. Each nub is unprotected metal in the path of the electrons. This is not good.

EMC2 has been studying the issue and may have a solution, but I have no idea what it might be. I made some suggestions early on but they may have been idiotic. Then again, maybe they have stolen my ideas. If so, more power to them!

[ladajo]

Wall Mounted coils = No nubs between coils

The mounts would be in the shadows of the coils, and the lengths of the mounts would be determined by the <oscillating?> plasma spikes from the cusps.

I, as stated several times, like the idea of wall mounting. And it certainly is also possible with higher order polyhedra, and certainly can be cost effective if each coil assembly is the same...economy of scale(ish).

Making them all true to the sphere would certainly be fun.

[Betruger]

That kind of custom-ish chamber config at a late production design stage would be pretty cool. And in the middle of a space ship, bulkheads etc