magrid configuration brainstorming

[happyjack27]

So, put all your ideas together. Suspend a micro-grid mesh inside a cryo-vacuum chamber and induce current in the mesh. Need to maintain the orientation of the grid so the alphas (we hope it's b-B11) come out aimed at the venetion blinds.

Could get fancy and levitate the micro-grid above magnets in the chamber walls. Orientation controlled by extensions from the grid to the chamber walls, etc.

you'd need electromagnetics to induce the fields in the first place. the electromagnetics that induce the currents could also provide for levitation.

but at that point, why don't you just provide power to some of the superconducting wires, thus increasing the net magnetic field strength felt by the ions? then you wouldn't need to levitate anything cause you can use the power lines as support structures.

point is, while all these ideas are cool and fun to think about, i'm not really sure they ultimately provide any net power gain. and failing that, there's really no reason to implement them on a commercial machine.

[rjaypeters]

There you go being all practical in the middle of a brainstorming thread.

[rjaypeters]

When I think about systems that self-organize, I wonder if they are sensitive to initial conditions which then makes we wonder if they are chaotic i.e. unsteady.

[happyjack27]

When I think about systems that self-organize, I wonder if they are sensitive to initial conditions which then makes we wonder if they are chaotic i.e. unsteady.

eh, ALL systems self-organize. that simply means their structure and dynamics emerges from local interactions. Which is true of all things, whether they are generally considered "self organizing systems" or not. lasers, the freezing of water, ecosystems, living tissue, dead tissue, the electrons in an electrical circuit, flocks of geese, socieities, galaxies.... the list goes on. only we have simpler descriptions for some systems - special cases, so to speak, in which certain feedback relations makes it so they asymptotically approach structures and dynamics that can be explained with much simpler equations.

anyways point is there's no reasons a self organizing systems should be chaotic or unstable. in fact, many things that are classically considered "self-organizing systems", are far more stable than more "classical" / "Newtonian" systems precisely _because_ they are "self-organizing".

take for example tcp/ip traffic across the internet. the ability of routers to adapt to local changes in their environment without global communication makes it much more robust against network outages. (in fact, that was the original design goal of the internet when it was still known as "arpa-net" (arpa standing for Advanced Research Projects Agency). that kind of resilency ("stability") simply wouldn't be possible in a classical newtonian system.

[rjaypeters]

In the case of a superconducting mesh for a fusion reactor, I don't think we want a chaotic system.

I'm well aware chaotic systems are often more robust than more rigidly controlled systems. The best example are heart beats. People with heart beats that chaotically vary from a metronome beat live longer, at least not die from heart attack, than those whose heart beats are close to metronomic regularity. Nobody knows why.

In the case of the superconducting mesh, we will probably never know.

[rjaypeters]

Twisted octahedron single-coil, 'cause it's pretty:

[ltgbrown]

Something I have been trying to wrap my head around is how to support any of these structures, even the simplest current magrid design. I am assuming at least two supports per coil. So, do you want magnetic fields around the supports to prevent recirculating electrons (or ions!) from impacting them? If so, how do you run the wires (or tubes) carrying the current to create the magnetic field through the supports without creating unequal magnetic fields around the magrid? Current flowing down a support to the magrid would create a magnetic field, but if the current returns back up the same support, wouldn't that negate (or at a minimum severly distort) the magnetic field? If so, then wouldn't all current flowing to the magrid have to be in a different support than the one the current flowing from the magrid is in? Wouldn't that create portions of a magrid face with one more turn of wire with current flowing through it and therefore a slightly stronger field?

Perhaps these affects are insignificant and only affect the final efficiency and not the ability to actually achieve fusion and eventually (we hope) net power at a reasonably sized reactor.

[rjaypeters]

Something I have been trying to wrap my head around is how to support any of these structures, even the simplest current magrid design. I am assuming at least two supports per coil.

Generally, yes, at least two supports.

So, do you want magnetic fields around the supports to prevent recirculating electrons (or ions!) from impacting them?

I think this is good idea.

If so, how do you run the wires (or tubes) carrying the current to create the magnetic field through the supports without creating unequal magnetic fields around the magrid?

One doesn't.

Current flowing down a support to the magrid would create a magnetic field, but if the current returns back up the same support, wouldn't that negate (or at a minimum severly distort) the magnetic field?

Indeed.

If so, then wouldn't all current flowing to the magrid have to be in a different support than the one the current flowing from the magrid is in?

Which is why I created the octahedron with twenty-four support arms:


the dodecahedron with one hundred twenty supports:


the cube:


and last, for fun, the dual hexagonal pyramid with thirty-six supports:


All of these configurations have four conductors parallel to each other running from the "rings" toward the chamber walls. Some have objected to this arrangement on the grounds of magnetic field considerations.

Wouldn't that create portions of a magrid face with one more turn of wire with current flowing through it and therefore a slightly stronger field?

I don't think so, but I haven't thought about it carefully.

Perhaps these affects are insignificant and only affect the final efficiency and not the ability to actually achieve fusion and eventually (we hope) net power at a reasonably sized reactor.

Now, we* don't know all the significant factors for polywell success. Later, we hope to know what is significant.

*The publicly responding members of this forum. Patience, as a virtue, is its own reward. But I want it now!

[KitemanSA]

this has all the bennies of those other MaGrids and doesn't need the forest of supports.

[rjaypeters]

I see six supports on one side of this geometry, there might be two I don't see. How will the SC cable be cooled on the opposite side from the supports? High pressure and velocity flows?

Edit: Don't get me wrong, I _like_ this design in part because one must provide only a strong baseplate in addition to the vacuum chamber.

[KitemanSA]

Please investigate the machine called the MPG. It was a single pass version of what you see here. In the MPG, one copper tube was bent into the entire MaGrid. The cooling fluid made a single pass thru the entire unit, carried directly by the tube. The problem is that the points of close approach need to be attached or they will splay outwards under the magnetic forces. In the original MPG, sodered (welded?) the close approaches together which put metal in one of the primary electron recirculation paths.

It is simple to split that tube into two loops for fluid flow (shortening the passage) while keeping the electrical circuit the entire length. Indeed, it is simple to make a multi-wind circuit with only two coolant loops. But the attachment issue remains. Now, picture that multi-circuit, dual loop overlaid with three more, each rotated 90 degrees about the vertical axis. Then, encase the set. Each successive layer provides the structural continuity to the others while the coolant loops remain short, unbranching, direct.

Easy.

[rjaypeters]

Please investigate the machine called the MPG. It was a single pass version of what you see here.

Thanks, I will.

In the original MPG, sodered (welded?) the close approaches together which put metal in one of the primary electron recirculation paths.

AIEEEE!

[rjaypeters]

KitemanSA, Thanks for bringing my attention back to the cuboctohedron. It makes a really nice heavily-forested-with-extensions single-coil configuration:

Much better, I think, than the dual hexagonal pyramid from a few days ago.

[rjaypeters]

Thinking about the cuboctahedron KitemanSA posted above and the "forest" comment, I decided on a compromise position. If not otherwise required, I don't like single-side supports for heavy masses or highly-stressed components, so I investigated a "columnar" concept for the cuboctahedron. I came up with two basic configurations: not-twisted and twisted:



The amount of twist refers to the longitudinal distance the pipe travels around on the way to the other pole. In the non-twisted version, the pipes return to the same longitude at the other pole. In the twisted version, the pipes wrap around to 180 relative longitude by the time they reach the other pole.

For structural reasons, I like the twisted version better. If I could have, I'd have arranged for a 360 degree excursion, but the cuboctohedron doesn't have enough segments, but other shapes might...

Anyway, I have previously done a columnar concept for the Tombo'd octahedrons before I got on the single-coil kick, so I won't be revisiting those.

EDIT: And Tombo did at least one "columnar" awhile back. I'd have remembered it earlier, but that's what you get when you work while you should be sleeping...

[rjaypeters]

A rhombicosidodecahedron in the columnar configuration with four pipes (which could be a single conductor if desired):



The single gray pipe that does twist all of the way around the sphere while traveling to the south pole:

Now if we could just find a way to hold the close approaches together without messing up recirculation or contaminating the vacuum.

Note there are two varieties of pipe configurations: gray and purple are the same (but rotated), green and orange are the same.

A more polar view: