Q&A : Major hurdles to overcome for Polywell Reactors

[Orionblade]

I was originally going to post this elsewhere, but the post I was replying to turned out to be off-topic, so I'll toss it in here.

I do understand that closed box reactors won't work because of cusp leakage and magnet impact. I think I have a way of making a workable closed box reactor, but that's not my primary goal - closed or open, the magnetic field provides several free paths that result in no net deceleration of the electrons once they escape the well - including paths that intersect the center of the coil itself, so I came up with the permanent magnet design so I could test things out without niobium and liquid nitrogen. Nb = $$$.

I've got a preliminary design using permanent magnets (just for an electron trap proof-of-concept) that also has some slightly off-axis Helmholtz coils around the large permanent magnets. These serve two purposes. For fuel and electron injection, I can turn the coils on and null most of the permanent magnet field, then once I have the low energy electrons in the trap, I can turn the coils off, and I have intense field compressing the electrons to the center of the trap and a higher potential (this is part of the demonstration aspect - I want to make sure my math works on this one in RL). The second purpose the coils serve is that I can pulse them at a very low voltage/current and wiggle the field of the magnet they're wound around - this is why they're off-axis slightly. If I control the phase of the wiggle properly, then the field will appear to rotate from the frame of reference of an electron escaping along a free flux path (cusp) and the cusp will then wiggle, thus turning the electron towards higher and higher field strength, and away from the cusp.

Further, I can increase the current to the coils and heat the plasma by wiggling the field and inducing plasma currents.

I think this would work with even a larger non-closed-box reactor, with the possible exception of the permanent magnets (kind of hard to get a ring magnet with field in the middle of it... The neat part of this whole thing is that you can charge up a superconducting magnet and let her rip, and effectively turn it off or "down" with the Helmholtz coil - pretty handy for making changes or refueling without quenching the magnet and making some really hot really expensive paperweights.

Any input?

[Orionblade]

I doubt the frog felt anything. The effect happens because all materials are diamagnetic - the electrons all tend to line up against the imposed field. So that means the water and the frog in it were all evenly lifted. I wouldn't mind the experiment of being floated in a 20 T field, but that would be one hell of a magnet!!

The forces are reasonably uniform at the atomic level - there is no way the nervous system would even notice.

There are several experiments with even rather low intensity magnetic fields that are pertinent not just to this side-topic, but the safety of working in the vicinity of high magnetic fields.

First, pulsed magnetic fields around 20Hz can induce flashing lights, not in your eyes, but via inductive coupling to the optic nerve.

Second, Even MRI level fields can cause disorientation if you turn your head quickly - the blood in your head is a conductor, as well as the fluid in your inner ear - when you move it in a magnetic field the diamagnetic properties and the conductive properties add up to produce oddball swirling movements. Move fast enough in a strong enough field and you can pass out due to changes in pressure and the like. I'd doubt you could rupture a blood vessel, but it's something to think about when you're getting dizzy and working with high voltage anything. *woozy* *Thwoomp* *KZAKK!*.... "Hey, where'd fred go?" : "Prolly went to get some paper towels to clean up all this carbon residue..."

Also, back in high school, I did an experiment where I supplemented various plants with iron-containing fertilizer - from cacti to broccoli. I then placed ceramic magnets on the stems and leaves of various test subjects. The plants were fine up to the point where the magnet rested. The upper part of the plant wound up running out of iron, couldn't bind oxygen, and died. The christmas cactus I used was a bit different - it formed concentric rings of very dark green in the leaf/stem around the magnet, but because the plant is flat and wide, it was able to pump enough nutrients past the quiescent field to keep the upper bits ticking. I only ran the experiment for a few weeks, so I didn't get any decent growth-rate variation, but the effect was pronounced even in the non-supplemented plants. We've got a ton more iron in our bodies (around a half a gram) than a plant does - think hemoglobin, myoglobin, etc. - so there's the whole paramagnetic effect of the iron, and the diamagnetic effect of all that water, as well as the MHD effects of the salt water we call blood.

I'm willing to bet you'd be calling ralph on the big white phone by the time you were done in that magnetic field, but I'd be willing to give it a shot as well. Know a good way of ballistically removing fillings from teeth?

Rion

[MSimon]

Orion,

Permanent magnets have the annoying feature of zero field gradient facing the poles. i.e. the poles would be equivalent to a point cusp. the magnet will get hot.

[Orionblade]

I think that's what I'm addressing with the field wiggle.

Isn't it pretty much the same problem with a coil?

There's a line (albeit a single point, rather than a large part of the face of a permanent magnet) coaxial to the coil that permits a charged particle to escape along a free path straight out of the reaction chamber. Open (coil) magnets let you recirculate the ions that would otherwise smack into the magnet itself and contribute to heating, but my thoughts were to wiggle the field with even a conventional BFR Magrid, so that this mean free path wiggled with a period shorter than the average time of flight of an electron attempting to escape along it, thus preventing any electrons from escaping along the path, since they would experience a non-parallel field vector as it "wiggled" around them.

My second question for this thread, is why must it be a spherically symmetrical trap? If we're just trapping electrons, does it really matter what shape the cloud is, as long as there's a single point that represents the maximum electrostatic potential? I had thoughts on how one might build a tall and skinny reactor, or a wide and short one, but then the ions would have a vastly different mean path than the electrons - any thoughts on that?

This is just a rambling as I drove home a few minutes ago with, appropriately, Juno Reactor thumping on the stereo, so don't think I've even come close to applying any real theory to it, just wondering if it's worth doing math.

*turns up the volume*

And thanks Msimon - I hadn't quite thought about the field geometry being that different, I was thinking the edges of the magnet would be getting hot, but now I presume it'll be the very center of the magnet that gets smacked the hardest.

Orion

[MSimon]

There's a line (albeit a single point, rather than a large part of the face of a permanent magnet) coaxial to the coil that permits a charged particle to escape along a free path straight out of the reaction chamber. Open (coil) magnets let you recirculate the ions that would otherwise smack into the magnet itself and contribute to heating, but my thoughts were to wiggle the field with even a conventional BFR Magrid, so that this mean free path wiggled with a period shorter than the average time of flight of an electron attempting to escape along it, thus preventing any electrons from escaping along the path, since they would experience a non-parallel field vector as it "wiggled" around them.

You are up in the GHZ range to make that happen. If it would work at all. And then you are dumping power into the machine that will tend to heat the electrons so it will be disruptive.

And don't forget that even in experimental machines the coils (or magnets in your case) are about .3 m across. Which is about 1 wavelength at 1 GHz. You are not going to get all the points of the magnet at the same potential.

Current carrying coils is the only way to go.

[Orionblade]

My original concept was a 1 cubic inch "reactor". I guess what you're saying is that this isn't a scalable technique?

Also, keeping in mind that the required input power for the wiggle wouldn't be but a few watts - you only need to move the field a nanometer to get an electron to start turning - there's no way to locate the secondary feild coils (helmholtz) outside the plasma chamber and make them smaller for adequate frequency response?

I had also thought about using these coils to tap power back out of the plasma, so that the plasma would oscillate, and those oscillations would be essentially an amplification of input power via fusion heating of the plasma.

I think I'm going to be learning alot more than I anticipated by building this thing.