Experiments with solid-state magnets

[MSimon]

Let me go even further Joe,

Why would a sharp guy like Dr. B have missed permanent magnets? They would greatly simplify things. Why would you think he didn't consider the idea?

[JoeStrout]

Why would a sharp guy like Dr. B have missed permanent magnets? They would greatly simplify things. Why would you think he didn't consider the idea?

If I ever get a chance to talk with him, I'll ask him.

He actually did use permanent magnets on some early machines, but the geometry was wrong. Once you realize the necessary geometry, it's no longer simple to use permanent magnets — you need compound magnets made of lots of little wedge-shaped pieces. These would probably have to be custom manufactured (or at least, custom ground), to get a high-quality field; off-the-shelf magnets would have gaps that would introduce significant complexities in the field shape. Winding big coils is almost certainly easier, and quite certainly faster, and much easier to model (since they can be perfectly round, more or less).

But there still may be a place for permanent magnets, especially when trying to do steady-state experiments on a smallish machine.

[MSimon]

Why would a sharp guy like Dr. B have missed permanent magnets? They would greatly simplify things. Why would you think he didn't consider the idea?

If I ever get a chance to talk with him, I'll ask him.

He actually did use permanent magnets on some early machines, but the geometry was wrong. Once you realize the necessary geometry, it's no longer simple to use permanent magnets — you need compound magnets made of lots of little wedge-shaped pieces. These would probably have to be custom manufactured (or at least, custom ground), to get a high-quality field; off-the-shelf magnets would have gaps that would introduce significant complexities in the field shape. Winding big coils is almost certainly easier, and quite certainly faster, and much easier to model (since they can be perfectly round, more or less).

But there still may be a place for permanent magnets, especially when trying to do steady-state experiments on a smallish machine.

I would really like to see how you get the magnetic fields to circle the magnets.

Magnetic fields circle current carrying coils. They go through solid magnets. Doesn't matter if it is "fixed" magnetism or induced magnetism.

[drmike]

The point of the mag field is to keep the electrons traped on them. So the magnetic
path has to be completely in free space. With a coil that's trivial. With a PM it's impossible.

Some magnetic path may go through material and that is the loss you want to avoid.

PM's just have too much electron loss, so a coil will always be more efficient.

[pstudier]

PM's just have too much electron loss, so a coil will always be more efficient.

I'm not positive about this. The Van Allen Radiation Belts are confined with a permanent magnet called the Earth. The mirror losses cause the northern lights. Don't know the confinement time or density or temperature. Maybe if we filled it with deuterium, we could make the earth into a sun.

[MSimon]

PM's just have too much electron loss, so a coil will always be more efficient.

I'm not positive about this. The Van Allen Radiation Belts are confined with a permanent magnet called the Earth. The mirror losses cause the northern lights. Don't know the confinement time or density or temperature. Maybe if we filled it with deuterium, we could make the earth into a sun.

You exactly pointed out the flaws - the poles. However, the reactor could be big enough to overcome the losses.

[drmike]

And the geometry: van allen belts are outside the earth, polywell fusion is in the center of the coils. If you can do fusion outside the PM then it'll work. If you want the center to be the fusion point, then it won't.

[Tom Ligon]

How seductive are solid state magnets? Well, there is a well-shielded steel potato-chip can over in one corner of my garage, containing 6 really bitchin' ceramic donut magnets, sufficient to build a WB-1 machine (the little toy with the solid state magnets) but the size of WB-3.

My intent was to possibly build a little desk-top toy that could be carted around to make glows for demonstrations, a successor to my personal Hirsch-Farnsworth machine. Dr. Bussard himself considered the idea of making up one or more for this purpose.

He was quite sure, for many reasons, that WB1 would not be a useful fusion machine. Still, it was SO easy to build, he couldn't resist. It served as a platform for getting the system set up for WB2, a wound copper machine. That allowed testing of cathodes, the high voltage system, basic mounting and insulation, etc. When they first tested it, it did light off a nice glow that was brighter inside the magrid than outside, and that actually surprised and delighted them.

I'd personally like to see a few high school science students build WB1 machines. A number have build Hirsch/Farnsworth fusors, and one of those took second place in the Intel Science Talent Search in 2003. A Polywell, even a flawed one, ought to make a nice science project.

If somebody nearby Northern Virginia wants my magnets for this purpose, let me know. I'm reluctant to ship the things out of consideration for magnetically-sensitive cargo getting near them.

[JoeStrout]

The point of the mag field is to keep the electrons traped on them. So the magnetic
path has to be completely in free space. With a coil that's trivial. With a PM it's impossible.

It's not impossible; it's just rather fiddly to manufacture. Magnetic fields are (like any field) additive. A single conventional PM always had field lines that pass through it, because of the way it's made. But you can combine a large number of wedge-shaped PMs so that when their fields add up, the lines of the resultant field circle them just like in a coil.

I know I've failed to explain this clearly — I did my best attempt a while back, and so far nobody seems to get it. I need to draw some pictures, or better yet, work up some simulations, but I just haven't had the chance.

Note that while this configuration is unusual, and substantially more complex than your typical donut magnets, it's not necessarily all that expensive, because it's made up of a large number of small, identical magnets. Smaller magnets are much cheaper to make than big ones. But you'd still need a magnet manufacturer to produce them for you, so it might be pricey in small volumes.

[JoeStrout]

It is actually worse than that. Dr. B tried the experiment and got burns on the magnets just where theory predicts.

No, Dr. B never tried this experiment. He used conventional solid magnets, where the field lines go through the material just as you described.

But I'm encouraged to see this response from you (even though I'm seeing it a bit late!), because it proves that you haven't understood what I'm proposing. I think your conclusions are absolutely right about the geometry you have in mind — it's just not the geometry I have in mind. And geometry does matter.

[MSimon]

It is actually worse than that. Dr. B tried the experiment and got burns on the magnets just where theory predicts.

No, Dr. B never tried this experiment. He used conventional solid magnets, where the field lines go through the material just as you described.

But I'm encouraged to see this response from you (even though I'm seeing it a bit late!), because it proves that you haven't understood what I'm proposing. I think your conclusions are absolutely right about the geometry you have in mind — it's just not the geometry I have in mind. And geometry does matter.

Yep. Geometry is everything in this work.

I'm familiar with Halbach arrays so maybe your idea can work.

I look forward to the simulations.

For those not familiar with Halbach Arrays here is a good explanation with pictures of the resulting field:

http://www.matchrockets.com/ether/halbach.html

BTW I don't think what you propose has ever been done. If it works you could become a rich man from just a few years of Polywell production.

[MSimon]

Joe,

File a patent before you disclose publicly. Also make sure any one you run it by signs a NDA.

Also MRI machines.

[JoeStrout]

OK, I don't know when I'll get enough time to find and learn a magnetic simulator, or even a 3D CAD program, so in grand tradition, I'm going to make do for now with crude pencil sketches.

Let's start with the magnetic field around a wire carrying a current, as seen in cross-section:


The field lines spiral around the wire, and if you add a bunch of them (from parallel wires all carrying current the same way) together, the resulting combined field spirals around the whole bundle. This is what we're trying to achieve with permanent magnets. But let's start with a single traditional magnet:

This is magnetized with (say) the N face on top and the S face on bottom; field lines pass through the magnet, as we all know.

Now, if we manufacture this as a wedge rather than a regular box or disk, it's not much different:


But the wedge shape lets us stack up several together. (To see how this works, look at the magnetic field around a stack of disc or donut magnets; I haven't sketched this here, but it's easy enough to find on the web.) The resulting field looks something like this (pardon my crude drawing):



Finally, let's continue stacking them all the way around. (They'll be happy to do this since we're sticking N face to S face at every point.) The resulting combined field should look like this:



Voila! The combined field spirals around the assembly, just like in the current-carrying wire. This is still a cross-section, exactly like the wire cross-section above. Imagine it extending outward from the screen and inward into the screen, and you have a cylinder just like a long bundle of wires.

Of course we'd need to then bend this cylinder into a ring, but if we can get a cylindrical section to behave like a section of wire, then bending it into a torus should be no big deal.

Hope these crude sketches help get the idea across — I would welcome help from anybody with a magnetic field simulator to help verify & clarify this idea.

[MSimon]

I think you may be correct.

However, the field outside the magnets is going to be weak because of the closed magnetic circuit.

[pstudier]

This toroidal magnet violates Maxwell's equations. A loop of magnetic field must interlock with a loop of current. In a permanent magnet, the current flows inside the material in loops that enclose the field. For example, in a cylindrical magnet with the poles on the ends, the currents are on the curved surface enclosing the fields lines inside the magnet.