And so it begins

The system is a rebuilt Leybold TOPS 151.  It's a very nicely built, compact turbo pumping system with a pretty large capacity.  Oil free.  :)  I opted for an upgrade to the 361 turbo pump.  It's rated at 345 liters/second, with ceramic bearings.  Air cooled.  Yum.  I'm pretty sure that the backing pump is undersized for this  for the system, but I'm not complaining.  10^-6 after 3 hour pump down on a non baked system - one that's had the aluminum chamber anodized just a day before, mind you.  And one of the ports of the bell jar was covered with a fairly ratty piece of Plexiglas to boot.  Finally, no vacuum grease on the bottom of the bell jar seal.  I figure if I take reasonably care I'll be able to be in serious vacuum pretty quickly.  The original specs for the TOPS 361 claim a pump down time of 3 minutes for a 25 liter chamber to 10^-5. Bitchin'.

The bell jar is truly huge, with a nice big rectangular Plexiglas port down the front.  It's like a zillion cubic liters, but it's truly lovely to behold.  Tomorrow I have to stop by Tap Plastics pick up an inch thick Plexiglas disk for the open port on the bell.  Larry (the fine gentleman who put this system together) gave me a steel blank I can face and drill, but I'm much to impatient for that.  Besides, if this system got to 10^-6 with a ratty piece of plastic, it'll be handy to have a nice big access hole I can just cover with a piece of plastic.

Mental note: install a ceiling hoist so the bell jar is easy to lift.

I also got the magnets from allmagnets.com.  These are big honkers - 4.5" OD.  Grade 5 ceramic, 3800 Gauss.  Tasty.  My poor wife didn't realize what they were and accepted the package and luckily left it downstairs because it was fairly heavy.  Heavy means it belongs in the workshop.  So luckily these 20 magnets didn't end up anywhere near sensitive magnetic hard drives.  Mental note.  Do not bring extremely powerful magnets anywhere near electronic equipment.

These magnets are very cool.  Magnetism has always been one of the most visceral of the fundamental forces for me.  You can definitely feel the fluidic nature of the force as you play with magnets.  20 of these suckers in serial is pretty powerful and you can really feel the fields interplay.  I can really see the allure of magnetohydrodynamics.

The system I'm going to play with is ridiculously simple.  Some amateur fusion experimenters have had some truly spectacular successes with Inertial Electrostatic Confinement (IEC) using spherical octagon grids.  These systems are incredibly cool and surprisingly low tech.  Playing with 25 KeV is non trivial, though and I'm constantly reminding myself of the dangers of high voltage and other nasty stuff.  But I think they have reached the limits of electrostatic confinement.  I've gotten a hold of Bussard's patents, and have been generally digging into current thinking in plasma physics.

In patent number 4,826,646, Bussard describes an elegant device for overcoming the limitations in devices based on the Fansworth, Hirsch and Meeks design class.  The basic problem with using grids is that you can't make the grid cross sectional area small enough.  At least that's the way it seems.  I'm not really sure because Bussard's other patent, number 5,160,695 has a number of grid based confinement systems diagramed.  I particularly enjoyed the spherical tetrahedral grid.  Very pleasing to my Synergetic sensibilities.

So anyways, Bussard's very clever idea is to create a magnetic bottle for containing the electrons.  Unlike other controlled fusion devices which use magnetic fields to contain the plasma, IEC uses electrostatic forces to create the ion densities necessary for fusion.  In Bussard's design, a polyhedral magnetic bottle is constructed to constrain electrons at the center of the sphere.  These electrons are injected via the faces of the polyhedron into the center.  Eventually you can build up quite a charge as these polyhedral magnetic bubbles are quite efficient at keeping the electrons you inject inside the system, squeezed down in the center of the device. 

Arranging the magnets strategically in a polyhedral array has the effect of creating a magnetic trap that only has point cusps (figure 1c on the diagram to the left).  This means that only electrons traveling in trajectory within a narrow solid angle can escape the system.  If you balance the electron's energy with the strength of the magnetic field you have very low losses within the system.

The very first thing I want to do to shake out the system is to simply inject some low energy electrons into the system - say 50 electron volts. Keep the pressure just below 10^-2 and see if I can get a glow discharge from the residual air.  I plan on getting a small bottle of neon or argon so I can get a really nice glow out of the system to make some pretty pictures.  I can use this to shake out the gas back filling subsystem I'll eventually need to put together.

I figure I can do a lot of experimentation with fairly low voltage and shake out my lab procedures and safety protocols.  I remember working with high voltage in my youth.  I vividly remember the smell of my burning skin when I caught an arc from a 12 kV TV fly back circuit on the back of my right hand.  Luckily I was obeying grounding procedure and I just ended up with a really nasty surface burn.  No scarring.  It's been a while since I've done such nutty stuff so sticking with the low end of the voltage and current scale for a while is only prudent.

I went to Triangle Machine surplus in San Jose yesterday and saw some huge high current power supplies.  The power output wasn't a simple screw terminal.  It was a big, thick plated bar. 8" x 0.5" x 1.0" - massive.  Truly terrifying to think about 200 amps being pumped out of these surplus beasts.  Even scarier was looking at the 30 kV, high current power supply they had sitting out in the yard.  This thing had meters up to 600 milliamps at this voltage.

Truly terrifying to behold.

So anyways, the plan is to now start building the magnetic trap.  The design on the right is pretty simple.  Underlying the system is a cage made of 0.25" square steel.  I chose the dodecahedron specifically because of all the polyhedra Bussard mentioned, this is the only one with five axis of symmetry.  I'll be saying why I think this is significant later, but let's just say that from my Synergetics studies, one thing that has been driven home is the importance of 5 fold symmetry

The 12 pentagons are also nearly circular, which means that I can approximate the system with circular ring magnets very nicely.  I don't want to mess with the currents necessary to produce the field strength to be interesting.  Permanent magnets are simple and relatively safe.  One major downside however, is that the ring magnets I'm using have a pretty large cross section.  They take up a pretty large percentage of the surface area of the containment sphere.  The upshot of this is that any fusion products are going to smack these surfaces around with some pretty high energy.  I'm not really sure how these ceramic magnets are going to hold up under this maelstrom.  They have an operating temperature upper bound of about 300 degrees Celsius.  Not spectacular.  But not shabby either.  Again, I figure I can still do a lot of useful experiments and check out some pet theories without actually doing actual fusion and blasting these magnets with 4 megavolt particle storms.  But then again, these magnets are cheap and can easily be replaced.

Anyways, I'll cross that bridge when I come to it.

One of the things I was concerned about in choosing the dodecahedron as the polyhedron was what the magnetic symmetry would work out.  I naturally have built a model out of the incredibly useful Zome modeling pieces, and found that there are several interesting configurations to explore.  One particularly interesting configuration doesn't use oppositely polarized faces.  It's an elegantly polarized system, but may turn out to be completely useless.  I'm pretty new at this plasma physics stuff and I'll just have to find out. 

Part of what I want to experiment with is my notions about using a magnetic confinement system based on the disdyakis triacontahedron.  This is a fascinating structure.  There are 120 triangular faces, divided into 60 positive spherical triangles and 60 negative spherical triangles.  Here I'm using the words positive and negative in the sense that a left glove is the "negative" of a right handed glove - i.e. left/right symmetry.  Anyways, this is a highly symmetric grid structure.

My first idea, when I read about IEC fusion was that I could use a grid based on this solid as the anode.  My feeling was that this was the most symmetrical grid I could imagine.  But the more I found in my digging into IEC, the more I realized that this grid is not very transparent.  I created an AutoCAD Inventor model of a grid and created a 1.5" prototype.  I uploaded the STL file to some wonderful people at SharedReplicators.com.  Five days later a FedEx box showed up on my doorstep with a plastic model of my hair brained scheme.

It was a truly surreal experience.

So about the time I got the prototype of the disdaikis triacontahedron grid back from the surreal web replicators, I found Bussard's patents and had started reading them.  I had suspected from some writings by the fusion researchers at the University of Wisconsin  that using a grid was not going to work in the long run.  They had mentioned something they referred to as Polywell^tm, which turned out to be what they use to refer to Bussard's ingenious polyhedral magnetic trap.  All of their reports and investigations seemed to point to the conclusion that the inner grid in an IEC device had to disappear in order for the system not to leak power like a sieve.

But I still can't get this beautiful solid out of my thinking.  My theory is that one can use the 15 great circles that form the basis of the disdaikis triacontahedron as a system of magnetic confinement of electrons.  In Bussard's designs, the faces of the polyhedra are magnets, the current being carried in the edges of the polyhedra.  However, Bussard doesn't ever mention using great circle transits as the magnetic fields, rather he describes systems where the magnetic circuits are not formed by great circles of a sphere.  My theory is that the 15 great circles of the icosahedron can trap the electrons in the same manner as a balloon, by the way charged particles would bounce around inside as system composed of these fifteen magnetic fields.

Again, only experimentation will tell me if I'm a lunatic or have found another novel way to magnetically confine electrons.  I really think there may be something here.  In any event, this fascination with five fold symmetrical magnetic trap structure is what led me to use the dodecahedron in my containment system design.

Well, that's it for tonight.  Let's hope that Tap Plastics has inch thick Plexiglas they can cut for me and I can start test pumping the system tomorrow night.

<heh>