More subtle -- Fusor-2

Yours and mine. This is where you can gas on about how you think the universe works. To a point, after that we'll expect you to actually test your stuff and report.

More subtle -- Fusor-2

Postby Doug Coulter » Fri Sep 03, 2010 9:17 am

Here is the idea I'm actually putting my time and money into right now, as I think it's the next step in fusor-like design and performance. I say fusor-like as there are some serious differences between this design and a normal fusor, or at least any I've ever seen so far. Only one other person seems to be really thinking along these lines, and I think he was too easily dismissed as a whack-job elsewhere, though I also think my idea is a good bit better than his -- but until we test, we only think, we don't know.

Here is a picture of a two grid arrangement (3 electrodes, counting the tank) seen end on, for a cylindrical arrangment. Even though this is easier to draw and visualize, I apologize for my lack of skills in making pretty drawings in advance. (click the pic for a larger version that shows my sloppy drawing better)
Fusor2.gif
Mass-spec derived fusor


Taking a page from our colleagues who design mass spectrometers, why not try something along these lines? Lets assume the tank and inner grid are initially grounded, and that there is a source of ions coming in. Further, we are doing this at very good vacuum, similar to where most beam type devices work, so we've got fairly pure ions, few neutrals and other junk. Say e-6 or e-7 millibar which gives a decent ion storage time in this context -- we don't need a very long time here, but want pretty good purity on the ones we retain in the trap. In fact, this is just an ion trap as shown, with altered geometry from the ring-endcaps design so commonly seen, but here we have 8 traps as drawn. The math for this sort of thing is well covered in "Quadrupole Mass Spectrometry and its Applications", edited by Peter H. Dawson, AIP press (AVS Classics) ISBN: 1-56396-455-4 The main difference is that the existing math does not take space charge effects into account, and as electron tube design shows, they can actually dominate over the deliberately applied field gradients because particle inertia can push them into a place where they make a local field gradient larger than that which started them. So we will have to do some learning here -- but not much, we have some shoulders to stand on. The same sort of knowledge was required to design the most successful beam power tube ever -- the venerable 6L6, which eliminated the need for suppressor grids with space charge effects used for benefit, rather than a disruption, of the normal vacuum tube operation.

Here is a good paper on beam power tube design, with space charge effects, by Schade in 1938. One could argue that most fusors are kind of an attempt to make an inside out triode, with a virtual center electrode. Obviously, electron tube design didn't stop there -- why should we?

The basic operation at this stage of the fusion cycle is with the outer grid driven by mixed AC and DC, same as an ion trap, and the red figure 8's (which I don't draw well) are the paths of the ions in this trap as seen end on when you get the drive correct for D+ ions. This will only trap a certain charge/mass ratio as designed, the rest are lost to the grid wires, the walls, neutralized and taken off by the pump. The idea is to gain purity, ion density, and space-time bunching. For example at a certain phase, all the ions will be centered between the outer grid wires at the same distance radially from center, and all heading toward the inner grid. By running this for a number of cycles, we can build up an amount of ions limited by space charge effects from a relatively wimpy ion source (not shown), and eliminate all that aren't what are desired (things other than D+ in this case).

Once equilibrium has been achieved (or earlier if that winds up getting better results, too much space charge may spoil this stew), and at the correct phase of the ion motions around the outer grid, we fire our big negative pulse on the inner grid. A square pulse might be used, but I'm guessing that for best net focus a fancier waveform will be required to account for space charge density increases as the ions approach the inner grid -- probably need something that looks more like an exponential rise, rather than a simple quick straight up rise to compensate the space charge. Any pair of outer grid rods and inner grid rods define an electrostatic lens in this design, but most of the existing math for this depends on being able to neglect space charge, and here we want to run as dense as we can, so we will have to adapt, and probably discover the best thing empirically -- which will of course be obvious in hindsight.

As the ions approach the inner grid, we have near ideal conditions for focusing possible, and as yet nothing inside the inner grid to defocus and repel them back out. Once they pass the inner grid wires, their trajectories have been set, and the full energy is already on them by the time they "see" each other. They will slow one another down some due to repulsion to be sure, but due to the symmetry here will not be deflected off the basic trajectories very much, producing a fine focus at inner grid center in space and in time. As soon as the ones that didn't fuse fly through and decelerate in the inner-outer grid field, we resume the first phase of operation -- rebunching in space and time in preparation for the next shot for a few cycles, with the inner grid at ground again.

In this design, we can leave the outer grid with the same AC/DC drive on it all the time, the old U + V terminology in the mass spec literature, as the field when we pulse the inner grid will totally swamp that and make it not matter.

I've been discussing this with some others here for a while now, it seemed time to draw it up and explain and discuss it since I'm about to actually do it now. I believe this solves some large problems compared to a normal fusor. We are not worried here with such high pressures as to have self-ionization, we don't run in a soup of various ion c/m ratios and free electrons, and we don't depend on some magic gradient being setup via emergent behavior to create recirculation. Further, we limit space charge density until the very last instant, so we don't have the Couloumb explosion messing up focus. So this is a sort of divide and conquer strategy in re space charge effects. By the time they become significant, it's simply too late for them to mess things up as they do in a normal fusor, and we avoid putting energy into ions that won't interact with high probability.

I am guestimating that something like this is behind the super high Q pulse modes we've been working with and measuring, as well as trying to actively drive with lesser subtlety. But with good effect. Minimal perturbation power anywhere near pulse mode conditions improves the Q of our otherwise Plain Jane fusor by factor of 3 -- with milliwatt extra input for the perturbation signal. I think I've learned enough from that work to take that from something that happens more or less by accident in a very narrow range of conditions to something we do "With Malice Aforethought". Hence my other discussion of things like arbitrary waveform generation, large video amplifiers and other related topics -- this is what was behind all that talk.

At any rate, this won't be a theory for long, I expect to get this tried before the year is out, with any luck. The mechanicals are built, the gear is mostly in-hand, and a few things need to be finished up to get going on it (the video amps, and more data acquisition).
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: More subtle -- Fusor-2

Postby chrismb » Sat Jul 02, 2011 7:08 am

Browsing patents and came across one of which this is a figure;

20110042562.jpg


Maybe the text contains something that might catch your imagination, Doug;

pat20110042562.pdf
(1.02 MiB) Downloaded 508 times
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm


Return to Theories and speculations

Who is online

Users browsing this forum: No registered users and 5 guests

cron