New simplifying approach -- beams

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.

New simplifying approach -- beams

Postby Doug Coulter » Mon May 23, 2011 4:03 pm

After a week of intense work with JonH here and some really good BS sessions, I'm going to propose a new way forward for my own little lab and local group. Although the fusor ->looks<- like a simple system, what's actually going on in there is pretty complex as things evolve with emergent behavior. Of course, I will continue to have a fusor, and to add ever more diagnostics to it to find out the whichness of the why and what's really going on, in parallel with any new effort here - it's ripe for that and it would be ignorant not to, but it seems hard to figure out how to manage that hive of charged ants to get to true understanding and significant improvements with the true complexity that seems to really be going on in there. Right now, its main value seems to be a real good reliable source of a lot of neutrons, but the next step, if it's not fusor II as described elsewhere here, is going to be fairly daunting to get going, and we need some more steps, reasonably fast, if we're to get to gain while I'm here to enjoy seeing it happen.

Jon and I discussed the cone trap tabletop beam device, modified to be a collider (it's cool that you can do that with the same ring and with no magnets, most of it doesn't care which way the particles fly) at some length, looking to simplify the system for better study and more-clear results. It is attractive that way, and would fit through any of the three doors on our systems already operating for testing.

But on further thought, even more simplification seems like a good idea, with the side bonus of being even easier to get going with. And that is - simple colliding beams. Just two electrostatic accelerators (gated) firing at one another. Pretty much all the stuff to do that is "at hand" except for a few pieces that need fabbed in the shop, no big deal at this point. And to do the other thing, you need a beam to inject into it anyway -- so no matter what, not wasted effort to get up to speed on beams, focus, bunching, and gating them in an even simpler system. Once that is going, we can then look at the other things we've discussed on the board -- conservation laws, Pauli's exclusion principle, selection rules, all that -- if we can't even do that in a simple once through system, I don't see how it'd be any easier in a recirculating one -- either the Cone trap beam ring, or a plain fusor.

So I am proposing a change of direction for my own lab -- as usual, what y'all do is on you -- but I like to get input from all the smart guys here, and we are lucky that way.

Curtis suggested that head on collisions might not work as well as glancing in some cases, intuiting those selection rules without actually knowing the underlying physics himself. Well, this kind of setup would make that easy to test. Finding out we need an extra foot of beam to do things like spin alignment would be no big deal in such a system -- just insert another section of beamline in the pipes to do that.

Since Chris is already "on" the idea of making recirculation and recovery from scattering his priority and has a decent enough plan for that (and I have some ideas to add once he actually posts his details here), it seems wise for us to work on making the reaction probability per pass higher - we can always combine the two approaches after each works separately, no? That's the value of a virtual team, all free to go their own way and follow their own noses but with collaboration (without the boring meetings and bureaucratic policy impediments) or so I think.

Mark Bickey has already made me the quartz fittings for a beam on target device -- adaptable to this work right off. And I have tons more tubing we can insert into a beam line at will, with "whatever" in it to manipulate the beam -- once we learn how to make and manipulate them generally -- gating and bunching are as yet unexplored here, and need to be looked at for example.

So, y'all chime in here -- as usual, we'll all do whatever it is we want to do anyway (no way I'm even trying to herd we squirrels), but if someone has an insight, why not share it?
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: New simplifying approach -- beams

Postby chrismb » Mon May 23, 2011 4:40 pm

I don't buy that you'll get many reactions at all. The fusion reactions in a DD fusor is because [and I stand by this as it is the only description of a fusor that stands up to all data I have read] the deuterium is whacking through a background of ionised gas [mostly gas] of deuterium. Fire beams at each other and I predict you'll see nothing much at all.

However... ( ;) ) ... I think the essence of what you want to do is cut back to the basics so as to reduce all the variables, and build up again to understand what beam manipulations you can perform. I'd suggest the best way to do this is to run singular beams into a solid target. Either lithium [remember this was the lowest hanging fusion fruit of all time...Cockroft and Walton] or follow one of the designs for neutron sources in that IAEA document that got posted up on fusor.net some time back, such as those that soak up deuterium fired at it in a target that then becomes the reactant for more deuterium fired into it. Brilliant document that one, lots of very interesting projects that could be followed up on parts of that!

[As far as I read the cross-section plots, p+6Li and p+7Li are similar to d+6Li and d+7Li, so you can start off with protons and work towards deuterons.]

Bottom line from my opinion is that if you haven't got enough beam energy to 'light up' a solid lithium target into fusion reactions, then you aren't going to have what it takes for beam-beam.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: New simplifying approach -- beams

Postby Doug Coulter » Mon May 23, 2011 4:56 pm

Always glad to hear from the resident skeptic. I don't expect a ton of reactions here -- and I don't yet care as I can measure pretty far down these days, and am interested more in what I can do to control the probability and improve it - so may as well start small - so as not to swamp our sensitive detectors when we get improvements. I tend not to make bets I think cannot win, hence the use of when, rather than if in the previous sentence.
Another way of saying that is: If I think big improvements are possible, I'd better start low or I won't be able to meter them well (or have to scale down in the middle of the game, with uncertainty), and I plan for success -- failure is what happens if you don't plan and execute, and sometimes (but less) even if you do. I am trying to load the dice of serendipitous discovery possibilities here.

As far as the other, yes, the Li reaction is my own favorite personally, but finding the hot alphas in a sea of protons seems like more work, and adds uncertainty, than using the working and trusty calibrated neutron detectors I now have. As far as beam on target, yeah, sure, I've been linking the Phillips borehole tube design up here (and on fusor.net) forever, this might be the tenth time here. Look at the chapter end. That's what Mark made up the quartz pieces for in fact. And that's easy to check, especially once I have a beam in proper focus to even try that, in this case a wide smooth beam ( don't point-heat the target as discussed in this old chapter) - and if we can't even do that, no point going for, well, point beams ;) Bill has already been on getting some pure H so we can tune the beam focus without getting burned up by neutrons -- note the expected reaction rates in that document. I see things like millions per microamp there...and yes, we have T if we want to play with that some but probably will not this year. No point, no need, we can tell from DD what we want to know with beam->target.

The important problem in all philosophy boils down to "what do I do next, right now". So in fact you seem to be agreeing with the approach, with the caveats you mentioned. I have them handled, I hope.
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: New simplifying approach -- beams

Postby chrismb » Mon May 23, 2011 5:08 pm

Yes, I agree with the principle of 'doing beams' intentionally, rather than allowing something-like-beams to form in a fusor but then wondering what the heck you've acutally got in there. I think beam-beam is too big a jump for stage 1, but you've thought about it and understand that's my point, so I've no need to labour it!

So, whether one or two [or...?] beams, next thing up is to say what beam-forming method you are planning to use?
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: New simplifying approach -- beams

Postby Doug Coulter » Mon May 23, 2011 5:21 pm

At first, yeah, beam on target -- whichever target, and I note that the "smart target" that couloub-focuses beams in a silicon lattice is also testable with this same hardware.

I'm still at the "gee whiz" stage here. I'd use a uwave ion source (good fraction of monatomic ions there) with electrostatic acceleration -- the first half of that '50's Phillips design, for which I have the parts and the power supplies to get to 125kv and more. I have a working 125kv positive supply, and that disc supply I was working on can be either polarity and 180kv, so I've got that more or less handled -- it will take some time to finish the bigger disc multiplier supply but it works great into a few discs already (tested to 60kv) -- just need to learn to handle the real high outputs.

I also have the stack and transformers from a negative 160 kv Spellman supply, complete with the sexy thick HV cables and connectors (thanks a million, or 160k, Cliff) to keep the whole biz out of air. So beam energy -- no problem. Luminosity is another question. As is gating/bunching. If I fire two electrostatic DC accelerators at one another 180 degrees (but that's not necessarily the plan) -- once I get beam quality where it's close -- there will be a buildup of complexity as the ions that miss and scatter increase in there. So it needs to be gated so you can see the evolution of that going on -- this is related to SteveS's reactor, but here I don't think there's going to be magic recirculation, just a spring-mass system, though as you said, as crud builds up in that -- there will be plenty beam on stationary neutral collisions once it does - hence the need to gate it and let that die off and be pumped out to see the real information before the crud builds up. Steve totally missed that whole issue.

So, for the moment, simple (!) ion source, simple E lens (two element) and some sort of gating grid as being able to control that seems really important. Then work from there.

And of course, any such apparatus as I'd have to start building would allow for a solid target in there for initial testing, even if that target is a mere phosphor screen to see the beam with.

I do think in the longer run that beam on beam of some form will be the way, unless you can move and replenish hard targets in there pretty quick. At real power, they'd be toast faster than you could see it happen!
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: New simplifying approach -- beams

Postby Joe Jarski » Mon May 23, 2011 7:40 pm

Beam on target is obviously the easier approach and a good place to start, but if you're looking for the effects of manipulation then... well, you don't really have any control of what's embedded in the target. So I'd have to agree that beam on beam would be the end goal to find out if there is anything to the theory.
User avatar
Joe Jarski
 
Posts: 231
Joined: Thu Sep 16, 2010 8:37 pm
Location: SouthEast Michigan

Re: New simplifying approach -- beams

Postby johnf » Mon May 23, 2011 7:59 pm

Doug

That was what I was going to do until that $%^%&$ took my beam line and scrapped it
hollow cathode ala conetrap and a anode layer source type ion sources to give and annulus beam to be focussed by an einzel after the ion source at each end, cathode in the centre of the beam line -50kV below earth, ion sources floating +33kV above earth giving two 80kV beams head on. This equivalent to a 160kV beam on target
From calcs here at these energies the beam would be travelling around 400km/sec so very little chance of beam spreading.

Beam on target Silicon probably not so good but Ti, Pd,Li all good as they do the hydride thing very well
johnf
 
Posts: 433
Joined: Sun Aug 08, 2010 3:51 pm
Location: Wellington New Zealand

Re: New simplifying approach -- beams

Postby Doug Coulter » Tue May 24, 2011 12:16 pm

I was thinking of the silicon (very thin, only a couple hundred atomic layers) as a focus device -- analogous to an array of funnels to concentrate the beams into a lot of little ones centered on the lattice "holes" via Rutherford scattering. The idea not being to trap D's in there -- too thin for that, I'd expect most to make it on through if they didn't make a head on collision inside one of the pipes. Of course target life would be a single shot for that one and you'd need a fancy step and repeat kind of thing to really make use of it. The only reason I was thinking of silicon is that picture and words I saw about it on Wikipedia where it seems for most implantation applications there are troubles caused by its transparency along that crystal axis. There are probably other substances that would be as good or better - that's just the one I noticed. For that one the idea is not to trap the D's and then hit them, but simply guide the two opposing beams into one another more efficiently (all the D's still moving at most of the top speed) -- increasing the effective luminosity of each by cutting the area to just the insides of the channels.

And on the rest -- yup, pretty close to what I envision myself. Right now, I'm looking at how to gate the beams pre-acceleration. As D's move about 60 times slower than electrons for a given voltage, it looks pretty hard to get anything like short pulses without quite a voltage swing on the "grid" or whatever is used for gating, and all that stuff would be well off ground.

I guess I need to be looking into some kind of reasonably fast fiber optics components to control things like that, and probably building up a copy of your shaft driven generator rig as well, as the ion sources need considerable power up there. I have the pieces in hand I think -- just need to do the mech design and make it. I think the gating is going to be important to seeing what I want to learn here -- letting accellerated ions build up in the thing and create a background of randomly moving ions and neutrals doesn't seem like what's wanted at all. Since this is for learning -- would no way be the final configuration -- I want the cleanest possible situation to learn in.

I'm thinking uWave ion sources for two reasons here. One is that they run way down in pressure, so I don't have to differentially pump the mess to get real nice long mean free paths. Here I've seen mine run down to e-6 mbar and maybe a little lower (once started at higher pressures). The other reason is that it's said they do a much better job of making monatomic ions than the usual magnet/dc type of source, and that's what I think I want, without having to have a charge/mass selector in there -- simple is good. Not to mention -- the one I've been using runs, and runs and runs and runs - it picked up hundreds of hours on the last installation with zero issues whatever -- it just works. When I took it out, it had embrittled the quartz tube and I broke that, but all things considered - looks like at least a thousand hour with zero maintenance thing.
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: New simplifying approach -- beams

Postby Doug Coulter » Tue May 24, 2011 1:27 pm

I think Chris pretty much hammered where I'm coming from with this:

Yes, I agree with the principle of 'doing beams' intentionally, rather than allowing something-like-beams to form in a fusor but then wondering what the heck you've actually got in there.


Nicely stated. :)

You can attempt to measure "what the heck you've actually got" and I've been attempting. But to the extent I find anything out, it's pretty unexpected. And also, to the extent I get what Jon just called "the crispest focus I've seen in a fusor" -- it works better, but that focus is still a joke (say, ~2mm) considering what ought to be possible, at least going by what you can see -- and it's not like you can turn some knob and adjust the mechanical structure of the grid to try and tune it in better (voltage doesn't seem to make that much difference in it, which is weird all by itself - but it does get worse with increased current, as expected). Most attempts to measure anything are thwarted by the probe messing things up by just being there, as well. I'd have to get really fancy to measure things well without perturbing the thing too much. Stuff like laser interferometry, microwaves, hyper narrow band optical filters to see doppler, junk like that. Just too much for me at the moment. And there's that "first one" problem -- how do I calibrate stuff like that so I know what it's telling me?

There is something going on in a regular fusor that's confusing me, and that's that when I perturb it, the response is really slow in terms of what transit times of things ought to be in there. And I mean really really slow. The very highest frequency at which things happen seems to be about 1.8 mhz, but in that pulse mode -- things like a couple kHz - slow. So it seems there's some sort of group behavior going on that takes some roundy-rounds to establish or change much.

When I tried that 2nd grid, tied to AC via a capacitor so it could self-rectify -- 60hz is about the upper frequency limit at which any fusor type action happens, even if a sea of ions is already present. Above that you get a blob of glow near the grid, no neutrons, no beams. At 60 hz, it did stimulate pulsing on the main grid, that at a couple khz, chaotic, starting milliseconds after the #2 grid "lit" on the negative half cycle....While Jon was here we hooked up a fast audio transformer to a 700w amplifier and tried up to 15khz with interesting but useless effects. Self rectification like I was getting pretty reliably at 60hz on down became intermittent, very strange. The transit times for the electrons one would expect create that self rectification should be in the mHz range -- but nope.

So though I don't understand what it is, it seems there is what I'll call emergent behavior -- just like one ant is stupid, but a colony is pretty smart. I discussed this with the gents who write the code for SIMION, and they say no way they can model something like this yet. Despite looking simple, this is super complex. I won't say that this emergent behavior can't be used to good, just that I can't figure out how without understanding it a lot better than I do from the hints I've been able to gather so far. I do know a few things from operational experience, but not enough to have a clear way forward in this configuration, and by golly the huge improvements that were almost a daily event have dropped off now. There is still that hard-to-get self pulsing mode that shows the really big Q (a record I think) but...it's hard to get and even that high Q isn't in range for "boiling that cup of tea" with fusion energy.

I did find one thing (which tends to agree with Chris' ideas). That's that a normal fusor keeps working better the less gas you can run in there, down to the point where it's just running on the ion generator at very low currents (micro-amps) -- a couple or 3 orders magnitude less pressure than the point it will self-run at. Q is smoothly rising going that direction (though output is falling -- just not as fast as power input), and it seems even the threshold voltage for detectable neutrons goes down a goodly bit, indicating to me that his idea of things mostly being ion on more or less stationary other ion or neutral is much of what's going on in a typical fusor under typical conditions, far from the ideal full energy beam on beam.

Those who won't be stopped can't be stopped. This is a reaction to that. If I can't get smarter, I'll make the system simpler until I really grok what's going on, then build the complexity back up step by step. I realize I'll have to push my detector sensitivity -- but frankly, I was in danger with the big fusor of swamping them and making them inaccurate anyway, so they'll be happier this way. The big guy is at the point of making the 3He's have pulse pileup, and I'm seeing events even during the "dead time" of the tubes -- that one really needs the activation measurements to get good numbers on, which aren't real time....And for me (perhaps I'm too impatient and not meticulous enough) that's a killer -- I have to be able to adjust this or that and see results right away to find sweet spots without months of tests (during which time, controlling all the other variables is pretty hard).

Given what I've seen so far, I even expect that with real beam on beam I'll see the cross section peak at some fairly low voltage, about the square root of what you see beam on target. This might be pretty nifty!
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: New simplifying approach -- beams

Postby chrismb » Tue May 24, 2011 2:01 pm

Not wishing to sound like I am just a big know-it-all, but I've said plenty about how I think a fusor works [on fusor.net] and I've seen precious little evidence to suggest anything I've hypothesised is wrong.

So let me throw in a little nugget out of my very first post which got exactly ZERO comments there, or since, which might feed your imagination!!

I suggested that the main operational behaviour of a fusor is 'bulk-beam' [in fact, you can search for that term]. That is, that ions run back-and-forth in the visible plasma beams as little packets of ions. These packets would tend to self-pinch against the electron-dominant background plasma of those visible beams, but they'd take a few 'moments' [in 'plasma-time'!] to form up. In part, these packets are fed by up-scattering of ions that then collide with the shell, ionise locally with a spray of electrons, then those ions are then positioned perfectly at the top of the electrostatic well to get a full run-up of energy, synchronous with that packet at its highest point in the e-well. It's a bit like a self-exciting laser - new ions are added as the packet reaches the top of the well, loses its upscattered ions, gains more new ones than it's lost, then heads back on in [but loses the net gain in scattering along its diametric route].

This explains a) why you can focus beams, but only up to a certain current, b) why they are self-perpetuating along particular beam lines, c) some other stuff I forgot long ago... ;)

Try the idea out, run with it as a mental experiment and see if it takes you anywhere...
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Next

Return to Theories and speculations

Who is online

Users browsing this forum: No registered users and 3 guests

cron