Page 2 of 3

Posted: Mon Apr 28, 2008 5:45 pm
by drmike
Here are the electron impact ionization cross section charts. Since the effective size of a particle goes as the DeBroglie wavelength, it makes sense that high energy alphas won't have much of a cross section with slow neutrals. A cold dense plasma will ionize pretty quickly though (and 100 eV is "cold" for our purposes). I would expect a neutral beam to be non-neutral by the time the particles get to the core. Not so sure about full ionization though.

Posted: Mon Apr 28, 2008 9:21 pm
by 93143
Whatever happened to ECR? Tuned to match the magnetic field near the edge of the wiffleball, it should produce near total ionization right at that depth, right?

Posted: Tue Apr 29, 2008 1:19 am
by cuddihy
So, to recap the problem is neutral fuel ionization at reasonable (~100 MW fusion) power levels, with four different suggested methods in this thread. Stop me when I sound stupid:

1. Continuous neutral gas flow (Msimon), with ionization assumed to occur by a combination particle collisions. Dr. Mike is convinced this won't work due to small alpha flux and cross section for collision w/ boron. But that leaves the question of whether the mix of fuel particles, electrons and high energy alphas could ionize the incoming fuel without causing too much spreading in their energies. I guess that's a question that requires a lot more knowledge about particle lifetimes, abundances, cross sections, and geometry. Perhaps, not calcuable without data from an operating WB rx.

2. "Neutral" particle beam injection (jmc) into the core. Seems like it would cause problems with deceleration grid and possibly thermal issues as well.

3. wakefield acceleration by electron or laser pulses(?) (Dr.Mike) I have no clue how this works, or what the pulses come from.

4. ECR --obviously this works for S/U. Does it work in an operating reactor, or would the ECR interfere with the potential well? Not sure I understand how to implement this in a polywell.

Posted: Tue Apr 29, 2008 2:16 am
by MSimon
If puff gas injection works I see no reason continuous flow will not work even if my suggested mechanism is incorrect.

Posted: Tue Apr 29, 2008 3:29 am
by drmike
@cuddihy - you're mixing too many metaphores (wrong word maybe, reminds me of the cartoon in the newspaper called "Crankshaft"). I think jmc's idea of using the neutral beam is to ensure the ion velocities stay focused on the center and fueling is a useful side effect.
In a way, if the plasma is thin and you don't ionize, the beam passes thru the system and can be recirculated on the other side.

The idea of POPS (and what I call "wakefield") is to create waves of ions that ride on an elelectrostatic (or electromagnetic) field created by the grids (or antennas) pushing the electrons around. The electrons are light, but the electric force is strong, so if you push the electrons around, the ions will follow. This has nothing to do with fueling the system.

Puffing gas in for refueling is yet another idea. Using microwaves to help ionize is still another method. And all of these methods can be used together in a final system - you could pulse in a neutral beam, and just when it gets to a specific point, zap it with microwaves to ionize it, but make the microwaves shaped to focus the resulting ions towards the center. If that changes your Q from 8x to 10x it may well be worth the effort.

I don't think we can say it is one thing or another at this point - it's better to keep all these ideas floating along so we can put them together in new ways to solve old and new problems. And let's hope we get lots and lots of new problems!!

Posted: Tue Apr 29, 2008 4:54 pm
by jmc
You have to remember that the edge is very, thin, it will be of order a debye length, even WB4 whose ion density was only of order 10^12/m^3 (a million times less than a tokamak) had an edge that was only about a centimetre and if you multiply the density a million fold you decrease the edge by a factor of 1000, how are you going to drop neutral gas in at exactly the same potential if the region at which the potential drops is less than 0.1mm?? Much easier to fire neutral particles in at uniform velocity and ionize them at the bottom of the well instead.

The second advantage is less electrons in the adiabatic region. If you ionize neutrals in a region where there is field then you leave electron behind in this region, these electrons then ionize more neutrals until eventually your plasma extends all the way to the magrid, the ionizing layer for neutral gas will continue to spread infact until it reaches somekind of scraper. The disadvantages of this are twofold:

1) heat gets conducted from like-like collissions all the way to the magrid

2) The potential well now exists in a region of magnetic field, this will ruin ion convergence.

I strongly suspect that the corners of the WB4 device actually scraped off plasma from this region possibly reducing this adiabatic population.




MSimon:
Why do you believe puff gas injection works? WB6 had far from a conclusive results.

Posted: Tue Apr 29, 2008 10:50 pm
by MSimon
Why do I believe puff gas works?

A very good reason is that experienced scientists who studied those results are recreating a more robust WB-6.

Dr. Mike and I are of the opinion that wakefield acceleration renormalizes particle energies no matter where in the reaction space they get ionized.

The Debye length is not important because the machine is not steady state (which was the way I conceived of it for a long time and thus was always unclear how it could possibly work). You have oscillating beams.

I don't believe much in the edge renormalization theory. I think the vast majority of renormalization is done by the wakefield of the oscillating beams.

The only way to be sure of what is going on is experiments. Which are going on as we speak.

Posted: Wed Apr 30, 2008 12:46 am
by cuddihy
I suppose what's got me utterly confused is the way the word "beam" is being used. I'm used to thinking of a beam as a particle stream with one origin and traveling in a tightly focused..well, beam, in a straight direction.

simon, which oscillating beams are you referring to? is that the electron beams being shot into the machine? or is that a description of ion and electron behavior in "steady state" in the machine?

Posted: Wed Apr 30, 2008 2:25 am
by drmike
I think of it as spherical compression waves. It is a "beam" in phase space - just like in a particle accelerator you have an emmittance that describes the particles in both position and velocity, the final focus is a point. After the particles pass the center focus, you want them to maintain their spherical shell shape as much as possible so you can turn them around and tidily smash them into the center along with all the other sides.

So the trick will be to use the incoming waves to slow the outgoing waves down without doing too much damage to either one. A very "interesting" physics problem!

There are other models of fusors where beams as you think of them are developed. They used multiple grids to form the beams and it helped the model quite a bit (it was a paper study as I recall, but I don't remeber where to look for it).

There's just a lot of fun physics to do, and not enough time to play with all of it!!

Posted: Wed Apr 30, 2008 3:20 am
by MSimon
Dr Mike,

http://ssl.mit.edu/publications/theses/ ... Thomas.pdf

http://ssl.mit.edu/publications/theses/ ... chCarl.pdf

If those guys got their codes right, what they saw was 6 beams (not well formed) oscillating in phase coherence. i.e. all going out and returning at the same time.

This could be confirmed in WB-7 by using a spectrum analyzer that can analyze over short bursts. Commonly used for cell phone testing, WiFi, etc. I believe they are called burst analyzers.

Ion beams would oscillate inside the reaction space. Electron beams would oscillate from the reaction space to the "dead" space and back.

Let me note that Joel (something) in a private e-mail got similar results.

This is unrelated to the electron beams used to get an initial net negative charge in the system.

Another possibility for testing would be to use a LeCroy (or similar) scope that can sample at 1 GHz real time. You then do an FFT on the samples to determine the frequencies produced.

Posted: Wed Apr 30, 2008 8:27 am
by jmc
Well, I suppose the fact that experienced scientists are recreating a more robust version of WB6 to investigate the inconclusive results from WB6 is a sign that gas puff has a chance of producing fusion.

Having said that they are strapped for cash and it may be they are using gas puff instead of neutral beams simply because they can't afford one at the moment rather than because it wouldn't enhance the chances of success.

Regarding wakefield acceleration renormalising particle energies, the closest analogue to waves influencing particle energies I can think of is Landau damping, this does not have the effect of accumulating particles energies at the wave velocity, rather it flattens particle energies on either side (although I suppose there could be some non-linear effect that does), in addition to this if your going to put an AC voltage on the thermionic emmitter, you better be sure the electrons really will stay in phase with the wave, because if they don't they come back at the emmitter at a different energy from their birth energy, this could lead to more losses.

I read thomas McGuires thesis its very interesting from an academic point of view and uncovered an entirely new phenomenon to do with IEC. As far as I'm aware he only modelled the ions, at extremely low densities, and not the electrons. While the Q-values he worked out were possible may have exceeded 1 the power densities were too low to make a credible reactor.

I haven't read Carl Dietrichs thesis although I intent to.

I got forwarded some of Joel's correspondences. I only recieved the ones showing Polywells had trouble converging in steady state, I didn't recieve his email talking about electron beam oscillations.

I'd be really interested to see the computational results for electrons since any device of credible power densities requires them.

Posted: Wed Apr 30, 2008 10:08 am
by tonybarry
Would it be appropriate to have two ion beams, diametrically opposed, such that the ejected particles run into each other and the collisions more or less wipe out the net kinetic energy?

I realise I know little about proton (or P-B11) collisions, and they may not behave correctly, but is it worth a try?

Regards,
Tony Barry

Posted: Wed Apr 30, 2008 10:17 am
by MSimon
Re: wakefield

Keep in mind that the beam is varying in velocity over its cycle.

==

Thomas McGuire's thesis re: beam oscillations has been confirmed in a private e-mail by Joel. I'd have to go back and see if he developed his own code. Also some one at the fusor forum did an RF check with a HF radio and found an RF peak at a frequency around the range I would expect from BOE calculations. Given that those folks are generally not excellent experimentalists I'd have to say the results are merely suggestive rather than definitive.

Carl Dietrich's thesis is more focused on the hardware aspects of the device.

At this point our ignorance vastly exceeds our knowledge. Which is a very fun place to be if you can stand the wait.

Posted: Wed Apr 30, 2008 11:27 am
by Keegan
^Nice to hear from you Tony

Simon do you have any links for the POPS work @ the fusor forum ? Im guessing the he must of did a frequency sweep with a HF radio and derived the results on a spectrum analyzer. You know that i have been thinking that same thing for a while now, but i realized that method had some deep flaws. Theres a much better way to do it, ill have it posted soon.

Posted: Wed Apr 30, 2008 11:57 am
by MSimon
Keegan wrote:^Nice to hear from you Tony

Simon do you have any links for the POPS work @ the fusor forum ? Im guessing the he must of did a frequency sweep with a HF radio and derived the results on a spectrum analyzer. You know that i have been thinking that same thing for a while now, but i realized that method had some deep flaws. Theres a much better way to do it, ill have it posted soon.
I'd have to go rooting around. You might want to try search terms:

RF M. Simon

or

Radio freq

As per usual I had something to say on the matter.

This might be a good place to start:

http://www.fusor.net/board/view.php?bn= ... dio%20freq