Quantification of Electron Temperature in Jet Mode.

For posts specifically relating to fusor design, construction, and operation.
Post Reply
bk8509a
Posts: 40
Joined: Sat Oct 31, 2009 8:47 pm
Real name:

Quantification of Electron Temperature in Jet Mode.

Post by bk8509a » Sat May 08, 2010 3:57 pm

So here it is. The grand finale of my project, there should be more data coming, but its the best yet. I posted this a week ago, but I've been busy with graduation and finals.

I managed to stick my Langmuir probe directly inside of the electron beam in jet mode. I then kept the pressure and operating voltage/current stable while I collected data for the next five minutes.

The results are interesting. There's an upper and lower bound on electron temperature inside of the beam which the majority of the points are at. This means there are two different different classes of energy in the beam. This can either be two ions (oxygen and nitrogen), or two breeds of electrons. I've done the calcs by using conservation of energy and the ratios of the velocities should match the ratio of oxygen and nitrogen mass, it doesn't. This makes me believe that I'm making two species of electrons, "hot and cold".

I need more data though. I assume mid summer i'll have this figured out.

In the first picture I show you electron temperature over time, I tried to keep the fusor as stable as possible during this run, but its hard.

In the next picture is the corresponding Maxwellian. Its bi-modal, which suggests two different energies of electrons.

Next is a picture of the probe in the jet.

Lastly is a repeated run, showing that although the fusor can be at a different operating condition (Large change in pressure), the bi-modal distribution continues to exist.

So science nerds, what explains this? I feel like i'm starting to really do 'research'. I'm getting to the point where I have no clue whats going on!
Attachments
TwoMaxwellian.png
Jet2.JPG
Jet2.JPG (158.09 KiB) Viewed 2471 times
Maxwellian.png
TemperatureMid.png

User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Quantification of Electron Temperature in Jet Mode.

Post by Chris Bradley » Sun May 09, 2010 2:45 pm

It is fine work and, as you say, is surely 'research' where it is beyond any 'common' explanation.

My two main concerns with the data are a) what are you measuring and b) is the measurement technique appropriate for this 'non-uniform-plasma' scenario.

On the whole, one might expect that the electron energy [in the beam] is the same as the field potential (wrt cathode) at the point of the measurement of an electron beam. At least, there should be some statistical distribution up to that value, given that some fraction of electrons will have lost some energy.

For those electrons and ions that do loose their energy to the background, via thermalisation processes (that are the thermodynamic brick-wall stopping a fusor from getting anywhere near 'over unity' fusion energy) then one can expect a plasma medium in the background, as evidenced by the glow discharge throughout.

If your measurement location is as imaged and that the probe tips are a little off-side to the electron beam, then two modes seem possibly explained by the fact that you have an excited background medium of Maxwellian-distributed electrons and you also have electrons that have meandered their way off the beamline by some random-walk process, each step of which has sucked out some energy.

Your range of energies, 40 to 100eV, seems a little high for the background (I would expect electron temp to be a little lower for such a bright glow) and seems a little low for the beam (I would expect it to be around one third of your drive voltage) so if I take your results at face value I would say there is some complex interaction between these two sources of electrons at you probes' tips that lead to the technique returning these value.

It is also worth bearing in mind that the background ionisation percentage may be extremely tenuous in a fusor, leading to sheath thicknesses that may be of the order of the separation distance of your probes which may compromise your results.

I would imagine that many electrons arriving at your probe off the beam-line would still have plenty of energy and that they'd penetrate the sheath around the probe tips very easily, but as to whether this second 'source' of electrons penetrating the sheaths can be practically measured by the technique you are using is, I would suggest, questionable. I think 'firing' charged particles at a probe whose theory and design is to measure uniform plasma conditions will lead to results requiring further indirect measurement work and interpretation to extract a correct meaning.

It might be interesting to move the probe towards and away from the beamline and repeat.

User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 7:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: Quantification of Electron Temperature in Jet Mode.

Post by Doug Coulter » Sun May 09, 2010 6:54 pm

I'm certainly liking seeing this data -- questions aside, I know how hard it is to get, so kudos, Brian!

I'd add a couple questions to look into. Why is there only one beam in the photo? When we get things nice here on our cylinder fusor, we get beams out of each grid opening. Do you get this too, and is this a special case of *not* having beams from each opening? In our setup, they are rarely all the same brightness, but they are all always there (but, interestingly, not a different color than the main plasma, camera artifact here? -- we only see multiple colors at higher pressures than we run and usually with contamination present).

Is there any RF showing up on your probes? Could something cyclic be going on? You can tell a sine wave from noise via a probability density function and this looks kinda like you got one in the noise and sampling errors. It may (probably is) be severely aliased at your sample rate, but you could always check with a fast scope...

We see various kinds of equlibria and instabilites here -- this may be a good way to explore that?

And I will definately second Chris -- put that thing on a wiggle stick, STAT!
Why guess when you can know? Measure!

User avatar
Richard Hull
Moderator
Posts: 12277
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Quantification of Electron Temperature in Jet Mode.

Post by Richard Hull » Mon May 10, 2010 1:41 pm

I will say that the image is not representative of anything close to fusor, "fusing" operation where things even out and many very dim beams are extant in the device. I have images of 13 beams coming out of the cathode and they taper off to invisibility quickly. In some instances during top fusor functioning (doing nuclear fusion) the beams are often not readily visible.

The single beam I see in the image supplied is more like that of "demo mode" where the device is not at a good operating pressure. (perhaps at 25 microns or so)

The applied voltage has to be rather low, too. My guess is well under 10kv.

Data taken from this may not have much relationship to the actual conditions when fusion is actually taking place. I doubt is anyone has ever fused with such a visual appearance (single bright beam).

Just observations from a guy who has been fusing in the fusor for 11 years.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Quantification of Electron Temperature in Jet Mode.

Post by Chris Bradley » Mon May 10, 2010 4:57 pm

Richard Hull wrote:
> I will say that the image is not representative of anything close to fusor, "fusing" operation where things even out and many very dim beams are extant in the device. I have images of 13 beams coming out of the cathode and they taper off to invisibility quickly.

I had presumed this was clear, and it was not an intention to measure 'fusing ion' beams....

This appears to be an attempt to measure an electron beam from a plasmoid, not reciprocating ion beams. If I have misunderstood Brian's attempt at measuring an electron beam then I presume he'll let us know in due course.

User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 11:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Quantification of Electron Temperature in Jet Mode.

Post by Carl Willis » Mon May 10, 2010 8:04 pm

Hi Brian,

Many thanks for your participation here and for sharing this data. It's a useful pioneering foundation for basic fusor measurements with Langmuir probes, and regardless of what the data means (I don't know), your approach has been carefully-controlled, well-documented, and the data is available for analysis by hardcore plasma theoreticians if we get any such visitors.

Are you going to upload your "thesis" (or final paper) on the forum? It would be nice to have your entire project--calculations, measurements, and techniques--all in one place. I hope the wrapping-up phase of your college years goes smoothly. What are your future plans with the fusor?

Something I can't appreciate is why you refer to Fig. 2 as "Maxwellian." It is clearly rather bimodal in velocity and not Maxwellian. But I wonder how much of your calculations might assume a Maxwell distribution of electron energies, and fall apart if the distribution is different? Regarding the bimodal question, seems to me that most fusors (or at least those generating neutrons and running at low pressures, high voltages, and low currents) have high electron energies depending mostly on radial position and cathode voltage, dominating any low-temperature tail that may indeed be Maxwellian in character. The Debye lengths are large, perhaps too large for accuracy with a Langmuir probe. From my position of little expertise on this kind of measurement, I am wary of its many limitations and the many assumptions that are built into the measurement theory. Perhaps your paper will make the justification convincingly for me.

Thanks again for your generous contributions.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277

DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 5:13 am
Real name:

Re: Quantification of Electron Temperature in Jet Mode.

Post by DaveC » Mon May 10, 2010 9:02 pm

Brian -

Thanks for posting these interesting results. It's a challenging task to measure almost anything in a plasma.

One thought that occurred to me, was that with the complexities of any plasma, it probably would be more theoretically manageable to use a single gas species, and go to "reasonably heroic" lengths to be sure it's pure.... mostly the cleaning process, and a good few 9's pure gas- like hydrogen, first.

I am almost certain you are correct in your surmise that there may be several Oxygen types.... there are almost surely O, O2 and O3, plus neutrals and singly or multiply ionized... and the same would go for nitrogen, and then the NOx combinations and electrons. Most likely there is quite a stew inside.

Keep up the very nice work. Greatly appreciated, here.

Dave Cooper

bk8509a
Posts: 40
Joined: Sat Oct 31, 2009 8:47 pm
Real name:

Re: Quantification of Electron Temperature in Jet Mode.

Post by bk8509a » Sun May 16, 2010 4:58 pm

Finally have time to give explanations/comments.

Chris: I need to move the probe around more and take more measurements and see if I get the same distribution. The hard part is the data measurement for these operations is tedious and we have not had the time to do it. The fusor needs to stay at constant conditions. I am especially concerned that we're not finding electrons with a temperature that corresponds to the driving voltage. Once again, more data should solve this.

Doug: The one beam is from jet mode. Not in star mode yet. Can't measure star mode because the probe starts arcing. I find it interesting that you've never seen one blue beam of electrons come out of your cylindrical fusor. It happens on all other fusors as far as I can tell. Check out my thesis for more info on this. Should be under "modes of operation". Also, I would like to take data with a fast scope but I need an extremely high impedance. Because of this I'm using a special DMM.

Richard: For reasons explained above, I've never been able to measure fusing conditions as the electrons have sufficient energy to arc through the probe and outside the chamber. I've found that measuring the center of the fusor, or fusing conditions, requires not only a probe that can handle these harsh environments, but actually a whole circuit that can literally withstand these conditions. I originally thought I could stick my probe into star mode, take data, and be done with it. Apparently, science wont cut me a break.

Carl: The thesis is uploaded. Not to toot my own horn, but anyone that comes here that is knowledgeable enough to understand it, and disciplined enough to read it will find it chock full of what to expect, why to expect it, and most importantly: how to do it. It also has an extensive section on Langmuir probes. In fig 2 I call the distribution Maxwellian because I initially assumed a Maxwellian distribution to convert the temperatures into velocities. Poor terminology on my part. I left this falsity because it gives 'some' idea of velocities (probably a shit ton of error). I could have just done a temperature histogram, but I felt a velocity distribution was more traditional. To comment on your last observation, I'm afraid that i'm driving in the dark with the Langmuir probe and although I understand the basics, I am really unsure of the theory. Its a catch 22. In order to figure out if the Debye sheath crosses into other probe tips I have to measure the electron temperature, in order to measure the electron temperature I have to use a Langmuir probe. Whats wrong with this picture? Everything.

Dave: This summer we're working on getting argon to put in there. I really don't care too much for hydrogen because of the neutron safety hazard and the paperwork that comes along with it at an institution. I feel that having a breed of one gas might change things and we all agree at AU that there's tons of different gases in there that are completely messing up our "control".

Thanks everyone for the advice and comments. I'll probably get some more data in and show it to you all.

-BK

User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 7:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: Quantification of Electron Temperature in Jet Mode.

Post by Doug Coulter » Sun May 16, 2010 6:13 pm

"Doug: The one beam is from jet mode. Not in star mode yet. Can't measure star mode because the probe starts arcing. I find it interesting that you've never seen one blue beam of electrons come out of your cylindrical fusor. It happens on all other fusors as far as I can tell. Check out my thesis for more info on this. Should be under "modes of operation". Also, I would like to take data with a fast scope but I need an extremely high impedance. Because of this I'm using a special DMM."

Yes, we see what we call "blue bugles" in that jet mode (generally during pumpdown), and to add something that might be interesting here we find that when we get our grid really precise, that it will bounce from aperture to aperture. When we get it very-very precise, we will sometimes see more than one at a time from different apertures. And the interesting part is that this translates to better neutron output in star mode and our pulsed mode too. Maybe funny, but this same kind of thing is why more modern electron tubes had so much better performance than the ones DeForest was playing with -- precision matters, a lot. Which is one of several reasons we're doing cylinders -- makes precision easier by quite a lot.

Check on the arcing. We get "interesting" things happening on ungrounded things in the tank in various modes. We've seen tens of KV picked up by an unused HV feedthrough far from the action sometimes. Usually, but not always negative polarity. Sometimes enough to make them arc over outside the tank! The effect seems to be very sensitive to the exact operating conditions. It's not something I've chased down very hard at this point, but it's there. In this case it was one in the main 14" diameter tank far, far from the 6" sidearm we were running the fusor in, and in no case in line with any of the visible output beams or even invisible ones we can see on our pinhole/scintillator camera we put on a wobble stick. Remember, you can't actually see a beam of electrons -- takes more than that to produce light outside of synchrotron radiation effects and we're all well below those kinds of energies here.

Marlin P. Jones, of all people, sells a fast scope probe with 100 megohm input impedance if you're interested, and it is fairly cheap and HV rated to boot. www.mpja.com. That's pretty high impedance, but for more yet of course there are many inexpensive opamps out now that are in the e16 ohm range and even have built in choppers for DC accuracy. Of course, for those you have to rig things so you stay in their common mode range or lose them.

Although a lower impedance scope probe might mess up your other measurement, it would most likely still tell you if there was significant RF/AC content just held near one of your sensor leads and purely capacitively coupled through the air. A little ingenuity can go a long way for "go, no go" kinds of indications, even if they aren't quantitative. Easy to try at any rate.

Here's the full link to that probe:
http://www.mpja.com/prodinfo.asp?number=18148+TE
Why guess when you can know? Measure!

User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Quantification of Electron Temperature in Jet Mode.

Post by Chris Bradley » Sun May 16, 2010 6:24 pm

Hello Brian,

It is appreciated that you posted your thesis. I did scan through your thesis and there were a few points I would've clarified, if it were mine.

You say "The gas will ionize due to thermal electrons that will eject
off the inner grid (22) towards the wall (21). The newly freed nuclei feel an extremely strong
attraction to the negatively charged inner grid (20) and accelerate towards the negative
potential (20)." This is, of course, correct but I do not believe it is a good argument for why you get significant nucleii at the drive voltage. There are reasons for a high proportion of near-full-drive ions, so those mechanisms are clearly very important for a fusor (as without them, it would barely do fusion at all); (1) that electrons sputter the outer shell and the sputtering process ionises the background high up in the field's potential, and (2) that as ions slow down to reciprocate so their ionsiation cross-sections shoot high and so preferentially ionise at the highest potentials, furthest away from the grid. Because the reason you quoted isn't as significant towards ions at fusion potential as these other two causes, I would emphasise these other two processes (if it were my thesis)

You say "At the center (34) the nuclei will oscillate back and forth where they will
either collide and fuse or capture an electron." I suggest the expression "will
either collide *and* fuse" is misleading and should certainly read "will either collide and *may* fuse". Otherwise, it is missing the dominant process which is colliding *without* fusing. When nucleii collide, they buy a lottery ticket each, and if they both win *then* they fuse. The odds on a collision resulting in fusion is >1E8 in a fusor - hence the abysmal efficiency. It seems many folks do not understand this point so go on to think a conventional fusor is actually a useful approach to fusion energy.

You say "When two particles are brought close enough to each other their nuclei have the potential to overcome the repulsive Coulomb force and enter the realm of the nuclear force." This is not tehcnically correct because the process by which fusion occurs is one of quantum tunneling of the Coulomb barrier and the nucelii need to, probabilistically, tunnel through the Coulomb barrier *before* they get anywhere near nucelar force distances. The potential required to overcome the Coulomb force is some 40MeV, so it is wrong to suggest it can be overcome in a fusor. It is, so to speak, "undermined".

In regards your conclusions, they are your own, of course. I did not follow what you were trying to conclude with the 'beam lensing' discussion, for example. But you seem to have excluded comments and conclusions regarding an environment which is a cold plasma background with hot particles running through it. Fusors are 'generally-cold, locally-hot' devices and it is worth re-iterating that to ensure comprehension - imho. The electron temperature range you are measuring is clearly dominated by the cold background.

It would've been worth going back to the equations on the Debye length and re-examining the plasma conditions you started off talking about, but did not quantify. Not sure of the relevance of s2.2, unless you use that information later on.

Good technique and experimental write-up.

Thanks for sharing.

Post Reply