Why is polywell supposed to be better than cusp confinement?

Discuss how polywell fusion works; share theoretical questions and answers.

Moderators: tonybarry, MSimon

JohnP
Posts: 296
Joined: Mon Jul 09, 2007 3:29 am
Location: Chicago

Post by JohnP »

drmike wrote:Interesting coincidences. Fred Ribe got me into plasma physics and grad school because I invited him to give a talk at Boulder. I finished at UW under Conrad.

To start with, check out Bussard's comparison of "Multi-cusp confinement" with cusp confinement: http://stinet.dtic.mil/cgi-bin/GetTRDoc ... tTRDoc.pdf

The main difference is presence of electrostatic fields.
I tried this link but the server says it's forbidden.

rnebel
Posts: 144
Joined: Mon Dec 24, 2007 12:15 am

Post by rnebel »

rcain

The Wiffle-ball was demonstrated on the WB-2 over 10 years ago. It's not so much that wiffle-ball confinement is so great but that mirror confinement in a cusp is so awful.

rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

Post by rcain »

scareduck wrote:
rcain wrote:To pstudier: On the other hand, worrying to hear Indrek has turned septic, what has turned him do you suppose? (Maybe we should ask him).
http://tech.groups.yahoo.com/group/IEC_ ... ssage/1407
hmm, thanks for that ... to quote from Indrek then:
...
There is little evidence for this kind
of virtual cathode formation and all the factors seem to be
against it.
...
What we've neglected here though is the wiffleball. Does
it exist? Does it work, does it help? If it does then technically
it could resolve this issue, right? So what we need is also
proper criticism of the wiffleball. Can you offer some?
...
re: virtual cathode - I was under the impression that this was fairly well known science (at least to Bussard's vacuum tube generation)

Although tinkering with the visualizations/rough models on http://www.falstad.com/vector3dm/ it all seemed terribly sensitive to precise configuration as to whether they formed inside, outside or on top of the magrid, or not at all.

re: WB - glad I'm not the only one uncertain about state of the art. Looking forward to futher news, obviously :)

rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

Post by rcain »

rnebel wrote:rcain

The Wiffle-ball was demonstrated on the WB-2 over 10 years ago. It's not so much that wiffle-ball confinement is so great but that mirror confinement in a cusp is so awful.
Dr Nebel. Thank you very much. Please excuse my lack of knowledge there. I'm doing my best to catch up :)

(Nice Plasma pics from WB7 by the way. Any neutrons yet?)

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

http://tech.groups.yahoo.com/group/IEC_ ... ssage/1407

There is little evidence for this kind
of virtual cathode formation and all the factors seem to be
against it.
I'm not sure why people still dispute the well forms. This has been shown experimentally. Experiment > simulation.

http://wwwsoc.nii.ac.jp/aesj/division/f ... hikawa.pdf
Actually, this has been the
central key issue for IECF researchers for
the past 30 years, until the first successful
direct measurement of the double-well
potential profile in the IECF device through
the laser- induced fluorescence (LIF)
method at Kyoto University [6] in 1999
with an approximately 200 V dip at the
center in the helium plasma core as will be
described below.
As the table in that study shows, there had been some doubt before, but laser-induced fluorescence seems pretty cut and dried. Unless I'm missing something...?

pstudier
Posts: 79
Joined: Thu Jun 28, 2007 11:37 pm

Re: Why is polywell supposed to be better than cusp confinem

Post by pstudier »

Art Carlson wrote:Greetings! After some interesting though inconclusive exchanges with some of you on cosmiclog.msnbc.msn.com, I thought it would be fun to drop by here.
If you haven't found this yet, here is the Wikipedia page: http://en.wikipedia.org/wiki/Polywell and here is the Wikipedia discussion page: http://en.wikipedia.org/wiki/Talk:Polywell and here is the archived old talk page http://en.wikipedia.org/wiki/Talk:Polywell/Archive_1 .

There used to be a Polywell wiki but it never saw much traffic and now it is gone.
Fusion is easy, but break even is horrendous.

scareduck
Posts: 552
Joined: Wed Oct 17, 2007 5:03 am

Post by scareduck »

pstudier -- Art has been active on the Polywell talk page you linked to, so it's safe to say he already knows about it.

pstudier
Posts: 79
Joined: Thu Jun 28, 2007 11:37 pm

Post by pstudier »

scareduck wrote:pstudier -- Art has been active on the Polywell talk page you linked to, so it's safe to say he already knows about it.
Sorry, I must have had a senior moment. I knew the name was familiar.

Does anyone know what happened to the Polywell Wiki?
Fusion is easy, but break even is horrendous.

dch24
Posts: 142
Joined: Sat Oct 27, 2007 10:43 pm

Post by dch24 »

rcain wrote:To pstudier: On the other hand, worrying to hear Indrek has turned sceptic, what has turned him do you suppose? (Maybe we should ask him).
Indrek wrote:...
There is little evidence for this kind
of virtual cathode formation and all the factors seem to be
against it.
...
What we've neglected here though is the wiffleball. Does
it exist? Does it work, does it help? If it does then technically
it could resolve this issue, right? So what we need is also
proper criticism of the wiffleball. Can you offer some?
...
rcain wrote:re: virtual cathode - I was under the impression that this was fairly well known science (at least to Bussard's vacuum tube generation)

Although tinkering with the visualizations/rough models on http://www.falstad.com/vector3dm/ it all seemed terribly sensitive to precise configuration as to whether they formed inside, outside or on top of the magrid, or not at all.

re: WB - glad I'm not the only one uncertain about state of the art. Looking forward to futher news, obviously :)
Indrek mentioned his question about well formation right here at talk-polywell.

He writes some of the best polywell simulation code available anywhere (except maybe in Los Alamos, NM ;)).

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Re: Why is polywell supposed to be better than cusp confinem

Post by Art Carlson »

pstudier wrote:There used to be a Polywell wiki but it never saw much traffic and now it is gone.
Too bad. I've been thinking that this BB format is sometimes hard to follow, and it's easy for content to get lost (and therefore at best rehashed). A wiki format with the Wikipedia rules against original research relaxed a bit (not too much!) would be a good way to organize knowledge about a technical and immature subject like the polywell.

seedload
Posts: 1062
Joined: Fri Feb 08, 2008 8:16 pm

Post by seedload »

rcain wrote:
rnebel wrote: (Nice Plasma pics from WB7 by the way. Any neutrons yet?)
I think we should stay away from asking Dr. Nebel about specific results that he would be reluctant or unable to talk about. We should respect the fact that he can't answer. The only thing asking will potentially do is make him nervous about sharing what he knows about the theory.

EDIT - answering rcain's next post without cluttering the thread more. Yes, I am pretty sure he would not answer. He has said that much before. I am simply suggesting we not be the kids in the back of the car laying on a constant barrage of "are we there yets".
Last edited by seedload on Sat Jun 28, 2008 3:30 am, edited 1 time in total.

rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

Post by rcain »

seedload wrote:
rcain wrote: (Nice Plasma pics from WB7 by the way. Any neutrons yet?)
I think we should stay away from asking Dr. Nebel about specific results that he would be reluctant or unable to talk about. We should respect the fact that he can't answer. The only thing asking will potentially do is make him nervous about sharing what he knows about the theory.
seedload, i know what you are trying to say, but my question was in parentheses because i didn't actually expect an answer. However, Dr Nebel is not a mouse, Im sure he can answer 'Im not at liberty to divulge'.

what you don't ask for, you don't get.

besides, i'm not even sure that would have been his answer, are you?
Last edited by rcain on Sat Jun 28, 2008 12:59 pm, edited 1 time in total.

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

George Vlases, now there is a name that I have not heard in a long long time.
I too went to UW and studied under him in 1975 but I did not get as far as you guys.
Then I took a job doing mechanical engineering and never looked back.
Is Mad Sciences Northwest still around?
George Vlases said the polywell (without a name) looked like I was trying to make a magnetic monopole, so I dismissed it too.
But it still bugged me from time to time, and then recently I found this project to be an interesting and serious effort along those lines.
I am afraid the tokomak will have space shuttle type problems due to its size and complexity, even if it were technically feasible.

Just what a Polywell is, is hard to pin down because it is a work in progress.
Also different people have different visions of it.
Hopefully your probing will improve the work.
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

jmc
Posts: 427
Joined: Fri Aug 31, 2007 9:16 am
Location: Ireland

Post by jmc »

Hello Art Carlson, I'm currently doing a Phd in tokamaks at Culham.

In a Polywell, you have six current hoops making up dipoles whose moments all point towards the centre, the fieldlines then flow in through the hoops and then leak out through the spaces between the hoop. Because the area inbetween the hoops is smaller than the area the hoops encircle, since all magnetic fieldines must be closed hoops, the field at the line cusp is considerably larger than the field at the point cusps.

I can't claim to be an expert in all forms of fusion and cusp machines are certainly not a full time profession but reading through a number of papers on experiments to do with spindle cusp machines, in all the papers I've come across the field at the line cusps is weaker rather than stronger than the field at the point cusps. This is a terrible waste of magnetic field as it is the line cusps where the dominant loss mechanism is found. I believe there may be a simpler solution to amplifying the field at the line cusp in a simple spindle cusp machine simply by placing two coils of wire with extremely high current following the line cusps while creating and intense localised magnetic field in this region. I have read a number of papers on cusp machines and have failed to find any that have attempted to implement this solution.

Another method of reducing cusp losses is via electrostatic plugging, here a electromagnetic wave generator is placed near the cusps and resonates with the excited particles increasing their perpendicular velocity and pitch angle until it is outside the loss cone in velocity space. A further method of reducing cusp losses was to charge the plates in contact with the plasma to a negative voltage with respect to main vaccuum vessel this can suppress both the energy lost by electrons and reduce the escape radius of the ions from the ion larmor orbit to the electron lamor orbit, as ions are now essential squezing through the equipotentials spikes that link the main plasma to the limited plate that are of order an electron larmor radius. These methods have been already used.

As you can see there are a number of ingenious ways of reducing cusps losses, there are even papers discussing the feasibility of cusp nuclear fusion plants. I'm not convince that cusp machines were hopeless at all,
they certainly achieved far higher beta values than tokamaks and were more MHD stable (against disruptions for example, one large one could break 10 billion dollars worth of ITER in a single shot).

I think tokamaks got far ahead of cusp devices on the Lawson triple product and cusp devices were axed as a result. This is probably the result of lack of funding to the nuclear fusion programme, if you don't much money to spend on heating power and you want to produce enough neutrons to make an impression then ofcourse you go for the machine with the longest confinement time, the tokamak. The history of nuclear fusion for the past 30 years, has been the transition from open field machines to closed field devices, in practice this has meant trading beta (open field ~100%, closed field ~3%) for confinement time, this is a trade off we may live to regret, especially since the tight squeeze in the centre limits the cross-section which the field coils may have increasing their resistance, it also limits the thickness which the blanket can be made forcing the structural material to take more neutron flux. The low beta values which closed field machines have achieved to date also prohibit the use of conventional conductors for any net power machine. This meant that instead you must use super conductors which are 1000's of times more expensive to manufacture (do we really want to sit around with our fingers crossed wished for a steep learning curve in the reduction of the price of superconductors?).

There was some hope originally that on top of the shrinking of the line cusps there would be convergence in the centre, that the electric field at the sheath surrounding the polywell, would all point toward the centre of the device, that all the ions would be ionized (created) in this sheath region and would all be accelerated on trajectories that pass through a similar point in the centre, this would lead to peaked densities at the centre and enhanced fusion power while cusps losses would remain unchanged. This would enhance both fusion power and the triple product by the convergence ratio.

Recent simulations and evidence suggests that achieving this will be difficult in the extreme, due to the equipotentials in the sheath not all pointing spherically inward and it being hard to ionize the atoms at a sufficiently similar energy for annealing to take place, and while it may perhaps not be impossible, it will probably be impossible in the absence of a far larger budget and a great deal more research.

I'm still not entirely sure why mirror like losses are so much worse that wiffle ball losses, as in a wiffle ball, the electrons and ions collide with the field aswell on each pass. Perhaps in mirror-like mode you have the worst of both worlds in that the magnetic moment is forgotten due to the larmor orbit being large in relation to the scale of the device, but the residual magnetic field still reduces the mirror ratio and "funnels" the particles into the cusps whereas in WB mode the flux is expelled from the centre of the device improving the mirror ratio and reducing the funnelling effect.

Just my two pence.

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Re: Why is polywell supposed to be better than cusp confinem

Post by Art Carlson »

Art Carlson wrote:More basically, it looks like the polywell should be classed as high beta cusp confinement. I'm not familiar with the details, but I know that this idea is as old as the hills and was rejected very early. Why is the polywell supposed to be better? Possibly it is the idea of using point cusps instead of line cusps. If that is the case, then I will amplify the reasons I have already indicated for believing that line cusps are unavoidable.
You guys haven't been doing a very good job of prosyletizing me. The closest thing I have to an answer here is that Rick Nebel "suspects" that the old cusp machines operated in a different regime. I've been spending my sleepless nights and long commutes thinking about cusps. I think I at least have a definition, namely field lines that end up in the zero-field region. I'm pretty good at doing 3-d visualizations in my head, but this problem is tough. Still, I'm pretty sure that any way you cut the cake, you will technically have line cusps along the edges of the cube. It seems to be the prevailing view that line cusps will leak like a slit with the width of a few gyroradius. I'm still haven't got a good picture of how cusps confinement works, but this result seems plausible. (It also has the nice property that the losses will be naturally close to ambipolar: The ions see a bigger hole, but they move through it more slowly.) It seems obvious to me that the losses along a line cusp should scale with the local gyroradius. The magnetic field near the corners of the cube shouldn't vary too rapidly, so even if part of the line cusp is closed off, a good fraction of it should still play a role. Up to now I haven't seen any argument from Bussard or anybody else why this should not hold. He just seems to have this picture of point cusps so applies the gyroradius squared scaling. This scaling is presumably correct for point cusps, but what happens to the line cusps? Bussard, by the way, is very clear about the fact that you can forget about any configuration with line cusps.
At this point we should seriously consider the possibility that I have my head up my ass. After all, Rick Nebel has said that Wiffelball confinement scaling (~B^2) has "clearly" been observed. Could somebody point me to the published data, please? I'd like to know if only the B^2 scaling has been observed, or also the dependencies on the other parameters, and also if the absolute value observed is close to that predicted. I'd also like to know if any more direct evidence of wiffle balls has been observed, e.g. has the electron and/or ion loss been localized and seen to be point-like, rather than line-like (on a scale compared to the gyroradius).
Of course, compared to a cusp machine, the polywell may also get some credit for density peaking and recirculation (although I will question both of these effects), but it seems certain that line cusps will be a killer, and it seems pretty sure to me that line cusps are unavoidable.
One more comment, if I may. The size of the coils compared to the faces of the cube is an important parameter. It has been conceded that the concept will not work if the coils touch, and the same has been indirectly conceded if they are too small (because then the fields at the mid-points of the edges of the cube are not large enough to suppress the line-like nature of the cusps). We are only discussing the case that the coils are big but not quite touching. Another loss channel I worry about in that case is the field lines that pass through the coils just a bit on the inside near those mid-points. The field lines probably do not penetrate completely to the zero-field region, but they are roughly radial and the field strength is probably fairly low along the whole length.
One more thing, then I'm let you go. Rick Nebel says that the gradient scale length of the field must be smaller than the (electron?!) gyroradius. I hope he doesn't really mean than. Even assuming resistivity/diffusion is classical, what is the timescale for broadening of a millimeter scale current sheath?

Post Reply