Has Wiffleball Been Created Ever?

Point out news stories, on the net or in mainstream media, related to polywell fusion.

Moderators: tonybarry, MSimon

kcdodd
Posts: 722
Joined: Tue Jun 03, 2008 3:36 am
Location: Austin, TX

Post by kcdodd »

KitemanSA wrote:
kcdodd wrote: That comes from seeing the results of the solvers Indrek and I made. I am not sure if one could come up with a coils where it would be in equilibrium though.
The graphic that had been at your site is now not available. Can you bring it back?
Hm, I'll have to search for it. Indrek's site is still up though. Here you can see the large variance of magnetic field strength for a spherical wiffleball. http://www.mare.ee/indrek/ephi/invwb/

All my solver did was find solution of shape for uniform magnetic field pressure on the surface.
Carter

93143
Posts: 1142
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

Robthebob wrote:
93143 wrote:If a particular ion were at high energy at the edge, it would see the wiffleball effect just like the electrons
I do not understand this. I'm under the impression that the WB effect is just the result of the ballooning effect of electrons at the core of the machine, since the ballooning effect changes the field structure to allow particles to come in easy but leave harder, electrons will effectively relocate themselves to the core.

If you inject fuel into the system when the WB effect is already in place, I can see the ions might go into the core and have a harder time leaving, but I dont see how this have to do with high energy ions at the edge, or something... I'm missing something here.
All I meant is that while the ions are nominally electrostatically confined, the energy spread means that some of them will encounter the magnetic field at the edge of the wiffleball at a substantially nonzero energy. The magnetic field will turn a fuel ion around just as effectively as it will turn an electron around, though it takes more distance because the ions are heavier. So the magnetic confinement works on ions too; just not as well because the ions see larger cusps. In fact, it works on fusion products, though most of the alphas are so high-energy (ie: large gyroradius) that they only get about 1e3 passes before they find a cusp.
Joseph Chikva wrote:
93143 wrote:
Joseph Chikva wrote:Leakage through the cusps will be observed at beta<1 too.
Of course. With Polywell, beta=1 means the point at which cusp losses are at a minimum, due to the wiffleball effect.
May be. But please provide corresponding reference. Better if that would be the experimental evidence.
Valencia paper, page 9.

You know what the state of play is regarding experimental evidence. I'd point you to the detailed WB-6 final report (not the three-page summary), but it's been withdrawn; it was apparently posted in error...
I am mechanical engineer by my first education.
Good for you. My first two degrees were in mechanical engineering.

Apropos of that, let me point out to all present that when I say "instability", I am using in in the technical dynamic-systems sense. In this context, I mean a mode in which a disturbance grows without bound, until the plasma configuration that gave rise to it is destroyed.
So:
Let’s consider the pressure vessel with several small enough holes allowing some quantity of being there gas to leak and also with several holes for injection of higher pressure gas inside. Let’s inside flow exceeds gas leakage. So, pressure in vessel should be constantly increased. Let’s call beta=1 when gas pressure inside vessel is equal to strength limit of vessel. What will happen with vessel if pressure will reach that limit? Can leakage hole preserve vessel from explosion?
Ever heard of the leak-before-burst design criterion?

Polywell (if I've understood it correctly) is the ultimate leak-before-burst magnetic configuration. The plasma can keep pushing harder and harder on the field, and it will just strengthen to keep pace right up until the surface of the diamagnetic volume passes the magnets themselves.

Of course, by that point, the cusps will have long since been forced wide open by plasma pressure, and losses will be extreme. I don't imagine that the fuel and electron feeds and power supply could drive the system anywhere near such a condition.
And what is beta>1 Not a case when pressure inside exceeds the strength limit.
Nope. In the case of Polywell, it's when the cusps start to open up again.
What feedback you talk about?
After the minimum-loss point (denoted beta=1), plasma pressure starts to open the cusps, and losses increase, so that a slight excess of plasma supply will not result in a plasma disruption, but rather an increase in losses to offset the excess supply.

I did say at one point "if I've understood this correctly". It is possible, as far as I know, that the system exhibits catastrophic hysteresis or some kind of severe nonlinearity that renders it impossible to recover the wiffleball once the cusps have blown out. This would make your objection to the beta=1 run condition correct, though still in my opinion somewhat pedantic. I suspect instinctively that this is not the case, but like everyone here, I await further experimental evidence.
...the geometry makes it immune to certain classes of macroinstability.
Please provide examples from fusion history.
Mirror machines. They aren't subject to some of the classic macroinstabilities that tokamaks have to guard against.
As I also heard these reasonings about advantages of convex fields (minimum beta machines) but we have clear examples that minimum B stellarators have not any advantage vs. TOKAMAK and also suffers macro-instabilities.
A stellarator is not a convex-field machine (this is trivially obvious, since it has no cusps, whereas any machine with uniformly convex fields must as a matter of geometrical necessity have cusps). It does, however, eliminate the risk of current disruptions, since it doesn't need a plasma current to maintain conditional MHD stability like a tokamak does.
kcdodd wrote:
KitemanSA wrote:
kcdodd wrote:That comes from seeing the results of the solvers Indrek and I made. I am not sure if one could come up with a coils where it would be in equilibrium though.
The graphic that had been at your site is now not available. Can you bring it back?
Hm, I'll have to search for it. Indrek's site is still up though. Here you can see the large variance of magnetic field strength for a spherical wiffleball. http://www.mare.ee/indrek/ephi/invwb/

All my solver did was find solution of shape for uniform magnetic field pressure on the surface.
Why does the plasma need to be a sphere? Instead of trying to configure the magnets to produce a perfectly spherical wiffleball, why not let the wiffleball decide what shape it wants to be?

If you try to hang a stick horizontally from a string attached to one end, it will not maintain a steady state. The best you can get is pulse mode. Remove the word "horizontally" and Bob's your uncle.

Joseph Chikva
Posts: 2039
Joined: Sat Apr 02, 2011 4:30 am

Post by Joseph Chikva »

93143 wrote:
Joseph Chikva wrote:
93143 wrote: Of course. With Polywell, beta=1 means the point at which cusp losses are at a minimum, due to the wiffleball effect.
May be. But please provide corresponding reference. Better if that would be the experimental evidence.
Valencia paper, page 9.

You know what the state of play is regarding experimental evidence. I'd point you to the detailed WB-6 final report (not the three-page summary), but it's been withdrawn; it was apparently posted in error.
I am not sure that this is experimental evidence. As:
When beta = unity is achieved, it is possible to greatly increase trapped electron density by modest increase in B field strength, for given current drive. At this condition, the electrons inside the quasi-sphere “see“ small exit holes on the B cusp axes, whose size is 1.5-2 times their gyro radius at that energy and field strength.
Analyses show that this factor can readily reach values of many tens of thousands, thus provides the best means of achieving high electron densities inside the machine relative to those outside the magnetic coils, with minimal injection current drive.
These two quotes are claims explaining inventions sense and not fact ascertaining of existence of wiffleball effect. How e.g. “size of exit holes” was measured if that is experimental data?
93143 wrote:
...the geometry makes it immune to certain classes of macroinstability.
Please provide examples from fusion history.
Mirror machines. They aren't subject to some of the classic macroinstabilities that tokamaks have to guard against.
Now I see you talk about “immune to some of the classic macroinstabilities”. Agreed if so. But also would like to add that even one of non-classic macroinstability is observed, that will not allow Polywell to run at beta=1.
The example of one of classical macroinstability occurring in mirror macine with convex field you can see here: http://ieeexplore.ieee.org/xpl/login.js ... %3D1061020
A small superconducting model of yin-yang coils for the Mirror Fusion Test Facility has exhibited a magnetoelastic buckling instability.
93143 wrote:
As I also heard these reasonings about advantages of convex fields (minimum beta machines) but we have clear examples that minimum B stellarators have not any advantage vs. TOKAMAK and also suffers macro-instabilities.
A stellarator is not a convex-field machine (this is trivially obvious, since it has no cusps, whereas any machine with uniformly convex fields must as a matter of geometrical necessity have cusps). It does, however, eliminate the risk of current disruptions, since it doesn't need a plasma current to maintain conditional MHD stability like a tokamak does.
Both cusp machines (convex field machines) and Stellarators provide “minimum B principle”. Convex lines can be described graphically and analytically using gradient. I think that gradiental expression of Stellarator’s field would be very similar to mirror machines’ field. And despite to satisfuction of mentioned priciple, Stellarators had not any advantages against TOKAMAKs not satisfying the same principle. So, for effective confinement of plasma this principle is less important. This is proved experimentally and this is all I would like to say.
Last edited by Joseph Chikva on Tue Jun 26, 2012 5:20 am, edited 1 time in total.

93143
Posts: 1142
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

Joseph Chikva wrote:I am not sure that this is experimental evidence.
I never said it was. I believe it's based on simulations. The experimental evidence is virtually all proprietary. The currently unavailable WB-6 final report had some details, but as mentioned before the diagnostics were a bit primitive...

Joseph Chikva
Posts: 2039
Joined: Sat Apr 02, 2011 4:30 am

Post by Joseph Chikva »

93143 wrote:I never said it was. I believe it's based on simulations. The experimental evidence is virtually all proprietary. The currently unavailable WB-6 final report had some details, but as mentioned before the diagnostics were a bit primitive...
Ok, thanks.

303
Posts: 114
Joined: Thu Jan 12, 2012 11:18 am

Post by 303 »

my eyes glazed over about two paragraphs into the abstract, but perhaps this might be of interest to you plasma fanboys

http://pop.aip.org/resource/1/phpaen/v1 ... horized=no

KitemanSA
Posts: 6188
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

kcdodd wrote:
KitemanSA wrote:
kcdodd wrote: That comes from seeing the results of the solvers Indrek and I made. I am not sure if one could come up with a coils where it would be in equilibrium though.
The graphic that had been at your site is now not available. Can you bring it back?
Hm, I'll have to search for it. Indrek's site is still up though. Here you can see the large variance of magnetic field strength for a spherical wiffleball. http://www.mare.ee/indrek/ephi/invwb/

All my solver did was find solution of shape for uniform magnetic field pressure on the surface.
Do you recall whether Indrek's calls had a virtual magnet representing the plasma effect?

kcdodd
Posts: 722
Joined: Tue Jun 03, 2008 3:36 am
Location: Austin, TX

Post by kcdodd »

93143 wrote:
kcdodd wrote:
KitemanSA wrote: The graphic that had been at your site is now not available. Can you bring it back?
Hm, I'll have to search for it. Indrek's site is still up though. Here you can see the large variance of magnetic field strength for a spherical wiffleball. http://www.mare.ee/indrek/ephi/invwb/

All my solver did was find solution of shape for uniform magnetic field pressure on the surface.
Why does the plasma need to be a sphere? Instead of trying to configure the magnets to produce a perfectly spherical wiffleball, why not let the wiffleball decide what shape it wants to be?
I don't think it would form into a sphere because a sphere would not be in equilibrium. I have uploaded the plots from my solver's iterations. It starts at a sphere solution as Indrek did, but high pressures go in and lower pressures come out to converge to the equilibrium solution where the pressure is uniformly equal to the "plasma pressure" (which I just arbitrarily chose at the time).

http://www.flickr.com/photos/81467326@N05/
Do you recall whether Indrek's calls had a virtual magnet representing the plasma effect?
Yes he used the image coil method to solve for where the magnetic field is tangent to the sphere at all points. I used current segments on the surface of a mesh to solve it (or current loops, I can't remember which one ended up working).
Carter

KitemanSA
Posts: 6188
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

Yup, that was them. Thanks!

kcdodd
Posts: 722
Joined: Tue Jun 03, 2008 3:36 am
Location: Austin, TX

Post by kcdodd »

I also found this among some plots. Looks like I did an electron trajectory in a spherical wiffleball. I'm sorry I don't remember much else about it.

http://www.flickr.com/photos/81467326@N05/7467696278/
Carter

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Post by D Tibbets »

kcdodd wrote:
93143 wrote:
kcdodd wrote:Hm, I'll have to search for it. Indrek's site is still up though. Here you can see the large variance of magnetic field strength for a spherical wiffleball. http://www.mare.ee/indrek/ephi/invwb/

All my solver did was find solution of shape for uniform magnetic field pressure on the surface.
Why does the plasma need to be a sphere? Instead of trying to configure the magnets to produce a perfectly spherical wiffleball, why not let the wiffleball decide what shape it wants to be?
I don't think it would form into a sphere because a sphere would not be in equilibrium. I have uploaded the plots from my solver's iterations. It starts at a sphere solution as Indrek did, but high pressures go in and lower pressures come out to converge to the equilibrium solution where the pressure is uniformly equal to the "plasma pressure" (which I just arbitrarily chose at the time).

http://www.flickr.com/photos/81467326@N05/
Do you recall whether Indrek's calls had a virtual magnet representing the plasma effect?
Yes he used the image coil method to solve for where the magnetic field is tangent to the sphere at all points. I used current segments on the surface of a mesh to solve it (or current loops, I can't remember which one ended up working).
It is innapropiate to call the Polywell plasma a sphere. That is impossible. Even if you could have a true monople surrounding magnet there would be two cusps/ projections from an idealized sphere. The plasma is often called quasipherical. It has irregularities, but these are symetrical and have an axis pointing towards the center. You can approach a sphere but never achieve it.

The reason a near sphere is desired (a spiky sphere if you like) is because a concentration of ions towards the center is possible. This has multiple desired consequences. Taken to extreams this would result in what has benn called a black hole effect. The fuel ions enter the coil but because of high density they would never leave. There would be alot of coulomb collisions, but befor they could travel far enough to leave the core, a fusion collision would also occur. The fuel ion never leaves, only the resultant fusion ion/ products. This is interesting in several regards, but of course this is an extream that is very improbable to obtain, but any movement towards has advantages for energy density (size of machine) and associated loss scaling, thermalization issues, etc.

Dan Tibbets
To error is human... and I'm very human.

kcdodd
Posts: 722
Joined: Tue Jun 03, 2008 3:36 am
Location: Austin, TX

Post by kcdodd »

Well, wiffleball is Bussard's description. A ball with straws coming out of it. One of the primary purposes was to increase electron confinement by making the cusps "smaller". But, I don't think that will really happen since everything gets pushed out, including the cusps, not just the "sides". This is the main difference between a sphere model, and what will probably actually happen. You notice the sides are still concave, not convex like a sphere. Huge difference even when talking just about single particle motion.
Carter

KitemanSA
Posts: 6188
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

While it is true the the cubeoctahedral WB6 was not quite spherical, being MORE spherical was one of Dr. B's stated R&D goals. He seemed to think that a more spherical unit would provide as much as 5 times the output as a similar sized cubeoctahedral unit. My thought is that more spherical means more focused in the middle thus higher density. Since the density factor is squared...

krenshala
Posts: 914
Joined: Wed Jul 16, 2008 4:20 pm
Location: Austin, TX, NorAm, Sol III

Post by krenshala »

KitemanSA wrote:While it is true the the cubeoctahedral WB6 was not quite spherical, being MORE spherical was one of Dr. B's stated R&D goals. He seemed to think that a more spherical unit would provide as much as 5 times the output as a similar sized cubeoctahedral unit. My thought is that more spherical means more focused in the middle thus higher density. Since the density factor is squared...
Thus the desire for the dodecahedral version.

KitemanSA
Posts: 6188
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

krenshala wrote:
KitemanSA wrote:While it is true the the cubeoctahedral WB6 was not quite spherical, being MORE spherical was one of Dr. B's stated R&D goals. He seemed to think that a more spherical unit would provide as much as 5 times the output as a similar sized cubeoctahedral unit. My thought is that more spherical means more focused in the middle thus higher density. Since the density factor is squared...
Thus the desire for the dodecahedral version.
. Or the square plan-form version of the C-O from which he also expected improved sphericity. Which leads to my BOWED square plan-form variant. There are a number of potential paths to improved sphericity.

Post Reply