Giorgio wrote:I suggest everyone gets a good read to Joel Rogers presentation.
He suggest that Q scales at power 3 of the radius instead of Bussard idea that Q scales at the 5th power of the radius. He than proceeds to estimate a reactor size of 150 meters for Break even....
Quite an interesting read.
very interesting, does anyone has anything to say about the scaling not being as good as what we originally thought?
on the other hand, Mattew Carr's report seem to be good news, seems like the WB effect does actually happen.
Rogers admits that his sim does not form a wiffleball, which would improve scaling. He also ignores the "eureka" discovery of rounded casings, simulating instead square cross section magnets. This is a low fidelity simulation.
Giorgio wrote:He suggest that Q scales at power 3 of the radius instead of Bussard idea that Q scales at the 5th power of the radius. He than proceeds to estimate a reactor size of 150 meters for Break even....
150 Meters for break even? A truncated cube with coils roughly 1.3 times larger than a football field? Which must all be contained within a vacuum chamber?
Giorgio wrote:I suggest everyone gets a good read to Joel Rogers presentation.
He suggest that Q scales at power 3 of the radius instead of Bussard idea that Q scales at the 5th power of the radius. He than proceeds to estimate a reactor size of 150 meters for Break even....
Quite an interesting read.
Fusion power in magnetic confinement devices is well accepted as scaling at R^3B^4. Dr. Buzzard suggested an engineering rule of thumb B ~R, and loss scaling in a polywell of of R^2. This gives power scaling of R^7 and Q scaling of R^5. To get only R^3 Q scaling would imply B scaling slower than R or much greater losses.
Update:
Read the paper. I saw no mention of magnetic field scaling or the basis of the R^5 Q proposal. His illustrations of field form don't look much like a wiffleball. The operating conditions in summary are wildly wrong, including 2:1 electron:ion ratio. Models total losses as ion loss, not dominant electron losses. Unable to get ions inside from external guns, possibly on account of magnet shape.
Skipjack wrote:Outch, lets hope that this does not turn out to be true. 150+ meters for break even does not seem very practical
sigh, nothing but bad news lately.
Bad news if WB-x weren't round casings. It's a sim. The actual polywell device is where the definitive news are. And even then, Bussard (as all experimenters) also went thru bad news before breakthrough.
Rick Nebel has commented here in the past (and on MSNBC, comment #21 below) that the full bounce-averaged Fokker-Planck equations are needed to accurately simulate Polywell. I don't think that's what Rogers is using.
R. A. Nebel, Los Alamos, NM
A quick comment on Mr. Katz's statement: I presume that he is referring to the work of Nevins and Rider from the early '90s. That work did not agree with the earlier papers of Bussard, Rosenberg and Krall which concluded that when you looked at the orbit averaged collisionality the system worked fine. Furthermore, the most complete treatise on this was published by Chacon, Barnes, Miley and Knoll in Physics of Plasmas in 2000. This work used the full bounce-averaged Fokker-Planck operator and concluded that IEC systems would indeed work.
So what should one conclude from this? When similar assumptions give you different answers, it means that the physics is sensitive to these effects (i.e. the devil is in the details). The only way to settle that issue is in the laboratory, which is what we intend to do. If we find that the collisionality is a problem, there are ways to innovate around it (see, for instance, Barnes and Nebel in Physics of Plasmas 1998).
Looks like Joel got some of the hypothesis for his work from actual results from experimental data.
Lot of stuff to dig into, and some of it is also contradictory.
Would be nice to have Joel tell us directly what part of the simulation was based on actual data (and what data) and what part is just a theoretical simulation.
Skipjack wrote:Outch, lets hope that this does not turn out to be true. 150+ meters for break even does not seem very practical
sigh, nothing but bad news lately.
It may be that we need some POPS to get to economic breakeven. But as interesting as Joel's models are, we really need to see what we get from WB-8. I suspect EMC2 has a significantly different model.
Rick Nebel has commented here in the past (and on MSNBC, comment #21 below) that the full bounce-averaged Fokker-Planck equations are needed to accurately simulate Polywell.
My understanding is that it's even worse than that -- there are 3D effects like shear flows which are so computationally intensive they apparently can't be modelled at reasonable time and cost in 2010.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...
With one quick read through, I didn't try to dig into his math too much, but. I did not see any discussion about magnetic scaling. B is included in his formula, so maybe it is accounted for, but size scaling without specified B scaling is not mentioned. Bussard felt that B field strength was scalable at ~ 10 X the size scaling (10X increase in radius may be accompanied by 100X increase in the B field). This is reasonable from a simple standpoint as the same aspect ratio magrid internal volume aviable for windings increases as the square of the linier size. Of course the need for cooling, insulation, superconductors,etc. complicates things.
He did mention some effect seen that might be consistent with a Wiffleball.
Also, he mentioned that the upscattered ions hitting the wall was a major loss mechanism. This is counter to Bussard's claims that electron losses dominate. I did not see anything that might be a factor to correct for Bussard's claimed Annealing Effect on the ions so that upscattering is much suppressed. I wonder if this claimed component of the Polywell would account for the large differences in scaling that he found.
One other question I have is his ion gun vs gas puff fed system. He mentions that the secondary electrons from the ionization cascade of the gas puffed neutrals accumulate in the cusps. Does he assume gas puffing through a cusp, or gas puffing into the Wiffleball from an off cusp location?
Giorgio wrote:He suggest that Q scales at power 3 of the radius instead of Bussard idea that Q scales at the 5th power of the radius. He than proceeds to estimate a reactor size of 150 meters for Break even....
150 Meters for break even? A truncated cube with coils roughly 1.3 times larger than a football field? Which must all be contained within a vacuum chamber?
Boy, sounds a bit like the ITER. If this is the case, then Polywell will be no more economical than ITER. Let us hope that Dr.B. had a better grip on this than Rogers.
one of the best answers I give to moon hoax believers is that definitly the filmings were made in vacuum (parabolic trajectory of kicked dust, flag doesnt wave even when astronauts pass nearby).
so I say that to build a large enough studio filled with vacuum, in the 60s, would be more difficult and expensive than sending men to the moon
That's a nice argument. Sort of sad it needs to made at all, but very elegant and accessible.
This 150m issue is why I keep saying the loss scaling numbers for WB-8 are the most critical piece of data right now. As Rick said, without wiffleball confinement we can kiss our butts goodbye.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...
