Remind me - why 10T field?

[MSimon]

Reminds me of the cult of Einstein. People are still quoting the dude.

Not sure why it reminds you so. Einstein's work is reproducible, and *provable*, from first principles. Polywell is a mishmash of hopeful claims - a bit like a religion, in fact.....

Uh. There is no way to know if Nebel's work is reproducible since as far as I know only one underfunded amateur in Brooklyn is trying to reproduce it on any scale.

But since "everyone" knows it can't work there is obviously no point in trying. A waste.

[D Tibbets]

TallD:

Are you being purposely evasive or is that just how it comes across? To avoid any possibility for further circumlocution, let's try and pin down one simple point here.

Q: Do we agree or not that a Polywell of a certain radius (say 1m coils) will have a maximum ion density that it can operate at?

(Accepting the assumption that the density across the entire radius is most likely some smooth functional relationship.)

Indeed there is a maximum limit. Based on what is mentioned in the patent application about the modest B fields necessary to maintain dynamic pressures of 10^22 to 10^23 particles / M3, the magnetic field strengths obtainable should not be limiting. Machine size and thermalization issues may force some limiting compromise.
But, from what Bussard, and Nebel have said, the Wiffleball confinement factor or gain is the critical issue. Due to the limit on allowable density outside the magrids before Pashin arcing occurs ( ~ 5 X 10^-6 atmospheres (~ X 10^19 to 10^20 particles / M^3)). This sets a density limit similar to Tokamaks, or I assume any machine where some mobile charge carrier separation from ground is required*. The Wiffleball confinement, if real, allows a density differential of ~ several thousand above this limit, and this is what limits any edge density within the Wiffleball to ~ 10^22 or 10^23/ M^3. The picture becomes more murky with confluence and effective densities above this in a core smaller than the Wiffleball radius. Nebel mentioned several orders of magnitude were possible. That would set a hard upper limit for the core density at ~ 10^25 charged particles / M^3 within a small confluent core.

* Why does the Wiffleball allow increase4d densities? I speculate it is due to multiple factors. Smooth surfaces so there is not charge build up. Gauss law properties such that the internal plasma does not see the positive charge on the magrid. Magnetic shielding against electrons being more significant than magnetic shielding against ions. This is a major difference from tokamaks. The electrons will not "slowly arc" to the magnetically insulated surfaces within the magrid as rapidly as ions due to their smaller gyroradii. The ions being contained electrostaticallly by the excess electrons allows for this dichotomy that I do not believe would be possible in neutral plasmas. Even with these advantages, I believe the Wiffleball confinement could not tolerate edge densities much above 10^22 to 10^23 charged particles / M^3 before arcing would occur within the machine. It all boils down to what tricks you can play to delay the onset of Pashin arcing as the density increases in the Wiffleball confined region.

Dan Tibbets

[TallDave]

Dan--

darn, now I'm curious -- I'll have to go back and read that LANL article about the guy who saved them six months or whatever with his arcing solution.

Anyways... plugging in B= .8 to the WB equation, I get 1.59E+20 density in WB-8 (someone please correct me if I did something horribly wrong there; I only varied B), so either WB-8 will never reach beta=1 or some theories of Polywell come off the shelf.

In fact, I get 2.48E+18 for WB-6/7 at .1T, which exceeds that 1E+17 mentioned earlier.

[D Tibbets]

Dan--

darn, now I'm curious -- I'll have to go back and read that LANL article about the guy who saved them six months or whatever with his arcing solution.

Anyways... plugging in B= .8 to the WB equation, I get 1.59E+20 density in WB-8 (someone please correct me if I did something horribly wrong there; I only varied B), so either WB-8 will never reach beta=1 or some theories of Polywell come off the shelf.

In fact, I get 2.48E+18 for WB-6/7 at .1T, which exceeds that 1E+17 mentioned earlier.

I think the ~10^ 18 particle/ M^3 ( = 10^-7 Torr) density represents the the starting density within the vacuum chamber for WB6. Using a conservative Wiffleball trapping factor of ~ 1000, would result in a internal density at the Wiffleball edge of ~ 10^21 particles / M^3 when the machine was operating a Beta=1. By increasing the B field strength to 8X would increase the density to ~ 6* 10^22 particles / M^3.
Already at the densities claimed for a larger breakeven machine.

But, this calculation assumes that the field strength at the cusps is what you are using. If the radius doesn't change it would be a direct comparison, but you have to correct for size. If WB 8 has 0.8 Tesla magnets and twice the diameter of WB6 then the strength of the B field at the point face cusps would be 8 * (1/r^2) because of the inverse square law.
That would be : 8 * ((1/ 30cm/ 15cm)^2) = 8 * (1/2^2) = 2 times the B field strength at the point cusps. As density increases as the square of this cusp field strength the density is 4X over WB6.
Lets see... Fusion scales as density squared and radius cubed. So WB 8 would produce ~ 16 * 8 = 128 times the fusion rate. this is less than the ~ 400X number often quoted, but that calculation assumed no change in size.

If you apply this same analysis to a 'Demo' machine with a magnetic field strength (at the coils) of 10 Tesla and a radius of 150 cm. then the magnetic field strengths at the cusps would be essentially unchanged compared to WB6.

This requires some more head scratching.

PS: The inverse square law applies to unopposed electromagnetic force spreading out. With two or more fields facing each other I'm not sure how the gradients would work out.

Dan Tibbets

[TallDave]

I think the ~10^ 18 particle/ M^3 ( = 10^-7 Torr) density represents the the starting density within the vacuum chamber for WB6.

Hmmm? I got 2.48E+18 applying Rick's WB beta equation, which should be the interior WB edge density.

If WB 8 has 0.8 Tesla magnets and twice the diameter of WB6 then the strength of the B field at the point face cusps would be 8 * (1/r^2) because of the inverse square law.
That would be : 8 * ((1/ 30cm/ 15cm)^2) = 8 * (1/2^2) = 2 times the B field strength at the point cusps. As density increases as the square of this cusp field strength the density is 4X over WB6

Hrm. Interesting idea, but I've never seen power given as B^4/r^2. If that were true, PW power would rise at something more like r^5 than r^7.

Here's Rick on the cusps:

Loss fraction = (summation (pi*rl**2))/(4*pi*R**2) where rl is the electron gyroradius and R is the coil radius. The summation is a summation over each of the point cusps. If you calculate rl from one of the coil faces, then there are "effectively" ~ 10 point cusps (fields are larger in the corners than the faces). The factor that your observed confinement exceeds this model is then lumped together as the cusp recycle factor.

I found this interesting too, if anyone was wondering mu0 comes from:

http://en.wikipedia.org/wiki/Vacuum_permeability

[D Tibbets]

I think the ~10^ 18 particle/ M^3 ( = 10^-7 Torr) density represents the the starting density within the vacuum chamber for WB6.

Hmmm? I got 2.48E+18 applying Rick's WB beta equation, which should be the interior WB edge density.


Hrm. Interesting idea, but I've never seen power given as B^4/r^2. If that were true, PW power would rise at something more like r^5 than r^7.

Here's Rick on the cusps:

Loss fraction = (summation (pi*rl**2))/(4*pi*R**2) where rl is the electron gyroradius and R is the coil radius. The summation is a summation over each of the point cusps. If you calculate rl from one of the coil faces, then there are "effectively" ~ 10 point cusps (fields are larger in the corners than the faces). The factor that your observed confinement exceeds this model is then lumped together as the cusp recycle factor.

I'm still confused about the stated density. If the starting pressure/ density within the vacuum chamber is ~ 10^-7 Torr or ~ 10^18 particles/M^3 in WB 6, then any Wiffleball trapping beyond this has to increase the numbers. The only thing that comes to mind, is that they anticipate maintaining a lower vacuum outside the machine with post WB6 machines. Perhaps if the vacuum vessel pressure is pushed to ~ 10^-9 Torr (which was the original target), the numbers are more consistent. Of course that would bring up vacuum pumping issues. Though if confinement improves, there will be less ions, electrons, and neutrals (better neutral conversion to ions) leaking out of the machine. So, there would be less relative vacuum pumping capacity needed. Ideally, the vacuum pumping capacity will be based largely on the rate of fusion ions leaving the system.

Concerning the Nebel formula you provided. If cusp confinement varies as the square of the gyroradius, and the gyroradius scales directly with magnetic field strength, then this would cancel out the r^2 (r**2 (?)) in the denominator, unless that is already part of the B^4 scaling calculation.

Dan Tibbets

[MSimon]

Vacuum pumping to 1E-9 refers to the chamber with no flow. With flow it goes up to 1E-6. i.e. out gassing impurity atoms are 1 in 1,000.

Then density goes up from there in the reaction space.

[TallDave]

Concerning the Nebel formula you provided. If cusp confinement varies as the square of the gyroradius, and the gyroradius scales directly with magnetic field strength, then this would cancel out the r^2 (r**2 (?)) in the denominator, unless that is already part of the B^4 scaling calculation.

Ah, thank you, that might explain it. I was circling around that gyroradius idea but hadn't tried to dig up support yet. Too much fun this weekend.

[WizWom]

There is no way to know if Nebel's work is reproducible since as far as I know only one underfunded amateur in Brooklyn is trying to reproduce it on any scale.

Its a machine - if you go by "needing another machine to be built" then ITER results won't be reproducible.

It is engineering - and you don't need to reproduce that, you just need to do it, and show that it does the job. Once you do the job, the physics is only important to eggheads.

[MSimon]

I dunno,

To me reproduce means to replicate. For solid replication you would want a second machine - although repeating the same results on a given machine might do.

In any case the costs of repeating the WB-8 results with a similar machine are not very high as these things go.

[MSimon]

I dunno,

To me reproduce means to replicate. For solid replication you would want a second machine - although repeating the same results on a given machine might do.

In any case the costs of repeating the WB-8 results with a similar machine are not very high as these things go.

Oh, it's gets even better, two previous eclipse observations were inconclusive when Einstein made the challenge to astronomer's worldwide.
So even if the formulae is not necessarily 100% exact, there should be some roadsigns that point at solution along the way?

What would be the roadsigns of data or results that would point to Richard Hull's lucky donkey?

Donkey hoof prints on his skull?

All the rest is speculation at this point.

[chrismb]

What would be the roadsigns of data or results that would point to Richard Hull's lucky donkey?

Stability, and scalability.

Stability = continuous neutron output in DD operation that can be turned up or down at will by controlling device net current [e.g. ion injection flux].
Scalability = detectable neutron output at 1kV CoM collision operating regime [2kV drive IEC, or 250eV thermal], in DD operation.