Talk-Polywell.org

While Googling on non neutral plasmas, I found this Japanese article that I had not seen before.

[EDIT]- Opps, forgot the link, it is:

http://www.jspf.or.jp/JPFRS/PDF/Vol5/jp ... 05-346.pdf


Dan Tibbets

Dan, do you have a link?
I remember reading some stuff from Japan early last year. I wonder if it is the same?

Thanks Dan,
Interesting read. Not seen that one before.
Seems a bit sparse on detail/explanation. Of interest to me:

Performance of spherical magnetic electrostatic
confinement fusion is analyzed. Magnetic field lines in
spherical cusps provide a chaotic scattering system that
improves the ion focusing due to the synergetic effect of
intermittent ion trajectories.

the language is a bit vague, but if i understand it right he is describing what i believe can only be the 'Wiffleball' effect, in terms of a stochastic (synergetic, ...what ever that is...) effect, rather than a purely geometric (Lorentz) effect which i had been hoping to find.

in particular...

The ion motion in this system refers to the
system known lrom nonlinear dynamics as a star-shaped
billiard, it belongs to the stochastic K-systems and many
examples of caustic formation in a stochastic scattering
systems are knorwn [ll,l2].

... is something i am inclinded to read up on, although...

The theoretical study in [4] illustrated efficient
operating regimes of spherical magnetic IEC (Penning
trap) with high ratio of fusion power to ion input power
(Q > 100) identifying beam-like solution for the ion
distribution

seems to be variance with (significantly above) Busards predictions, iirc.

if only we had some real data.

Since it's not the same coil geometry as a Bussard magrid, I don't think it can be directly compared to a Wiffleball. More like a rotationally symmetric fusor grid. All of the cusps here are coaxial rings on a sphere, not points, except for the pair on the symmetry axis. If the fields were high enough, and leakage/losses low enough for fusion, the alphas would exit along cones. Might be useful for some applications, but I doubt the ring cusps would have low-enough leakage for p11B. Just a guess, though.

DeltaV wrote:Since it's not the same coil geometry as a Bussard magrid, I don't think it can be directly compared to a Wiffleball. More like a rotationally symmetric fusor grid. All of the cusps here are coaxial rings on a sphere, not points, except for the pair on the symmetry axis. If the fields were high enough, and leakage/losses low enough for fusion, the alphas would exit along cones. Might be useful for some applications, but I doubt the ring cusps would have low-enough leakage for p11B. Just a guess, though.

yes, i had a bit of a problem visualising their geometry. i take your point. but i think it is pretty similar to a Polywell - its MIEC, quasi spherical, convex everywhere and it has point and line cusps (and ok 'ring' cusps - are those not similar to point cusps?). It lacks Polywell's 'funny' cusps it seems.

All the basics are the same. even down to the b=1 criterion (for WB definition) and the well and shell structures. Seems to me we are looking at very similar animals.

What we dont see here is any treatement of cusp constriction due to WB inflation, postulated by Bussard, though he does brielfy look at cusp plugging and edge effects.

I have no experience in physics.
But the drawing seems to indicate the magnets are solid as in any electrons following the magnetic lines end up smacking into the surface of the magnet.

Or am I misinterpreting the drawing?

vernes wrote: But the drawing seems to indicate the magnets are solid as in any electrons following the magnetic lines end up smacking into the surface of the magnet.

Or am I misinterpreting the drawing?

No, the magnets are all circular current loops, with a common axis, with alternating current directions.

Can someone sketch this without all field lines and stuff? I am just not getting it.

Pick a diameter of a sphere and then arrange a series of planes perpendicular to that diameter. The circles where the planes intersect the sphere define the coil positions. Coil polarity alternates from one to the next, so N is facing N, S is facing S, except at the ends of the diameter. The ends are "point" cusps. The others are circular cusps.

DeltaV wrote:Pick a diameter of a sphere and then arrange a series of planes perpendicular to that diameter. The circles where the planes intersect the sphere define the coil positions. Coil polarity alternates from one to the next, so N is facing N, S is facing S, except at the ends of the diameter. The ends are "point" cusps. The others are circular cusps.

So we are talking about coils along lines of latitude, e.g. at 5N, 15N, 25N, etc and 5S, 15S, 25S, etc. The coils at 15S, 5N, 25N, etc point north, the coils at 25S, 5S, 15N etc point south. There are cusps at 20S, 10S, the equator, 10N, 20N, etc. as well as the poles.

That about right?

That sounds right. That's a higher coil density then his Figs. 1 & 2, which have coils at 0, 30, 60, 90 deg.

Field lines topology for spherical system of 5 ring cusps and point cusps at the axes is shown in Fig. 2. It illustrates the minimum-B configuration that offers a good magnetohydrodynamical stability.

A similar configuration:
http://www.iccworkshops.org/icc2010/upl ... _paper.pdf

DeltaV wrote:A similar configuration:
http://www.iccworkshops.org/icc2010/upl ... _paper.pdf

Is my conclusion correct when I say there are no funny cusps?

DeltaV wrote:A similar configuration:
http://www.iccworkshops.org/icc2010/upl ... _paper.pdf

Very interesting. I'm not sure I understand the two concentric layers of coils. Do they plan to use the outer coils to induce currents in the inner coils (mirror currents)?

They don't mention putting the coils at some electric potential (i.e. implementing a magrid).

I think this coaxial coil geometry does offer one advantage over the polywell. Both have line cusps. For the polywell, the peak field varies along the cusp, so I expect the leakage to be largest where the cusp is furthest from the coils. If the field is just strong enough there, then it's stronger than necessary everywhere else. For the coaxial geometry, the peak field is the same all along a given cusp. Looks more efficient.

vernes wrote:

DeltaV wrote:A similar configuration:
http://www.iccworkshops.org/icc2010/upl ... _paper.pdf

Is my conclusion correct when I say there are no funny cusps?

That seems so.