Talk-Polywell.org

Hope this works.

Used Femm 4.2 for this, an idea thats been rattling round for a day or two, I like the way the flux lines loop straight back to the centre. Not sure what would happen when the top and bottom coils go on though.

Hello to everyone, by the way.

Indeed an interesting 2D layout.

The cusps turn into reentry paths, and the central space seems to remain nearly spherical. The problems are: is there any concave surfaces towards the core?
The field lines come very close to the surfaces in areas. This may allow for to much ExB drift, though if electrostatic charge reverses most of the electrons before they reach this relatively exposed surfaces (are they more eposed or represent larger areas than the funny cusps in the truncated cube Polywell? How would this effect the upscattered electrons that you actually want to get rid of?

Dan Tibbets

One modification that might be interesting and possibly improve the standoff distance of the cusp bordering field lines from the non corner magnets would be to decrease the strength of the corner magnets somewhat (a slight to modest amount?). This should change the geometry of the field lines (magnetic isobars) so that they do not approach the non corner magnets quite so close.

Dan Tibbets

yes, interesting. it looks from the field lines that this would have move grid losses than the more traditional symmetric arrangement.

Almost looks like closed field lines in 2D, until you trace the field lines facing the plasma and see they intercept the magnet casings.

Increased the spacing slightly between coils, still a few more things to try. I'd really like a 3d version to play with, I think the top and bottom coils will have an 'interesting effect', if i'm thinking right.

That helps the B isobars representing the cusps approaching magnet surfaces, but decreases the sphericity of the central space (more spiky). Adjusting the spacing of the corner- non corner magnets and/ or adjusting the strengthsof both may help the central facing fields. In any case you will have concave fields towards the center- a big no- no. That these are hiding past the point where the recirculation might occur may provide sufficient forgiveness. I don't know if this would be a problem.

I down loaded the software. If you have the input parameters available, submission would be appreciated both for educational and playing purposes.

Dan Tibbets