Here is an idea I had while thinking about the Polywell system (please forgive my rough paint.net skills):
It's the same basic design, except the green lines in the drawing represent a container system that completely encloses the cusp lines. The container system should be wound with wire so that it loops the plasma back around from the faces to the corners. It's a continuation of the magnetic field created by the MaGrid, so that it is purposefully directed back around and back into the grid. The MaGrid remains positively charged but additionally a high negative potential is also applied at the narrow most portions for additional acceleration. The plasma would end up forming a clover leaf pattern in the machine. I'm sorry if it's unclear what is going on but it is difficult to draw. The enhanced case forms a manifold that has 4 chambers going into the faces and 3 going into the corners.
I was under the impression that even without the tokamak addition, the plasma flow of the polywell already followed the clover leaf pattern. The tokamak addition simply forces the recirculating ions to concentrate into beams before reentering the main chamber.
The tokamak portion would be a continuous winding from face to corner. The toroid would be wound for the ring faces as per a normal polywell but then the winding would continue from the toroid in a tube/manifold that loops back to the corner. Each face would have four of these tubes extending to each corner, so at some point the four tubes would have to open up into one chamber and connect to the toroid. The same goes for the corners except there would be three tubes into one chamber. It may even be better to use four wires to wind the toroid but as you get to the outside use each individual wire to continue to each corner, braiding and winding some of it as you go into the tubes/manifolds.
If there is a problem with the winding of the tokamak portion's magnetic field disrupting the sphere shaped field in the middle as I think you are saying, it should be possible to place ferrous materials around portions of the outside of the tokamak portion in strategic locations to minimize the problem.
Sorry about possible thread necromancing here, but I had a bit of a brain-fart yesterday:
To what extent is it possible to combine the 'best of both worlds' for Tokamak and Polywell-type devices into a more feasible concept? More specific, I was thinking of two ideas:
* Applying the high trapping factor of a Polywell to a Tokamak device, e.g. installing a MaGrid along the Tokamak walls which would confine the plasma more than a simple set of confinement magnets would. However, I am not sure whether an IEC trapping method would work in a Tokamak topology.
* The other way round, applying Tokamak tech to Polywells: Heating the plasma in the core by means of (AC) magnets around the machine or microwave beams. However, this would instantly thermalize the whole lot inside the core and possibly render the entire Polywell concept useless.
You are forgetting a very important consideration with the Polywell. There is no magnetic recirculation of ions. Once an ion escapes through a cusp it is lost, in part due to the positive charge on the magrids. Placing negative potentials somewhere in the cusps might recirculate some of these ions, but only at the cost of increased electron losses, and that is already the dominate energy loss mechanism. The whole purpose of the Polywell is to magnetically confine/ recirculate electrons and use these confined electrons to form an electrostatic virtual cathode which then confines the ions. Magnetic confinement of ions (even with cusp pinching/ recirculation) is not very good. Cross field transport through the magnetic fields is a major loss mechanism for ions. That is why Tokamaks need to be so large. Electrons are much more forgiving in this regard. It is the trick of exploiting this electron magnetic confinement advantage and using the confinement of electrons electrostatically to confine the ions. The claimed advantages of accelerating the ions to the center and delaying thermalization are seperate issues.
Finally, in a working Polywell it is anticipated that the electrostatic confinement of the ions is already greater than their expected lifetime befor fusion. There is no need to confine them or recirculate them longer. Also, the 'few' upscattered ions that do escape the electrostatic confinement and thus have a chance to escape the much less effective magnetic confinement are evil and you want to get rid of them anyway.
The open grid design in WB6 and subsequent improvements confine and recirculate the electrons already. Other interacting surfaces outside the magrid need to be kept to a minimum, at least close to the magrids) In the 2008 patent application, this cusp management efficiency for electrons (ions are unimportant in this regard) might be good enough that the electron losses through the cusps may optimistically be similar to the electron losses due to crossfield transport (IE: it is as good as it can get). That is why efforts to modify, or eliminate the exposed nubs (connections) between the magrids may be very important.
