Elmore- Tuck- Watson fusor was from 1959 I believe. I think it was a theoretical work. Other than myself playing with demo fusors ( I was surprised their picture of the plasma, was similar to mine- I thought the dim glow meant very poor confinement, but perhaps not), and the Polywell, the only experimental work I know of is this:
http://www.plasma.ee.kansai-u.ac.jp/iec ... s/Tuft.pdf
I have seen a web site where they were planning to magnetically shioeld the peripheral anode grid, but without the geometry of the Polywell that results in the Wiffleball effect.
Note that in the above link they report formation of a potential well though the deep well only persists for a few microseconds (pulsed mode?). In the Polywell, older works had similar appearing potential well graphs- time scales. The electron loss rate is just too ridiculously large to stretch out the lifetime of the deep potential well past a few transits (perhaps ~ few hundred passes (guess- grid a few cm wide, electron speed ~ 700- 800 million cm/second, means that there would be ~ 100-500 passes in a few microseconds. That would be good as non shielded central cathode grids in fusors get in the way after ~ 10-20 passes. I'm guessing the larger anode grid size results in greater transparency (wire surface area makes up a smaller percentage of the total sphere surface area) as the sphere enlarges.
Magnetically shielding these anode grid wires should, and apparently do increase the effective transparency of the grid.
The later work with claimed improving electron lifetimes (perhaps up to several hundred microseconds) apparently permits maintainance of deep potential wells at correspondingly smaller electron input current. Throw in recirculation and the ion density advantages obtainable with the Wiffleball that allows for the smaller electron current, and viola, the net effects that are claimed to make the Polywell viable.
I still don't know how the thermalization times of the electrons are extended from a few microseconds, to in excess of the electron confinement time of a few hundred microseconds. I don't know if there may be some 'annealing' process similar to ions (except it occurs near the center of the spherical geometry) or if the preferential escape of the developing high energy tail of the electrons can contribute significantly.
The high energy tail- upscattered electrons leave earlier through a cusp (same number of passes, but each pass is faster), and is not recirculated because their outward energy/ speed is greater than the positive charge on the magrid. The upscattered electrons travel the same distance as the non recirculated non up scattered electrons, but the mean free path of the up scattered electrons is also longer, so the net effect, is that they will effectively leave earlier, without as much of their abhorrent energy being transferred to other electrons.
It seems a large extrapolation for this to result in 1-2 orders of magnitude increase in thermalization times, and I don't know how the math would work out. Also, keep in mind that the Coulomb collision crossection and thus the inverse of the thermalization time falls rapidly as the energy/ speed increases. This may be why ~ 80 KeV is mentioned for a D-D Polywell, rather than ~ 15 KeV where the fusion crossection change is the steepest- ie: the fusion gain/ input power is at it's best at ~ 15 KeV, but at the higher energy the fusion power (at constant size and B) is higher and possibly most importantly the thermalization times are longer also.
Lets see- I'm guessing the above experiments were done at ~ 10,000 Volts (they had brief potential wells of ~ 7 ,000 volts) The Coulomb collision rate would be ~ 1/60th as fast (I'm using the rough estimate that coulomb collision crossection changes (decreases) as the square of the temperature. That means that at ~ 10,000-12,000 accelerating volts, if the thermalizing time was ~ 5 microseconds, then at ~ 80,000 V the thermalization time would be ~ 300 microseconds. Perhaps the removal of the upscattered tail in the Polywell does not need to do so much after all. It may be more important from a Bremsstrulung perspective, especially with P-B11 fuel!
Dan Tibbets