Page 1 of 1

Has anyone built a working Polywell Fusion device

Posted: Sun Mar 04, 2012 9:31 pm
by CH007
Hi,

I have done a fair amount of reading and wondered if anyone has built a working polywell fusor? I appreciate some have done 0.5s pulses but has anyone done minutes long running and if not why not?

I am fairly sure in my limited knowledge that the coils would rapidly overheat from plasma. Is this the reason for very short runs? What is being done to improve these problems? I am just trying to figure out how something is going to be scaled up if it can't be run in small form?

Please don't take this post as negative the theory looks very good for the Polywell I am just wishing to understand more and stimulate ideas and thought .

Colin

Posted: Mon Mar 05, 2012 12:47 pm
by KitemanSA
Hi Colin,

The biggest, baddest Polywell in the known world is running at EMC2 ( WB-8 ) and it is a pulsed machine as far as we know.

The main reason for the pulse is not the plasma but the current needed to make the magnetic field strong enough. that current heats up the copper windings and if run too long, the windings will start to melt. Until a Polywell is made with superconducting coils, it seems likely that they will always be pulsed machines.

Dr. Bussard made statements to the effect that even though they were "pulsed" to our viewpoint, to the veiwpoint of the particles involved, they were steady-state. I take that to mean that the length of the pulse was longer than the lifetime of the typical particle in the machine.

Just for info's sake, the plasma should never touch the coils except for what is hoped to be minor leakage. That was the great step forward between WB5 and WB5. In WB6, the coil cans were made field conformal so as not to present metal surfaces to the plasma More specifically, the electrons.

Posted: Mon Mar 05, 2012 1:10 pm
by CH007
That makes more sense than the issues I had imagined :)

I would agree in that case that the magnets could well be pulsed at rapid enough intervals to maintain confinement.

Does anyone know or is there info on what strength field is effective for confinement?

Posted: Mon Mar 05, 2012 6:43 pm
by ladajo
I would agree in that case that the magnets could well be pulsed at rapid enough intervals to maintain confinement.
Some may argue that could be called "POPS". Others may even add that there was no accident that Dr. Nebel, and Dr. Parks got on the Polywell Project.

Posted: Mon Mar 05, 2012 6:47 pm
by ladajo
Does anyone know or is there info on what strength field is effective for confinement?
That is a big question. It is driven by many factors. Things to consider are primarily coil diameter, intended well depths, and of course available power. For your sake, we know in the public domain that EMC2 is runnig the WB8 device at .8T. What is not public is the diameter and other specifics. We can assume it is at least a 30cm machine (based on WB6/7), and that they are driving to 10KV (or more) based on again, WB6 & 7.
Read Dr. Bussard's IAC Paper, and he gives a decent run up to what field is needed and why. It is available via the EMC2 website.

Posted: Mon Mar 05, 2012 6:57 pm
by KitemanSA
CH007 wrote: I would agree in that case that the magnets could well be pulsed at rapid enough intervals to maintain confinement.
Superconducting magnets won't need to be pulsed, their field can be maintained at the high strength desired without the coils heating from resistance. The only heating would be the limited plasma interaction, and tat whould be FAIRLY easy to handle.

Posted: Tue Mar 20, 2012 2:11 am
by D Tibbets
I'm not certain that Ohmic (or inductive if you prefer) heating of the magnetic coils is the limiting factor on the 'pulse' length of the Polywell test machines. WB4 was water cooled and I understand that magnetic fields were maintained for multiple seconds. Even in WB6 there is mention of the magnetic coil current being on for perhaps several seconds. For the gas puff machines like WB4(?) and WB6 the time limiting factor was directly related to the buildup of excess neutral gasses outside the machine that lead to very rapid arcing discharge of the capacitor banks. The ability to inject enough neutral gas towards the magrid so that ionization could proceed at rates necessary for Wiffleball inflation, also led to quick buildup of neutral gas throughout the vacuum chamber (neutral gas is not trapped by the Polywell) as the efficiency of ionization was not 100%. This is an inherent problem of smaller radii machines and the shortcomings of the relatively cheap gas puffers used. Two work arounds is a much tighter regulation of the gas puffers and larger radii for the magrid. The ionization process is time dependent, and greater size means the neutral gas molecules spend more time inside of the magrid before completing a passage. Ion guns is an alternate method of introducing ions into the magrid, but they have their own problems.

IE: the resistive heating of the copper windings in the tests was trivial over a period of a few milliseconds. The accumulation of gasses inside the vacuum chamber within a few milliseconds (or less) was significant. The test may have started with a pressure of perhaps 10^-7 atm. Once the gas was puffed and the non ionized fraction diffused throughout the vacuum chamber (Faraday cage in WB6) the pressure climbed past ~ 5 * 10^-6 atm. This pressure at ~ 12,000 volts was sufficient for Paschin Arc breakdown to begin. This destroyed the potential well while discharging the capacitor banks, thus effectively ending the test. The voltage, the pressure, magnetic shielding of the magrid, lack of shielding outside the magrid, sharp corners, etc all play a role.

Bussard, etel was budget constrained so they worked with less than ideal equipment. Adaptations / improvements available for WB7 and subsequent machines is unknown. This is why I suspect effective 'pulse' times may have already far exceeded those of WB 6. ~ 0.25 milliseconds for WB6 vs 10s of milliseconds for WB7.1 and/ or WB8.

The important physics time scales may have been a few microseconds for electrons, thus Bussard's claim that even the brief "steady state" time of ~ 250 microseconds in WB6 was adequate to prove the system. That may have been an optimistic assessment in some regards. I think it was sufficient to prove confinement times and dynamics for the electrons. I'm uncertain if it was sufficient to fully answer ion thermalization issues, as the average ion lifetime till fusion or escape may be in the 10s of milliseconds. Also, the thermalization issues of recirculating electrons would be better demonstrated (not just predicted) on multiple millisecond time scales.

Dan Tibbets

Posted: Tue Mar 20, 2012 8:47 pm
by Roger
Wasn't there a comment about someone seeing a 6 ft dewer @EMC2 a while back?

Of course many of us thought about MSimons LN2 cooled thoughts.

SO if WB-8 is LN2 cooled, EMC2 could have a quicker cool down between runs, or could be seeing runs of seconds vs ms. I've always wondered about that.

Posted: Tue Mar 20, 2012 9:53 pm
by KitemanSA
The WB8 was supposed to have a field strength of .8T which was ~ 8 times that of WB6&7. One way to help achieve that is to run the magnet at LN temperatures where the resistivity is much lower and the current can be much higher.

Posted: Wed Mar 21, 2012 6:19 am
by Roger
KitemanSA wrote:The WB8 was supposed to have a field strength of .8T which was ~ 8 times that of WB6&7. One way to help achieve that is to run the magnet at LN temperatures where the resistivity is much lower and the current can be much higher.
Yes, thanks for reminding me.

If the coils of WB-8 are twice the size of WB-6&7 then one could fit 4 times the copper wire.. no?

SO WB-8 might be heating (ballpark) just as much as WB-6? Vs heating way more....