the problem of pumping electrons into the Polywell

rcain · Post by **rcain** » Sat Nov 03, 2012 2:22 pm

rcain wrote:
KitemanSA wrote: Yes, the WB-8 doesn't have X-Cusps.
ah. maybe i'm confusing with corner Y-cusps.

sorry to be a dunce. remind me under what config/conditions X-cusps appear?

- just answering my own question - 'Octahedron', obviously.

though, btw: i recall Icarus suggesting X-cusps bad idea - since null field would give charged particles clear 'view of metal' - though injecting through each might change that i suppose.

anyway, not currently available - so academic.

happyjack27 · Post by **happyjack27** » Sat Nov 03, 2012 2:27 pm

i believe the solution to this problem might be found through dimensional analysis.

fusion reaction volume is determined by the volume of the wiffleball, whereas electron loss through the wiffleball is determined by the surface area of the wiffleball divided by the field strength (roughly). so if you increase the size of the machine without increasing the magnetic field strength, you increase the ratio of the fusion volume to the rate at which you have to pump in electrons, without increasing the difficulty of injecting electrons.

however, if the electrons have to travel a greater distance to get in the wiffleball - that might reduce injection efficency enough to obviate the scaling advantage.

happyjack27 · Post by **happyjack27** » Sat Nov 03, 2012 2:31 pm

what if you shoot a helical beam of electrons in counter-rotation to the magnetic field?

ladajo · Post by **ladajo** » Sat Nov 03, 2012 2:38 pm

hmmm.

hanelyp · Post by **hanelyp** » Sat Nov 03, 2012 2:49 pm

MSimon wrote:Actually not harder. Just better beam control.

My understanding of the wiffleball magnetic configuration is that an electron orbiting the right magnetic field line has a very good chance of getting through a cusp, while an electron heading towards the cusp not orbiting that field line is reflected. Which means the injector has to have a fairly small effective radius to work efficiently.

As far as recirculation, an electron getting out is already on the right field line, and should get back in if not scattered elsewhere.

The negative ion idea, to my knowledge, has never been tried in an experiment applicable to the polywell. The one liability that comes to mind is the high ion energy entering the well.

CharlesKramer · Post by **CharlesKramer** » Sat Nov 03, 2012 2:55 pm

Nice summary, Rc!

rcain wrote: i suspect they will need to crack the problem with stealth and smart thinking/innovation rather than 'brute force' (keV's?).

Which is another way of saying tweaks and "baby-steps" won't work -- something more fundamental may be needed.

I am fascinated by the idea beryllium is transparent to x-rays (at some frequencies, anyway). For dense plasma focus that might give an incentive to use beryllium in the electrodes. Wouldn't a beryllium Polywell also behave differently than one made of the usual materials?

There are bunches of variables, and when theory fails (the way Tesla claimed he invented) the Thomas Edison brute approach applies: full speed ahead trying everything!

Which doesn't change my disappointment (justified or not) that the old joke that fusion is 10 years away (and always will be) for the moment remains valid.

hanelyp · Post by **hanelyp** » Sat Nov 03, 2012 2:59 pm

happyjack27 wrote:what if you shoot a helical beam of electrons in counter-rotation to the magnetic field?

If fired in a burst that could widen the cusp momentarily as the burst arrives. As a steady stream, it would open the cusp for electrons to escape more easily.

On reflection, if the beam has some randomness in electron momentum, aka heat, that would translate to counter rotation in the magnetic field. From that perspective a hot filament injector sounds good.

happyjack27 · Post by **happyjack27** » Sat Nov 03, 2012 3:25 pm

hanelyp wrote:
happyjack27 wrote:what if you shoot a helical beam of electrons in counter-rotation to the magnetic field?
If fired in a burst that could widen the cusp momentarily as the burst arrives. As a steady stream, it would open the cusp for electrons to escape more easily.

On reflection, if the beam has some randomness in electron momentum, aka heat, that would translate to counter rotation in the magnetic field. From that perspective a hot filament injector sounds good.

"Electron wave packets"

GIThruster · Post by **GIThruster** » Sat Nov 03, 2012 3:37 pm

CharlesKramer wrote:I am fascinated by the idea beryllium is transparent to x-rays (at some frequencies, anyway). For dense plasma focus that might give an incentive to use beryllium in the electrodes. Wouldn't a beryllium Polywell also behave differently than one made of the usual materials?

The Polywell doesn't use materials for its cathode. In the Poly, the cathode is virtual, it's a magnetic mirror, so point in fact, it is transparent to X-rays of all frequencies.

rcain · Post by **rcain** » Sat Nov 03, 2012 4:54 pm

CharlesKramer wrote:Nice summary, Rc!

rcain wrote: i suspect they will need to crack the problem with stealth and smart thinking/innovation rather than 'brute force' (keV's?).
Which is another way of saying tweaks and "baby-steps" won't work -- something more fundamental may be needed.

I am fascinated by the idea beryllium is transparent to x-rays (at some frequencies, anyway). For dense plasma focus that might give an incentive to use beryllium in the electrodes. Wouldn't a beryllium Polywell also behave differently than one made of the usual materials?

There are bunches of variables, and when theory fails (the way Tesla claimed he invented) the Thomas Edison brute approach applies: full speed ahead trying everything!

Which doesn't change my disappointment (justified or not) that the old joke that fusion is 10 years away (and always will be) for the moment remains valid.

just glad to see a decent technical discussion stirred again. been a while.

don't think it can be 'that fundamental' an alteration, otherwise we might end up with 'impossible geometry' - more 'crafty' i was thinking.

very much like the ideas being touted about here. (sort of ion-(ballistic?magnetic?)-diode). but maybe what they are trying already can still work - EMC2-Polywell team are very smart people after all.

re: Beryllium - to my knowledge always planned for FocusFusion DPF approach - very different. cant think of an advantage in Polywell - see also GIThrusters point.

'bunches of variable' - yes - though for me, Polywell at present more resembling jigsaw puzzle with all but last three pieces in place - choice is, to 'hammer em in and make em fit' somehow, or....

'disappointment' - equally shared by most here who have found themselves waiting on that promise for the last 50 years already - at least we are now talking about 10 rather than another/the same 50 again. (albeit i think we might still find ourselves waiting on ITER to deliver it - unless that is one of the alt-fusion approaches (Polywell, FoFu, et al) finally has a lucky break - fingers crossed).

even if we get such a break tomorrow however, will still take decades to come to commercialisation/maturity.

keep the faith

DeltaV · Post by **DeltaV** » Sat Nov 03, 2012 5:03 pm

happyjack27 wrote:
hanelyp wrote:
happyjack27 wrote:what if you shoot a helical beam of electrons in counter-rotation to the magnetic field?
If fired in a burst that could widen the cusp momentarily as the burst arrives. As a steady stream, it would open the cusp for electrons to escape more easily.

On reflection, if the beam has some randomness in electron momentum, aka heat, that would translate to counter rotation in the magnetic field. From that perspective a hot filament injector sounds good.
"Electron wave packets" :-)

Casual musings (translation: not yet run through my 1024 GPU supercomputer and coffee warmer):

-- Modulation of injector current? "Neural/genetic/fuzzy/swarm" search for optimum injector waveform? Maybe used with POPS?
-- Injectors involving antimagnets and/or double- or multi-helix magnets and/or bulk superconductors and/or field-free line formation?

KitemanSA · Post by **KitemanSA** » Sat Nov 03, 2012 7:51 pm

rcain wrote:
rcain wrote:
KitemanSA wrote: Yes, the WB-8 doesn't have X-Cusps.
ah. maybe i'm confusing with corner Y-cusps.

sorry to be a dunce. remind me under what config/conditions X-cusps appear?

- just answering my own question - 'Octahedron', obviously.

though, btw: i recall Icarus suggesting X-cusps bad idea - since null field would give charged particles clear 'view of metal' - though injecting through each might change that i suppose.

anyway, not currently available - so academic.

Lackwing was incorrect. Since there is a hole where the cusp is (see graphic), there is no grounding path.

The X-Cusp is centrally located in this graphic.

D Tibbets · Post by **D Tibbets** » Sun Nov 04, 2012 2:00 am

Some information about electron injection is contained in the 2008 EMC2 patent application.

http://www.askmar.com/Fusion_files/Pate ... 187086.pdf

Pages 11 and especially the bottom of page 14 and top of page 15 gives some hints, though details of EMC2 experience is sparse.

Dan Tibbets

rcain · Post by **rcain** » Sun Nov 04, 2012 3:43 am

KitemanSA wrote:...
... Lackwing was incorrect. Since there is a hole where the cusp is (see graphic), there is no grounding path.
...
The X-Cusp is centrally located in this graphic.

thanks Kite. i wondered what he was talking about.

btw: found your your magfield maps for cusps: viewtopic.php?p=49219&highlight=xcusp#49219

(strange, you and Icarus were bickering on that page too - what is it with you two ?

)

i've been trying to re-acquaint myself with the whole cusp geometry discussion again - naming conventions (eg: funny cusps, corner cusps, etc still a tad confusing).

i referred to our present truncated cube config corner cusp as 'Y-cusp' few posts ago - cause i couldn't think what else to call it - seems a logical name to me, - is it?

all field lines (from corner/point/(Y-) cusp point outward - yes?
(trying to get head around how it all reconnects and what exactly happens in the middle).

happyjack27 · Post by **happyjack27** » Sun Nov 04, 2012 3:40 pm

if you could shoot plasma pinches into it, timed to collapse when they enter the wiffleball...