I wanted to throw out a question and see what some general consensus/opinion on the Polywell fusion power system is.
After I first read about the precursor (Farnsworth fuser) and was initially interested greatly as I had similiar thoughts before really researching fusion.
Personally, I'm not a big fan of the ITER project being pursued currently... but like us all I have hope that it might work.
So, what are your thoughts, comments, criticism?
Also, why havn't particle accelerators as a method of fusion been pursued?
ZapperZ
Sep25-08, 07:19 PM
Also, why havn't particle accelerators as a method of fusion been pursued?
thanks for replying to my posts and guiding me like this. I tried doing a search on "polywell" but couldn't find any threads.
I'm a second year aerospace engineering student with a weird imagination... (don't we all have one?)
You?
Astronuc
Sep25-08, 08:09 PM
The polywell/fusor systems have been around a long time. They basically use an electrostatic field to accelerate ions (radially) into the center of the chamber, as opposed to magnetic compression. They have drawbacks. Neutron howizters using d+t fusion (in which deuterons are accelerated in electrostatic fields into tritiated targets) were used as neutron sources, but they have very low reaction rates, i.e. low power density.
Accelerators would be inefficient (due to scattering) since one is only looking for 1-300 keV range for fusion temperatures. On the other hand, neutral beam injectors are one method of fueling and adding energy to a magnetically confined plasma. In neutral beam injectors, atoms are stripped and the nuclei accelerated electrostatically, and then neutralized (recombined with electrons) so that they pass through the magnetic field, and the energetic neutrals collide with the plasma and become ionized while dispersing their energy. Obviously the neutral beam energy must be greater than the nominal plasma energy (temperature).
The plasma confinement system is constrained by the strength of materials, which limits the plasma pressure, which limits the plasma temperature and density by virtue of pressure being proportional to nkT, where n is the particle density, k = Boltzmann's constant, T = ion/electron temperature. Also, the plasma temperature limited in order to reduce energy losses due to cyclotron and brehmsstruhlung radiation.
mheslep
Sep27-08, 01:35 AM
R. Nebel of Los Alamos has taken over the Polywell work this past summer. He's been blogging a little about his motivation. Apparently he believes Polywell is worth the time because of its advantages as practical reactor vs a tokamak, mainly in power density. He asserts the Polywell has the potential to magnetically confine electrons per
n*kBolt*Te = B**2/(2*mu0)
That is, where the electron pressure is in balance the with magnetic bottle pressure.
Ions are in turn confined electrostatically by the well set up by the energetic electrons so that the plasma is quasineutral. If Te=10^4eV, B=10T, density is 2.5e22/m**3 which is ~ 10^4 greater than ITER's planned density. Nebel is not counting on any convergence of beam ions in the center here, just a confinement slightly out of LTE. Then there are other advantages like no external neutral beam or RF heating required as for a tokamak, in a Polywell the drive energy is supplied by the electron source. The big unknown appears to be the scale of the losses from the cusps in the quasi spherical magnetic bottle.
TallDave
Sep30-08, 01:01 PM
Lots of relevant discussion at Talk Polywell, in the Theory forum.
Art Carlson came by and laid out a lot of objections. Nebel has replied that many of them don't apply in a non-LTE environment, and explained why he believes the others can be solved.
Here's where Carlson originally weighed in before the dicussion moved to T-P.
The relatively recent POPS research that might give an IEC scheme orders of magnitude more ion density also bodes well for Polywell's chances of eventually producing an economically practical fusion reactor, something the ITER path probably can't achieve (due to the poor power density) until competing fission/fossil fuels run out (at least 1,000 years away for the former).
TallDave
Sep30-08, 01:17 PM
The big unknown appears to be the scale of the losses from the cusps in the quasi spherical magnetic bottle.
This is a common point of confusion (and contention). The electron "losses" from the cusps are not losses per se, because the machine recirculates them (they do matter, because the machine must have a 1000:1 ratio inside the "bottle" versus outside). According to Bussard, the only actual losses are electron losses to unshielded areas of the machine, and cross-field transport.
What's really unknown is what losses look like at 1.5m vs 15 cm. Bussard believed they scaled as r^2, but the history of problems with predicting transport scaling in tokamaks suggest this prediction is fraught with uncertainty.
mheslep
Sep30-08, 02:08 PM
...
The relatively recent POPS research that might give an IEC scheme orders of magnitude more ion density also bodes well for Polywell's chances...Yes I see comments that Nebel still thinks POPs viable, but my impression from the series of POPs paper authored by his team was no, POPs won't work in this case. The last one in series Space charge neutralization in inertial electrostatic confinement plasmas, Nebel et al, Physics of Plasmas 14, 2007
said in essence that while a steady state plasma might achieve compression in the core, oscillating plasmas (ie POPs) can not achieve high density through periodic compression:
...Results indicate that there are limits on the amount of
compression that can be achieved by oscillating plasmas while simultaneously maintaining space charge neutralization and parabolic background potential.The problem has to do with the spherical geometry chosen by a Polywell/fusor which has a work around
What cannot be done is to simultaneously completely mitigate space charge effects with large compressions and maintain a pure harmonic-oscillator potential [POPS]in a sphere....
...It is also possible to get around the geometrical convergence limits altogether by using a cylindrical system rather than a spherical one.
But then they are back to the mirror confinement problem of the 70's.
Please comment if TD if you have other information.
mheslep
Sep30-08, 02:12 PM
Here's where Carlson originally weighed in before the dicussion moved to T-P.
As to POPS, not sure where that stands at the moment. IIRC the original POPS paper stated 10^4 compression was possible, and I'm not sure what the referenced limit worked out to. Even a considerably smaller improvement would be helpful I suppose.
Nebel has said ion focus probably isn't that important in a Polywell. Maybe that obviates the need for a "parabolic" potential well as mentioned above and allows POPS to work better.
Nebel recently commented on another use for resonance: removing ash.
The idea is pretty simple. You oscillate the amplitude of the virtual cathode at the resonant transit frequency for the desired ion species you want to pump energy into. It selects species by q/m. For a harmonic oscillator potential, it's pretty trivial. It's described by Mathieu equations (driven harmonic oscillators). We demonstrated this experimentally (that's what's in the Phys Rev Lett.). This will be more involved for potentials other than harmonic oscillators, but it is probably doable.
Hopefully we'll get to find out how all this works in a larger machine. The funders are reportedly reviewing the WB-7 results.