Talk-Polywell.org

Hello all,

We know that radiation is a huge loss mechanism in the polywell. The Larmor formula tells us that every time a charged particle speeds up or slows down it will radiate light. This light is going to be visible, x-ray, ect… and there is going to be allot of it.

We can reflect the light. We may even reflect the x-rays. But this is a moot point. The electrons in the cloud are not dense enough to reabsorb the reflected light. Thus, radiation is still a loss mechanism.

We are trying to make the fusor more efficient. The Polywell reduces the conduction losses from the fusor. Light reflection and reabsorption reduces radiation losses. Maybe we can reduce losses enough to get net power at lower plasma temperatures?


Is there any way to change the wavelength of the light to something the low density electrons could absorb?



http://en.wikipedia.org/wiki/Larmor_formula

http://www.nature.com/nphoton/journal/v ... 1.197.html

http://en.wikipedia.org/wiki/Electromag ... ctron_wave

We know that radiation is a huge loss mechanism in the polywell

Do we? How?

From any photos you have seen of an operating Polywell, that is recombinations creating the glow. The actual power reaction does not produce light.

In Bussard's and Nebel's work, they used PEDs to count neutrons, but also measured recobination light to measure peak confinement and corrosponding currents/voltages to validate Wiffleball. As we understand, the more current devices use much better diagnostics than a "flash of light" for proving wiffleball.


I think you are wasting your time chasing "reflecting light". There are other things more pressing to worry about, like scaling laws.

Light emission is a significant energy loss mechanism. . Though mentioning Lorentz make me think that you are considering cyclotron radiation. I don't know the percentage , but it is low and Bussard was not concerned about it.

Bremsstruhlung radiation is a different matter. It can be very painful or even devastating, at least for high Z fuels like boron. That is why considerations of diluting the boron with excess hydrogen and dynamic distribution of electron speeds in the Polywell are critical elements in profitable aneutronic fusion possibilities. This was, I believe, Rider's primary criticism for p-B11 fusion. The fusion power could never reach the Bremsstruhlung losses. This is well accepted physics, but it does not account for the two above considerations.

VISIBLE light radiation occurs at energy levels associated with orbital changes of electron orbits - as electrons fall to a lower energy orbits, they emit light, often at visible wavelengths. Once you reach plasma where the electron KE is so much that they cannot be considered as orbiting (captured by) nuclei; when they lose energy by curving around a nucleus, or magnetic field (cyclotron radiation), the energy released is more in the UV or X-ray range. Only when these energy drops are small enough is the light emission in the visible spectrum, and this mostly occurs after the recombination events which occur at rates dependent on several factors and accounts for only a tiny (?) fraction of the collisional events. Sometimes the shorter light wavelengths can be converted to longer (less energetic) wavelengths through fluorescence or other mechanisms (?) but this is probably not amenable to significant energy manipulation or recovery.

As mentioned there has been significant efforts to decrease radiation production (Bremsstruhlung). Efforts to recover the energy of the X-rays have not been pursued much in the Polywell community. First off, the KE of the ions in the Polywell is due to electrostatic (electrodynamic if you prefer) forces represented by the potential well, which in turn is due to the injection of high energy electrons. This allows for claimed non Maxwellian thermalized plasma which is a desired component. The Polywell is an accelerator/ amplifier powered by high energy electrical potentials, not collisional heating as in Tokamaks, or any other thermalized approach. Capturing / reflecting x-rays would have to incorporate these thermalizing concerns, so, I suspect, are inappropriate for the Polywell.

Recovering the energy of the x-rays is another matter. Generally, the x-rays give up their energy as heat in the walls, and coolant layers. This heat can be converted to useful (recycled) energy through a steam cycle at perhaps ~ 30% efficiency. This is not enough to overcome the losses as envisioned by Rider, but it brings things closer. Eric Lerner's scheme of converting x-rays to electricity through a bulky photovoltaic process might recover 80-90% of the energy. But this becomes increasingly bulky as machine size increases, and has to consider intervening materials like walls, and coolent layer absorption. It may work for a small device like the Dense Plasma Focus, but is more problematic for machines like the Polywell with magrids, direct conversion grids, etc. In the DPF low x-ray absorption materials like beryllium needs to be used, and there are other tradeoffs that have to be made to work.

In short, I do not see an ability or desire to reflect x-rays, which is a difficult process at best (consider the Chandra x-ray telescope). Recovering the x-ray energy is already at a baseline of ~ 30% if a steam cycle is used. The coolant is necessary for other waste heat anyway , the question is if it is worth running a steam turbine, or just dumping the heat- if direct conversion of the alpha particles delivers sufficient power. Higher conversion efficiency of the x-ray output would be desirable as it could deliver more useful output energy, and perhaps avoid a steam plant. Lerner's patented scheme is the best alternative that I know of, though whether it can be made to work is an open question.

http://www.physicsessays.com/doc/s2005/ ... encies.pdf

Dan Tibbets

A review of my understanding of the Bremsstruhlung problem follows.

Bremsstrulung radiation is mostly dependant on the electron speeds. At the same KE the electron is traveling much faster than the ion (perhaps 60 times as much or more.). This is why the fast on the edge and slow in the center electron dynamic motion is significant. If the ions are more dense near the center, the chances of ion- electron close encounters goes up, but with slower electron speeds in this region, the rate of Bremsstruhlung generating vs fusion events stay the same, but the magnitude of the Bremsstruhlung radiation per event is significantly less. With a non convergent ion distribution, I believe this benificial effect is not present.

Bremsstrulung scales as the 1.75th power of the temperature/ KE. At an energy of 100 KeV, if the Bremsstruhlung is 1 unit/s, then at 200 KeV the Bremsstruhlung is ~ 3 units/s.

Bremsstruhlung also scales as the square of the Z of the ions. Deuterium has a Z of 1, Boron has a Z of 5 (Atomic number).

So, a mixture of 0ne Hydrogen plus one Boron results in a Z of (1 +25)/2 or 13.
With a mixture of 10 H plus one B, the Bremsstruhlung is ((10*1)+25)/11 or ~3.3
One hydrogen (or Deuterium) plus one hydrogen has a Bremsstruhlung of 1.

[EDIT- This may be a misleading comparison. I considered dropping the boron density 10 fold, while maintaining the hydrogen density. You could also maintain the boron density and increase the hydrogen density. In this case fusion rate may increase, while the Bremsstruhlung increases less. I'm unsure if the two scenarios result in the same fusion to Bremsstruhlung ratios.]


This is an obvious way to greatly decrease the Bremsstruhlung losses. I am unsure how it effects the fusion rate. From the Boron's perspective, it should not be a problem at all, but from the hydrogen's perspective, the chances of a Boron impact is 10 times less. I don't know how the final fusion rate is effected, though I speculate that it would be ~ 1/2the rate

Taking those numbers . the dilution would result in a Bremsstruhlung / fusion ratio of ~ 13 / 1 vs 3.3 / 0.5 or 13 vs 6.6. This would double the fusion power gain per unit of Bremsstruhlung losses.

This guestimate would bring Rider's Q estimate to near unity. Any further gain would come from x-ray energy recovery and/ or ion confluence (which might be very significant). From this I take Nebel's statement that confluence (central ion focus) is not needed as applicable to deuterium fusion, but perhaps not to P-B11 fusion.


PS: This is why I like the idea of running at ~ 60 KeV*, instead of ~ 200 KeV. There is a narrow resonance peak in the P-B11 fusion cross section curve at that energy. IF the 'monoenergetic ' range is small enough, the fusion rate may approach that of the D-D fusion reaction, which is ~ an order of magnitude above the baseline curve of P-B11. This is still 2-3 times less than the rate at the target of 200 KeV, but the corresponding Bremsstruhlung is (200/60) ^1.75 = 3.3 ^1.75 or ~ 8 times less. The fusion to Bremsstruhlung ratio is improved ~ 3 fold. Also improved would be the fuel to fusion energy difference, which depending on electrical energy convesion efficiency could add additional modest gains. Thermal loads on the walls would be less, and even the machine size may decrease. The total fusion yield would be less, but the Bremsstruhlung loss would decrease even more, so the net gain per unit volume might even be positive (?).

*M.Simon has a more accurate calculation of the beam- beam KE needed at the P-B11 resonance peak.

Dan Tibbets

Dan,
I am sorry I have not gotten back to you sooner. Time constraints.

I need to go through all this stuff with a fine tooth comb.

But generally we agree: Lots of energy will come off as light. Plan:


1. Identify mechanisms which generate the light. Use simple math to estimate losses. These include:

A. Visible Light – If it comes from electrons in orbitals, than I expect this to be near 0.
B. Light only from acceleration/deceleration. Use Larmor formula. This should be tiny.
C. X-rays. Rider and Spitzer both give equations for estimating this, given plasma parameters.
D. Other Mechanisms….


What I would really love is an estimate of radiation losses (energy/time) as a plasma cloud increases in average temperature (eV). Oh man, that would be awesome. If anyone has this, I would be interested.

Please look up "recombination". Please understand how plasmas really exist.