looking for an equation, where is the main FAQ for polywell?

[ohiovr]

Well I guess that settles it then. Unless this were to be run in pulsed operation superconducting magnets are essential. We can use the Stefan-Boltzmann law to find radiation heat transfer between surfaces in a vacuum. The electromagnets would have to be cooled by liquid helium, and around that a dewar of liquid nitrogen would be needed, and around that a chilled water jacket dewar would run and around that and then a pressure vessel for containing steam pressure water would be needed. Its complicated but do able.

http://hyperphysics.phy-astr.gsu.edu/hb ... an.html#c3

How much energy does it cost to run a superconducting electromagnet? I know little about them. Since there is no ohmic heating there is no loss of energy from wire resistance and extremely high currents are possible with low voltages. I guess a big part of it depends on the machines that condense liquid helium and nitrogen. What else is there to consider?

An 8.3 tesla magnet with a 5 meter radius polywell could gross 1.03 gigawatts of power.

Lets say the surface area of all 6 electromagnets will be 773 square meters and the emissivity of our dewars will be only .03. The inner chilled water jacket at 363.15 K radiating heat to the next layer, the liquid nitrogen layer at 77 K will be absorbing 7,382 watts of heat. Then the liquid nitrogen layer will be transferring only 46 watts of heat if there are not considering physical contacts.

I'm just trying for some ball park figures. Pardon me if I seem a bit optimistic. I only have a laymans understanding of some of this and no understanding in other areas.

[ohiovr]

.... such as how does the electromagnets actually accelerate the ions? And why they don't thermalize into a Maxwell–Boltzmann distribution.

[KitemanSA]

.... such as how does the electromagnets actually accelerate the ions?

They don't (at least not much). They accelerate and confine the electrons which create a potential well which accelerates the ions.

And why they don't thermalize into a Maxwell–Boltzmann distribution.

Annealing. (This one is fun!) Thermalization at the potential well edge (zero kinetic) means everything is random at zero so when they fall back into the well, they are mono-energetic (IIUIC). So, thermalization prevents thermalization.

[D Tibbets]

.... such as how does the electromagnets actually accelerate the ions?

They don't (at least not much). They accelerate and confine the electrons which create a potential well which accelerates the ions.

And why they don't thermalize into a Maxwell–Boltzmann distribution.

Annealing. (This one is fun!) Thermalization at the potential well edge (zero kinetic) means everything is random at zero so when they fall back into the well, they are mono-energetic (IIUIC). So, thermalization prevents thermalization.

My understanding is that the magnetic fields do not significantly accelerate the ions or electrons. Dr Bussard mentioned two possible ways of acellerating the electrons in WB6. First have a high potential on the electron guns so that they are injected through a cusp at high energy (eg 12,000 eV). The magrid is at ground. The other way (actually used) is to eject the electrons from the electron guns at low energy. They are then acellerated to high energy as they approach the positive potential on the magrid.

In addition to providing the high potential to acellerate new electrons, I'm guessing the positive charge on the magrid helps to recirculate the cusp loss electrons, and also might help to encourage any escaped ions to quickly fly to the vacuum vessel wall. This might reduce the tendency of the low energy escaped ions to collect near the external surfaces of the magrid (if grounded)and thereby increasing the chances for arching.


Dan Tibbets

[ohiovr]

so the electrons are shot in and the electromagnets confine them. then ions are shot in and they are accelerated to the center due to their attraction to the electrons. Why don't they slow down and neutralize? Excess electrons?

[ohiovr]

and thanks all for trying to explain this to me for the nth time.

[MSimon]

so the electrons are shot in and the electromagnets confine them. then ions are shot in and they are accelerated to the center due to their attraction to the electrons. Why don't they slow down and neutralize? Excess electrons?

"Low" density and excess electrons.

[Art Carlson]

Well I guess that settles it then. Unless this were to be run in pulsed operation superconducting magnets are essential.

I'd leave the question open until some detailed engineering is done. For reasons of both power balance and power density, contrary to widespread fantasies, this thing will produce neutrons. To keep them out of your superconductor you will need at least 50 cm of shielding. If you have the space, then go for it.

On the other hand, a polywell is supposed to use its field much more efficiently than a tokamak (beta = 1 vs. beta = 10%), so the ratio of ohmic heating power in the coils to fusion power might be small enough to live with.

Does pulsing make a difference? Ohmic heat production scales with I^2*f, where f is the duty factor. If you are limited by heat transfer out of the coils and run with f=0.01, then you can make your current and therefore your field 10 times higher. This allows 100 times the pressure and 10^4 times the reactivity, so the average fusion power increases by a factor of 100. In addition, if the cusp (or other) losses are reduced by a higher field (as most of us seem to believe), then your power balance with respect to those losses can also be improved. Pulsed operation appears to offer some big advantages. It introduces some problems, but they are "merely" engineering problems. Among these are higher forces, material fatigue through cycling, and eddy currents.

[Art Carlson]

And why they don't thermalize into a Maxwell–Boltzmann distribution.

I know it sounds crazy. Probably because it is. But the proponents want to operate in a mode where the loss time is shorter than the thermalization time. (Or at least where the electron residence time is shorter than the thermalization time, although not the isotropization time.) If I correctly remember the results of my own calculation, that implies that 99.99% of the energy of departing electrons has to be recovered.

[ohiovr]

50 cm of shielding

How is the field going to penetrate half a meter of shielding? Unobtainium with vacuum permeability? Yikes!

[MSimon]

I'd leave the question open until some detailed engineering is done. For reasons of both power balance and power density, contrary to widespread fantasies, this thing will produce neutrons. To keep them out of your superconductor you will need at least 50 cm of shielding. If you have the space, then go for it.

My admittedly BOE calculation says that 5 cm of H2O is enough to thermalize the neutrons and a few mm of B10 is enough to absorb them. Assuming X-Rays are not a problem for the coils. Of course you have the added heat load from B10 neutron absorption - but as Arts says. That is just engineering.

[MSimon]

And why they don't thermalize into a Maxwell–Boltzmann distribution.

I know it sounds crazy. Probably because it is. But the proponents want to operate in a mode where the loss time is shorter than the thermalization time. (Or at least where the electron residence time is shorter than the thermalization time, although not the isotropization time.) If I correctly remember the results of my own calculation, that implies that 99.99% of the energy of departing electrons has to be recovered.

Doesn't that correspond well with Dr. B saying electron losses must be kept below about 1E-4 or 1E-5 or so to make the device work?

[TallDave]

Electron losses are separate from what's necessary to have a non-Maxwellian distribution of electrons.

OTOH, Polywells should preferentially suck down the coldest electrons, so I'm not sure I buy Art's calculation. Still, this is another situation where we really need to know more, and this is one of the objections that can't be refuted with WB-7 results (as maxwellianization won't affect smaller machines much).

Maybe Rick Nebel will stop by and enlighten us at some point. I'm sure it's something they've considered.

[ohiovr]

Polywell FAQ part I

What is a maGrid?
The maGrid is short for magnetic grid. It is the circular electromagnets that contain the electrons and ions in the potential well. All polywell magrids so far have had 6 electromagnets. Future designs may employ more than 6 (how many? Why is more better?)

Why is polywell claimed to have a non Maxwellian distribution of ion energies?
According to Dr. Bussard's lecture at google video (22:25) all the ions in the potential well are of the same energy distribution. This is unlike a Maxwellian or thermal distribution of energies which vary widely across a broad spectrum for a given temperature. A Maxwellian distribution looks like an elongated bell curve while polywell's distribution is a spike. Only a fraction of the ions in the most optimal Maxwellian distribution are optimal for fusion. Where as the polywell distribution (correct thing to say?) all ions can fit the best velocity profile for optimal fusion conditions. And this is because of some process called annealing. And the loss time is shorter than the thermalization time. And??

Ok ions in the polywell are of a single energy level theoretically most of the time. What energy level is optimum for T+D reactions, D+D reactions and P+11B reactions and the corresponding coil drive voltages?
...P+11B reaction requires a 100KV coil drive voltage but since the 11B ion has a charge of 5 it will receive an energy of 500KV at the bottom of the potential well (source google video lecture 23:21).

How many nuclei collisions are required for one fusion event in the polywell?

Say boron 11 proton fusion were achieved in polywell. Those ions with several MeV energy levels will zoom out of the well and will be harnessed with deceleration grids or they will hit the apparatus. Many millions of collisions per square centimeter will be occurring every second on the electromagnets sputtering the surface away and contaminating the potential well with metalic ions. How will this be dealt with? What about collisions on the surface of the deceleration grid wires and electron guns?
?

How will fresh fuel be injected and the spent thermonuclear exhaust be removed? Fuel is injected via ion guns. It is exhausted by using? What about the helium nucli flying around inside of the polywell, what will be done with them?
?

What is the ion and electron density of the inside of the polywell? Is it a gradient? What about outside the polywell where ions are decelerated?
?

What is a deceleration grid and is it a polywell only invention? I mean does it have any prior use before polywell?
?

What does the WB in WB 6 mean?
Whiffle Ball

What is the WB-100 I some times hear mentioned?

How are ions accelerated towards the center of the polywell?
The ions are accelerated by the electric forces of the electrons circulating in the center of the polywell.

How are they confined?
They are confined by the maGrid (magnetic grid).

What are cusps and cusp losses?
?

What is a funny cusp?
?

WB-6 produced a .1 miliwatt output for a 7,500,000 watt input. If bigger is better and can lead to net power, just how big does it have to be for x amount of gross power? Is there a way to find out how much power output and input from an equation due to size?
This is still a topic of debate.

Could the polywell have applications besides fusion power (in case this has no use for power generation)? Xray source perhaps?
?

[Art Carlson]

Electron losses are separate from what's necessary to have a non-Maxwellian distribution of electrons.

Not according to Rick Nebel. His plan to maintain a non-Maxwellian distribution is to extract electrons before they have a chance to (fully) equilibrate. What's yours?

OTOH, Polywells should preferentially suck down the coldest electrons, so I'm not sure I buy Art's calculation.

If you are thinking of losses parallel to the field, then electrons that slow down will fill up the bottom of the potential well (near the magrid). That is something to worry about. If you are thinking of cross-field losses, they are either due to collisions - which will thermalize the electrons at least as fast as they move them - or due to turbulent fluctuations. Turbulence will probably also thermalize the velocities, but even if it doesn't, it is not usually something to wish for.