What is the best material for the rings?

[mattman]

Hello Guys,


What is the best material for the rings? I am sure someone else has looked at this before, and feedback would be appreciated.


The Ideal material would have:

1.Have a magnetic permeability close to vacuum or 1.256E-6 [Henrys/Meter].

2. Have a high melting point.

3. Have a low, or no, cross section at the following energies:

- 64,000 eV if the machine is setup to fuse Deuterium and Tritium.

- 1,250,000/1,750,000 eV if the machine is to fuse Deuterium and Deuterium.

- 550,000 eV if the machine is to fuse Boron.

4. Have a low price.

5. Have an extremely low electrical conductivity, to guard against arching.



===================
For durability against neutrons, there is an equation from todays’ nuclear industry to estimate how fast a material breaks down. It is measured in displacements per atom. The equation is below, with an example calculation of the DT reaction hitting a 15 feet diameter reactor wall made of 316 stainless steel:

Displacements per atom = Neutron Fluence * Cross Section of wall (@ Energy) * time.

0.08 dpa per year = [~2.5E+16 neutrons/ second*cm^2] * [~0.1 barn @ 64,000 KeV] * [1E-24 cm^2/1 barn] * [31,536,000 seconds/year]


We know that above 1 dpa embrittlement becomes a problem in fission reactors. The neutrons from the Polywell should be at an energy roughly ten times higher than those in Fission reactors, because there is no moderator.
====================



Here are some suggested ring materials:

- Aluminum
- 316 Stainless Steel
- Molybdenum
- Neodymium
- Tungsten-Carbide

Are there any other suggestions? Did someone already look at this problem?

[D Tibbets]

Your energy levels are inappropriate for D-D fusion. And the energy for D-T may be a little high. The energy for P-B11 is perhaps appropriate. In any case, why do you want the surface material of the magrid to be transparent to x-rays? Assuming the copper coils or superconductors absorb x rays to produce unwanted heat, it would be best to stop the x-rays before they reached this depth. The cooling problem (at least for the magnetic coils is more easily managed in this case.
In another thread I explained why drive energies much above ~ 100 KeV are actually losing ground because of bremsstrulung and the shallower slope of the D-D crossection graph.

A point I did not make was a concept presented by Dr Bussard. This is to utilize D,T, and He3 fuels. I believe he called it the D-D one half catalyzed process. D-D fuses producing T +Photon. or He3 +Neutron.
These fusion produced ions are too energetic to be confined by the potential well and soon escape without contributing to heating the plasma much (this is not an ignition machine). These escaped tritium and He3 ions, after cooling (hitting the walls) and neutralizing, are collected from the vacuum pump exhaust, processed and fed back into the reactor along with additional deuterium. For each two D-D reactions you can now get one D-T and one D-He3 reaction. This produces considerably more energy, For these three fuel combinations, ~ 80 KeV is a good compromise for utilizing the different crossections. If the D-D reaction turns out to fall a little short of breakeven, this cycle may allow for profitable power production- especially the D-T reaction. The He3 may actually be more valuable if separated and used for other purposes. This has been discussed in the past. A D-D reactor, perhaps with tritium utilization might make for a good general purpose power plant for grid use. The He3 could be harvested, and saved for dedicated D-He3 reactors ( for sea ships or space ships). This would eliminate the need for very problematical and expensive harvesting of He3 on the Moon. This can be a backup option for 'near' neutron free reactors if P-B11 proves unfeasible (but the D-D, and D-He3 reactor is feasible).

Dan Tibbets

[hanelyp]

Since the entire surface of the toroids will be at a uniform electric potential, an insulating material here wouldn't help much with arcing. But a poor conductor shouldn't hurt here. Graphite, silicon carbide, or boron are all worth considering.

An insulating cover for the standoffs would help, producing a more uniform potential gradient between the magrid and the surrounding.

[MSimon]

An insulating cover for the standoffs would help, producing a more uniform potential gradient between the magrid and the surrounding.

A slightly conductive coating (megohms per square) might serve better.

[D Tibbets]

As MSimon has pointed out in the past, letting the standoff surfaces 'float'- iE have poor conduction to ground and separated from the positive charge on the magrid proper allows them to acquire a net negative charge as electrons hit them. This will then inhibit further electron impacts. This may have some advantages concerning arcing, provided that electrons are the dominate charge carrier. I suspect it will not make much difference on recirculation as any electrons that reach the vicinity of the back of the magrid coils where the standoffs attach, are upscattered and have decelerated as much as they will and have little chance of reentering the Wiffleball. This is in contrast to nubs that bridge between magrid coils only a few mm past the midline of the coil (where the recirculation potential starts exerting it's effect).
The magrids themselves are insulated by the magnetic fields. Additional insulation on the surface would have two effects. First it might impead the potential well forming magrid potential. Depending on which portions of the magrid had this physical insulation, the more important effect is that the protion of electrons that penetrate the magnetic field through ExB diffusion, ground on the positive magrid. This costs some in terms of maintaining the potential on the magrids. But, with physical insulation these electrons would accumulate on this insulating layer, effectively shielding the positive potential of the magrid. In the first case you can compensate by feeding more current to the magrid surface. In the second situation it would be difficult to overcome this harmful shielding effect. The situation is different for the standoffs as they are not serving as the electron accelerator.
Because of the magnetic shielding I do not think arcing is a dominate concern for the magrid itself ( so long as acute angles are avoided) as arcing in other areas would occur first. The bigger problem would be arcing between external structures that do not have this magnetic shielding- things like electron guns, ion guns, vacuum vessel wall, standoffs, direct conversion grids, etc. Having an induced (?) negative charge on the standoffs could help some (electrostatic insulation instead of magnetic insulation) except care would be needed at the junction where the standoffs connect to the magrid coils. Here there would have to be a physical insulator, and arcing across this presumably short distance could be a problem. That this junction may still be well within the magnetic insulation region of the magrid would help.

Dan Tibbets

[D Tibbets]

My previous post brings up another point. If the potential well is ~ 80% of the drive potential on the magrid- 10,000 V/ 12,000 V in WB6, then a non upscatered electron would see a 2,000 voltage differential decelerating the electron as soon as it passed the midplane of the magrid. Better to say that it's 10 KeV KE would see a decellerating force ~ 1.2 times it's KE. If there was no difference between the potential well voltage and the drive/ recirculating voltage then the electron would never turn around until it had traveled an almost infinate distance- with field lines this would allow for reentry through another cusp if enough space is provided. This would be fine, except it negates the thermalization inhibiting effects of recirculation through a single cusp. WB 5 may have addressed this with its electron repellers in the major cusps, but as Bussard pointed out, this caused a fatal ion loss through the same cusps. How far would the electron travel before stopping and reversing.? A mm, a cm, 10 cm...?. An up scattered electron with an energy of 11,9999999999 eV would stop eventually, but at a very long distance from the magrid. In fact if it didn't hit something it would probably follow a magnetic field line back through another cusp. The problem with this is that it would then retain it's upscattered energy and feed it back into the general Wiffleball contained plasma. This is bad as it causes faster thermalization, Bussard mentioned requirements nessisary to prevent this.
I wonder what these standoff distance considerations for perscribed recirculation properties is.

Dan Tibbets

[93143]

The electron KE is equal to the potential difference between the magrid and the emitters, not to the well depth.

[MSimon]

The purpose of the high resistance coating of the insulators is to prevent localized hot spots. Which have been noted in vacuum tube work (glass tubes) with the hot spots eventually melting. i.e. the accumulation of charge does not seem to repel further electron hits.

A little flow will tend to even out the gradients. Which is a very good thing from the standpoint of arcing.

[ladajo]

Very much concur. This even occurs on grids themselves due to uneven heating and secondary emissions.

[krenshala]

The purpose of the high resistance coating of the insulators is to prevent localized hot spots. Which have been noted in vacuum tube work (glass tubes) with the hot spots eventually melting. i.e. the accumulation of charge does not seem to repel further electron hits.

A little flow will tend to even out the gradients. Which is a very good thing from the standpoint of arcing.

Off the wall interested amateur question: Would using some type of high-temp glass work for the coils of the magrid? You know, make the Polywell even -more- like a vacuum tube machine?

Or would doing so lead to structural problems due to the differences between a glass and the currently used/proposed metals?

[ladajo]

The more brittle the material, and higher NDT values, the more risk with heat up and cooldown cycles.

[D Tibbets]

The electron KE is equal to the potential difference between the magrid and the emitters, not to the well depth.

Indeed. This referrs to an external electron from a new electron or an electron being recirculated. The important point is that the potential well (eg:wb6) is only ~ 80% of this. It is this potential well voltage (~10,000 volts) that drives the electrons outward, Gauss's law precludes the magrid potential from having an effect on internal electrons (or ions).When the outward moving electron exit at a cusp they have this ~ 10KeV average energy. These non upscattered electrons then sees the +12,000 V magrid potential which is ~ +2000 volts (eV) greater than the electrons outward 10KeV KE. This slows the electron to a stop within some defined distance. Now this recirculating electron starts from a standstill and is acellerated to ~ 12 KeV just like a new electron. The electon speed(KE) and/or vector is changed as it enters the Wiffleball space (due to scattering and curved magnetic field lines) resulting in the net potential well of ~ 10KeV. This potential is the result of the motions of the electrons (radial vs transverse in addition to their initial KE), not just the speed of the electrons. If the electrons maintained perfectly radial motions the potential well depth would equal the accelerating potential, but this cannot happen due to geometry of the Wiffleball borders, in addition to the space charge considerations (i ppm difference in the electron/ ion populations. That is why the potential well is less than the electron velocities would suggest. Unless there is some waste heat generated, the power of the electron injection would not change but the voltage* current relationship would. This is possible because the current is a dynamic flow within the Wiffleball. It is not simply a current from point A to B, it is a bidirectional flow A to B to A to B.....

In any case the accelerating force pushing the electrons out is less than the force pushing electrons in. This serves as a buffer for recovering upscattered electrons (up to 12KeV energies) and recirculating them efficiently to the final 10 KeV potentials within the Wiffleball. The only variable is the distance beyond the midplane of the magrid where this turn around occurs. This has engineering consequences to where standoffs are placed, vacuum vessel wall and other external structures are placed. The electrons need to be stuck to field lines and these field lines have to terminate on structures so that the excessively upscattered electrons cannot loop around into another cusp. This i mentioned in the 2008 patent application.

There is a tradeoff. the closer the potential well depth can approach the magrid accelerating voltage, the less input cost (at least in terms of optimizing the KeV vs fusion rate crossections). But at the same time the margin between the potential well depth and the upscattered electron more closely approaches the magrid potential. Thus upscattering electron losses increase. It might be more economical with a potential well depth of 80-90% instead of ~99%. The low voltage E-gun current would need to be increased to replace the lost electrons. Perhaps more importantly if an electron is upscattered by 1 KeV it would be recovered by recirculation in an 80% well. The upscattered energy would also be recovered (as I have said before the magrid is an excellent example of a direct conversion grid). But if the potential well was 99% of the magrid potential, then this 1 KeV would be lost. In this case only the energy of standard or down scattered electrons would be recovered. The half of the electrons that are upscattered by even a small amount would be lost along with their remaining energy. This would also allow for a higher flow of electrons into the external spaces, which would add to arcing concerns.

Bussard, etel struggled mightily to push the potential well depth (as a ratio of the accelerating voltage) from small values up to the ~ 80-85% level of WB6. They may have reached the ideal performance level once the importance of recirculation was recognized. Improvements past this point may be counterproductive.

Dan Tibbets

[palladin9479]

You should definitely go with unobtainium.

Sorry couldn't resist :p

[93143]

The electron KE is equal to the potential difference between the magrid and the emitters, not to the well depth.

Indeed. This referrs to an external electron from a new electron or an electron being recirculated. The important point is that the potential well (eg:wb6) is only ~ 80% of this. It is this potential well voltage (~10,000 volts) that drives the electrons outward, Gauss's law precludes the magrid potential from having an effect on internal electrons (or ions).When the outward moving electron exit at a cusp they have this ~ 10KeV average energy.

Wrong.

The electrons have 12 keV passing the magrid. They have 2 keV in the centre. When they get back to the altitude of the magrid they have 12 keV again. (Ideally; no collisions.) Conservation of energy.

You made a bum assumption somewhere in there, and it ruined the rest of your reasoning...

...

[Alternately, you could say that the well depth of ~80% refers only to the wiffleball well depth, rather than the total potential difference between the bottom of the well and the magrid. I don't think he meant that, and I can't be bothered to hunt through the papers to make sure... but in neither case is your description right.]

[D Tibbets]

The electron KE is equal to the potential difference between the magrid and the emitters, not to the well depth.

Indeed. This referrs to an external electron from a new electron or an electron being recirculated. The important point is that the potential well (eg:wb6) is only ~ 80% of this. It is this potential well voltage (~10,000 volts) that drives the electrons outward, Gauss's law precludes the magrid potential from having an effect on internal electrons (or ions).When the outward moving electron exit at a cusp they have this ~ 10KeV average energy.

Wrong.

The electrons have 12 keV passing the magrid. They have 2 keV in the centre. When they get back to the altitude of the magrid they have 12 keV again. (Ideally; no collisions.) Conservation of energy.

You made a bum assumption somewhere in there, and it ruined the rest of your reasoning...

...

[Alternately, you could say that the well depth of ~80% refers only to the wiffleball well depth, rather than the total potential difference between the bottom of the well and the magrid. I don't think he meant that, and I can't be bothered to hunt through the papers to make sure... but in neither case is your description right.]

I disagree (I think). In a simple system like analyzing the movement of a point charge A relative to B in a one dimensional system. With more particles and in a 2 dimensional system things are more complex (elliptical or parabolic orbits, etc.) and even more complex in 3 D systems. THe KE of the particle have two componets- the radial vector and the transverse angular momentum vector. This may change the voltage as manifested by the radial KE (or not- I'm not sure).

In any case your simple analysis ignores other particles, ions, neutral gas and interactions with them.
With gas puffers a portion of the injected electron's KE is consumed in ionizing the gas and heating the secondary electrons.

A couple of EMC2 publications that perhaps includes pertinent information. The second in general probably provides all the details you might desire. The system is a complex dynamic system and applying true but simple concepts can be misleading.


PHENOMENOLOGTCAL MODELLING OF POLYWELLm/SCIF
MULTI-CUSP INERTIAL-ELECTROSTATIC CONFINEMENT SYSTEMS

In fact; if
the well depth were to become exactly equal to the electron injection energy (a physical
impossibility so long as ions are present), the cusp fields would not confine the electrons at
all. The MR mode works, but only poorly if the system has a deep potential well.

Forming and Maintaining a Potential Well
in a Quasispherical Magnetic Trap
by Nicholas A. Krall, Krall Associates,

5. Energy distributions
The model proposed in Section IV.B.2 predicts that the
background plasma will force a decrease in the potential
to a value below the gun energy, and that the energy of
the background will be comparable to this potential.
The hot electron population will maintain an energy
somewhat less than its injection energy, because of energy
conservation. Figure 11 clearly shows this trend,
starting at times well below a ms, but becoming most
apparent at the times that correspond to the rise in the
background plasma density, a few ms.

Dan Tibbets