Ejection of byproducts

[BenTC]

I'm not quite sure I've got this right, but here goes..

Apart from the ejection of high energy alphas, I seem to remember some discussion about how to eject particles from the core that don't have any charge. Are there some neutrals created by recombination of ions and electrons? I thought that their speeds differed sufficiently that their "cross-section" for collision was small (is that the correct use of that "cross-section", or does that only apply to the nuclear interactions ?)

Anyway, having just read this by chrismb, "IF" the density in the core is atmospheric, while a deep vacuum is nearby, won't the non-charged particles naturally diffuse into that vacuum?

[D Tibbets]

I'm not quite sure I've got this right, but here goes..

Apart from the ejection of high energy alphas, I seem to remember some discussion about how to eject particles from the core that don't have any charge. Are there some neutrals created by recombination of ions and electrons? I thought that their speeds differed sufficiently that their "cross-section" for collision was small (is that the correct use of that "cross-section", or does that only apply to the nuclear interactions ?)

Anyway, having just read this by chris, "IF" the density in the core is atmospheric, while a deep vacuum is nearby, won't the non-charged particles naturally diffuse into that vacuum?

Read some of the articles by Bussard and others. Neutrals do not have any net charge so they will not respond to magnetic or electrical fields (except where the electric field may strip away an electron, which I believe is less significant than ionization by collisions with high energy electrons). The primary purpose of having the fuel ionized as much as possible is so that they are kept within the magrid primarily due to the magnetically confined electrons. If the population of neutral (but ionizable) atoms exceeds ~ 10e18 (?) per cc arcing can occur outside the magrid. For useful fusion rates you need ~ 10e20 or even 10e22 ions per cc. Bussard has stated that this concentration of the ions inside the machine is the most critical aspect of the wiffleball effect on the ions. The energy lost due to ion escape is not as important as the ions that escape are at the top of thier potential well so not much energy is lost. The recirculation of the electrons also help to reduce the significance of Wiffleball leakage of them.

Any neutrals will diffuse evenly throughout the vacuum vessel and no concentration would be possible. Also, as CB likes to point out, Beam -Beam collisions (ions- ions) are needed to achieve some of the physics advertised for the Polywell. For this to occur the ions in the reaction space needs to significantly exceed the numbers of neutrals present (if neutrals predominate most fusion collisions would be ion- neutral in nature).
That is why ion guns may be needed. Neutral gas injection would work if mechanisms are in place to quickly ionize them before they can transit across the magrid diameter. Bussard claimed that in the WB6 the neutral gas that was injected was ionized quickly by collisions with the injected electrons (initial electrons and secondary electrons from subsequently generated ions- one high energy electron could eventually ionize several hundred neutral atoms/molecules if given enough time). WB 6 struggled to do this due to it's size, larger machines would be more efficient because it would take a neutral atom longer to drift to the otherside of the magrid.

[EDIT]
Ions can and do recombine with free electrons to form neutral atoms (this is the the source of the visible glow in the fusor (or florescent lamp). But most of these recombinations would quickly be reionized very quickly. It is a dynamic enviornment. The recombined neutrals that do survive long enough to travel past the confinement space can accumulate outside the magrid and they need to be accounted for (through pumping) or again arcing would eventially occur.

Any material fusion product would be charged particles, excepting neutrons which leave the reactor quickly and do not play a role. Except of course when the neutron hits the wall it can knock off/sputter alot of ions. These would also need to be accounted for. An aneutronic fuel would suffer from this problem much less as hopefully the charged particles are slowed almost to a stop (direct conversion) before thay hit anything. The removal of these decelerated fusion ions is a topic that has been debated in several threads.



Dan Tibbets

[BenTC]

Thanks for your comprehensive reply. Not everything that I read in various articles sticks together properly. It makes sense at the time but there are too many details that are new and things get mixed up later. I'm planning to get some of the books MSimon mentioned recently.

The way I had remembered it was that the problem was neutrals inside the core. I gather from your reply that this isn't the problem. Apart from them being ionized quickly - and thinking about this in a new light, "perhaps" some neutrals inside the core would be beneficial. Would a collision between a stationary neutral and a fast travelling ion be easier than two fast travelling ions of the same charge? Half the velocity but less repulsion (though I suppose momentum ~ velocity^2 would be a factor).

These atoms/cc figures were new to me...
* arcing can occur outside the magrid ~10e18(?) per cc
* useful fusion rates at ~10e20 or even 10e22 ions per cc.

I found the following for comparison [ ref ] :
* normal earth atmosphere ~10e19 molecules/cc
* high order man made vacuum in a dewar contains ~10e11 molecules/cc
* Intergalactic Space by comparison ~3 molecules/cc.

[D Tibbets]

My calculations could very easily have been wrong.
Lets see:
One mole of gas as STP = 22.4 liters and 6.03 * 10e23 molecules
One cubic meter contains 1000 liters/ 22.4 liters per mole = ~50 moles
50 moles= ~6*10e23 molecules * 50 = 3 *10e25 molecules / cubic meter at STP (one atmosphere). This is ecual to ~ 10e19 molecules per cc at one atmosphere, so yes I screwed up.
Arching - glow discharge becomes a problem at ~5-10 microns or ~
10e-5 atmospheres or ~ 10e14 molecules per cc. This is then the maximum pressure density allowable outside the magrid.

If the Wiffle ball traping allows a gradiant of 1000 the concentration inside the magrid would be ~10e17. A million fold ratio would be needed to reach a density of ~ 10e20 molecules (or neutral atoms or ions).
~ a year ago R. Nebel mentioned that either 10e20 or 10e22 ions/cc was expected or actually measured (I don't rember which . It is in one of the threads)

As far as ion - neutral collisions benifiting from the electron shielding, I supose that there is an effect, but I suspect it is trivial. Remember that there is an excess of free electrons present in the plasma. These will fly past the ions frequentyl in hyperbolic orbits (they cannot be captured due to thier speed). Actual captured electrons in a neutral atom may not spend much more time near the nucleas than any free electrons in the electron excess soup in which the ions are embeded.
Even low energy neutals will drift across the internal volume of the magrid in a tiny fraction of a second, and if not ionized will diffuse to the outside. Using the above calculation, if the neutrals inside the machine exceed ~ 1 part per million, the concentration gradient could not be maintained at net power producing levels. So even if ion- neutral collisions had a very slight advantage in fusion crossection, it could not contribute more than ~ one millionth of the total fusion output and still allow the machine to work.

Hopefully, I have not compunded any of my prevous errors


Dan Tibbets

[TallDave]

R. Nebel mentioned that either 10e20 or 10e22 ions/cc was expected or actually measured (I don't rember which . It is in one of the threads)

I remember we something like that, but I think he was just talking about an extrapolation from the ITER numbers to show a comparable Polywell would have 62500 times greater power density. It was purely hypothetical.

viewtopic.php?p=4911&highlight=iter#4911

Actually, you need to click on “read more” under the design section, then “main parameters” then on the “more” button. What you will find is that the average density of ITER is ~ 1.0e20/m**3. If you use the formula I sent you for the Polywell, you will get a density ~ 2.5e22/m**3. The upshot of this is that the Polywell has a power density that is ~ 62500 times bigger than ITER EVEN IF THERE IS NO ION CONVERGENCE! Thus, a Polywell should far outperform a Tokamak even with a constant density Maxwellian plasma. Even

Or maybe you were thinking of something else. I'm just glad to be saying something in Theory again. The lack of Polywell news is making my neurons cry.

[chrismb]

Maybe there is some interesting news from Tokamak-world to interest your neutrons!?!?

Can't say I've heard any. In fact, despite the recent warm-words from the politicos, it feels like no fusion efforts have been gained recently.

Does anyone know if any fusion projects were won through that recent DoE Stimulus Act round? FF was left out in the cold, and I don't think there were any IEC pitches.

(Sorry for the thread-subject-irrelevance, but seemed an approproate response to TD's post)

[chrismb]

For useful fusion rates you need ~ 10e20 or even 10e22 ions per cc. Bussard has stated that this concentration of the ions inside the machine is the most critical aspect of the wiffleball effect on the ions.

An aneutronic fuel would suffer from this problem much less as hopefully the charged particles are slowed almost to a stop (direct conversion) before thay hit anything. The removal of these decelerated fusion ions is a topic that has been debated in several threads.

The first comment has been discussed. It probably meant 10E21/m^3, as 'n' is referenced in cubic meters and not cc.

In regards the deceleration of fusion product ions, 'direct conversion' is still a dream to be demonstrated. Why is is different for magnetic plasmas? As I say, so very often, this is a mis-overheard conversation by an IEC person over two magnetic plasma people. In an IEC device, there is NO plasma that will decelerate ions in the way you describe. If there were, then the electric fields would not penetrate it and the device wouldn't be an IEC device. The fusion products, be they ions or neutrons, will pummel the containment vessel and ablate it until the reaction volume is contaminated to the point of non-function. Positively charged ions sputter far far more effectively than neutrons - it's why neutrons can get so far in a shield but ions stop almost immediately, they have to give their energy up to the reaction-volume-facing material surface and do so in quick order.

[KitemanSA]

I'm just glad to be saying something in Theory again. The lack of Polywell news is making my neurons cry.

I've been thinking the same thing for several months now.

And now; we wait!

Code: Select all

                           da
                            da
                             da
  da       da        da  da   da    da
                               da
                                da
da  da  da  da    da  da         da   da…


      da

Final Jeopardy anyone?

[D Tibbets]

For useful fusion rates you need ~ 10e20 or even 10e22 ions per cc. Bussard has stated that this concentration of the ions inside the machine is the most critical aspect of the wiffleball effect on the ions.

An aneutronic fuel would suffer from this problem much less as hopefully the charged particles are slowed almost to a stop (direct conversion) before thay hit anything. The removal of these decelerated fusion ions is a topic that has been debated in several threads.

The first comment has been discussed. It probably meant 10E21/m^3, as 'n' is referenced in cubic meters and not cc.

In regards the deceleration of fusion product ions, 'direct conversion' is still a dream to be demonstrated. Why is is different for magnetic plasmas? As I say, so very often, this is a mis-overheard conversation by an IEC person over two magnetic plasma people. In an IEC device, there is NO plasma that will decelerate ions in the way you describe. If there were, then the electric fields would not penetrate it and the device wouldn't be an IEC device. The fusion products, be they ions or neutrons, will pummel the containment vessel and ablate it until the reaction volume is contaminated to the point of non-function. Positively charged ions sputter far far more effectively than neutrons - it's why neutrons can get so far in a shield but ions stop almost immediately, they have to give their energy up to the reaction-volume-facing material surface and do so in quick order.

Units of CC or cubic Meters, I'm always forgeting which is being used. Hopefully while the units are wrong, my ratios are still reasonable.

Looking at it another way- Assuming it takes 1 billion fusions per second to produce 1 microwatt of power, then 10e20 fusions per second are needed to generate 100 million watts. In a ~3 meter diameter Polywell at 10e22 ions per cubic meter, there wound be ~3*10e23 total ions(~30 cubic meters volume in the machine). This means that one out of every ~ 3000 ions would need to fuse within one second. This is within an order of magnitude of CB's estimate of ~ ion mean time to fusion of ~ 8000 seconds, so I suppose the numbers are reasonable (?). If the machine arcs at ~ 7 microns pressure (~10e-5 atm.) the limiting ion density outside the machine would be ~ 10 e20 ions per cubic meter (10 e14/cc). This means the Wiffleball traping only has to maintain a 100 fold gradient inside the magrid. This is much more forgiving than my above (apparently wrong) estimates.

As to direct conversion, the streams of alpha (or other charged fusion product ions) is not a net neutral plasma (like in a Tokamak), and not even the quasineutral plasma within the magrid of the Polywell (~1 ppm electron excess) The electrons are are kept back within the Wiffleball. Within the limits of tolorable charge seperation ( how this plays out is beyond me) the isolated pos. ions streaming twoards the walls (presumably in beams through the cusps) in theory can be decellerated in a process similar but opposite to particle accelerators. The charge inside the magrid is maintained by the continous injection of new ions as the leaked ions and fusion ions neutralize on the walls, so the balance remains the same. In the case where neutral gas is injected instead of ions it seams a huge amount of electrons would be left behind. How this would be handled is again beyond me. Of course every thing else makes perfect sense.


Dan Tibbets

[MSimon]

Dan,

Convention when writing a number like 30,000 in text exponential form is

3E4. Not 3 *10E4. The 10 is assumed. Of course if you can do exponents then 3 10⁴ is perfectly proper.

[chrismb]

Dan,

Convention when writing a number like 30,000 in text exponential form is

3E4. Not 3 *10E4. The 10 is assumed. Of course if you can do exponents then 3 10⁴ is perfectly proper.

Not wishing to drag a boring bit of pedantery too far, but why not when the boards are so quiet, but Dan is potentiall less ambiguous than you are (though has equally mixed the issue): 3E4 has, indeed, become adopted as meaning 30,000 but in point of fact the convention of saying 30k is far more established. If it were, say 30,500 then you'd write 30k5. That being said, and because 'E' is the SI symbol for 10^18, writing 3E4 would more readily mean "3 quintillion 400 quadrillion". It's all about context, I guess...

[KitemanSA]

What in the world is thirty kilo-fives? Did you mean 30.5k? And three Exa-fours is equally baffling. But 3E4 is almost universal compu-speak. Heck, even Excel accepts it.

[blaisepascal]

Dan,

Convention when writing a number like 30,000 in text exponential form is

3E4. Not 3 *10E4. The 10 is assumed. Of course if you can do exponents then 3 10⁴ is perfectly proper.

Not wishing to drag a boring bit of pedantery too far, but why not when the boards are so quiet, but Dan is potentiall less ambiguous than you are (though has equally mixed the issue): 3E4 has, indeed, become adopted as meaning 30,000 but in point of fact the convention of saying 30k is far more established. If it were, say 30,500 then you'd write 30k5. That being said, and because 'E' is the SI symbol for 10^18, writing 3E4 would more readily mean "3 quintillion 400 quadrillion". It's all about context, I guess...

I'm not certain I've ever seen the prefix-replacing-decimal-point convention as an established convention. I've not sure I've ever seen a good convention for counts, except to simply drop the units. So 1000000 fusions per second would probably be written as 1000000/s or perhaps 1M/s, maybe.

Perhaps the paragraph in question should be written as:

Looking at it another way- Assuming it takes 1.7amol/s fusions to produce 1 µW of power, then 170µmol/sec fusions are needed to generate 100 MW. In a ~3 m diameter Polywell at 1.7mmol/m^3, there would be ~500mmol total ions(~30 m^3 volume in the machine). This means that one out of every ~ 3000 ions would need to fuse within one second. This is within an order of magnitude of CB's estimate of ~ ion mean time to fusion of ~ 8000 seconds, so I suppose the numbers are reasonable (?). If the machine arcs at ~ 7 microns pressure (~100µatm.) the limiting ion density outside the machine would be ~ 170 µmol/m^3 (170pmol/cm^3). This means the Wiffleball traping only has to maintain a 100 fold gradient inside the magrid. This is much more forgiving than my above (apparently wrong) estimates

[chrismb]

I'm not certain I've ever seen the prefix-replacing-decimal-point convention as an established convention. I've not sure I've ever seen a good convention for counts, except to simply drop the units.

What in the world is thirty kilo-fives? Did you mean 30.5k? And three Exa-fours is equally baffling. But 3E4 is almost universal compu-speak. Heck, even Excel accepts it.

Interesting you should say that, and therein lies the 'cause'. In the age when 'computer' was still a job description the old-and-bold scientists would be scribbling away on paper, and it is dead easy to write a superscript. Then came an interlude between SI units and computer ubiquity when the nomenclature gradually became more lazy and one-lining. Now we're into a new era of software that has been principally written for the accounting classes and engineers have adapted yet again.

There's nothing set in stone on this, and nomenclature matures and changes, but I assure you that 3E4 would have meant something different some years ago. Go ask an electronics engineer, today, what it means (after telling him the symbol for 10^18, 'Exa', is 'E').

Dan's nomenclature, 10e4, is less confusing as there is no 'e' in SI and clearly hints towards an exponential form, and separating off the significand is even more distinct and obvious. Not sure if I write 'E' or 'e' now, but frankly I'm boring myself too much to bother checking, so will stop this pedantery....!! (Fusion boards are sooo quiet these days, maybe it really is time to focus on fission research...)

[D Tibbets]

........................

Perhaps the paragraph in question should be written as:

More percise discription, but I believe 7 microns =~ 10 micro armospheres, not 100. (760,000 microns per atmosphere).

Dan Tibbets