From the point of view of cooling, shielding and supporting the coils a closed box design is far simpler. In addition to that corner losses can be more conveniently eliminated through repeller plates.
All that efficiency means is the ammount of energy you can get out in electricity in proportion to the fuel you put in. Efficiency is important for the proton-boron reaction, but for the D-T reaction the energy that comes out will be far in excess of the energy that is put in, since the cost of fuel is negligable to the cost of operation and maintenance the main drive will be to construct a simple, cheap machine that can get fusion energy out.
As a result it may well be preferable to build a machine that can be made of cheap materials that can be maintained, that runs through thermal conversion then to build an advanced machine where the conversion mechanism is more efficiennt but the materials aswell as operation and maintenance are more complex.
* * * *
With regards to rockets it may well be that an open box machine is best even if the materials requirements are more advanced, but that doesn't mean that the same engine will be used in run of the mill power plants.
This is in the same way that solar cells used in calculators and watches aren't always the same as the ones in spacecraft. Or the way compact reactors that use highly enriched Uranium for naval vessels aren't the same as the large reactors that use 4% enriched uranium for commercial power preoduction.
It is of value to look into both open and closed box designs.
* * * * * * *
MSimon, perhaps your right about the problems in a closed box design for direct conversion, but if you could build the thing for cheaper and shield the coils more easily aswell, then that might outweigh the advantages of direct conversion for a run-of-the-mill power station.
BTW even if Rider is right about Bremstrahlung losses (and there are good reasons to believe he isn't) then a D-T burning Polywell could still work. Whats more getting breakeven in D-T is likely to be alot easier than getting it using boron, so even if boron burning is hypothetically possible with a Polywell it would be wiser to successfully design a net power producing D-T plant first, show it off to the world and then when people throw money at fusion from right, left and centre take on the considerably more challenging task of designing a boron fusion polywell. If you try to design a boron power plant first time and fail funding might dry up and you might not get a second chance.
If your going to use DT the coils need to be shielded, which requires an closed box design.
Earthed Plate screening for WB5 repellors
Closed box designs can't work because they do not allow electron recirculation. Without recirculation and the whiffle ball effect, a machine would require superconducting magnets made of unobtanium. In fact - the problems with the closed box design do not allow the whiffle ball effect to work well and the fields are not conformal to the grids.
The magnets in the chamber do several things for you: shield the grid, whiffle ball effect, make electrostatic power recovery possible - because the larger volume would not require bigger magnets.
BTW I've worked out a way (on paper) to shield MgB superconductor coils for 40 years of operation in a high neutron flux environment (one thing to do is to make the MgB from high purity B11). The #1 difficulty will be extracting the heat from the n B10 > 2.8 MeV reaction.
For pBj the heat loads would be insignificant and a very little shielding material ought to give a coil life of 100 years.
The magnets in the chamber do several things for you: shield the grid, whiffle ball effect, make electrostatic power recovery possible - because the larger volume would not require bigger magnets.
BTW I've worked out a way (on paper) to shield MgB superconductor coils for 40 years of operation in a high neutron flux environment (one thing to do is to make the MgB from high purity B11). The #1 difficulty will be extracting the heat from the n B10 > 2.8 MeV reaction.
For pBj the heat loads would be insignificant and a very little shielding material ought to give a coil life of 100 years.
Engineering is the art of making what you want from what you can get at a profit.
Look, MSimon, this repeller idea would effectively cause recirculation by intensifying the electrostatic potential at the cusps so the electrons recirculate by being reflected back in through the cusps. You would have some losses to the ion screen, but possibly you could come out ahead of the traditional machine in terms of electron loss power. The wiffle-ball would be unaffected, Tom Ligon's comment on the PXL-1 in the thread "Space Magnetism" shows this is true.
I know it won't work with electrostatic decelleration of the alphas, but this might be a good step to use with a smaller-scale D-T machine. It might be able to demonstrate breakeven on a smaller budget than pB11, which could be good for funding. I agree that I'd rather see pB11 used because it is more efficient and causes less radioactive waste, but as an intermediate step something like this might really save this work if it could provide an impressive demonstration and bring investment/grants.
I know it won't work with electrostatic decelleration of the alphas, but this might be a good step to use with a smaller-scale D-T machine. It might be able to demonstrate breakeven on a smaller budget than pB11, which could be good for funding. I agree that I'd rather see pB11 used because it is more efficient and causes less radioactive waste, but as an intermediate step something like this might really save this work if it could provide an impressive demonstration and bring investment/grants.
D-T is a waste of time unless you are willing to go with very large machines with lithium blankets for T generation. Which means larger magnets, thermal conversion to electricity plant, etc.
Plus - lithium is scarce. Using it in batteries is probably better than burning it up.
I'd go with D-D. And if I can make D-D work with a little more effort I can get pBj to go.
Plus - lithium is scarce. Using it in batteries is probably better than burning it up.
I'd go with D-D. And if I can make D-D work with a little more effort I can get pBj to go.
Engineering is the art of making what you want from what you can get at a profit.
I completely dont follow this line if thinking.jmc wrote:Whats more getting breakeven in D-T is likely to be alot easier than getting it using boron, so even if boron burning is hypothetically possible with a Polywell it would be wiser to successfully design a net power producing D-T plant first, show it off to the world and then
WB-6 & 7 are 30 cm devices.
Break even, even with DT is likely to mean a 1.5 meter device.
It is likely that a 30cm size device could do PB-11 fusion.
There are 2 issues at hand:
1) Configuration. Truncated cube vs Dodec.
2) Scaling laws.
They can be solved easily by building:
1) 30cm Truncated cube & Dodec.
2) 60cm devices.
If PB-11 is a 3rd issue, a 30 cm device is reasonable.
At that point we would know what the scaling laws are and whether or not the cube or DoDec is best, and that PB-11 fusion has been proven.
Once we know these things we can consider building a net power producing proof of concept. Why stop at break even ? That is silly, I think the logical steps are to go from 30cm to 60cm to net power.
I like the p-B11 resonance peak at 50 KV acceleration. In2 years we'll know.
I don't see the need. Electrons with energies below the grid voltage will recirculate. Those above the grid voltage will be lost.Solo wrote:Look, MSimon, this repeller idea would effectively cause recirculation by intensifying the electrostatic potential at the cusps so the electrons recirculate by being reflected back in through the cusps. You would have some losses to the ion screen, but possibly you could come out ahead of the traditional machine in terms of electron loss power. The wiffle-ball would be unaffected, Tom Ligon's comment on the PXL-1 in the thread "Space Magnetism" shows this is true.
The only way to keep the magnetic field lines from intercepting metal is to put the magnets in the reactor vessel.
Engineering is the art of making what you want from what you can get at a profit.
I think for now the fusion effort should concentrate on magrid designs such as WB7.
There is an incorrect statement above. WB5 was the only one of the WB series to be a closed box machine. I would have named it PXL-2. Since WB means wiffleball, and Dr. Bussard expected wiffleballs could form in either magrid or box machines, he evidently decided to call 5 a wiffleballer.
Once we've saved the world and have plenty of resources, I'd love to see what a few clever grad students could do with closed-box machines, but I would caution them against trying repeller plates. Sticking anything in the corners seems to cause trouble.
Stick a metal plate there charged to repel electrons, and it will attract ions. The ions will hit the plate, cause secondary electron emission, pick up electrons to release neutral gas, and dump heat, and then you'll have hot repeller plates that emit electrons.
We even had trouble with insulators. On the MPG (single-turn copper tube magrid) machines, forming the magrid by using ceramic couplings at the corners caused fireworks. You would think alumna would be well behaved, but evidently it acts as a recombination catalyst. If it charges up with electrons, it attracts ions, and sparks start to fly ... literally.
A cloud of low-energy electrons hanging in the corners, OTOH, would act as a repeller without offering a surface on which recombination could occur.
I described the bizarre behavior of PXL-1 in the Space Magnetism thread. If anything can be done with these machines, the trick lies not in repeller plates or insulated corners, but by exploiting their native tendency to drop into that mysterious low-loss mode. That is probably a 2-color electron trick of some sort, in which low-energy electrons block the loss path.
It probably does make sense to open up the corners, putting a deep port on each corner for such tricks as diagnostics, pumping, emitters, etc, and to go to larger face ports. I believe that describes WB5.
IIRC, WB5 did make some fusion. Maybe some breakthru will eventually turn these into useful machines. The benefits of having the magnets outside are certainly great, and they have always been very seductive because of this. It could even be that superconducting magnets of a Tesla or two will achieve the desired effect. But for now, consider the closed box machines to be sirens trying to lure you onto the rocks.
There is an incorrect statement above. WB5 was the only one of the WB series to be a closed box machine. I would have named it PXL-2. Since WB means wiffleball, and Dr. Bussard expected wiffleballs could form in either magrid or box machines, he evidently decided to call 5 a wiffleballer.
Once we've saved the world and have plenty of resources, I'd love to see what a few clever grad students could do with closed-box machines, but I would caution them against trying repeller plates. Sticking anything in the corners seems to cause trouble.
Stick a metal plate there charged to repel electrons, and it will attract ions. The ions will hit the plate, cause secondary electron emission, pick up electrons to release neutral gas, and dump heat, and then you'll have hot repeller plates that emit electrons.
We even had trouble with insulators. On the MPG (single-turn copper tube magrid) machines, forming the magrid by using ceramic couplings at the corners caused fireworks. You would think alumna would be well behaved, but evidently it acts as a recombination catalyst. If it charges up with electrons, it attracts ions, and sparks start to fly ... literally.
A cloud of low-energy electrons hanging in the corners, OTOH, would act as a repeller without offering a surface on which recombination could occur.
I described the bizarre behavior of PXL-1 in the Space Magnetism thread. If anything can be done with these machines, the trick lies not in repeller plates or insulated corners, but by exploiting their native tendency to drop into that mysterious low-loss mode. That is probably a 2-color electron trick of some sort, in which low-energy electrons block the loss path.
It probably does make sense to open up the corners, putting a deep port on each corner for such tricks as diagnostics, pumping, emitters, etc, and to go to larger face ports. I believe that describes WB5.
IIRC, WB5 did make some fusion. Maybe some breakthru will eventually turn these into useful machines. The benefits of having the magnets outside are certainly great, and they have always been very seductive because of this. It could even be that superconducting magnets of a Tesla or two will achieve the desired effect. But for now, consider the closed box machines to be sirens trying to lure you onto the rocks.
Roger,
Good plan from an engineering standpoint.
I'd do one more thing: build them in lots of threes (if the funding was there). Most of the small quantity cost of custom production is set up. Making 3X might cost only 1.5X as much as making one.
Then you have one machine doing experiments. One cooking out. One being refitted. Or alternate between two machines with the "third" kept as spare parts.
I'd add one other point. All future machines should be able to run for at least a 5 minutes a shot. Preferably one hour.
Good plan from an engineering standpoint.
I'd do one more thing: build them in lots of threes (if the funding was there). Most of the small quantity cost of custom production is set up. Making 3X might cost only 1.5X as much as making one.
Then you have one machine doing experiments. One cooking out. One being refitted. Or alternate between two machines with the "third" kept as spare parts.
I'd add one other point. All future machines should be able to run for at least a 5 minutes a shot. Preferably one hour.
Engineering is the art of making what you want from what you can get at a profit.
The reason why closed box machines make sense IMO is because, I do not see how you can protect the coils in and ope box machine from neutrons. If this cannot be done then a Polywell cannot be made to burn DT.
Yes, any machine that burns DT will be fairly large, but it still might be far better than ITER (steady state, the primary heating mechanism a thermionic cathode rather than a neutral beam, higher beta, lower field, etc.etc.)
And the fact of the matter is that DT fusion is alot easier than proton boron fusion, bremmstrahlung is not a problem at the lower temperatures, low electron energies reduce cusp losses etc. on top of that if you design a reactor that is only capble of conducting aneutronic fusion you'll freak everyone out. If you design a reactor that can do DT aswell then people will give it more credance and will be more likely to build one and once you have a well designed, well diagnosed polywell you'll be able to veryify all the physics concepts required to make boron-proton fusion possible.
I would disagree with putting off the design of a closed box machine, I believe it would be adviseable to pursue both approaches at the same time. You can design a closed box DT Polywell reactor capable of producing net energy, I don't think this is possible with an open box design.
Here are some ideas I can think of to make repeller plates work better.
1) Place an earthed screen in front of them with narrowholes at point cusps and narrow slits at the line cusps of width 3-5 electron gyroradii.
2) Make the repeller plated in a deep V-shape so that the odd ion that wanders in through the slits hits them obliquely spreading the deposited heat over a large area. I don't understand the neutral emmision bit, sure any secondaries creates will go back out through through cusps an the slits.
3) Another idea I had (which bussard may also have had) would be to have the thermionic electrode at a less negative potential to the others, this would have the effect of only allowing electrons that were upscattered to the extreme to touch the other plates (the feasibility of this all depends on whether the repeller plates can be shielded from the ions), this might possibly reduce the effective cusp loss holes to one. In addition to that after 'filling up' the polywell with electrons you might cool down the thermionic electrode and push up the voltage og the electron source, completely isolating the electrons from the solid surfaces, except by crossfield transport (Provided the ions could be shielded from the plates)
Yes, any machine that burns DT will be fairly large, but it still might be far better than ITER (steady state, the primary heating mechanism a thermionic cathode rather than a neutral beam, higher beta, lower field, etc.etc.)
And the fact of the matter is that DT fusion is alot easier than proton boron fusion, bremmstrahlung is not a problem at the lower temperatures, low electron energies reduce cusp losses etc. on top of that if you design a reactor that is only capble of conducting aneutronic fusion you'll freak everyone out. If you design a reactor that can do DT aswell then people will give it more credance and will be more likely to build one and once you have a well designed, well diagnosed polywell you'll be able to veryify all the physics concepts required to make boron-proton fusion possible.
I would disagree with putting off the design of a closed box machine, I believe it would be adviseable to pursue both approaches at the same time. You can design a closed box DT Polywell reactor capable of producing net energy, I don't think this is possible with an open box design.
What about the repeller plates? Were they in deep ports? WB5 looked fairly boxy from the outside.Tom Ligon wrote: It probably does make sense to open up the corners, putting a deep port on each corner for such tricks as diagnostics, pumping, emitters, etc, and to go to larger face ports. I believe that describes WB5.
Here are some ideas I can think of to make repeller plates work better.
1) Place an earthed screen in front of them with narrowholes at point cusps and narrow slits at the line cusps of width 3-5 electron gyroradii.
2) Make the repeller plated in a deep V-shape so that the odd ion that wanders in through the slits hits them obliquely spreading the deposited heat over a large area. I don't understand the neutral emmision bit, sure any secondaries creates will go back out through through cusps an the slits.
3) Another idea I had (which bussard may also have had) would be to have the thermionic electrode at a less negative potential to the others, this would have the effect of only allowing electrons that were upscattered to the extreme to touch the other plates (the feasibility of this all depends on whether the repeller plates can be shielded from the ions), this might possibly reduce the effective cusp loss holes to one. In addition to that after 'filling up' the polywell with electrons you might cool down the thermionic electrode and push up the voltage og the electron source, completely isolating the electrons from the solid surfaces, except by crossfield transport (Provided the ions could be shielded from the plates)
I can see how to protect the coils in a high neutron machine: proper shielding. H2O and B10 should do the trick.The reason why closed box machines make sense IMO is because, I do not see how you can protect the coils in and ope box machine from neutrons. If this cannot be done then a Polywell cannot be made to burn DT.
Since you need the H2O for cooling anyway - no problem.
I think I did a piece on this at the IEC Fusion Tech blog. I haven't revised the estimate I posted there, but at this time I believe an engineering solution could be worked out.
Engineering is the art of making what you want from what you can get at a profit.
<<What about the repeller plates? Were they in deep ports? WB5 looked fairly boxy from the outside. >>
I really don't know how it was put together, except to say it had physical modifications to allow tricks of that sort. PXL-1 had fairly plain corners except for the pump port and one port opposite the pump port which we exploited for such tricks as injecting microwaves.
Honestly, any physical trick you stick in the corners is not going to work as well as you think. That was the lesson. All this stuff kept coming up in HEPS, PXL-1, and WB-5, and the conclusion was it was not going to work. Had they gone straight to WB6 and not screwed around with WB-5, we might be building the big machines right now. I know Dr. Bussard spent a lot of time in his last two years regretting that choice.
The one thing that, to me, held promise, was PXL-1's apparent ability to make its own "virtual" electron repeller plates. If you learn to work with the physics of these machines, they'll probably do it on their own.
If you try to cheat, they'll bite you. You are free to try stuff like this if you wish, but my advice is to build repeller structures in a manner that allows you to conveniently remove them.
I really don't know how it was put together, except to say it had physical modifications to allow tricks of that sort. PXL-1 had fairly plain corners except for the pump port and one port opposite the pump port which we exploited for such tricks as injecting microwaves.
Honestly, any physical trick you stick in the corners is not going to work as well as you think. That was the lesson. All this stuff kept coming up in HEPS, PXL-1, and WB-5, and the conclusion was it was not going to work. Had they gone straight to WB6 and not screwed around with WB-5, we might be building the big machines right now. I know Dr. Bussard spent a lot of time in his last two years regretting that choice.
The one thing that, to me, held promise, was PXL-1's apparent ability to make its own "virtual" electron repeller plates. If you learn to work with the physics of these machines, they'll probably do it on their own.
If you try to cheat, they'll bite you. You are free to try stuff like this if you wish, but my advice is to build repeller structures in a manner that allows you to conveniently remove them.
Why would removing them be good? I thought that in Bussard's Askmar paper he stated that WB5 could get something like 1000 times the ion density as the HEPS machine did without repeller plates. So it seems to indicate to me that he did manage to improve things by a great deal by adding them.Tom Ligon wrote:If you try to cheat, they'll bite you. You are free to try stuff like this if you wish, but my advice is to build repeller structures in a manner that allows you to conveniently remove them.
P.S. I think I've got a clever if slightly overcomplicated structure for protecting the repeller plates.
Because you can get the same effect with lower losses with stronger magnets. If the WB effect is large enough the losses from repeller plates will exceed the gains (overall).jmc wrote:Why would removing them be good?Tom Ligon wrote:If you try to cheat, they'll bite you. You are free to try stuff like this if you wish, but my advice is to build repeller structures in a manner that allows you to conveniently remove them.
With this particular device nothing is our most effective component. If we could reduce the MaGrid to nothing we would be in really good shape.
We must make the plasma do our work for us using as much nothing as possible.
Engineering is the art of making what you want from what you can get at a profit.
jmc,
The abilitity to remove the repeller plates is good because if (most likely when) you discover they are screwing the thing up, you will wish to do so.
There is nothing difficult about this. Install CF vacuum ports on all the corners and faces, the largest diameter you can fit, and have a good stock of copper gaskets of those sizes. And make sure you keep a set of blank plates on hand. That way, you can try complicated combinations of stuff to whatever degree your budget allows, and then remove them to see what it will do without that stuff. You can also use the ports to try extra electron emitters, microwave or laser plasma density diagnostics, Langmuir probes, POPS antennae, etc.
The other nice thing about this structure is, with the magnets outside, you are at liberty to play with those, as well. This would be a nice structur for trying out high temperature superconductor magnets, for instance.
The abilitity to remove the repeller plates is good because if (most likely when) you discover they are screwing the thing up, you will wish to do so.
There is nothing difficult about this. Install CF vacuum ports on all the corners and faces, the largest diameter you can fit, and have a good stock of copper gaskets of those sizes. And make sure you keep a set of blank plates on hand. That way, you can try complicated combinations of stuff to whatever degree your budget allows, and then remove them to see what it will do without that stuff. You can also use the ports to try extra electron emitters, microwave or laser plasma density diagnostics, Langmuir probes, POPS antennae, etc.
The other nice thing about this structure is, with the magnets outside, you are at liberty to play with those, as well. This would be a nice structur for trying out high temperature superconductor magnets, for instance.
@Tom Ligon:
What kind of voltage would we be looking at to repell all electrons? If I'm right, you are saying that there will still be some impacts due to upscattered electrons having more energy than others, so the repelling potential will have to be larger than the average energy. But with enough voltage, it should be possible to repell all electrons and prevent that loss current, correct? And the ion screens will prevent that recombination catalysis that you had with the alumina insulator/repellers.
In any event, the closed box does look more promising for lower-budget work, maybe something a university could do. PRobably not a private individual like the fusor people, though.
What kind of voltage would we be looking at to repell all electrons? If I'm right, you are saying that there will still be some impacts due to upscattered electrons having more energy than others, so the repelling potential will have to be larger than the average energy. But with enough voltage, it should be possible to repell all electrons and prevent that loss current, correct? And the ion screens will prevent that recombination catalysis that you had with the alumina insulator/repellers.
In any event, the closed box does look more promising for lower-budget work, maybe something a university could do. PRobably not a private individual like the fusor people, though.