Harvey wrote:Okay, I get it; this is a "only Polywell works" discussion group, not a general scientific discussion group.
But to share what I have read:
Again it is about cost; your electromagnetic drivers require a huge power supply and are very expensive. My guess is that will be about 1000x what pneumatic impact will cost...i.e. you will need nearly a billion dollars so it will not be funded by government or private industry.
MTF at LANL is not a "spare no expense" place. They are quite under funded and pay huge tariffs to the National Lab. Exact dollars etc are available on the web.
You don't control compressed steam, you servo the pistons. Regarding the Navy; they do not have any constraints on cost.
Proton Boron while "officially" aneutronic (less than 1% of energy from neutrons) unfortunately produces about .1% neutrons. Even more unfortunately these will kill you. In comparison General Fusion with their lead filled sphere will have much less neutrons than any p-B11 scheme.
GF might very well turn out not to work but it might also be an affordable way to get net gain. Certainly it is a new approach. I give them a higher chance of success than any p-B11 proposal...
Yes, pneumatics will be cheap in theory.
Getting 200 or 500 pneumatic cylinders to all operate to within 1 uSec or less precision will run up the cost and even then it is uncertain that it can be done. Then there is the wear out factor.
BTW lead is not very good at stopping neutrons. Stopping neutrons is best done by first thermalizing them and then absorbing them. Boron is a good absorbent. The result is low energy gammas. Which are primarily reduced by mass. Lead is good if you want to minimize volume of the mass. Concrete is good if you want to minimize cost.
MTF is obviously well funded compared to what Dr. B has had to work with.
Yes. You servo the piston. I'm not aware of any pneumatic servo that can attain 1 USec precision. The speed of sound limits you. For the sake of argument let us stipulate that the speed of sound in the compressed air is 1,000 m/second. That means that to get 1 uSec precision you must control the various distances in the machine to better than 1 mm. Tough. Then there is the wear out.
Honestly, if I thought this approach had better prospects I'd start to work on the engineering ASAP. In any case I absolutely would not make the prototype dependent on getting the compressed air cylinders working. That would be phase 2 after I proved it with electrical drivers.
What the rig requires is in fact 200 or 500 servo controlled oscillators. Not only kept identical in frequency but also phase. At 1,000 Hz operating frequency the phase would need to be controlled to within .1 degree or better. Which implies a system response out to 10 MHz. Do you off hand know of any pneumatic servo systems with a Bode plot that has no resonances and is 3 dB down at 10 MHz? Quite a trick if you can pull it off. Heck, it will be very hard to do with electromagnets. I'll go easy on you. Do you know of any pneumatic servos that extend out to even 1 MHz?
Consider this. If the servo piston is more that 1 mm away from the gas valve you can't control the system out to 1 MHz. The speed of sound kills you by adding a lag to the system that is significant relative to the control frequency. For good control without the system running away into oscillation you want a bode plot that extends to at least 3X your required frequency response and 10 X is better. It also means your piston can't be more than 2 mm in diameter or stroke. What gas pressure will be required to get enough force from a 2 mm dia piston? Just because I can conceive of something doesn't mean I can engineer it.
I'm always looking for better devices and improvements in the field. This one has some problems as conceived. How about the requirement to get the liquid metal spinning so as to create a vortex induced void in the center? That will be the first trick required even before considering compression. Then you have to do high pressure gas injection without distorting the sphere. I don't think they have that one solved.
The neat thing about the Bussard machine is that similar machines are already producing fusions. What Dr. B has done is to reduce the losses to make such a machine a net energy generator. Plus the controls are all electrical. Electrical power supplies these days come it at $1 to $2 a watt for low volume production. It goes down to $.25 a watt for high volume. What does an air compressor cost in terms of $ per watt?
By trying to optimize the cost of the power supplies required for compression you suboptimize a LOT of other things.
Dr. B's compressor uses fixed magnetic fields and fixed drive voltages. The compression can be varied by controlling electron injection. I can do that easily at up to 100 MHz.
The technology Dr. B is using is common in semiconductor processing. And linear accelerators (becoming common in industry for other uses). Air cylinders with 1 MHz response frequency are common where exactly?
Net gain is always about energy of compression vs power out. Adiabatic compression is not very efficient compared to linear accelerators. So right away you start out with a big handicap. The same one ITER is up against. Since you are using adiabatic compression you are limited to D-T fuel. Because only the thermal tail gets hot enough to fuse.
Dr. B's machine uses linear accelerators for "compression" which is why he will be able to burn pBj and adiabatic machines can't. It is just a matter of getting the drive voltages up. Not too tough.
I look forward to your response.