Spheroidal Foci and POPS?
DeltaV - Not so much on topic, but a lot of background theory papers can be found on ASKMAR, which he pulled from DTIC.
http://www.dtic.mil/
If you search Polywell on DTIC, you can find a lot of the background research Dr Bussard did to set the stage for today.
Polywell seems to be an careful operator induced balance system. I am sure we will see some interesting results from Famulus now that he has his Mass Flow Controller up and running. His reactor has entered the entirely digital age (unlike, as far as I can tell, Dr Nebel's). Once Famulus transitions to a Polywell Field from fusor, it should be more interesting, (However, I think that gas mix and vacuum chamber size will present immediate physical challenges to him).
Polywell may seem robust on the surface, however, it seems the actual fruitful operating regions lay in a narrow band. (See ASKMAR and DTIC, and the foundation research Dr. Bussard accomplished.) As MSimon may say, simple physics can lead to complex engineering. I have been going back recently and re-reading alot of the papers to compare them to what we now think we understand. Interesting to follow the growth. It may yet be that we all come to accept that "cusp recirculation" becomes "cusp oscillation", and that the tuning of a Polywell becomes a complex endeavor. What happens when you PID a PID on a PID? Chaotic Fuzzy Logic Controllers?
It is fun to muse, but I think looking for Mach Effect in Polywell is very far down the road, if ever. The system balances and tuning would be painful at the least. The inherent foci shifts this would probably induce could result in detrimental effects to efficiency and increased losses. (Lots of hand waving here). Shoving the "Spikeyball"(I like saying that) and directing Alphas one way or another using the B-Field is a stimulating mental exercise that butts against the physical construct of the machine as we know it. In the mean time I think we should stick to simpler things like: Is it better to eliminate the nubs between coils, and go with vessel/wall based mounts to feed power to each coil thus leaving no heating/loss target between them? Asked before, but not explored so well (unless Dr. Nebel hasn't told us that is part of WB8 design).
As you said, let's get the baseline machine working first.
Dan T - Back to your Pressure cooker experiment. How about enhancing your microwave donuts with some copper turns and another power source to feed them? You could even add a Pot and play "dial-a-field", may make for some great video. I looked at your pics as you upped your B Fields by changing magnets, and as you noted, there is a definate difference in the plasma configuration (Spikeyness). Recombinations given, (you must have free electrons and ions to do so), there seems to be validity in a better understanding to exploring the visual compression effect induced by the combined magnetic fields. This came to mind while re-reading Dr. Bussard's "PHENOMENOLOGICAL MODELLING OF POLYWELL/SCIF MULTI-CUSP INERTIAL-ELECTROSTATIC CONFINEMENT SYSTEMS" (DTIC AD-257944) paper where it discussed the device transition point from Mirror to Whiffleball.
http://www.dtic.mil/
If you search Polywell on DTIC, you can find a lot of the background research Dr Bussard did to set the stage for today.
Polywell seems to be an careful operator induced balance system. I am sure we will see some interesting results from Famulus now that he has his Mass Flow Controller up and running. His reactor has entered the entirely digital age (unlike, as far as I can tell, Dr Nebel's). Once Famulus transitions to a Polywell Field from fusor, it should be more interesting, (However, I think that gas mix and vacuum chamber size will present immediate physical challenges to him).
Polywell may seem robust on the surface, however, it seems the actual fruitful operating regions lay in a narrow band. (See ASKMAR and DTIC, and the foundation research Dr. Bussard accomplished.) As MSimon may say, simple physics can lead to complex engineering. I have been going back recently and re-reading alot of the papers to compare them to what we now think we understand. Interesting to follow the growth. It may yet be that we all come to accept that "cusp recirculation" becomes "cusp oscillation", and that the tuning of a Polywell becomes a complex endeavor. What happens when you PID a PID on a PID? Chaotic Fuzzy Logic Controllers?
It is fun to muse, but I think looking for Mach Effect in Polywell is very far down the road, if ever. The system balances and tuning would be painful at the least. The inherent foci shifts this would probably induce could result in detrimental effects to efficiency and increased losses. (Lots of hand waving here). Shoving the "Spikeyball"(I like saying that) and directing Alphas one way or another using the B-Field is a stimulating mental exercise that butts against the physical construct of the machine as we know it. In the mean time I think we should stick to simpler things like: Is it better to eliminate the nubs between coils, and go with vessel/wall based mounts to feed power to each coil thus leaving no heating/loss target between them? Asked before, but not explored so well (unless Dr. Nebel hasn't told us that is part of WB8 design).
As you said, let's get the baseline machine working first.
Dan T - Back to your Pressure cooker experiment. How about enhancing your microwave donuts with some copper turns and another power source to feed them? You could even add a Pot and play "dial-a-field", may make for some great video. I looked at your pics as you upped your B Fields by changing magnets, and as you noted, there is a definate difference in the plasma configuration (Spikeyness). Recombinations given, (you must have free electrons and ions to do so), there seems to be validity in a better understanding to exploring the visual compression effect induced by the combined magnetic fields. This came to mind while re-reading Dr. Bussard's "PHENOMENOLOGICAL MODELLING OF POLYWELL/SCIF MULTI-CUSP INERTIAL-ELECTROSTATIC CONFINEMENT SYSTEMS" (DTIC AD-257944) paper where it discussed the device transition point from Mirror to Whiffleball.
Piling on PIDs can work if the time constants are adjusted right.
The core process gets fast control (what ever that means in context), the next level up slow, and the level above that slowest.
Small machines are hard because control of the core process may be MHz. The fuel flow in 10s of KHz. And other things that may need control would have to be in between.
Bigger machines can be slower. For instance fuel flow control might be adequate at the 100 Hz (Bode 3 db) level. Grid control might be adequate at 100 KHz.
Where you do get chaos (without feed forward) is when all the things needing to be controlled operate on separate controllers with similar time scales.
And of course there are speed of light considerations where fast processes that are not co-located (relatively) in space need to be controlled.
But of course without knowing what has to be done it is hard to determine how to do it. It could be easy (POPS, electron injection, grid voltage, fuel flow - which forms a reasonable hierarchy) or it could be difficult due to additional rqmts (keeping the grid voltage exact to stay on the pB11 resonance peak in the face of varying POPS frequencies caused by well variations - a kind of re-entrant or recursive problem).
Of course I have worked out some of this in preliminary paper research - for instance staying on the pB11 resonance peak using SSB phase control methods. But integrating all that with lots of frequencies running loose could get very tricky. And then you have the problem of extracting a useful signal in the face of 100 MW of noise, corona, HV arcing, delays caused by the separation of detectors from the process [time of travel]. etc. etc. etc.
I would love to be working on this - a very interesting problem.
The core process gets fast control (what ever that means in context), the next level up slow, and the level above that slowest.
Small machines are hard because control of the core process may be MHz. The fuel flow in 10s of KHz. And other things that may need control would have to be in between.
Bigger machines can be slower. For instance fuel flow control might be adequate at the 100 Hz (Bode 3 db) level. Grid control might be adequate at 100 KHz.
Where you do get chaos (without feed forward) is when all the things needing to be controlled operate on separate controllers with similar time scales.
And of course there are speed of light considerations where fast processes that are not co-located (relatively) in space need to be controlled.
But of course without knowing what has to be done it is hard to determine how to do it. It could be easy (POPS, electron injection, grid voltage, fuel flow - which forms a reasonable hierarchy) or it could be difficult due to additional rqmts (keeping the grid voltage exact to stay on the pB11 resonance peak in the face of varying POPS frequencies caused by well variations - a kind of re-entrant or recursive problem).
Of course I have worked out some of this in preliminary paper research - for instance staying on the pB11 resonance peak using SSB phase control methods. But integrating all that with lots of frequencies running loose could get very tricky. And then you have the problem of extracting a useful signal in the face of 100 MW of noise, corona, HV arcing, delays caused by the separation of detectors from the process [time of travel]. etc. etc. etc.
I would love to be working on this - a very interesting problem.
Engineering is the art of making what you want from what you can get at a profit.
Sorry, I wasn't clear in reference to chaotic control. Brain faster than fingers. I was thinking of alternate control methods per esempio Stochastic Control. If you can't control the actual feedback, set boundaries and let the system wander within.
I fully agree that this is a multilayer and integrated control problem that would be fascinating. In the end we can hope for a self balancing design where you can "move the rods" or "adjust steam flow" and everything else follows.
I fully agree that this is a multilayer and integrated control problem that would be fascinating. In the end we can hope for a self balancing design where you can "move the rods" or "adjust steam flow" and everything else follows.
I'm an old fission guy myself. Mare Island, Idaho, DLG(N)-25.ladajo wrote:Sorry, I wasn't clear in reference to chaotic control. Brain faster than fingers. I was thinking of alternate control methods per esempio Stochastic Control. If you can't control the actual feedback, set boundaries and let the system wander within.
I fully agree that this is a multilayer and integrated control problem that would be fascinating. In the end we can hope for a self balancing design where you can "move the rods" or "adjust steam flow" and everything else follows.
===
Just don't open the steam valves or yank the rods too fast.
Engineering is the art of making what you want from what you can get at a profit.
-
alexjrgreen
- Posts: 815
- Joined: Thu Nov 13, 2008 4:03 pm
- Location: UK
I wonder if it might be practical to create a college demo. What I have in mind is a 4 inch diameter glass round bottomed flask with six 1 1/2 inch diameter circular "trenches" dropped into it to create a polywell with its long diagonal along the centre line of the flask.ladajo wrote:In the mean time I think we should stick to simpler things like: Is it better to eliminate the nubs between coils, and go with vessel/wall based mounts to feed power to each coil thus leaving no heating/loss target between them? Asked before, but not explored so well (unless Dr. Nebel hasn't told us that is part of WB8 design).
As you said, let's get the baseline machine working first.
The magnet coils could then be dropped into the "trenches" from outside.
Not sure whether the vacuum rating would be adequate. Perhaps an engineer can comment.
Ars artis est celare artem.
At 10KV and up there is no such thing as a safe supply.KitemanSA wrote:Personally, I think one of the best thing this forum could do is design and market a "safe" power supply for these types of systems (fusor, Polywell). I suspect that one factor limiting the general exploration of these technologies is the concern over building the power supply.
BTW three nukes in a row in the comments - amazing. Welcome aboard lads.
Engineering is the art of making what you want from what you can get at a profit.
-
alexjrgreen
- Posts: 815
- Joined: Thu Nov 13, 2008 4:03 pm
- Location: UK
Perhaps it's not even necessary to inset the coils for a college demo.alexjrgreen wrote:I wonder if it might be practical to create a college demo. What I have in mind is a 4 inch diameter glass round bottomed flask with six 1 1/2 inch diameter circular "trenches" dropped into it to create a polywell with its long diagonal along the centre line of the flask.
The magnet coils could then be dropped into the "trenches" from outside.
Not sure whether the vacuum rating would be adequate. Perhaps an engineer can comment.
The e/m of the Electron Apparatus gives the basic idea. Just use six Helmholtz coils instead of two and cross-wire them.
This caught my attention:
The vacuum tube has a downward pointing electron gun in an evacuated bulb that has a little helium added so that the path of the electron in tube is visible. The helium gas added to the tube fluoresces when struck be the moving electrons and produce a bright, clear view of their circular path, Thus, the circular tracing of the electron path is undisturbed by the previously emitted electrons, contributing to a more accurate measurement.
Ars artis est celare artem.
-
alexjrgreen
- Posts: 815
- Joined: Thu Nov 13, 2008 4:03 pm
- Location: UK
In fact I'm going to claim that this fluorescence is what you can see here, which would mean we already know what a wiffleball looks like...alexjrgreen wrote:This caught my attention:
The vacuum tube has a downward pointing electron gun in an evacuated bulb that has a little helium added so that the path of the electron in tube is visible. The helium gas added to the tube fluoresces when struck be the moving electrons and produce a bright, clear view of their circular path, Thus, the circular tracing of the electron path is undisturbed by the previously emitted electrons, contributing to a more accurate measurement.
Ars artis est celare artem.
No need to "claim" it. That fact has been stated in prior papers (or maybe it was just a posting).alexjrgreen wrote: In fact I'm going to claim that this fluorescence is what you can see here, which would mean we already know what a wiffleball looks like...
-
alexjrgreen
- Posts: 815
- Joined: Thu Nov 13, 2008 4:03 pm
- Location: UK
Something like this?KitemanSA wrote:Which is why I put the word safe in quotes. What is needed is a power source that is as safe as experts can make it, without any of the potentially fatal mistakes that the non-expert may introduce.MSimon wrote: At 10KV and up there is no such thing as a safe supply.
A High Voltage Power Supply for your Lab
Ars artis est celare artem.