Further 'cheap' tests

Discuss how polywell fusion works; share theoretical questions and answers.

Moderators: tonybarry, MSimon

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Further 'cheap' tests

Post by D Tibbets »

Making the assumption that some limited funds are aviable to continue work with the current infrastructure EMC has, what tests can be done without further large capital investments (other than new magrids)?

Increased power to test reaction rates -vs- voltage/current. Magnetic field strength effects on wiffle ball size and tightness (loss rates)

Mildly larger diameter (if the vacuum chamber is large enough) to look more at scaling.

Increased number of grids- 'dodecahedron(?)' leading to increased efficiency. Do the corner grids need to be the same size as the current 'face' grids. Can they be smaller and placed slightly closer or further away than the 'face' grids? Can square or truncated square grids work better than circular grids?

POPS

D-T, D-3He, or even H-11B fuel.

Developing computer models to test the above configurations.

etc???


Dan Tibbets


ps: which is the correct nomenclature - B11 or 11B?
To error is human... and I'm very human.

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

Scaling equations could be tested with a smaller Magrid in the same apparatus.
I know this does not help test the system any closer to the desired operational conditions but it would be a quick and relatively easy way to get another point(s) on the curve(s) to test the scaling laws.
In some ways it would be a better experiment than a bigger machine because it would change the least number of variables.
i.e. the vacuum chamber, electron guns, instrumentation, vacuum pumps, power supplies, control systems, etc. all remain the same AFAIK.
Even the predictive computer model should be easy to adjust to a smaller Magrid.

Other issues:
It also gives the team one more practice round near term to gain experience that could be useful in designing larger devices.
It could be a chance to give the apprentices a chance to do a round while the masters are working on the peer review and larger issues/devices.
It would even be a chance to train newcomers to the team so when the project takes off there will be more people already up to speed.
It is also less bureaucratic risk because no one expects better performance.
There would be nobody breathing down their necks wanting to know results every hour.
(Well fewer bodies and breathing less heavily anyway.)
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Post by D Tibbets »

I hadn't considered shrinkage. I'm guessing that measurements of electron confinement time, varous other plasma measurements could be usefull. But, measurement of neutron output would be problamatical. The 30? cm WB6 only produced ~3 detected neutrons in a sub microsecond period . Smaller sizes would result in smaller counts with greater uncertainity due to noise, unless the run time or drive energy was increased. Being able to have lots of runs to average the results with greater presion would also help.

Since the input power requirement possibly scales as the square of the size, the heating of the magnet would be less and the mass of the smaller magnet (to absorb the heat) would less. Would the compitition between the two allow for longer, shorter, or unchanged run lengths (where heat build up is the limiting facter)? The smaller size should be less demanding on the power supply.


Dan Tibbets
To error is human... and I'm very human.

rnebel
Posts: 144
Joined: Mon Dec 24, 2007 12:15 am

Post by rnebel »

Going smaller to check scaling usually isn't a real good idea. It usually gives you overly optimistic results. The concern is that as you go to higher field and larger size the dominant transport mechanism may change (the fastest mechanism always dominates) so going smaller may mask the limiting mechanism at high field and large size.

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

I was thinking maybe only 5-10% smaller.
This should give a large change in measurements due to the scaling exponent of 5 or 7.

For that matter if it would be possible to shoehorn one just 2% bigger into the existing chamber, would give a 10% signal which should be easily readable.

Another useful (in my opinion necessary) & hopefully cheap test (of the engineering not of the science) is to incorporate active cooling in the magrid.

I expect the magrid to already be vacuum tight, to mitigate wire insulation out-gassing.
Now it "just" :wink: needs to be plumbed for cooling gas or liquid.
This would allow far more shots, which would give a better chance of seeing neutrons, even from a smaller device.

I assume that emc2 sweated every last mm of size into the chamber already, but they had to leave some safety factors for unknowns.
Now that it works, a 2-5% larger magrid could possibly be squeezed out of those safety factors.
I also assume they pushed the chamber as large as possible too, with the hopes of a phase 2 test of scaling laws in the same chamber.
So, the extra chamber size might be there already.

If you want me to take a crack at making the machine shop drawings in AutoCAD just send me the details and give me the go ahead.
(Well, a guy can dream can't he?)
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

Torulf2
Posts: 286
Joined: Fri Sep 21, 2007 9:50 pm
Location: Swedem

Post by Torulf2 »

Maybe its better to make it only 5-10% bigger.

Aero
Posts: 1200
Joined: Mon Jun 30, 2008 4:36 am
Location: 92111

Post by Aero »

Guys, Changing the size of the magrid is not using wb-7. It is building a new machine of difference size. It will cost the same as another wb-7. Frankly, if it will fit, I'd vote for salvaging wb-6 and wb-7 to make wb-8, about twice the radius of wb-6 and wb-7, but if it will fit into the chamber ....
My other choice is to run as much current through wb-7's coils as it will tolerate. Remember, scaling is really R^3 * B^4. Increase the magnetic field strength by a few percent, say X percent, and you increase the overall "power" of wb-7 by a factor of four-sevenths X. If X = 7% then power increases by 4%, and if X=14%, power increases by 8% according to the scaling laws. That might tell something.
Aero

blaisepascal
Posts: 191
Joined: Thu Jun 05, 2008 3:57 am
Location: Ithaca, NY
Contact:

Post by blaisepascal »

Aero wrote: Remember, scaling is really R^3 * B^4. Increase the magnetic field strength by a few percent, say X percent, and you increase the overall "power" of wb-7 by a factor of four-sevenths X. If X = 7% then power increases by 4%, and if X=14%, power increases by 8% according to the scaling laws. That might tell something.
If scaling is R^3*B^4, then why will increasing B by X percent increase it only a factor of 4/7ths. I'd expect increasing the field strength by 7% would, by that scaling law, increase the power by 31%, as 1.07^4= 1.31, and increasing the field strength by 14% would increase power by 69%.

And with that large scaling factor, if it were easy to adjust the field strength of the wb-7 and make B-field and power measurements, it wouldn't surprise me if Dr. Nebel didn't already do that experiment. After all, it has been implied that they did multiple runs/day, for days. Surely they weren't running the same conditions over and over again.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

The power scaling isn't really too controversial. What needs to be proven is that losses can be held to roughly r^2, that brem can be beat, and that new problems don't arise at 1.5m radius.

I've began wondering lately how much value there would be in an uncooled 1.5m Polywell, basically WB-7 design but WB-100 sized. That would presumably cost a lot less than the $100M, and it sounds like the physics are fast enough that the results might be applicable to a cooled machine. Not sure how feasible that is from an engineering perspective, though, or exactly how much cheaper.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

blaisepascal
Posts: 191
Joined: Thu Jun 05, 2008 3:57 am
Location: Ithaca, NY
Contact:

Post by blaisepascal »

TallDave wrote:
increasing the field strength by 14% would increase power by 69%.
Remember, Bussard detected 3-4 neutrons per test. I'm not sure they would even bother looking for anything less than an order of magnitude.
And we have nothing beyond speculation as to what Dr Nebel detected. Here's a reasonable speculation:

Dr Bussard build WB-6 at the end of his contract, pushing things at or beyond the reasonable capabilities of his people and equipment. In the end, he detected 3-4 neutrons per test, for a short series of short tests.

The Navy, finding the WB-6 results intriguing, but not completely convincing, hires Dr. Nebel to, in essence, redo the WB-6 experiment, but this time to do it right. Dr. Nebel builds an identical device, the WB-7, but with a better power supply and better instrumentation. Because he has the time and money dedicated to this project,he doesn't have to cut corners as Bussard probably had to do. As such, they get better data.

Under such a speculation -- which seems to match what we know -- it's entirely reasonable to Dr. Nebel to have better neutron detectors, and/or run the reactor for longer for each test, and as such instead of detecting 3-4 neutrons/test, detect in the range of 30-40/test for the same nominal power output.

At 3-4 neutrons, it's low enough to the noise floor to trust only order of magnitude increases. At 30-40 neutrons, I can see looking for, and accurately detecting, 60% increases.
In any case, the power scaling isn't really controversial. What needs to be proven is that losses can be held to roughly r^2 and that brem can be beat.
How does brem scale? Is it with B or with R? If with B, it can be tested with a single-scale machine.

icarus
Posts: 819
Joined: Mon Jul 07, 2008 12:48 am

Post by icarus »

rnebel wrote:Going smaller to check scaling usually isn't a real good idea. It usually gives you overly optimistic results. The concern is that as you go to higher field and larger size the dominant transport mechanism may change (the fastest mechanism always dominates) so going smaller may mask the limiting mechanism at high field and large size.
rnebel: By "dominant transport mechanism", would I be correct in assuming that you are referring to the convective-turbulence transport processes of the plasma that are dominating the losses?

If so, what would be the current working model for the scaling laws of the turbulence processes dominating the losses? I can probably add something in this area, if that is necessary.

Of course, smaller/bigger geometry, stronger/weaker field, density ratios, energy ratios, etc, could all be possible experiments to validate those laws, if they are formulated correctly in the first place.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

and/or run the reactor for longer for each test,
I don't think it works that way. As I recall, Busard got fusion only at the quarter-millisecond where beta was approx 1. I don't think there was ever anything like the control necessary to extend that time in WB-7 (as Tom's analogy put it, it lacks a carburetor and ignition timing), though if anyone knows better I'd be happy to be wrong.
How does brem scale? Is it with B or with R? If with B, it can be tested with a single-scale machine.
I think it only becomes an issue when power is somewhere near Q. Also, I'm not sure the losses are even measurable in a WB-7 sized machine, where fusion power is only .001 watts. I'm also not sure a small machine can be driven to the well depths necessary for the p-11B brem tests.

Anyways, as Rick notes above there are a lot of problems with extrapolating results from smaller machines. I recall his earlier comment about how the transport loss equations in tokamaks changed every couple of years as they built bigger ones.
Last edited by TallDave on Sun Sep 28, 2008 7:31 pm, edited 1 time in total.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

Roger
Posts: 788
Joined: Fri Jul 06, 2007 2:03 am
Location: Metro NY

Post by Roger »

TallDave wrote:
I've began wondering lately how much value there would be in an uncooled 1.5m Polywell, basically WB-7 design but WB-100 sized. That would presumably cost a lot less than the $100M,
Me too. A cost guess... 1.6M size DD, pulse mode.. 25 million ?

The upside is, well.... net power... I think it makes the history books, no?
Downside, scaling is wrong, or worse....
I like the p-B11 resonance peak at 50 KV acceleration. In2 years we'll know.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

/agree with Roger.

If the physics are so fast the pulse can prove the concept, could be a cheap way to bet on this.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Post by Art Carlson »

blaisepascal wrote:How does brem scale? Is it with B or with R? If with B, it can be tested with a single-scale machine.
P_Br [W/m^3] = [n_e / 7.69 \times 10^18 m^-3]^2 * T_e[eV]^1/2 * Z_eff

The n^2 scaling of the power density is the same as for the fusion power, so, assuming a fixed temperature, the ratio of bremsstrahlung power to fusion power is a constant. For D-T the ratio is 0.7 percent, for p-B11 it is about 1.7! These numbers assume the Z_eff corresponding to pure fuel, which is unrealistic. ITER, like other tokamaks, will have to fight to keep Z_eff from getting too large, and is expected to operate with a few percent bremsstrahlung losses. I would expect a polywell to have a harder time, since the coils are very close to the plasma and there is no divertor.

Post Reply