Aviation Week on the Lockheed Skunkworks CFR

Point out news stories, on the net or in mainstream media, related to polywell fusion.

Moderators: tonybarry, MSimon

hanelyp
Posts: 2261
Joined: Fri Oct 26, 2007 8:50 pm

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by hanelyp »

Stability isn't strictly from convex vs. concave magnetic fields, but from magnetic field gradient as the plasma pushes out. Convex fields are associated with stability because they usually have favorable field gradient, while concave fields usually have an unfavorable gradient. Where the magnetic field passes around the inside magnetics may have a favorable gradient on both sides of the plasma, despite the one side being concave.
The daylight is uncomfortably bright for eyes so long in the dark.

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

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by D Tibbets »

hanelyp wrote:Stability isn't strictly from convex vs. concave magnetic fields, but from magnetic field gradient as the plasma pushes out. Convex fields are associated with stability because they usually have favorable field gradient, while concave fields usually have an unfavorable gradient. Where the magnetic field passes around the inside magnetics may have a favorable gradient on both sides of the plasma, despite the one side being concave.
I disagree, based on my flimsy knowledge. The stability of a B field against edge instabilities, macro instabilities, or what ever name is used is at its most basic a potential energy question. With aconvex field against the plasma, a local instability may occur due to various chaotic movements of the local plasma. This can push into the plasma and form a bulge. But this requires the application of kinetic energy as you are building potential energy in the area. The natural tendency is for the potential energy to decrease. This returns KE back to the internal plasma. More importantly it implies that for the bulge/ instability to continue to grow, energy has to be applied. The energetically natural thing is for the bulge to flatten out because this represents a lower potential energy state- it is more stable.

With the B field concave towards the plasma, the reverse is true. As the bulge grows the potential energy is decreasing, the natural tendency is again for the potential energy to decrease and it does so by expanding the bulge. Eventually there is such concentrated kinetic energy the magnetic insulation fails and a prompt rupture of the magnetic containment results and a concentrated packet of plasma rushes outward to hit a wall.

I like the water sieve analogy. A bowl shaped size (fine mesh) allows water to seep through. Once most of the water is gone a film is left that is quasi stable due to water tension. If the sieve is turned bowl side up, the water will not form a droplet in the center, any tendency for water to move (gravitational potential energy in this case) is towards the edge. There water could drop off and this is equivalent to the cusps. This is more diffuse as the flow is towards larger surface area. The water is still lost, but the mechanism is different and the magnitude over time may also be different. This illustrates also, that the only way to avoid concave B fields is to have cusps (or a perfect spherical monopole, which is outside of accepted physics).

With the sieve bowl pointing down, now the gravitational potential energy is converted to KE of the water collecting at the lowest point- the center of the bowl. It seep through the sieve and forms a drop, this drop continues to grow until the surface tension is overcome (the magnetic equivalent) and the drop falls = a macro instability.

What becomes confounding is the plasma- B field orientation in a complex system where inside and outside are mixed up. In a quasi spherical geometry the outside is always- well, outside. It is easy to compare. In the Polywell with the Wiffleball condition met, it is also easy to compare as the inside is essentially B field free with a well defined transition boundry.
In a more complex geometry like the Lockheed design seems to be. The concave vs convex picture has to be considered in the local area. Inside all of the cusp geometry magnets the convex B fields are always stable towards the more central plasma. But, after traversing a cusp, the shape of the fields relative to the local plasma reverses. In these peripheral areas the convex B field is still orientated towards the magnets, but they are now inside of the plasma, the B fields are now concave outside of the local plasma and edge instabilities are a natural consequence- the plasma can squirt through the B field much like the water dropping from the center of a sieve. This plasma moves further towards the periphery of the machine- eventually hitting a surface.

What I see as a modifier for this situation compared to a tokamak, where essentially all of the plasma faces concave fields always (in the direction towards the periphery of the machine), or cylindrical solenoid mirror machines (which alternate the field curvature along the length) is the relative densities. The edge instabilities grow exponentially with density. As such, inside the magnets of the Polywell , and the inner magnets of the Lockheed machine the density is stable against edge effects/ macro instabilities. The density with a neutral or non neutral plasma can thus be greater if the macro instabilities is the limiting containment issue. This is at least in part why the Polywell can operate at higher densities than a Tokamak. This is convenient, and actually essential as this allows for the useful high densities that leads to high Beta conditions with corresponding drops in cusp losses. In the Lockheed design, the plasma (mostly electrons presumably- if there is a potential well*) that escapes through a cusp is now exposed to concave B fields (concave on one side and convex on the other), This may be of much less significance though because if Wiffleball type high Beta confinement is achieved, this escaped plasma may be as much as several hundred thousands times less dense. The macro instabilities may not contribute much to the overall balance .

Also, with layered magnets, the escaped plasma may undergo macro instabilities, but in doing so move further towards the peripheral magnets where they would again face convex fields towards the center. This onion skin layering of magnets might mitigate macro instabilities , while also, perhaps aiding in recirculation, or at least operator interventions.
As with with Tokamaks,local inputs of energy might control edge instabilities without depending on the always relaxed condition of always convex B fields. The Polywell though, is by far the simplest possible solution. As such it seems to me to be the most promising, if it can be made to work. If additional layers of control have to be imposed on the squirmy plasma, the complexity and cost rapidly increases, possibly to the point where there is little to no advantage over tokamaks, or at least spheromaks.

In magnatized plasma machines like tokamaks, the plasma creates its own B field and is embeded in it. As such all of the plasma is exposed to both concave and convex B fields. Due to the torus shape the effects may not be equal, but simplistically, if a local turbulence leads to a macro instability pointed towards the core of the plasma the confinement is not directly effected. Only those macro instabilities that point towards the outside are significant. This may be an even more important consideration for spheromaks with their convoluted B fields.


* If the Lockheed machine does not have a potential well/ non neutral plasma with excess energetic electron injection , then ion containment will probably be due to a combination of ion cusp losses and ion ExB losses. And the ion ExB losses may actually dominate. Bussard insisted in his Google talk, that for a small machine, especially a small machine with adequate fusion output and thus density, ion ExB losses had to be greatly reduced, otherwise the machine could not be any smaller than a tokamak and most, if not all of the advantages would be lost.

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

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

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by D Tibbets »

TheRadicalModerate wrote:Dan--

Here's the main patent. There are others mentioned up-thread, but I haven't gone through those.

The patent never mentions a potential well, although it does refer to a magnetic well. (I just did a search through the text.)
The patent application mentions lots and lots of embodiments... how do you patent that?
In the patent the only heating mechanism mentioned is neutral beam.

They do mention varying the plasma shape asI have pointed out in the past.

"For example, the central confinement well can be expanded into a more spherical shape, increasing its volume and suitability for non-thermal fusion schemes. Another example is the stretching of the central region into a more elongated cigar shape, perhaps for easier integration into aerospace vehicles or for easier power conversion or surface wall effects such as breeding. "


"The addition of two axial ‘mirror’ coils (i.e., mirror coils 160) serves to decrease the axial cusp losses and more importantly makes the recirculating field lines satisfy average minimum-p, a condition not satisfied by other existing recirculating schemes. In some embodiments, additional pairs of internal coils "

The external magnets do compress the field lines as they loop around between cusps. This may help in recirculating electrons and fuel ions. But it may have more importance for recirculating fusion ions- primarily alphas. They mention in the video that it is a D-T fueled ignition machine. As per Dr Nebel, this implies that the alphas mush hang around long enough to thermalize with the fuel ions. As per Dr Nebels, the alphas in the Polywell last only about a few thousand passes- a few thousand meters of travel distance, before they escape through a cusp. In the Polywell, they would be expected to hit a wall and not be recirculated. The B field loop is to large- the radius of the B field line would hit a wall before the alpha could loop around. With external magnets, perhaps the mildly lower energy alphas from D-T fusion may be recirculated through neighboring cusps multiple times to provide the necessary dwell time for thermalization with the fuel ions. Defiantly no direct conversion, only heat transfer for powering electricity generation.

The issue of ion ExB diffusion is not answered. They do mention something about ion gyroradius multiples needing to be met. Somewhere in the 5-7 range? This is actually comparable to the electron gyro radius clearance needed in the Polywell. But, being ions, the clearance would have to be greater by a factor of ~ 60 or more- either increased B field strength, or greater machine size. This is a major consideration that Bussard pointed out . How they can claim a small high density machine like the Polywell without the fuel ions being decoupled from Magnetic ExB loss mechanism s a mystery. That they apparently only propose D-T fusion, a target plasma temperature of perhaps 20 KeV versus 100 Kev for D-D fuel mitigates the ExB jump distance issue somewhat.

The original video which started all of this-

http://aviationweek.com/technology/skun ... or-details

"The CFR is expected to have a beta limit ratio of one. “We should be able to go to 100% or beyond,” he adds."

"This crucial difference means that for the same size, the CFR generates more power than a tokamak by a factor of 10. This in turn means, for the same power output, the CFR can be 10 times smaller. The change in scale is a game-changer in terms of producibility and cost, explains McGuire. “It’s one of the reasons we think it is feasible for development and future economics,” he says. “Ten times smaller is the key."

“We also have a recirculation that is very similar to a Polywell concept,” he adds, referring to another promising avenue of fusion power research. A Polywell fusion reactor uses electromagnets to generate a magnetic field that traps electrons, creating a negative voltage, which then attracts positive ions. The resulting acceleration of the ions toward the negative center results in a collision and fusion. "

The prototype would demonstrate ignition conditions and the ability to run for upward of 10 sec. in a steady state after the injectors, which will be used to ignite the plasma, are turned off. “So it wouldn’t be at full power, like a working concept reactor, but basically just showing that all the physics works,” McGuire says.

https://www.youtube.com/watch?v=JAsRFVbcyUY

The video mentions plasma heating via microwaves. Admittedly, I though it had mentioned a potential well, but no... Mention of D-T reaction, lithium breeding, etc. And, is this supposed to fit on a truck or in a plane? Lots of luck on that.


So, my current appreciation for the Lockheed design is that it is like a Polywell in that it confines plasma in a cusp magnetic field and can have a central point- 'quasi spherical'. Recirculation of fuel plasma, not just electrons is probable. recirculation may be more efficient (has to be) and also applies to escaping fusion alphas- in order to reach ignition conditions. This removes the opportunity for direct conversion.
Heating is by conventional neutral beam/ microwave heating. No mention of a potential well, and thus no decoupling of fuel ion containment from magnetic containment, no central confluence possible, and no monoenergetic conditions possible. MHD stability/ edge stability issues are probably not significant for reasons I postulated in my previous post.
No possibility of advanced fuel usage except small chance of D-D.

ExB ion issues seem paramount. If Bussards, Nebels, , etc. understandings are pertinent, I am leery of Lockheeds expectations.

This current Lockheed design, based on the limited information released, and my understanding seems tentative at best.

They have stressed though that they are concentrating on magnetic confinement issues. If they introduce a electrostatic potential well as part of the design at a latter stage, the idealized advantages of the Polywell could be met, at least for a thermal machine driving a gas turbine. Direct conversion possibilities would seem to be excluded. And, I am uncertain of Bremsstruhlung consequences.

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

hanelyp
Posts: 2261
Joined: Fri Oct 26, 2007 8:50 pm

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by hanelyp »

Relevant to stability of the CFR configuration: When I was running simulations of the configuration in XOOPIC, a test I'd sometimes run is set for steady plasma injection and run until I get a blowout. The margins between the minimum that looked to me like beta=1 and blowout could be insane, the latter containing many times the plasma of the former. As extra plasma was added beyond the minimal beta=1 condition, the plasma pushed out into stronger magnetic fields closer to the coils. When blowout was reached it was always symmetric, not on just one end. And after blowout the system settled into a new stable leaky configuration until plasma quantity dropped back to much less than triggered the blowout.

The one simulation result that concerns me about the configuration is that I always got a fair density of near neutral plasma passing outside the inner magnets, where it would have trouble with the magnet supports.
The daylight is uncomfortably bright for eyes so long in the dark.

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

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by D Tibbets »

First, what does CFR stand for?

In a cusp configuration, when Beta= 1 is exceeded, I would not expect a blow out of impressive proportions, the effective cusp hole size to total surface area would open up in a near symmetrical fashion comparable to the ratio as Beta =1 is approached from the low side. Some fluctuations may occur due to leakage rate being near input rate, as may have been seen with Dr Parks Mini-B machine. A rapid "blow out" would represent either a massive plasma pressure being much greater than magnetic pressure (Beta>>1), plasma density variations locally that might be seen with POPS type effects-moving plasma pressure waves, or when the plasma pressure has pushed out the B field to the extent that metal surfaces of the magnets are exposed and the plasma shorts directly to the magnet cans.

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

hanelyp
Posts: 2261
Joined: Fri Oct 26, 2007 8:50 pm

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by hanelyp »

A quirk of the configuration I was modeling is that the magnetic field starts with field lines running in on end, around the outside of the inner magnets, then out the other end. When pressure from the plasma injection rose sufficiently there was a magnetic reconnection on those field lines so they no longer ran end to end, but had cusps between coils that plasma could stream through. The field didn't just become leaky through existing cusps, the configuration changed to produce new much larger escape cusps. I haven't simulated if the field recovers if plasma injection is cut off after blowout.

As for beta=1, that's simply a statement of plasma pressure = magnetic pressure. But when the magnetic field becomes stronger moving outward from the center you have to ask where that boundary is encountered. So beta=1 can cover a range of plasma volumes and pressures within a machine.
The daylight is uncomfortably bright for eyes so long in the dark.

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

Re: Aviation Week on the Lockheed Skunkworks CFR

Post by D Tibbets »

hanelyp wrote:...

As for beta=1, that's simply a statement of plasma pressure = magnetic pressure. But when the magnetic field becomes stronger moving outward from the center you have to ask where that boundary is encountered. So beta=1 can cover a range of plasma volumes and pressures within a machine.
Where the Beta=1 border may lie has been a question I have pondered (well...lightly wondered about)
I suppose you could calculate the border radius based on plasma pressure and B field pressure using a drop off rate of d^3 for the B field gradient in opposing magnet configurations. That takes work though, so I was appreciative when Dr Park;s revieled the border was about 1/2 the diameter of the "Mini-B machine- ~ 10 cm. Still plenty of room for the field to be pushed out before the cans was encountered (depending on your expectations for electron ExB diffusion blurring the margins). So, I expect the cusps opening back up is the dominate variable effecting plasma loss well past Beta=1. What is 'well past Beta=1'? Well, perhaps 2 or 3. This is just a guess. It depends on how conformal the cans are to the B field generating wires, and on the thickness of the insulating/ cooling layers within the can are.

Of course, in a perfectly spherical magnetic monopole, I think that Beta >1 means magnetic confinement is fully overcome, like a balloon bursting. In a cusp geometry though, the Beta=1 condition is where the cusp loss cones are smallest, before opening back up. Incidentally the Grad described effect of the B field gradient being greater than the defined gyroradius (of the electron) at the defined radial border- meaning the electrons effectively bounce off the border like a hard surface instead of being entrapped in a near circular local and spiraling orbit. Interesting things are occurring as Beta approaches one in a quasi spherical cusp geometry.

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

Post Reply