Aviation Week on the Lockheed Skunkworks CFR

[crowberry]

That suggests the dynamic field variations are being exploited as the primary compression mechanism (see my Oct 21 post on this thread).

A Polywell wiffleball can, in theory, be static. This, maybe not...

Maybe they are interlacing B field oscillations (coil currents) with E field oscillations (coil case potentials). The limiting factor for oscillation frequency is the inductive reactance of the coils, as MSimon pointed out years ago. The potentials can be varied much faster than the currents.

Use agile E to make up for sluggish B?
An E "holding action" on the speedy particles until the B "sledge hammer" can be dropped?

(Sorry for all the plasma physics jargon :\)

If Lockheed Martin would use electric fields, then it most likely would be mentioned in their patent applications, but that is not the case. For this reason I doubt that they plan to use electric fields like in a Polywell.

[choff]

In the meanwhile, can't go to EMC2's webpage, maybe the pending updates aren't going to be pending for much longer.

[DeltaV]

If Lockheed Martin would use electric fields, then it most likely would be mentioned in their patent applications, but that is not the case. For this reason I doubt that they plan to use electric fields like in a Polywell.

Unless the E part was separated into a classified patent...

[mvanwink5]

I thought the voltage gradient resulted from the use of a neutral beam, then loss of positive ion due to poor positive ion magnetic retention and good electron retention ( thus creating the voltage field gradient. As the voltage gradient builds, positive ions begin to be retained. In fact, that is what I thought EMC2 was migrating to, instead of biasing the magnetic toruses.

[D Tibbets]

I thought I heard something about a potential well somewhere, but I cannot reference any source. Perhaps I misinterpreted the mention of a magnetic well in the Aviation Week article.

They do mention high Beta and that implies high density, especially if D-T fuel is envisioned. The target temperature is less than any other fusion fuel, and thus the density contribution to pressure is relatively greater. If a Wiffleball effect is envisioned, along with small size; then high density is required for useful fusion output. All of this leads me to think that densities will be similar to Polywell. D-T fuel is easier so smaller density/ weaker B fields may be used, but I suspect not much change. Perhaps it would work at densities of 10^21 instead of 10^22/ meter cubed, and still get enough fusion energy from D-T, In any case, the machine is advertized as 10 times smaller than tokamaks (I assume in linear dimensions). Ion ExB losses would therefor be at least 100 times greater per unit of surface area, assuming similar temperatures and 10X greater density. This would seem to be pushing magnetic ion containment. I suppose that comparisons must include ExB diffusion issues and edge instabilities. If the edge instabilities contibute more to required tokamak sizes, there may be some manuvering room. If there is an electrostatic potential well that confines ions away from the magnetic 'walls' this concern is greatly lessened..

If there is an electrostatic potential well, it would confine fuel ions, but not fusion ions unless the cusp confinement, even under Wiffleball conditions is much better. Alternately , an additional layer of magnets might constrain looping plasma- fuel ions, fusion ions, and electrons such that the B field lines they follow are not so large that they intercept the vacuum vessel wall. Such would have to be the case, I think, for a cusp magnetic system to contain the high energy fusion ions long enough for them to thermalize with the fuel ions and reach ignition conditions . Ignition is claimed for the machine. If the fusion ions thermalize with the fuel ions, the plasma has to be fully thermalized. There would be no monoenergetic benifits.

There may be a central dense core with essentially Wiffleball confinement , the rest of the 'recirculating plasma' (fusion ion, fuel ion and electron populations) is at significantly lower densities so that ExB diffusion, and edge instabilities are less painful as only a tiny portion of the plasma is exterior to the central core at any given time (ratio of the Wiffleball confinement). If the densities inside the core and in the looping flows are similar, I don't see how the machine could be smaller than tokamaks, without also having proportionately greater ExB losses. Edge instability management would be another issue. They mention good edge stability in the core (?), but I do not think this could apply outside the core. With magnetic fields on both sides, and the plasma itself being magnetized in this region (?) edge instabilities are unavoidable.

Ignition implies relatively long confinement times - some primary magnetic confinement time plus very efficient recirculation of cusp losses. I don't know if fuel ion annealing, and eventual fusion ion annealing is anticipated, but I suppose such might be invoked, though I don't see how it would work out. In the Polywell Bussard claimed that ion thermalization lifetimes was longer than the time to fusion. If ignition is part of the Lockheed machine, such cannot be claimed. The fuel ions have to hang around long enough to thermalize with the fusion ions, which take longer because of the initial high energy of the fusion ions (thermalization occurs at a rate of ~ 1/ KE^2). In an ignition machine, the total population of ions is contained. This implies a need for some means to extract the fusion ash, otherwise the system becomes poisoned (fusion ions once thermalized with the fuel ions do not contribute further to heating, but they do continue to contribute to cooling of the plasma). They also dilute the portion of the plasma that can be fuel ions. Tokamaks use unproven diverter technology to remove the heavier fusion products- I think somewhat like a mass spectrometer. How would this machine handle this question? Only pulsed operation may be possible.

The Lockheed device as a modification of the Polywell concept, which is narrowed cusp magnetic confinement of excess electrons with favorable B field geometries, coupled to resultant potential well electrostatic confinement of fuel ions (not fusion ions), is easy to follow based on what I have learned about the Polywell. The additional layers of magnets, and much (?) prolonged secondary magnetic containment and recirculation of electrons and ions(?) with the additional resultant loss mechanisms (ExB and edge instabilities) seems to introduce multiple competing mechanism that are difficult for me to envision as being overall beneficial.

If there is an electrostatic potential well for fuel ion containment, then better electron recirculation of say 99% rather than 90% may be useful. It depends on recirculation versus electron/ plasma injection interactions. and only if the ExB and edge instabilities of this external (to the core) electron recirculation does not consume any gains. I don't know, but this approach may share somewhat with pulsed field reversed configuration concepts.

Other than the issue of an electrostatic potential well, the idea that this is an ignition machine opens a large can of worms, that is avoided in the Polywell, at least for a steady state approach.

Dan Tibbets

[D Tibbets]

I thought the voltage gradient resulted from the use of a neutral beam, then loss of positive ion due to poor positive ion magnetic retention and good electron retention ( thus creating the voltage field gradient. As the voltage gradient builds, positive ions begin to be retained. In fact, that is what I thought EMC2 was migrating to, instead of biasing the magnetic toruses.

I think EMC2 concepts have not changed, except possibly for further reducing any cold electron plugging near the cusps that might be associated with escaping electron stoppage and turnaround near the magrid due to positive charge on the magrid. I personally, in my inexpert openion , think this may be an issue, but perhaps best addressed by a compromise as opposed to only one or the other option. Especially if electron injection efficiency dominates the picture.

Energetic electron injection is the key. This is either by having the acceleration grids on distant E- guns, or near the magrid radius. The final electron acceleration is the same. The details are different but the end result is the same. That is, excess energetic electrons produce the ion confining, centrally directed potential well due to space charge effects.

Remember that the potential well is effected somewhat by the presence of ions. The space charge electron potential well dominates the ion motions, but locally (collisions) the ions tend to drag electrons along with them. This changes the shape of the potential well , which Bussard mentioned in his Google talk. Applied from the other side- that is an excess of energetic ions escaping magnetic confinement faster than electrons (presumably due to differences in ExB diffusion rates) leaving behind an excess electron population which results in an equalibrium that then retards further ion escape, is reasonable from a space charge perspective. But the local dragging of electrons by the ions would lead to some increased electron magnetic losses also. Im not sure of the magnitude, but I suspect it would be intolerable. I think this is the picture that A. Carlson had when he argued that ions would follow electrons out of the machine. Remember that ExB diffusion refers to gyroradius jumps in a magnetic field. This is dominate under certain conditions of field strength and particle energy and density. The non gyroradius jumping collisional effects can become more dominate and mixed with certain conditions. The gyro radii jump per collision can dominate, but does not have to be the only diffusion mechanism. It is a mixed picture, thus electrons could be dragged through the magnetic field almost as fast as ions if the ions are the dominate species as the magnetic field region is reached. I think the establishment of a negative potential well- importantly a deep negative potential well through ion depletion is impractical, if not impossible. The ions have much more inertia/ momentum than electrons so transient local effects (collisions between ions and electrons) have more effects on electrons than ions. The ions can drag electrons around, but not so much for electrons dragging ions around. The electrons have to gang up on the ions (space charge effects) to get the ions attention.
Also, The ions have to be at minimal energy at the edge from the start, preferably not encountering the pushed out B field at all.

PS: Under Wiffleball conditions the cusp holes are essentially equal to the gyroradius of the charged particle. I'm not sure how the profile of the ion gyroradius with retained KE compared to the fast electron equates into cusp hole size for each, but for at least up scattered ions or initially energetic ions the cusp loss hole size is larger, perhaps matchnig the relative hole size for electrons. So this non diffusion mechanism would allow for both ion and electron losses at equal rates, no gradient would build up or be maintainable without the significant electron excess induced potential well. This would make the development of a deep potential well through ion depletion even more difficult. Cusp losses normally dominate electron losses (by a ratio of up to ~ 100:1 over ExB losses in the Polywell. Ion losses without a retaining potential well may have similar loss rates via both cusps and ExB. Any slightly increased ion losses (up to 2X?) would leave behind a negative space charge, but with the relatively huge cusp holes the electrons would exit the cusps more rapidly. Even without the ions dragging the electrons through the B fields, this vulnerable cusp hole loss mechanism would lead to equilibrium at probably very low potential well values. Always the electron injection must exceed the electron losses, along with the replacement of slower ion losses, in order to build up to Wiffleball conditions, then the equilibrium must be maintained. In this instance the chicken must always come before the egg.

I don't think the injection of neutral low energy plasma in the EMC2 mini B machine created a potential well through subsequent ion depletion. This pulse of neutral relatively cold plasma was added on top of an existent high energy electron beam. The negative bias of the plasma was present at the start, and was maintained briefly until the excess electrons were lost through the cusps faster than they could be replaced. Through the electron losses and cooling, the Wiffleball pressure could not be retained, Presumably at all stages of these tests the ion losses were less than the electron losses., at least until the potential well was completely washed out- or rather a new equilibrium at lower density was reached (which was much lower due to the loss of Wiffleball electron confinement).

Dan Tibbets

[mvanwink5]

Yes, and Park has shown that just to start up and achieve cusp plugging, high power electron injection is needed first. Still, the Polywell's magnetic torus array does not need to be biased as the well electron population could just be increased to achieve the same thing, no?

I don't see Lockheed showing electron injection for start up... yet? So, I can see you would have issue or at least doubts that Lockheed would be able to achieve an effective electron over population created potential well with just neutral beam injection.

[D Tibbets]

Yes, and Park has shown that just to start up and achieve cusp plugging, high power electron injection is needed first. Still, the Polywell's magnetic torus array does not need to be biased as the well electron population could just be increased to achieve the same thing, no?

I don't see Lockheed showing electron injection for start up... yet? So, I can see you would have issue or at least doubts that Lockheed would be able to achieve an effective electron over population created potential well with just neutral beam injection.

"Shudder" I hate the term of 'Cusp plugging' in the Polywell. I equate this to WB5 experience. The cusps loss cones are shrunken , at least in relation to the plasma confining surface area, but there is nothing blocking the cusps like a plug of cold electrons (or a negative electrode located very near the midplane of the Magrid radius). This does not work, as demonstrated by WB5.

I'm not sure a Willeball condition cannot be achieved with the injection of a neutral beam only, it may or may not be somewhat more difficult. But, I suspect that a deep potential well is impossible without excess energetic electron injection over whatever neutral beam injection may be used. A deep potential well might be achieved with excess e- gun input and cold ion gun input, or with the energetic e- gun and corresponding cold neutral beam input. The dynamics are somewhat different, but the results are the same. The ion and electron average temperatures are the same, but the distribution of the energy of each species is determined by the establishment at some point of a DEEP electrostatic potential well. Without this you could have a hot mixture of ions and electrons under Wiffleball confinement (high Beta), but several beneficial aspects of the Polywell are absent: ion annealing would not be possible, Bremsstruhlung suppression is absent, no confluence, suppression of ion ExB diffusion would not occur, no simple heating of ions (in the core region) by the potential well, and probably several other issues. As such, D-T fusion may be the only option without these considerations in an otherwise similar machine.

I don't know just what the Lockheed approach is, but the one illustration shows two injectors, one a neutral beam injector, and the other an electron beam injector (if I remember right). [EDIT]. The illistration I was thinking of shows two injectors , both labeled an neutral beam injectors. different colors are used so one might be a ion injector while the other is an electron injector. Together they would serve as a neutral beam injector. But, if the electron gun was turned up in current (and voltage?) it could create a potential well. There is flexibility in what they could do. If the ion gun voltage was also turned down, the results might be similar to a Polywell, provided the guns were positioned very near a presumed Wiffleball border.

Dan Tibbets

[hanelyp]

My impression of the wiffleball effect is it doesn't care about plasma charge, just beta approaching 1 with a suitable cusp.

[D Tibbets]

My impression of the wiffleball effect is it doesn't care about plasma charge, just beta approaching 1 with a suitable cusp.

True. In the Polywell though- at least as described in the patent application, the pressure against the confining B field is primarily from the electrons. This may have some impact on the B field strength nessisary to contain a given non neutral plasma, specifically the ion pressure in the center due to confluence of hot ions. It is all related to the potential well. The ions are cold on the edge so do not exert much pressure against the B field. The electrons are hot though, so at nearly equal total numbers compared to the ions, it is the electron pressure that pushes against the B field and inflates the Wiffleball. It seems counter intuitive, but apparently if there is significant ion confluence (central focus) the core ion pressure can actually be considerably higher than the pressure exerted on the Wiffleball border (by the electrons) the average pressure taken over the entire volume is identical (?), but the working pressures on the edge and core are dominated by electrons and ions respectively.

Since the fusion rate is the square of the density, and the 4th power power of the confining B field in a homogenous ion mix over the internal volume, the central focus/ central hot ion density increase leads to the local core fusion rate being shifted to even higher levels. It depends on the relative increased central density and the corresponding volume of the core. In the patent application, it is pointed out that theoretically (but not realistically) the confluence could be so great that the ions would fuse in one pass through the core. That would result in stupendous fusion rates at only modest confining B field strengths- no need to go to 50 or 100 Tesla, even 10 Tesla may not be needed. In reality the central virtual anode formation would make this piratically impossible, but even modest gains may be important when considering advanced fuels.


Dan Tibbets

[hanelyp]

I believe pv=nRT still applies to plasma, even with the potential well. The potential well has a strong effect on energy and possibly density distributions. The potential well may need to be accounted for in measuring T.

[D Tibbets]

I believe pv=nRT still applies to plasma, even with the potential well. The potential well has a strong effect on energy and possibly density distributions. The potential well may need to be accounted for in measuring T.

Indeed. Here I am somewhat uncertain, but if you only consider the edge region where the plasma is actually pushing against the B field- inflating the Wiffleball or surface of the balloon if you will. the temperature/ KE of the electrons are high, at the bottom of their internal potential well, while the ions are at low temperature/ KE, at the top of their potential well. pv=nRT. For the electrons on the edge the T is large, so with constant volume, the pressure in the edge region is high. For ions the T is low so the pressure due to ions in that region is low. The average temperature of ions and electrons is the same (assuming a good deep potential well), but when location is considered, the picture is different.

What I am not confident of, is considering the average energy o fall particles inside a balloon as the indicator of the balloon inflation. Or, just considering the energy and density of the particles that are actually hitting the balloon wall and pushing it out. If the plasma/ gas is homogenous in energy distribution, as is the case for most gasses/ plasma the picture is the same, but with dynamic opposing energy distributions of two species the picture different. The n or density or particles present in an interval of time also plays a role. In a shapeless mass of plasm, the cold species would hang around longer, so the ballance may be the same as considering an energy homogenous plasma. But, with quasi spherical geometry, with at least a little bit of ion confluence, this ion speed versus dwell time in different radii defined regions is different. The ion density can be greater in the core despite shorter dwell times for each individual ion. The balance depends on the degree of confluence/ central focus, effects of forming central virtual anode, etc. It is mentioned in the EMC2 patent application though, so I assume the effect is significant (or at least can be significant). I think saying the electron pressure dominates the Wiffleball inflation, and thus permitting greater ion densities has to be qualified with the ion density in the core versus the confining B field strength of the Polywell as being the important point.

Dan Tibbets

[DeltaV]

Fissioneers react:

Opinion: Many Reasons To Be Skeptical Of Skunk Works’ Fusion Project
Dec 1, 2014 By John T. Holland and Joe P. Holland | Aviation Week & Space Technology
http://aviationweek.com/technology/opin ... on-project

Yet another problem is the energy carried by the alpha particles. They deliver a real sucker punch to the inner wall of a fusion reactor. Nearly mono-energetic alpha particles will plow into that wall, all stopping at virtually the same depth. The alpha flux will produce helium-filled blisters covering the wall and they will shed metal flakes into the vacuum chamber, causing steady erosion of that wall.

"The reactor walls are falling! The reactor walls are falling!"

Apparently they missed the whole idea of direct conversion of p-11B alphas into UHVDC.

Alphas... Good!
Fission... Bad!

[D Tibbets]

Sputtering of surfaces due to high energy products does not only aply to alphas from P-B11 fusion, but also any charged fusion particles, such as tritium and He3 nuclei from D-D fusion, or alphas and protons (?) frrom D-T fusion. Direct conversion of any of these products that escape before thermalization with the fuel ions will cause sputtering problems. Direct conversion would considerably ease this problem, even if there is not much energy harvesting gain.

This is not a problem for Polywells only though. Spalling/ sputtering is a significant consideration for any high energy plasma machine. Even the fuel ions carry enough energy to create considerable impact energies. Tokamaks suffer this in the form of fuel ions, etc. hitting the walls through ExB diffusion and also through macro instabilities, where single events may deposit as much energy to the walls in a small packet that a small (very small?) grenade analogy may be applied. This ion impact problem may be extreamly troubsom in the diverter required in a tokamak. With a lithium blanket the sputtered material would be mostly neutral lithium atoms and molecules and clumps, and lithium ions. These can poison the reactor just as much as heavier metals. In more compact machines the reactor volume to surface area is smaller, so any sputtering may be more significant. In the Polywell at least, this may be mitigated by any sputtered ions from the chamber walls being outside the magrid. they would not enter the reaction space / core , atleast if the Magrid has a significant positive voltage. Negative ions and sputtered neutrals is a different matter.. If the vacuum pumping can keep up, and the sputtered material does not result in problematic coating of surfaces, then the functional impact may be considerably mitigated in the Polywell.

Dan Tibbets

[mvanwink5]

General Fusion has a liquid lead first wall so they have no issue with the high energy fusion 'ash.'