Making sure I have the basics right

[jsbiff]

Hey all,

I hope you won't mind me double checking my assumptions here. I think I understand the basic theory behind the polywell, but after reviewing some of the FAQs, I'm still not 100% sure my understanding is correct, so let me lay this out, and make sure I've at least got the basics right:

The polywell uses a very powerful magnetic field (well, configuration of overlapping magnetic fields, plural) to confine a lot of electrons at the center of the reactor. This produces a very strong negative electric field, which is omni directional.

Positively charged ions (nuclei which have had their electrons stripped off) are then injected into the reactor, and because the positively charged ions are attracted by the negative field created by the trapped, focused electrons, they accelerate towards the center of the well, much like an object in space might be attracted by the gravitational field of a planet, and accellerate towards the planet.

Most of those ions will not actually 'collide' with the electrons at the center of the well, but will basically get trapped into a high-speed, high-energy 'orbit' around the 'planet' of electrons, much like satellites orbit the earth.

Because there are so many ions orbitting, and orbitting at very high speeds, it becomes statistically likely that some number of collisions will happen as the orbits of different ions 'cross paths'. When the ions collide while in orbit like that, they will sometimes undergo fusion (but other times, they will simply 'bounce' off each other and continue to orbit).

Is that basically right?

I have seen some discussion on this forum about thermalization and annealing. Am I understanding this right:

Thermalization is where you start with a bunch of the ions orbitting at approximately the same speed/kinetic energy. As discussed above, some collisions don't result in fusion, but instead the ions bounce off each other, and when this happens, sometimes one of the ions will speed up after the collision, while the other ion slows down. Those ions may undergo additional collisions with other ions (because we have billions or trillions of ions in the reactor at a time), trading speed with various other ions in the mixture, until you get a wide distribution of speed/energy, with some ions going much slower than the 'average' speed, some ions going much faster than the average speed, and a whole lot of ions having speeds somewhere in the middle.

Apparently, Thermalization decreases the likelyhood of fusion significantly, so could be a hindrance to efficient fusion (you end up with very hot plasma, but little fusion happening)?

Annealing is a hypothesis posited by Dr. Bussard which states there's some kind of process whereby the 'faster than average' ions will be slowed down while the 'slower than average' ions speed up, bringing the entire mixture back to being about the same speed, so that fusion will still occur at a significant/useful rate?

Do I have all that about right?

[KitemanSA]

... The polywell uses a very powerful magnetic field (well, configuration of overlapping magnetic fields, plural) to confine a lot of electrons at the center of the reactor.

Almost, but not quite. Polywell uses magnetic fields to protect a positively charged grid which together confine elec...

This produces a very strong negative electric field, which is omni directional.

Yup.

Positively charged ions (nuclei which have had their electrons stripped off) are then injected into the reactor, and because the positively charged ions are attracted by the negative field created by the trapped, focused electrons, they accelerate towards the center of the well, much like an object in space might be attracted by the gravitational field of a planet, and accellerate towards the planet.

So far so good.

Most of those ions will not actually 'collide' with the electrons at the center of the well, but will basically get trapped into a high-speed, high-energy 'orbit' around the 'planet' of electrons, much like satellites orbit the earth.

Perhaps more like items orbiting a black hole. (I.e., there is nothing at the center of the well)

Because there are so many ions orbitting, and orbitting at very high speeds, it becomes statistically likely that some number of collisions will happen as the orbits of different ions 'cross paths'. When the ions collide while in orbit like that, they will sometimes undergo fusion (but other times, they will simply 'bounce' off each other and continue to orbit).

Is that basically right?

Other than as corrected, seems so.

PS: What improvements would you suggest for the FAQ?

[BenTC]

Well here goes my first attempt at answering a technical query based on what I've picked up, and not just asking questions myself.

Most of those ions will not actually 'collide' with the electrons at the center of the well, but will basically get trapped into a high-speed, high-energy 'orbit' around the 'planet' of electrons, much like satellites orbit the earth.

While ions may gain some rotational component, the majority of motion in oscillating through the core.

Because there are so many ions orbitting, and orbitting at very high speeds, it becomes statistically likely that some number of collisions will happen as the orbits of different ions 'cross paths'. When the ions collide while in orbit like that, they will sometimes undergo fusion (but other times, they will simply 'bounce' off each other and continue to orbit).

Yes - but not "orbiting"

Thermalization is where you start with a bunch of the ions orbitting at approximately the same speed/kinetic energy. As discussed above, some collisions don't result in fusion, but instead the ions bounce off each other, and when this happens, sometimes one of the ions will speed up after the collision, while the other ion slows down. Those ions may undergo additional collisions with other ions (because we have billions or trillions of ions in the reactor at a time), trading speed with various other ions in the mixture, until you get a wide distribution of speed/energy, with some ions going much slower than the 'average' speed, some ions going much faster than the average speed, and a whole lot of ions having speeds somewhere in the middle.

Apparently, Thermalization decreases the likelyhood of fusion significantly, so could be a hindrance to efficient fusion (you end up with very hot plasma, but little fusion happening)?

Annealing is a hypothesis posited by Dr. Bussard which states there's some kind of process whereby the 'faster than average' ions will be slowed down while the 'slower than average' ions speed up, bringing the entire mixture back to being about the same speed, so that fusion will still occur at a significant/useful rate?

For some reason just now I'm thinking that annealling is mainly related to the electrons. The higher speed electrons would escape rather than recirculate causing losses - which is a major component of the power balance for whether Polywell is viable - but I can't think how the process works.

Actually, its coming back to me. Annealing is related to the ions. Ions moved fast at the centre and slow at the edge of their oscillation through the core. At the edge, the ions do thermalise - but the spread of speeds around the "slow" average value is insignificant to the speed they gain accelerating towards the core. The transit of ions through the core is not long enough for significant thermalisation to occur there, before being annealed at the edge each transit.

[jsbiff]

... The polywell uses a very powerful magnetic field (well, configuration of overlapping magnetic fields, plural) to confine a lot of electrons at the center of the reactor.

Almost, but not quite. Polywell uses magnetic fields to protect a positively charged grid which together confine elec...

Ahh, that makes sense. I've not got any physics degree (as may be obvious from my posts), but I've taken some undergraduate level physics classes (not the 'easy' ones the liberal arts majors take, but the "General Physics for Scientists and Engineers" course which goes into a bit more depth, which I'm sure does NOT impress any of you, but the point is, I've got some decent understanding of *basic* physics, and advanced physics are built upon the foundation of basic physics), passed them with decent marks, and remember a lot of the basic theory (even if I can't remember some of the formulae).

I couldn't figure out how a magnetic field would confine electrons, but a positive electric field, that makes sense. I'm still not sure how the magnetic field protects the grid. I recall from the section on electromagnetism that when a charged particle is in motion, it creates a magnetic field which is (perpendicular?) to the direction of motion, so I suppose the magnetic field doesn't interact with the (basically) stationary electrons in the center (which I believe you guys call a virtual cathode), but if any electrons become 'unconfined' and start to move towards the grid, that motion would generate a magnetic field, which would then interact with the MaGrid, and cause repulsion?

Most of those ions will not actually 'collide' with the electrons at the center of the well, but will basically get trapped into a high-speed, high-energy 'orbit' around the 'planet' of electrons, much like satellites orbit the earth.

Perhaps more like items orbiting a black hole. (I.e., there is nothing at the center of the well)

I'm afraid I don't know that much about the details of black holes to understand the analogy? I think I remember reading somewhere that, because of gravitational distortion of timespace around a black hole, you can never actually reach the center of the black hole, but that doesn't mean nothing is there - in fact, there is *so much* mass there that it warped space, but it's there.

In any case, as applied to the polywell, what does that statement mean? Aren't there electrons at the center of the polywell? So, how does that reconcile with, "(I.e., there is nothing at the center of the well)?"

Because there are so many ions orbitting, and orbitting at very high speeds, it becomes statistically likely that some number of collisions will happen as the orbits of different ions 'cross paths'. When the ions collide while in orbit like that, they will sometimes undergo fusion (but other times, they will simply 'bounce' off each other and continue to orbit).

Is that basically right?

Other than as corrected, seems so.

PS: What improvements would you suggest for the FAQ?

As for improvements, I'll spend some time reviewing the FAQ this weekend, and perhaps suggest some improvements. In general, it seems like both the Wikipedia article linked to by the FAQ, and the FAQ itself, assume a fairly high knowledge of physics terminology and theory.

I have just enough physics education that I was able to basically understand most of it, as you can see, with some misconception of how the MaGrid worked (although, looking at the Wikipedia article again, I see it does mention the grid being positively charged; I guess I missed that when reading previously). However, it might be nice for less technical people who are interested in finding out about this, to have a slightly more expanded explanation of the physics principles and what's happening.

I think my post (modified as needed), might be a good starting place for a less technical description of the basics.

[jsbiff]

I think I should maybe clarify one point. When I describe the Ions as orbitting, I'm picturing a very long, eccentric elliptical orbit, not a mostly circular orbit. But, the reason I call it an orbit is that, I don't think it could *possibly* be a straight line, could it? In particular, as Ions collide, even if the 'oscillations' started off in a nice straight line, the collisions would introduce random vectors into the motion of the ions, resulting in elliptical orbits, right?

[D Tibbets]

Ideally, very narrow elliptical orbits for the ions.
Also, thermalization can occur in the energy spread- upscattering and down scattering; and in angular momentum(sideways scattering) which could 'circulize' the orbits so that there is not as much confluence or central focus of the ions.
The electrons are not at the center exactly, but enough of them spend most of their time closer to the center of the well, than the unscattered ions that began their life near the Wiffleball border. and this creates the potential well (also, the excess total electrons play a role). The inertia of the ions can actually carry them towards the center despite their charge buildup, and this will create a virtual anode if the confluence is good enough. This has important consequences in several aspects of the Polywell.
The electrons are contained fairly well (considering the relatively shorter times required in the Polywell compared to Tokamaks) by the magnetic field. The random walk process is slower for electrons than the more massive ions with their larger gyroradii. That is why the magnetic fields do not need to be so large (in volume). I believe that is what drives Tokamaks to such large sizes. This allows for adequate containment in Polyewlls, despite the cusps (but only when recirculation works). This shorter containment of the electrons and the costs associated is adequate for useful fusion because the fusion rate is faster in a Polywell ( due to increased density, monoenergetic ions, and possibly some degree of confluence.

The positive charge on the magrid plays two roles- first to accelerate the new electrons into the magrid space at high energy. Those electrons that escape through a cusp are recirculated (effectively increasing the confinement time ~ 10-100 X. Those upscattered electrons that escape (a good thing as upscattered electrons are bad) give back that portion of their initial energy as they climb away from the positive magrid. This limits the energy cost of removing these bad electrons, and serves to maintain the narrow range of electron temperatures (at least on the high temperature side).

I like the brief discription of ion annealing by BenTC.


Dan Tibbets

[93143]

One thing about this, that I've seen before with others, is that jsbiff appears to have the impression that the electrons are confined to a tight space near the centre, and the ions then oscillate around outside that space. This is not true.

The "wiffleball" is a condition in which the electrons are pushing back on the magnetic field strongly enough to generate a fairly sharp boundary (not really spherical; more probably somewhat spiky) with essentially no magnetic field inside. The rapid turnaround of the electrons at the edge is what generates the counterfield.

The ions ideally never leave this spiky ball.

In fact, at least one fueling scheme involved injecting neutrals and ionizing them at the boundary of the electron ball.

The density and velocity distributions of the ions and electrons inside the wiffleball are subjects of ongoing debate. But the idea of a "planet" of electrons with "orbiting" ions, while the physical principle is roughly correct, does sound like you've got the wrong picture in your head...

[KitemanSA]

Most of those ions will not actually 'collide' with the electrons at the center of the well, but will basically get trapped into a high-speed, high-energy 'orbit' around the 'planet' of electrons, much like satellites orbit the earth.

Perhaps more like items orbiting a black hole. (I.e., there is nothing at the center of the well)

I'm afraid I don't know that much about the details of black holes to understand the analogy? I think I remember reading somewhere that, because of gravitational distortion of timespace around a black hole, you can never actually reach the center of the black hole, but that doesn't mean nothing is there - in fact, there is *so much* mass there that it warped space, but it's there.

Sorry, tried to maintain an analogy that wasn't correct to start with. My intent was to basically point out that "orbit, i.e., go around a big object" wasn't correct and that the ions would fall almost straight down and back out not hitting much at all except other ions. Sorry for any confusion.

[icarus]

93413:

The rapid turn around of the electrons at the edge is what generates the counter-field.

Aren't the plasma wiffleball surface loop currents involved in 'generating the counter-field' ?

[MSimon]

In fact, at least one fueling scheme involved injecting neutrals and ionizing them at the boundary of the electron ball.

Joel Rogers is doing some work on that. Interesting stuff. I am giving him the engineering perspective which has added another knob to his "tuning" kit.

The "knob" is gas injection temperature.

[D Tibbets]

Re reading Joel's paper, it seems his model had the ions at ~ equal energy and density(?) within the Wiffleball, so he is not assuming any confluence. This would mean his fusion performance is not as high as possible. It also begs the question about thermalization issues under those conditions.

Dan Tibbets