thread for segments files and parameters for simulation runs

[ladajo]

2.) very clear ion wiffleball formation. fairly close to the center, too. which is good for containment, i suppose, but it means: the voltage on the potential well should really be measured from wiffleball to wiffleball, NOT from electron wifflebal to grid. but if the ions are really packing themselves in at that level, then perhaps that actually turns out to be a higher voltage?

I think this is a good point. Traditional discussions I believe have not really addressed the PE aspect of ION oscillation verses the e- core. We (at least I) tend to think in tterms of the absolute e- core vice the relative. Hmmm.
I also tought it was interesting that your sim to date has shown an e- shell around the e- core. I believe we have discussed this several times elsewhere. That the well may not be a clean curve.

[TallDave]

anycase it seems one could save a lot of trouble - one could increase the effective electron lifetime orders of magnitude - if one aimed the e-guns just right and sent them off with just the right velocity to stop in the center.

It could help focusing, but they won't want to sit there long, as it's the top of their potential hill

very clear ion wiffleball formation. fairly close to the center, too.

I'm not 100% sure what you mean here -- usually wiffleball refers to the magnetic field that confines the electrons. I think you're saying you get a sphere of ions?

the voltage on the potential well should really be measured from wiffleball to wiffleball, NOT from electron wifflebal to grid. but if the ions are really packing themselves in at that level, then perhaps that actually turns out to be a higher voltage?

Well, I think typically well depth is measured from ion insertion point (edge) to the center of the virtual cathode because that's what the ions see when you drop them in. I don't think you can get a higher potential than the Magrid from a virtual anode from the ion shell at the edge, because something in the system would have to drive that potential (virtual anodes have been measured, but they're really a dip in the virtual cathode potential at the core), though as ladajo alludes to concentric shells have been raised before in discussions of how Polywells can be generally quasineutral but locally non-neutral -- 93143 has a nice theoretical description of such around here in Theory somewhere -- so that seems promising.

[happyjack27]

It could help focusing, but they won't want to sit there long, as it's the top of their potential hill

from my simulations, they sit there VERY long, as the magnetic fields confine them there. as others have put it, the electron core is COLD. and that's precisely why you want to start them off there, because you add that huge amount of time onto their life. esp. compared to placing them randomly where where they have a very small chance of getting into the cold core before escaping, and in that relatively short time, they don't create much of a potential well, relatively speaking, anyways.

bla bla bla ion wiffleball bla bla bla

I'm not 100% sure what you mean here -- usually wiffleball refers to the magnetic field that confines the electrons. I think you're saying you get a sphere of ions?

potato, potato. hmmm, that doesn't exactly work out the same when written. yes, i get a sphere of ions in the sims. same geometry as the electron wiffleball but a much larger radius.

the voltage on the potential well should really be measured from wiffleball to wiffleball, NOT from electron wifflebal to grid. but if the ions are really packing themselves in at that level, then perhaps that actually turns out to be a higher voltage?

Well, I think typically well depth is measured from ion insertion point (edge) to the center of the virtual cathode because that's what the ions see when you drop them in. I don't think you can get a higher potential than the Magrid from a virtual anode from the ion shell at the edge, because something in the system would have to drive that potential (virtual anodes have been measured, but they're really a dip in the virtual cathode potential at the core), though as ladajo alludes to concentric shells have been raised before in discussions of how Polywells can be generally quasineutral but locally non-neutral -- 93143 has a nice theoretical description of such around here in Theory somewhere -- so that seems promising.

a lot of those ions that you shoot in are just going to be deflected by the magnetic field. you can see from the video that regardless of where they are born ions spend most of their life (the ones that live long at least) in the ion non-wiffleball, with a near 0 KE at the edge. my point is if the goal is to measure the average KE of ions in the collision region, well the vast majority of those ions started out at 0 KE at the edge of the non-wiffleball (or closer). so the vast majority of ions that pass through the collision region are going to have KE proportional to the difference in voltage from the collision region to where most of them had about 0 KE. and most of them had about 0 KE at that non-wiffleball edge.

[happyjack27]

i'm seeing the polywell dynamics differently. from the simulations i see two different wiffleballs, if you will, or as you'd rather have me put it much more awkwardly -- err... i don't even want to try that one -- an electron wb and an ion wb. now they are both formed by a balance between electric kinetic pressure and magnetic pressure. but since the ions are much heavier, that balance point is much further out (i.e. much in favor of electric kinetic pressure, relative to the electrons) so now you have two concentric cores, at different radii and of opposite polarity.

now as we know particles don't like that. so that voltage difference wants to pull the two cores in to the same radii. but there are a number of problems with that.

a) the mag fields aren't having it, esp. for high charge to mass ratio particles and esp. for outward expansion. which means esp. esp. for electrons.
b) it's a point charge (roughly), which makes it approx. 1/r^2, which makes it quite liable to overshoot or at best orbit. in any case exchanging PE and KE in an oscillatory fashion like a pendulum. and likewise never actually reaching 0KE 0PE at the center, save some kind of damping phenomena. and having highest KE, 0PE at the center and vice-versa.

so what you get is the a bunch of ion pendulums swinging through the same center, with greatest distance equal to the ion wiffleball radius, and greatest speed equal to the PE at that point. which, ofcourse, is the difference in voltage from that point to the center.

[happyjack27]

in fact, you can think of the electron core as a planet, and mass as your force, since it's 1/r^2 attractive. thus, if you want all trajectories that hit the core, it's clear that you're looking precisely for all sub-orbital velocities. ion-ion and magnetic forces aside, it's clear that in the point-charge approximation EVERY ion with sub-orbital velocity will oscillate in a spiral fashion through the EXACT center of the core, and EVERY ion with a greater velocity won't ever see the center unless some other force changes its orbit. again, neglecting ion-ion and magnetic forces, and presuming an exact point charge.

[happyjack27]

decided to work on the e-gun velocity calculation problem. made some good progress but ran into a roadblock. here's my research trail. this is a conservation of energy problem so i figured looking up the unit of energy would be a good starting point. i recalled it is the "Joule":

http://en.wikipedia.org/wiki/Joule

"work required to move an electric charge of" assumed charge of q0 = 1 coloumb, we want "voltage"

http://en.wikipedia.org/wiki/Voltage

voltage = diff of electrostatic potential

http://en.wikipedia.org/wiki/Electric_potential_energy


(all qiqj's in which neither qi or gj = my point are irrelevant because they are assumed to not change position)
electro-static potential of q0 on qi = coloumb's constant * q0*qi/r(0 to i)

perhaps i can sum this up when i'm applying the static fields to charges. i'll also have to add in the static fields! i'll look that up later.

electro-static potential at origin =sum(qi/ri) = sum(qi*reciprocal norm(position vectori))
electro-static potential at point p. =sum(qi/ri) = sum(qi*reciprocal norm(position vectori - point p))

... (pop stack) (pop stack)

now we have voltage, multiply this by the charge of our electron = work required = joules of energy (E).

now convert this into KE. easy. divide by mass and stuff.

http://en.wikipedia.org/wiki/Kinetic_energy

bla bla bla... E = mv^2/2. oh yeah. (old high school physics) now some basic algebra:

v = sqrt(2E/m)

viola!

okay, not so fast. we have to go back and do the potential from static field from the wire segments still.

okay umm... don't see it anywhere.

somehow ephi was able to calculate this.

anyone?

[TallDave]

from my simulations, they sit there VERY long, as the magnetic fields confine them there.

Electrons held at the center of the WB by the magnetic field? Hmmm... Nebel and Bussard have described a PW with no field at the center and electrons being deflected at the edge of the WB. Additionally, the electrons should be bouncing around pushing back the B field, with their average position creating the virtual cathode at the center.

My other comment was related to electrons trapped in the wiffleball. Over most of their orbit there is little or no magnetic field (i.e. Larmor radius bigger than the device size) with the electrons turning when they hit the barrier magnetic field. The electron behavior is stochastic since there are no invariants. We don't have any direct measure of the internal magnetic fields, but we do know the density and have a pretty good idea what the electron energy is. High beta discharges should expel the magnetic field. The vacuum fields should be in a mirror regime (as was the DTI device) while the wiffleball fields should transition to better confinement. There is about 3 orders of magnitude difference in the predicted confinement times so it's pretty easy to see which regime the device operates in (unless, of course, the cusp recycle is truly enormous).

a lot of those ions that you shoot in are just going to be deflected by the magnetic field. you can see from the video that regardless of where they are born ions spend most of their life (the ones that live long at least) in the ion non-wiffleball, with a near 0 KE at the edge.

I think the idea is to inject them along field lines. I know that's what is done for electrons.

On the other hand, cusped systems are open field line systems. You can inject charged particles in these systems along the field lines, which is what is done with the electrons in the Polywell. The injection rate of ions doesn't have to be the same as the electrons. If you inject ions and electrons at the same rate, you get the ambipolar result which will likely have small electrostatic potentials and the confinement likely isn't what you desire. On the other hand, if you flood the device with electrons you will get deep potential wells and the ions will be extremely well confined and at very high energy. This is what you want to do in a Polywell. This even works in the quasi-neutral limit where the number of electrons and ions are approximately the same (with just a few more electrons to make the potential well).

[happyjack27]

from my simulations, they sit there VERY long, as the magnetic fields confine them there.

Electrons held at the center of the WB by the magnetic field? Hmmm... Nebel and Bussard have described a PW with no field at the center and electrons being deflected at the edge of the WB. Additionally, the electrons should be bouncing around pushing back the B field, with their average position creating the virtual cathode at the center.

that's what i mean. when you hold something in place you don't put your hand inside it.

a lot of those ions that you shoot in are just going to be deflected by the magnetic field. you can see from the video that regardless of where they are born ions spend most of their life (the ones that live long at least) in the ion non-wiffleball, with a near 0 KE at the edge.

I think the idea is to inject them along field lines. I know that's what is done for electrons.

i.e. through the center of a cusp. from my sims you need to get very near the exact center to have any chance of getting the electron to stay.
yes. and the stronger the magnetic field, the harder it is. I dont' know how accurate eguns are (average angular error? average offset position error?) (i'd like to).

but the question remains, assuming you can shoot them with decent accuracy, at what VELOCITY do you fire in the electrons at? too slow and they'll bounce right off. too fast and they'll go right through. okay, so then what's ideal?

i would argue that ideal is just the right speed so they stop in the exact center. i.e. with kinetic energy equal to the difference in electric potential between the center and where you shoot them from. now once you calculate this you may want to tune it a little to compensate for the magnetic field.

but the point is, from what i've seen in my sims, an electron that starts inside the small region in the center at low enough velocity that it doesn't just fly through has many ORDERS OF MAGNITUDE longer lifespan than a particle that misses that region, even by a relatively small amount. and aren't electron guns supposed to be able to fire at very precise velocities?

[TallDave]

i.e. through the center of a cusp. from my sims you need to get very near the exact center to have any chance of getting the electron to stay.

I'm still not sure whether you're talking about them staying in the WB itself, or in the exact middle of the WB. Are you thinking they would shoot through and exit an opposing cusp if not lined up properly? I don't think that's a major concern. They should bounce around about 100,000 times before exiting.

[icarus]

happyjack:

in your electron only simulation I am amazed at how spherical, i.e., not spiky, the inner core of electrons is. Yet it also seems to show bunching of electrons along the cusp lines (cube projected onto a sphere effect) or am I seeing things there? If this is a real physical feature it indicates there must some secondary effect causing the electrons trapped in the core to 'puff-out' into a perfect sphere (minimal surface shape), like a surface tension due to some global interaction of the particles/plasma.

If the six-coil arrangement can achieve a near perfect sphere electron core it means there would be no need to go to higher-order coil arrangements. In fact, it then would be interesting to see if such a spherical core could be achieved by just two coils in opposition, i.e., the simplest technical arrangement, would be a huge win.

Your ion WB larger than electron WB is also a very interesting observation ... would change the game completely if that is real.

[happyjack27]

i.e. through the center of a cusp. from my sims you need to get very near the exact center to have any chance of getting the electron to stay.

Are you thinking they would shoot through and exit an opposing cusp? I don't think that's a major concern. They should bounce around about 100,000 times before exiting.

if you shoot them with perfect accuracy susbtantially faster than what is needed to reach the center, then neglecting b fields and coloumb forces they'll go right past and reach the velocity they started out at at the exact opposite point. but ofcourse it's not perfectly accurate and there are magnetic fields and coloumb forces. so "going straight through" ofcourse isn't really going to happen. i was speaking roughly. if they dont' fast enough to get through the magrid, they'll just bounce off with out "bouncing around" even once. if you get them somewhere in the vicinity, they'll bounce around like you said maybe 100,000 times. but if you get them right inside the cold electron core, they'll "bounce around" a billion times before exiting.

[TallDave]

Ah, OK.

I think the problem you have there is that the velocity is the voltage of which the well depth is a some fraction around 80%, so if you keep adjusting the e-gun voltage down to the well depth to get your electrons to just the center of your well, the well will keep getting smaller.

We may not want electron lifetimes that long anyway -- we have to drive the electron population continunously to keep the electron distribution nonthermal.

But it's an interesting thought.

[happyjack27]

this is what i'm talking about:

http://www.youtube.com/watch?v=dm0En2EE6Ko

that's after hours of simulation time, just having new electrons sprout randomly whenever an electron gets lost (beyond 3x the magrid radius)

you can see the electrons are much denser in the center, and given how they accumulated, that can only imply that their lifetimes are much longer. so really how many electrons do you want to pump in to the system? i thought electron losses where the biggest problem. that you wanted to avoid spending the energy pumping in new electrons as much as possible. also ofcourse, you want the electrons concentrated in the center. well there you go, electrons concentrated in the center. and are they thermalized? no. if you zoom in you can see that they're very cold compared to the others. and you can't have a large variance in speed if you don't have large individual speeds. so it's everything you want.

[icarus]

Why does the electron distribution have to be "non-thermal" (you mean non-Maxwellian right) again?

If the electrons are confined in the center in enough numbers to produce the well isn't that job done?

[happyjack27]

"electron distribution nonthermal" maybe by distribution he means positions rather than velocity. but in that case, if you've got to keep pumping in electrons to do that, then clearly the wiffleball is not working. you mean you have to pump in electrons to replace ones that escaped so you can keep your deep potential well. well that what you want is for them not to to escape, don't you? that's what it means to have long lifetimes. and i'm talking long lifetimes in a very small and contained region in the center. i.e. i'm talking nonthermal electron distribution.