Hypothesis on Electron and Ion Behavior Inside the Polywell.

[happyjack27]

that parameter ("global loss radius") (aka inner radius of the surrounding vacuum chamber), as well as many other parameters, will be set via a configuration file. there will be two configuration files, one for the particles (including injection points), and one for the static fields (aka magrid)

here's a description of the file formats

here's a sample static fields file. this one's for a perfect cuboctahedral grid with radius 3.

sample particles file - injection on faces of above static fields file

same as above, injection on corners


and that note, i guess i have a few questions on parameters...

what are the approximate scales? e.g. should max injection rate be in volts/sec?
should certain things by differences instead instead of ratios? e.g. should the number of electrons be set to create a constant net charge, rather than a constant ratio with the other particles?

things like that. if anyone finds anything off in the configuration file format like that, please let me know and i'll change it accordingly.

[D Tibbets]

"what are the approximate scales? e.g. should max injection rate be in volts/sec? "

Hopefully not off base, but wouldn't rate be in Amps, or Coulombs per second, or particles per second (corrected for Z)? Volts (eV) describes the energy of the particles, not their abundance or density.

Dan Tibbets

[happyjack27]

"what are the approximate scales? e.g. should max injection rate be in volts/sec? "

Hopefully not off base, but wouldn't rate be in Amps, or Coulombs per second, or particles per second (corrected for Z)? Volts (eV) describes the energy of the particles, not their abundance or density.

Dan Tibbets

thanks. i was actually thinking coloumbs per second. ideally the rate in whatever unit is chosen would be close to 1. that's what i mean by "correct scale". and since time is per second and particles have very small charges and masses, i figure we're talking millions per second, give or take a few orders of magnitude. and thus coloumbs per second would be better. (same as amps? i'll look that up.) i guess my question is (questions are) put most succinctly as: what is easiest / simplest for the end user? and what are typical values?

thanks.

[happyjack27]

new video:

http://www.youtube.com/watch?v=vAft1qpDUv8

ohh... pretty...

you guys do not know how fun it is to watch these thing in 1080p. it's really starting to put a damper on my productivity.

[happyjack27]

static magnetic field test

does this look right?

it's a single long line current going through a hollow sphere of electrons and ions. (net neutral)

i'm not even sure i know what to look for. clearly the electrons are behaving erratically in the presence of the field.

[Aero]

Its beautiful. Its probably right for what you have modeled. A Polywell is not net neutral, though, (there are about a million more electrons than ions, but I don't recall any discussion about whether 11B counts as one ion or 5 + charges) and magnetic field lines don't penetrate the well once it has formed.

Someone who knows needs to jump in here and talk about the physics of a Polywell. I would, but Kiteman would correct me, as would others so maybe they should try to describe it for you, mathematically.

[icarus]

Yes, Kiteman should describe it mathematically, he seems to know everything about how it works.

[D Tibbets]

...Pollywell is not net neutral, though, (there are about a million more electrons than ions, but I don't recall any discussion about whether 11B counts as one ion or 5 + charges) and magnetic field lines don't penetrate the well once it has formed....

My understanding is that there is ~ 1 ppm excess negative charges (electrons). If there are 1 * 10^20 charged particles (deuterium positive ions and electrons) per cubic meter, then there are actually 4.999999 *10^19 ions and 5.000001 *10^19 electrons. Adjust for higher Z ions like boron11 or Helium3, or lithium6, etc. Then further adjust for the ratio of different ions (like 10 protons for each boron).

In the above example there are actually a lot more than 1 million excess electrons. There would actually be ~10^14 more electrons, which is one millionth of 10^ 20 total charged particles per cubic meter.*
If there was only one million excess electrons out of a population of 10^20 charged particles, then it would be so close to net neutrality, that it could be considered neutral within a deviation of one part per 100 trillion. Probably unmeasurable and completely inadequate to electrostatically confine ions.

*Assuming I'm not confused again


Dan Tibbets

[happyjack27]

alright, yeah, i think i got it right. i put a negative charge on the wire and the electrons went backwards along the wire as they came off. that makes sense. if the electrons are traveling radially and the field lines are concentric then the lorentz force should push them in the one remaining direction: along the length of the wire.

short video of charged particles in the presence of a charged wire with current

so yeah that means my static e-field test worked too. though it actually didnt seem to so for now i'm modeling all wires as infinite length to make it simpler. i'll have to look into it again. but with that caveat, i'm ready for modeling.

thou bear in mind two things:
1) in addition to currently being limited to infinite length wires, i don't have particle sources and sinks in yet.
2) i only have like 14k particles. so e.g. if they're all electrons that's only about 1x10^(-15) coloumbs. now i can set a "clumping" parameter to model say every 1,000 particles as 1 particle w/1,000 times the charge and mass, but a) so now that's 1x10^(-12), big whoop., and b) i don't know how valid the simulation is for large "clumping" values. in any case i'd rather just scale the other parameters down instead. the point is not to achieve fusion densities, just to have a scale model of the electrodynamics.

[krenshala]

Very nice videos. Can't wait to see more of them.

From a clarity of view standpoint you might consider changing electrons from yellow, which blend in with the white boron particles for me, with something else such as red or green.

[happyjack27]

Very nice videos. Can't wait to see more of them.

From a clarity of view standpoint you might consider changing electrons from yellow, which blend in with the white boron particles for me, with something else such as red or green.

done. it was 100% red 100% green 0% blue (aka yellow), i toned down the green to about 60%, so now it's just a little on the yellow side of orange. much more distinct. i want to save the darker red range for diagnostic stuff like visualizing electron loss areas. and i want to save the "cool" colors (blue/green/white) for ions.

btw if anyone has a cuda capable video card (nvidia 8800 or better), i can upload an executable to sourcefoge, and they'll be able to set all the particle types and colors as well as the static fields via configuration files. then they can make their own videos and play around with it as much as they like.

[happyjack27]

and for my next exhibit...

"Variations on the Theme of Two Parallel Lines"

||
||
||
||

[KitemanSA]

... I would, but Kiteman would correct me,...

[happyjack27]

the anti-baseball (2 currents, skew to each other):

http://www.youtube.com/watch?v=JfnU8LZ0CLU

and with oppositely charged wires:

http://www.youtube.com/watch?v=bpMPexE18tA

and a positively charged square coil:

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

so many variations, so little video upload bandwidth.


i've disabled the "damping" and "softening" sliders 'cause they're not applicable. i think i'm going to turn them into "voltage" and "current" sliders, so you can change the static field strengths on-the-fly.

anycase, i'm taking requests.

and if anyone wants to just send me a series of line segments, e.g.

1.0,0.0,0.0,-1.0,0.0,0.0

would be a 2-meter wire down the x-axis.

i can mess around w/the voltage and current values to find interesting areas, and i'll make a video of it and post it.

[icarus]

happyjack:

I assume you have already made sure that a test case single charge around a wire has the correct gyro-radius?
Similarly a test charge moves away from a charged wire with correct acceleration?

Best to validate the building blocks before constructing larger scale simulations in my experience.