OK, so there's trapped electrons in the center and ions streaming in toward them.
What's to keep the ions from merely sweeping up all the electrons and simply forming neutral atoms?
What happens to free electrons/ions?
The energy level. The relative motion between the ions and electrons is measured in 100's to 1000's of electron volts (eV), but the binding energy of electrons to the atom is measured in 1 to 10 eV. To be captured, the electrons and ions have to be moving at the same speed and in roughly the same direction so the relative energy difference is small.
It still happens, but the probability is really small.
It still happens, but the probability is really small.
Or in even simpler(less accurate) laymen terms: The electromagnetic force between charged particles has an escape velocity much like gravity. If electrons are going fast enough they don't slow down enough going past an ion to be caught by it's electromagnetic force.drmike wrote:The energy level. The relative motion between the ions and electrons is measured in 100's to 1000's of electron volts (eV), but the binding energy of electrons to the atom is measured in 1 to 10 eV. To be captured, the electrons and ions have to be moving at the same speed and in roughly the same direction so the relative energy difference is small.
It still happens, but the probability is really small.
Yargh...
I finally got some code to generate mag field descriptions. But nubmers are no fun to read, so I went looking for some 3D data viewers that might run under my linux box. By mid day, I had gotten a few downloads to fail and then in the process of trying to get one program to compile, "fixed" my computer so it wouldn't even boot!
It didn't take too long to get my system back, but it's proof again that "if you want something done right you have to do it yourself". Or it's pretty good proof I'm a nut case
I'll just have to generate my own pictures and post them to a web site. Browsers can read PNG files no problem, but the whole data set won't be interactive.
The MaGrid fields are straight forward to compute and display. The electron density is a bit more tricky, but should be "straight forward" for no collisions. That's not realistic, but it's a start. I haven't worked out the formulas to include ion density, but I suspect that's pretty similar to electrons - at least in form.
If I can find an easy way to code the energy density, it will be fairly obvious why neutrals can't form. The number density of electrons cold enough will be pretty darn close to zero.
I finally got some code to generate mag field descriptions. But nubmers are no fun to read, so I went looking for some 3D data viewers that might run under my linux box. By mid day, I had gotten a few downloads to fail and then in the process of trying to get one program to compile, "fixed" my computer so it wouldn't even boot!
It didn't take too long to get my system back, but it's proof again that "if you want something done right you have to do it yourself". Or it's pretty good proof I'm a nut case
I'll just have to generate my own pictures and post them to a web site. Browsers can read PNG files no problem, but the whole data set won't be interactive.
The MaGrid fields are straight forward to compute and display. The electron density is a bit more tricky, but should be "straight forward" for no collisions. That's not realistic, but it's a start. I haven't worked out the formulas to include ion density, but I suspect that's pretty similar to electrons - at least in form.
If I can find an easy way to code the energy density, it will be fairly obvious why neutrals can't form. The number density of electrons cold enough will be pretty darn close to zero.
I haven't got very far with 3D visualization. I've experimented with chimera:
http://www.rbvi.ucsf.edu/chimera/
Here's a 200x200x200 bfield magnitude map:
http://www.pld.ttu.ee/~indrek/ephi/mapgen.sit.zip
And here are some images:
http://www.pld.ttu.ee/~indrek/ephi/chimera1.png
http://www.pld.ttu.ee/~indrek/ephi/chimera2.png
http://www.pld.ttu.ee/~indrek/ephi/chimera3.png
Figured out how to write situs file format just an hour ago. Early days.
Indrek
http://www.rbvi.ucsf.edu/chimera/
Here's a 200x200x200 bfield magnitude map:
http://www.pld.ttu.ee/~indrek/ephi/mapgen.sit.zip
And here are some images:
http://www.pld.ttu.ee/~indrek/ephi/chimera1.png
http://www.pld.ttu.ee/~indrek/ephi/chimera2.png
http://www.pld.ttu.ee/~indrek/ephi/chimera3.png
Figured out how to write situs file format just an hour ago. Early days.
Indrek
No, I haven't tried those. I've used VisIt from LLNL which works on windows, but I'm running a linux distro.
Computing the mag field is pretty simple, just computationally time consuming. Once I've got the data, I want to check that it looks reasonable - visualization is the simplest way to do that. It's actually pretty easy to pick a view point and then draw vectors on a screen - but I need to pick a "number density" that makes sense. I'll have to play with the volumes I average over so the number of vectors is "reasonable".
It's really fun to make the computations and visualization work. I think my attack will be to check that one coil produces what I expect (and it'll be a trivial "infinitly thin" wire), then add the remaining coils one at a time, then add electrons. I suspect lots of numerical problems at that point, so it should be fun and interesting.
My graphics won't look great - but it will be sufficient for debugging.
I'm also going to see if I can get a different linux distro to boot on the same machine - I've got an empty partition. Unfortunatly, that's a fun time sink too
Computing the mag field is pretty simple, just computationally time consuming. Once I've got the data, I want to check that it looks reasonable - visualization is the simplest way to do that. It's actually pretty easy to pick a view point and then draw vectors on a screen - but I need to pick a "number density" that makes sense. I'll have to play with the volumes I average over so the number of vectors is "reasonable".
It's really fun to make the computations and visualization work. I think my attack will be to check that one coil produces what I expect (and it'll be a trivial "infinitly thin" wire), then add the remaining coils one at a time, then add electrons. I suspect lots of numerical problems at that point, so it should be fun and interesting.
My graphics won't look great - but it will be sufficient for debugging.
I'm also going to see if I can get a different linux distro to boot on the same machine - I've got an empty partition. Unfortunatly, that's a fun time sink too
Very interesting stuff. I have seen Indreks YouTube movies and I liked them so much I saved them to my computer hdd and converted them to *.avi's. Maybe you could put your finished work up on YouTube aswell DrMike?
Btw I have a question. I know why Polywell would need a ion gun. To get higher density of electrons inside the machine than outside. But why is it that fusion need a electron gun anyway? What is the purpose of electrons to begin with? I though polywell fusion was intended to be fueld with proton and boron!?!?
I have also thought about creating a private skypecast each night for one hour or so where everyone from this forum can join and talk about polywell fusion. Give me a hint if this is something that would interest you people?
//Rexxam62
Btw I have a question. I know why Polywell would need a ion gun. To get higher density of electrons inside the machine than outside. But why is it that fusion need a electron gun anyway? What is the purpose of electrons to begin with? I though polywell fusion was intended to be fueld with proton and boron!?!?
I have also thought about creating a private skypecast each night for one hour or so where everyone from this forum can join and talk about polywell fusion. Give me a hint if this is something that would interest you people?
//Rexxam62
Basic Idea
Both fusors and the Polywell use electrons to contain and accelerate the fuel ions, be it DD or pB11. The core principal is as simple as opposite charges attracting. They use a strong negative charge in the middle of a cloud of positively charged ions. If the charges are strong enough the ions will be attracted to a high enough speed to fuse if they collide. The principal of getting fusion with such a configuration is simple. The problem is that high electron densities generally cost a lot of energy to maintain.rexxam62 wrote:Very interesting stuff. I have seen Indreks YouTube movies and I liked them so much I saved them to my computer hdd and converted them to *.avi's. Maybe you could put your finished work up on YouTube aswell DrMike?
Btw I have a question. I know why Polywell would need a ion gun. To get higher density of electrons inside the machine than outside. But why is it that fusion need a electron gun anyway? What is the purpose of electrons to begin with? I though polywell fusion was intended to be fueld with proton and boron!?!?
I have also thought about creating a private skypecast each night for one hour or so where everyone from this forum can join and talk about polywell fusion. Give me a hint if this is something that would interest you people?
//Rexxam62
I'm running a linux system as well. Still these days recompiling the kernel no longer gets me excited (as it did 10 years ago), rather it gets me annoyed 
Here are some more strange pictures, they're not really showing anything specific or that useful, just some fun-looking graphics chimera generates:
http://www.pld.ttu.ee/~indrek/ephi/chimera4.png
http://www.pld.ttu.ee/~indrek/ephi/chimera5.png
http://www.pld.ttu.ee/~indrek/ephi/chimera6.png
The chimera2.png and chimera5.png show the equimagnitudal magnetic field shape at the center where it gradually declines. Of course there's more to a magnetic field than its magnitude - it is a vector field after all.
I can calculate a 250x250x250 grid in 26 seconds (single current loops as coils, quad cpu). So at around 600,000 field samples per second (not including writing them to disk). In ephi-0.1 it would have taken couple of hours and at several orders of magnitude less precision - so I've come a long way. But I think I can probably increase the speed 5-10x (with some loss in precision).
If you want to explore the magnetic field magnitude in a home-grown way then here's a web application:
http://www.mare.ee/indrek/ephi/bfm/
Just click on buttons to move at different levels. This simulation also has more realistic coils formed by multiple current loops.
Here are some more strange pictures, they're not really showing anything specific or that useful, just some fun-looking graphics chimera generates:
http://www.pld.ttu.ee/~indrek/ephi/chimera4.png
http://www.pld.ttu.ee/~indrek/ephi/chimera5.png
http://www.pld.ttu.ee/~indrek/ephi/chimera6.png
The chimera2.png and chimera5.png show the equimagnitudal magnetic field shape at the center where it gradually declines. Of course there's more to a magnetic field than its magnitude - it is a vector field after all.
I can calculate a 250x250x250 grid in 26 seconds (single current loops as coils, quad cpu). So at around 600,000 field samples per second (not including writing them to disk). In ephi-0.1 it would have taken couple of hours and at several orders of magnitude less precision - so I've come a long way. But I think I can probably increase the speed 5-10x (with some loss in precision).
If you want to explore the magnetic field magnitude in a home-grown way then here's a web application:
http://www.mare.ee/indrek/ephi/bfm/
Just click on buttons to move at different levels. This simulation also has more realistic coils formed by multiple current loops.