It Is A Different Machine

[KitemanSA]

Wouldn't any ion shot in from the outside have enough energy to exit the other side? I thought we wanted to generate the ions effectively at the well edge.

[ravingdave]

Wouldn't any ion shot in from the outside have enough energy to exit the other side? I thought we wanted to generate the ions effectively at the well edge.

This has been something that I used to wonder about. The ion will "Fall" towards the center with sufficient velocity to go out the other side. The only explanation I can think of for why this might not happen is that it is impossible to guarantee that a "dropped" ion will go precisely through the center and reach apogee at precisely the neutral focus of the magnet coil. If it deviates ever so slightly, it will go into a sort of orbit with a teeny tiny bit of lateral movement that will cause it to miss the hole in the magnetic field that will allow it to exit.


This effect of "Missing the exit hole" could be enhanced by injecting ions from opposite bores. As the ions encounter each others field, they will start to repel each other, creating an effect like two fire hoses pushing against each other.


Add the inevitable collisions with the electrons and other ions and they will be reduced in velocity somewhat, thereby precluding escape velocity.

Dunno. Just an idea.


David

[KitemanSA]

I am thinking that (using the above idea) an ion would start traveling towards the center of the well, only to be immediately affected by the magnetic field. It would probably look something like this...

....

I don't agree. The force should be along the LENGTH of the coil, not around it. It is v CROSS B, not v DOT B. Since the ion can't curve back towards the casing, evenually it should drift toward the center and get caught into the well, no?

I am thinking that starting the ion out with a gyro effect isn't as good as starting it out moving in a nice straight line towards the center. It seems like that would cause too much lateral motion as opposed to radial.

I don't get where it will pick up ANY radial. Longitudinal maybe, but not radial.

In actuality, I think it would be impossible to get this thing to work at all if you don't get the ions converging on the center.

Concur, but I don't see where the introduction of ions just inside the MaGrid wouldn't result in ions converging to the center.

For this reason I think the ions pretty much have to be released from some area that is sorta neutral magnetically like the bore of the magnet or a corner cusp.

Don't concur. I suspect such a system would add needless complexity, but what do I know. I am a mechanical engineer!

[ravingdave]

If the path of the ion cuts across field lines there will be a force generated on it. If the field is non uniform along the path of the Ion, it can't help but cut across some field lines.

It is my understanding that the field surrounding the torus is distorted by the adjacent fields and the wiffleball. (other people have posted pictures of this.) As the field is distorted, the path of the ion must cut across some field lines imparting some degree of deflection.

Here's my amateurish drawing of what i'm talking about.




I think you are doing a cross product for a uniform field with no component across the direction of motion of the ion. This won't work for a non uniform field. The variations in field strength will create a component across the direction of motion of the ion.


This is, however, not the case for an Ion injected straight through the bore of the magnet. All the varying components of magnetic field strength cancel out in the very center of the bore.

At least this is how I understand it.

David

[KitemanSA]

The distintion is that your prior graphics had the ion deflecting round the minor diameter of the torus, along the field lines. Mine says the deflection will be into or out of the page, along the MAJOR circumference of the torus. Yours has it ending up orbiting the minor diameter. Mine has it drifting to the field edge (the wiffleball) then falling down the well.

So does anyone have a definitive answer for which is correct?

Edit (diameter to circumference, boy did I mess THAT one up!)

[MSimon]

It's going to be some kind of gee whiz kit. A direct energy conversion screen that catches the alphas streaming out of the cusps, thus protecting the injection guns, but that allows the injected ions/electron to go through in the opposite direction.

One of those hi-sci-fi techno naming moments ..... "space charge diode"?
energy flux diverter ... differentiated space-charge diverter ...

go to town .... then you've got to build the thing.

The Alpha Flux Diverter should do the trick.

It may be possible - since the alphas will be focused by the magnets - to mount the electron injectors off axis to avoid the majority of the alpha flux.

[imaginatium]

It's going to be some kind of gee whiz kit. A direct energy conversion screen that catches the alphas streaming out of the cusps, thus protecting the injection guns, but that allows the injected ions/electron to go through in the opposite direction.

One of those hi-sci-fi techno naming moments ..... "space charge diode"?
energy flux diverter ... differentiated space-charge diverter ...

go to town .... then you've got to build the thing.

The Alpha Flux Diverter should do the trick.

It may be possible - since the alphas will be focused by the magnets - to mount the electron injectors off axis to avoid the majority of the alpha flux.

Can you be more specific about where off access you would put them, and the trajectory of the electron/ion flow?

[MSimon]

It's going to be some kind of gee whiz kit. A direct energy conversion screen that catches the alphas streaming out of the cusps, thus protecting the injection guns, but that allows the injected ions/electron to go through in the opposite direction.

One of those hi-sci-fi techno naming moments ..... "space charge diode"?
energy flux diverter ... differentiated space-charge diverter ...

go to town .... then you've got to build the thing.

The Alpha Flux Diverter should do the trick.

It may be possible - since the alphas will be focused by the magnets - to mount the electron injectors off axis to avoid the majority of the alpha flux.

Can you be more specific about where off access you would put them, and the trajectory of the electron/ion flow?

You put the electron guns where they will work the way you want. ;-)

Think of a funnel where you are shooting peas out the funnel bell by injecting them into the stem. Most of the peas are gong to come straight out. Now aim a stream of BBs into the wide mouth of the funnel off axis and the funnel will guide the BBs to the stem.

The question is: how much will the angular motion of the electrons affect operation?

[KitemanSA]

Folks,
If the electrons and ions are streaming out of the cusp, put a magnetic field cross ways against the beams and the ions will go one way and the electrons will go the other, and vice versa coming in.

[MSimon]

Folks,
If the electrons and ions are streaming out of the cusp, put a magnetic field cross ways against the beams and the ions will go one way and the electrons will go the other, and vice versa coming in.

That will work for the electrons coming in. However since the ions are somewhat more randomly distributed re: velocity the ions will not all go to the left while electrons go to the right. Due to velocity distributions ions will go to the left and right.

My idea works because electrons are more easily deflected by a magnetic field. So even off axis they will get steered into the stem of the funnel.

It really is an optical problem. The cusps tend to act as collimators with different indexes of refraction for electrons and ions.

Can it be made to work? My math is not good enough for an estimate and even then experiments may be different than theory. And then you have the problem of electrons neutralizing the ions. It may be necessary to have one or two magnets much stronger than the others limiting ion exit and allowing on axis electron injection.

[KitemanSA]

Folks,
If the electrons and ions are streaming out of the cusp, put a magnetic field cross ways against the beams and the ions will go one way and the electrons will go the other, and vice versa coming in.

That will work for the electrons coming in. However since the ions are somewhat more randomly distributed re: velocity the ions will not all go to the left while electrons go to the right. Due to velocity distributions ions will go to the left and right.

??? I can accept that the ions may spread out in the direction opposite the way the electrons move, but none will turn the same way that the electrons will, unless you posit negative ions in the mix. And as to spreading out, I think that would be a good thing. It would allow the "venetian blind" generators to work more efficiently on more narrowly grouped ion energies.

[MSimon]

Folks,
If the electrons and ions are streaming out of the cusp, put a magnetic field cross ways against the beams and the ions will go one way and the electrons will go the other, and vice versa coming in.

That will work for the electrons coming in. However since the ions are somewhat more randomly distributed re: velocity the ions will not all go to the left while electrons go to the right. Due to velocity distributions ions will go to the left and right.

??? I can accept that the ions may spread out in the direction opposite the way the electrons move, but none will turn the same way that the electrons will, unless you posit negative ions in the mix. And as to spreading out, I think that would be a good thing. It would allow the "venetian blind" generators to work more efficiently on more narrowly grouped ion energies.

It is a cross product. B X v. If the velocity is right and left then ions will be deflected left and right (given the appropriate pole arrangement) While electrons will be deflected right and left. If the velocity distribution of the electrons is narrower (approximately unidirectional) then they will get deflected in one direction only.

[Art Carlson]

If the electrons and ions are streaming out of the cusp, put a magnetic field cross ways against the beams and the ions will go one way and the electrons will go the other, and vice versa coming in.

That is true if you have individual particles, but not if you have a plasma. (That is, not if the Debye length is smaller than the beam diameter.) The particles will "try" to move in the vXB direction as you say, but as soon as they start to move apart, the charge separation will produce an electric field in that direction. The particles will then experience an EXB drift in the (vXB)XB direction, i.e. in the direction of the initial motion v. The net effect is that both ions and electrons will more or less straight across the magnetic field.

This effect is one way to look at the need for a rotational transform (twisting of field lines) in a toroidal machine (tokamak or stellarator). The individual particles first are happy to spiral along the toroidal field lines. But then the curvature and gradient of the field lines starts to push the ions up and the electrons down (or vice versa). As in the case of a beam, this results in a large vertical electric field which pushed the plasma outward (fast!). By twisting the field lines, you short out this electric field. Otherwise toroidal confinement wouldn't have even the half-chance it does.

[Art Carlson]

Think of a funnel where you are shooting peas out the funnel bell by injecting them into the stem. Most of the peas are gong to come straight out. Now aim a stream of BBs into the wide mouth of the funnel off axis and the funnel will guide the BBs to the stem.

My idea works because electrons are more easily deflected by a magnetic field. So even off axis they will get steered into the stem of the funnel.

It really is an optical problem. The cusps tend to act as collimators with different indexes of refraction for electrons and ions.

I'm not sure if you are just waving your hands here, but this picture is false if you take it literally.

You can get BBs to go down a funnel if you have dissipation and gravity. If you have perfectly reflecting walls, they will bounce repeatedly, and with each bounce their angle to the axis will increase. Eventually they will be turned around (or at least most of them will).

Optically you have the same situation. Put two plane mirrors at an angle to each other with a small gap at the apex. If you look into the V, there is a small stripe around the apex where you see - possibly after several reflections - through the gap. Most everywhere you look you will see a reflection of your side of the room.

[KitemanSA]

... The particles will "try" to move in the vXB direction as you say, but as soon as they start to move apart, the charge separation will produce an electric field in that direction. The particles will then experience an EXB drift in the (vXB)XB direction, i.e. in the direction of the initial motion v. The net effect is that both ions and electrons will more or less straight across the magnetic field. ...

Thank you for the clarification.
If the ions have MeVs of energy and the electrons have 10s to 100s of keV, won't they have enough to overcome the attraction?

Oh, and they would still be in the ~100keV MaGrid field too. How would that effect things?