Magnetic mirror: pentagon or hexagon toroid?

[artis]

I know that the problem of magnetic mirrors is that they leak out the tight ends of the mirror, on the other hand the main problem of toroids is that magnetic field line curvature creates a better confinement on the inner diameter and lesser on the outer diameter so needs a strong plasma current to keep the plasma confined.

Has anyone ever attempted to create a magnetic mirror where the mirrors are joined end to end and bent slightly in the mirror "bottleneck" part so that after 5 or 6 of such mirrors the circle is closed and all the mirrors are ends to end forming a pentagon or hexagon or any other higher number profile?

In such a device in theory the part where the plasma path is bent is very short and most of the path is straight.

 

[trurle]

I know that the problem of magnetic mirrors is that they leak out the tight ends of the mirror, on the other hand the main problem of toroids is that magnetic field line curvature creates a better confinement on the inner diameter and lesser on the outer diameter so needs a strong plasma current to keep the plasma confined.

Has anyone ever attempted to create a magnetic mirror where the mirrors are joined end to end and bent slightly in the mirror "bottleneck" part so that after 5 or 6 of such mirrors the circle is closed and all the mirrors are ends to end forming a pentagon or hexagon or any other higher number profile?

In such a device in theory the part where the plasma path is bent is very short and most of the path is straight.

You cannot connect an odd number of identical plasma mirrors in closed loop without creating zero-field seam. Therefore, "pentagon" configuration is inferior to "hexagon".

 

[artis]

@trurle although I just sketched 5 mirrors in a pentagon on a piece of paper and I fail to see why there should be a problem? I mean the ends or the "bootlenekcs" have coils with current going in the same direction and they are connected end to end so N-S-N-S,

 

[berkeman]

@trurle although I just sketched 5 mirrors in a pentagon on a piece of paper and I fail to see why there should be a problem? I mean the ends or the "bootlenekcs" have coils with current going in the same direction and they are connected end to end so N-S-N-S,

I believe he was making the point that odd and even numbers of mirrors in the path result in different plasma instabilities that you have to deal with.

Are you familiar with technical treatments of plasma physics like Chen's book?

https://www.amazon.com/dp/3319223089/?tag=pfamazon01-20

 

[artis]

I just got one by Jeffrey Freidberg, called plasma physics for fusion energy, hope it's good enough

 

[PeterDonis]

I just got one by Jeffrey Freidberg, called plasma physics for fusion energy, hope it's good enough

He taught one of the plasma physics courses I took at MIT. He was a good teacher, so I expect his book should be pretty good.

 

[artis]

Ok, here is what interests me. One of the biggest problems with a torus like that of a tokamak is the need for large plasma currents and additional coils because the constant bend that the torus represents makes the plasma have a weaker confinement on it's outer diameter vs it;'s inner one.
Have there ever been any attempts to say create anything similar to what I am talking about here, where straight plasma mirrors are arranged in a circle?

With many such mirrors arranged, the bend angle between each two mirrors would get smaller, so most of the "torus" is not straight except for the band parts, but they are also the ones where the B field is the strongest due to there being magnets, so the plasma "takes a turn" but for a short distance.

From what I know the single biggest issue with magnetic mirror is the end loss, so I wonder how would it compare to other magnetic concepts if arranged in such geometry.
ps. @trurle I am not sure I fully understood your argument about the "zero field seam" because each mirror has two end magnets whose polarities are N-S, how can the field be disturbed because no mater how many coils you put in a toroid fashion as long as each one is the same polarity they all add up don't they?
What would be the different between say (N-S: N-S: N-S) vs (N-S: N-S: N-S: N-S) for example?