Mirror machines, together with stellarators, were the strongest proposals for plasma confinement devices when fusion research started in the early 1950s. The effort was led by R. F. (Dick) Post, who is still actively pursuing the mirror concept today. Unfortunately, the mirror program at Livermore was canceled in 1986 by the USA in order to concentrate on the tokamak. Research on mirror confinement has con­tinued in Russia, under the guidance of Dmitri Ryutov, and in Japan, with the Gamma 10 machine. Reactors using the mirror principle would have the great advantage of direct conversion of energy to electricity without a thermal cycle, the same advantage that hydro, wind, and solar power have over other power sources.

A mirror machine is a leaky magnetic bottle. In Fig. 10.20, a pair of coils gener­ates a magnetic field that bulges out between them. An ion or electron will gyrate in a Larmor orbit around the field lines, as shown in Fig. 4.10. As the orbit approaches the strong-field region at the ends, the Larmor orbit becomes smaller and smaller. To conserve angular momentum, the particle has to gyrate faster and faster, just as a skater does when she pulls her arms in during a spin. But energy is conserved, and to get this extra rotational energy, the particle has to take it out of its translational motion. It slows down in its efforts to escape out the end. Finally, all its translational energy is lost, and the particle has to turn around and go back.

It is reflected by the magnetic mirror, which is just the strong-field region. The particle bounces back and forth between the mirrors at each end.

When a plasma is created, ions (and electrons) have both translational energy and gyrational energy. The ones with lots of translational energy and little gyra — tional energy are lost out the ends. If the mirrors are strong, meaning that the mirror ratio between the fields at the throat and at the midplane is large, only a few particles are lost; and the rest of the plasma is confined. However, this plasma is not in thermal equilibrium; some of the velocities are missing. These velocities are in the loss cone. The plasma will then devise an instability to regenerate those missing veloci­ties and fill the loss cone. There is then a continuous loss, giving magnetic mirrors a short confinement time. Such microinstabilities, however, are not the main prob­lem in mirrors.