In order to improve the ion radial confinement and to improve the macroscopic stability of the GAMMA10 tandem mirror, we are designing a magnetic divertor in the GAMMA10 central cell.

There is an ion radial loss even in the axisymmetric mirror cell. For example, a plug potential can be created in a tandem mirror experimentally without neutral beam injection at the plug region, which is accompanied by a thermal barrier potential formation. The formation of a thermal barrier potential requires a quick loss of ions trapped in the thermal barrier potential. Experimentally, the ion density observed at the thermal barrier is much less than that in the central cell, indicating the exsistence of large ion loss there. A mechanism of ion radial loss in the thermal barrier is considered to be a non-axisymmetric electrostatic potential being formed in the axisymmetric end mirror cell.

A magnetic divertor is anticipated to make the electrostatic potential axisymmetric in the central cell. In order to install a magnetic divertor in the central cell of GAMMA10 at low cost, mid two solenoid coils in the central cell reverse their current direction by joining the wire of the two solenoid coils to the race-track wire system. The current of race-track wire system can be changed independently of that of central solenoid wire system. The currents of race-track wire system and the central solenoid wire system are adjusted to recircularize a magnetic field line. That is, a circular magnetic flux tube started at central cell midplane becomes circular again at the anchor midplane and circular again at the thermal barrier.

The MHD stability due to the magnetic divertor is carried out including non-paraxial magnetic filed line curvature and finite ion Larmor radius. In the conference we will present on a magnetic divertor in GAMMA10 and its contribution to the MHD stability.