Stabilized Liner Compressor: 21/2-D multiphysics simulations

Summary form only given. NumerEx's Stabilized Liner Compressor (SLC) is a mechanical driver that rotates and implodes a liquid metal vortex - our liner - to compress magnetic fields and plasma to fusion energy densities. The rotation conquers drive asymmetries, and makes the liner rebound for immediate reuse after the implosion. The liquid liner will serve as a first wall in a reactor, absorbing neutrons, protecting other components, and breeding tritium. The time is right for SLC because magnetized plasma target lifetimes are increasing dramatically. However SLC's design poses significant challenges. We must make the implosion fast enough, the structure strong enough, and the flow repeatable enough to recapture the liquid liner. We will surmount these challenges first in a fully functional laboratory-scale model designed to implode a 10 cm diameter cylindrical volume to 1 cm diameter in 25 μs. We are designing this device using multi-physics computer simulation. We are using 2½-d MACH2, with its 2-d domain and ∂/∂θ assumed zero but all three components of each vector field included, as is essential for problems with fluid rotation. The driving energy in SLC comes from a plenum of high pressure gas released by fast valves to a chamber behind a free annular piston. The piston pushes on low melting-point liquid metal that converts the piston's swept volume to the implosion of an inward-facing rotating free surface. Our simulations follow the compressible motion of the high pressure gas into the chamber and the elastic behavior of the piston as it transmits the pressure to the liquid metal, producing an initially nearly incompressible fluid flow. As the free surface of the liquid liner is driven to smaller radius and higher radial and azimuthal velocity, our simulations capture the increasingly compressible fluid behavior. We will show simulations of low pressure experiments done with water and with eutectic NaK at NRL in the late 1970's. The relatively complete data from the water experiment are well matched by our simulations in radius and symmetry. The NaK data are less complete and less well matched. We will show very high pressure simulations of Na and NaK in geometries similar to those used at NRL, with similarly symmetric results. We will also show simulations of axially longer geometries that fail to achieve the necessary symmetry to show that this is clearly a work in progress!