Stabilized Liner Compressor For Low-Cost Controlled Fusion: Progress And Issues

The Stabilized Liner Compressor (SLC) concept uses annular free-pistons, driven by high-pressure helium, to implode a rotating liquid metal liner that compresses a plasma/field target to fusion conditions [1], [2]. The free-pistons accelerate and re-capture the liquid metal, avoiding Rayleigh-Taylor instability by eliminating the free outer surface of the liquid. Sufficient rotation at the inner surface of the liquid prevents Rayleigh-Taylor instability there during the compression and re-expansion of the liner at peak energy-densities of a 50 – 100 T field. Such field levels correspond to operation near the cost-minimum for controlled fusion systems [1], [2]. The combination of free-piston drive and rotation permits stable exchange of energy between pneumatic energy storage and the plasma target. In the fusion reactor concept, modest losses in the liner implosion system are replaced by work done by magnetically-confined alpha-particles from D-T fusion reactions, thereby reducing the required nuclear gain for an economical reactor. Our project for ARPA-E is focused on extending the earlier success with the stabilized liner techniques, demonstrated at the Naval Research Laboratory c. 1979, to the higher drive-pressures (25 kpsi) and implosion speeds (>1 km/s) required to compress plasma targets anticipated in the near future. This technique would enable the frequent repetition of plasma compression experiments needed to develop the liner and plasma system for breakeven tests and fusion reactor design. We use the MACH2 code [3] for calculation of the liner dynamics and also for obtaining estimates of mechanical stresses in the very high pressure SLC prototype. Development of the SLC represents a departure from conventional pulsed electrical power techniques, substituting compact energy storage as highpressure gas (at >100 MJ/m3) for capacitive energy storage, fast gas-valves instead of electrical switches, and pulsed hydrodynamic convergence instead of multiple electrical stages for power density multiplication.