Prototype of the compression system recently built in General Fusion Inc. aims to test full size pistons (powered by compressed air) and algorithms of their synchronization. Currently we have 14 full size pistons mounted on the steel sphere with inner radius of 0.5m (rings of 7 pistons above and below the equator of the sphere). Interior of the sphere is filled with tangentially pumped liquid lead (Pb) such that evacuated cavity (vortex) is formed in the middle of the sphere (similar to bathtub vortex). Each piston consists of two main parts: “hammer” piston and floating “anvil” piston. A 100kg hammer piston (accelerated to the velocity up-to 50 m/s) impacts the “floating” anvil that is in contact with the liquid lead. As a result of this impact the pressure wave propagates through the anvil and reaches the interface between steel and liquid lead. Due to the close acoustic match between steel and lead most of the pressure wave is transmitted into the liquid lead. Discrete pressure waves produced by the individual pistons merge into a converging pressure wave as they propagate towards the evacuated cavity. When a combined converging wave hits the lead-plasma interface (evacuated cavity) it is almost entirely reflected because of the severe mismatch between the acoustic impedance of liquid lead and vacuum. This interaction results in a rapid inward acceleration of the interface. In the final design magnetized plasma target trapped inside evacuated cavity, is compressed as the interface moves inwards and accelerates further. At the same time a pressure wave is reflected from the interface as a rarefaction wave. This puts the liquid into tension and initiates formation of cavitation regions in the liquid lead.
Compression efficiency in our system relies heavily on our ability to engineer uniform collapse of the evacuated cavity. This task is not easy as a Richtmyer-Meshkov instability (and a Rayleigh-Taylor instability at late stages) may develop during the compression due to imperfections of the liquid lead-vacuum (plasma) interface. Imperfections may originate from different sources: (i) surface ripples resulting from the design of the pumping system and (ii) non-uniformity of the converging wave front coming from slight miss-timing of the pistons, potential instability of the converging wave itself and also because the converging wave is built of discrete waves. In this study we focus on the later one where a particular attention will be given to the structure (temporal and spatial) of the pressure wave produced by the individual piston, as this will have a strong effect on the structure of the combined wave. High fidelity structural Y code based on a novel finite-discrete element approach, is used for the structural simulations. Pressure distribution obtained in this way is then used as a boundary condition to simulate wave propagation in the liquid metal for the geometry of our prototype. At this stage this part will be carried out using OpenFOAM software with particular attention devoted to the role of the equation of state in the liquid metal.
This computational procedure has been already implemented to obtain preliminary results. In those simulations a simplified temporal structure and uniform pressure distribution across the surface of the piston have been used. Results of those simulations provide useful overall picture of what happens in our prototype but further investigation is necessary to get the accurate details and to provide criteria for the efficsient compression.
(a) (b)
Fig 1. (a) Current prototype of the compression system in General Fusion Inc.; (b) liquid lead vortex inside the compression system ( view from the top).
(a) (b) (c)
Fig 2. (a) Simplified computational model used in our preliminary simulations; (b) & (c) Pressure wave propagation inside the liquid metal in geometry of our current compression system prototype.
