gas turbine. The abovementioned fluoride salts have high latent heats of fusion and relatively high densities, yet have low thermal conductivity (of the order of 1 W m-1 K-1 and large volumetric changes upon melting (10-30%). Due to the latter, small separated containment canisters with thick walls are required in construction of the LTES. In this paper, two new LTES methods are proposed. An advanced receiver configuration is also presented. Encapsulated Thermal Energy Storage Principle of Stress Relaxation One of the serious problems encountered when PCMs are encapsulated in hard canisters is the generation of a void due to volumetric changes of the PCM upon melting. Most of the candidate fluoride salts mentioned above are known to have large volumetric changes upon melting, and thus pose difficulties of this sort. The formation of voids affects both melting and solidification heat transfer processes, and induces mechanical stress on the containment walls. A typical example of phase change behavior within a canister is illustrated in Fig. 1. Void placement under 1 g and microgravity conditions is shown in Figs 1(a) and 1(b), respectively. When heat is added, the liquid in the melted region expands at the beginning of the phase change. In both cases, the liquid is trapped inside a solid section and cannot expand into the void(s), resulting in mechanical stress on the wall. Eventually, a deformation and/or rupture of the canister may occur. This problem can be resolved by thickening the walls of the containment canister, but this also significantly increases the system's weight. The principle of our new method, which relaxes the stress, is illustrated in Fig. 2. The inside of the containment canister is divided by the addition of volumetric variable fins which remain unwetted by the PCM. Under such conditions, it is expected that melting will also occur preferentially along the fins, creating channels from the liquid up into the void(s). The liquid phase can thus be transferred to the center of the canister via such channels. This not only relaxes the mechanical stress but also enhances the heat transfer rate. If the fin material is wetted by the PCM, the liquid phase permeates the inside of the fin structure and then solidifies. In such a case, the gap is not created. It is thus essential that the correct materials are selected for the fin. Fortunately, we obtained a good material: carbon is not wetted by fluoride salts and has a high thermal conductivity and low density. Moreover, it resists corrosion by fluoride salts even at elevated temperatures. Carbon cloth or carbon felt has been the most promising form of the material. A cylindrical canister with a carbon fin and the canister after sealing by electron-beam welding are shown in Figs 3 and 4, respectively. Each canister is 20 mm in diameter, 30 mm in height, and has molybdenum walls (of thickness 0.8 mm). Visualization Experiment In order to obtain experimental information on phase changes inside the containment canister, an in situ visualization experiment using X-ray computer tomography (CT) was performed. X-ray CT is effective for continuous observation of the curved upper surface of the liquid (the meniscus) inside the metallic canister. The experimen-
RkJQdWJsaXNoZXIy MTU5NjU0Mg==