Conceptual design of a liquid-metal divertor for the European DEMO

• A realistic liquid metal divertor concept is designed for the European DEMO reactor. • An armor of 3D-printed porous tungsten, filled with liquid tin, is used. • Heat loading capability is increased compared to baseline designs, in both steady state operation and during slow transients. • Resilience against disruptions is not outside the realm of possibility, but requires experimental testing. Liquid metal (LM) divertors are considered for the European DEMO reactor, because they may offer improved performance compared to the tungsten monoblock concept. The goal of this work is to provide a concept design, and explore the limitations of liquid metal divertors. To this end, a set of design requirements was formulated in close collaboration with the EUROfusion Power Plant Physics and Technology team (responsible for the design of the EU-DEMO). Tin was chosen as the preferred liquid metal, because unacceptable Tritium retention issues arise when lithium is used in DEMO. A concept design was then chosen that consists of water cooled pipes that are square on the outside and round on the inside, a corrosion barrier, and a 3D-printed porous tungsten armor layer filled with liquid tin. The porous armor layer acts as a Capillary Porous System (CPS). The design was analyzed using thermo-mechanical FEM simulations for various armor thicknesses and heat sink materials: Densimet, W/Cu composites, and CuCrZr. The highest heat loading capability achieved is 26.5 MW/m 2 in steady state (18.9 MW/m 2 when taking into account a safety margin of 1.4). This is achieved using a CuCrZr pipe, with a 1.9 mm thick armor. When increasing the armor layer to 3 mm thick, more than 80 MW/m 2 can be withstood during slow transients thanks to vapor shielding, but at the same time the steady-state capability is reduced to 18 MW/m 2 . Resilience against disruptions cannot yet be proven, but is deemed within the realm of possibility based on estimates regarding the behavior of vapor shielding. This should be further investigated. Overall, the concept is considered a significant improvement compared to the original specifications (which are also the specifications to the tungsten monoblocks: 10 MW/m 2 in steady state, and ∼ 20 MW/m 2 during slow transients). Moreover, the possibility of withstanding disruptions is regarded as a potentially major improvement.