Numerical modelling and optimization of an impeller immersed in a molten salt bath under microwave heating for thermal energy storage applications

This thesis work has been performed in the framework of a research contract involving ENEA and the Energy department at Politecnico di Torino, which deals with thermal energy storage in the context of a hybridized CSP-PV system. The electricity produced by the PV system must be transformed into heat in order to be accumulated in a single-medium (molten salt) TES. For this purpose, microwave heating has been proposed as an efficient way to heat up the molten salt by using the electricity generated from the PV panels. However, the experience of heating up the molten salt by using microwaves is very limited; for this reason, an experiment is being carried out at ENEA to prove its feasibility. The experimental device consists of a tank filled with molten salt which is connected to a microwave generator by means of a waveguide. The numerical simulations carried out during the design phase suggested that an impeller, immersed in the salt bath, should help in homogenizing the salt temperature by increasing the mixing. The model of the impeller in the salt bath has been developed using COMSOL. First, the computational domain has been defined. It corresponds to the total salt bath composed by a rotating domain that simulates the rotation of the impeller and a large one that is around it. The impeller instead is out off the computational domain. Then, the boundary conditions have been applied on the walls of the tank, the blades and the free surface. At this point, a specific study has dwelled on mesh structure, also. After building it, a mesh independence has been made especially to gain confidence in the accuracy of model. At the end, the mesh refinement that has been applied was a compromise between error respect to the true solution and computational cost. About physics interface, different turbulent closures have been compared; finally, the SST model has been adopted. In addition, the physic model includes the action of gravity in the problem and treats the molten salt as incompressible fluid. Since mixing is a transitory problem, a specific way to elaborate a stationary study is needed. A computationally more efficient method is to simulate the flow using the frozen-rotor approach that is a modeling concept that treats the rotor as fixed, or frozen in space. Afterwards, a parametric analysis was carried out on the geometry of impeller and its position. This phase of analysis has been carried out considering graphical (streamlines and velocity field) and numerical results (punctual velocity values). The second part of the thesis deals with the development of a simplified model of the impeller, which aims at reducing the CPU time of frozen rotor approach. Indeed, the impeller is a source of complexity in the model, involving a not negligible computational cost. According to the literature, a simplified model can be obtained replacing the impeller with a system of volume forces applied to a subdomain. At the end, a thermofluidynamic stationary study has been carried out comparing the simplified model and the original (impeller) one. In this way, it has been observed what are the differences between the results obtained by the two models. Drawing the conclusions of this work, the results described in the thesis can result the starting point for future analysis and research for other type of studies about impeller performance but also especially to improve the simplified model presented.