CFD optimization of an immersed particle heat exchanger

An innovative immersed particle heat exchanger has been recently proposed by the authors: it makes use of very small solid particles as intermediate medium to perform heat transfer between two gas flows at different temperature. The potential of such heat exchanger has been already demonstrated by the authors, who have developed a 1D model for the computation of the pipe length that ensures a prescribed heat exchange and for the evaluation of the influence of particle characteristics. As completion of the heat exchanger design procedure, this paper provides a numerical procedure for the design optimization of the other geometric parameters defining the heat exchanger, such as diameters and angles of inlet and outlet pipes and particle injection mode. The objective of this numerical optimization is to maximize the heat exchanger efficiency by maximizing the dispersion of the particles falling in countercurrent within the flow. Optimized configurations have been obtained using two different optimization algorithms: these results demonstrate that a numerical optimization based on 3D thermo-fluid-dynamic simulations is needed to complete the design of the heat exchanger: such approach, which incorporates 3D flow paths and particle trajectories, allows to limit the undesired horizontal shifts of the falling particles near elbows and thus to retrieve a thermal efficiency closer to the theoretical value calculated by the 1D model. The experimental validation of the employed 3D model has been performed on the initially-guessed configuration, where the 3D features cited above are more evident.

► Heat exchangers are considered a key driver in the development of recuperated and externally fired gas turbine cycles.
► The authors developed and tested an innovative heat exchanger in the past using solid particles.
► The prototype design highlights a non-optimal geometric configuration of the heat exchanger.
► The design of the innovative heat exchanger is enhanced in this paper using CFD optimization.