Design of compact automotive heat exchanger, analysing the effects of RANS models and utilising Additive Manufacturing

Abstract: The analytical modelling of complex turbulent airflow remains one of the great unsolved mysteries of physics, but in this paper two widely used Reynolds Averaged Navier-Stokes models (k-$\epsilon$ and k-$\omega$ SST) are compared while designing a heat exchanger for the KTH Formula Student electric race car. CAD software was used to design lattices for the heat exchanger core and theorise about how to increase heat transfer while also taking into account the utilisation of metal additive manufacturing. The models were then analysed using Computational Fluid Dynamics to determine their characteristics as well as the effects of the two turbulence models.  It was found that the first iteration of the second design performed best in terms of pressure drop and generating turbulent kinetic energy closely followed by the second iteration of the second design and the second iteration of the first design. When comparing the turbulence models the results indicated agreement with their theoretical foundations. The first model overestimating turbulent kinetic energy relative to the second, which picked up more detail of near-wall turbulence thanks to better boundary layer formulation. Future work includes improving the simulation setup, correlating the results with wind tunnel testing and further evaluating more complex designs.