Hypersonic intake design using a cfd data-driven multi-objective optimisation strategy

The present study proposes a robust optimisation strategy for the automatic shaping of inlet intakes in strong compressible conditions. Reynolds Averaged Navier-Stokes databases are mined inside Genetic Algorithms to generate optimal solutions concerning a priori defined goals. Based on an intake geometry available in literature, a validated steady-state model is used for multi-objective optimisation with the aim to reduce the drag coefficient while increasing the static and the total pressure ratios. The Pareto optimal solutions are analysed, looking at the flow patterns behaviour responsible for the improvements. Although the procedure is applied to a specific case, the method is entirely general, providing a smart flowchart for super/hypersonic intake design. In particular, since CFD data drive the optimisation, the process accounts for the whole complexities of such devices, including shock/boundary-layer interactions, recirculation bubbles and separate flow portions from the earliest design stages.