Daniel Lindblad (Imperial College London)

Jet noise has remained the dominant source of aircraft noise during take-off since jetliners were introduced over half a century ago. To date, most reductions in jet noise have been achieved by increasing the diameter of the engine, thereby reducing the jet velocity while keeping the thrust constant. In the future, it will be hard to maintain this trend since the space under the wing is limited. Therefore, in order to continue reducing the noise levels from aircraft, and thereby minimize the negative impacts of aviation, new noise reduction concepts must be developed. To this end, high-fidelity numerical simulations that can predict the complex noise generation mechanisms found in a turbulent jet are needed. The purpose of this work is to explore the benefits of using the compressible flow solver in Nektar++ for this purpose. The compressible flow solver in Nektar++ uses the high-order discontinuous Galerkin method to discretize the unfiltered Navier-Stokes equations on unstructured grids. By using a high-order discretization in combination with unstructured grids, we should be able to predict higher frequencies than the current state of the art. This is in turn important to predict the total noise levels, and thereby assess the benefits of noise reduction technologies, for high Reynolds number jets. The simulations performed with Nektar++ are coupled with the Ffowcs Williams – Hawkings method implemented in the Antares library developed by CERFACS to compute the far-field acoustic signature of the jet. The complete framework is then applied to a single-stream jet, for which experimental data obtained at the University of Southampton exists for comparison.