Numerical prediction of jet-noise based on “first principles”

A brief outline is presented of a numerical system developed by the authors during the last five years and aimed at building a non-empirical (i.e., based on “first principles”) tool for prediction of the noise radiated by the exhaust jets of aviation engines with an industrially useful accuracy of 2–3 dB and as wide frequency range as possible. In this system the computation of the aerodynamic and turbulent characteristics of the jets is performed with the use of Large-Eddy Simulation (LES) coupled with an implicit high-order finite-volume scheme implemented on structured multi-block grids in curvilinear coordinates. For the far-field noise computation the system uses the integral method of Ffowcs–Williams/Hawkings (FWH). Thanks to a number of original features in the implementation of both the aerodynamic and aeroacoustic parts of the system, even on relatively coarse grids with only 2–5 million nodes, the system yields results which tangibly surpass in terms of accuracy those obtained by other authors on grids with tens of millions of nodes. The capabilities of the system are demonstrated by examples of computation of aerodynamic and noise characteristics of simple round jets within a wide range of Mach number (from 0.4 to 2) and temperature ratio (up to ${\sim}3$). Other than that, examples are presented of applications to the noise prediction of a number of more complex jets, gradually getting closer to the real exhaust jets of modern engines. In particular, we consider under-expanded sonic jets with intense shocks, jets from dual nozzles with staggered exits, and, finally, an exhaust system including a dual nozzle with an extended central body. The accuracy of the noise prediction in all the considered cases is close to the “target” accuracy of 2–3 dB for both the integrated noise and its directivity and the spectral characteristics. The well-resolved frequency range corresponds to a maximum diameter Strouhal number ranging from 2 to 5, depending of the size of the grid and the jet's parameters.