Direct Computation of Aerodynamic Noise

Abstract

In his pioneering work Lighthill formulated the problem of predicting the noise radiated by an unsteady flow in terms of an acoustic analogy. However, in order to predict the far-field noise the acoustic source terms of such acoustic analogies must be known accurately. Such detailed information is usually not available and empirical models of the acoustic sources are utilized to arrive at an empirical prediction of the far-field noise. Recent advances in direct and large eddy simulations of turbulent flows provide an alternative to this empiricism. The far-field noise can be obtained from first principles by using the ‘exact’ source terms from DNS/LES in an acoustic analogy. Developing the methods to make this possible and studying the nature of particular flow processes responsible for the most intense noise are the principal objectives of the research underway at Stanford.