Numerical Methods for Aeroacoustic Analysis of Turbomachinery

Air transportation is today an integral part of many peoples life and it continuously grow to serve more of us in new and ingenious ways. There are however negative sides to this growth as well, such as increased noise around airports and more CO2 emissions, which must be addressed if aviation is to be part of a sustainable future. A key ingredient to successfully addressing these issues is accurate and reliable numerical tools that can be used to investigate and/or optimize new and existing technologies.This thesis presents a computational framework for predicting the noise generated by a fuel efficient aircraft engine propulsor known as the Counter Rotating Open Rotor (CROR). The framework solves the Favre-Averaged Navier Stokes equations using the Finite Volume method for spatial discretization and the Harmonic Balance method for temporal discretization in order to obtain a numerical solution of the transient flow field around the CROR blades. This solution is further coupled with an acoustic analogy in which a convective form of the Ffowcs Williams-Hawkings equation for permeable integration surfaces is used to determine the far field noise signature of the CROR at any arbitrary observer location. The underlying theory of both the temporal discretization scheme (Harmonic Balance method) and the acoustic analogy (Ffowcs Williams-Hawkings equation) are presented in detail in order to investigate the applicability of these methods for predicting CROR noise. It is found that the Harmonic Balance method performs well for predicting turbomachinery noise in general, but also that issues with aliasing might impair the accuracy of the results when the flow-field becomes strongly nonlinear. The acoustic analogy is found to perform well for propagating the types of noise generated by a CROR. Issues may however occur when wakes and shock waves cross the integration surface. This makes the placement of the integration surface a complicated task for the particular formulation of the Ffowcs Williams - Hawkings equation used in this wor