Uncertainty Quantification in Direct Noise Computations of Cavity Feedback
* Presenting author
Abstract:
The underlying physical mechanisms of aeroacoustic cavity feedback originate from a highly non-linear and multi-scale interaction between the turbulent hydrodynamics and acoustics. This multi-scale nature causes a strong sensitivity of the noise generation process with respect to small perturbations of the incoming flow or the cavity geometry. In consequence, small uncertainties of boundary conditions may have a large influence on the resulting acoustics. This may lead to unreliable numerical predictions and to discrepancies between experimental and numerical results. The research area of uncertainty quantification provides mathematical tools to quantify the effect of such uncertainties on the numerical solution and derived quantities. This talk is devoted to introduce a framework of different uncertainty quantification methods and their application in cavity noise simulations. We present two polynomial-based methods, the stochastic Galerkin method and the non-intrusive spectral projection method, and an improved sampling-based method, the multilevel Monte Carlo method. All three methods are compared and discussed in terms of applicability in turbulent aeroacoustic simulations. Subsequently, the most promising method, the non-intrusive spectral projection method, is used in combination with a zonal LES framework. We conduct turbulent aeroacoustic cavity simulations with uncertain boundary conditions to investigate the resulting stochastic moments of the aeroacoustic signal.