Efficient low-rank model order reduction of complex vibroacoustic benchmark problems for aircraft cabin noise assessment

Innovative aircraft concepts will drastically impact the interior noise inside cabin.Therefore, it is sensible to push for an inclusion of vibroacoustic assessments in early design phases using high-fidelity wave-resolving models, as to yield exhaustive insights into the pressure distribution inside the cabin.However, in an early design stage, there is missing crucial data, so these large-scale simulations have to be evaluated multiple times.This contribution aims to present an approach for efficiently considering uncertain loads due to a turbulent boundary layer in a vibroacoustic setting by extending a recently proposed method to a more complex vibroacoustic benchmark problem.Benchmark models are important in understanding problems, because they contain a similar underlying structure, while often entailing less computational effort than a full vibroacoustic aircraft model.Therefore, it is feasible to evaluate the performance of a method on benchmark models, gradually increasing complexity until the full model can be examined.In this contribution, low-rank approximation and model order reduction are combined to allow for a rapid holistic model evaluation of a stiffened plate coupled to a cavity excited by many load vectors.It becomes evident that gradually increasing complexity of models when developing methods can be helpful.