A Matrix-Free Model Order Reduction Scheme for Vibro-Acoustic Systems including Complex Noise Control Treatments

This paper presents a Matrix-Free rational Krylov interpolation model order reduction framework as an efficient tool to predict vibro-acoustic field variables of interest in complex systems including (multi-)layers of damping treatments. The model order reduction scheme does not require knowledge on the underlying mathematical model structure and works based on given transfer functions between the system inputs and outputs. Hence, it can just as well be considered as a black box model, not requiring access to the system matrices and thus justifying the naming "Matrix-Free". The method can be applied to any type of problem, being bounded or unbounded, damped or undamped. A strategy is proposed to iteratively expand the reduced order model until the requested accuracy is achieved. The method allows to overcome the computational cost of highly detailed vibro-acoustic models with frequency-dependent noise control treatments which require a high number of degrees of freedom to accurately capture the different variables of interest given the complex near-field dynamics. The potential of the method is illustrated on different examples, including a car interior.