Probabilistic Estimation of the Oblique Incidence Sound Absorption Coefficient - A Numerical Concept Study

Abstract

The in situ absorption characteristics of sound absorbing panels can be inversely estimated from sound field measurements above a specimen under free-field conditions or in a semi-anechoic chamber. In any case, the measurement and the inference process are subject to numerous sources of uncertainty. In this contribution, a Bayesian approach is applied to assess the uncertain, frequency-dependent material properties (complex-valued wavenumber and density) of a layer made of melamine foam. In the initial phase of inference, the quantities are inferred at a discrete frequency based on a diffuse prior. In the sequential phase, the posterior mean at an analyzed frequency serves as the prior mean at the subsequent frequency. The uncertain oblique plane wave incidence absorption coefficient then follows from the identified material properties. A virtual source antenna, modeled as a single monopole source, which is sequentially moved parallel to the specimen's surface, is used to excite the sound field and a combined pressure-particle velocity probe records the specific impedance above the specimen. We demonstrate the application of the framework for simulated noisy measurements in a semi-anechoic chamber and assess the uncertain material properties and the absorption coefficient.