Validation of a uniaxial structure-borne sound benchmark with emphasis on power and phase accuracy

Abstract The evaluation of energetic quantities by means of structure-borne sound power is an attractive approach in analyzing coupled vibroacoustic systems and material characterization. When cyclic mean power quantities are determined for non-resonant frequencies of a system, the accuracy of the relative phases between force and acceleration signals can be crucial for both measurement and simulation-based procedures. In particular, in the case of light damping, the frequency response of the mean power is sensitive to small errors in the phase. This paper first discusses the problem of the phase-correct determination of power quantities from a theoretical perspective. Secondly, a benchmark structure is introduced in order to validate corresponding experimental and numerical methods. For this purpose, the benchmark has a finite element representation, which correlates to hardware realization. All the data required to perform the validation are provided. Based on this benchmark structure, the numerical solution for structure-borne power is first reviewed and generalized. Thereafter, for uniaxial vibration an experimental methodology to determine power quantities is developed in the frequency domain. It is observed that filtering the measurement data with a frequency-dependent phase correction, which includes the predetermined phase error of the individual measurement chain, can lead to physically significant and accurate results of structure-borne sound power. Based on these experimental results, damping values are extracted for any resonant and non-resonant frequency. Additionally, the benchmark case aids in explaining effects, such as apparent non-passive system behaviors that are observed in the measurements with impedance heads.