Energy based correlation criteria in the mid-frequency range

Abstract

Aircraft structures are characterized by their lightweight design. As such, they are prone to vibrations. Numerical models based on the Finite Element Method often show significant deviations when the mid-frequency range is considered, where strong interaction between vibrations and acoustics is present. Model validation based on experimental modal data is often not possible due to the high modal density that aircraft fuselage structures exhibit in this frequency range. Classical correlation criteria like the Modal Assurance Criterion require mode shapes and can therefore not be applied. Other correlation criteria using frequency response data, such as the Frequency Domain Assurance Criterion, are highly sensitive to even small structural modifications and fail to indicated the correlation between test and analysis data in the mid-frequency range. Nevertheless, validated numerical models for the mid- to high-frequency ranges are a prerequisite for acoustic comfort predictions of aircraft cabin. This paper presents a new method for the correlation of response data from test and analysis in the mid-frequency range to support model validation in the mid-frequency range and to enable the usage of finite element models in this frequency range. The method is validated on a stiffened cylindrical shell structure, which represents a scale-model of an aircraft fuselage. The correlation criterion presented here is inspired by Statistical Energy Analysis and is based on kinetic energies integrated over frequency bands and spatially integrated over surface areas of the structure. The objective is to indicate frequency bands where the finite element model needs to be adjusted to better match with experimental observations and to locate the areas where these adjustments should be applied.