Statistical Investigations of Uncertainty Impact on Experiment-Based Identification of a Honeycomb Sandwich Beam

Experimentally, errors on measurement points’ coordinates, among others, could affect identification results. These errors can be committed by engineer or result from measuring tools and conditions. Resulting coordinates’ variability is modeled in this work by uncertainties and is included into an experiment-based identification process to identify, in a wave propagation framework, the wavenumber and the wave attenuation of a honeycomb sandwich beam. The proposed process combines a Variant of the Inhomogeneous Wave Correlation (V-IWC) method and a sample-based uncertainty propagation method: the Latin Hypercube Sampling. Vibratory fields, which are used as inputs of the identification process, are computed experimentally. Both deterministic and statistical investigations of identified wavenumber and damping are performed. Results prove the efficiency of the proposed V-IWC method on wide frequency ranges and the robustness of identification against uncertainties. Moreover, if some measured vibratory fields do not match associated measurement points’ coordinates, no damping sensitivity to such uncertainty is detected.