Stochastic Modeling and Uncertainty Quantification in the Operational Modal Analysis of a Honeycomb Sandwich Panel

Resumo

The aeronautical and aerospace industries have been carried out researches to develop structures that combine cost-effectiveness with excellent mechanical properties. In this context, honeycomb sandwich panels stand out due to their high stiffness and low structural weight. In aerospace applications, these panels are often subjected to complex dynamic behavior, requiring advanced methods to analyze modal parameters under operational conditions and ensure structural integrity. Operational Modal Analysis (OMA) emerges as a robust approach that complements models based on the Finite Element Method (FEM), which are commonly used for simulating structures with dynamic behavior. OMA uses stochastic models to identify modal parameters accurately, even without knowledge of excitation forces. However, its application is subject to uncertainties associated with system variables such as nominal mechanical properties, geometric tolerances, manufacturing effects, and acquisition equipment errors, which may affect the accuracy of modal parameter estimation. Given the importance of accurately estimating modal parameters under real operational conditions, this study proposes an uncertainty analysis to quantify the influence of variability in the modal parameter estimation procedure for a honeycomb sandwich panel. The tested panel represents a structural element of SGDC 1, the first Brazilian geostationary satellite. The SSI-COV and SSI-DATA methods were used to extract modal parameters and, to quantify uncertainties, laboratory tests were conducted and the Monte Carlo method was used to generate synthetic responses with added noise, forming a randomized sample space. The SSI methods were applied to each dataset using computational routines, allowing modal parameter estimation. The goal is to quantify the impact of noise uncertainty on modal parameters estimation, identify the most sensitive variables, and support modeling strategies and the application of OMA to improve result reliability in aerospace projects. The results showed that the SSI-COV method yielded more reliable results, being less sensitive to noise and having lower computational cost. Although more sensitive to noise, the SSI-DATA method also provided reliable results and is recommended for cases using only direct structural response data