Dynamic Model Updating of Al-Al Honeycomb Sandwich Panels for Aerospace Applications

Resumo

In aerospace environment, it is crucial to develop reliable dynamic models to accurately predict the
structural behavior of aircraft. The established approach involves constructing numerical models using the Finite
Element Method and utilizing experimental data for model updates and improvements. This paper focuses on the
construction and updating process of dynamic models applied to Al-Al honeycomb sandwich panels, which serve
as the main structure of the Brazilian Geostationary Satellite. Two numerical models are proposed to replicate the
honeycomb plate's geometry, including a simple equivalent laminated plate, and a face plate-equivalent solid core
model. Experimentally obtained parameters are utilized to update the numerical models using a Bayesian

optimization algorithm, which finds equivalent values for physical parameters enhancing the numerical-
experimental correlation of natural frequencies. Since this process is probabilistic, Monte Carlo simulations are

performed to ensure convergence of the obtained values. The results demonstrate that even the lower complexity
equivalent plate model can adequately represent the panel, making it suitable for preliminary analysis and saving
computational time compared to the higher complexity model. Overall, this paper presents a comprehensive
approach to constructing and updating dynamic models of honeycomb sandwich panels, demonstrating their
effectiveness in accurately capturing the dynamical behavior of aerospace structures.