Sensitivity-Based Model Update for Estimating Generalized Proportional Damping Parameters in a Finite-Element Model

Abstract

Geometrical simplifications, the choice of boundary conditions, modeling of connections and the determination of material property values are the steps that constitute the process of idealizing a finite element model. Therefore, there are many unknown and uncertain variables that affect how the numerical model results differ from the measured experimental data. Updating mass and stiffness by experimental modal data has been extensively studied. But identifying damping characteristics represents a next level of difficulty due to their not well-established sources. Generalized proportional damping may solve this issue by using information on damping ratio measurements, because an arbitrary variation of damping ratios can be modeled accurately by using this approach. The sensitivity method is one among several methodologies to quickly perform a model identification. Here, a prototype metallic framed structure has its mode frequencies and damping ratios measured and used to update a finite element model. Results of a test campaign of the experimental results along with the finite element model update steps are presented and show little computational effort to obtain good agreement for numerical and experimental results. A cloud of sample plots for measured parameters and the identified ones and covariance confirm the good agreement especially for the damping parameters.