Uncertain boundary condition Bayesian identification from experimental data: A case study on a cantilever beam. (February 2016)

• Record Type: Journal Article
• Authors: Ritto, T.G.; Sampaio, R.; Aguiar, R.R.
• Is Part Of: Mechanical systems and signal processing. Volume 68/69(2016)
• Page Start: 176
• Page End: 188
• Publication Date: 2016-02
• Subjects: Uncertain boundary condition -- Clamped beam -- Bayesian identification -- Stochastic dynamics -- Structural dynamics; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.08.010
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: In many mechanical applications (wind turbine tower, substructure joints, etc.), the stiffness of the boundary conditions is uncertain and might decrease with time, due to wear and/or looseness. In this paper, a torsional stiffness parameter is used to model the clamped side of a Timoshenko beam. The goal is to perform the identification with experimental data. To represent the decreasing stiffness of the clamped side, an experimental test rig is constructed, where several rubber layers are added to the clamped side, making it softer. Increasing the number of layers decreases the stiffness, thus representing a loss in the stiffness. The Bayesian approach is applied to update the probabilistic model related to the boundary condition (torsional stiffness parameter). The proposed Bayesian strategy worked well for the problem analyzed, where the experimental natural frequencies were within the 95% confidence limits of the computed natural frequencies probability density functions. Abstract : Highlights: Uncertain boundary condition. Identification of the stiffness of the boundary condition of a beam structure. Bayesian updating with experimental data obtained by constructing a test rig. The uncertain torsional stiffness has a great impact on response amplitude. The random natural frequencies are strongly positive correlated.