Constrained maximum likelihood modal parameter identification applied to structural dynamics. (May 2016)

• Record Type: Journal Article
• Authors: El-Kafafy, Mahmoud; Peeters, Bart; Guillaume, Patrick; De Troyer, Tim
• Is Part Of: Mechanical systems and signal processing. Volume 72/73(2016)
• Page Start: 567
• Page End: 589
• Publication Date: 2016-05
• Subjects: Reciprocity -- Real (normal) mode shapes -- Modal parameters -- Maximum likelihood -- Constraints -- Modal analysis; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.10.030
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: A new modal parameter estimation method to directly establish modal models of structural dynamic systems satisfying two physically motivated constraints will be presented. The constraints imposed in the identified modal model are the reciprocity of the frequency response functions (FRFs) and the estimation of normal (real) modes. The motivation behind the first constraint (i.e. reciprocity) comes from the fact that modal analysis theory shows that the FRF matrix and therefore the residue matrices are symmetric for non-gyroscopic, non-circulatory, and passive mechanical systems. In other words, such types of systems are expected to obey Maxwell–Betti׳s reciprocity principle. The second constraint (i.e. real mode shapes) is motivated by the fact that analytical models of structures are assumed to either be undamped or proportional damped. Therefore, normal (real) modes are needed for comparison with these analytical models. The work done in this paper is a further development of a recently introduced modal parameter identification method called ML-MM that enables us to establish modal model that satisfies such motivated constraints. The proposed constrained ML-MM method is applied to two real experimental datasets measured on fully trimmed cars. This type of data is still considered as a significant challenge in modal analysis. The results clearly demonstrate the applicability of the method to real structures with significant non-proportional damping and high modal densities. Highlights: We introduce a constrained maximum likelihood modal parameters estimator. The applied constraints are: the FRFs reciprocity & Real mode shapes estimation. The proposed estimator is validated by means of two industrial applications. This estimator outperforms the classical way used to apply those constraints.