Automatic operational modal analysis: Challenges and practical application to a historical bridge

An automated modal identification procedure is described in the paper. The proposed methodology comprehends two key points: (a) selection of the optimal input parameters (affecting the quality of the estimated modal parameters) for the well-known Covariance-driven Stochastic Subspace Identification technique; (b) automatic analysis of stabilization diagrams, performed though a first check of modal complexity control and subsequent clustering. The proposed approach is applied to the data collected on a historic iron arch bridge. The complexity of the structure also highlights some identification problems related to the modal splitting phenomenon. In the last part of the paper, the estimates of natural frequencies provided by the proposed procedure are compared to those independently obtained with a manual interpretation of stabilization diagrams produced by the Data-driven Stochastic Subspace Identification technique.