Damage Identification in PC Bridges by Operational Modal Analyses

In the context of the activities outlined in the agreement between the Italian High Council of Public Works and the ReLUIS Consortium, the authors examined several case studies involving dynamic identification of scaled structural elements and real bridges, both with and without damages. The experimental activities aimed to reduce model uncertainties and provide tools for monitoring bridge structures. This paper presents the results of an experimental dynamic identification of a highway overpass characterized by a very common structural type (prestressed concrete beams simply supported over a span of about 25 m), which has incurred impact damage. The analysis of the results is used to assess the ability of modal parameters to determine the damage state of a structure when the anomaly is already present and to localize the deterioration. The efficiency of modal analysis is compared to the potential of an AI algorithm to recognize a different local response of the structure. The study results help to better understand that identifying existing damage is challenging with both approaches, highlighting the potential for improved performance in permanent monitoring systems and emphasizing the necessity of a numerical model for structure identification.