Assessing the Behaviour of Engineering Structures: Application to the Dallet Steel Bridge

Abstract Structural health monitoring is a fast-growing field, playing a key role in optimizing the service life of structures and maintaining reliable, safe service conditions. This multidisciplinary field has undergone extensive development over the last three decades, notably through advanced vibration-based methods. In this context, the Institut Pascal and CIDECO have jointly developed an acquisition system and a series of robust analysis algorithms, enabling the automated detection and localization of structural damage. The aim of this work is to present the experimental results obtained on metal structures. The Dallet bridge is a riveted steel structure dating back to 1899, which underwent major renovation work between 2023 and 2024. An initial, one-day acquisition campaign was carried out prior to the renovation work. This preliminary campaign enabled the updating of a simplified numerical model of the structure (stiffness of elements and elasticity of supports). Subsequently, a second campaign was carried out, monitoring the progressive demolition of the deck. A modal analysis and the updating procedure of a simplified finite element model allowed for tracking the evolution of the structural behaviour during these critical construction phases. In addition, a detailed numerical model enabled the comparison between mode shapes obtained from an advanced model and those derived from the updated simplified model.