Advanced calibration of a 3D masonry arch bridge model using non-destructive testing and numerical optimisation

Historical masonry arch bridges constitute the backbone of many existing transportation networks in different countries in Europe and worldwide. They represent valuable cultural heritage assets and play an essential social and economic role. Since construction, old masonry bridges have accumulated structural damage from traffic and environmental actions. Furthermore, depending on their geometrical and mechanical characteristics, they may be particularly vulnerable to extreme events like earthquakes. Thus, accurate structural assessment under different loading conditions is critical for the conservation of these structures. Realistic assessment requires suitable numerical models to represent the characteristic 3D behaviour. The complexity of this task is further compounded by the practical difficulty in obtaining essential information on the internal bridge structure and the masonry mechanical parameters, which are vital to achieve accurate response predictions against service and extreme actions. This paper presents an advanced calibration procedure for a refined macroscale bridge model, allowing for the anisotropic nature of the masonry material. The proposed calibration approach is applied to an actual multi-span masonry viaduct, where sonic, ultrasonic, and ground penetrating radar tests are conducted to investigate the internal structure of the viaduct and determine the elastic properties of the masonry materials. In addition, the dynamic characteristics of the bridge are evaluated through in-situ measurements under environmental vibrations and used for model validation. The results from a standard simplified model calibration and an enhanced calibration are compared considering the vibration modes of the bridge. Simplified calibration is carried out using the results from in-situ tests, while a statistic inference procedure and numerical optimisation are adopted in the refined calibration to achieve improved accuracy. Although the paper focuses on a specific case study, the adopted methodology can be easily applied to studying other masonry bridges and cultural heritage masonry structures.