An Integrated Structural Health Monitoring and Digital Twin Framework for Condition Assessment of Gravity Dams

This study presents a Digital Twin-based Structural Health Monitoring framework for the assessment of existing concrete gravity dams. The proposed methodology integrates distributed sensing, automated signal processing, finite element modelling, model calibration and validation, degradation sensitivity analysis, and digital visualization within a unified monitoring framework. A network of inertial sensors is deployed along the dam to characterize its dynamic and quasi-static response. Hourly monitoring records are processed to identify natural frequencies using Fast Fourier Transform, Power Spectral Density, and Peak Picking techniques, while acceleration measurements are also used to derive displacement and inclination indicators. A 3D finite element model developed in ANSYS Mechanical APDL is calibrated by iteratively updating the effective Young’s modulus and density until the relative error between numerical and experimentally identified modal frequencies is below 5%. Independent validation using inclination measurements resulted in relative differences between 0.74% and 4.92%. The calibrated Digital Twin is subsequently used to simulate progressive stiffness degradation and establish relationships between degradation, modal properties, and structural deformation. The resulting model-informed thresholds are integrated into a digital monitoring platform for automated condition assessment and anomaly detection. The framework provides a physically interpretable approach for complementing conventional dam safety inspections with long-term quantitative structural health information.