Abstract (EN):
Vibration-based health monitoring systems have been successfully implemented in many different large civil engineering structures such as bridges, wind turbines and bell-towers, providing important data about their dynamic behaviour and allowing the establishment of damage detection methodologies based on the evolution of modal properties. Allying this knowledge with the emergence of very sensitive low noise sensors and high-resolution digitizers during the past few decades, an opportunity arose to implement continuously operating dynamic monitoring systems in dams and validate the suitability of these systems to monitor such massive structures with the goal of detecting damage. The detection of abnormal structural behaviour, that may indicate the occurrence of damage, can be based on control charts and is associated with shifts in the modal parameters values that are not explained by other physical phenomena but a change in the structure¿s stiffness. To test these statistical tools, modal parameters obtained from field measured time series may be contaminated with simulated damages. In this context, the ability of the control charts to identify the emergence of abnormal structural behaviour in dams is studied in this paper using data obtained from the continuous dynamic monitoring of Baixo Sabor arch dam, a concrete dam in Portugal that started operating in late 2015. Since it was not expected to find damage in such a recent structure, a set of plausible damage scenarios were simulated using a numerical model and their influence in the dam dynamic properties was used to contaminate the values of natural frequencies obtained from continuous monitoring.
Language:
English
Type (Professor's evaluation):
Scientific