Operational modal analysis for testing and monitoring of bridges and special structures

Dynamic testing and continuous monitoring based on Operational Modal Analysis can play an important role in the management of important civil infrastructures, since these experimental tools permit the periodic or continuous assessment of bridges, dams, high-rise buildings, stadia or offshore structures through the analysis of their response to ambient excitation, while they are in normal operation. Nowadays, the performance of dynamic tests has become very practical, even when carried out on large structures, as there are solutions that avoid the use of long electrical cables. Furthermore, the installation of continuous dynamic monitoring systems with remote control and data retrieving has become much more straightforward and economical due to the diversity of commercially available solutions. However, this highly sophisticated hardware has to be combined with powerful routines that should be able to efficiently extract useful information from data collected during dynamic tests or continuously recorded by permanent installations. Therefore, this work was firstly focused on the development and implementation of processing procedures adequate to extract accurate modal parameter estimates from the time series recorded during ambient vibration tests of large civil engineering structures. The efficiency of the presented algorithms to simultaneously process several datasets was then validated with the analysis of databases collected in the dynamic tests of the Millau Viaduct and Humber Bridge. Additionally, the accuracy of modal damping ratio estimates provided by ambient vibration tests was evaluated using data from five relevant structures where both ambient and forced vibration tests were performed. Afterwards, routines suitable for the on-line processing of data continuously collected by dynamic monitoring systems were developed and integrated in a monitoring software package. One of the most important features of the software is the automatic identification of modal parameters from the collected acceleration time series through the adoption of a new procedure suitable to be used together with the state-of-the-art algorithms: SSICOV and p-LSCF. This is complemented by algorithms that minimize the effects of environmental and operational factors on the identified natural frequencies, and by statistical tools that provide indexes adequate to automatically identify changes in the natural frequencies due to abnormal causes, as for instance, the occurrence of damages. The implemented monitoring tools were tested with two applications: one in a large arch bridge (Infante D. Henrique Bridge) and another in the roof of a stadium (Braga Stadium). In particular, the system installed in the bridge, in continuous operation for more than two years, permitted to demonstrate the usefulness and robustness of all the components included in the monitoring software.