Abstract (EN):
A continuous dynamic monitoring system of an arch bridge in operation for more than 2 years has
permitted to create a very complete database with the time evolution of the first 12 natural frequencies of
the bridge. Firstly, the time evolution of the modal parameters automatically identified from the data
continuously collected by the monitoring system is used to test alternative static and dynamic regression
models for the elimination of environmental and operational effects on the natural frequencies.
Afterwards, it is evaluated the ability of Shewhart control charts to detected realistic damages, simulated
with a numerical model. This statistical tool is applied to the natural frequencies corrected by the
previously selected dynamic regression model and post-processed by a Principal Components Analysis. It
is demonstrated that with an adequate selection of the control chart parameters, this permits the detection
of very small frequency shifts (few milli-Hz), as the ones due to the four simulated damage scenarios
(localized bending stiffness reductions of 10%).
Language:
English
Type (Professor's evaluation):
Scientific
No. of pages:
16