|
Vibration-based methods of structural health monitoring are generally founded on the principle that localized damage to
a structure would exhibit changes within the global dynamic response. Upon this basis, accelerometers provide a unique
health monitoring strategy in that a distributed network of sensors provides the technical feasibility to isolate the onset of
damage without requiring that any sensor be located exactly on or in close proximity to the damage. While in theory this
may be sufficient, practical experience has shown significant improvement in the application of damage diagnostic
routines when mode shapes characterized by strongly localized behavior of specific elements are captured by the
instrumentation array. In traditional applications, this presents a challenge since the cost and complexity of cable-based
systems often effectively limits the number of instrumented locations thereby constraining the modal parameter
extraction to only global modal responses. The advent of the low-cost RF chip transceiver with wireless networking
capabilities has afforded a means by which a substantial number of output locations can be measured through referencebased
testing using large-scale wireless sensor networks. In the current study, this approach was applied to the Prairie du
Chien Bridge over the Mississippi River to extract operational mode shapes with high spatial reconstruction, including
strongly localized modes. The tied arch bridge was instrumented at over 230 locations with single-axis accelerometers
conditioned and acquired over a high-rate lossless wireless sensor network with simultaneous sampling capabilities.
Acquisition of the dynamic response of the web plates of the arch rib was specifically targeted within the
instrumentation array for diagnostic purposes. Reference-based operational modal analysis of the full structure through
data-driven stochastic subspace identification is presented alongside finite element analysis results for confirmation of
modal parameter plausibility. Particular emphasis is placed on the identification and reconstruction of modal response
with large contribution from the arch rib web plates.
|
