Field test investigations for condition monitoring of a concrete culvert bridge using vibration responses

Abstract

Summary Over the last two decades, vibration‐based structural health and condition monitoring has gradually become one of the main methods for monitoring changes in a structure on a global scale. This method has been applied to damage detection in many areas, including aerospace, mechanical and structural systems. However, it has not been widely used for identifying condition change of the structure itself. This paper demonstrates the feasibility of utilising acceleration and strain responses measured from a short‐span concrete culvert bridge located in New South Wales, Australia, to identify the effects of structural condition changes made to the deck due to bridge widening. In addition to the operating structural responses, two controlled traffic tests were carried out, before and after the bridge widening, in support of verifying the results from the modal analysis. A comparative study is presented evaluating the results obtained by two identification algorithms, namely, frequency‐domain decomposition (FDD) and frequency–spatial domain decomposition (FSDD), from which the link between the decrease in modal frequency and the event of bridge expansion as condition change is successfully established. The contribution of this work is to emphasise the application of FDD and FSDD on a short‐span concrete culvert bridge, in order to drive the interest from large‐scale bridges to short‐span bridges so as to increase the coverage of current bridge health monitoring.

A web‐based and design‐oriented structural health evaluation system for long‐span bridges with structural health monitoring system

“…There are several extreme value distribution models available 11,23 . If the time duration of available measurement data is short (e.g., more than 2 years but less than 20 years), the generalized Pareto distribution (also called GPD or POT) can be adopted 23 . If the time duration of available measurement data is long enough (e.g., more than 20 years), the generalized extreme value (GEV) distribution can be used 11 …”

“…To have a good prediction of the site‐specific extreme values, all the relevant and quality measurement data shall be included in the prediction. There are several extreme value distribution models available 11,23 . If the time duration of available measurement data is short (e.g., more than 2 years but less than 20 years), the generalized Pareto distribution (also called GPD or POT) can be adopted 23 .…”

“…The early SHM technology focuses on structural damage detection including system identification and model updating, 13,14 but it turns out to be very difficult for detecting damage of long‐span bridges because their sizes are huge, the environment around and loads on their structures are complex, but the damage often begins locally 15,16 . Many investigations are thus devoted to analyzing the data recorded by SHM systems to get a realistic understanding of environments and loads as well as their effects on the bridges 17–24 . Some studies are also performed to evaluate the bridge performance based on health monitoring data.…”

“…15,16 Many investigations are thus devoted to analyzing the data recorded by SHM systems to get a realistic understanding of environments and loads as well as their effects on the bridges. [17][18][19][20][21][22][23][24] Some studies are also performed to evaluate the bridge performance based on health monitoring data. For example, the fatigue performance and life of long-span bridges have been evaluated using SHM data.…”

Extracting the time‐dependent resonances of a vehicle–bridge interacting system by wavelet synchrosqueezed transform

“…He et al 7 proposed a mass-normalized mode shape identification method to enrich the mode shape resolution for bridge structures by using a parked vehicle on the bridge under environmental excitation. Lin et al 8 identified the modal frequencies of a short-span concrete bridge before and after bridge deck widening by applying frequency domain decomposition techniques to the bridge under a blocked traffic condition. Spiridonakos et al 9 proposed a method to consider the problem of parametric output-only identification of a time-varying structures based on autoregressive moving average technique to capture the varying frequencies of a laboratory scaled bridge-like structure.…”