Civil Engineering Reference
In-Depth Information
12.2 Review of Previous Work
In this section, we review some of the most recent and relevant work published by researchers in the area of vibration based
structural health monitoring of steel beam-concrete slab bridges. We focus on two specific aspects: (1) identification of
structural mechanisms of traffic load transfer (also referred to in the literature as load distribution factors) and (2)
identification of dynamic response characteristics, namely, natural frequency, mode shapes and damping.
12.2.1 Estimation of Load Transfer Mechanisms from Measured Data
Kim and Nowak [
2
] used strain measurements induced by operational traffic loads to investigate conventional I-girder
bridges load distribution factors and to assess AASHTO recommendations on dynamic impact factors [
3
]. The strains were
measured on the bottom flange of I-girders. The study was conducted on two similar bridges during a span of 48 h and
recorded data induced by approximately 900 trucks. The researchers concluded that the measured load distribution factors
(LDF) and dynamic impact factors are consistently lower than the AASHTO recommendation. Bre˜a et al. [
4
] monitored a
conventional I-girder type highway overpass using strain measurements under a controlled live-load test. Strain
measurements were used to estimate LDFs and these results were compared with the finite element model (FEM) load
distribution factors. The researchers found that although the bridge did not have explicit shear transfer members (such as
shear studs), the strain measurements across the cross section (assuming Bernoulli's hypothesis of linear strain distribution)
was consistent with the condition of I-girders acting as composite with the reinforced concrete slab. Chakraborty and
DeWolf [
5
] developed and implemented a continuous strain monitoring system on a three-span conventional composite
I-girder overpass. The study reported on data over a period of 5 months. This study analyzed the LDFs using AASHTO
recommendations, field measurements, and FEM analysis. The study also included the determination of the location of
the height of the neutral axes when large trucks travel across the bridge. One of the conclusions of this study was that the
measured strain levels and LDFs are typically significantly below those recommended by AASHTO. The authors stated that
this is a byproduct of conservative simplifications typically used in conventional designs, such as not including
redundancies, connection restraints, and the way in which loads are distributed to different parts of the structure. This
conclusion is in agreement with previous studies [
2
]. Finally, Jauregui et al. [
6
] conducted controlled loading tests on a
standard I-girder bridge and used strains measured at three points through the depth of the I-girders. Results of the
investigation show that even though the girders did not have explicit elements to transfer shear forces in through the
interface with the concrete slab, they behaved as partially composite sections in the positive moment region up to the onset of
yielding.
In conclusion, the results presented in the existing literature typically use strain gauges to determine load distribution
factors and the extent of composite action in typical slab-on-girder bridges. Strain data was measured either in controlled
load tests or during operational traffic. In some cases LDFs from measurements were compared with FEMs and the
AASHTO code.
12.2.2 Finite Element Model Updating
Finite element model (FEM) updating refers to the collection of computational methods that attempt to solve the inverse
problem of estimating user-defined free parameters of a finite element model based on the minimization of a user-defined
objective function of the residual between features extracted from measured vibration data on a system and model prediction
of those features [
7
]. It is assumed throughout the process that the finite element model is consistent with the system, i.e., that
there exists a set of free parameters that will minimize the residuals to a level consistent with the measurement noise
(Morozov discrepancy principle). Numerous studies have been conducted in the realm of FEM updating for bridge
evaluation and diagnosis. Here we will focus on discussing some of the most recent literature on FEM updating regarding
slab-on-girder bridges.
Sanayei et al. [
8
] developed a baseline finite element model for bridge management using strain data collected during
controlled static load tests. Feng et al. [
9
] applied a neural network-based system identification technique to establish and
update a baseline FEM based on acceleration and strain measurements during operational traffic-induced vibrations.
Brownjohn et al. [
10
] assessed the effects of a bridge strengthening. Experiment-based structural assessments consisting
