摘要: |
This paper presents a method for identifying the flexural rigidity of bridges with limited structural information using modal frequencies identified from measured acceleration data. The output of this study provides a simple approach that can be adapted for condition assessment, the bridge load rating process, and nondestructive evaluation. The overall methodology is generic and can adapted for different types of bridge, but in this paper the approach is evaluated for skewed reinforced concrete slab bridges because the national database of bridges indicates that slab bridges represent the largest subset among bridge types. A large number of slab bridges with different structural dimensions, such as skew angle, span, width, and thickness, were first analyzed using the finite-element method to obtain their modal frequencies and their corresponding nondimensional frequency parameters, which play an important role in identifying the flexural rigidity of slab bridges. This population of generated data was then used to create an artificial intelligence model, which was developed to predict the nondimensional frequency parameters for slab bridges with different geometrical characteristics. Moreover, an algorithm based on variational mode decomposition was presented to identify the modal frequencies and damping ratio of a bridge. The method was first validated with a set of numerical studies, and it was then applied to a highly skewed reinforced concrete slab bridge in the Commonwealth of Virginia for load rating purposes. The bridge was instrumented with wireless accelerometers; the vibration responses of the bridge under ambient loading and impact hammer testing were recorded. Finally, the flexural rigidity of the bridge was identified from the established relationship between the modal frequencies and the flexural rigidity. Results show that the proposed method is capable of predicting flexural rigidity and can be used as a basis for the load rating of bridges without complete structural information. |