Structural identification and its application in condition assessment for constructed facilities

Nondestructive evaluation of constructed facilities requires the integrated analytical and experimental work. The experimental results are used in the analytical model calibration. The physical properties of every component of a structure and its associated boundaries need to be identified with certain accuracy. The procedure of utilizing experimental results to identify the analytical model is called structural identification (SI). The goal of nondestructive evaluation is to find the residual capacities of a structure so that a decision can be made to maintain safety. Since the accuracy of the analytical model depends on the accuracy of the experimental results, it is extremely important to obtain valid experimental results which represent current state of the constructed facility under concern. In this thesis, the new SI classification incorporating both analysis and experiment is given in order to have a better view of various SI methods. Both modal testing and static testing were applied to the Cross-county Highway Bridge in Cincinnati area. A filter to obtain valid modal test data from raw data of multi-reference impact test is developed. The quality of experimental results can be ensured via this filter. A methodology to conceptually calibrate analytical model is proposed. The application of this approach leads to a successful structural identification of the Cross-County Highway Bridge. The overall percentage error for 18 frequencies of the 3-D model is around 6%. Considering that the experimental frequency error can be up to 5%, this kind of match indicates substantial analysis/test agreement. In order to explore a simpler representation of the bridge, a 2-D grid model was built, calibrated and optimized using MATLAB programming. The analytical results of this optimized grid model correlate to experiment well. To locate damages, the damage indices are developed. These new approaches are very sensitive to the damage which can detect the initial damage. The new method together with the existing methods can give a complete and reliable damage location for the constructed facilities.