| Accurate condition assessment of bridge decks and detection of early signs of deterioration is essential for economic management of aging highway bridges. One of the most common problems in concrete bridge decks is corrosion induced deck delamination. Impact echo (IE) is an ultrasonic high frequency seismic method, which has been successfully implemented in detection of delaminations in concrete bridge decks. The IE test is non-destructive, simple, accurate, and more importantly, it allows detection of delaminations at various stages of progression. However, the bridge deck condition assessment using IE in many cases is qualitative and, therefore, can not be automated. Well established criteria quantifying the detectability of delaminations and the progress of deterioration do not exist. Also, there is a strong need to use more advanced analysis methods for analyzing the IE signals.; To develop the detectability criteria for the IE test, an extensive finite element analysis was carried out to simulate IE test results on the surface of delaminated decks. It was found that the detectability of delaminations depends on their depth, size to depth ratio and their position relative to the IE receiver. It was also shown that the impact duration, the source and receiver spacing in the IE test setup, and their relative location with respect to the delaminated area, are important in accurate interpretation of IE results. To investigate the potential of the IE test in long term monitoring of bridge decks, two probable scenarios of deterioration progression were simulated. It was shown that the gradual worsening condition of the deck will reflect in IE test results and, therefore, can be recognized and quantified.; Time-frequency methods were proposed as complementary analysis methods to enhance the interpretation of IE test results. The advantages of using the time-frequency methods over the traditional Fourier-based spectrum in analyzing non-stationary IE signals were pointed out. Continuous and discrete wavelet transforms were introduced and their applications in providing an efficient time-frequency representation and "denoising" of IE signals were illustrated by a number of examples. The concept of Hilbert-Huang transform was demonstrated by analyzing synthetic signals. Hilbert spectrum has a great potential in better representing the frequency content of IE signals. |
