Damage detection in bridge structures including environmental effects

Damage identification for two real big bridges in Luxembourg is carried out in this paper. Vibration responses were captured from different types of excitation such as measurable and adjustable harmonic swept sine excitation and hammer impact. Before, different analysis methods were applied to the data measured from these structures showing interesting results. However, some difficulties are faced, especially due to environmental influences (temperature and soil-behaviour variations) overlaid to the structural changes caused by damage. These environmental effects are investigated in detail in this work. First, the modal parameters are identified from the response data by means of Wavelet Transform. In the next step, they are stochastically collected and processed through Principal Component Analysis. Damage index is based on outlier analysis.