Real-time damage detection of RC columns from their acceleration responses by estimating dominant Lipschitz-Hölder exponent

Abstract

Quick state judgement of important infrastructures such as bridges is required after earthquakes. We focused on accelerometers which are easy to be installed and already installed on many infrastructures. We tried to detect the damage of structures by applying singularity analysis based on Multi-fractal analysis to the acceleration data which are measured on RC column at large shaking table tests. Some singularity points of acceleration responses were detected by wavelet transform when the damages judged from their strain data and video record occured. However, their wavelet spectrum reaction was not clear. Therefore, we computed dominant Lipschitz-Hölder exponent of the acceleration waveform from singularity spectrum which can quantitatively evaluate singularity, and it was shown that the moments of some damage occurrences were detected form the dominant Lipschitz-Hölder exponents.