Abstract
Acoustic emission method is a most useful method for detection of cracks generated in concrete structure. Burst-type AE signals from cracks are to be discrinated from environmental noises under noisy conditions from water flow in a concrete canal bridge. Consequently, FFT, AR model, and AIC method are applied to classify AE signals. Results show that the frequency ranges and the amplitudes of AE signals become high in the case of low water level. AE signals from cracks are effectively discriminated by combination with AIC method. The detection ratio is improved by employing the peak frequencies in AR model. Thus, it is promising to detect crack damages by AEsignals with the relation to water levels in concrete canal bridges.
