WAVELET DENOISING FOR LOW STRAIN PILE INTEGRITY TESTING

Abstract

Low strain impact test is widely used to assess the structural integrity of reinforced concrete piles due to its high productivity and cost effectiveness. However, a low signal-to-noise ratio may prevent proper evaluation of pile length and integrity using the standard data analysis approaches. In this paper, we propose a technique for the time-frequency analysis of low strain test data, which allows us to separate the useful components of a signal from the unwanted ones. The technique is based on the continuous wavelet transform with the complex Morlet wavelet, which is shown to be the most suitable for the low strain test data decomposition. To filter the signal, the moving median of the square modulus of the continuous wavelet transform is calculated in the frequency band of the initial impact pulse. The capabilities of the technique are illustrated by the low strain test signals with the artificial noise and the results of a field test of a 30m long bored pile.