ANALYSIS OF SIGNAL ACIVITY DETECTION IN ACOUSTIC EMISSION

Abstract

In recent years, the dynamic behavior of solid structure defects is extremely important as a small defect that is growing may well be more
significant than a larger stable defect. Acoustic Emission (AE) is the method used to investigate the behavior of defects under stress. The
importance of the AE is to determine the source location when it occurs. It is a real time monitoring technique. Identifying the actual sources of
elastic waves during rapid local stress relaxation in solids under load is the major point in acoustic emission non-destructive testing, seismology
and soon. This relies heavily on the accuracy of the arrival time detection process. Conventionally block thresholding technique is used to detect
the Acoustic Emission, but accuracy is less in this method. The main focus of this work is to increase the accuracy of the real time signal
detection and to verify actual phase picking transient waveforms of minimum amplitude, using novel wavelet transform based algorithm. This
algorithm relies on the fact that noise commonly manifests itself as fine-grained structure in the signal, and Wavelet Transform (WT) provides a
scale-based decomposition. This algorithm was evaluated in different types of acoustic emission tests, demonstrating the excellent temporal
localization of the phases picked, even for the signals with minimum signal to noise ratio (SNR) and time of arrival (TOA) of the signal is
detected exactly. This method is applied for different signals having different frequency sampling (Fs). In this work signal activity and time of
arrival is determined using wavelet transform and block thresholding method. The results are obtained for signal activity and time of arrival for
both techniques. From these results the accuracy is high for wavelet transform method when compared to block thresholding algorithm.
Comparing to block threshold algorithm the wavelet based approach is applied to the signals with low amplitude and low SNR. These results will
be helpful to find whether the signal will intrusive or not.