The AE Source Location Method For Rock-like Materials Based On Wavelet Analysis

Acoustic emission (AE) activity reflects internal damage evolution in rock-like material. Therefore, the real-time monitoring of the AE activities during the failure process has become an important measure to evaluate the internal damage in the rock-like materials. AE events always occur at the location where the deformation localizes, and also the damage level is of course related to the AE energy release. AE signals may contain so many important details about micro fractures, such as the location, mode, magnitude (strength), and energy release rate and so on. However, due to the complexity and non-stationary characteristics of AE signals, it’s difficult to use traditional signal analysis theory to recognize useful information and thus restrict its application for damage evaluation. In this paper, the wavelet analysis is used to analyze the AE data, which is of great significance in theory and practice for deriving the real AE source location information.Based on the laboratory experiments, theoretical analysis and numerical modeling the AE characteristics during rock damage are studied. Aiming at enhancing precise of the AE signals analysis, the wavelet analysis is used to process the acquired experimental AE data, and the more accurate AE source location is achieved when the improved location algorithm based on time differences is adopted. The main contents of this work are as follows:(1) AE technology can be used for real-time tracking and monitoring of failure process of rock materials under uniaxial compression. The AE characteristics during the different stage of deformation and failure process of rock have been analyzed, based on which, the failure mechanism of rock, as characterized as damage initiation, damage evolution and the ultimate failure process, is revealed. The rock damage is characterized by cumulative AE rate and ralative AE energy releasing rate.to indicate the inconsistence between them. Meanwhile, the RFPA3D is used to simulate the AE characteristics of rock under 3-dimensonal loading conditions, and the numerical results are compared well with the experimental ones.(2) According to the non-stationary feature of AE signals, an adaptive filer model based on the wavelet analysis is established by combining the adaptive filting and wavelet analysis together. This model uses the noise components decomposed from wavelet transforms as the input of adaptive filter, and the excess noise having overlapping spectrum with the original signals can be filtered to the greatest extent. This method can be applied in the processing of AE signals to verify the effectiveness of the denoising method.(3) Since the choice of initial values exerts influence on the positioning accuracy and convergence rate, a Geiger iterative optimization algorithm based on the least square method is proposed. Making full use of the estimation of property of least square, the initial values problems of Geiger iterative method can be solved effectively, and convergence of iterative algorithm is confirmed as well as the convergence rate is accelerated.(4) Based on the wavelet analysis for the study on acoustic emission signal power spectral density, the time-frequency energy distribution of acoustic emission signals is derived from using wavelet time-frequency energy analysis. Also, the time-delay estimation about waves of the same mode and the same frequency is made according to the energy distribution differences in different time and at different mode waves. Furthermore, more precise positioning of the time difference for the sound source is achieved, which can provide an effective approach for identification of AE sources by wavelet analysis.(5) The multi-scale wavelet decomposition is applied on the original signals to decompose the received signs into the waveforms of a single mode. Then, by picking up some wavelet coefficients of corresponding frequency band, and making correlation analysis to the waveform of a single mode, the time difference between AE signals can be determined by cross-correlation analysis.This dissertation provides a more reliable method to analyze and process the AE signals, to effectively decompose the noise from source signals, and to precisely inverse the accurate AE information for the damage monitoring of the rock mass.