A Novel Method for Estimating Acoustic Emission b Value Using Improved Magnitudes

The coustic emission (AE) b value plays an important role in studying structural damage evolution and disaster precursors. The b value can be obtained from the least square fit between AE event number and magnitude (the Gutenberg-Richter relationship), where the magnitude is generally calculated through the maximum amplitude of an AE signal. However, the magnitudes obtained from amplitudes of different AE signals may be quite different, due to the influence of coupling state between the sensor and specimen (expressed as the sensor sensitivity coefficient, ${Sc}$ ). Thus, the b values obtained from a single sensor and without magnitude correction may not be reliable. To address this, this study used pencil-lead breaks and far-field theory to obtain ${Sc}$ values. One ${Sc}$ value may be as much as double another, demonstrating that signal maximum magnitudes should be corrected before b value estimation. The ${Sc}$ -corrected median magnitude is then treated as the AE event magnitude and used to estimate the b value. Application to uniaxial compression AE data from a red sandstone indicates that the improved magnitude reduces the instability compared to magnitude based on a single sensor. The b value time series among sensors exhibit large differences, while the improved b values agree well with AE event locations and rock specimen failure mode. This research demonstrates for the first time the necessity of amplitude correction in AE magnitude and b value computation.