Acoustic emission (AE) analyses have been used for decades for rock mechanics testing, but because AE systems are not typically calibrated, the absolute sizes of dynamic microcrack growth and other physical processes responsible for the generation of AEs are poorly constrained. We describe a calibration technique for the AE recording system as a whole (transducers + amplifiers + digitizers + sample + loading frame) that uses the impact of a 4.76 mm free-falling steel ball bearing as a reference source. We demonstrate the technique on a 76 mm diameter cylinder of westerly granite loaded in a triaxial deformation apparatus at 40 MPa confining pressure. In this case, the ball bearing is dropped inside a cavity within the sample while inside the pressure vessel. We compare this reference source to conventional AEs generated during shear loading of a saw-cut simulated fault in a second granite sample at confining pressures up to 120 MPa. All located AEs occur on the saw-cut surface and have moment magnitudes ranging from M -5.7 down to at least M -8. Dynamic events that rupture the entire simulated fault surface (stick-slip events) have measurable stress drop and macroscopic slip, and radiate seismic waves similar to those from a M -3.5 earthquake. The largest AE events that do not rupture the entire fault are M -5.7. For these events, we also estimate the corner frequency (200- 300 kHz), and we assume the Brune earthquake source model to estimate source dimensions of 4-6 mm. These AE sources are larger than the 0.2 mm grain size and smaller than the 76 × 152 mm fault surface. Finally, we compare our results to other calibrated AE studies performed on different loading machines and discuss reasons for the observed maximum AE magnitude.


Acoustic emissions (AEs) are tiny seismic events thought to be caused by microcracking or slip instability on the grain scale. They are sometimes recorded during rock mechanics experiments to monitor fracture and faulting processes [1]. In slow loading experiments on rock samples containing pre-existing artificial faults, AEs tend to cluster around stick-slip instabilities (dynamic events that involve slip of the entire fault surface) in a manner reminiscent of foreshocks and aftershocks. It has long been assumed that AEs are in some sense small-scale versions of earthquakes and that they can provide insights into earthquake mechanics [2- 5]. Yet, while earthquakes are routinely quantified by their seismic moment, only rarely is the absolute size of an AE determined. This is because AE recording systems are not typically calibrated.

This content is only available via PDF.
You can access this article if you purchase or spend a download.