Acoustic emission monitoring of rockbursts during TBM-excavated headrace tunneling at Jinping Ⅱ hydropower station

To better understand the mechanical properties of marble at Jinping II hydropower station, this paper examines the changes of brittle rocks in excavation damaged zones(EDZs) before and after excavation of tunnel with the tunnel boring machine(TBM). The paper attempts to employ the acoustic emission(AE) to study the AE characteristics and distribution of rockburst before and after TBM-excavated tunnel. It is known that the headrace tunnel #2, excavated by the drill-and-blast(D&B) method, is ahead of the headrace tunnel #3 that is excavated by TBM method. The experimental sub-tunnel #2–1, about 2000 m in depth and 13 m in diameter, between the two tunnels is scheduled. In the experimental sub-tunnel #2–1, a large number of experimental boreholes are arranged, and AE sensors are installed within 10 m apart from the wall of the headrace tunnel #3. By tracking the microseismic signals in rocks, the location, frequency, quantity, scope and intensity of the microseismic signals are basically identifed. It is observed that the AE signals mainly occur within 5 m around the rock wall, basically lasting for one day before tunnel excavation and a week after excavation. Monitoring results indicate that the rockburst signals are closely related to rock stress adjustment. The rock structure has a rapid self-adjustment capacity before and after a certain period of time during tunneling. The variations of rock stresses would last for a long time before reaching a fnal steady state. Based on this, the site-specifc support parameters for the deep tunnels can be accordingly optimized.

Fracture evolution around pre-existing cylindrical cavities in brittle rocks under uniaxial compression

The development of fracture around pre-existing cylindrical cavities in brittle rocks was examined using physical models and acoustic emission technique. The experimental results indicate that when granite blocks containing one pre-existing cylindrical cavity are loaded in uniaxial compression condition, the profiles of cracks around the cavity can be characterized by tensile cracking （splitting parallel to the axial compression direction） at the roof-floor, compressive crack at two side walls, and remote or secondary cracks at the perimeter of the cavity. Moreover, fracture around cavity is size-dependent. In granite blocks containing pre-existing half-length cylindrical cavities, compressive stress concentration is found to initiate at the two sidewalls and induce shear crack propagation and coalescence. In granite blocks containing multiple parallel cylindrical cavities, the adjacent cylindrical cavities can influence each other and the eventual failure mode is determined by the interaction of tensile, compressive and shear stresses. Experimental results show that both tensile and compressive stresses play an important role in fracture evolution process around cavities in brittle rocks.

A rapid audio event detection method by adopting 2D-Haar acoustic super feature vector

For accuracy and rapidity of audio event detection in the mass-data audio pro- cessing tasks, a generic method of rapidly recognizing audio event based on 2D-Haar acoustic super feature vector and AdaBoost is proposed. Firstly, it combines certain number of con- tinuous audio frames to be an ＂acoustic feature image＂, secondly, uses AdaBoost.MH or fast Random AdaBoost feature selection algorithm to select high representative 2D-Haar pattern combinations to construct super feature vectors; thirdly, analyzes the commonality and differ- ences between subcategories, then extracts common features and reduces different features to obtain a generic audio event template, which can support the accurate identification of multi- ple sub-classes and detect and locate the specific audio event from the audio stream accurately. Experimental results show that the use of 2D-Haar acoustic feature super vector can make recog- nition accuracy 5% higher than ones that MFCC, PLP, LPCC and other traditional acoustic features yielded, and can make tile training processing 7 20 times faster and the recognition processing 5-10 times faster, it can even achieve an average precision of 93.38%, an average recall of 95.03% under the optimal parameter configuration found by grid method. Above all, it can provide an accurate and fast mass-data processing method for audio event detection.