Identification of acoustic emission sources in injected rock assisted by pattern recognition techniques

Recording and analysis of noises produced by disturbance of rock mass allow the identification of instable zones. This is due to the fact that a rupture is generally preceded by disorders of different nature, which can be detected with geophones. This paper present an experimental study of identification of acoustic emission sources, emanating from different mechanisms in an injected rock mass under load. The different mechanisms occurring in an injected rock under load were modelled in the laboratory by different tests and samples. The injected rock was modelled with a square plate (200×200×20 mm) in which was inserted a grout disk of 20 mm diameter. The statistical pattern recognition techniques were applied to classify the recorded signals.