Physical Sciences and Mathematics Commons^™

Detection Of Small-Scale Rockfall Incidents Using Their Seismic Signature, Achilleas Tripolitsiotis, Antonis Daskalakis, Stelios Mertikas, Dionysios Hristopulos, Zacharias Agioutantis, Panagiotis Partsinevelos

Mining Engineering Faculty Publications

Several algorithms have been effectively used to identify the seismic signature of rockfall incidents, which constitute a significant threat for human lives and infrastructure especially when occurring along transportation networks. These algorithms have been mostly evaluated using data from large scale rockfall events that release a large amount of energy. However, low-energy rockfall events (< 100 Joules) triggered by small-sized individual rocks falling from small heights can be severely destructive. In this study, a three-parameter algorithm has been developed to identify low-energy rockfall events. An experimental setup was implemented to 1) validate the results obtained by this algorithm against visual inspection of seismic signals records, 2) define the optimal algorithm parameterization to minimize false alarms, and 3) investigate whether tri-axial vibration monitoring can be replaced by a uniaxial device in order to reduce the installation cost of a real-time rockfall monitoring system. It was found that the success rate of the proposed algorithm exceeds 80% independently of the parameters used, while event identification at a maximum distance with minimal false alarms was achieved when using mean ± 3σ as the threshold criterion and 6 ms and 4 ms as the trigger and event window parameters respectively. Finally, it was found that for the specific experimental setup, a uniaxial device could be used for rockfall event identification.