*Kotoyo TSUCHIDA1, Hironori KAWAKATA1, Eiichi FUKUYAMA2, Futoshi YAMASHITA2, Kazuo MIZOGUCHI3
(1.Ritsumeikan University, 2.National Research Institute for Earth Science and Disaster Prevention, 3.Central Research Institute of Electric Power Industry)
Keywords:stick-slip event, rupture propagation, large-scale biaxial frictional experiment
Pre-slip was expected to occur prior to large earthquakes, since a pre-slip model was proposed by Ohnaka and Kuwahara (1990) based on their rock frictional experiments. The pre-slip accelerates toward an unstable sliding event. However, such phenomena have never been clearly observed for natural earthquakes. Ohnaka and Kuwahara (1990) observed a 1-D strain distribution along a sample surface, and estimated the apparent rupture propagation speed. In addition, the fault was narrow, and the rupture growth might be affected by free surfaces at the edge of the sample, though the free surface effect is not so common for natural earthquakes. Therefore, we closely observed two-dimensional rupture propagation on a wider fault during rock frictional experiments. We carried out meter-scale rock frictional experiments (Fukuyama et al., 2013), and investigated rupture propagation of stick-slip events and some of their characteristics, using AE (acoustic emission) and strain records. The fault consisted of an interface of two Indian gabbro blocks. Their width and height were 0.5 m, and the length of upper and lower blocks were 1.5 m and 2.0 m, respectively. The arrays of strain gauges and AE sensors were installed within the lower block in order to understand two-dimensional rupture propagation. Twenty four sets of AE sensors and biaxial strain gauges were attached 60 mm below the sliding surface at intervals of 150 mm parallel to the slip direction and at intervals of 75 mm perpendicular to the slip direction. We analyzed time series of strain and AE data, and found stick-slip events accompanied with slow and accelerating strain decrease that propagated at a speed much slower than elastic wave speed. This study was supported by NIED research project "Development of monitoring and forecasting technology for crustal activity" and JSPS KAKENHI Grant Number 23340131.