Large-scale Biaxial Friction Experiments Using a NIED Large-scale Shaking Table: Design of Apparatus and Preliminary Results

摘要

To investigate the scale dependency of friction and the characteristics of dynamic rupture propagation, we developed a large-scale biaxial friction apparatus using the large-scale shaking table (15 m×14.5 m) at the National Research Institute for Earth Science and Disaster Prevention in Japan. We carried out a series of large-scale friction experiments using this apparatus in the spring of 2012. In the apparatus, the actuators of the shaking table were used as the engine of the constant speed loading force applied to a pair of specimens made of Indian gabbro. A1.5-m-long quadrangular prismatic specimen overlaid on a 2-m specimen was used. Their height and width were 0.5 m. The lower 2-m-long specimen moves with the shaking table and the upper 1.5-m-long specimen is fixed to the basement of the shaking table by a reaction force bar. The shaking table can generate a maximum displacement of 0.4 m with a velocity ranging between 0.025 mm/s and 1m/s. Before the series of experiments, the sliding surface was flattened with less than 0.01 mm undulation using a large-scale surface grinder. However, the surface roughness evolved with the subsequent experiments. Normal stress was applied up to1.3 MPa. The stiffness of this apparatus was measured experimentally and was of the order of 0.1GN/m. We measured the coefficients of friction under a constant loading velocity between 0.025 and 100 mm/s. The coefficient of friction was about 0.75 under the normal stress between 0.13 and1.3 MPa for the loading velocity of 0.1mm/s and1mm/s. As the loading velocity increased, the coefficient of friction decreased. The obtained coefficient of mction was consistent with those obtained by previous studies using smaller specimens. We then monitored the stick-slip events by observing the shear stress change evolution measured by strain gauges as well as by the high-frequency wave emission measured by piezoelectric transducers. These sensors were attached at the edge of the slipping area. We found that there were many stick-slip events that nucleated inside the sliding surface but did not reach the edge of the sliding surface until the termination of slip. The locations of these high-frequency events correspond to the place where the slip was nucleated. We found that these locations were on the edge of the grooved surfaces that were created in the previous experiments.%岩石摩擦のサイズ依存性と破壊伝播の詳細を調べるため，大型二軸摩擦試験機を開発し，2012年の 春に，防災科学技術研究所所有の大型振動台（テーブルサイズ15 m ×14.5 m)を用いて，大型摩擦実 験を行った.この実験においては，振動台の起震機を利用して岩石試料間に一定の食い違い速度を与 えた.岩石試料は，インド産斑糲岩を用い，長さ1.5 mの四角柱試料を長さ2 mの試料に重ねて実験に 用いた.各試料の高さと幅はそれぞれ0.5 mである.下側試料は振動台に固定されており，振動台とと もに動く.上側試料は反カバーを介して振動台基礎と接続されている.振動台は，最大変位0.4 m， 0.025 mm/sから1m/sの速度を出す事ができる.実験に先立って，岩石試料のすべり面は，大型研削盤を用 いて凹凸が0.01 mm以下になるように整形した.しかしながら，実験回数を重ねるに従い，すべり面 の荒さは進化していった.垂直応力は，1.3 MPaまでかけて実験を行った.試験機の強度を実験中に測 定を行い，0.1GN/mであった.スティックスリップが顕著に発生しない0.1mm/s-1 mm/sの載荷速度 における摩擦係数を測定したところ，法線応力が0.13〜1.3 MPaの間において，約0.75であり，この 値は，小さいサイズの岩石試料を用いた実験で得られた値とほぼ一致する値であった.さらに，ステイツ クスリップイベントの破壊伝播をひずみゲージアレイを用いて調べ，破壊がすべり面中央部から始ま り岩石試料の両端に達しないイベントが多数発生している事がわかった.これらのイベントは，すべ り面が摩耗してできた溝の端から始まっている事がわかった.