UTokyo Repository 東京大学
 

UTokyo Repository >
131 地震研究所 >
東京大学地震研究所彙報 >

このページ(論文)をリンクする場合は次のURLを使用してください: http://hdl.handle.net/2261/12796

タイトル: Frequency Dependence of Acoustic Emission Activity in Rocks under Incremental, Uniaxial Compression
その他のタイトル: 一軸圧縮増加応力下における岩石のアコースティック・エミッション発生頻度の周波数依存性
著者: Ohnaka, Mitiyasu
Mogi, Kiyoo
著者(別言語): 大中, 康譽
茂木, 清夫
発行日: 1981年8月31日
出版者: 東京大学地震研究所
掲載誌情報: 東京大学地震研究所彙報. 第56冊第1号, 1981.8.31, pp. 67-89
抄録: Acoustic emission generated during the deformation of rock under compression was monitored simultaneously through different frequency windows under a fixed condition of the monitoring channels throughout experimental runs, and the count rate of the emission events monitored through a low frequency window was compared with that of those monitored through a higher frequency window throughout the whole process from application of loading to failure. The deformation process of brittle rock under incremental compression may be commonly divided into the following five stages in terms of the emission rates. At low stresses in the region where a flurry of acoustic emission activity occurs, the activity is mostly due to low frequency acoustic emissions, presumably resulting from the closing of preexisting cracks with low aspect ratio at large angles to the compression axis (first stage). At higher stresses, the activity dies down to a very low level (second stage). As the stress is further increased gradually, again acoustic emission activity begins to build up and steadily increases (the emission rate increases exponentially with time when the applied stress rate is constant): both the emission rates n(l) and n(h) monitored through a low frequency window and a higher frequency window, respectively, increase with time at statistically the same proportion; that is, n(l)/{n(l)+n(h)} is nearly constant through- out this stage (third stage). As the failure time approaches, both n(l) and n(h) increase supraexponentially with time, and the increase rate of n(l) becomes more rapid than that of n(h) until immediately before failure (fourth stage); in other words, this stage is characterized by an increase in n(l)/ {n(l)+n(h)} with time or by a higher level of n(l)/ {n(l)+n(h)}. The final (fifth) stage is characterized by a very rapid acceleration of higher frequency acoustic emission activity immediately before and during failure; as a result, failure occurs immediately after n(l)/ {n(l)+n(h)} is lowered. One interesting feature in these stages is that the increase rate of the acoustic emission counted through a lower frequency window becomes more rapid with time in the fourth stage. This is not due to the change in an applied stress level itself, but due to the change in structure within the deforming rock; the two possible explanations are given for this in this text. The beginning of the fourth stage is not concurrent with the onset of dilatancy: appreciable dilatancy always begins to occur prior to the beginning of the fourth stage. The effect in this stage is consistent with the fact that a greater number of emission events with larger amplitude increase in relative terms (namely, the decrease in m-value) as rock approaches failure, since crackings resulting in larger size tend to generate acoustic waves with both larger amplitudes and lower frequency components. Thus, it is presumable that the decrease in m-value and the frequency dependence of acoustic emission rate are different manifestations of an aspect of the fracturing process to failure. A greater number of higher frequency emission events observed in the fifth stage may be thought to be those originating around the transducers as a result of local stress concentrations near the transducers during the process of forming final macroscopic cracks. The effect found here may be used for understanding structural instability of the region where earthquakes (or rock bursts in mines) can occur potentially, and for predicting the time when a main earthquake occurs in the region, in terms of frequency contained in seismic waves if a sizable microearthquake population is observed through transducers placed near the fault.
常温・一軸圧縮応力下における岩石の変形過程で発生するアコースティック・エミッション(AE)の頻度に周波数依存性がみられるかどうか,もしみられるとすれば試料内部の微小破壊の進行と共にどう変化するかを実験的に詳細に調べた.一般に,PZTのような圧電素子では分極軸に直交する方向でも電気機械結合係数が分極方向と実際には同程度故,厚み,直径(円板状の場合)の両サイズに対応した共振周波数でそれぞれ大きな感度を持つ.同時にそれらの高調波もあらわれる.さらに,この圧電素子を金属容器に固定して換振器とすると容器との結合による共振が別にあらわれる.したがって,このような換振器の周波数特性は一般に極めて複雑で広範囲にわたり感度差が著しい.筆者らは,圧電素子による換振器のこのような特徴をむしろ積極的に利用することにし,特定の周波数付近のピークを残し,他はすべてフィルターで消すことによって,限られた周波数範囲の窓L(~30kHz),I1(~250kHz),I2(250kHz~400kHz),H(~1MHz)を用意し,実験毎に各窓を通じて得られるAEイベントの発生率を加荷重開始点から破損に至る全過程を通じて計数した.イベント計数法ではデッドタイム及びしきい値の設定を的確にしなければならない.各周波数窓から得られるAE波形の継続時間は異なる故,予備実験により多数のAE波形を観測し,適当なデッドタイムを各周波数の窓毎に定めた.しきい値が低いと,AE最頻発時には波形が互いに重なり合い見かけ上一連の波形となって,イベントの識別が不可能になる.これを避けるため,しきい値をノイズレベルよりはるかに高く設定した.AEは1イベント毎に波形(振幅,周波数)が異なる故,周波数依存性を調べるためには統計的処理が不可欠である.本方法のユニークな点は,破損までの全過程を通じチャンネル毎に一定の条件を満たすイベント全てを計数するので,チャンネル間の統計的比較が容易にできることである.試料として直径50mm,高さ125mm.の円柱状に整形された新小松安山岩,万成花崗岩を用い,換振器を試料側面の中央部又は端部に固定した.
URI: http://hdl.handle.net/2261/12796
ISSN: 00408972
出現カテゴリ:東京大学地震研究所彙報
東京大学地震研究所彙報

この論文のファイル:

ファイル 記述 サイズフォーマット
ji0561004.pdf1.88 MBAdobe PDF見る/開く

本リポジトリに保管されているアイテムはすべて著作権により保護されています。

 

Valid XHTML 1.0! DSpace Software Copyright © 2002-2010  Duraspace - ご意見をお寄せください