The failure mechanisms of engineering rock masses primarily involve tensile and shear failure. Differentiating between the acoustic emission (AE) signals generated during the tensile and shear damage processes in rock can provide a scientific basis for the classification of acoustic signals in field rock fracture monitoring. This paper presents a study on acoustic emission monitoring during the direct tensile testing of granite, proposing a method for classifying AE signals based on the damage and failure processes of the samples. Additionally, the classification of tensile and shear AE signals is explored. The main conclusions are as follows. The proportion of low-frequency signals (frequency <200 kHz) and high-frequency signals (frequency >200 kHz) in all AE signals was found to be 81.6 % and 19.4 %, respectively. Based on an integrated classification and statistical method for AE signals in rock tensile failure, which involves steps such as “denoising the raw waveform, time-frequency domain data transformation, fuzzification processing, extraction of dominant frequency and corresponding amplitude, and identification of secondary dominant frequencies,” the AE signals were categorized into two types, A and B. Type A signals accounted for an average of 7.6 %, while Type B signals made up 92.4 %. Based on the polarity determination method, the focal mechanisms of AE (Acoustic Emission) events were identified. In tensile events, the average proportion of Type A signals was 8.34 %, while the average proportion of Type B signals was 91.66 %. The Brazilian splitting test also yielded classification results similar to those obtained from direct tensile testing. Thus, it was preliminarily concluded that Type A signals, characterized by the presence of both a primary and secondary frequency, correspond to shear signals, whereas Type B signals, which only exhibit a primary frequency without a secondary frequency, correspond to tensile signals.

中文翻译：

---

直接拉伸破坏下岩石的声发射演化和断裂机制


工程岩体的破坏机理主要涉及拉伸和剪切破坏。区分岩石拉伸和剪切损伤过程中产生的声发射 （AE） 信号，可为现场岩石断裂监测中声信号的分类提供科学依据。本文对花岗岩直接拉伸试验过程中的声发射监测进行了研究，提出了一种根据样品的损伤和失效过程对声发射信号进行分类的方法。此外，还探讨了拉伸和剪切 AE 信号的分类。主要结论如下。低频信号（频率 <200 kHz）和高频信号（频率 >200 kHz）在所有 AE 信号中的比例分别为 81.6 % 和 19.4 %。基于岩石拉伸破坏中声发射信号的综合分类统计方法，涉及“原始波形去噪、时频域数据转换、模糊处理、主频和相应振幅提取、次级主频识别”等步骤，将声发射信号分为 A 型和 B 型两种类型，A 型信号平均占 7.6%。 而 B 型信号占 92.4%。基于极性确定方法，确定了 AE （声发射） 事件的局灶机制。在拉伸事件中，A 型信号的平均比例为 8.34 %，而 B 型信号的平均比例为 91.66 %。巴西劈裂试验还产生了与直接拉伸试验相似的分类结果。 因此，初步得出结论，以同时存在初级和次级频率为特征的 A 型信号对应于剪切信号，而 B 型信号（仅表现出初级频率而没有次级频率）对应于拉伸信号。