2.1 Experimental equipment and rock samples

The experiment was completed on the creep perturbation power-impact loading experimental machine that can facilitate external heterogenous perturbation as shown in Figure 1. The equipment has high testing accuracy and stable performance; the parameters of the host loading system in the static loading unit are as follows: a vertical load range of 0–800 kN; a static load loading displacement rate of up to 150 mm/min; a dynamic loading unit maximum vertical load of 100 kN; a dynamic load loading displacement rate of up to 100 mm/min; and a maximum waveform perturbation frequency of 10 Hz. The experimental machine adopts the Deminzer magnetostrictive displacement transducer with a range of 200 mm and an accuracy of 0.002 mm for measuring the piston displacement. In this paper, the acoustic emission equipment is utilized to monitor the experimental process and collect the acoustic emission signals; the threshold of the acoustic emission system is set at 43 dB; and the sampling frequency is 1 MHz. Before the experiment, lead break coupling experiments were conducted on the sensors to ensure that the amplitude of the probe signals was above 90 dB and that the coupling effect between the acoustic emission sensors and rock samples was evident. The end of the sensors was coated with petroleum jelly and fixed to the specified position before the experiment.

               

In this paper, AS, MS, and FS stand for muddy siltstone, medium-grained siltstone, and fine-grained siltstone, respectively. These three rock samples used in the experiments were obtained from the No. 1 coal mine in New Shanghai. All samples were taken from the same rock so as to reduce the dispersion of rock samples. The processed sandstone rock samples are shown in Figure 2. The stress–strain relationship curves and damage fracture diagrams of the samples under uniaxial loading conditions are displayed in Figure 3 to provide a basis for the determination of the parameters of the subsequent cyclic perturbation experiments.

               

               

2.2 Experimental program

The rocks were subjected to uniaxial compression and cyclic disturbance, which can investigate the deformation response characteristics of rocks near the proximal face of a deep mine to cyclic dynamic disturbance. Three types of rocks, namely, muddy siltstone, fine-grained siltstone, and medium-grained siltstone from the No. 1 Coal Mine of Shanghai Temple were subjected to the experiment of “uniaxial compression + cyclic disturbance” to investigate the deformation response characteristics of rocks near the proximal face of a deep mine to cyclic dynamic disturbance. The specific experimental parameter settings are shown in Table 1, and the experimental loading path is shown in Figure 4. The stress ratios of 60%, 70%, and 80% were considered high, medium-high, and ultrahigh stress levels, respectively. The specific experimental steps are as follows: before the experiment, the processed rock samples were numbered and grouped into specific loading programs as shown in Table 2. The average uniaxial compressive strengths of muddy siltstone, fine-grained siltstone, and medium-grained siltstone were measured to be 27.25, 38.05, and 43.86 MPa, respectively. First, the axial force was given to the preset level σm (average of the upper limit of the cyclic disturbing stress and the lower limit of the stress), and the perturbation stress was then loaded to the preset level σm (average of the upper limit of the perturbation stress [denoted as σmax] and the lower limit of the perturbation stress [denoted as σmin]; the amplitude of perturbation was calculated as ∆σ = σmax – σmin). The rock samples were rested for 5 min at this stress level to stabilize the redistribution of the stresses inside the rock samples. Since the rest time was short, the rheological effect was ignored accordingly. The rock samples were loaded to a frequency of 5 Hz with displacement loading control mode, with a loading rate of 0.1 mm/min, control cycle amplitude of 5 MPa, and cyclic perturbation effect under the sinusoidal wave. If the deformation was still stable after 1000 times cycle, then the axial force was to be applied until the rock sample was destroyed. The experiment was completed in approximately 18–25 min.

               

3.1 Mechanical property

3.1.1 Stress–strain curve analysis

The stress–strain curves of the specimens are shown in Figure 5, where point O represents the origin, point A represents the beginning of specimen disturbance under the preset stress level, point B represents the end of the specimen cycle 1000 times, point C represents the peak stress, and point D represents the end of the experiment. Different stress levels affect the damage pattern of the specimens differently. For the fine-grained and medium-grained specimens, in the early OA stage of the experiment, the specimens witness a compression stage, and the compression stage curve exhibits an upward bending shape. When the specimens enter the elastic stage, they are subjected to sinusoidal cyclic perturbation, which corresponds to the AB stage. Further application of axial stress corresponds to the BC stage, and the destruction of the specimens corresponds to the CD stage. For muddy siltstone, the damage occurs after 1000 times cycle at the stress level of 60% and before 1000 times cycle at the stress levels of 70% and 80%. The hysteresis loop corresponding to the stress level of 70% is more intensive than that corresponding to 80%, and the number of cyclic perturbations under the stress level of 80% is less than that under 70%. These findings indicate that with an increase in the stress level, the load-bearing capacity of the specimens decreases and damage occurs at stress levels of 80% and 70%. For fine-grained specimens, the specimens are destroyed after 1000 times cycle of cycling under the stress levels of 60% and 70%. With the stress level further increasing to 80%, the specimens are destroyed after nine times of cyclic perturbation. For medium-grained specimens, they are destroyed after nine times of cyclic perturbation under stress levels of 60%, 70%, and 80%. However, damage is not observed after 1000 times cycle of sinusoidal cyclic perturbation under the stress levels of 60%, 70%, and 80% and has only occurred after the continuous application of axial stress. This finding indicates that muddy and fine-grained siltstone are more sensitive to sinusoidal cyclic perturbation than medium-grained siltstone, and fine-grained siltstone is the most sensitive to sinusoidal cyclic perturbation.

                 

3.2 Damage characterization

Figure 6 and Table 3 show the postdamage pictures of muddy siltstone, fine-grained, and medium-grained samples under different stress levels. The surface cracks of the samples are also sketched. When the stress level is 60%, the muddy siltstone samples show X-shaped shear damage. With the gradual increase in the stress level, long steeply inclined cracks appear in the middle and upper part of the sample, implying heightened tension and damage. The fine-grained sample mainly shows splitting into two parts along the direction of applied axial force; the rupture surface is rough, indicating the damage of the sample with fragment crumbling and strong sound. This splitting is accompanied by the collapse of the sample debris and a violent sound; the ruptured surface becomes rough, and brittle damage characteristics are observed. When the stress level is 60% and 80%, the sample shows compression-shear damage. When the stress level is 70%, the sample exhibits single-beveled shear damage. For medium-grained samples, the damage characteristics are more abundant than those observed in mud sandstone, sandstone, and other sandstones. For the medium-grained sample, the damage is more complicated than that for the muddy siltstone and fine-grained samples. In particular, the main crack extends from the upper end to the lower end of the sample, and the secondary crack extends from the middle and lower end of the sample to the edge of the right upper end. The sample exhibits compression-shear damage when the stress level is 60%, 70%, and 80%. As shown in the macroscopic damage diagrams of the samples, the degree of sample fragmentation increases with the applied stress level, implying that stress level is one of the important factors affecting the damage pattern of the samples.

               

3.3 Acoustic emission analysis of damage processes in specimens with different lithologies

The trends of acoustic emission energy, counts, and cumulative counts during deformation and damage were obtained by monitoring the acoustic emission of muddy siltstone, fine-grained, and medium-grained samples as shown in Figure 7. The acoustic emission ringer counts can reflect the degree of damage within the sample, and the acoustic emission energy can be used to reflect the process of crack development and expansion within the sample. As shown in Figure 7, the trend of the acoustic emission activity in terms of energy and ringing counts over time can be divided into three phases I, II, and III, which correspond to the weakening, stabilizing, and surging periods, respectively.

               

In the weakening period of acoustic emission activity (phase I), no acoustic emission activity is observed and an increase in the cumulative ringing counts is not evident due to the small number of internal cracks in the specimen at the beginning of the experiment. When the axial stress reaches the preset stress level, the acoustic emission energy and counts start to increase and the acoustic emission activity passes a steady phase II during cyclic perturbation. The specimen undergoes macroscopic damage: the cracks inside the specimen start to develop slowly and gradually penetrate through the specimen. When the axial stress reaches the peak, a large number of acoustic emission events are generated. At this time, the acoustic emission counts surge, the energy also increases gradually, and the acoustic emission activity enters into the surging phase III. The acoustic emission activity also serves as a precursor to the damage of the specimen. The acoustic emission counts generated in phase III increase with the stress level when the mudstone reaches the peak stress or during the last cycle. The same is true for fine-grained and medium-grained sandstones. During the sinusoidal cyclic perturbation, the acoustic emission counts from the mud siltstone, fine-grained sandstone, and medium-grained sandstone increase with increasing stress levels in Acoustic Emission Stage II. As illustrated in the figure, the acoustic emission characteristics are generally consistent among the specimens with different lithologies at different compression stages.

3.4 Acoustic emission waveform characterization

The acoustic emission equipment can monitor the elastic waves during the development of the internal fissures of the specimens and respond to the energy changes in the three different lithological specimens during cyclic perturbation. Thus, the main frequency eigenvalues of the key points were obtained by extracting key points N1 and N2 in the acoustic emission characteristics, transforming the random signals from the time domain to the frequency domain by using the Fast Fourier Transform, and obtaining amplitude–frequency–time data by using the MATLAB 3D diagram. The effect of dynamic perturbation on the acoustic emission characteristics of muddy siltstone, fine-grained, and medium-grained sandstone was then analyzed.

As shown in Figure 8, the 3D plots of the main frequency eigenvalues of the acoustic emission from the key points N1 and N2 of three different lithological specimens under different stresses are obtained by Fourier transform. The points corresponding to the amplitude in the plots are the main frequency eigenvalues as shown in Table 4 and Figure 8. Under different cyclic perturbation experiments, the specimen is transformed from the high main frequency of key point N1 to the low main frequency of key point N2, and the decrease in the main frequency eigenvalue is large. The high-frequency eigenvalue of the waveform signal represents the small-scale damage of the specimen, and the low-frequency eigenvalue of the waveform signal represents the large-scale damage of the specimen.

               

               

The main frequency eigenvalues of key points N1 and N2 for different types of rock samples are shown in Figure 9. For the muddy siltstone specimen, the main frequency eigenvalue of key point N1 is in the range of 192–201 kHz. In particular, the lowest main frequency eigenvalue of 192 kHz corresponds to a stress level of 70%, and the highest main frequency eigenvalue of 201 kHz corresponds to a stress level of 60%. Meanwhile, the main frequency eigenvalue of key point N2 is in the range of 76–87 kHz, with the lowest main frequency eigenvalue of 76 kHz corresponding to a stress level of 80% and the highest main frequency eigenvalue of 87 kHz corresponding to a stress level of 60%.

               

For the fine-grained specimen, the main frequency eigenvalue key point N1 is in the range of 169–178 kHz. In particular, the lowest main frequency eigenvalue of 169 kHz corresponds to a stress level of 80%, and the highest main frequency eigenvalue of 178 kHz corresponds to a stress level of 60%. Meanwhile, the main frequency eigenvalue of key point N2 is in the range of 73–83 kHz, with the lowest main frequency eigenvalue of 73 kHz corresponding to a stress level of 80% and the highest main frequency eigenvalue of 83 kHz corresponding to a stress level of 60%.

For the medium-grained specimen, the main frequency eigenvalue range of key point N1 is in the range of 158–165 kHz. In particular, the lowest main frequency eigenvalue of 158 kHz corresponds to a stress level of 80%, and the highest main frequency eigenvalue of 165 kHz corresponds to a stress level of 60%. Meanwhile, the main frequency eigenvalue range of key point N2 is in the range of 70–74 kHz, with the lowest main frequency eigenvalue of 70 kHz corresponding to a stress level of 80% and the highest main frequency eigenvalue of 74 kHz corresponding to a stress level of 60%.