Abstract
Uniaxial compression experiments and acoustic emission (AE) tests were conducted on weakly consolidated sandstone samples with bedding angle of 0°, 45°, and 90° collected from Xiaojihan coal mine in Shaanxi Province, China; the acoustic emission (AE) characteristics of bedding sandstone were investigated, including the influencing mechanism of bedding angle on characteristics, spatial distribution, and value changes of AE activity. According to the research, AE ringing counting rate and energy rate are the highest and energy release is intense during the loading process when the bedding angle is 0°, and these values decrease and the energy release is gentle with rising of the angle. Based on spatial distribution, the number of AE events was the largest and cluster phenomenon occurs between the bedding plane when the bedding inclination was 0°, which decreased and sprouted along the plane of bedding plane at 45° and which was smaller and evenly distributed in the space at 90°. With the increase of bedding angle, the proportion of large-scale microcracks decreases and the variation range of microcracks narrows down. Therefore, under the same uniaxial compression conditions, the value increases by the increase of bedding angle. The research results offer reliable reference to carry out monitoring work on acoustic emission and microseism of mine with layer rock (rock mass).
1. Introduction
With the popularization of mechanization and high-intensity mining of coal mines in western China, prevention of mine roof and floor accidents and disasters had become an important issue in coal mine safety production. Sandstones widely exist in the roof and floor of coal mines in this area, a large part of which has a bedding structure and a weak cementation, and are important to mine safety production. Compared with common rocks, bedded rocks have some special characteristics. Some sedimentary rocks (such as sandstone, limestone, and shale) and metamorphic rocks (such as granite, basalt, and granulite) have obvious bedding structure, and the rocks show manifest anisotropy. Studies have shown that [1–5] the strength of the bedding plane in these rocks is generally weak, which has a significant effect on the mechanical properties of the rock. Therefore, study of the mechanical characteristics of this type of rock has important theoretical and practical values for the safe production of coal mines.
A large number of experimental studies have been utilized for studying layered rock, of which the Brazilian disk-splitting test, shear test, and the compression test are more common. Chen et al. [6] studied the anisotropy of tensile strength of bedded rock under Brazilian disk-splitting test conditions. From the theoretical and experimental standpoints, the systematic analysis shows that the tensile strength is closely related to the angle between bedding and loading direction. Hou et al. [7] conducted the Brazilian disk-splitting tests on black shale from different bedding angles for study of the anisotropic characteristics of tensile strength and failure mode and the variation of absorption energy in deformation in the failure process. The results show that the bedding angle is closely related to the tensile strength and the absorption energy. Zhou et al. [8] conducted a Brazilian splitting experiment on layered rocks and established an equivalent continuous model of layered rocks and verified it by numerical simulations. They believed that layered rock masses exhibit significant anisotropy in both the deformability and strength. Chong et al. [9] studied the anisotropy of the stratigraphic shale by numerical simulation of Brazilian splitting experiments and proposed the AMBBM model. They found found that the ratio of cohesion to tensile strength of SJ mainly affects the number of cracks formed, which further leads to different failure modes. Liu and Zhang [10] studied the anisotropic characteristics of layered rocks through shear tests and obtained the changes of and values, stress states, and the relative orientations of layers and external loads on the direction of shear failure surfaces. Qi et al. [11] used the frozen rock structure with a bedding structure to perform shear experiments and studied its shear deformation characteristics. Heng et al. [12] analyzed the mechanical properties of the bedding plane and the anisotropic characteristics of the shear strength of the shale under the influence of the experiment and analyzed the causes of the characteristics from different angles.
In addition to the Brazilian disk-splitting test and share test, scholars have also studied the bedded rocks under compressive loading. Cheng et al. [13] performed uniaxial and triaxial compression tests on coal measure shale under dry and saturated conditions. The results show that the anisotropy and water content of the shale bedding have important influence on the strength and deformation of the coal measure shale. Guo et al. [14] performed a fatigue test on layered shales under cyclic loading and found that the rock failure process is related to bedding and can be divided into three stages, namely, initial damage stage, constant velocity damage stage, and accelerated damage stage. Hou et al. [15] conducted uniaxial compression tests on the bedded shale with bedding orientations. The results show that different bedding angles lead to different types of rock failure, and the correlation between elastic modulus and longitudinal wave velocity is good, and both decrease with the increase of bedding angle. Li et al. [16] studied the mechanical properties of layered brittle shale and its failure characteristics under hydraulic fracturing by numerical simulation. In the uniaxial compression test of shale, Liu et al. [17] used X-ray CT to scan the specimens, which illustrated the micromechanics of the failure process in the anisotropic shale. Xu et al. [18] performed a uniaxial compression experiment on limestone at 0° and 90° bedding and studied the effect of bedding angle on strength characteristics and conductivity anisotropy.
The above studies are of great significance for understanding the bedding angle and the mechanical properties of bedded rock, the mechanism of failure, and the evolution of internal energy. However, these studies are still based on the conventional description of the relationship between stress and strain and have failed to conduct an in-depth study in the mechanism of inside mesodamage of rock under the external load. Acoustic emission (AE) technology can effectively and continuously monitor generation and expansion of fine cracks in brittle materials in real time under load effects and achieve positioning of the cracks. It has become an important monitoring method for rock or rock mass deformation and failure. Wasantha et al. [19] investigated the mechanical behavior and energy releasing characteristics of bedded-sandstone under dry and water saturated conditions. Liu et al. [20] and Yang et al. [21] carried out experiments on rocks under confining pressure and studied the initiation and evolution of damage and their destabilizing behavior during loading. Zhang et al. [22] studied the AE characteristics of rock-like material under different loading rates and obtained the effect of loading rate on acoustic emission signals. Liu et al. [23] studied the initiation and propagation of cracks in the rock under pressure and made relevant assessments through the acoustic emission technique. Liu et al. [24] studied the occurrence and failure mechanism of collapse columns through the numerical simulation of acoustic emission characteristics. Dong et al. [25–27] proposed novel localization methods to improve location accuracy of AE/microseismic sources in rocks.
Based on the above studies, the present study conducted uniaxial compression tests on the bedded weakly cemented sandstones with AE monitor. Experiments were conducted to study the characteristics of AE activities and energy changes, as well as the temporal and spatial evolution of AE events and the variation of values. It is of academic value to reveal the mechanism of stratified rock in depth by means of acoustic emission.
2. Tested Rock Material and Testing Method
2.1. Sandstone Material
The rocks used for the experiment were collected from Xiaojihan coal mine, Yuyang District, Yulin City, Shaanxi Province, China, which is located in Hengyu Mine Area of northern Shaanxi. First, massive rocks without obvious defects were selected for processing. The angle of the core drilling rig was adjusted to create an included angle of 0°, 45°, and 90° between the bedding direction and axial direction so as to drill the rock core; then the massive rock was processed to a cylinder of 50 mm × 100 mm according to ISRM standards after passing through the cutting and grinding procedure. The wave speed of the specimens is tested with the ultrasonic wave velocity measuring instrument produced by Proceq Company of Switzerland. Test pieces with similar wave speed were selected to carry out tests, in order to reduce the dispersion degree of the test results. The surface of the sandstone specimens had a visible layer of bedding. The sandstone specimens used in this experiment are shown in Figure 1.
The specimens were analyzed by XRD (X-ray diffraction analysis) test. The results showed that the mineral components contained in the sample were ordered by probability: (1) quartz; (2) albite; (3) iron dolomite; (4) red uranium ore; (5) mastic chlorite. In addition there are micro-hematite, siderite, plagioclase, illite, copper chloride aluminum vanadium, and other minerals, but the content is minimal. The content of mineral components can be roughly judged. The specimens are argillaceous cemented sandstone, along with some iron cementation. Meanwhile, the mesoscopic images of the sandstone specimens were observed using a polarizing microscope. The distribution and appearance of quartz, feldspar, and cement can be clearly seen in Figure 2. A detailed composition of this rock is described as follows: Quartz grains are angular and subangular, with a content of 55%–60%. Feldspar grains are angular, with a content of 20%–25%. The content of mica and other rock fragments is approximately 10%, with a particle size of 0.2–0.5 mm. Cement exhibits mostly pelletization cementation properties and also has a small amount of carbonate cements.
(a) Magnification of 50
(b) Magnification of 20
The tests were carried out with three test pieces whose bedding angles were 0°, 45°, and 90°, respectively. See Table 1 for statistics of physical and mechanical parameters of each test piece, of which there are no obvious differences in basic physical properties such as density and wave speed. But the average uniaxial compressive strength of test pieces with different angles is 68.14 MPa, 40.14 MPa, and 59.52 MPa, respectively, under the effect of bedding angle.
2.2. Testing Equipment and Experimental Method
The uniaxial compression test used a HUALONG-600 microcomputer control compression-testing machine and the acoustic emission monitoring used a PAC’s PCI-II acoustic emission acquisition system, including PCI-2 host and Nano30 sensor and Mistras preamplifier. Through real-time acquisition of acoustic emission events, the waveform of events was recorded, combined with built-in A/D transition card, transited into a digital signal and stored in the computer. The main parameters (occurrence time, momentary magnitude, energy, maximum amplitude, etc.) were analyzed by using the AEwin Acoustic Emission Handler and were stored in the database. The indoor test system was as shown in Figure 3.
During the test, the operation of loading system and the acoustic emission monitoring system shall be kept synchronously. The uniaxial compression loading control was loaded at a rate of 0.15 mm/min using a displacement loading method until the specimen was broken. Acoustic emission monitoring used 8 Nano30 sensors to collect the signal at the same time. A rubber belt was used to evenly fix the sensor around the specimen. The sensors were 25 mm from the upper cross section and lower cross section of the specimen. In order to ensure that the acoustic emission signal could be effectively received by the sensor, butter shall be applied on the contact position of the specimen and the sensor for coupling. At the same time, the test set amplitude threshold of the acoustic emission system to 45 dB to eliminate the effect of ambient noise on the acoustic emission test. The sampling frequency of the acoustic emission system was set to 1 MHz, the main amplifier was 40 dB, and the probe resonant frequency was 20–400 kHz. Schematic diagram of AE and loading system and the transducer arrangements were as shown in Figure 4.
3. Acoustic Emission Characteristics of Specimens with Different Bedding Angles
3.1. Acoustic Emission Activity Characteristics
As a result of the limitation of space, a representative specimen shall be selected for analysis and description of each bedding angle. Respectively, they were 0004 (bedding angle of 0°), 4503 (bedding angle of 45°), and 9003 (bedding angle of 90°). Figure 5 shows relations of rings rate-strain, accumulated rings-strain, and stress-strain curves of specimens with different bedding angles. In order to clearly reflect the acoustic emission ringing rate and ringing cumulative number, the vertical axis unit in the figure was not normalized.
(a)
(b)
(c)
It could be seen from accumulated rings number-strain curve that when the bedding angle was 0°, that is, when the bedding direction was perpendicular to the axial direction, the accumulated rings number was 16.81 × 105 times; when the bedding angle was 45° and the bedding direction obliquely crossed the axial direction, the accumulated rings number was 7.02 × 105 times; and when the bedding angle was 90°, that is, the bedding direction was parallel with the axial direction, the accumulated rings number was only 1.02 × 105 times. It could be seen that with the increase of the bedding angle the overall activity level of AE was decreasing significantly, and when the bedding angle was 90°, the numerical value was only 6.1% of the numerical value when the bedding angle was 0°. The test results of the remaining specimens also showed the same trends, as shown in Table 2.
Because when the bedding angle was 0° stress concentration was likely to occur between the bedding planes, the original crack was constantly compacted and new crack occurred continuously, acoustic emission activities were frequent, and the ringing cumulative numbers were the most; when the bedding angle was 45°, the new crack was more likely to occur along the bedding plane, but under the same stress condition the activity level of AE was lower than the activity level of AE when the bedding angle was 0°, and the ringing cumulative number was low; when the bedding angle was 90°, that is, the bedding direction was parallel with the axial direction, the bed plane reduced the rock tension-compression ratio, the specimen was likely to generate crack along the bedding plane under tension, consolidation in the horizontal direction basically disappeared, AE activity was placid, and the ringing cumulative numbers were the least.
From the perspective of ringing counting rate-strain curve, during the stage of which AE event occurred, after the AE several extremely large ringing rates were removed, the average value of ringing count rates under the three kinds of bedding angle is as follows: 1176 times/s (0°), 758 times/s (45°), and 211 times/s (90°). It showed that the ringing times in unit time were more when the bedding plane was perpendicular to the axis, and the intensity of activity of AE was the highest, and when the bedding plane was parallel with the axis, the intensity of activity of AE was the lowest.
Because when the bedding angle was 0° stress concentration area occurred between horizontal bedding planes under the stress effect, resulting in the original cracks closing and the generation, development, and eventually integration of a large number of new cracks, the ringing count rate increased sharply during integration and AE ringing count rate suddenly increased; when the bedding plane was parallel to the axial direction, the vertical bedding plane was equivalent to the weak plane existing along the axial direction, and the tension effect caused by the early axial stress could easily crack it, but the rock still had a certain axial bearing capacity and would lead to damage in later period under the action of shear force, so the ringing count rate during the whole process was relatively low; when the bedding plane and axial direction were 45°, the specimen would occur with crack under compressive stress and slid for friction along the bedding plane under force action. During this period, ringing count rate slowly increased and achieved peak value while being disrupted. The activity intensity was between 0° and 90°.
3.2. Characteristics of Acoustic Emission Energy
Figure 6 shows the energy rate-strain, energy-strain, and stress-strain curves of the specimen with different bedding angles. The total energy of AE was 4.03 × 10−9 J with the bedding angle of 0°; the total energy of AE was 1.54 × 10−9 J with the bedding angle of 45°; and the total energy of AE was 2.03× 10−10 J with the bedding angle of 90°, indicating that the energy released by the densification, initiation, expansion, and penetration of the crack was the largest when the bedding plane was perpendicular to the axis during the specimen loading; with the increase of bedding angle, the energy was gradually reduced. The variation trend of energy showed the activity level of weakly cemented sandstone.
(a)
(b)
(c)
From the energy rate-strain curve of the different bedding angles in Figure 6 it can be concluded that the time of increasing of the energy rate is less when the bedding angle is 0°, which occurs at the strain of 0.012 and break, but under high strain the energy is sharply released in a short time, because the horizontal bedding specimen can withstand greater load, and more energy is required to make the cracks between the bedding surface expansion and cut-through; when the bedding angle is 45°, the energy required for sliding failure along the bedding surface due to transpression is less than 0° of the bedding angle, so the increasing energy rate before the damage is more dispersed and the amplitude is not large, and high rate of energy is concentrated in the vicinity of the destruction phase; when the bedding angle is 90°, the high rate of energy is scattered in the whole process of deformation and the amplitude is low, and the latter is more centralized than the previous period; this is because the axial direction is the same as the bedding plane, and the nonpenetrating part inside the rock can withstand low loads. The crack has begun to expand, penetrate, and release energy when the energy has not yet reached a higher value, and such a process has been repeated until the specimen is damaged, during which each propagation and coalescence of crack correspond to a high energy value.
4. Spatial Distribution of AE Locations with Different Bedding Angle
AE localization can reproduce the internal damage process during rock compression from space; it is important to study the temporal and spatial distribution of AE for characterizing the internal failure zone and damage point of rock. Figure 7 shows the spatial distribution of AE locations for three kinds of bedding angle specimens at different strain levels during uniaxial loading. The following could be seen from the number of AE locations in three different bedding angle specimens: 1646 at most at 0°, 1339 at 45°, and 560 at 90°. This phenomenon coincides with the change in the number of bells and the energy discussed in the previous article; that is, when the bedding direction is perpendicular to the axis, the AE activity is more intense. The color of the pellets in the model represents the magnitude of the AE locations, with red showing the higher magnitude and blue showing the lower one.
(a) 0°
(b) 45°
(c) 90°
(d) Magnitude chromatic scale
The AE locations in Figure 7(a) are distributed substantially along the horizontal direction and have a significant correlation with the bedding angle. Before ( represents the UCS) the locations showed a slow growth trend, and balls were mostly blue, indicating a small magnitude event; the locations at the top of the specimen suddenly increased to 691 as it was up to 0.7, and the balls were yellow-green, indicating the magnitude increased; when it reached 0.9, the number of AE locations in the lower part of the specimen increased sharply; the number of balls was 1.3 times that of 0.9 during peak stress, and the new AE locations were mostly green and red, indicating the magnitude was larger. This indicated that the AE location increased sharply, and a lot of energy was suddenly released before the specimen with the bedding angle of 0° close to be damaged. This was because the strain of the specimen was continuously concentrated along the layer during compression; firstly the cracks appeared in the middle part and orderly developed and formed AE location cluster, with the increasing strain, it began to expand toward the upper and lower part and formed AE location cluster at the upper and lower part, respectively, AE location magnitude also gradually increased, and finally the lower part of the specimen suddenly occurred in AE location with large magnitude; the crack quickly penetrated upper part, and the rock had unstable failure.
Figure 7(b) showed that the AE location was distributed along the bedding angle of 45°. AE locations firstly occurred in the middle and formed balls clusters; it began to expand toward 0.4 and formed balls clusters along the bedding plane at 0.9. The numbers of AE locations were 763 at that moment, and the number of balls increased sharply to more than 1300 during peak stress, while it could be seen that new balls included many higher magnitudes of yellow and red balls, which indicated that near the peak stress the produced AE location and energy accounted for a large proportion in the whole process, and the crack penetrated from the middle to the bottom left; finally, the rock had unstable failure. The reason is that, under the influence of bedding angle, the strain firstly concentrated on the weak surface between the bedding and the crack continually produced, and the crack was more likely to expand along the bedding plane with the increasing strain. Under the action of friction, the strain was concentrated in the lower part of the specimen, and the increase of the number of crack expansion led to the accumulation of AE locations. The damage speed of the rock was intensified and the crack penetrated through the upper right finally and the specimen was broken.
Figure 7(c) shows the distribution of AE locations when the bedding angle was 90°; compared with (a) and (b), the distribution was scattered and there was no obvious direction. During the entire loading process, there was no sudden increase in AE locations, but balls accelerated after 0.8. The number of AE locations was smaller than the previous two kinds of bedding angle, and the number of red balls in the middle and high magnitudes was less. This was because when the bedding angle was parallel to the axial direction, the specimen was more prone to tensile damage under tensile stress; due to the less energy required for tensile damage, the number of middle and high-level AE locations was lower. Although the event was cracked, the specimen still withstood pressure, so the lower part of the specimen generated the stress concentration under the action of axial stress, the crack propagated and penetrated, and the specimen was broken.
5. Evolution of Value of Specimens with Different Angles of Bedding via Acoustic Emission
In 1941, the statistical relation between earthquake magnitude and frequency was put forward by Gutenberg and Richter in the study of the seismic activity of the world. The famous G-R relationship [28, 29] is expressed in the following formula:where refers to the magnitude, refers to the acoustic emission times greater than the magnitude , and and are constants, of which value is the function of the relative magnitude distribution of acoustic emission.
In rock mechanics, value can be used as the function of crack expansion scale to represent the ratio of different amplitude acoustic emission events or the ratio of different scale cracks in the process of rock failure. According to the research of Lockner et al. and Zeng et al. [30, 31] and other scholars, the increase of value shows that the proportion of small-scale cracks increases, which are mainly microcracks; the value invariant indicates that the size-scale crack distribution is unchanged, and the state of different scale microcracks is relatively constant; the decrease of value means the increase of large-scale cracks in ratio. If value changes in a small range, it represents the fact that the state of microcracks changes slowly; and it is an asymptotic, stable expansion process; if value has a burst transition in a large extent, it indicates a burst change of microcrack state and a burst failure.
In this manuscript, value was calculated by the Matlab program by means of discrete frequency method. A number of 1000 of AE data pieces were selected as a group, and sampling calculation was conducted as per 200 event slides; by doing so, excessive calculation errors resulting from the lower number of AE data in a certain magnitude were avoided. When the change rule of value with the time is obtained and the variation regularity of value with the stress level is gained through combining the relationship between stress and time, the value-stress horizontal curve of test piece with different bedding angles is shown in Figure 8.
(a) 0°
(b) 45°
(c) 90°
It can be seen from Figure 8 that the development trend of value in AE event had obvious bedding effect, and the value-stress horizontal curve in the deformation of specimen with different angles of bedding showed different patterns. When the bedding angle was 0°, the curve presented the trend of rise first and drop next, and the increase and decline ranges were large, indicating that the value was increasing and then had a rapid decline when reaching a certain stress level; when the bedding angle was 45°, the curve changed to be mild from the load, and when the stress level increased to 60%, the curve slowed down, which indicated that value had little change in the early stage, and it decreased at the late stage; when the bedding angle was 90°, the curve was basically in horizontal fluctuation state, and there was a small range of decline after the stress level reached to 80%; that is, the change of value was kept within a stable scope. This is because when the bedding surface was in horizontal state, with the increase of stress, the original cracks in the specimen were compacted and the new initiated microcracks had a rapid growth, AE event was mainly small-scale cracks, and value rose; when the stress reached 60%, the expansion and communicating of microcrack rapidly led to a larger crack; at this time, there were mainly large-scale cracks, and value decreased, and the crack scale differentiation was obvious; when the bedding angle was 45°, microcracks along the bedding surface began to sprout, the number was increasing but the crack scale remained stable, and small-scale cracks accounted for a high ratio, so the change range of value was small; with the increase of shear force, large-scale cracks were increasing, with a high the proportion, and value decreased; when the bedding angle was 90°, the expansion of microcracks of the specimen is stable during the loading process, and the number of cracks is lower than the former two angles. The only decrease observed in the proportion of large-scale cracks was before the damage, and the value had a small range of decrease.
In addition, the value in AE event also showed bedding effect. It can be seen from the figure that, in the case of the same stress, when the bedding angle was 90°, the mean value of the value was the highest; when the bedding angle was 0°, the mean values of the value are the lowest; and when the bedding angle was 45°, the mean value of value was medium. For example, when the stress level is 40%, the values at 90, 45, and 0 degrees are 1.68, 2.04, and 2.37, respectively, while when the stress level is 80%, the values are 1.32, 1.81, and 2.34, respectively. This is because, under the same stress, when the bedding angle was 90°, the small-scale cracks had a higher proportion and the energy of AE event was low; when the bedding angle was 0°, the energy of AE event was high and the large-scale cracks had a higher proportion. This conclusion is also consistent with the previous trend of energy change in AE event with different angles of bedding.
6. Conclusion
Based on the AE test of weakly cemented sandstone samples with different bedding plane under uniaxial compression, the influence of bedding plane on the compression strength and spatial distribution of AE events, energy release, and value during rock failure was studied, and the conclusions are summarized as follows.
(1) Bedding plane has an important influence on the compression strength of the rock. With different bedding angles, the order of uniaxial compression strength of the rock specimen is as follows: 0° > 90° > 45°.
(2) Bedding plane is related to the AE activities during the loading process of rock. With the increase of the bedding angle, the AE rings rate decreases and the AE activity is less. The rock specimen with large bedding angle is more prone to rupture along the bedding plan when it is destroyed. Simultaneously, AE energy rate is higher when the bedding angle is smaller, and the energy sudden release will occur before the specimen failure.
(3) The spatial distribution rule of AE location event is closely related to the bedding angle. With the increase of bedding angle, the number of AE location events inside the specimen decreases, and the location is easier to occur along bedding plane.
(4) With the increase of the bedding angle, the expansion of the microcrack initiation in the rock specimen gradually weakens, and the proportion of the large-scale cracks increases. Therefore, the variation of value progressively decreases, and the average value increases gradually at the same stress level.
Conflicts of Interest
The authors declare no conflicts of interest.
Authors’ Contributions
Hong Wang and Tianhong Yang put forward the research ideas and designed the experiment. Hong Wang performed the experiments. Hong Wang and Yujun Zuo analyzed the data and wrote the manuscript.
Acknowledgments
Financial support was obtained from the National Key Research and Development Plan of China (Grant no. 2016YFC0801602) and Guizhou Science and Technology Plan Project (Provincial Science and Technology Cooperation Plan, Grant no. LH [2016]7465), and the National Natural Science Foundation of China (Grant nos. 51574093, 51774101).