(China Youshi University (East China) Open Laboratory of Geochemistry and Lithospheric Dynamics.

About the author: Li Yue, female, 1979, was born in Cangzhou, Hebei. She graduated from Youshi University in China in 2002 with a bachelor's degree and is a doctoral student. Her research direction is geological resources and geological engineering, and her email address is lyysy _ 79 @163.com.

On the basis of monitoring the fracture process of granite samples by MEMS technology, the application of MEMS technology in gas explosion prediction is discussed by applying the fracture monitoring principle. In the experiment, four batches of microcracks were observed by continuously pressing the rock samples with a press. Three batches of micro-fractures before the main fracture are the result of the gradual concentration and mutual penetration of cracks in rock samples, which can be regarded as precursors before the earthquake. The appearance of the main fracture produces cracks on the macro level. Based on the above principles, the application of this technology to predict mine earthquakes caused by mining and mine explosions caused by natural earthquakes will also produce good results.

MEMS technology; Fracturing; Micro-cracks; Coal mine disaster

Application of MEMS in Coal Mine Gas Disaster Prediction

, Zhou

(Open Laboratory of Geochemistry and Lithospheric Dynamics, China Shiyou University, Dongying 25706 1)

Abstract: Based on the monitoring of the fracture process of granite samples by MEMS, the application of MEMS in gas explosion prediction is discussed by applying the monitoring principle. In this experiment, four series of micro-fractures were observed by constantly applying pressure to the sample. The first three groups of micro-fractures before the main fracture are considered as precursors of earthquakes because of the concentration and penetration of cracks in the samples. The main fracture produces macro cracks. Based on the above principles, it is concluded that this technology has a good effect on predicting mine explosions caused by mining and natural earthquakes.

Key words: MEMS fracturing; Micro fracture; Coal mine disaster

foreword

MEMS (micro-electro-mechanical systems) is usually called micro-electromechanical system technology, which refers to micro-devices or systems that can be manufactured in batches and integrate micromechanics, micro-sensors, micro-actuators and signal processing and control circuits, including interfaces, communication and power supply. [ 1]

In recent years, mine accidents have accounted for a considerable proportion of major casualties, and work-induced gas explosions and earthquakes have brought great threats to workers. This paper mainly discusses the application of MEMS technology in coal mine disaster prediction on the basis of experiments.

1 experiment

The experiment mainly uses the sensitive characteristics of MEMS technology to observe the instantaneous response of the sensor when micro-fracture occurs by monitoring the fracture process of granite.

1. 1 Introduction of sampling and observation system

The samples are collected from Laizhou, Shandong Province, and belong to Yanshanian granite. Processing into 50× 15×7.5 cm3 experimental samples. Granite has a uniform grain structure, mainly composed of syenite, feldspar, biotite and a small amount of heavy minerals. The largest phenocryst of feldspar can reach about 5mm, and the average particle size is 0.5 ~ 3 mm The biotite is usually distributed linearly along the edge of feldspar particles (see figure 1).

Figure 1 Microstructure of Granite (Orthogonal Polarization× 50)

Sensors are four ME MS- 122 1 L single-component acceleration sensors produced by Dongying Gan Wei science and technology development company. Its sensitivity is 2 V/G, its resolution is 10-4G, and its frequency band range is 0 ~ 0~ 1000Hz. Data acquisition and analysis system is a general data monitoring and analysis software RBH-General developed by Dongying Gan Wei Science and Technology Development Company.

The fracturing experiment uses WE-300 press in the Mining Machine Laboratory of the School of Electromechanical Engineering of Youshi University in China (Figure 2). The observation system is shown in Figure 2(b) and Figure 3.

Fig. 2 Experimental pressure and observation system

A is WE-300 experimental press, and B is the sensor placement and pressing support position of rock sample observation system.

Fig. 3 plan view of observation system

Among them, the numbers 1, 2, 3 and 4 are four sensors, and the sensors 1 and 4 are close to the edge of the rock block. The four sensors are on a horizontal line. The center distance between sensor 1 and sensor 2 is 10cm, and the distance between sensor 3 and sensor 4 is the same. The radius of the sensor is 2.5cm.

1.2 experimental process and data discussion

1.2. 1 experimental process

Firstly, the rock sample is laid flat on the press, and the distance between the two ends of the rock sample and the support line is equal, and the distance between the two sides is recorded at the same time, so as to clarify the arm of force when the rock sample is compressed; Then put four sensors on the rock sample in turn (Figure 3) and record their respective positions. At the same time, the sensor is connected with the data acquisition and analysis system to record the signals of micro-cracks in different parts.

Time recording starts from 0 seconds, and the data sampling frequency is 4000Hz. The process of pressure application is gradual, and the pressure increases gradually from 0 kN. Observe the change of data, first record the noise spectrum. When the internal structure of the rock sample changes due to the increase of pressure, the spectrum changes immediately. The process of spectrum change will be discussed below, in which red represents the spectrum of sensor 1, black represents the spectrum of sensor 2, blue represents the spectrum of sensor 3 and yellow represents the spectrum of sensor 4. In the fracturing process of nearly 360 seconds, the fracturing of real rock samples was completed in the last minute, that is, 302.290~303.826 s; They are; 305.599 ~ 307. 135s； During 3 16.793 ~ 3 18.329s and 357.923~360.258s, four batches of micro-cracks appeared in rock samples. Except for the last batch of microcracks lasting more than 2s, the last three batches of microcracks last less than1.5s. Each batch of microcracks consists of a group of dense microcracks, and the duration of a single microcrack is generally less than 50ms..

1.2.2 fracturing process data recording and analysis

In the following, the spectral characteristics recorded in 10 representative time periods are selected in chronological order for discussion. Due to technical reasons, the accuracy of the sensor currently used is not enough to distinguish the exact time of receiving the signal when the rupture occurs. We will solve this problem step by step in our future work.

Noise spectrum of (1) 0.291~ 31.826s after compression (Figure 4): Shortly after compression, although the noise received by each sensor is different, the main frequency of the noise is concentrated in the low frequency range of 50 ~ 300 Hz and the high frequency range of 400 ~ 750 Hz. Because the No.4 sensor is far away from the oil pump, and its frequency is distributed in the lower and higher areas, 20 ~ 20~200Hz and 600 ~ 750 Hz. The noise difference recorded by different sensors is mainly related to their different positions.

(2) Noise spectrum of 31.990 ~ 33.526s (Figure 5): Compared with the compressed noise spectrum of 0.29 1 ~ 3 1.826 s, the amplitude of noise is nearly doubled, but the frequency is still concentrated in the low frequency region, and the amplitude of high frequency is somewhat suppressed relative to the low frequency region.

Fig. 4 noise spectrum after 0.291~ 31.826s compression.

Fig. 5 3 1.990 ~ 33.526s noise spectrum

(3) Noise spectrum from 300.665 to 302.20 1 s (Figure 6): Before micro-rupture, the noise level further decreased, especially the positions of No.2,1and No.4 sensors decreased obviously. The noise level at position 3 is relatively high.

Figure 6 300.665 ~ 302.438+0s noise spectrum

(4) The frequency spectrum of micro-cracks in 302.290 ~ 303.826 s (Figure 7): This is the frequency spectrum feature of the first micro-crack in the rock sample. It can be clearly seen that the amplitude is abnormal, and the data obtained by different sensors are different: 1, the frequency range of No.2 sensor is about 700 ~ 800 Hz, and No.3 and No.4 sensors, especially No.3 sensor, are greatly affected by noise and have no obvious response to micro-fracture. The frequency range of No.3 sensor is about 500 ~ 600 Hz, and the frequency range of No.4 sensor is about 650 ~ 750 Hz. The first batch of micro-cracks only changed the fine structure of rock samples, and there was no macro change.

Fig. 7 spectrum of microcracks from 302.290 s to 303.826 s.

(5) The frequency spectrum of micro-cracks in 305.599 ~ 307.135s (Figure 8): Compared with the frequency spectrum of micro-cracks in 302.290~303.826s, it obviously moves to the low frequency direction, and the frequency range is about 650 ~ 750 Hz.

Fig. 8 305.599 ~ 307. 135s microcrack spectrum.

(6) Noise spectrum of 307.612 ~ 309.147s (Figure 9): The rock sample that continues to be pressurized after micro-fracture will not fracture again temporarily, which is basically the same as the noise spectrum at the beginning, but the high-frequency noise is relatively higher than the low-frequency noise, indicating that the internal structure of the rock sample has changed.

Fig. 9 Noise spectrum of 307.612 ~ 309.147s

(7) Spectrum of 316.793 ~ 318.329s microcracks (Figure 10): Compared with the first two batches of microcracks, the third batch of microcracks has greater strength and enhanced amplitude. With the increase of pressure, on the basis of the previous fracture, when the internal cracks of the rock sample develop and penetrate again, the rock sample will fracture. The spectrum characteristics of different sensors are quite different, and their frequency ranges are also different. Among them, the frequency range of micro-fracture recorded by 1 sensor is about 350 ~ 500 Hz, that of No.2 sensor is about 450 ~ 550 Hz, that of No.3 sensor is about 400 ~ 500 Hz and that of No.4 sensor is about 650 ~ 750 Hz.

Fig.10316.793 ~ 318.329s spectrum of microcracks.

(8) 8) Noise spectrum of 326.534 ~ 328.070s (Figure 1 1): After the third batch of micro-fracturing, the rock sample has already cracked, and the continuous pressurization will not have a great impact on the rock sample in a short time, so it still shows the spectral characteristics of pressure noise.

Figure11326.534 ~ 328.070s noise spectrum

(9) Frequency spectrum of main fracture at 358.723 ~ 360.258 s (Figure 12): After continuous pressurization, the rock sample is more strongly fractured on the basis of the previous micro-fracture, that is, the main fracture. From the data we collected, the amplitude of this rupture is much larger than that of the previous rupture, and the peak value has an obvious trend of moving to the low frequency region. The frequency range of each sensor is obviously different: 1 sensor has a frequency range of 300 ~ 500~700Hz, sensor No.2 has a frequency range of 200 ~ 300 ~ 500Hz, sensor No.3 has a frequency range of 300 ~ 500Hz and sensor No.4 has a frequency range of 500 ~ 700Hz. Because the final fracture surface is located between No.2 and No.3 sensors, and the final fracture extends to No.2 sensor, the amplitude and frequency of micro-fracture recorded by No.2 and No.3 sensors are relatively low, especially No.2 sensor. However, the amplitude and frequency of microseisms recorded by 1 and 4 sensors which are relatively far away from the fracture surface are much higher. This may be related to the smaller the rock sample, the farther the sensor is from the fracture surface, and the greater the displacement.

Fig.12 main rupture spectrum of 358.723 ~ 360.258s.

(10) 361.335 ~ 362.871s noise spectrum after the main fracture (figure 13): the pressure exerted after the main fracture has no effect on the rock sample, because the rock sample has been completely fractured, and at this time we can clearly see that there is a crack in the appearance of the rock sample. Keep pressuring. However, compared with the noise spectrum at the beginning of pressure application, because the rock sample has cracked, the noise of the oil pump is transmitted to the sensor through the rock sample, and the crack has an impact on the transmission of noise, resulting in a greatly weakened high-frequency noise and a relatively enhanced low-frequency noise.

Fig. 1336438+0.335 ~ 3238+0s noise spectrum after main rupture.

1.2.3 Variation characteristics of micro-fracture spectrum

By analyzing the experimental process of rock samples pressurized by a press, it can be seen that the frequency range and amplitude of the spectrum are different when four batches of microcracks are produced (see table 1).

Table 1 Frequency range and spectrum peak value received by different sensors when four batches of microcracks occur.

When four batches of rupture occur, the frequency range is not only concentrated in the range listed in table 1, but also in a relatively concentrated area. However, due to the low peak frequency or narrow range in other areas, they are not listed one by one, and only the main frequency range is listed in the table. As can be seen from the data in the table, for a sensor, with the increase of pressure, the frequency range of four ruptures decreases in turn, that is, with the increase of rupture times, the frequency decreases gradually; For the same micro-crack, the data obtained by the two sensors near the pressure application point in the first two batches are smaller than those obtained by the sensors far away from the pressure application point, while the frequency range of only the No.4 sensor is obviously larger than the other three sensors when the main crack occurs, indicating that the closer to the crack, the lower the frequency value. From this phenomenon, the following laws can be summarized: with the increase of pressure, the frequency value decreases; The larger the crack, the smaller the frequency value. Moreover, due to the small volume of the rock sample itself, the position is not accurate enough when it is placed. A slight difference will lead to a slight inclination of the rock sample during compression, which will lead to a great difference between the data of 1 and No.4, No.2 and No.3 geophones in symmetrical position. From the spectral peak of each rupture, the spectral peak emitted by the sensor near the pressure action point is larger in the first two ruptures, but the situation is just the opposite in the last two ruptures. This may be due to the small scale of micro-fracture in the first two fractures, but the internal structure has not changed much. The scale of micro-fracture in the last two fractures has increased relatively, and the fourth batch of micro-fracture even caused macro-fracture of rock samples.

Discussion on the experimental results of 1.3

In recent years, seismologists have realized that earthquakes are a kind of fracture behavior of earth materials with cracks, and explored the gestation and occurrence of such fractures in the process of studying the formation of microcracks in general solid materials including rocks. At present, all the basic hypotheses about earthquake preparation regard the evolution of the earth rupture as the key to finding and solving earthquake precursors and solving earthquake prediction [2- 10]. The main fracture is caused by the fracture of the rock sample under the conditions of front and constant pressure, which makes the internal cracks accumulate and increase, and finally achieves the result of mutual penetration, and the rock sample produces a crack approximately parallel to the pressure direction on the macro level. The following is a detailed discussion on the data record of a main micro-crack selected from four batches of micro-cracks:

(1) microseismic records produced by the first batch of major faults (Figure 14): The signals sent by the four sensors are reflected in the figure respectively. The first batch of micro-cracks occurred when the compressive strength of granite samples first reached the limit, and enough cracks accumulated inside, and first penetrated in the direction of principal compressive stress, thus cracking occurred.

Figure 14 microseismic records of main cracks in the first batch of microcracks.

(2) microseismic records of the second batch of micro-cracks when the main fracture occurs (figure 15): the second batch of micro-cracks developed on the basis of the first batch of micro-cracks, and the frequency of fracture is mainly concentrated in the low frequency region. And the frequencies of red and yellow spectra are higher than those of black and blue spectra, which shows that the frequency of sensors placed near cracks is low. That is, the closer to the sound source, the lower the frequency.

Figure 15 microseismic records produced when the main rupture occurred in the second batch of micro-ruptures.

(3) The microseismic records produced by the main fracture in the third batch of micro-fractures (Figure 16): The third batch of micro-fractures is due to the continuous development of internal cracks in rock samples under continuous pressure, the strength is much stronger than that of the second batch, and the frequency range tends to shift to the low frequency region, which can be regarded as an important micro-fracture before the earthquake.

Figure 16 microseismic records produced by the main fracture in the third batch of micro-fractures.

(4) The microseismic records produced by the main cracks in the fourth batch of micro-cracks (Figure 17): The fourth batch of micro-cracks is the main crack and the final crack of the rock sample under compression. This fracture is due to the continuous development and high concentration of internal cracks in rock samples with the increase of pressure (the final pressure reaches 10.4 kN), which leads to macro cracks and stress concentration in rock samples. If this is applied to earthquake prediction, the occurrence of cracks at this time can be defined as the occurrence of earthquakes. The frequency spectrum obtained by sensors close to the source is low.

Figure 17 microseismic records produced by the main fracture in the fourth batch of micro-fractures.

Most rocks have cracks such as joints and cleavage, and some still have large weak structures such as cracks. When the pressure is increased to a certain extent, these cracks will rupture in a concentrated way. The fracture model of granite can be summarized as avalanche unstable fracture formation model, also known as the model of Institute of Geophysics of the Soviet Academy of Sciences. The model is based on two phenomena: the interaction of fracture stress field and the local concentration of crack formation. Under the long-term effect of slowly changing load, any material, including rock, will inevitably produce these two phenomena before failure. The long-term strength theory is based on the gradual development of the number and size of cracks under the slow action of "subcritical" (less than the instantaneous strength of materials) stress. When the fracture density reaches the critical density state value, the material will transition to the rapid macroscopic fracture stage. If the distribution of cracks in the medium is uniform from a statistical point of view, the number and size of cracks will gradually increase under the action of slowly increasing load or under the influence of active medium, and some more favorable cracks will connect with each other to form larger cracks. If Griffith's theory and some theories derived from it are applied to earthquake sources, it is considered that a small number of long cracks are gradually produced during the formation of avalanche cracks, and these long cracks collude and converge with each other, resulting in macro-fracture of rocks (earthquake) [1 1].

Application of 2 in coal mine gas disaster prediction

The earthquake induced by coal mining (called rockburst in mining industry) is one of the dynamic geological disasters induced by mining. Under the influence of mining activities and regional stress field, mine earthquake makes the stress in and around the mining area unstable, and part of the energy accumulated locally in the mining area is released through impact or gravity, resulting in rock mass vibration. According to incomplete statistics, since 1980s, the level of mine earthquakes in Northeast China, such as Beipiao in Liaoning, Liaoyuan in Jilin, Hegang in Heilongjiang, Hanniaoxi in Shuangyashan and Qitaihe, has gradually increased, and the losses caused by some mine earthquakes are quite serious. It has attracted the attention of earthquake, coal system and researchers at all levels. The occurrence of mine earthquake is closely related to the tectonic environment and regional tectonic stress field besides mining factors [12].

Coal mining makes the distribution of underground stress change with the increase of mining depth. Under the influence of regional tectonic activities, tectonic stress makes the old and new structures inherit and regenerate in different degrees. Some underground fault structures gradually move or creep from a stable state, which is the internal dynamic environment of mine earthquake [13].

Earthquake is the deformation of underground rock mass under the action of stress, which causes rock mass fracture, relative displacement, sliding, fault and seismic wave radiation. The location of mine earthquake is the vibration of underground rock mass in mining area, and the seismic records are similar to natural seismic records in many places. The focal depth of mine earthquake is shallow, which can be approximated as the random fluctuation of surface focal point in a large range.

Under the action of regional tectonic forces, coalbed methane will be produced and accumulated in some specific directions. When the produced coalbed methane overflows and accumulates in the local area of the mine, if the local temperature of the mine reaches the ignition point of coalbed methane, it may cause an explosion. Coal mine gas explosion and seismic activity are synchronous in time [14- 15]. Therefore, accurate prediction of seismic activity plays an important role in preventing gas explosion in coal mines.

Based on the conclusions of the above experiments and the relationship between seismic activity and coal mine gas explosion, MEMS 122 1 L single-component acceleration sensor can be used to predict cracks caused by mine earthquakes and mining and natural earthquakes. Thereby reducing the disasters caused by gas explosion in coal mines.

We put the sensors in different positions in the coal mine, and at the same time connect the sensors to the computer observation and analysis system to record the signals sent by the sensors at different times. According to our above experimental process, in the process of continuous mining, the machine will exert great force on the ore body. When the rock structure inside the ore body changes, the sensor will change obviously, and we will see that the recorded frequency spectrum signal will change suddenly. After two or three such mutations, the ore body is very likely to collapse. Therefore, in the first mutation, we should strengthen prevention and take corresponding measures to prevent the occurrence of fractures.

Similarly, when an earthquake occurs underground, we can also take precautions according to this principle. Most seismologists believe that there is a process of stress concentration in the earthquake source area before the earthquake, which is called earthquake preparation process or earthquake preparation process. When this process develops to a certain stage, the rocks in the earthquake-prone area may appear micro-fracture or plasticity, which will lead to the change of seismic wave spectrum. In addition, the change of focal dynamic parameters of small earthquakes in the earthquake-producing area may also cause some changes in seismic wave spectrum. These are the physical basis of prediction research based on seismic wave spectrum anomalies. Before the main rupture, a series of seismic waves with small amplitude and low frequency often appear, which can be regarded as precursor seismic waves. In this experiment, the seismic waves produced by three micro-ruptures before the main rupture can be regarded as precursor seismic waves. The occurrence of these seismic waves is the energy accumulation of the main seismic waves. When the energy accumulates to a certain extent, earthquakes will inevitably occur.

3 Conclusion

Under uniaxial pressure, (1) granite produces four relatively concentrated brittle fractures, and the strength of these four fractures tends to increase with the increase of pressure. When micro-fracture occurs, the frequency tends to shift to the low frequency region, and the larger the crack, the lower the frequency.

(2) Three batches of micro-fractures before the main fracture are the result of the gradual concentration and interconnection of internal cracks in rock samples, which can be regarded as precursors before the earthquake. The occurrence of main rupture can be regarded as the occurrence of earthquake because it produces cracks on the macro level;

(3) The near-source observation records of fracturing experiments show that MEMS technology has high sensitivity in monitoring cracks, so it will achieve good results in coal mine disaster prediction, thus reducing mine disasters caused by mining and natural earthquakes.

Acknowledgement: Thanks to the technical support provided by Dongying Gan Wei Science and Technology Development Co., Ltd. and the press equipment provided by the laboratory of the School of Mechanical and Electrical Engineering of China Youshi University (East China). In the process of completing the thesis, I got the help from my other disciples, and I would like to express my gratitude.

refer to

John Crabte .. 1968. Synthesizing layered media from acoustic wave transmission response: Geophysics, 33, 264~269.

Dr. Danes Waal, Dr. Clay and Dr. Savage .. 1995. Passive seismic imaging using microseisms, Geophysics, 60,1178 ~186.

[3]M.Reza Daneshvar, Passive seismic imaging with microseisms, Geophysics, 60(4)

De Yan Draganov, 2004, Passive Seismic Imaging in the Presence of White Noise Sources, Frontier, September.

Zhang Shan, Liu Qinglin, Zhao Qun, etc. 2002. Application of microseismic monitoring technology in oilfield development, petroleum geophysical exploration, 4 1(2), 226 ~ 23 1.

[6] Andy Juppe waiting, translated by Tian Zengfu. 1999. microseismic monitoring: listen to oil reservoir, petroleum geophysical translation series, 5, 17 ~ 20.

Liu Jianzhong, Wang Chunyun and others, 2004. Monitoring oilfield production performance by microseismic method, oil exploration and development, 3 1(2), 7 1 ~ 73.

[8] Andy Juppe, translated by Li. 1999. microseismic monitoring reservoir, natural gas exploration and development, 44 ~ 48.

[9] Zhu Pei A, Cowles J, Jones R.. 1998. Microseismic monitoring: Listen to oil reservoir, world oil, 219 (12):17/kloc-0.

[10] Dong Shitai, Gao Hongxia.2004. Microseismic monitoring technology and its application in oilfield development, Petroleum Instruments, 18(5), 5 ~ 8.

[1 1] Feng Deyi, Chen Huaran, Ding Weiguo. 1994. Study on the abnormal characteristics of seismic wave spectrum before large earthquakes, earthquake research, 17(4), 3 19 ~ 329.

[12] Zhang Fengming, Yu Zhongyuan, Xu Xiaoyan and others, 2005. Study on earthquake induced by mining in Hegang coal mine, Journal of Natural Disasters, 14( 1), 139 ~ 143.

Zheng, Wang Yong, 2004. Discussion on seismic activity factors in coal mine gas disasters, prevention and control of geological disasters in China, 15(4), 54 ~ 59.

Yang. 1996. formation and disaster of ground fissures in wangjiashan coal mine, Gansu geology, 5(2), 9 1 ~ 95.

Zhang, Zhang Huaxing, Yue. 2003. observation and analysis of cracks in coal seam mining, geotechnical mechanics, 24 (supplement), 4 14 ~ 4 17.