Hard and brittle rock local instability identification method based on distributed optical fiber technology

Abstract

The application provides a hard and brittle rock local instability identification method based on distributed optical fiber technology, and relates to the technical field of rock instability identification. First, a rock sample is connected with a distributed optical fiber data monitoring system, data collection is carried out, the collected data is subjected to noise reduction treatment and data correction is carried out according to the accuracy requirement, then a statistical analysis method or a coordinate-based deformation localization analysis method is selected to post-process the data, rock local instability is identified through the change of corresponding indexes, and the continuously and rapidly increased degree of rock deformation localization indicates the local instability of the rock.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rock instability identification, and particularly relates to a hard and brittle rock local instability identification method based on distributed optical fiber technology. BACKGROUND

[0002] In the construction process of China's deep major projects, hard and brittle rocks such as granite are often encountered, such as granite tunnels in the Sichuan-Tibet railway, and the energy in the hard and brittle rock may suddenly release to cause rock burst and other engineering disasters during construction. In order to ensure the safe construction of the project and the safety of the lives and property of the construction personnel, accurate prediction of rock local instability failure is extremely important. Traditional displacement sensors are mainly single-point sensors, and the prediction of rock local instability through single-point displacement monitoring mainly depends on artificial experience. When the rock deformation monitored by the displacement sensor significantly increases, the rock may fail. However, the instability and failure of hard and brittle rock is a process of elastic energy accumulation and sudden release. Due to the high elastic modulus of hard rock, the small deformation of the rock during the accumulation of elastic energy cannot provide sufficient early warning information. At the same time, the local instability deformation of the rock may occur at any position due to the heterogeneity of the rock, and it is difficult for the conventional displacement sensor to monitor the large-scale deformation. In addition, the failure of hard and brittle rock is sudden, and once the deformation or stress of the rock reaches a certain critical level, the rock will suddenly release energy and fail, which requires the sensor to have high-frequency dynamic monitoring capability. However, the intermittent monitoring mode in the traditional method cannot realize long-term and high-frequency dynamic displacement monitoring of the rock. As a novel monitoring technology, distributed optical fiber can realize contact strain measurement by being attached to the measured object, and has the advantages of high monitoring accuracy and high frequency. By reasonably arranging the optical fiber, the deformation of the rock in a large range can be monitored. The acoustic emission technology, digital image processing technology, computer tomography technology and polarized light microscope have proved that the deformation and failure of hard and brittle rock are mainly controlled by internal crack propagation, and there is a clear process of crack initiation, propagation and aggregation into macroscopic cracks before the peak strength of the rock, accompanied by obvious deformation localization. Therefore, with the rapid development of full-field deformation monitoring technology, capturing and quantitatively analyzing the deformation localization characteristics of the rock has become an effective method for predicting rock local instability. SUMMARY

[0003] In view of the deficiencies of the prior art, the present application provides a hard and brittle rock local instability identification method based on distributed optical fiber technology. The distributed optical fiber technology is applied to the local instability identification of hard and brittle rock, and the full-field deformation data collected by the distributed optical fiber is post-processed to solve the problem that it is difficult to predict the instability of hard and brittle rock through deformation. The present method realizes the prediction of the instability and failure of the test piece through the small deformation data of the rock surface collected by the optical fiber.

[0004] A hard and brittle rock local instability identification method based on distributed optical fiber technology, comprising the following steps:

[0005] Step 1: coupling the rock sample with the distributed optical fiber;

[0006] Specifically, the rock sample is processed into a cylindrical shape, wherein the cross-sectional radius of the cylindrical rock sample is greater than the minimum curvature radius of the optical fiber;

[0007] The distributed optical fiber is wound on the surface of the rock sample, and the winding mode of the distributed optical fiber includes horizontal winding and spiral winding, the spiral wound distributed optical fiber forms a spiral optical fiber, and a set of pre-stress is maintained during winding, and the distributed optical fiber is pasted on the rock surface using quick-drying glue as the coupling agent;

[0008] The inclination angle of the spiral optical fiber tends to 0, and the maximum spiral optical fiber inclination angle is less than Where v is the Poisson's ratio.

[0009] Step 2: connecting the rock sample with the distributed optical fiber data monitoring system to collect data;

[0010] The distributed optical fiber monitoring system comprises an optical fiber, an optical fiber data demodulator and a computer host;

[0011] Step 2.1: place the rock sample in the mechanical loading device, and draw one end of the distributed optical fiber from the mechanical loading device as an inlet end to input optical signals, and the other end of the distributed optical fiber as an output end of the optical signals, and knot the output end of the distributed optical fiber;

[0012] Step 2.2: fuse the inlet end with the optical fiber of the optical fiber data demodulator, and detect the distributed optical fiber using a laser pen or the optical fiber data demodulator;

[0013] Step 2.3: set the data parameters of the distributed optical fiber monitoring system on the computer host, including the strain transfer coefficient and temperature transfer coefficient of the distributed optical fiber, the spatial resolution and time resolution of data acquisition;

[0014] Step 2.4: detect the light loss in the distributed optical fiber again, and calibrate the position by pressing the distributed optical fiber to determine the position starting point and ending point of data acquisition;

[0015] Step 2.5: pre-load the rock mechanically, and after pre-loading, zero the axial force, displacement data in the mechanical loading device and the surface strain data collected by the distributed optical fiber;

[0016] Step 2.6: load the rock sample with the wound distributed optical fiber with compression force, and start collecting axial stress, axial strain and optical fiber strain data at the same time.

[0017] Step 3: Correction of fiber strain data;

[0018] The noise contained in the fiber strain data includes values exceeding the strain range and sudden change values in the continuous change process of the data; the fiber strain data after denoising meets the strain continuous change and has no sudden change value;

[0019] According to the inclination of the spiral fiber, it is judged whether the fiber strain data obtained by the spiral fiber needs to be corrected, and the judgment formula is:

[0020]

[0021] Wherein, ε f is the strain data obtained by the spiral fiber, θ is the fiber inclination, v is the Poisson's ratio, and ε c is the hoop strain value of the sample; when θ is less than 5°, the theoretical error between the fiber strain data and the true hoop strain value is less than 1%, and no correction is needed;

[0022] If the direct error between the measured strain value caused by the inclination of the spiral fiber and the true hoop strain value is greater than the set error, the corrected hoop strain distribution is calculated using the hoop strain correction formula, and the hoop strain correction formula is:

[0023]

[0024] Wherein, θ1, θ2 are the two inclinations of the spiral fiber, and ε1, ε2 are the corresponding fiber strain values;

[0025] Step 4: Fiber data deformation localization analysis method based on statistical distribution, extract rock local instability failure precursor characteristics;

[0026] The hoop strain values obtained by the distributed fiber are arranged in time sequence, one time point is a group, and the concentration degree of rock deformation at each time point is calculated according to the calculation formula of discrete Gini coefficient; when the strain value in a certain area in the monitoring range increases rapidly, the Gini coefficient increases, indicating that large deformation occurs in the area, that is, deformation localization, which is a precursor of local failure;

[0027] The rapid increase of the strain is determined by the following method: the strain data collected at two time points are standardized according to the following formula, and the standardized strain result of the next time point of the same monitoring point is subtracted from the standardized strain result of the previous time point; the monitoring range of the monitoring point with positive subtraction is the area with increased strain value, that is, ε t,i -ε t-1,i >0, ε t,i -ε t-1,i The greater, the faster the strain increases;

[0028]

[0029] wherein ε i is the strain value of the i-th monitoring point, and σ is the average value and standard deviation of the set of data, ε t,i is the normalized result of the i-th monitoring point at time point t;

[0030] wherein the calculation formula of the Gini coefficient is as follows:

[0031]

[0032] wherein G is the Gini coefficient, W i and P i is the proportion of the strain of the i-th monitoring point in the total strain value of the whole field, and W i is the proportion of the area monitored by the i-th monitoring point in the total monitoring area; i Considering that the distributed optical fiber is a linear sensor, the monitoring length of each monitoring point is the same, so P i = 1 / N, N is the number of monitoring points in the monitoring range;

[0033] The value range of the Gini coefficient is (0, 1), when the strain is evenly distributed on each monitoring point, G = 0, the more concentrated the high strain value is, the higher the degree of deformation localization is, and the larger the Gini coefficient is;

[0034] For hard and brittle rocks, the abnormal high strain value caused by surface cracks is captured to assist in judging the precursor of local instability of the rock, wherein the calculation formula of the skewness coefficient is as follows:

[0035]

[0036] S is the skewness coefficient, ε i is the strain value of each monitoring point, is the average value of a set of strain data, and σ is the standard deviation of the set of data;

[0037] In the process of predicting local damage of hard and brittle rocks, the joint determination method of the Gini coefficient and the skewness coefficient can be used to determine the precursor of rock instability and damage when the Gini coefficient and the skewness coefficient start to continuously increase at the same time;

[0038] Step 5: Fiber data deformation localization analysis method based on monitoring point coordinates, extract rock local instability and damage precursor characteristics;

[0039] When the monitoring data corresponds to the monitoring point coordinates one by one, the coordinate-related characterization method can also be used for deformation localization quantitative analysis:

[0040] Firstly, the degree of deformation localization is calculated by the global Moran index. When strain values of similar size are aggregated in space, the spatial correlation of monitoring points with similar strain values is enhanced, and the Moran index increases. Conversely, when the spatial correlation of monitoring points with similar strain values is weakened, the Moran index decreases.

[0041] The specific calculation formula of the Moran index is as follows:

[0042]

[0043] I is the Moran index, N is the number of monitoring points in the considered spatial range, X i and is the strain value of the i th monitoring point and the average strain value of the N monitoring points, w ij is the weight matrix.

[0044] The value range of the Moran index is (-1, 1). When the monitored high strain value monitoring points are aggregated in space, the Moran index is positive and continuously increases, showing the strain localization characteristics.

[0045] When the hard and brittle rock is close to instability and failure, the deformation near the crack significantly increases and presents the deformation localization characteristics. When the Moran index continuously increases, it proves that the rock starts to have obvious localized deformation.

[0046] The beneficial effects produced by the above technical solutions are as follows:

[0047] The present application provides a hard and brittle rock local instability identification method based on distributed optical fiber technology, which has the following beneficial effects:

[0048] 1. A deformation localization quantification method based on statistics is provided, which solves the problem that the traditional method must rely on coordinates to evaluate the degree of deformation localization.

[0049] 2. A method for quantitatively predicting hard and brittle rock local instability according to deformation localization is provided. Compared with the traditional method which can only judge rock damage and predict failure according to axial, hoop and volumetric stress-strain curves, the method proposed in the present application fully considers the heterogeneous deformation in the rock compression and failure process, and is suitable for data obtained by distributed optical fiber and other full-field deformation monitoring means. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 is the overall flowchart of the hard and brittle rock local instability identification method in the embodiment of the present application;

[0051] Figure 2 is the distributed optical fiber data monitoring system and optical fiber layout path in the triaxial high pressure test process in the embodiment of the present application;

[0052] Figure 3 Gini coefficient and skewness coefficient analysis results in the process of grading creep of granite under 60MPa confining pressure in the embodiment of the present application;

[0053] Figure 4 Probability distribution curve of normalized creep strain of all monitoring points under different stress levels in the embodiment of the present application;

[0054] Figure 5 Moran index analysis results in the process of grading creep of granite under 60MPa confining pressure in the embodiment of the present application;

[0055] Figure 6 Strain distribution visualization results corresponding to different stresses in the embodiment of the present application;

[0056] Wherein (a) is the strain distribution on the surface of the rock at the early stage of loading at 26% of the peak stress; (b) is the strain distribution on the surface of the rock at the early stage of loading at 60% of the peak stress; (c) is the strain distribution on the surface of the rock at the late stage of loading at 98% of the peak stress. DETAILED DESCRIPTION

[0057] The specific embodiments of the present application will be further described in detail below in combination with the drawings and examples. The following examples are used to illustrate the present application, but not to limit the scope of the present application.

[0058] A hard and brittle rock local instability identification method based on distributed optical fiber technology, as shown in FIG. 1, comprises the following steps: Figure 1

[0059] Step 1: coupling the rock sample with the distributed optical fiber;

[0060] Coupling rock and distributed optical fiber monitoring system:

[0061] Specifically, the rock sample is processed into a standard cylinder, wherein the cross-sectional radius of the cylindrical rock sample is greater than the minimum curvature radius of the optical fiber, so as to ensure that the optical fiber can normally transmit optical signals;

[0062] The distributed optical fiber is densely wound on the surface of the rock sample to obtain as much rock surface deformation data as possible. The winding mode of the distributed optical fiber includes horizontal winding and spiral winding. The spiral wound distributed optical fiber forms a spiral optical fiber. A predetermined prestress is maintained during the winding process to ensure that the distributed optical fiber is in full contact with the rock surface to be measured. A high modulus quick-drying adhesive is used as a coupling agent to paste the distributed optical fiber on the rock surface. It is ensured that the optical fiber is not damaged during the pasting process;

[0063] The inclination angle of the spiral optical fiber tends to 0, so that the optical fiber remains in tension. The maximum inclination angle of the spiral optical fiber should be less than Wherein v is the Poisson's ratio. ​

[0064] Step 2: connect the rock sample with the distributed optical fiber data monitoring system for data acquisition;

[0065] The distributed optical fiber monitoring system comprises an optical fiber data demodulator and a computer host; in the embodiment, as shown in the figure, the distributed optical fiber is G652b single-mode optical fiber, and the surface coating layer of the distributed optical fiber should be as thin as possible to reduce strain hysteresis and obtain more accurate strain data; Figure 2

[0066] Step 2.1: place the rock sample in the mechanical loading device, and draw one end of the distributed optical fiber from the mechanical loading device as an inlet end to input an optical signal, and the other end of the distributed optical fiber as an output end of the optical signal, and knot the distributed optical fiber at the output end to reduce the influence of optical loss at the optical fiber break on the test result;

[0067] Step 2.2: fuse the inlet end with the optical fiber of the optical fiber data demodulator, ensure that the optical loss is less than 0.01 dB, and use a laser pen or an optical fiber data demodulator to detect the distributed optical fiber to ensure that the optical fiber can normally transmit signals;

[0068] Step 2.3: set the data parameters of the distributed optical fiber monitoring system on the computer host, including the strain transfer coefficient and the temperature transfer coefficient of the distributed optical fiber, the spatial resolution and the time resolution of data acquisition; the equipment selected in the application is based on optical frequency domain reflectometry (OFDR), and the sampling accuracy of data is 1με;

[0069] Step 2.4: detect the optical loss in the distributed optical fiber again to ensure that the distributed optical fiber can normally transmit strain signals, and calibrate the position by pressing the distributed optical fiber to determine the position starting point and the position ending point of data acquisition;

[0070] Step 2.5: pre-load the rock mechanically, and after the pre-loading is completed, zero the axial force, displacement data in the mechanical loading device and the surface strain data collected by the distributed optical fiber;

[0071] Step 2.6: perform mechanical loading, apply compression force to the rock sample with the wrapped distributed optical fiber, and simultaneously start collecting axial stress, axial strain and optical fiber strain data.

[0072] Step 3: optical fiber strain data correction;

[0073] ​The noise contained in the fiber strain data includes values exceeding the strain range and sudden values in the continuous change process of the data. The fiber strain data after noise removal satisfies the continuous change of strain and has no sudden values. The data must be completely removed because excessive or insufficient noise data will seriously affect the results of subsequent rock deformation localization analysis.

[0074] According to the inclination of the spiral optical fiber, it is determined whether the optical fiber strain data obtained by the spiral optical fiber needs to be corrected to obtain more accurate hoop strain. The judgment formula is:

[0075]

[0076] wherein, ε f is the strain data obtained by the spiral optical fiber, θ is the inclination of the optical fiber, v is the Poisson's ratio, and ε c is the hoop strain value of the sample. When θ is less than 5°, the theoretical error between the optical fiber strain data and the true hoop strain value is less than 1%, and no correction is needed.

[0077] If the direct error between the measured strain value caused by the inclination of the spiral optical fiber and the true hoop strain value is greater than the set error, the corrected hoop strain distribution is calculated using the hoop strain correction formula:

[0078]

[0079] wherein, θ1 and θ2 are the two inclinations of the spiral optical fiber, and ε1 and ε2 are the corresponding optical fiber strain values. For the non-uniform deformation field of rock, the average strain value of the adjacent region is preferably used when the strain is corrected by the above formula.

[0080] Step 4: Fiber data deformation localization analysis method based on statistical distribution, extracting rock local instability failure precursor characteristics;

[0081] The hoop strain values obtained by the distributed optical fiber are arranged in time sequence, one time point is a group, and the concentration degree of rock deformation at each time point is calculated according to the calculation formula of the discrete Gini coefficient. The Gini coefficient reflects the distribution characteristics of the data. When the strain value in a certain region within the monitoring range rapidly increases, the Gini coefficient increases, indicating that large deformation occurs in the region, i.e. deformation localization, which is a precursor of local failure.

[0082] The rapid increase is determined by the following method. The strain data collected at two time points are standardized according to the following formula. The standardized strain result at the latter time point is subtracted from the standardized strain result at the former time point at the same monitoring point. The monitoring range of the positive subtraction is the region where the strain value increases, i.e. ε t,i - ε t-1,i > 0, and ε t,i-ε t-1,i The greater, the faster the strain increases;

[0083]

[0084] Wherein, ε i is the strain value of the i th monitoring point, and σ is the average value and standard deviation of the data set, ε t,i is the normalized result of the i th monitoring point at time point t;

[0085] Wherein, the calculation formula of Gini coefficient is as follows:

[0086]

[0087] Wherein, G is the Gini coefficient, W i and P i is to sort all the optical fiber strain data from low to high, calculate the proportion of the strain of the i th monitoring point to the total strain value W i and the proportion of the area monitored by the i th monitoring point to the total monitoring area P i , considering that the distributed optical fiber is a linear sensor, the monitoring length of each monitoring point is the same, so P i =1 / N, N is the number of monitoring points in the monitoring range;

[0088] The value range of Gini coefficient is (0, 1), when the strain is evenly distributed on each monitoring point, G=0, the more concentrated the high strain value is, the higher the degree of deformation localization is, and the greater the Gini coefficient is;

[0089] It should be noted that the Gini coefficient measures the relative gap of strain distribution, if the strain data of each monitoring point is doubled, although the absolute value of the deformation of the relatively weak position of the rock mass increases much more than that of the relatively hard position of the rock mass, because the proportion of each position increases is the same, the relative gap is still the same, and the calculated Gini coefficient is also the same, only when an abnormal large deformation occurs at a certain position, the Gini coefficient will continuously increase, as shown in the embodiment; Figure 3 ;

[0090] For hard and brittle rock, the progressive failure of rock is caused by the initiation, expansion and penetration of cracks, and the abnormal high strain value caused by surface cracks can be used to assist in judging the precursor of local instability of rock, and the abnormal high strain value will cause serious uneven distribution of data, and the uneven strain data distribution caused by abnormal high strain value can be identified by skewness coefficient; wherein, the calculation formula of skewness coefficient is as follows:

[0091]

[0092] S is the skewness coefficient, εi the strain value of each monitoring point, the average value of a group of strain data, and σ is the standard deviation of a group of data;

[0093] The skewness coefficient measures the degree of skew distribution of data. When the data presents a symmetric normal distribution, the skewness coefficient is close to 0. When a small number of strain values are very large, the probability distribution function curve of the data is dragged very long on the right side, that is, it presents a positively skewed distribution. The longer the right tail, the thicker the skewness coefficient, and the larger the skewness coefficient. In the embodiment, as shown in the formula (3), the skewness coefficient calculation result can be positive or negative. When a positively skewed distribution is presented, the skewness coefficient is positive; Figure 4

[0094] When a crack appears at the monitoring position of the optical fiber, the abnormally high strain value makes the data probability density curve present a clear positively skewed distribution. The positively skewed distribution becomes more and more obvious with the expansion of the crack, and the skewness coefficient continuously increases;

[0095] In the local damage prediction process of hard and brittle rocks, the joint determination method of Gini coefficient and skewness coefficient is adopted to avoid the error caused by a single calculation method, and the rock instability and damage precursor is determined when the Gini coefficient and the skewness coefficient start to continuously increase at the same time;

[0096] Step 5: Based on the optical fiber data deformation localization analysis method of the monitoring point coordinates, the rock local instability and damage precursor characteristics are extracted;

[0097] The above method of using Gini coefficient and skewness coefficient to represent strain localization is applicable to complex situations where the monitoring point coordinates cannot be determined. When the monitoring data and the monitoring point coordinates are one-to-one corresponding, the deformation localization quantification analysis can also be performed by using the coordinate-related representation method:

[0098] Firstly, the spatial deformation localization degree is calculated by the global Moran index. When strain values of similar sizes are aggregated in space, the spatial correlation of monitoring points with similar strain values is enhanced, and the Moran index increases. Conversely, when the spatial correlation of monitoring points with similar strain values is weakened, the Moran index decreases;

[0099] The specific calculation formula of the Moran index is as follows:

[0100]

[0101] I is the Moran index, N is the number of monitoring points in the considered spatial range, X i and is the strain value of the i-th monitoring point and the average strain value of the N monitoring points, w ij ​is a weight matrix, and the form of the weight matrix is selected as needed, such as the weight being inversely proportional to the Euclidean distance from the strain center point, or a binary weight matrix is taken, that is, the weight value is 1 within a certain threshold range from the center point, and the weight value is 0 outside the range, and the distance threshold range can be adjusted according to different field conditions;

[0102] The value range of the Moran index is (-1, 1), when the monitored high strain value monitoring points are spatially aggregated, the Moran index is positive and continuously increases, and the strain localization feature is shown, and in the embodiment, as shown in FIG. 5, the global Moran index simultaneously considers the distribution of high and low strain value monitoring points, and the aggregation of low strain value monitoring points also increases the Moran index. Figure 5

[0103] When the hard and brittle rock is close to instability and failure, the deformation near the crack significantly increases and presents the deformation localization feature, and the Moran index starts to continuously increase, which proves that the rock starts to have obvious localized deformation; it should be noted that the Moran index is simultaneously affected by the distribution state of the low strain value, and when the low strain value is widely and discretely distributed and the high strain value is highly concentrated, the Moran index may slightly decrease;

[0104] It is worth noting that the analysis method provided above is applicable to strain localization analysis of different rocks, but only hard and brittle rocks with strong correlation between deformation and cracks will present obvious strain localization phenomenon before failure due to the generation and expansion of macro cracks, and other types of rocks may have different deformation localization features before failure, such as marble, which presents obvious strain delocalization feature before instability and failure, and the analysis method can also be applied to other rocks for failure precursor identification after detailed research, and in the embodiment, as shown in FIG. 6, the strain distribution visualization results corresponding to different stresses are shown, wherein (a) is the strain distribution on the rock surface at the early loading stage of 26% peak stress, (b) is the strain distribution on the rock surface at the cracking stress (60% peak stress) in the middle loading stage, and (c) is the strain distribution on the rock surface at the late loading stage of 98% peak stress. Figure 6

[0105] The above description is only the preferred embodiments of the present disclosure and the explanation of the applied technical principles. Those skilled in the art should understand that the scope of the application involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combinations of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or equivalent features without departing from the above inventive concept. For example, the above features are replaced with the technical features disclosed in the embodiments of the present disclosure (but not limited to) having similar functions to form technical solutions.​​

Claims

1. A method for identifying local instability in hard and brittle rocks based on distributed optical fiber technology, characterized in that, Includes the following steps: Step 1: Couple the rock sample to a distributed optical fiber; Step 2: Connect the rock sample to the distributed fiber optic data monitoring system to acquire data; The distributed optical fiber monitoring system includes optical fiber, optical fiber data demodulator, and computer host. Step 3: Correction of fiber optic strain data; The fiber optic strain data is denoised to remove noise from the data, including values ​​exceeding the strain range and abrupt changes during continuous data variation. The denoised fiber optic strain data then satisfies the condition of continuous strain variation without abrupt changes. Step 4: Extract precursor features of local instability and failure of rocks based on the fiber optic data deformation localization analysis method according to statistical distribution; Step 5: Extract the precursor features of local rock instability and failure based on fiber optic data deformation localization analysis method using monitoring point coordinates; Step 1 specifically involves processing the rock sample into a cylindrical shape, wherein the cross-sectional radius of the cylindrical rock sample is greater than the minimum curvature radius of the optical fiber. Distributed optical fibers are wound around the surface of a rock sample. The winding methods for the distributed optical fibers include horizontal winding and helical winding. The helically wound distributed optical fibers form a helical optical fiber. During the winding process, a set prestress is maintained, and quick-drying adhesive is used as a coupling agent to attach the distributed optical fibers to the rock surface. The tilt angle of the spiral fiber tends to 0, and the maximum tilt angle of the spiral fiber is less than 0. Where v is Poisson's ratio; Step 2 specifically includes the following steps: Step 2.1: Place the rock sample into the mechanical loading device, and lead one end of the distributed optical fiber out of the mechanical loading device as the inlet to input the optical signal, and the other end of the distributed optical fiber as the output of the optical signal. Tie the distributed optical fiber at the output end into a knot. Step 2.2: Splice the fiber at the input end to the fiber optic data demodulator, and use a laser pointer or fiber optic data demodulator to detect the distributed fiber optic network. Step 2.3: Set the data parameters of the distributed optical fiber monitoring system on the computer host, specifically including the strain transfer coefficient and temperature transfer coefficient of the distributed optical fiber, the spatial resolution of data acquisition, and the temporal resolution; Step 2.4: Detect the optical loss in the distributed optical fiber again, and determine the start and end points of data acquisition by pressing the distributed optical fiber to establish its position. Step 2.5: Apply mechanical preloading to the rock, and after the preloading is completed, reset the axial force, displacement data and surface strain data acquired by the distributed optical fiber in the mechanical loading device to zero. Step 2.6: Apply mechanical loading by applying compressive force to the rock sample wound with distributed optical fibers, and simultaneously start collecting axial stress, axial strain, and optical fiber strain data; Step 3 specifically involves: The determination of whether the fiber strain data obtained from the spiral fiber needs correction is based on the tilt angle of the spiral fiber. The formula is as follows: ； in, The strain data obtained from the helical optical fiber. The fiber tilt angle, Poisson's ratio, The circumferential strain value of the specimen, when When the angle is less than 5°, the theoretical error between the fiber strain data and the actual circumferential strain value is less than 1%, and no correction is required. If the direct error between the measured strain value and the actual circumferential strain value caused by the tilt angle of the helical fiber is greater than the set error, then the circumferential strain correction formula is used to calculate the corrected circumferential strain distribution. The circumferential strain correction formula is as follows: ； in, , These are the two tilt angles of the helical optical fiber. , This corresponds to the fiber strain value; Step 4 specifically involves: arranging the circumferential strain values ​​acquired by the distributed optical fiber in a time sequence, with each time point as a group; calculating the concentration of rock deformation at each time point according to the formula for calculating the discrete Gini coefficient; when the strain value in a certain area within the monitoring range increases rapidly, the Gini coefficient increases, indicating that a large deformation has occurred in that area, i.e., deformation localization, which is a precursor to local failure. The rapid increase in strain is determined as follows: strain data collected at two time points are standardized using the following formula; the standardized strain result at the second time point is subtracted from the standardized strain result at the first time point for the same monitoring point; the monitoring range of the monitoring point with a positive subtraction is the region of increased strain value. , The larger the value, the faster the strain is considered to increase. ； in, Let be the strain value at the i-th monitoring point. .and Here are the mean and standard deviation of this set of data. This represents the standardized result of the i-th monitoring point at time t. The formula for calculating the Gini coefficient is as follows: ； Where G is the Gini coefficient. and It sorts all fiber optic strain data from low to high and calculates the proportion of the strain at the i-th monitoring point to the total strain value of the entire field. The ratio of the area monitored by the i-th monitoring point to the total monitored area. Considering that the distributed optical fiber is a linear sensor and the monitoring length of each monitoring point is the same, therefore... N is the number of monitoring points within the monitoring range; The Gini coefficient ranges from (0,1). When the strain is evenly distributed at each monitoring point, G=0. The more concentrated the high strain values ​​are, the higher the degree of deformation localization, and the larger the Gini coefficient. For hard and brittle rocks, abnormally high strain values ​​caused by surface cracks are used to help determine the precursors of local rock instability. The formula for calculating the skewness coefficient is as follows: ； S is the skewness coefficient, representing the strain value at each monitoring point. The average value of a set of strain data. The standard deviation of a set of data; In the process of predicting local failure of hard and brittle rocks, the Gini coefficient and skewness coefficient are used to determine the simultaneous and continuous increase of the Gini coefficient and skewness coefficient as a precursor to rock instability and failure. Step 5 specifically involves: when the monitoring data corresponds one-to-one with the coordinates of the monitoring points, using a coordinate-related characterization method to perform deformation localization and quantitative analysis. First, the degree of spatial deformation localization is calculated using the global Moran index. When strain values ​​of similar magnitudes cluster in space, the spatial correlation of monitoring points with similar strain values ​​is enhanced, and the Moran index increases; conversely, when the spatial correlation of monitoring points with similar strain values ​​is weakened, the Moran index decreases. The specific formula for calculating the Moran index is as follows: ； I represents the Moran index, and N represents the number of monitoring points within the considered spatial range. and Let be the strain value at the i-th monitoring point and the average strain value at the N monitoring points. This is the weight matrix; The Moran index ranges from (-1, 1). When the high strain monitoring points are clustered in space, the Moran index is positive and continuously increases, showing strain localization characteristics. When hard and brittle rocks are close to instability and failure, the deformation near the crack increases significantly, showing the characteristics of localized deformation. When the Moran index begins to increase continuously, it proves that the rock has begun to have obvious localized deformation.

Images