A rock slope failure precursor identification method based on ae and dic multifractal characteristics

Abstract

The present application relates to the technical field of geological engineering, and more particularly to a rock slope failure precursor identification method based on AE and DIC multifractal characteristics. The steps are as follows: S1: collecting signals during the loading process of the rock slope or sample, and constructing the collected signals into discrete time series in chronological order; S2: performing multifractal analysis on the AE and DIC signal time series; S3: extracting the multifractal characteristic parameters of the spectral width Δα and the spectral width difference Δα0 and determining whether the rock mass is damaged. The rock slope failure precursor identification method based on AE and DIC multifractal characteristics provided by the present application combines acoustic emission (AE) and digital image correlation technology (DIC), utilizes the multifractal characteristics of both, and realizes a method for objectively identifying rock slope failure precursors, so as to improve the accuracy and reliability of rock slope instability early warning.

Description

Technical Field

[0001] This invention relates to the field of geological engineering technology, and in particular to a method for identifying precursors of rock slope failure based on AE and DIC multifractal characteristics. Background Technology

[0002] Rock slope stability has always been a key issue in the field of geological engineering, especially in mining, transportation, and urban construction projects, where the stability of rock slopes directly affects the safety of the project and the safety of people and property. Existing methods for monitoring rock slope instability mostly rely on single signals, such as acoustic emission or displacement monitoring, but these methods have the following shortcomings: A single signal is susceptible to noise interference; it is difficult to simultaneously account for internal damage and external deformation. The identification of warning signs of damage is highly subjective.

[0003] Furthermore, the rock mass failure process exhibits significant nonlinear and multi-scale characteristics, making it difficult for traditional statistical methods to effectively capture its complex evolutionary patterns.

[0004] Existing technologies for rock mass monitoring utilize either acoustic emission monitoring or digital image correlation (DIC) techniques, each with its own advantages and disadvantages. Acoustic emission technology: It can monitor the generation and propagation of microcracks inside rock masses, but it is easily affected by noise and is difficult to accurately reflect external macroscopic strain.

[0005] Digital image correlation (DIC) technology can accurately capture the evolution of external strain fields, but it cannot effectively reflect the microscopic evolution of internal damage.

[0006] To this end, a method for identifying precursors of rock slope failure based on the multifractal features of acoustic emission (AE) and digital image-correlated strain signals is designed. This method utilizes the multifractal features of acoustic emission signals and digital image-correlated strain signals to identify precursors of rock slope failure, thereby improving the accuracy and reliability of rock slope instability early warning. Summary of the Invention

[0007] Therefore, it is necessary to provide a method for identifying precursors of rock slope failure based on the multifractal characteristics of AE and DIC, in order to solve the technical problems mentioned above.

[0008] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for identifying precursors of rock slope failure based on AE and DIC multifractal features, comprising the following steps: S1: During the loading process of the rock slope or the sample, the signals are collected and the collected signals are constructed into a discrete time series in chronological order; S2: Perform multifractal analysis on the time series; S3: Extract the spectral width of the feature parameters of the multifractal. and spectral width difference And determine whether the rock mass is damaged.

[0009] As a preferred embodiment of the rock slope failure precursor identification method based on AE and DIC multifractal features provided by the present invention, the signals collected in step S1 include: Acoustic emission ringing count or energy signal; Surface strain signals obtained based on digital image correlation technology.

[0010] As a preferred embodiment of the rock slope failure precursor identification method based on AE and DIC multifractal features provided by the present invention, in step S3, multifractal analysis is performed on the time series, and the steps are as follows: The multifractal spectrum of internal and external signals is calculated using box counting. The time series of the collected information is taken as the dataset {Ti}, and the time series is divided into N consecutive subsets, each of length n. The normalized probability measure of each subset is given by the following expression: ; in, P i To normalize the probability density, S i For the first i Cumulative sum over a period of time; By using power-law weighting, the multifractal dimension can be divided into multiple regions to different degrees, and after Legendre transformation, the fractal spectral function f(q) and the average singular intensity α(q) can be calculated.

[0011] As a preferred embodiment of the rock slope failure precursor identification method based on AE and DIC multifractal features provided by the present invention, the expression of the fractal spectrum function f(q) is as follows: ; The expression for the average singularity intensity α(q) is as follows: ; In the formula: u i Let q be the partition function, q be the statistical moment order, and n be the scale length. P i (n) is the normalized probability measure.

[0012] As a preferred embodiment of the rock slope failure precursor identification method based on AE and DIC multifractal features provided by the present invention, a generalized fractal dimension D is introduced. q To determine whether a signal has multifractal characteristics, the expression is as follows: ; If D q The signal decreases significantly with increasing q and exhibits a nonlinear relationship, indicating that the signal has multifractal characteristics.

[0013] As a preferred embodiment of the rock slope failure precursor identification method based on AE and DIC multifractal features provided by the present invention, in step S3, when acoustic emission... and When both numerical values ​​simultaneously show a sudden increase (≥0.2) and a sudden decrease (≤-0.1), the rock mass is determined to have entered a critical state of failure; when DIC... and When both numerical values ​​simultaneously show a sudden increase (≥0.02) and a sudden decrease (≤-0.005) in abnormal signals, the rock mass is determined to have entered a critical state of failure. When acoustic emission and DIC... and After all the above abnormal signals appear in numerical values, it is determined that the specimen is about to fail (the thresholds for sudden increases and decreases in the above numerical values ​​are derived from...). Figures 2-4 The experimental curves can be calibrated through preliminary experiments based on different lithologies or samples.

[0014] It is clear without a doubt that the technical solution described above in this application can solve the technical problem that this application aims to address.

[0015] Meanwhile, through the above technical solutions, the present invention has at least the following beneficial effects: This invention provides a method for identifying precursors of rock slope failure based on the multifractal features of acoustic emission (AE) and digital image correlation (DIC). By combining acoustic emission (AE) and digital image correlation (DIC) techniques and utilizing their multifractal features, this method achieves objective identification of precursors of rock slope failure, thereby improving the accuracy and reliability of rock slope instability early warning.

[0016] This invention achieves synergistic characterization of internal damage and external deformation; utilizes multifractal methods to depict the nonlinear failure process of rock mass; makes precursor identification objective and quantifiable; and is applicable to different lithologies and different fracture conditions; thus improving the reliability of rock slope instability monitoring.

[0017] This invention combines acoustic emission (AE) technology and digital image correlation (DIC) technology to capture the characteristics of different stages of rock mass failure through multifractal analysis. Experimental results show that the changes in the multifractal characteristics of AE and DIC signals before rock mass failure exhibit strong regularity, and the method can accurately identify the initial stage of rock mass failure and the critical instability state. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the generalized fractal spectrum of the present invention; Figure 2 This is a schematic diagram of the time-varying evolution curves of the multifractal parameters during the entire loading process of the samples 00-1 and 10-2 of this invention. Figure 3 This is a schematic diagram of the time-varying evolution curves of the multifractal parameters during the entire loading process of the samples 20-3 and 30-2 of this invention. Figure 4 This is a schematic diagram of the time-varying evolution curves of the multifractal parameters during the entire loading process of the samples 40-2 and 50-3 of this invention. Figure 5 This is a schematic diagram of the three-segment locking landslide generalization model of the present invention; Figure 6 This is a schematic diagram showing the mineral composition and geometric dimensions of the sample of the present invention; Figure 7 This is a schematic diagram of the test system program of the present invention; Figure 8 This is a schematic diagram of the stress-strain curves of specimens with different fracture segment lengths under uniaxial loading according to the present invention. Figure 9 This is a schematic diagram showing the ringing count rate, cumulative time, maximum principal strain on the surface, and crack propagation evolution process of the samples 00-1 and 10-2 of this invention. Figure 10 This diagram illustrates the ringing count rate, cumulative time, maximum principal strain, and crack propagation evolution process of samples 20-3 and 30-2 in this invention. Figure 11 This is a schematic diagram showing the ringing count rate, cumulative time, maximum principal strain on the surface, and crack propagation evolution process of the 40-2 and 50-3 samples of this invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0022] It should be noted that, unless otherwise specified, the embodiments and features and technical solutions in the present invention can be combined with each other.

[0023] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0024] Example 1 Reference Figures 1-4 A method for identifying precursors of rock slope failure based on AE and DIC multifractal features, the steps of which are as follows: Step 1: Signal Acquisition Simultaneous data collection during rock slope or sample loading process: Acoustic emission ringing count or energy signal; Surface strain signals obtained based on digital image correlation technology.

[0025] Step 2: Time Series Construction The collected signals are constructed into a discrete time series in chronological order.

[0026] Step 3: Multifractal Analysis This invention utilizes box counting to calculate the multifractal spectrum of internal and external signals. The time series of collected information is used as a dataset {Ti}, which is then divided into N consecutive subsets of length n. The normalized probability measure for each subset is: (1) In the formula: P i To normalize the probability density, S i For the first i Cumulative sum over a given interval.

[0027] The partition function is: (2) In the formula: τ ( q () is a quality index. q For statistical moment order; n The scale length. By continuously changing... n The value is calculated repeatedly from ln( u i ( q , n )) and ln( n The quality index is obtained by taking the slope of a double logarithmic function on the fitted curve. τ ( q ),when τ( q )along with q When changes occur, time series exhibit multifractal characteristics.

[0028] Through power-law weighting, the multifractal dimension can be divided into multiple regions to varying degrees. After Legendre transformation, the calculation process of the fractal spectral function f(q) and the average singular intensity α(q) is as follows: (3) (4) In the formula: u i Let q be the partition function, q be the statistical moment order, and n be the scale length. P i (n) is the normalized probability measure.

[0029] At the same time, the generalized fractal dimension is introduced. D q To determine whether a signal has multifractal characteristics ，D q It can be calculated using relation (5): (5) If D q The value decreases significantly with increasing q and exhibits a non-linear relationship, indicating that the signal possesses multifractal characteristics. Figure 1 To obtain the generalized fractal spectrum of internal AE signal and external DIC strain under different trailing edge tensile cracks.

[0030] analyze Figure 1 (a) It can be seen that the AE signal D q The -q curve exhibits significant nonlinear characteristics, D q The signal decreases rapidly with increasing q, indicating that the AE signal exhibits strong multifractal properties. Analysis Figure 1 (b) It can be seen that the DIC signal D q The nonlinear characteristics of the -q curve are relatively weaker than those of the AE signal, D q The changes are gradual. The multifractal nature of the AE signal is more pronounced because it directly reflects the non-uniform evolution of internal damage; the DIC signal, on the other hand, reflects the overall distribution characteristics of surface strain more and has lower complexity. This indicates that the evolution of the external strain field is controlled by the macroscopic mechanical response of the sample, and its statistical characteristics have high self-similarity on a spatial scale, which may be related to the brittle failure mode of locked sandstone.

[0031] Spectrum width The calculation formula is as follows: (6) In the formula, It is the maximum value of the singularity index. It represents the minimum singularity index. The magnitude of the value represents the degree of unevenness of the probability subset.

[0032] Spectral width characterizes the uniformity of the distribution of physical parameters across the entire fractal structure. The larger the value, the greater the signal fluctuation, the more obvious the multifractal, and the higher the degree of non-uniformity, and vice versa.

[0033] The formula for calculating the subset of the spectral measure is as follows: (7) In the formula, q( ) refers to when α= The value of q at time , similarly, q( ) refers to when α= The q value at that time.

[0034] Δf This reflects the probability of the occurrence of large and small measure subsets. Δf The larger the value, the more stable the fracture is. Δ f The smaller the value, the greater the proportion of signals at the large rupture scale, and vice versa.

[0035] The formula for calculating the spectral bandwidth difference is as follows: (8) In the formula, The difference in spectral width; α p Peak value f max (α) corresponds to the value of α.

[0036] Spectral morphology is generally classified into: left-skewed distribution, symmetrical distribution, and right-skewed distribution. The left-skewed distribution indicates that larger fracture events are dominant, while the right-skewed distribution indicates that smaller crack events are dominant.

[0037] Step 4: Feature Parameter Extraction Extract the following parameters from the multifractal spectrum: Spectrum width ; Spectrum width difference .

[0038] Step 5: Identification of Early Signs When sound is emitted and When abrupt decreases or increases in numerical values ​​occur, it indicates that the rock mass has entered a critical state of failure. Therefore, when acoustic emission... and When both numerical values ​​simultaneously show a sudden increase (≥0.2) and a sudden decrease (≤-0.1), the rock mass is determined to have entered a critical state of failure; when DIC... and When both numerical values ​​simultaneously show a sudden increase (≥0.02) and a sudden decrease (≤-0.005) in abnormal signals, the rock mass is determined to have entered a critical state of failure. When acoustic emission and DIC... and After all the above abnormal signals appear in numerical values, it is determined that the specimen is about to fail (the thresholds for sudden increases and decreases in the above numerical values ​​are derived from...). Figures 2-4 The experimental curves can be calibrated through preliminary experiments based on different lithologies or samples.

[0039] Example 2 Reference Figures 5-11 Based on the above embodiment one, its specific application is disclosed.

[0040] Slopes are classified into two main categories: locked and unlocked. Locked slopes, a typical type of rock slope, are primarily controlled by the mechanical properties of the locked sections, resulting in a typical three-stage failure mode: creep-tensile cracking-shear fracture. The failure process initially manifests as tensile cracking at the top and creep failure at the bottom, ultimately leading to shear fracture in the middle rock bridge or locked area. Therefore, exploring the damage evolution, fracture failure mechanism, and pre-fracture characteristics of locked rocks is of significant engineering importance.

[0041] 1. Experimental Design 1.1 Sample Preparation The three-stage locked landslide, characterized by "creep-tensile cracking-shearing," is a typical failure mode in rock slopes. Its simplified conceptual model is as follows: Figure 5 As shown, the root cause of this type of landslide is the failure of the locking segment. Therefore, it is generalized as a locking slope sample to further explore the instability and failure mechanism and crack evolution law of this type of rock slope.

[0042] Given the relatively good homogeneity and ease of processing of sandstone, particularly blue sandstone, a common geotechnical material in geotechnical engineering, which exhibits strong sudden brittle failure, this experiment used relatively homogeneous blue sandstone to construct a three-segment locked-in test model to represent a brittle rock slope. The main component content and geometric dimensions are as follows: Figure 6 As shown in (a) and (b).

[0043] Depend on Figure 6(a) shows the mineral composition of the sandstone. The sandstone contains a high content of quartz and albite, approximately 40% and 35% respectively. These two minerals have high hardness and strength. It can be preliminarily inferred that the sandstone sample may have high compressive strength and wear resistance. These inferences can provide a reference for subsequent mechanical property experiments.

[0044] To minimize variability among samples, all test samples were taken from the same large sandstone block with few natural fissures, unaffected by weathering and erosion, and maintaining a high degree of integrity. The sandstone samples were processed into slab-shaped specimens with dimensions of length (60 mm) × width (30 mm) × height (120 mm), and samples with similar mass and density were selected to meet the specific requirements of subsequent tests. To facilitate variable control and obtain reliable data, high-precision machine tool rock cutting technology was used, referring to… Figure 6 (b) Model dimensional diagram was used to prepare six different trailing edge tensile crack lengths of locked specimens, with three specimens in each group. The trailing edge tensile crack lengths of the specimens were precisely machined in the vertical direction to be 0, 10, 20, 30, 40 and 50 mm, and the numbering method was as follows: the first specimen in the group with a trailing edge tensile crack length of 10 mm was numbered 10-1. The basic physical parameters of the specimens are shown in Table 1.

[0045] Table 1: Basic physical parameters of the sample

[0046] 1.2 Test System Figure 7 This is a schematic diagram of the test system. The test used a SAW-2000 microcomputer-controlled electro-hydraulic servo rock pressure testing machine manufactured by Changchun Xinke Instrument Co., Ltd., with a maximum load of 1000KN and a pressure accuracy of ±0.5%. Figure 7 As shown in (a1) to (a4); a DIC acquisition and software analysis system manufactured by Thousand Eyes Wolf Co., Ltd. was used, equipped with a CCD camera with a resolution of 3376 pixels × 2704 pixels and a high-brightness ultraviolet fill light. Figure 7 As shown in (c1) to (c4). Acoustic emission signal monitoring uses a DS5-8B multi-channel monitoring system manufactured by Beijing Ruandao Co., Ltd., with a sampling frequency of 100kHz, a sampling length (number of points) of 2048 (approximately 0.2ms), a waveform threshold of 40dB, and a preamplifier gain of 40dB. Figure 7 As shown in (b1) to (b4).

[0047] 1.3 Experimental Procedure (1) Before the test, the rock sample was weighed, numbered and photographed. In order not to affect the DIC observation, the AE sensor was placed on the side and back of the sample, and an appropriate amount of coupling agent was applied to the contact surface. The sample was placed in the center of the loading platform.

[0048] (2) Turn on all instruments and equipment. The press uses displacement control loading method and the rate is set to 0.03 mm / min. The DIC monitoring system is set to soft trigger and acquisition rate of 3 sheets / s.

[0049] (3) The DIC, acoustic emission and uniaxial compression systems start synchronously. After the sample becomes unstable and fails, all equipment stops at the same time, the data is saved and exported.

[0050] 2. Study on mechanical properties and fracture mechanism of locked specimens 2.1 Mechanical Property Analysis Statistical analysis of the experimental data showed that the mechanical parameters and acoustic emission signal response characteristics of the sandstone samples under the same working conditions were quite similar. Due to space limitations, one sample from each group was selected as an example for analysis and discussion. Figure 8 Table 2 shows the variation of uniaxial compressive strength of specimens with different trailing edge tensile fracture lengths.

[0051] analyze Figure 8 It can be seen that the compressive strength of the intact specimen is significantly higher than that of the specimen containing the locking section, indicating that the rock bridge in the locking section has a significant impact on the mechanical properties of the rock mass. Subsequently, as the length of the tensile fracture section increases, the strength of the specimen shows a trend of "..." The peak strength of the specimen is positively correlated with the length of the fractured section when the diameter is 10-30 mm; negatively correlated when the diameter is 30-50 mm; and the peak strength of the locked specimen is the highest when the diameter is 30 mm.

[0052] Table 2: Physical and mechanical parameters of specimens with different trailing edge fracture lengths

[0053] As shown in Table 2, the average peak uniaxial compressive strength of the samples under the six working conditions (00, 10, 20, 30, 40, and 50 mm) were 82.81, 37.02, 37.86, 40.63, 37.13, and 28.46 MPa, respectively. The uniaxial compressive strength of each group of samples differed by less than 3 MPa, which indicates that each group of samples has good uniformity.

[0054] Figure 2 middle, Figure 2 (a) is a workpiece with sample number 00-1. It is a schematic diagram showing the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone when the working condition is 00. "-1" is the sample corresponding to the most suitable parameter selected according to Table 2 when the working condition is 00.

[0055] Figure 2(b) is a workpiece with a sample size of 10-2. It is a schematic diagram showing the specific changes in the mechanical parameters, acoustic emission signal and DIC optical signal of the sandstone under working condition 10. "-2" is the sample corresponding to the most suitable parameters selected according to Table 2 under working condition 10.

[0056] Figure 3 middle, Figure 3 (a) is a workpiece with sample 20-3, which is a schematic diagram of the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone under working condition 20. "-3" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 20.

[0057] Figure 3 (b) is a workpiece with sample 30-2, which is a schematic diagram of the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone under working condition 30. "-2" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 30.

[0058] Figure 4 middle, Figure 4 (a) is a workpiece with sample 40-2, which is a schematic diagram of the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone under working condition 40. "-2" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 40.

[0059] Figure 4 (b) is a workpiece with a sample of 50-3. It is a schematic diagram showing the specific changes in the mechanical parameters, acoustic emission signal and DIC optical signal of the sandstone under working condition 50. "-3" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 50.

[0060] Figure 9 middle, Figure 9 (a) is a workpiece with sample number 00-1. It is a schematic diagram showing the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone when the working condition is 00. "-1" is the sample corresponding to the most suitable parameter selected according to Table 2 when the working condition is 00.

[0061] Figure 9 (b) is a workpiece with a sample size of 10-2. It is a schematic diagram showing the specific changes in the mechanical parameters, acoustic emission signal and DIC optical signal of the sandstone under working condition 10. "-2" is the sample corresponding to the most suitable parameters selected according to Table 2 under working condition 10.

[0062] Figure 10 middle, Figure 10(a) is a workpiece with sample 20-3, which is a schematic diagram of the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone under working condition 20. "-3" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 20.

[0063] Figure 10 (b) is a workpiece with sample 30-2, which is a schematic diagram of the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone under working condition 30. "-2" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 30.

[0064] Figure 11 middle, Figure 11 (a) is a workpiece with sample 40-2, which is a schematic diagram of the specific changes in mechanical parameters, acoustic emission signal and DIC optical signal of blue sandstone under working condition 40. "-2" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 40.

[0065] Figure 11 (b) is a workpiece with a sample of 50-3. It is a schematic diagram showing the specific changes in the mechanical parameters, acoustic emission signal and DIC optical signal of the sandstone under working condition 50. "-3" is the sample corresponding to the most suitable parameter selected according to Table 2 under working condition 50.

[0066] Implementation steps: 1) Acquire microcrack propagation data inside the rock mass using acoustic emission sensors.

[0067] 2) Use the DIC system to monitor the strain field on the rock surface in real time.

[0068] 3) Perform multifractal analysis on the acquired AE and DIC signals and calculate their fractal dimension.

[0069] 4) The multifractal parameters showed significant abrupt changes before destruction, successfully enabling the identification of precursors to destruction.

[0070] Results: Compared with single acoustic emission or DIC methods for monitoring precursory information of sample failure during experiments, the joint analysis of multifractal characteristics of AE and DIC signals can accurately predict precursory signals of rock mass failure.

[0071] 3. Characteristics of internal and external damage evolution during crack propagation The transient elastic wave phenomenon generated by the rapid energy release during loading of rock samples is commonly referred to as AE (acoustic emission), which can be used to identify micro-cracks and damage within rock materials. The strain evolution of crack propagation on the rock sample surface and the changes in the cumulative acoustic emission count curve exhibit a certain synergistic effect. Therefore, this invention combines multiple perspectives, including stress-strain curves during axial loading of locked-type samples, acoustic emission internal damage monitoring, and DIC (disseminated intradermal signaling) external strain field monitoring, to analyze the internal and external damage evolution characteristics during crack propagation in locked-type samples. The results are as follows: Figures 9-11 As shown.

[0072] analyze Figures 9-11 It can be known that: (1) Combining the cumulative number of AE events of the sample, the evolution characteristics of the maximum principal strain curve on the surface and the strain displacement cloud map, it was found that the sample has progressive failure characteristics. The sample fracture process can be divided into the initial damage stage (I), the damage stabilization accumulation stage (II), and the unstable failure stage (III).

[0073] (2) Initial damage stage (I): The maximum principal strain curve increases slowly, and the surface strain field cloud map is evenly distributed, which more intuitively reflects the compaction state of the original cracks in the sample. The AE signal is mainly generated by the contact, friction, and slip behavior of the crack surface. It is worth mentioning that in this stage, compared with the intact sample, the strain field cloud map of the locked sample gradually and slowly migrates from the global average distribution to the tip of the creep segment.

[0074] (3) Damage stabilization accumulation stage (II): DIC shows a "butterfly-shaped" localized high strain accumulation area, that is, the specimen with a trailing edge tensile crack length of 10-30 mm is located at the tip of the creep section, and the specimen with a length of 40-50 mm is located at the creep section and the tip of the tensile crack. During this stage, the AE signal shows a slow and steady increase due to the high strain energy accumulation effect at the crack tip.

[0075] (4) Instability failure stage (III), DIC captures the macroscopic crack propagation path. For specimens with a trailing edge tensile crack length of 10-30 mm, the crack extends upward from the tip of the creep section. For specimens with a trailing edge tensile crack length of 40-50 mm, the crack extends from the creep section and the tip of the tensile crack towards each other, and the locked section is connected. During this stage, the AE signal and the maximum principal strain increase rapidly. The analysis is that the small-scale cracks extend and connect during this stage to form a failure band, which also indicates that the specimen is about to become unstable and fail.

[0076] Overall, compared with locked-in specimens, intact specimens exhibit a longer damage stabilization accumulation stage (II), characterized by sparse AE signals and a higher degree of elastic energy accumulation. Conversely, the unstable failure stage (III) is shorter, exhibiting more pronounced sudden failure characteristics. Locked-in specimens, due to stress concentration at the crack tip inducing local micro-fractures, reflect the gradual accumulation of damage, exhibiting "energy dissipation" characteristics and significantly regulating the evolution of each stage of the failure process. Specifically, this manifests as a shortened damage stabilization accumulation stage (II) and a prolonged unstable failure stage (III), demonstrating a significant three-stage progressive failure characteristic. For locked-in specimens, the transition time from stage II to III decreases with increasing crack length (approximately 1200s for 10mm → approximately 800s for 50mm). This is because the effective bearing area of ​​the locked section in long cracks (40-50mm) is reduced, making stress concentration more likely to reach the fracture critical value, thus accelerating the instability process.

[0077] 4. Discussion As a typical heterogeneous material, rock exhibits significant nonlinear and stochastic characteristics in its crack initiation and instability failure processes. Furthermore, the acoustic emission (AE) signals and diffraction (DIC) strain information show high density and volatility in the time domain, making it difficult to intuitively determine their evolutionary patterns and pre-fracture information. Therefore, employing nonlinear theoretical methods to reveal the fracture mechanism of locked sandstone under loading is particularly important. Compared to single fractals, multifractals can characterize the complexity and nonlinear characteristics of acoustic emission signals in more detail.

[0078] 4.1 Coupling Analysis of Acoustic-Optical Signal Precursor Information In previous studies on multifractal characteristic parameters, many scholars have adopted... Δf and The coupled analytical method captures short-term precursor information of rock mass failure. Through comparative analysis of the entire loading process, the author found that compared with... Δf and The combination and The temporal evolution characteristics of the parameters exhibit more significant symmetry and abrupt change, providing a better quantitative indicator for accurately identifying the precursors of failure in locked-in specimens. Accordingly, this invention calculates the multifractal parameters for both acoustic emission signals (characterized by ring count) and DIC optical signals (characterized by the maximum principal strain on the specimen surface). and , drawing Figures 2-4 The figure shows the time-varying evolution curves of the multifractal parameters throughout the loading process.

[0079] Depend on Figures 2-4 Overall analysis shows that the acoustic signal of the sample in the initial stage of loading... and All are in a state of low-amplitude fluctuation. The value is small (<0.5) and The value is close to zero, indicating that the internal damage is mainly characterized by micro-crack frictional slip, and the damage exhibits a homogeneous feature; the DIC signal... and The fluctuations are smoother, and there is even a short-term signal loss in the initial stage of loading. This is attributed to the fact that the strain field on the sample surface has not yet formed a localized accumulation region, resulting in low macro-fractal complexity. With increasing axial stress, the acoustic signal parameters show no significant change; the optical signal... A slight increase began to appear. Increased amplitude and denser fluctuations following the increase in tensile cracks at the trailing edge correspond to the formation of a "butterfly-shaped" high-strain concentration zone at the creep section or the tip of the tensile crack, indicating a transition from a globally uniform to a locally concentrated surface strain field. Near failure, the acoustic signal... Sudden increase The decrease in symmetry indicates that small-scale cracks rapidly propagate to form macroscopic damage bands, the difference in multi-scale events reaches its peak, and the internal damage enters a critical instability state. At this point, the stress level is approximately 80% to 87% of the peak intensity (σ = 0.80σ). c ~0.87σ c The optical signal appears earlier than the external signal abrupt change, reflecting the advanced response characteristics of internal micro-damage; subsequently, the precursor point of the edge tensile crack increases earlier; and The abrupt change lags behind the acoustic signal and The increase is less than that of the acoustic signal, and the stress level is approximately 89% to 94% of the peak intensity (σ = 0.89σ). c ~0.94σ c This corresponds to the visible propagation of macroscopic cracks on the surface, reflecting the "hysteretic response" of the external strain field to internal damage—the internal cracks need to penetrate to a certain extent before significant surface deformation is triggered.

[0080] During the period from the occurrence of the warning point to the failure of the sample, the sample is in a critical state of significant high stress self-organization. For acoustic emission signals, this directly captures the transient elastic waves of internal crack initiation, propagation, and friction. and The dynamic changes can accurately characterize the non-uniform evolution process of "microcrack nucleation-propagation-through" within the locked segment; for the DIC signal, it reflects the macroscopic distribution of the surface strain field. and The changes reflect the gradual characteristics of external deformation from uniformity to localization, and the two together constitute a multi-dimensional representation of "internal-external" damage.

[0081] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for identifying precursors of rock slope failure based on AE and DIC multifractal features, characterized in that, The steps are as follows: S1: During the loading process of the rock slope or the sample, the signals are collected and the collected signals are constructed into a discrete time series in chronological order; S2: Perform multifractal analysis on the time series of AE and DIC signals; S3: Extract the spectral width of the feature parameters of the multifractal. and spectral width difference And determine whether the rock mass is damaged.

2. The method for identifying precursors of rock slope failure based on AE and DIC multifractal features according to claim 1, characterized in that, In step S1, the signals collected include: Acoustic emission ringing count or energy signal; Surface strain signals obtained based on digital image correlation technology.

3. The method for identifying precursors of rock slope failure based on AE and DIC multifractal features according to claim 1, characterized in that, In step S3, multifractal analysis is performed on the time series, and the steps are as follows: The multifractal spectrum of internal and external signals is calculated using box counting. The time series of the collected information is taken as the dataset {Ti}, and the time series is divided into N consecutive subsets, each of length n. The normalized probability measure of each subset is given by the following expression: ； in, P i To normalize the probability density, S i For the first i Cumulative sum over a period of time; By using power-law weighting, the multifractal dimension can be divided into multiple regions to different degrees, and after Legendre transformation, the fractal spectral function f(q) and the average singular intensity α(q) can be calculated.

4. The method for identifying precursors of rock slope failure based on AE and DIC multifractal features according to claim 3, characterized in that, The expression for the fractal spectral function f(q) is as follows: ； The expression for the average singularity intensity α(q) is as follows: ； In the formula: u i Let q be the partition function, q be the statistical moment order, and n be the scale length. P i (n) is the normalized probability measure.

5. The method for identifying precursors of rock slope failure based on AE and DIC multifractal features according to claim 4, characterized in that, Introducing the generalized fractal dimension D q To determine whether a signal has multifractal characteristics, the expression is as follows: ； If D q The signal decreases significantly with increasing q and exhibits a nonlinear relationship, indicating that the signal has multifractal characteristics.

6. The method for identifying precursors of rock slope failure based on AE and DIC multifractal features according to claim 1, characterized in that, In step S3, when acoustic emission... and When abnormal signals of sudden decrease or increase appear in the numerical values, it is determined that the interior of the rock mass has entered a critical state of failure; when DIC... and When numerical values ​​show a sudden decrease or surge in abnormal signals, the rock mass is determined to have entered a critical state of failure; when acoustic emission and DIC... and When abnormal signals of sudden decrease or increase appear in the numerical values, it is determined that the specimen is about to fail.

Images