A damage early warning device and method for rock uniaxial compression test

Abstract

The present application relates to the technical field of rock mechanics test monitoring and safety early warning, and specifically discloses a failure early warning device and method for rock uniaxial compression test, which comprises the following steps: real-time acquisition of load data and displacement data of rock samples in the process of rock uniaxial compression test; preprocessing of the load-displacement data and determination of the optimal sliding window length; calculation of real-time tangent stiffness and stiffness degradation rate of the rock samples in the process of loading; conversion of the load-displacement data into stress-strain data, obtaining of dissipated energy and dissipated energy change rate; construction of continuous damage variable and obtaining of damage evolution rate; weighted combination of the stiffness degradation rate, the dissipated energy change rate and the damage evolution rate, obtaining of a comprehensive early warning index and analysis of the evolution trend, determination of whether the rock sample enters the instability early warning range, and output of early warning information. The present application can realize early identification of the instability development trend of rock, significantly improve the safety of the test, and has a wide range of application.

Description

Technical Field

[0001] This invention relates to the field of rock mechanics test monitoring and safety early warning technology, and specifically discloses a failure early warning device and method for uniaxial compression tests of rocks. Background Technology

[0002] Uniaxial compression testing of rocks is an important method for studying the mechanical properties of rocks, and it is widely used in geotechnical engineering, mining engineering, underground engineering, and geological disaster research. During the test, an axial load is applied to the rock sample using a loading device, and a load-displacement curve can be obtained to analyze the rock's deformation characteristics, peak bearing capacity, and failure mode, among other mechanical properties.

[0003] During rock loading, its internal structure typically undergoes stages such as crack compaction, elastic deformation, stable crack propagation, and unstable failure. As the rock approaches its peak load-bearing capacity, internal microcracks rapidly propagate and eventually connect, ultimately leading to sudden failure. This process is usually characterized by significant brittleness, with the moment of failure potentially accompanied by fragmentation, splashing, and rapid energy release, posing safety risks to testing equipment and personnel. Currently, rock mechanics testing systems primarily rely on peak loads, extreme displacement values, or load variation trends to determine the rock's failure state. However, when the rock reaches or approaches its peak load-bearing capacity, its internal damage has usually already progressed to an unstable stage, with a short warning time, making it difficult to take timely and effective protective measures. Therefore, identifying the evolution trend of rock before it becomes unstable and achieving early warning has become a crucial issue in the safety monitoring of rock mechanics testing.

[0004] During loading, as microcracks propagate and damage accumulates, the overall stiffness, energy dissipation, and damage evolution of the rock all undergo significant changes. Stiffness changes reflect the degradation process of the rock's internal structure; energy dissipation often increases significantly and exhibits abrupt changes near the instability stage; damage variables and their evolution rates also increase significantly during the rapid crack propagation stage. These characteristics can characterize the development process of rock from stable damage to unstable failure from different perspectives. However, existing methods mostly focus on qualitative analysis of experimental curves, lacking effective methods for real-time calculation of stiffness evolution, energy dissipation, and damage evolution characteristics using basic experimental data and for achieving safety early warning. Furthermore, some monitoring technologies rely on additional equipment such as acoustic emission and digital image correlation, resulting in complex and costly systems that are difficult to promote and apply in conventional rock mechanics testing systems. Therefore, it is necessary to propose a failure early warning device and method for uniaxial compression tests of rocks. By performing real-time calculation and comprehensive analysis on the characteristics of rock stiffness deterioration, energy dissipation, and damage evolution, the evolution process of rocks from the stable damage stage to the unstable stage can be identified, thereby realizing real-time safety early warning for uniaxial compression tests of rocks and improving the safety and reliability of the test. Summary of the Invention

[0005] This invention addresses the problems of difficulty in timely identification of rock instability development trends, short early warning time, insufficient test safety, and high cost in existing uniaxial compression tests of rocks by proposing a failure early warning device and method for uniaxial compression tests of rocks.

[0006] This invention provides a failure early warning method for uniaxial compression tests of rocks, comprising the following steps:

[0007] S1. Real-time acquisition of load and displacement data of rock specimens during uniaxial compression tests;

[0008] S2. Preprocess the collected load and displacement data to establish load-displacement data and determine the optimal sliding window length;

[0009] S3. Based on the optimal sliding window length, the sliding window regression algorithm is used to calculate the real-time tangential stiffness and stiffness degradation rate of the rock sample during the loading process; the load-displacement data is converted into stress-strain data, and the dissipated energy and the rate of change of dissipated energy are obtained based on the stress-strain data.

[0010] S4. Construct continuous damage variables based on the real-time tangential stiffness, and obtain the damage evolution rate based on the continuous damage variables;

[0011] S5. The stiffness degradation rate, dissipation energy change rate and damage evolution rate are weighted and combined to obtain a comprehensive early warning index. The evolution trend of the comprehensive early warning index is analyzed to determine whether the rock sample has entered the instability early warning range.

[0012] S6. If the rock sample is determined to have entered the instability warning range, then output the warning information to complete the damage warning.

[0013] According to some embodiments of this application, a failure early warning method for uniaxial compression tests of rocks, in step S2, determining the optimal sliding window length includes:

[0014] S201. The preset candidate window length set is shown in formula (1):

[0015] (1)

[0016] in, Represents the set of candidate window lengths. Indicates the minimum window length. Indicates the window step size. Indicates the maximum window length;

[0017] S202. At the current sampling time For any candidate window length Select the most recent The load-displacement data of each sampling point are shown in formula (2):

[0018] (2)

[0019] in, Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point;

[0020] Local linear fitting is performed on the load-displacement data corresponding to each candidate window length, as shown in Equation (3):

[0021] (3)

[0022] in, Represents load data, Indicates the candidate tangent stiffness. Represents displacement data. Indicates the regression intercept;

[0023] Candidate tangent stiffness is obtained using the least squares method. As shown in formula (4):

[0024] (4)

[0025] in, Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point;

[0026] Regression intercept As shown in formula (5):

[0027] (5)

[0028] in, This represents the average load data corresponding to the candidate window length. , This represents the average displacement data corresponding to the candidate window length. ; This represents the average candidate tangent stiffness corresponding to the candidate window length.

[0029] S203. Calculate the mean squared fitting error and goodness of fit for each candidate window length, wherein the mean squared fitting error is as shown in formula (6):

[0030] (6)

[0031] in, This represents the mean squared fit error;

[0032] The goodness of fit is shown in formula (7):

[0033] (7)

[0034] in, Indicates goodness of fit;

[0035] S204. Construct a comprehensive evaluation function based on the mean squared fitting error and goodness of fit for each candidate window length, as shown in formula (8):

[0036] (8)

[0037] in, This represents the comprehensive evaluation function. This indicates the median of the local goodness of fit corresponding to the candidate window length. This indicates the median of the local mean square fitting error corresponding to the candidate window length. , ,and Denotes the weight coefficients, and satisfies , The value range is 0.4 to 0.5. The value range is 0.3 to 0.4. The value range is 0.2 to 0.3;

[0038] S205. Select the candidate window length with the smallest comprehensive evaluation function as the optimal sliding window length, as shown in formula (9):

[0039] (9)

[0040] in, This represents the optimal sliding window length.

[0041] According to some embodiments of this application, a failure early warning method for uniaxial compression tests of rocks is provided. In step S3, the load-displacement data within the optimal sliding window length is locally linearly fitted using the least squares method to obtain the real-time tangential stiffness, as shown in formula (10):

[0042] (10)

[0043] in, Indicates real-time tangential stiffness. This represents the number of sampling points within the optimal sliding window length;

[0044] The stiffness degradation rate is obtained from the real-time tangential stiffness, as shown in formula (11):

[0045] (11)

[0046] in, This indicates the rate of stiffness degradation.

[0047] According to some embodiments of this application, a failure early warning method for a uniaxial compression test of rock, in step S3, converting load-displacement data into stress-strain data includes obtaining stress based on the load data and the cross-sectional area of ​​the rock specimen, as shown in formula (12):

[0048] (12)

[0049] in, Indicates the first Stress at each sampling point Represents the cross-sectional area of ​​the rock sample;

[0050] The strain is obtained based on the displacement data and the cross-sectional area of ​​the initial height of the rock sample, as shown in formula (13):

[0051] (13)

[0052] in, Indicates the first Strain at each sampling point Indicates the initial height of the rock sample;

[0053] The total input strain energy of the rock sample is obtained from the stress-strain data, as shown in formula (14):

[0054] (14)

[0055] in, This represents the total input strain energy. Indicates real-time stress. Indicates real-time response. Indicates the first Stress at each sampling point Indicates the first Stress at each sampling point Indicates the first Strain at each sampling point Indicates the first Strain at each sampling point;

[0056] The elastic properties of the rock specimen are obtained from the stress-strain data, as shown in formula (15):

[0057] (15)

[0058] in, Indicates elasticity. This represents the initial elastic modulus of the rock sample;

[0059] The dissipated energy is obtained from the total input strain energy and elastic energy, as shown in formula (16):

[0060] (16)

[0061] in, Indicates dissipated energy;

[0062] The rate of change of the dissipated energy is obtained from the dissipated energy, as shown in formula (17):

[0063] (17)

[0064] in, This represents the rate of change of dissipated energy.

[0065] According to some embodiments of this application, a method for early warning of failure in a uniaxial compression test of rock, in step S4, a continuous damage variable is constructed based on the real-time tangential stiffness, as shown in formula (18):

[0066] (18)

[0067] in, Represents continuous damage variables, Indicates the initial stiffness of the rock sample;

[0068] The damage evolution rate is obtained from the continuous damage variables, as shown in formula (19):

[0069] (19)

[0070] in, This indicates the rate of damage evolution.

[0071] According to some embodiments of this application, a failure early warning method for uniaxial compression tests of rocks is provided. In step S5, the stiffness deterioration rate, dissipation energy change rate, and damage evolution rate are weighted and combined to obtain a comprehensive early warning index, as shown in formula (20):

[0072] (20)

[0073] in, This indicates the comprehensive early warning index. The weighting coefficients representing the rate of stiffness degradation. Weighting coefficients representing the rate of change of dissipated energy. Weighting coefficients representing the rate of damage evolution. , The value range is 0.4 to 0.5. The value range is 0.3 to 0.4. The value range is 0.2 to 0.3.

[0074] According to some embodiments of this application, a method for early warning of failure in a uniaxial compression test of rock is provided. In step S5, determining whether the rock sample has entered the instability warning range includes observing whether the comprehensive warning index shows abnormal growth, abrupt change, or inflection point characteristics during the loading process. If abnormal growth, abrupt change, or inflection point characteristics are observed, it is determined that the rock sample has entered the instability warning range from the stable damage stage.

[0075] According to some embodiments of this application, a failure early warning method for a uniaxial compression test of rock is provided. In step S6, the early warning information includes one or more output forms such as text prompts, sound alarm signals, light alarm signals, and sending a shutdown or load reduction command to the test control system.

[0076] According to some embodiments of this application, a method for early warning of failure in a uniaxial compression test of rock includes light alarm signals ranging from blue, yellow, orange and red, in order of increasing strength.

[0077] The present invention also provides a failure early warning device for uniaxial compression tests of rocks, comprising:

[0078] The data acquisition module is used to acquire load and displacement data of the rock specimen during the uniaxial compression test in real time.

[0079] The data preprocessing module is used to preprocess the collected load data and displacement data, establish load-displacement data, and determine the optimal sliding window length.

[0080] The stiffness calculation module is used to calculate the real-time tangential stiffness and stiffness degradation rate of the rock sample under load using a sliding window regression algorithm based on the optimal sliding window length.

[0081] The energy analysis module is used to convert load-displacement data into stress-strain data, and obtain the dissipated energy and the rate of change of dissipated energy based on the stress-strain data.

[0082] The damage variable calculation module is used to construct continuous damage variables based on the real-time tangential stiffness and obtain the damage evolution rate based on the continuous damage variables.

[0083] The multi-parameter coupled early warning module is used to weight and combine the stiffness degradation rate, dissipation energy change rate and damage evolution rate to obtain a comprehensive early warning index, analyze the evolution trend of the comprehensive early warning index, and determine whether the rock sample has entered the instability early warning range.

[0084] The early warning output module is used to output early warning information and complete the damage warning.

[0085] The present invention provides a failure early warning device and method for uniaxial compression tests of rocks, which has the following beneficial effects:

[0086] 1. Ability to identify rock instability development trend in advance: By analyzing the changes in stiffness degradation rate, dissipation energy change rate and damage evolution rate of rock samples during loading, the evolution process of rock from stable damage stage to instability stage can be identified. Compared with the method that only relies on load change for judgment, the trend of rock instability development can be identified earlier.

[0087] 2. Significantly improve the safety of rock mechanics tests: By monitoring the changes in rock mechanics parameters in real time, early warning signals can be issued in a timely manner when the rock enters the critical stage of instability, thereby reducing the risks of fragmentation and splashing during sudden rock failure and improving the safety level of test equipment and personnel.

[0088] 3. Excellent real-time monitoring capability: The sliding window regression algorithm is used to perform real-time calculations on the load-displacement data collected during the test. It can continuously update parameters such as rock stiffness, dissipated energy and damage variables during the loading process, so as to realize dynamic monitoring of the rock damage evolution process.

[0089] 4. Wide range of applications: Early warning analysis can be achieved by relying solely on the load and displacement data routinely collected during the uniaxial compression test of rock, without the need for additional complex monitoring equipment, resulting in low cost. It is applicable to uniaxial compression tests of rocks with different lithologies and under different loading rate conditions. Attached Figure Description

[0090] Figure 1 This is a schematic flowchart of a failure early warning method for a uniaxial compression test of rock according to Embodiment 1 of the present invention;

[0091] Figure 2 This is the load-displacement data curve in Embodiment 2 of the present invention;

[0092] Figure 3 This is the real-time tangential stiffness variation curve in Embodiment 2 of the present invention;

[0093] Figure 4 This is the stiffness degradation rate change curve in Embodiment 2 of the present invention;

[0094] Figure 5 This is the energy dissipation variation curve in Embodiment 2 of the present invention;

[0095] Figure 6 This is the curve showing the rate of change of dissipated energy in Embodiment 2 of the present invention;

[0096] Figure 7 This is the continuous damage variable curve in Embodiment 2 of the present invention;

[0097] Figure 8 This is the damage evolution rate curve in Embodiment 2 of the present invention;

[0098] Figure 9 This is the trend line of the comprehensive early warning index in Embodiment 2 of the present invention. Detailed Implementation

[0099] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0100] Example 1: This example provides a failure early warning method for uniaxial compression tests of rocks, such as... Figure 1 As shown, it includes the following steps:

[0101] S1. Real-time acquisition of load and displacement data of rock specimens during uniaxial compression tests; During uniaxial compression tests, load and displacement data of rock specimens during loading can be acquired in real time using load sensors and displacement sensors.

[0102] S2. Preprocess the collected load and displacement data to establish load-displacement data and determine the optimal sliding window length.

[0103] S3. Based on the optimal sliding window length, the sliding window regression algorithm is used to calculate the real-time tangential stiffness and stiffness degradation rate of the rock sample during the loading process; the load-displacement data are converted into stress-strain data, and the dissipated energy and the rate of change of dissipated energy are obtained based on the stress-strain data.

[0104] S4. Construct continuous damage variables based on real-time tangent stiffness, and obtain the damage evolution rate based on the continuous damage variables.

[0105] S5. The stiffness degradation rate, dissipation energy change rate and damage evolution rate are weighted and combined to obtain a comprehensive early warning index. The evolution trend of the comprehensive early warning index is analyzed to determine whether the rock sample has entered the instability early warning range.

[0106] S6. If the rock sample is determined to have entered the instability warning range, the warning information is output to complete the damage warning, realize the real-time safety warning of the uniaxial compression test process of rock, and improve the safety, real-time performance and reliability of the test process.

[0107] Example 2: This example provides a failure early warning method for uniaxial compression tests of rocks, including the following steps:

[0108] S1. Real-time acquisition of load and displacement data of rock specimens during uniaxial compression tests; data can be continuously recorded according to a set sampling frequency and transmitted in time series form. The acquired data mainly includes load and displacement data corresponding to each sampling moment during the test, which are used to reflect the stress state and deformation response of the rock specimens during loading.

[0109] S2. Preprocess the collected load and displacement data to establish load-displacement data, and obtain load-displacement data curves, such as... Figure 2 As shown, the optimal sliding window length is determined.

[0110] More specifically, determining the optimal sliding window length includes:

[0111] S201. The preset candidate window length set is shown in formula (1):

[0112] (1)

[0113] in, Represents the set of candidate window lengths. Indicates the minimum window length. Indicates the window step size. This indicates the maximum window length.

[0114] S202. At the current sampling time For any candidate window length Select the most recent The load-displacement data of each sampling point are shown in formula (2):

[0115] (2)

[0116] in, Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point.

[0117] Local linear fitting is performed on the load-displacement data corresponding to each candidate window length, as shown in Equation (3):

[0118] (3)

[0119] in, Represents load data, Indicates the candidate tangent stiffness. Represents displacement data. This represents the regression intercept.

[0120] Candidate tangent stiffness is obtained using the least squares method. As shown in formula (4):

[0121] (4)

[0122] in, Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point.

[0123] Regression intercept As shown in formula (5):

[0124] (5)

[0125] in, This represents the average load data corresponding to the candidate window length. , This represents the average displacement data corresponding to the candidate window length. ; This represents the average stiffness of the candidate tangent corresponding to the candidate window length.

[0126] S203. Calculate the mean squared fitting error and goodness of fit for each candidate window length to evaluate the fitting effect under different candidate window lengths. The mean squared fitting error is shown in formula (6):

[0127] (6)

[0128] in, This represents the mean squared fit error.

[0129] The goodness of fit is shown in formula (7):

[0130] (7)

[0131] in, This indicates the goodness of fit.

[0132] S204. Construct a comprehensive evaluation function based on the mean squared fitting error and goodness of fit for each candidate window length, as shown in formula (8):

[0133] (8)

[0134] in, This represents the comprehensive evaluation function. This indicates the median of the local goodness of fit corresponding to the candidate window length. This indicates the median of the local mean square fitting error corresponding to the candidate window length. , ,and Denotes the weight coefficients, and satisfies Taking into account local linear fitting accuracy, error control, and the ability to preserve stiffness abrupt changes, emphasis is placed on the goodness of local linear fitting. The value range is 0.4~0.5; attention should be paid to fitting error to ensure stable stiffness calculation. The value range is 0.3 to 0.4; the window length constraint should be appropriately considered to avoid noise amplification caused by an excessively small window. The value range is 0.2 to 0.3.

[0135] S205. Select the candidate window length with the smallest comprehensive evaluation function as the optimal sliding window length, as shown in formula (9):

[0136] (9)

[0137] in, This represents the optimal sliding window length.

[0138] Using the above method, the optimal sliding window length can be automatically determined based on real-time load-displacement data, so that the calculated tangential stiffness has both high fitting accuracy and can maintain sensitivity to the abrupt changes in rock stiffness characteristics, thereby improving the stability and reliability of subsequent calculations of stiffness degradation rate, damage evolution rate and comprehensive early warning index.

[0139] S3. Based on the optimal sliding window length, the sliding window regression algorithm is used to calculate the real-time tangential stiffness and stiffness degradation rate of the rock sample during the loading process; the load-displacement data are converted into stress-strain data, and the dissipated energy and the rate of change of dissipated energy are obtained based on the stress-strain data.

[0140] More specifically, the real-time tangent stiffness is obtained by performing local linear fitting on the load-displacement data within the optimal sliding window length using the least squares method, as shown in formula (10):

[0141] (10)

[0142] in, Indicates real-time tangential stiffness. This represents the number of sampling points within the optimal sliding window length;

[0143] The stiffness degradation rate is further obtained through real-time tangential stiffness, as shown in formula (11):

[0144] (11)

[0145] in, This indicates the rate of stiffness degradation.

[0146] Stiffness degradation rate is used to characterize the change in the overall load-bearing capacity of rock as a function of loading. When a rock sample approaches instability, such as... Figure 3 and Figure 4 As shown, the real-time tangential stiffness usually decreases significantly, and the corresponding stiffness degradation rate also changes abnormally.

[0147] In step S3, after obtaining the load data and displacement data, the original data can be converted and calculated based on the geometric parameters of the rock specimen, converting the load-displacement data into stress-strain data, including obtaining the stress based on the load data and the cross-sectional area of ​​the rock specimen, as shown in formula (12):

[0148] (12)

[0149] in, Indicates the first Stress at each sampling point This represents the cross-sectional area of ​​the rock sample.

[0150] The strain is obtained from the displacement data and the cross-sectional area of ​​the initial height of the rock sample, as shown in formula (13):

[0151] (13)

[0152] in, Indicates the first Strain at each sampling point This indicates the initial height of the rock sample.

[0153] The total input strain energy of the rock specimen is obtained from the stress-strain data. Under discrete data conditions, it can be calculated using the trapezoidal integral form. The total input strain energy can be continuously updated during the test. This energy reflects the total work input to the rock specimen by the external loading system. As shown in formula (14):

[0154] (14)

[0155] in, This represents the total input strain energy. Indicates real-time stress. Indicates real-time response. Indicates the first Stress at each sampling point Indicates the first Stress at each sampling point Indicates the first Strain at each sampling point Indicates the first Strain at each sampling point.

[0156] The elastic properties of the rock specimen are obtained from the stress-strain data, as shown in formula (15):

[0157] (15)

[0158] in, Indicates elasticity. This represents the initial elastic modulus of the rock sample.

[0159] The energy consumed within a rock due to processes such as microcrack propagation, particle friction, structural slippage, and irreversible damage is defined as dissipated energy. It can be obtained from the total input strain energy and elastic energy, as shown in formula (16):

[0160] (16)

[0161] in, This represents dissipated energy. In the initial stage of loading, if the rock sample is mainly in the elastic response stage, most of the external input energy is stored in the form of elastic energy, and the dissipated energy is relatively small. As loading continues, microcracks gradually initiate and propagate, frictional slip and damage evolution intensify, and the dissipated energy gradually increases. When the rock sample approaches the critical state of instability, cracks rapidly propagate, connect, and penetrate, and the dissipated energy often exhibits a significant rapid growth characteristic. Therefore, dissipated energy can serve as an important indicator reflecting irreversible damage activity within the rock.

[0162] The rate of change of dissipated energy is obtained from the dissipated energy, as shown in formula (17):

[0163] (17)

[0164] in, This represents the rate of change of dissipated energy.

[0165] During the process of rock being subjected to compressive loading, part of the energy input from the outside is stored inside the rock in the form of elastic energy, and the other part is converted into dissipated energy consumed by irreversible processes such as crack propagation and frictional slip, such as... Figure 5 and Figure 6 As shown, the dissipation energy and the rate of change of dissipation energy change, which can be used to identify the rapid propagation stage of cracks inside rocks.

[0166] S4. Construct continuous damage variables based on real-time tangent stiffness, and obtain the damage evolution rate based on the continuous damage variables;

[0167] More specifically, a continuous damage variable is constructed based on the real-time tangent stiffness, as shown in formula (18):

[0168] (18)

[0169] in, Represents continuous damage variables, This indicates the initial stiffness of the rock sample.

[0170] The damage evolution rate is obtained from the continuous damage variables, as shown in formula (19):

[0171] (19)

[0172] in, This indicates the rate of damage evolution.

[0173] As the loading process proceeds, internal damage to the rock gradually accumulates, such as... Figure 7 As shown, the variable of continuous damage increases continuously. (As...) Figure 8 As shown, when the damage evolution rate increases significantly, it indicates that the internal crack propagation of the rock accelerates.

[0174] S5. In order to improve the stability of early warning judgment, this embodiment adopts a multi-parameter coupling method to weight and combine the stiffness deterioration rate, dissipation energy change rate and damage evolution rate to obtain a comprehensive early warning index. The evolution trend of the comprehensive early warning index is analyzed to determine whether the rock sample has entered the instability early warning range.

[0175] More specifically, the stiffness degradation rate, dissipation energy change rate, and damage evolution rate are weighted and combined to obtain a comprehensive early warning index, as shown in formula (20):

[0176] (20)

[0177] in, This indicates the comprehensive early warning index. The weighting coefficient represents the stiffness degradation rate and controls the influence of stiffness degradation characteristics on the comprehensive early warning system. The weighting coefficient represents the rate of change of dissipated energy and controls the influence of the characteristics of dissipated energy change. Weighting coefficients representing the rate of damage evolution control the influence of damage evolution characteristics; Taking into account both the direct characterizing effect of stiffness degradation on load-bearing capacity degradation and the auxiliary characterizing effects of dissipation energy change rate and damage evolution rate on crack propagation process, The value range is 0.4 to 0.5. The value range is 0.3 to 0.4. The value range is 0.2 to 0.3. Furthermore, sensitivity analysis of different weight combinations can verify the stability of the comprehensive early warning index with respect to the weight values. The comprehensive early warning index is calculated in real time during the loading process, and its evolution trend is analyzed.

[0178] In step S5, determining whether the rock sample has entered the instability warning range includes observing whether the comprehensive warning index shows abnormal growth, abrupt changes, or inflection point characteristics during the loading process, such as... Figure 9 As shown, if the comprehensive early warning index shows abnormal growth, sudden change characteristics or inflection point characteristics, it is determined that the rock sample has entered the instability early warning range from the stable damage stage.

[0179] S6. If the rock sample is determined to have entered the instability warning range, then output the warning information to complete the damage warning.

[0180] More specifically, the warning information includes one or more output forms such as text prompts, sound alarm signals, light alarm signals, and sending shutdown or load reduction commands to the test control system.

[0181] The light alarm signals, from weakest to strongest, include blue, yellow, orange, and red warnings. When necessary, the system can automatically send control commands to the testing machine to stop loading or slow down the loading rate, thereby reducing the risk of sudden rock failure to the testing equipment and personnel.

[0182] Through the above steps, the present invention can monitor the rock damage evolution characteristics in real time during the uniaxial compression test of rock, and realize test safety early warning through multi-parameter coupling analysis, thereby improving the safety and reliability of the rock mechanics test process.

[0183] Example 3 discloses a failure early warning device for uniaxial compression tests of rocks, characterized in that it includes a data acquisition module, a data preprocessing module, a stiffness calculation module, an energy analysis module, a damage variable calculation module, a multi-parameter coupled early warning module, and an early warning output module. The system comprises the following modules: a data acquisition module for real-time acquisition of load and displacement data during uniaxial compression testing of rock samples; a data preprocessing module for preprocessing the acquired load and displacement data, establishing load-displacement data, and determining the optimal sliding window length; a stiffness calculation module for calculating the real-time tangential stiffness and stiffness degradation rate of the rock sample under load using a sliding window regression algorithm based on the optimal sliding window length; an energy analysis module for converting load-displacement data into stress-strain data, and obtaining dissipated energy and its rate of change based on the stress-strain data; a damage variable calculation module for constructing continuous damage variables based on real-time tangential stiffness, and obtaining the damage evolution rate based on the continuous damage variables; a multi-parameter coupled early warning module for weighted combination of stiffness degradation rate, dissipated energy change rate, and damage evolution rate to obtain a comprehensive early warning index, analyzing the evolution trend of the comprehensive early warning index, and determining whether the rock sample has entered the instability early warning range; and an early warning output module for outputting early warning information to complete the damage early warning.

[0184] More specifically, the data acquisition module is used to acquire the raw mechanical response data of the rock specimen during the uniaxial compression test in real time, providing basic data for subsequent load-displacement data processing, stiffness evolution analysis, energy calculation, damage variable calculation, and early warning judgment. The data acquisition module may include a load sensor and a displacement sensor. The load sensor is used to acquire the axial load data applied to the rock specimen by the testing machine in real time; the displacement sensor is used to acquire the axial displacement data of the rock specimen during compression. The load sensor and displacement sensor can be integrated into the uniaxial compression testing machine body or externally connected to the testing machine control system, and connected to the device of this embodiment through a data acquisition card, serial communication module, industrial bus, or other data interface to achieve real-time transmission of test data. The data output by the data acquisition module may include load values, displacement values, and corresponding sampling times in a time series, and is transmitted to the data preprocessing module for subsequent stiffness calculation, energy analysis, and damage variable calculation.

[0185] More specifically, the data preprocessing module reads, sorts, and synchronously verifies the acquired load and displacement data, and establishes continuous load-displacement data according to the sampling time sequence. Simultaneously, the data preprocessing module can also read the geometric parameters of the rock sample, including sample height, diameter or side length, and cross-sectional area, for subsequent energy calculations and data calibration. Furthermore, it can automatically determine the optimal sliding window length based on real-time load-displacement data, ensuring that the calculated tangential stiffness has both high fitting accuracy and maintains sensitivity to abrupt changes in rock stiffness characteristics, thereby improving the stability and reliability of subsequent calculations of stiffness degradation rate, damage evolution rate, and comprehensive early warning index.

[0186] More specifically, the stiffness calculation module solves for the local slope of the load-displacement data during the uniaxial compression test of the rock in real time, thereby obtaining the real-time tangential stiffness of the rock specimen at the current loading stage. This characterizes the evolution of the overall bearing capacity of the rock specimen with the loading process, and further extracts the stiffness degradation rate to characterize the evolution of the rock specimen's bearing capacity with the loading process. During the loading process of the rock specimen, the load-displacement relationship of the rock specimen usually exhibits obvious nonlinear characteristics. In order to obtain the equivalent stiffness at the current stage in real time during the test, this embodiment uses the sliding window linear regression method to locally fit the load-displacement data, thereby calculating the real-time tangential stiffness. As the loading process progresses, when the microcracks inside the rock gradually expand and cause structural damage, the calculated real-time tangential stiffness will gradually decrease, and exhibit obvious accelerated degradation characteristics in the near-instability failure stage. Therefore, by continuously monitoring the real-time tangential stiffness and its evolution trend, important basis can be provided for subsequent stiffness degradation rate calculation, damage variable calculation, and multi-parameter coupled early warning analysis. By continuously calculating the real-time tangential stiffness sequence, the stiffness degradation rate curve of the rock sample during the entire loading process can be obtained. When the rock is in the fracture compaction stage or elastic stage, the stiffness degradation rate is usually small; however, when the rock enters the damage propagation stage, the stiffness degradation rate will gradually increase, and will show obvious abrupt changes or rapid decline characteristics when approaching the instability and failure stage.

[0187] The output of the stiffness calculation module mainly includes real-time tangential stiffness and stiffness degradation rate. Real-time tangential stiffness reflects the immediate load-bearing capacity and overall structural stiffness characteristics of the rock specimen at the current loading stage; stiffness degradation rate characterizes the speed and evolution trend of the overall stiffness of the rock specimen as it changes during loading. During uniaxial compression loading of rock, when the rock specimen is in the fracture compaction stage or elastic deformation stage, the internal structure of the rock specimen is relatively stable, and the real-time tangential stiffness generally changes relatively slowly, with the corresponding stiffness degradation rate usually fluctuating within a small range. As the loading process continues, when microcracks gradually initiate and propagate within the rock specimen, the overall structure of the rock specimen begins to undergo damage evolution, the real-time tangential stiffness gradually decreases, and the amplitude of the stiffness degradation rate also gradually increases. When the rock specimen gradually approaches the critical stage of instability and failure, due to the rapid propagation and penetration of internal cracks, the overall rock structure exhibits significant stiffness degradation. At this time, the real-time tangential stiffness often shows a significant sudden drop, and the corresponding stiffness degradation rate also shows significant abnormal changes. Therefore, by continuously monitoring and analyzing the real-time tangential stiffness and stiffness degradation rate, the evolution process of the internal structure of rock samples from the stable stage to the unstable damage stage can be effectively identified, providing important basic criteria for the subsequent damage variable calculation module and multi-parameter coupled early warning module.

[0188] More specifically, the energy analysis module analyzes the deformation and damage evolution characteristics of rock samples during loading from the perspective of energy conservation and dissipation. During compression, not all external energy input to the rock sample is converted into recoverable elastic energy; a portion of the energy is dissipated due to irreversible processes such as crack initiation, propagation, frictional slip, and structural damage. Compared to analysis solely from the perspective of stress or strain, energy analysis can more directly reveal the evolution of the rock's internal structure and the state of crack activity, thus playing a crucial role in identifying the transition from a stable damage stage to an unstable stage. During the uniaxial compression test of the rock, the raw data acquired by the data acquisition module consists of load and displacement data. To analyze the rock's energy evolution process using strain energy theory, the load-displacement data must first be converted into stress-strain data.

[0189] The main parameters output by the energy analysis module include total input strain energy, elastic energy, dissipated energy, and the rate of change of dissipated energy. Dissipated energy reflects the degree of irreversible damage within the rock, while the rate of change of dissipated energy reflects the intensity and speed of damage activity. Compared to stiffness indices, energy indices focus more on the crack activity process within the rock itself, providing another dimension to characterize the evolution of rock instability development. The energy analysis module, together with the stiffness calculation module, forms a crucial foundation for instability identification.

[0190] More specifically, the damage variable calculation module constructs a continuous damage variable reflecting the degree of internal damage accumulation in the rock based on the real-time tangential stiffness output by the stiffness calculation module during the uniaxial compression test. It further calculates the damage evolution rate to identify the transition from a stable damage stage to a rapid damage development stage. In the initial loading stage, the damage variable changes little, and the damage evolution rate is usually at a low level. As the rock sample enters the stable crack propagation stage, the damage gradually develops, and both the continuous damage variable and the damage evolution rate gradually increase. When the rock sample approaches instability, internal crack activity is intense, and stiffness decays rapidly, leading to a significant accelerated growth in the continuous damage variable. Correspondingly, the damage evolution rate also shows a sudden increase or abnormal rise. This pattern effectively reflects the transition from slow accumulation to rapid development of internal damage in the rock. The damage variable calculation module outputs the continuous damage variable and the damage evolution rate. Unlike the stiffness calculation module, which directly reflects changes in bearing capacity, the damage variable calculation module focuses more on describing the degree of internal structural deterioration in the rock from the perspective of continuous damage mechanics. Compared to the energy analysis module, the damage variable has the advantages of simple expression, clear physical meaning of parameters, and ease of real-time implementation. Therefore, combining the damage variable calculation module with the stiffness calculation module and the energy analysis module can improve the stability and reliability of identifying the critical state of rock instability.

[0191] More specifically, the multi-parameter coupled early warning module is used to comprehensively process the stiffness degradation rate, dissipation energy change rate, and damage evolution rate to construct a unified comprehensive early warning index, thereby achieving real-time identification of the critical instability state of rock samples. Since a single parameter may fluctuate under different lithologies, loading rates, or discrete sample conditions, relying solely on one parameter for identification is easily affected by noise interference or local anomalies. Therefore, this embodiment employs a multi-parameter coupling approach for comprehensive early warning to improve the stability and robustness of the identification. Through the common characterization of the three types of parameters, the comprehensive early warning index can simultaneously reflect information on three aspects: the decline in rock bearing capacity, the enhancement of internal irreversible energy dissipation, and the acceleration of damage accumulation.

[0192] The evolution of the comprehensive early warning index follows this pattern: In the initial stage of rock sample loading, due to small changes in stiffness, low dissipated energy, and slow growth of damage variables, the comprehensive early warning index is generally at a low level and changes slowly. As the test enters the damage propagation stage, the three indicators gradually strengthen and increase. When the rock sample approaches the critical instability state, stiffness rapidly degrades, dissipated energy increases significantly, and damage evolution accelerates. The comprehensive early warning index typically exhibits characteristics such as significant rise, abrupt change, inflection point, or abnormal fluctuations. The multi-parameter coupled early warning module does not rely on whether a single parameter reaches a fixed value, but rather identifies whether the rock has entered the instability warning range through the overall evolution trend of the comprehensive index. This method is more suitable for scenarios where there are differences between different lithological samples and a single threshold is not easily uniformly determined.

[0193] The multi-parameter coupled early warning module can monitor the changes in the comprehensive early warning index in real time. When the comprehensive early warning index shows obvious abnormal growth, abrupt changes, or inflection point characteristics during the loading process, or when its growth trend changes from gradual to significant enhancement, it can be determined that the rock sample has entered the instability early warning range from the stable damage stage. At this time, the module enters the early warning state and can output early warning information according to preset rules.

[0194] The multi-parameter coupled early warning module is the core module of this device. It integrates characteristic information from three dimensions: stiffness, energy, and damage. This allows early warning judgment to move beyond the instantaneous changes of a single mechanical quantity and instead establish a comprehensive identification mechanism based on the synchronous evolution of multi-source characteristics. Compared to existing methods that rely solely on stress peaks, sudden load drops, or single acoustic emission events, the coupled early warning approach more comprehensively reflects the evolution of the rock's internal structure, reduces the interference of single-parameter fluctuations on early warning results, improves the stability and reliability of identifying instability critical states, and facilitates real-time application in conventional uniaxial compression testing systems.

[0195] The stiffness calculation module, energy analysis module, damage variable calculation module, and multi-parameter coupled early warning module provided in this embodiment work together to monitor the mechanical response, energy evolution, and damage development of rock samples in real time during uniaxial compression tests. They also identify the state of rock samples entering the instability warning range through a comprehensive early warning index, thereby achieving real-time safety early warning during uniaxial compression tests and improving the safety, real-time performance, and reliability of the test process.

[0196] More specifically, after the multi-parameter coupled early warning module determines that the rock sample has entered the instability warning range, the early warning output module outputs early warning information to the experimenters or the test control system according to preset rules, realizing safety prompts and interventions during the uniaxial compression test of the rock. The early warning output module is connected to the multi-parameter coupled early warning module and triggers corresponding early warning responses according to preset logic to provide safety prompts to the experimenters before sudden failure of the rock sample. The early warning output module can output various forms of early warning signals, including text prompts, graphical interface warnings, audible alarms, visual alarms, vibration warnings, or send control commands to the test control system to stop loading, reduce the loading rate, or maintain the current loading state.

[0197] Furthermore, the early warning output module can also set different levels of early warning status based on the magnitude or rate of change of the comprehensive early warning index, such as blue warning, yellow warning, orange warning and red warning, so as to take different safety control measures at different risk stages.

[0198] Furthermore, the early warning output module can be configured with a tiered early warning mechanism. Based on the magnitude or degree of change in the comprehensive early warning index, the early warning status can be divided into different levels, each corresponding to a different prompting method and control measure. For example, a prompt message can be issued when the early warning index shows an initial abnormality; an audible and visual alarm can be activated when the abnormality intensifies; and a stop loading command can be sent to the testing machine when the system determines that the sample is approaching an unstable state.

[0199] In practical applications, the early warning output module can also include a human-machine interface unit, an alarm execution unit, and a control linkage unit. The human-machine interface unit is used to display the load-displacement curve, the comprehensive early warning index curve, and the current early warning level; the alarm execution unit is used to drive the buzzer or warning light to issue an alarm signal; and the control linkage unit is used to communicate with the testing machine control system to perform operations such as pausing loading, stopping loading, or reducing load when preset conditions are met.

[0200] In addition, the early warning output module may also include data recording and storage functions to save the early warning trigger time, early warning level, comprehensive early warning index and related test data for subsequent test analysis and early warning model optimization.

[0201] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A method for early warning of failure in uniaxial compression tests of rock, characterized in that, Includes the following steps: S1. Real-time acquisition of load and displacement data of rock specimens during uniaxial compression tests; S2. Preprocess the collected load and displacement data to establish load-displacement data and determine the optimal sliding window length; S3. Based on the optimal sliding window length, the sliding window regression algorithm is used to calculate the real-time tangential stiffness and stiffness degradation rate of the rock sample during the loading process; the load-displacement data is converted into stress-strain data, and the dissipated energy and the rate of change of dissipated energy are obtained from the stress-strain data; S4. Construct continuous damage variables based on the real-time tangential stiffness, and obtain the damage evolution rate based on the continuous damage variables; S5. The stiffness degradation rate, dissipation energy change rate and damage evolution rate are weighted and combined to obtain a comprehensive early warning index. The evolution trend of the comprehensive early warning index is analyzed to determine whether the rock sample has entered the instability early warning range. The stiffness degradation rate, dissipation energy change rate, and damage evolution rate are weighted and combined to obtain a comprehensive early warning index, as shown in formula (1): （1） in, This indicates the comprehensive early warning index. The weighting coefficients representing the rate of stiffness degradation. Weighting coefficients representing the rate of change of dissipated energy. Weighting coefficients representing the rate of damage evolution. , The value range is 0.4 to 0.

5. The value range is 0.3 to 0.

4. The value range is 0.2 to 0.3; S6. If the rock sample is determined to have entered the instability warning range, then output the warning information to complete the damage warning.

2. The failure early warning method for uniaxial compression tests of rock according to claim 1, characterized in that, In step S2, determining the optimal sliding window length includes: S201. The preset candidate window length set is shown in formula (2): （2） in, Represents the set of candidate window lengths. Indicates the minimum window length. Indicates the window step size. Indicates the maximum window length; S202. At the current sampling time For any candidate window length Select the most recent The load-displacement data of each sampling point are shown in formula (3): （3） in, Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point; Local linear fitting is performed on the load-displacement data corresponding to each candidate window length, as shown in Equation (4): （4） in, Represents load data, Indicates the candidate tangent stiffness. Represents displacement data. Indicates the regression intercept; Candidate tangent stiffness is obtained using the least squares method. As shown in formula (5): （5） in, Indicates the first Load data at each sampling point Indicates the first Displacement data of each sampling point; Regression intercept As shown in formula (6): （6） in, This represents the average load data corresponding to the candidate window length. , This represents the average displacement data corresponding to the candidate window length. ; This represents the average candidate tangent stiffness corresponding to the candidate window length. S203. Calculate the mean squared fitting error and goodness of fit for each candidate window length, wherein the mean squared fitting error is as shown in formula (7): （7） in, This represents the mean squared fit error; The goodness of fit is shown in formula (8): （8） in, Indicates goodness of fit; S204. Construct a comprehensive evaluation function based on the mean squared fitting error and goodness of fit for each candidate window length, as shown in formula (9): （9） in, This represents the comprehensive evaluation function. This indicates the median of the local goodness of fit corresponding to the candidate window length. This indicates the median of the local mean square fitting error corresponding to the candidate window length. , ,and Denotes the weight coefficients, and satisfies , The value range is 0.4 to 0.

5. The value range is 0.3 to 0.

4. The value range is 0.2 to 0.3; S205. Select the candidate window length with the smallest comprehensive evaluation function as the optimal sliding window length, as shown in formula (10): （10） in, This represents the optimal sliding window length.

3. The method for early warning of failure in uniaxial compression tests of rock according to claim 2, characterized in that, In step S3, the load-displacement data within the optimal sliding window length is locally linearly fitted using the least squares method to obtain the real-time tangent stiffness, as shown in formula (11): （11） in, Indicates real-time tangential stiffness. This represents the number of sampling points within the optimal sliding window length; The stiffness degradation rate is obtained from the real-time tangential stiffness, as shown in formula (12): （12） in, This indicates the rate of stiffness degradation.

4. The failure early warning method for uniaxial compression tests of rock according to claim 3, characterized in that, In step S3, converting the load-displacement data into stress-strain data includes obtaining the stress based on the load data and the cross-sectional area of ​​the rock specimen, as shown in formula (13): （13） in, Indicates the first Stress at each sampling point Represents the cross-sectional area of ​​the rock sample; The strain is obtained based on the displacement data and the cross-sectional area of ​​the initial height of the rock sample, as shown in formula (14): （14） in, Indicates the first Strain at each sampling point Indicates the initial height of the rock sample; The total input strain energy of the rock sample is obtained from the stress-strain data, as shown in formula (15): （15） in, This represents the total input strain energy. Indicates real-time stress. Indicates real-time response. Indicates the first Stress at each sampling point Indicates the first Stress at each sampling point Indicates the first Strain at each sampling point Indicates the first Strain at each sampling point; The elastic properties of the rock specimen are obtained from the stress-strain data, as shown in formula (16): （16） in, Indicates elasticity. This represents the initial elastic modulus of the rock sample; The dissipated energy is obtained from the total input strain energy and elastic energy, as shown in formula (17): （17） in, Indicates dissipated energy; The rate of change of the dissipated energy is obtained from the dissipated energy, as shown in formula (18): （18） in, This represents the rate of change of dissipated energy.

5. A method for early warning of failure in uniaxial compression tests of rock according to claim 4, characterized in that, In step S4, the continuous damage variable is constructed based on the real-time tangential stiffness, as shown in formula (19): （19） in, Represents continuous damage variables. Indicates the initial stiffness of the rock sample; The damage evolution rate is obtained from the continuous damage variables, as shown in formula (20): （20） in, This indicates the rate of damage evolution.

6. The failure early warning method for uniaxial compression tests of rock according to claim 5, characterized in that, In step S5, determining whether the rock sample has entered the instability warning range includes observing whether the comprehensive warning index shows abnormal growth, sudden change characteristics, or inflection point characteristics during the loading process. If abnormal growth, sudden change characteristics, or inflection point characteristics are observed, it is determined that the rock sample has entered the instability warning range from the stable damage stage.

7. The failure early warning method for uniaxial compression tests of rock according to claim 6, characterized in that, In step S6, the warning information includes one or more output forms such as text prompts, sound alarm signals, light alarm signals, and sending shutdown or load reduction commands to the test control system.

8. A method for early warning of failure in uniaxial compression tests of rock according to claim 7, characterized in that, The light alarm signals, from weakest to strongest, include blue warning, yellow warning, orange warning, and red warning.

9. A failure early warning device for uniaxial compression tests of rocks, characterized in that, include The data acquisition module is used to acquire load and displacement data of the rock specimen during the uniaxial compression test in real time. The data preprocessing module is used to preprocess the collected load data and displacement data, establish load-displacement data, and determine the optimal sliding window length. The stiffness calculation module is used to calculate the real-time tangential stiffness and stiffness degradation rate of the rock sample under load using a sliding window regression algorithm based on the optimal sliding window length. The energy analysis module is used to convert load-displacement data into stress-strain data, and obtain the dissipated energy and the rate of change of dissipated energy based on the stress-strain data. The damage variable calculation module is used to construct continuous damage variables based on the real-time tangential stiffness and obtain the damage evolution rate based on the continuous damage variables. The multi-parameter coupled early warning module is used to weight and combine the stiffness degradation rate, dissipation energy change rate and damage evolution rate to obtain a comprehensive early warning index, analyze the evolution trend of the comprehensive early warning index, and determine whether the rock sample has entered the instability early warning range. The stiffness degradation rate, dissipation energy change rate, and damage evolution rate are weighted and combined to obtain a comprehensive early warning index, as shown in formula (21): （21） in, This indicates the comprehensive early warning index. The weighting coefficients representing the rate of stiffness degradation. Weighting coefficients representing the rate of change of dissipated energy. Weighting coefficients representing the rate of damage evolution. , The value range is 0.4 to 0.

5. The value range is 0.3 to 0.

4. The value range is 0.2 to 0.3; The early warning output module is used to output early warning information and complete the damage warning.

Images