Methods for constructing rock creep models, apparatus, and methods for determining rock failure modes.

Abstract

This application relates to the field of computer technology and discloses a method, apparatus, and method for determining rock failure modes in constructing a rock creep model. The method includes acquiring a dataset; the dataset includes input and output data, where the input data includes creep holding time, stress data of the sample rock, initial macroscopic damage data, initial microscopic damage data, long-term strength, and creep initiation threshold; the output data is the true strain value. The initial rock creep model is fitted using the dataset to obtain the final rock creep model. The initial rock creep model includes at least one of the following: a damaged elastic body strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. Each strain term includes a macroscopic-microscopic nonlinear coupled damage term, which is generated based on the initial microscopic damage data and the initial macroscopic damage data. The model constructed in this application can realistically reflect the nonlinear mechanical response of rock mass under the combined action of multiple-scale initial defects.

Description

Technical Field

[0001] This application relates to the field of computer technology, and more specifically, to a method and apparatus for constructing a rock creep model and a method for determining rock failure modes. Background Technology

[0002] Rock creep models are key tools in rock mechanics, crucial for the safe construction and long-term operation and maintenance of various rock mass engineering projects such as tunnels, mine roadways, and slopes. These models can quantitatively describe the evolution of rock mass deformation over time under constant loads, clearly presenting the complete process from instantaneous elastic deformation, deceleration-steady-state creep to accelerated creep and ultimately instability and failure. This provides core theoretical basis for predicting long-term deformation trends, assessing surrounding rock stability, and forecasting instability and failure time. Furthermore, based on the analysis results of rock creep models, engineers can develop targeted and scientifically sound support schemes and disaster early warning mechanisms to effectively avoid engineering disasters such as collapses, rockbursts, and slope slippage, ensuring the safe and stable operation of rock mass engineering projects.

[0003] Existing creep models mainly include viscoelastic-viscoplastic component combinations such as Nishihara, Burgers, and Poynting-Thomson, as well as extended models based on empirical formulas or nonlinear viscous bodies. These models can describe instantaneous creep, decelerating creep, and accelerating creep behavior to varying degrees.

[0004] However, as mineral resource extraction and underground space development extend to deeper areas, the geological environment of the surrounding rock becomes increasingly complex, exhibiting characteristics of both high ground stress and strong structural surfaces (joints and fissures). Deep rock masses not only contain primary macroscopic geological defects such as joints and fissures, but also microscopic damage caused by excavation unloading or historical tectonic stress, such as microcracks and pores. The presence of these multi-scale initial defects makes the creep characteristics of rocks highly nonlinear, often leading to catastrophic processes from "decelerated deformation" to "sudden instability." Existing studies have shown that rock creep behavior is not only affected by the loading level but is also highly susceptible to the control of the degree of initial damage, especially under conditions of pre-existing macroscopic fissures or significant unloading damage, where the creep rate, accelerated creep initiation time, and long-term strength of the rock all change significantly. In this context, most existing rock creep models treat macroscopic structural damage and unloading-induced microscopic stress damage separately, making it difficult for the models to accurately reflect the nonlinear mechanical response of deep rock masses under the combined action of multi-scale defects. Therefore, it is urgent to solve this technical problem. Summary of the Invention

[0005] In view of the above situation, this application provides a method, apparatus and method for constructing a rock creep model and a method for determining rock failure modes, which aims to solve the above problems or at least partially solve the above problems.

[0006] In a first aspect, embodiments of this application provide a method for constructing a rock creep model, the method comprising:

[0007] Obtain the dataset; each sample in the dataset includes model input data and model output data, wherein the model input data includes creep holding time, stress data of sample rock, initial macroscopic damage data and initial microscopic damage data of sample rock, as well as preset long-term strength and preset creep initiation threshold, and the model output data is the strain true value;

[0008] Each sample in the dataset is used as a target sample. The model input data of the target samples is processed using an initial rock creep model to generate strain prediction values. The initial rock creep model includes at least one of the following: a damaged elastic body strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. Each strain term includes a macro-micro nonlinear coupled damage term, which is generated based on initial micro-damage data and initial macro-damage data. The initial micro-damage data is determined based on the degradation of the elastic modulus of the intact rock sample after experiencing loading and unloading paths, and the initial macro-damage data is determined based on the strength or stiffness attenuation rate of the sample rock relative to the intact rock sample.

[0009] Using a preset loss function, a loss value is calculated based on each predicted strain value and the corresponding true strain value. The initial rock creep model is then fitted according to each loss value to obtain the final rock creep model.

[0010] Secondly, embodiments of this application also provide a method for determining rock failure modes, the method comprising:

[0011] Obtain the actual stress data and actual creep deformation data of the target rock;

[0012] Initialize the initial macroscopic damage data of the target rock to obtain the initial macroscopic damage data to be adjusted;

[0013] Using the target rock creep model obtained by the rock creep model construction method described in the first aspect, the initial macroscopic damage data to be adjusted and the actual stress data are processed to generate theoretical creep deformation data;

[0014] Based on the theoretical creep deformation data and the actual creep deformation data, the initial macroscopic damage data to be adjusted is adjusted to obtain the initial macroscopic damage prediction data.

[0015] Based on the initial macroscopic damage prediction data and the creep damage rate parameters in the target rock creep model, the failure modes of the target rock are determined; wherein, the failure modes include structurally controlled failure, stress-controlled failure, and combined failure.

[0016] Thirdly, embodiments of this application also provide a device for constructing a rock creep model, the device comprising:

[0017] The acquisition module is used to acquire a dataset; each sample in the dataset includes model input data and model output data, wherein the model input data includes creep holding time, stress data of sample rock, initial macroscopic damage data and initial microscopic damage data of sample rock, as well as preset long-term strength and preset creep initiation threshold, and the model output data is the strain true value;

[0018] The prediction module is used to take each sample in the dataset as a target sample, process the model input data of the target sample using an initial rock creep model, and generate strain prediction values. The initial rock creep model includes at least one of the following: a damaged elastic body strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. Each strain term includes a macro-micro nonlinear coupled damage term, which is generated based on initial micro-damage data and initial macro-damage data. The initial micro-damage data is determined based on the elastic modulus degradation of the intact rock sample after experiencing loading and unloading paths, and the initial macro-damage data is determined based on the strength or stiffness attenuation rate of the sample rock relative to the intact rock sample.

[0019] The fitting module is used to calculate the loss value based on each of the predicted strain values ​​and the corresponding true strain values ​​using a preset loss function, and to fit the initial rock creep model according to each of the loss values ​​to obtain the final rock creep model.

[0020] Fourthly, embodiments of this application also provide a rock failure mode determination device, the device comprising:

[0021] The acquisition module is used to acquire the actual stress data and actual creep deformation data of the target rock;

[0022] An initialization module is used to initialize the initial macroscopic damage data of the target rock to obtain the initial macroscopic damage data to be adjusted.

[0023] The generation module is used to process the initial macroscopic damage data to be adjusted and the actual stress data using the target rock creep model obtained by the construction method of the rock creep model as described in the first aspect, and generate theoretical creep deformation data.

[0024] The adjustment module is used to adjust the initial macroscopic damage data to be adjusted based on the theoretical creep deformation data and the actual creep deformation data to obtain the initial macroscopic damage prediction data.

[0025] The determination module is used to determine the failure mode of the target rock based on the initial macroscopic damage prediction data and the creep damage rate parameters in the creep model of the target rock; wherein the failure mode includes structurally controlled failure, stress-controlled failure and combined failure.

[0026] Fifthly, embodiments of this application also provide an electronic device, including: a processor; and a memory arranged to store computer-executable instructions, wherein the executable instructions, when executed, cause the processor to perform the steps of the above-described method for constructing a rock creep model or the above-described method for determining rock failure modes.

[0027] By means of the above technical solutions, the rock creep model construction method, apparatus, and rock failure mode determination method provided in the embodiments of this application, the rock creep model construction method first obtains a dataset consisting of multi-dimensional model input data such as creep holding time, stress data, initial macroscopic damage data, initial microscopic damage data, and long-term strength, and strain true values ​​as model output data. This provides comprehensive data support for model construction and fitting that closely matches the actual geological characteristics of deep rock masses, ensuring that the modeling data can reflect the correlation characteristics between multi-scale initial defects and core mechanical indicators of rock creep. Then, each sample in the dataset is used as a target sample, and the model input data of the target sample is processed by the initial rock creep model incorporating macroscopic and microscopic nonlinear coupling damage terms to generate strain prediction values. Damage terms can be combined with initial microscopic damage data and initial macroscopic damage data to generate macroscopic-microscopic nonlinear coupled damage. Each strain term in the initial model is incorporated into the above-mentioned macroscopic-microscopic nonlinear coupled damage, thereby breaking the existing model's fragmented treatment of macroscopic structural damage and unloading-induced microscopic stress damage. It realizes the nonlinear coupled characterization of macroscopic-microscopic damage at the strain term level, allowing the initial model to predict strain based on the mechanical characteristics of the combined action of multi-scale defects in deep rock masses. Subsequently, a loss value is calculated by combining each predicted strain value with the corresponding true strain value through a preset loss function. The initial model is then fitted and optimized based on the loss value. By using a data-driven approach, the deviations in the macroscopic-microscopic coupled damage characterization and strain prediction process of the model are corrected, so that the model's prediction results continuously approach the actual creep strain law of the rock. Ultimately, this embodiment yields a rock creep model that better reflects the geological environment of deep, high-stress, and strong structural surfaces. This model can realistically reflect the nonlinear mechanical response of deep rock masses under the combined action of multi-scale defects, accurately characterize the control effect of the initial damage level on rock creep behavior, effectively describe the instantaneous creep, deceleration creep, and acceleration creep behavior of rocks affected by initial macroscopic and microscopic damage, and accurately predict the rock creep rate, acceleration creep initiation time, and creep strain variation law.

[0028] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0029] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0030] Figure 1 A flowchart illustrating the method for constructing a rock creep model provided in an embodiment of this application is shown.

[0031] Figure 2 The following diagrams illustrate the nonlinear creep constitutive model provided in the embodiments of this application: (a) Nishihara model; (b) nonlinear creep damage model proposed in this application; (c) nonlinear creep damage model incorporating switching elements; (d) strain-time curves.

[0032] Figure 3 The following are comparative results of creep curve fitting for different initial macroscopic damage specimens provided in the embodiments of this application;

[0033] Figure 4 The following diagram illustrates the creep fitting effects of different models provided in the embodiments of this application: (a) Dm=0.08, σ1=40.11MPa; (b) Dm=0, σ1=48.88MPa; (c) Dm=0.27, σ1=35.61MPa; (d) Dm=0.44, σ1=25.87MPa (7th step), σ1=27.39MPa (8th step).

[0034] Figure 5 A schematic diagram of the triaxial creep fitting curve of a specimen with macroscopic initial damage provided in an embodiment of this application is shown; (a) D m =0.06, σ 3 = 10 MPa; (b) D m =0.23, σ 3 = 10 MPa;

[0035] Figure 6 The uniaxial creep curves and model prediction results of sandstone samples with different initial micro-damage levels provided in the embodiments of this application are shown; (a) fitting results of 0D1; (b) fitting results of 0D3 and 0D4 and prediction results of 0D2.

[0036] Figure 7 A flowchart illustrating the rock failure mode determination method provided in an embodiment of this application is shown.

[0037] Figure 8 A schematic diagram of the structure of the device for constructing a rock creep model provided in an embodiment of this application is shown;

[0038] Figure 9 A schematic diagram of the rock failure mode determination device provided in an embodiment of this application is shown;

[0039] Figure 10 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0041] 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.

[0042] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such use can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the term "comprising" and its variations should be interpreted as open-ended terms meaning "including but not limited to."

[0043] As mentioned earlier, existing creep models mainly include viscoelastic-viscoplastic component combinations such as Nishihara, Burgers, and Poynting-Thomson, as well as extended models based on empirical formulas or nonlinear viscous bodies. These models can describe instantaneous creep, decelerating creep, and accelerating creep behavior to varying degrees.

[0044] However, as mineral resource extraction and underground space development extend to deeper areas, the geological environment of the surrounding rock becomes increasingly complex, exhibiting characteristics of both high ground stress and strong structural surfaces (joints and fissures). Deep rock masses not only contain primary macroscopic geological defects such as joints and fissures, but also microscopic damage caused by excavation unloading or historical tectonic stress, such as microcracks and pores. The presence of these multi-scale initial defects makes the creep characteristics of rocks highly nonlinear, often leading to catastrophic processes in the surrounding rock from "decelerated deformation" to "sudden instability." Existing studies have shown that rock creep behavior is not only affected by the loading level but is also highly susceptible to the control of the degree of initial damage, especially under conditions of pre-existing macroscopic fissures or significant unloading damage, where the creep rate, accelerated creep initiation time, and long-term strength of the rock all change significantly. In this context, most existing rock creep models treat macroscopic structural damage and unloading-induced microscopic stress damage separately, making it difficult for the models to accurately reflect the nonlinear mechanical response of deep rock masses under the combined action of multi-scale defects. Based on this, this application proposes a method, apparatus for constructing a rock creep model, and a method for determining rock failure modes. The following is a detailed description of this application through specific embodiments.

[0045] To facilitate understanding of this embodiment, a method for constructing a rock creep model disclosed in this application will first be described in detail. The execution entity of the rock creep model construction method provided in this application is generally a computer device with certain computing capabilities. This computer device may include, for example, a terminal device, a server, or other processing devices. The terminal device may be a user equipment (UE), a mobile device, a user terminal, a terminal, a computing device, etc. In some possible implementations, the rock creep model construction method can be implemented by a processor calling computer-readable instructions stored in memory.

[0046] Figure 1 This paper illustrates a flowchart of a method for constructing a rock creep model according to an embodiment of this application. Figure 1 It can be seen that the embodiments of this application include at least steps S101-S103:

[0047] S101: Obtain the dataset; each sample in the dataset includes model input data and model output data, wherein the model input data includes creep holding time, stress data of the sample rock, initial macroscopic damage data and initial microscopic damage data of the sample rock, as well as preset long-term strength and preset creep initiation threshold, and the model output data is the strain true value;

[0048] S102: Each sample in the dataset is taken as a target sample, and the model input data of the target sample is processed using the initial rock creep model to generate strain prediction values; the initial rock creep model includes at least one of the following: damaged elastic body strain term, damaged viscoelastic body strain term, and damaged viscoplastic body strain term; each strain term includes: macro-micro nonlinear coupled damage term, which is generated based on initial micro-damage data and initial macro-damage data, wherein the initial micro-damage data is determined based on the elastic modulus degradation of the intact sample rock after experiencing the loading and unloading path, and the initial macro-damage data is determined based on the strength or stiffness attenuation rate of the sample rock relative to the intact sample rock;

[0049] S103: Using a preset loss function, calculate the loss value based on each of the predicted strain values ​​and the corresponding true strain values, and fit the initial rock creep model according to each of the loss values ​​to obtain the final rock creep model.

[0050] As can be seen, the method for constructing the rock creep model proposed in this application first obtains a dataset consisting of multi-dimensional model input data such as creep holding time, stress data, initial macroscopic damage data, initial microscopic damage data, and long-term strength, along with strain true values ​​as model output data. This provides comprehensive data support for model construction and fitting that closely matches the actual geological characteristics of deep rock masses, ensuring that the modeling data can reflect the correlation characteristics between multi-scale initial defects and core mechanical indicators of rock creep. Then, each sample in the dataset is used as a target sample, and the model input data of the target samples is processed using an initial rock creep model incorporating macroscopic and microscopic nonlinear coupled damage terms to generate strain prediction values. This macroscopic and microscopic nonlinear coupled damage term can combine initial microscopic damage data and... Initial macroscopic damage data generates macro- and micro-scale nonlinear coupled damage. Each strain term in the initial model incorporates this macro- and micro-scale nonlinear coupled damage, thereby breaking the existing model's fragmented treatment of macroscopic structural damage and unloading-induced micro-stress damage. This achieves nonlinear coupled characterization of macro- and micro-scale damage at the strain term level, enabling the initial model to predict strain based on the mechanical characteristics of the combined action of multi-scale defects in deep rock masses. Subsequently, a loss value is calculated by combining each predicted strain value with the corresponding true strain value using a preset loss function. The initial model is then fitted and optimized based on the loss value. By using a data-driven approach, deviations in the macro- and micro-scale coupled damage characterization and strain prediction process are corrected, ensuring that the model's prediction results consistently closely approximate the actual creep and strain characteristics of the rock. Ultimately, this embodiment yields a rock creep model that better reflects the geological environment of deep, high-stress, and strong structural surfaces. This model can realistically reflect the nonlinear mechanical response of deep rock masses under the combined action of multi-scale defects, accurately characterize the control effect of the initial damage level on rock creep behavior, effectively describe the instantaneous creep, deceleration creep, and acceleration creep behavior of rocks affected by initial macroscopic and microscopic damage, and accurately predict the rock creep rate, acceleration creep initiation time, and creep strain variation law.

[0051] The following provides a detailed explanation of S101-S103.

[0052] Regarding the above S101:

[0053] To construct the final rock creep model, the first step is to obtain a dataset for fitting the model. Specifically, the dataset includes model input data and model output data. The model input data includes creep holding time, stress data of the sample rock, initial macroscopic damage data of the sample rock, initial microscopic damage data, preset long-term strength, and preset creep initiation threshold. The model output data is the true strain value.

[0054] To accurately quantify the initial state of the rock, this application defines two damage variables: initial macroscopic damage variable and initial mesoscopic damage variable. The initial macroscopic damage variable characterizes the macroscopic fracture geometry of the rock. In practice, the initial macroscopic damage data is defined based on the strength or stiffness attenuation rate of a rock sample with pre-existing fractures relative to a intact rock sample, characterizing the weakening effect of fracture geometry on rock integrity. Specifically, for example, the uniaxial compressive strength of an intact rock sample can be measured through a uniaxial compression test. and the strength of rocks with fractures at different dip angles Introducing weighting coefficients Adjusting the degree of joint influence (e.g., to 1), the initial macroscopic damage data of the sample rock is calculated based on the strength reduction principle:

[0055] (A1)

[0056] For example, for sandstone samples with pre-fabricated fracture dip angles of 90°, 60°, 45°, 30°, and 0°, the strength was found to gradually decrease through testing, and thus the progressively increasing strength was calculated. The value quantifies the degree to which macroscopic defects weaken the integrity of the rock mass.

[0057] The initial microscopic damage variable is defined based on the stress history experienced by the rock (such as the degradation of elastic modulus caused by loading and unloading paths during excavation). This variable characterizes the degree of isotropic microcrack damage generated inside the material matrix by high ground stress environment or construction disturbance, reflecting the "internal injury" state of the rock material.

[0058] Specifically, the initial elastic modulus E0 of intact rock and the unloading modulus after experiencing a specific stress history can be measured through uniaxial compression and unloading tests. Based on the principle of elastic modulus degradation, the initial mesoscopic damage data are calculated as follows:

[0059] (A2)

[0060] The preset long-term strength is the maximum stress threshold at which a sample rock will not undergo macroscopic failure under long-term constant load. It is the critical stress for the rock to transition from steady-state creep to accelerated creep, and also the critical stress for macroscopic failure. When the applied stress is below this threshold, the rock can maintain steady-state creep without macroscopic failure. When the stress is above this threshold, the rock will enter the accelerated creep stage, internal cracks will rapidly penetrate, and macroscopic failure will eventually occur. In practice, for example, the isochronous curve method can be used, which involves conducting single-stage creep tests at different stress levels on multiple groups of rock samples, plotting stress-strain isochronous curves at different holding times, and taking the stress corresponding to the obvious inflection point of the curve as the preset long-term strength. Alternatively, the steady-state creep rate method can be used, which calculates the steady-state creep rate under different stresses and plots the stress-steady-state creep rate curve. When the steady-state creep rate increases sharply with stress, the corresponding stress is the preset long-term strength.

[0061] The preset creep initiation threshold is the lower limit of stress at which microcracks begin to stably propagate within a rock during creep. It serves as a marker for the initiation of micro-damage. When the applied stress is below this threshold, only initial microcracks exist within the rock, and no new cracks propagate; creep deformation is primarily elastic and viscoelastic. When the stress exceeds this threshold, microcracks begin to continuously initiate and propagate within the rock, entering the stage of micro-damage accumulation. In practice, this can be determined using acoustic emission monitoring. During creep tests, the acoustic emission signals of the sample rock are continuously monitored. When the event rate and energy rate of acoustic emission show significant abrupt changes, the corresponding stress is the creep initiation threshold. Alternatively, the strain rate mutation method can be used, where the stress corresponding to the transition of the creep strain rate from the deceleration stage to the steady-state stage is taken as the creep initiation threshold.

[0062] In practice, for example, sample rocks with different initial macroscopic and microscopic damage can be prepared first. At the same time, the long-term strength and creep initiation threshold of the intact sample rocks can be determined through relevant mechanical tests. Then, creep tests are carried out on sample rocks with different damage states. During the test, specific stress is applied to the sample and the creep holding time is continuously recorded. At the same time, the true creep strain value at each holding time point is collected and obtained using professional testing equipment. Then, the creep holding time, stress data applied to the sample rock, initial macroscopic damage data, and initial microscopic damage data of the sample rock corresponding to each holding time point, combined with the long-term strength and creep initiation threshold of the intact rock, are used as the model input data for a single sample. The true creep strain value collected at that time point is used as the corresponding model output data. Then, relevant test data at different time points, different stress levels, and different damage states in all creep tests are extracted in this way to complete the generation of all samples in the dataset.

[0063] Regarding S102-S103 above:

[0064] The rock creep model provided in this application is based on the improved Nishihara Model, and its parameters evolve in real time with the total macroscopic and microscopic damage variables. Specifically, the total macroscopic and microscopic damage variables are introduced into the parameters of the damaged elastomer, damaged viscoelastic body, and damaged viscoplastic body in the improved Nishihara Model. Figure 2 A schematic diagram of the nonlinear creep constitutive model provided in an embodiment of this application is shown, in conjunction with... Figure 2 The rock creep model proposed in this application is generated based on at least one of the following three parts:

[0065] (1) Damaged elastomer: describes instantaneous elastic deformation, whose elastic modulus is modified by macroscopic and microscopic nonlinear coupling damage. Figure 2 -d-Ⅰ);

[0066] (2) Damaged viscoelastic bodies: describe deceleration and steady-state creep, including generalized stress-switching elements. (Only when the stress data of the sample rock is available) Exceeding the preset creep initiation threshold σ cr (Activated at time), the viscosity coefficient deteriorates with the accumulation of damage ( Figure 2 -d-Ⅱ);

[0067] (3) Damaged viscoplastic body: describes accelerated creep, only when Triggered when the preset long-term strength is exceeded; its nonlinear viscosity coefficient decays exponentially over time, driving the generation of the accelerated creep stage. Figure 2 -d-Ⅲ).

[0068] In some embodiments, the rock creep model in this application includes a damaged elastomer strain term. In other embodiments, the rock creep model in this application includes a damaged elastomer strain term and a damaged viscoelastic body strain term. In still other embodiments, the rock creep model in this application includes a damaged elastomer strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. The damaged elastomer strain term, damaged viscoelastic body strain term, and damaged viscoplastic body strain term are obtained according to the constitutive equations of the damaged viscoelastic body, damaged viscoelastic body, and damaged viscoplastic body, respectively.

[0069] In some embodiments, if the stress data of the sample rock is less than the preset long-term strength, then the step of processing the model input data of the target sample using the initial rock creep model to generate strain prediction values ​​includes:

[0070] Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data;

[0071] Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer.

[0072] Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data.

[0073] If the stress data of the sample rock is less than or equal to the preset creep initiation threshold, then the strain data of the damaged elastomer is used as the strain prediction value.

[0074] If the stress data of the sample rock is greater than the preset creep initiation threshold, the strain prediction value is calculated based on the strain data of the damaged elastomer and the strain data of the damaged viscoelastic body.

[0075] In this embodiment, when the stress data of the sample rock is less than the preset long-term strength, the sample rock is in the non-accelerated creep stage. In this stage, if the stress data of the sample rock is less than or equal to the preset creep initiation threshold, the sample rock is in the low-stress stage; if the stress data of the sample rock is greater than the preset creep initiation threshold, the sample rock is in the medium-stress stage. In the low-stress stage, the strain prediction value is the strain data of the damaged elastomer; in the medium-stress stage, the strain prediction value is the sum of the strain data of the damaged elastomer and the strain data of the damaged viscoelastic body.

[0076] In this application, both the strain data of the damaged elastomer and the strain data of the damaged viscoelastic body are corrected by the total macroscopic and microscopic damage data. Therefore, in this embodiment, the total macroscopic and microscopic damage is first calculated based on the initial macroscopic damage data and the initial microscopic damage data of the target sample.

[0077] In some embodiments, if the sample rock is a single-fracture rock mass or an isotropic rock mass, the total macro- and micro-damage data are calculated based on the initial macro-damage data and initial micro-damage data of the target sample using the following formula:

[0078] (A3)

[0079] (A4)

[0080] in, Represents macro- and micro-level total damage data in scalar form. Represents the initial macroscopic damage data in scalar form. This indicates a detailed view of the total damage data. This represents the initial mesoscopic damage data in scalar form. This represents microscopic creep damage data, where A represents the creep damage rate parameter and t represents the creep holding time.

[0081] If the sample rock is a complex jointed rock mass containing multiple sets of joints, then the total macroscopic and microscopic damage data are calculated using the following formula based on the initial macroscopic damage data and initial microscopic damage data of the target sample:

[0082] (A5)

[0083] in, Represents macro- and micro-level total damage data in tensor form. Represents the identity matrix. Represents the initial macroscopic damage data in tensor form.

[0084] This embodiment provides two formulas for calculating the total macroscopic and microscopic damage data when the stress data applied to the sample rock is less than the preset long-term strength.

[0085] Specifically, embodiments of this application introduce a time-evolving mesoscopic creep damage variable. Based on existing research, Assuming it is over time t The cumulative negative exponential function form is as follows:

[0086] (A6)

[0087] In this context, parameter A is the acceleration factor α during the steady creep stage, and can be adjusted to a larger acceleration factor during the accelerated creep stage. β This variable describes the propagation process of microcracks over time under constant load. α and β are model parameters, determined through model fitting.

[0088] Taking into account both the initial mesoscopic damage and the mesoscopic creep damage, and assuming that the two are coupled in series at the mesoscopic scale, the total mesoscopic damage data is given by the above formula A4.

[0089] Next, based on set theory and the Lemaitre equivalent strain principle, the macro- and micro-damage nonlinear coupling equations for single-fracture or isotropic rock masses are derived. That is, the calculation formula for the total macro- and micro-damage variables is given by formula A3 above. Here, 0 ≤ ≤1.

[0090] This formula shows that the presence of macroscopic fractures nonlinearly amplifies the impact of microscopic damage on the overall mechanical properties of the rock. When D m When the damage is large, even a small increase in microscopic damage can lead to an increase in total damage. D w The significant increase is consistent with the physical mechanism of damage propagation induced by stress concentration at the macroscopic crack tip.

[0091] Specifically, during the low-stress stage, no new cracks initiate within the rock, and the damage remains constant, that is:

[0092] (A7)

[0093] This represents the initial macroscopic and mesoscopic total damage data at time t=0.

[0094] During the intermediate stress stage, the rock undergoes deceleration and steady-state creep, and the microscopic damage increases with time at a rate... Accumulation. At this point, a closer look at the total damage is needed. The total damage coupling formula is:

[0095] (A8)

[0096] During the high-stress stage, the rock enters accelerated creep, and damage increases over time at a rate... Rapid accumulation. At this point, a closer look at the total damage is necessary. The total damage coupling formula is:

[0097] (A9)

[0098] If the sample rock is a complex jointed rock mass, a second-order damage tensor can be constructed first using Kawamoto's geometric damage theory, based on the measured joint geometry information (occurrence, density, size). For example, assuming the sample rock contains... M Group advantage joint, for the first k The joints have a bulk density of The average diameter is The normal vector is The initial macroscopic damage data in tensor form is calculated as follows:

[0099] (A10)

[0100] This tensor It comprehensively includes anisotropic information of the macroscopic structure of the rock mass (the principal value represents the degree of damage, and the principal direction represents the dominant direction of damage).

[0101] Maintain detailed damage D t As an isotropic scalar (assuming matrix damage is non-directional), it is compared with the macroscopic tensor. By coupling the macroscopic and microscopic damage, based on Lemaitre's effective stress principle, the nonlinear coupling equation for macroscopic and microscopic damage when the sample rock is a single-fracture or isotropic rock mass can be derived. That is, the formula for calculating the total macroscopic and microscopic damage data is Equation A5 above. This formula achieves physical unification and nonlinear superposition of damage at different scales and of different types. The formula shows that the total damage is a tensor that evolves over time, with each component evolving at a different rate. This means that even in an isotropic external stress field, the rock mass will still experience damage due to… The presence of this leads to anisotropic creep deformation; and this anisotropic deformation increases with microscopic damage. D t The increase (i.e., the passage of time) further amplifies this. The nonlinear creep model uses the total damage tensor... Introduced into the elasticity matrix and viscosity coefficient tensor, the model is able to describe the anisotropic creep deformation of complex jointed rock masses under constant load.

[0102] After obtaining the total damage data at both macroscopic and microscopic levels, specifically in some embodiments, the predicted strain value can be calculated according to the following formula:

[0103] (A11)

[0104] (A12)

[0105] in, Indicates the predicted strain value. This represents the strain data of the damaged elastomer. This represents strain data of a damaged viscoelastic body. The stress data of the sample rock (if the sample rock is a single-fracture rock mass or an isotropic rock mass, then it is uniaxial stress), E1 represents the elastic modulus of the viscoelastic body, and η1 represents the viscoelastic viscosity coefficient.

[0106] The derivation process of the above strain prediction calculation formula A11 is explained below.

[0107] For damaged elastomers, the constitutive equation is the above formula A12. For the The corrected elastic modulus.

[0108] For damaged viscoelastic materials, the constitutive equation is:

[0109] (A13)

[0110] in, For viscoelastic strain rate, For viscoelastic strain data, For the Corrected viscoelastic modulus express The corrected viscoelastic viscosity coefficient.

[0111] After obtaining the constitutive equations for damaged elastomers and damaged viscoelastics, the above strain prediction calculation formula A11 can be derived using the Laplace transform and inverse transform.

[0112] In other embodiments, if the rock mass in geotechnical engineering is under triaxial stress, the one-dimensional creep damage model can be extended to a three-dimensional model based on elastoplastic theory. That is, when the sample rock mass is under triaxial stress, the strain prediction value can be calculated according to the following formula:

[0113] (A14)

[0114] (A15)

[0115] in, For the maximum principal stress, For the minimum principal stress, The shear modulus of the damaged elastomer. Bulk modulus This refers to the shear modulus of the damaged viscoelastic body.

[0116] The derivation process of the strain prediction calculation formula A14 is explained below. Based on elastoplastic theory and combined with the generalized Hook's law, the total strain under triaxial compression is... It can be represented as:

[0117] (A16)

[0118] in , , The strain contribution from the elastic element, viscoelastic element, and viscoplastic element models is not considered when the stress level does not exceed the long-term strength.

[0119] (1) Damage to elastomers:

[0120] (A17-1)

[0121] in, For the deviatoric stress tensor, It is the Kronecker function. It is the spherical stress tensor. According to the spherical stress tensor... and deviatoric stress tensor The stress tensor can be obtained. :

[0122] (A17-2)

[0123] in, This represents the intermediate principal stress.

[0124] (2) Damage to viscoelastic materials:

[0125] (A17-3)

[0126] After obtaining the three-dimensional constitutive equations for the damaged elastomer and the damaged viscoelastic body, the above-mentioned formula A14 for calculating the strain prediction value can be derived.

[0127] This embodiment targets single-fracture or isotropic rock masses. It employs scalar initial macroscopic damage data and total mesoscopic damage data, calculating the total macroscopic and mesoscopic damage data through a coupling formula. The total mesoscopic damage data is generated based on initial mesoscopic damage and creep damage. The coupling formula subtracts the overlapping portion of the two to avoid double calculation, and then converts it into an exponential form to accurately fit the cumulative law of damage with load holding time. This is then combined with the initial macroscopic damage to obtain the total damage. For complex jointed rock masses containing multiple joints, considering the anisotropy of macroscopic damage, a tensor form is used to represent the initial macroscopic damage and total damage. A matrix operation formula is used to calculate the total macroscopic and mesoscopic damage data to effectively characterize the collaborative damage behavior among multiple joints. Ultimately, this embodiment yields total macroscopic and mesoscopic damage data that is adaptable to different rock mass types and more closely reflects the actual damage state, providing more accurate damage parameter support for subsequent creep constitutive model establishment and long-term mechanical property evaluation of the rock mass.

[0128] In some embodiments, if the stress data of the sample rock is greater than or equal to the preset long-term strength, then the step of processing the model input data of the target sample using the initial rock creep model to generate strain prediction values ​​includes:

[0129] Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data;

[0130] Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer.

[0131] Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data.

[0132] Using the aforementioned damaged viscoplastic strain term, the damaged viscoplastic strain data are calculated based on the stress data of the sample rock, the preset long-term strength, the creep holding time, and the total macroscopic and microscopic damage data.

[0133] The strain prediction value is calculated based on the strain data of the damaged elastomer, the strain data of the damaged viscoelastic body, and the strain data of the damaged viscoplastic body.

[0134] In this embodiment, if the stress data of the sample rock is greater than or equal to the preset long-term strength, the sample rock is in a high-stress stage, and the strain prediction value is the sum of the strain data of the damaged elastomer, the strain data of the damaged viscoelastic body, and the strain data of the damaged viscoplastic body. Specifically, in some embodiments, the strain prediction value can be calculated according to the following formula:

[0135] (A18)

[0136] (A19)

[0137] in, This represents strain data in damaged viscoplastic volumes. To preset long-term strength, It is the viscoplastic viscosity coefficient.

[0138] The derivation of the strain prediction calculation formula A18 is explained below. For the damaged viscoplastic body in the model, its constitutive equation is:

[0139] (A20)

[0140] in, It represents the viscoplastic strain rate.

[0141] In existing technologies, classic linear element models (such as the Nishihara model and the Burgers model) have constant parameters and cannot describe the accelerated creep stage at all; existing nonlinear models are often implemented by stacking empirical formulas (such as exponential laws and power laws), lacking a clear physical mechanism explanation and failing to reflect the essential process of the gradual deterioration of rock material properties over time. However, in formula A20 provided in the embodiments of this application, the viscoplastic strain rate and... Proportional to each other, over time The increase approaches 1, leading to a sharp increase in strain rate, thus physically reproducing the nonlinear accelerated creep failure process of rocks.

[0142] After obtaining the three-dimensional constitutive equations for damaged elastomers, damaged viscoelastics, and damaged viscoplastics, the above-mentioned formula A18 for calculating strain prediction values ​​can be derived.

[0143] In other embodiments, if the rock mass in geotechnical engineering is under triaxial stress, the one-dimensional creep damage model can be extended to a three-dimensional model based on elastoplastic theory. That is, when the sample rock mass is under triaxial stress, the strain prediction value can be calculated according to the following formula:

[0144] (A21)

[0145] (A22)

[0146] The derivation process of the strain prediction calculation formula A21 is explained below. For the damaged viscoplastic body in the model, its three-dimensional constitutive equation is:

[0147] (A23-1)

[0148] (A23-2)

[0149] Where F represents the yield function of the rock, This represents the initial value of the rock yield function. During implementation, to simplify the calculation process, the sensitivity coefficient of the viscoplastic rate is used. N The value can be 1.

[0150] Based on formulas A17-1, A17-2, A17-3, A23-1, and A23-2, the formula A21 for calculating the strain prediction value can be derived.

[0151] In this embodiment, the rock creep model satisfies the following conditions:

[0152] (A24)

[0153] This indicates the stress experienced by the damaged elastomer. This indicates the stress experienced by a damaged viscoelastic body. This represents the stress experienced by the damaged plastic body. Each damaged element will satisfy the following relationship:

[0154] (A25)

[0155] Indicates The corrected initial elastic modulus, :through Corrected elastic modulus of damaged viscoelastic body :through Corrected viscosity coefficient of damaged viscoelastic body :through The corrected viscosity coefficient of the damaged viscoplastic body.

[0156] Studies have found that traditional rock creep models typically neglect the weakening effect of initial damage on the long-term load-bearing capacity of rocks. In fact, the higher the degree of initial damage, the weaker the rock's ability to resist long-term loads. Based on this, in some embodiments of this application, the preset long-term strength is generated according to the following method:

[0157] Multiple data sets are acquired, each data set including independent variable values ​​and dependent variable values. The independent variable values ​​are the initial macroscopic and microscopic total loss data of the sandstone sample, and the dependent variable values ​​are the long-term strength of the sandstone sample.

[0158] By fitting the multiple data sets, a long-term intensity dynamic weakening function is obtained;

[0159] The initial values ​​in the macroscopic and microscopic total damage data are processed using the long-term intensity dynamic weakening function to obtain the preset long-term intensity.

[0160] In this embodiment, a long-term strength dynamic weakening function considering initial damage is first constructed. In practice, the long-term strength can be established based on experimental data. Compared with initial macroscopic and mesoscopic total damage data Functional relationship between :

[0161] (A26)

[0162] in, σ s0 For the long-term strength of the intact sample rock, g ( D ini ) is a decay function (such as a linear decreasing function). This formula reflects the controlling effect of rock mass deterioration on its long-term bearing capacity threshold, that is, the more fractured the rock mass, the lower its threshold value for resisting creep.

[0163] For the initial macroscopic and mesoscopic total damage data, we can set t=0 and substitute it into the above macroscopic and mesoscopic damage nonlinear coupling equation to obtain the initial macroscopic and mesoscopic total damage data containing only the initial defect:

[0164] (A27)

[0165] In practice, it can be based on different degrees of damage (different D ini Analysis of graded loading creep test data for sandstone specimens; determination of long-term strength of each specimen using the isochronous curve method or the transition creep method. σ s This results in multiple data sets. Among them, graded loading is a multi-stage incremental mechanical test loading method, which means that instead of applying a constant stress until the specimen fails all at once, the stress is divided into multiple levels. Each level is kept at a constant load for a period of time (to allow the specimen to produce stable creep). After the creep of that level stabilizes, the stress is increased to a higher level, and this process is repeated until the specimen fails or reaches the preset stress limit.

[0166] Next, the data sets are fitted. Specific fitting methods can be found using existing methods, which will not be elaborated upon here. The fitting yields the long-term strength. σ s Compared with the initial total damage D ini They exhibit an approximately linear negative correlation:

[0167] (A28)

[0168] in k Here, represents the material fitting constant. With initial total damage The monotonically decreasing trend of the initial structural defects indicates the weakening effect of the rock's long-term bearing capacity.

[0169] After obtaining the constructed long-term strength dynamic weakening function, the initial values ​​in the macroscopic and microscopic total damage data of the sample rock are processed using the long-term strength dynamic weakening function to obtain the preset long-term strength.

[0170] By using this dynamic preset long-term strength as a dynamic threshold to determine whether accelerated creep has begun, "negative feedback" control of structural damage on the stress threshold is achieved.

[0171] Accordingly, the strain data used in the aforementioned calculation of damaged viscoplastic bodies... Specifically, it can be For example, formula A19 can be transformed into:

[0172] (A29)

[0173] This embodiment considers the weakening effect of initial damage on the long-term bearing capacity of rock and constructs a dynamic weakening function for long-term strength. This function is used in conjunction with the initial values ​​of macroscopic and microscopic total damage data to generate a preset long-term strength, making the obtained long-term strength more consistent with the actual long-term mechanical property changes of the material and the long-term strength value more accurate. This improves the accuracy of subsequent predictions of the long-term stability and failure time of deep fractured rock masses.

[0174] In some embodiments, the method further includes:

[0175] Acquire the stress data, initial macroscopic damage data and initial microscopic damage data of the rock to be predicted, and preset the long-term strength;

[0176] Using the final rock creep model, based on the stress data, initial macroscopic damage data and initial microscopic damage data of the rock to be predicted, strain prediction data of the rock to be predicted is generated by presetting the long-term strength.

[0177] This embodiment acquires stress data, initial macroscopic damage data, initial microscopic damage data, and long-term strength of the rock to be predicted; then, it integrates and calculates these data using a well-fitted final rock creep model, giving full play to the model's ability to characterize the coupling effect of macroscopic and microscopic damage and the entire creep process, and finally obtains strain prediction data that is more in line with the actual creep law.

[0178] This application also utilizes the aforementioned rock creep model to study different... D m (Controlled by different fracture dip angles) and different D s The creep test curves of sandstone under graded loading (induced by different unloading degrees) were fitted and predicted.

[0179] (1) Validation of the one-dimensional model under different macroscopic initial damage:

[0180] Sandstone samples containing pre-existing fractures at different dip angles were selected for uniaxial graded loading creep tests. Based on the macroscopic damage definition method described in this application, the initial macroscopic damage values ​​of the selected samples were determined. D m The values ​​were 0 (intact sample), 0.08, 0.27, 0.34, 0.44, and 0.50, respectively.

[0181] The model parameters obtained by least squares inversion (including elastic modulus E1, viscosity coefficient) η 1, η 2. Creep rate parameters α , β (etc.) See Table 1 below for details.

[0182] Figure 3 The accompanying diagram shows the fitting comparison results of creep curves for different initial macroscopic damage specimens provided in the embodiments of this application. The black dashed line represents the fitting result of the model in this application, and the colored data represents the experimental results. According to... Figure 3 It can be seen that the model in this application can accurately describe the deformation characteristics of the entire process, including deceleration, steady state and accelerated creep.

[0183] Table 1. Creep parameters obtained by fitting the model to experimental data.

[0184]

[0185] The nonlinear creep model (scalar coupling mode) constructed in this application was used to fit the creep curves of the above-mentioned samples throughout the entire process, with particular emphasis on analyzing the data entering the accelerated creep stage. D mThe values ​​were 0, 0.08, 0.27, and 0.44, respectively. Meanwhile, the classic Nishihara Model was introduced as a control group for comparison, and it was found that the model in this application has better creep simulation performance. Figure 4 The different models provided in the embodiments of this application demonstrate the accelerated creep fitting effect. See also Figure 4 As shown, in high damage ( D m Under the condition of 0.44, the model of this application can accurately capture the nonlinear characteristics of a sharp increase in strain rate, and the goodness of fit is significantly higher than that of the Nishihara model, which cannot describe accelerated creep.

[0186] (2) Validation of the three-dimensional model under different macroscopic initial damage:

[0187] To verify the applicability of the rock creep model in this application under triaxial stress, triaxial creep test data of the specimens under a confining pressure of 10 MPa were selected. The test data are from the paper "Creep constitutive model considering nonlinear creep degradation of fractured rock". Initial macroscopic damage of the specimens. D m The values ​​are 0.06 and 0.23, respectively. The experimental data were fitted and analyzed using the three-dimensional creep constitutive equation from this application to identify the triaxial creep model parameters (including shear modulus). G 0. Bulk Modulus K (e.g., 0) See Table 2 below for details.

[0188] Table 2. Comparison of triaxial creep curve fitting results for specimens with different initial macroscopic damage.

[0189]

[0190] Figure 5 A schematic diagram of the triaxial creep fitting curve of a specimen with macroscopic initial damage provided in an embodiment of this application is shown. See also Figure 5 As shown, the prediction curve of the model in this application is in high agreement with the experimental data points. This demonstrates that by introducing a macroscopic damage scalar... D m The method of modifying triaxial constitutive relations can effectively describe the nonlinear creep behavior of deep rock masses under confining pressure.

[0191] (3) Model validation under different initial microscopic damage:

[0192] Sandstone samples with initial mesoscopic damage were selected after undergoing different unloading paths. Based on the definition of mesoscopic damage (based on elastic modulus degradation) described in this application, their corresponding initial mesoscopic damage was determined. Ds The values ​​are 0.036 (experimental data for 0D1), 0.227 (experimental data for 0D2), 0.383 (experimental data for 0D3), and 0.505 (experimental data for 0D4), respectively. Here, 0D1, 0D2, 0D3, and 0D4 describe sandstone specimens with different initial stress-induced microstructural damage under a confining pressure of 0 MPa. The source of each experimental data is the paper "Research on the Time-dependent Deformation Damage Mechanism and Control Strategies of Soft Rock Tunnel Surrounding Rock Considering Initial Damage Effect".

[0193] A rigorous verification strategy of "partial fitting-partial prediction" was adopted. Inversion model parameters were retrieved using experimental data from 0D1, 0D3, and 0D4 to summarize the evolution of parameters with damage. Based on this pattern, the creep behavior of the 0D2 experimental data specimen was predicted, and the predicted curve was compared with the measured data from the 0D2 experimental data. Model parameters under different microscopic damage states are detailed in Table 3 below.

[0194] Table 3. Results of fitting parameters for uniaxial creep curves of specimens with different initial microscopic damage.

[0195]

[0196] Figure 6 The uniaxial creep curves and model prediction results of sandstone samples with different initial micro-damage levels provided in the embodiments of this application are shown. Figure 6 The results show that the model in this application can not only fit known data, but also accurately predict the creep deformation trend under unknown microscopic damage conditions. This proves that the introduction of initial microscopic damage variables in the model is effective. D s The physical necessity of the model is demonstrated, and the model has good generalization ability for historical stress damage.

[0197] The above verification process shows that:

[0198] 1. The rock creep model proposed in this application can accurately describe the deformation characteristics of the deceleration creep and steady-state creep stages under low stress levels;

[0199] 2. By introducing σ s ( D ini The weakening function of this application model accurately predicts the phenomenon of accelerated failure of highly damaged specimens at lower stress levels, which solves the problem that traditional fixed-value models cannot distinguish the failure time of rocks with different damage.

[0200] 3. The failure time calculated by the model in this application is in high agreement with the experimental results, which verifies the effectiveness and reliability of the scalar coupling mechanism and constitutive equation.

[0201] The above embodiments are based on uniaxial and triaxial graded loading creep tests on sandstone containing pre-fabricated fractures, illustrating the specific implementation process of initial macro- and micro-damage quantification, nonlinear coupling equation construction, and model parameter identification in this application. It should be noted that these embodiments do not limit the application scope of this application. The macro- and micro-damage coupling mechanism and nonlinear creep constitutive framework proposed in this application are universal, applicable not only to sandstone but also to other brittle rocks such as granite and marble; not only to simplified working conditions with pre-fabricated single fractures, but also, through tensile expansion, to deep rock engineering with complex fracture networks. Engineers can adjust and apply the methods of this application according to the geological characteristics (such as joint occurrence and geostress level) and mechanical parameters of the actual rock mass on site to achieve scientific prediction of the long-term stability of deep engineering projects.

[0202] There are two different mechanisms for the time-dependent failure of deep rock masses: one is structure-controlled failure, where failure mainly occurs along macroscopic dominant structural planes, with relatively little influence from stress levels, manifesting as block instability; the other is stress-controlled failure, where failure is mainly caused by the propagation and penetration of microcracks within the rock matrix driven by high stress, often accompanied by brittle failure characteristics such as rockbursts and plate cracking. Research has found that existing models lack a unified parameter system to quantify this competitive relationship, making it impossible to quantitatively identify the dominant failure mode of the rock mass. This leads to a lack of targeted support design and difficulty in achieving precise control. For example, blindly increasing the thickness of the shotcrete layer in structure-controlled roadways without strengthening the anchoring of dominant joints results in material waste and poor effectiveness; or, in stress-controlled roadways, incorrectly implementing large-scale decompression can damage the self-supporting capacity of the surrounding rock and induce new instability.

[0203] Based on this, this application also provides a method for determining rock failure modes. Figure 7 This paper illustrates a flowchart of a rock failure mode determination method provided in an embodiment of this application. Figure 7 It can be seen that the embodiments of this application include at least steps S701-S705:

[0204] S701: Obtain the actual stress data and actual creep deformation data of the target rock;

[0205] S702: Initialize the initial macroscopic damage data of the target rock to obtain the initial macroscopic damage data to be adjusted;

[0206] S703: Using the target rock creep model obtained by the construction method of the rock creep model provided in any of the foregoing embodiments, the initial macroscopic damage data to be adjusted and the actual stress data are processed to generate theoretical creep deformation data;

[0207] S704: Based on the theoretical creep deformation data and the actual creep deformation data, the initial macroscopic damage data to be adjusted is adjusted to obtain the initial macroscopic damage prediction data;

[0208] S705: Based on the initial macroscopic damage prediction data and the creep damage rate parameters in the target rock creep model, determine the failure mode of the target rock; wherein the failure mode includes structurally controlled failure, stress-controlled failure, and combined failure.

[0209] In this embodiment, to determine the failure mode of the target rock, for example, the actual triaxial stress data it bears is first measured using a geostress testing device, including key indicators such as the maximum principal stress and the minimum principal stress. Then, monitoring devices such as a total station and strain sensors are used to continuously record data such as rock mass creep settlement and convergence at different time points, and the data are compiled into complete actual creep deformation data.

[0210] Next, the initial macroscopic damage data of the target rock is initialized to obtain the initial macroscopic damage data to be adjusted. For example, for single-fracture or homogenized rock masses, test data of similar rock masses can be referenced to set an initial macroscopic damage value in scalar form; for complex jointed rock masses containing multiple sets of joints, the principal value, principal direction and other parameters of the initial macroscopic damage tensor can be inferred based on the approximate distribution of the joints to obtain the initial macroscopic damage data to be adjusted.

[0211] Subsequently, the initial macroscopic damage data to be adjusted and the actual stress data are substituted into the target rock creep model constructed through the corresponding embodiment. The model performs calculations through core modules such as macroscopic and microscopic damage coupling equations and long-term strength weakening functions, outputting theoretical creep deformation at different time points, and integrating them to form theoretical creep deformation data. Finally, the error between the theoretical creep deformation data and the actual creep deformation data at each corresponding time point is calculated. If the error exceeds the preset accuracy range, the numerical value or tensor parameters of the initial macroscopic damage data to be adjusted are adjusted according to the error situation. The adjusted data is then substituted back into the model to generate new theoretical creep deformation data. This adjustment and calculation process is iterated repeatedly until the error between the theoretical and actual creep deformation data meets the requirements. The adjusted data obtained at this time is the initial macroscopic damage prediction data.

[0212] After obtaining the initial macroscopic damage prediction data, the failure mode of the target rock can be determined based on the initial macroscopic damage prediction data and the creep damage rate parameters in the target rock creep model. Here, the failure modes of the target rock include structurally controlled failure, stress-controlled failure, and combined failure. In specific implementation, if the initial macroscopic damage prediction data is large and the creep damage rate parameters in the target rock creep model are small or medium, the failure mode of the target rock is determined to be structurally controlled failure; if the initial macroscopic damage prediction data is small and the creep damage rate parameters in the target rock creep model are extremely large, the failure mode of the target rock is determined to be stress-controlled failure; if both the initial macroscopic damage prediction data and the creep damage rate parameters in the target rock creep model are in the intermediate transition zone, the failure mode of the target rock is determined to be combined failure.

[0213] Here, the two types of thresholds compared with the initial macroscopic damage estimation data and creep damage rate parameters can be set according to the actual situation, and this application embodiment does not limit this. In specific implementation, the two types of thresholds can be obtained by conducting statistical analysis on rock masses under similar geological conditions in the target engineering area; for example, damage inversion data from historical rock mass instability cases can be extracted, and cluster analysis (such as K-means algorithm) or statistical distribution laws can be used to divide the parameter boundary intervals characterizing different failure modes, which can be used as the comparison thresholds in this application.

[0214] It should be noted that for structurally controlled failure, the core triggering premise is large initial macroscopic damage. This characteristic directly leads to a significant reduction in the long-term strength threshold of the rock, allowing the rock mass to fail under relatively low stress. The dominant mechanism of this type of failure is the slippage of existing weak surfaces in the rock mass. This type of failure develops gradually, rather than being a sudden damage in a short period of time, and does not require a rapid damage accumulation process. Therefore, reflected in the creep model of the target rock, the creep damage rate parameter, which characterizes the speed of damage development, naturally exhibits a small or moderate characteristic. Thus, "large initial macroscopic damage inference data and small or moderate creep damage rate parameter" is used as the criterion for identifying structurally controlled failure. For stress-controlled failure, small initial macroscopic damage (relatively intact rock mass) is a fundamental characteristic. This results in a high long-term strength threshold for the rock, allowing it to withstand extremely high stress. When the stress exceeds this high threshold, sudden damage occurs within the rock mass, characterized by instantaneous energy release and a rapid increase in microcracks. This type of sudden damage requires rapid accumulation of damage within a short period. In the target rock creep model, this is reflected in the extremely large creep damage rate parameter. Therefore, "small initial macroscopic damage prediction data and extremely large creep damage rate parameter" is used as the criterion for identifying stress-controlled failure.

[0215] During implementation, engineering support strategies can be automatically generated based on the failure mode. For example, if the failure mode is structurally controlled failure, the output engineering support strategy can be: prioritize "anchoring" and "stitching" methods, using high-strength anchor bolts or high-prestressed anchor cables to penetrate the dominant fracture surface, increasing the normal stress of the joint surface to improve its shear strength, and suppressing block slippage instability and disturbance deformation between joint surfaces. If the failure mode is stress-controlled failure, the engineering support strategy can be: prioritize "confining pressure reinforcement" or active "pressure relief" measures. On the one hand, a combined support method of high-strength shotcrete, steel arch frame, or high-strength closely spaced steel arch frame with high-prestressed anchor cable mesh can be used to provide high confining pressure and strong radial support pressure, limiting the opening and expansion of microcracks inside the rock matrix. On the other hand, borehole pressure relief can be implemented in stress concentration areas to actively release stress energy and reduce the level of deviatoric stress.

[0216] This embodiment first acquires the actual stress and creep deformation data of the target rock; then initializes the initial macroscopic damage data to obtain the data to be adjusted, providing a reasonable starting point for subsequent iterative optimization; subsequently, using the constructed target rock creep model, theoretical creep deformation data is generated by combining the initial macroscopic damage data to be adjusted with the actual stress data; based on the theoretical and actual creep deformation data, the initial macroscopic damage data to be adjusted is adjusted to eliminate initial assumption bias, making the obtained initial macroscopic damage prediction data fit the actual damage state of the rock mass; finally, based on the initial macroscopic damage prediction data of the target rock and the creep damage rate parameters in the target rock creep model, a more accurate rock failure mode determination result that fits the actual working conditions can be obtained, providing a targeted basis for engineering support design and avoiding material waste and instability risks caused by blind support.

[0217] Those skilled in the art will understand that in the above-described method of the specific embodiments, the order in which the steps are written does not imply a strict execution order, but constitutes no limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0218] It should be noted that in practical applications, all the above-described possible implementation methods can be combined in any way to form possible embodiments of this application, and will not be described in detail here. The information (including but not limited to device information, user information, etc.) and data (including but not limited to data used for analysis, storage, and display) involved in this application are all information and data authorized by the user or fully authorized by all parties. The software tools or components appearing in the embodiments of this application are merely illustrative examples and do not represent actual use.

[0219] Based on the same concept, this application also provides a device for constructing a rock creep model, which corresponds one-to-one with the method for constructing the rock creep model in the above embodiments. Figure 8 A schematic diagram of the structure of the device for constructing a rock creep model provided in an embodiment of this application is shown. See also Figure 8 As shown, the rock creep model construction apparatus 800 provided in this application embodiment includes:

[0220] The acquisition module 801 is used to acquire a dataset; each sample in the dataset includes model input data and model output data, wherein the model input data includes creep holding time, stress data of sample rock, initial macroscopic damage data and initial microscopic damage data of sample rock, as well as preset long-term strength and preset creep initiation threshold, and the model output data is the strain true value;

[0221] The prediction module 802 is used to take each sample in the dataset as a target sample, process the model input data of the target sample using an initial rock creep model, and generate strain prediction values. The initial rock creep model includes at least one of the following: a damaged elastic body strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. Each strain term includes a macro-micro nonlinear coupled damage term, which is generated based on initial micro-damage data and initial macro-damage data. The initial micro-damage data is determined based on the elastic modulus degradation of the intact sample rock after experiencing loading and unloading paths, and the initial macro-damage data is determined based on the strength or stiffness attenuation rate of the sample rock relative to the intact sample rock.

[0222] The fitting module 803 is used to calculate the loss value based on each of the predicted strain values ​​and the corresponding true strain values ​​using a preset loss function, and to fit the initial rock creep model according to each of the loss values ​​to obtain the final rock creep model.

[0223] In some embodiments, in the above-described apparatus, if the stress data of the sample rock is less than a preset long-term strength, the prediction module is used to:

[0224] Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data;

[0225] Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer.

[0226] Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data.

[0227] If the stress data of the sample rock is less than or equal to the preset creep initiation threshold, then the strain data of the damaged elastomer is used as the strain prediction value.

[0228] If the stress data of the sample rock is greater than the preset creep initiation threshold, the strain prediction value is calculated based on the strain data of the damaged elastomer and the strain data of the damaged viscoelastic body.

[0229] In some embodiments, in the above-described apparatus, if the stress data of the sample rock is greater than or equal to the preset long-term strength, the prediction module is used to:

[0230] Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data;

[0231] Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer.

[0232] Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data.

[0233] Using the aforementioned damaged viscoplastic strain term, the damaged viscoplastic strain data are calculated based on the stress data of the sample rock, the preset long-term strength, the creep holding time, and the total macroscopic and microscopic damage data.

[0234] The strain prediction value is calculated based on the strain data of the damaged elastomer, the strain data of the damaged viscoelastic body, and the strain data of the damaged viscoplastic body.

[0235] In some embodiments, in the above-described apparatus, if the sample rock is a single-fracture rock mass or an isotropic rock mass, the prediction module 802, when calculating the total macroscopic and microscopic damage data based on the initial macroscopic damage data and initial microscopic damage data of the target sample, specifically uses the following formula to calculate the total macroscopic and microscopic damage data:

[0236]

[0237]

[0238] in, Represents macro- and micro-level total damage data in scalar form. Represents the initial macroscopic damage data in scalar form. This indicates a detailed view of the total damage data. This represents the initial mesoscopic damage data in scalar form. This represents microscopic creep damage data, where A represents the creep damage rate parameter and t represents the creep holding time.

[0239] If the sample rock is a complex jointed rock mass containing multiple sets of joints, then the total macroscopic and microscopic damage data are calculated using the following formula based on the initial macroscopic damage data and initial microscopic damage data of the target sample:

[0240]

[0241] in, Represents macro- and micro-level total damage data in tensor form. Represents the identity matrix. Represents the initial macroscopic damage data in tensor form.

[0242] In some embodiments, the apparatus further includes a computing module for:

[0243] Multiple data sets are acquired, each data set including independent variable values ​​and dependent variable values. The independent variable values ​​are the initial macroscopic and microscopic total loss data of the sandstone sample, and the dependent variable values ​​are the long-term strength of the sandstone sample.

[0244] By fitting the multiple data sets, a long-term intensity dynamic weakening function is obtained;

[0245] The initial values ​​in the macroscopic and microscopic total damage data are processed using the long-term intensity dynamic weakening function to obtain the preset long-term intensity.

[0246] In some embodiments, the apparatus further includes an application module for:

[0247] Acquire the stress data, initial macroscopic damage data and initial microscopic damage data of the rock to be predicted, and preset the long-term strength;

[0248] Using the final rock creep model, based on the stress data, initial macroscopic damage data and initial microscopic damage data of the rock to be predicted, strain prediction data of the rock to be predicted is generated by presetting the long-term strength.

[0249] This application provides a device for constructing a rock creep model. First, a dataset is acquired, consisting of multi-dimensional model input data (creep holding time, stress data, initial macroscopic damage data, initial microscopic damage data, long-term strength, etc.) and strain true values ​​as model output data. This dataset provides comprehensive data support for model construction and fitting, closely reflecting the actual geological characteristics of deep rock masses. It ensures that the modeling data reflects the correlation between multi-scale initial defects and core mechanical indicators of rock creep. Then, each sample in the dataset is used as a target sample. An initial rock creep model incorporating macroscopic-microscopic nonlinear coupled damage terms is used to process the model input data of the target samples to generate strain prediction values. These macroscopic-microscopic nonlinear coupled damage terms can combine initial microscopic damage data and initial macroscopic damage data. The model generates macro- and micro-scale nonlinear coupled damage from observed damage data. Each strain term in the initial model incorporates this macro- and micro-scale nonlinear coupled damage, thereby breaking the existing model's fragmented treatment of macro-structural damage and unloading-induced micro-stress damage. It achieves nonlinear coupled characterization of macro- and micro-scale damage at the strain term level, enabling the initial model to predict strain based on the mechanical characteristics of the combined action of multi-scale defects in deep rock masses. Subsequently, a loss value is calculated by combining each predicted strain value with the corresponding true strain value using a preset loss function. The initial model is then fitted and optimized based on the loss value. By using a data-driven approach, the model corrects deviations in the macro- and micro-scale coupled damage characterization and strain prediction process, ensuring that the model's prediction results consistently closely approximate the actual creep and strain laws of rocks. Ultimately, this embodiment yields a rock creep model that better reflects the geological environment of deep, high-stress, and strong structural surfaces. This model can realistically reflect the nonlinear mechanical response of deep rock masses under the combined action of multi-scale defects, accurately characterize the control effect of the initial damage level on rock creep behavior, effectively describe the instantaneous creep, deceleration creep, and acceleration creep behavior of rocks affected by initial macroscopic and microscopic damage, and accurately predict the rock creep rate, acceleration creep initiation time, and creep strain variation law.

[0250] Specific limitations regarding the apparatus for constructing the rock creep model can be found in the limitations on the construction method of the rock creep model described above, and will not be repeated here. Each module in the aforementioned apparatus for constructing the rock creep model can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0251] Based on the same concept, this application also provides a rock failure mode determination device, which corresponds one-to-one with the rock failure mode determination method in the above embodiments. Figure 9 A schematic diagram of the rock failure mode determination device provided in an embodiment of this application is shown. See also: Figure 9As shown, the rock failure mode determination device 900 provided in this application embodiment includes:

[0252] The acquisition module 901 is used to acquire the actual stress data and actual creep deformation data of the target rock.

[0253] Initialization module 902 is used to initialize the initial macroscopic damage data of the target rock to obtain the initial macroscopic damage data to be adjusted;

[0254] The generation module 903 is used to process the initial macroscopic damage data to be adjusted and the actual stress data using the target rock creep model obtained by the construction method of any of the aforementioned rock creep models, and generate theoretical creep deformation data.

[0255] The adjustment module 904 is used to adjust the initial macroscopic damage data to be adjusted based on the theoretical creep deformation data and the actual creep deformation data to obtain the initial macroscopic damage prediction data.

[0256] The determination module 905 is used to determine the failure mode of the target rock based on the initial macroscopic damage prediction data and the creep damage rate parameters in the creep model of the target rock; wherein the failure mode includes structurally controlled failure, stress-controlled failure and combined failure.

[0257] This application provides a rock failure mode determination device. First, it acquires the actual stress and creep deformation data of the target rock. Then, it initializes initial macroscopic damage data to obtain data to be adjusted, providing a reasonable starting point for subsequent iterative optimization. Next, it uses a pre-constructed target rock creep model, combined with the initial macroscopic damage data to be adjusted and the actual stress data, to generate theoretical creep deformation data. Based on the theoretical and actual creep deformation data, it adjusts the initial macroscopic damage data to eliminate initial assumption biases, making the obtained initial macroscopic damage prediction data closely match the actual damage state of the rock mass. Finally, based on the initial macroscopic damage prediction data of the target rock and the creep damage rate parameters in the target rock creep model, it obtains a more accurate rock failure mode determination result that better reflects actual working conditions, providing a targeted basis for engineering support design and avoiding material waste and instability risks caused by blind support.

[0258] Specific limitations regarding the rock failure mode determination device can be found in the limitations of the rock failure mode determination method described above, and will not be repeated here. Each module in the aforementioned rock failure mode determination device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0259] Figure 10 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Figure 10 As shown, at the hardware level, this electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or it may include non-volatile memory, such as at least one disk drive. Of course, this electronic device may also include other hardware required for other business operations.

[0260] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 10 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0261] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0262] The processor reads the corresponding computer program from non-volatile memory into memory and then runs it, forming a device for constructing a rock creep model or a device for determining rock failure modes at the logical level. The processor executes the program stored in memory and specifically performs the aforementioned method.

[0263] The processor may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method.

[0264] This electronic device can execute the rock creep model construction method provided in several embodiments of this application, and realize the rock creep model construction device in Figure 8 The functions of the illustrated embodiment, and the rock failure mode determination device in Figure 9 The functions of the embodiments shown are not described in detail here.

[0265] This application also proposes a computer-readable storage medium that stores one or more programs, the programs including instructions that, when executed by an electronic device including multiple applications, enable the electronic device to perform the rock creep model construction method or rock failure mode determination method provided in several embodiments of this application.

[0266] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0267] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0268] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0269] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0270] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0271] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0272] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0273] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0274] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0275] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for constructing a rock creep model, characterized in that, The method includes: Obtain the dataset; each sample in the dataset includes model input data and model output data, wherein the model input data includes creep holding time, stress data of sample rock, initial macroscopic damage data and initial microscopic damage data of sample rock, as well as preset long-term strength and preset creep initiation threshold, and the model output data is the strain true value; Each sample in the dataset is used as a target sample. The model input data of the target samples is processed using an initial rock creep model to generate strain prediction values. The initial rock creep model includes at least one of the following: a damaged elastic body strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. Each strain term includes a macro-micro nonlinear coupled damage term, which is generated based on initial micro-damage data and initial macro-damage data. The initial micro-damage data is determined based on the degradation of the elastic modulus of the intact rock sample after experiencing loading and unloading paths, and the initial macro-damage data is determined based on the strength or stiffness attenuation rate of the sample rock relative to the intact rock sample. Using a preset loss function, a loss value is calculated based on each predicted strain value and the corresponding true strain value. The initial rock creep model is then fitted according to each loss value to obtain the final rock creep model. If the stress data of the sample rock is greater than or equal to the preset long-term strength, then the process of using the initial rock creep model to process the model input data of the target sample and generate strain prediction values ​​includes: Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data; Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer. Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data. Using the aforementioned damaged viscoplastic strain term, the damaged viscoplastic strain data are calculated based on the stress data of the sample rock, the preset long-term strength, the creep holding time, and the total macroscopic and microscopic damage data. Based on the strain data of the damaged elastomer, the strain data of the damaged viscoelastic body, and the strain data of the damaged viscoplastic body, the predicted strain value is calculated; If the sample rock is a single-fracture rock mass or an isotropic rock mass, then the total macroscopic and microscopic damage data are calculated using the following formula based on the initial macroscopic damage data and initial microscopic damage data of the target sample: in, Represents macro- and micro-level total damage data in scalar form. Represents the initial macroscopic damage data in scalar form. This indicates a detailed view of the total damage data. This represents the initial mesoscopic damage data in scalar form. This represents detailed creep damage data. A The parameter represents the creep damage rate, and t represents the creep holding time. This indicates the strength of rocks containing fractures at different dip angles. This represents the uniaxial compressive strength of the intact rock sample. Indicates the weighting coefficient. This represents the unloading modulus after experiencing a specific stress history. This represents the initial elastic modulus of a complete rock. If the sample rock is a complex jointed rock mass containing multiple sets of joints, then the total macroscopic and microscopic damage data are calculated using the following formula based on the initial macroscopic damage data and initial microscopic damage data of the target sample: in, Represents macro- and micro-level total damage data in tensor form. Represents the identity matrix. Represents the initial macroscopic damage data in tensor form.

2. The method for constructing a rock creep model according to claim 1, characterized in that, If the stress data of the sample rock is less than the preset long-term strength, then the process of processing the model input data of the target sample using the initial rock creep model to generate strain prediction values ​​includes: Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data; Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer. Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data. If the stress data of the sample rock is less than or equal to the preset creep initiation threshold, then the strain data of the damaged elastomer is used as the strain prediction value. If the stress data of the sample rock is greater than the preset creep initiation threshold, the strain prediction value is calculated based on the strain data of the damaged elastomer and the strain data of the damaged viscoelastic body.

3. The method for constructing a rock creep model according to claim 1, characterized in that, The preset long-term strength is generated according to the following method: Multiple data sets are acquired, each data set including independent variable values ​​and dependent variable values. The independent variable values ​​are the initial macroscopic and microscopic total loss data of the sandstone sample, and the dependent variable values ​​are the long-term strength of the sandstone sample. By fitting the multiple data sets, a long-term intensity dynamic weakening function is obtained; The initial values ​​in the macroscopic and microscopic total damage data are processed using the long-term intensity dynamic weakening function to obtain the preset long-term intensity.

4. The method for constructing a rock creep model according to claim 1, characterized in that, The method further includes: Acquire the stress data, initial macroscopic damage data and initial microscopic damage data of the rock to be predicted, and preset the long-term strength; Using the final rock creep model, based on the stress data, initial macroscopic damage data and initial microscopic damage data of the rock to be predicted, strain prediction data of the rock to be predicted is generated by presetting the long-term strength.

5. A method for determining rock failure modes, characterized in that, The method includes: Obtain the actual stress data and actual creep deformation data of the target rock; Initialize the initial macroscopic damage data of the target rock to obtain the initial macroscopic damage data to be adjusted; Using the target rock creep model obtained by the construction method of the rock creep model as described in any one of claims 1-4, the initial macroscopic damage data to be adjusted and the actual stress data are processed to generate theoretical creep deformation data; Based on the theoretical creep deformation data and the actual creep deformation data, the initial macroscopic damage data to be adjusted is adjusted to obtain the initial macroscopic damage prediction data. Based on the initial macroscopic damage prediction data and the creep damage rate parameters in the target rock creep model, the failure modes of the target rock are determined; wherein, the failure modes include structurally controlled failure, stress-controlled failure, and combined failure.

6. A device for constructing a rock creep model, characterized in that, The device includes: The acquisition module is used to acquire a dataset; each sample in the dataset includes model input data and model output data, wherein the model input data includes creep holding time, stress data of sample rock, initial macroscopic damage data and initial microscopic damage data of sample rock, as well as preset long-term strength and preset creep initiation threshold, and the model output data is the strain true value; The prediction module is used to take each sample in the dataset as a target sample, process the model input data of the target sample using an initial rock creep model, and generate strain prediction values. The initial rock creep model includes at least one of the following: a damaged elastic body strain term, a damaged viscoelastic body strain term, and a damaged viscoplastic body strain term. Each strain term includes a macro-micro nonlinear coupled damage term, which is generated based on initial micro-damage data and initial macro-damage data. The initial micro-damage data is determined based on the elastic modulus degradation of the intact rock sample after experiencing loading and unloading paths, and the initial macro-damage data is determined based on the strength or stiffness attenuation rate of the sample rock relative to the intact rock sample. The fitting module is used to calculate the loss value based on each of the predicted strain values ​​and the corresponding true strain values ​​using a preset loss function, and to fit the initial rock creep model according to each of the loss values ​​to obtain the final rock creep model. If the stress data of the sample rock is greater than or equal to the preset long-term strength, the prediction module is used to: Based on the initial macroscopic damage data and initial mesoscopic damage data of the target sample, calculate the total macroscopic and mesoscopic damage data; Using the strain term of the damaged elastomer, the stress data of the sample rock and the total macroscopic and microscopic damage data are processed to obtain the strain data of the damaged elastomer. Using the strain term of the damaged viscoelastic body, the strain data of the damaged viscoelastic body are calculated based on the creep holding time, the stress data of the sample rock, and the total macroscopic and microscopic damage data. Using the aforementioned damaged viscoplastic strain term, the damaged viscoplastic strain data are calculated based on the stress data of the sample rock, the preset long-term strength, the creep holding time, and the total macroscopic and microscopic damage data. Based on the strain data of the damaged elastomer, the strain data of the damaged viscoelastic body, and the strain data of the damaged viscoplastic body, the predicted strain value is calculated; If the sample rock is a single-fracture rock mass or an isotropic rock mass, the prediction module, when calculating the total macro- and micro-damage data based on the initial macro-damage data and initial micro-damage data of the target sample, specifically uses the following formula to calculate the total macro- and micro-damage data: in, Represents macro- and micro-level total damage data in scalar form. Represents the initial macroscopic damage data in scalar form. This indicates a detailed view of the total damage data. This represents the initial mesoscopic damage data in scalar form. This represents detailed creep damage data. A The parameter represents the creep damage rate, and t represents the creep holding time. This indicates the strength of rocks containing fractures at different dip angles. This represents the uniaxial compressive strength of the intact rock sample. Indicates the weighting coefficient. This represents the unloading modulus after experiencing a specific stress history. This represents the initial elastic modulus of a complete rock. If the sample rock is a complex jointed rock mass containing multiple sets of joints, then the total macroscopic and microscopic damage data are calculated using the following formula based on the initial macroscopic damage data and initial microscopic damage data of the target sample: in, Represents macro- and micro-level total damage data in tensor form. Represents the identity matrix. Represents the initial macroscopic damage data in tensor form.

7. A device for determining rock failure modes, characterized in that, The device includes: The acquisition module is used to acquire the actual stress data and actual creep deformation data of the target rock; An initialization module is used to initialize the initial macroscopic damage data of the target rock to obtain the initial macroscopic damage data to be adjusted. The generation module is used to process the initial macroscopic damage data to be adjusted and the actual stress data using the target rock creep model obtained by the construction method of the rock creep model as described in any one of claims 1-4, and generate theoretical creep deformation data. The adjustment module is used to adjust the initial macroscopic damage data to be adjusted based on the theoretical creep deformation data and the actual creep deformation data to obtain the initial macroscopic damage prediction data. The determination module is used to determine the failure mode of the target rock based on the initial macroscopic damage prediction data and the creep damage rate parameters in the creep model of the target rock; wherein the failure mode includes structurally controlled failure, stress-controlled failure and combined failure.

8. An electronic device, comprising: processor; And a memory arranged to store computer-executable instructions, characterized in that, when executed, the executable instructions cause the processor to perform the steps of the method for constructing a rock creep model as claimed in any one of claims 1-4 or the steps of the method for determining rock failure modes as claimed in claim 5.

Images