Visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twinning

Abstract

The invention discloses a rock structure and three-dimensional stress-strain field evolution visualization method based on digital twinning. The rock digital twinborn model is constructed by fusing an in-situ loading CT scanning experiment, three-dimensional digital image correlation method (3D-DIC) surface deformation field measurement, digital volume image correlation method (DVC) internal deformation field measurement, digital modeling and numerical simulation technologies and methods. And obtaining mesomechanics parameters of the numerical model in real time, synchronously performing mirror image simulation, generating a four-dimensional dynamic body image with space-time continuity, and realizing visualization of the whole evolution process of the rock internal structure and the three-dimensional stress-strain field.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for visualizing the evolution of rock structure and three-dimensional stress-strain field based on digital twinning TECHNICAL BACKGROUND

[0002] In the field of rock mechanics, studying the internal deformation evolution of an object is crucial for understanding material properties and mechanical behavior. In the prior art, in-situ loading computed tomography (CT) tests combined with digital volume correlation (DVC) are widely used to obtain the three-dimensional deformation field inside the object. However, the scanning process is time-consuming, and the loading process is usually divided into discrete stages (e.g., scanning at each stable loading state), which results in a lack of temporal and spatial continuity in the obtained deformation field data, making it difficult to capture the dynamic evolution process.

[0003] As a supplement, three-dimensional digital image correlation (3D-DIC) can capture high temporal and spatial continuity of surface deformation fields by using high-speed cameras to capture sequential images of speckle features on the specimen surface. However, this method is limited to surface measurement and cannot reflect the deformation behavior of internal structures. In digital rock mechanics applications, constructing numerical models using CT scanning to obtain the microstructure of the core (through image segmentation and vectorization processing) and combining with macroscopic mechanical experimental parameters for simulation analysis has become a common method for evaluating the mechanical properties of the core. However, due to the heterogeneity of rock, structural discontinuity, and lack of direct data support for internal deformation measurement, the numerical simulation results often differ significantly from the experimental macroscopic stress-strain curve, limiting the accuracy of the model.

[0004] In summary, the main problems of the prior art are: insufficient temporal and spatial resolution of internal deformation observation (such as the DVC method), inability to extend surface deformation measurement (such as the 3D-DIC method) to the interior, and lack of effective fusion between numerical modeling and experimental data. This limits the realization of four-dimensional dynamic volume images (i.e., visualization of structure evolution with temporal and spatial continuity), and there is an urgent need for a technical solution that integrates multi-source data to overcome these deficiencies.

[0005] The experimental device structure used in the experiment refers to the technical solution described in the applicant's granted patent number CN119043228A, which was published on November 29, 2024. SUMMARY

[0006] The present application proposes a method for visualizing the evolution of rock structure and three-dimensional stress-strain field based on digital twinning to observe and analyze the internal structure of rock and the evolution process of three-dimensional stress-strain field.

[0007] To achieve the above-mentioned purpose, the present application provides a method for visualizing the evolution of rock structure and three-dimensional stress-strain field based on digital twinning, comprising:

[0008] Obtaining surface profile image f of the specimen at time T0 before loading s (x s ) and volume image f0(x) at time T0 before loading;

[0009] Establishing a geometric model from the shape of the specimen and meshing the geometric model;

[0010] Applying one or more stages of load to the specimen to obtain the specimen after loading;

[0011] Obtaining complete surface profile image g of the specimen during loading s (x s ’,t) and volume image g(x’,T i ) of the specimen at time T i after loading;

[0012] Combining the mesh, the surface profile image f s (x s ) and volume image f0(x) at time T0 before loading, the complete surface profile image g s (x s ’,t) and volume image g(x’,T i ) after loading of the specimen, calculating and simulating to obtain a digital model and generate virtual volume image g’(x’,t) at different times.

[0013] s represents surface;

[0014] T i represents the i-th loading stage, i = 1, 2, 3, …;

[0015] t represents loading time

[0016] g’ represents virtual volume image

[0017] x s ’ is the coordinates (x s ’, y s ’, z s ’) of each point on the surface of the specimen after loading, x s ’ = x’ + u s (x s );

[0018] x is the coordinates (x, y, z) of each point inside the specimen;

[0019] x’ is the coordinates (x’, y’, z’) of each point inside the specimen after loading, x’ = x + u(x);

[0020] u s (x s ) is the displacement vector (u s(x s ),v s (x s ),w s (x s ));

[0021] u(x) is the displacement vector of each point of the test piece (u(x), v(x), w(x));

[0022] Preferably, the surface profile image f0(x) at T0 before loading is obtained s (x s ) and the volume image f0(x) are specifically

[0023] The speckle is made on the surface of the test piece or the natural texture of the test piece is used, at T0, the surface profile image of the test piece before loading is obtained by multi-camera, and the surface profile image before loading is obtained by the calculation workstation. s (x s );

[0024] The rays are emitted to the test piece before loading by the ray source, and the rays passing through the test piece loaded to a certain load are collected by the image acquisition device and converted into the projection image before loading, and the volume image before loading f0(x) is obtained by the calculation workstation.

[0025] The surface displacement field measurement uncertainty and the overall displacement field measurement uncertainty are obtained, specifically,

[0026] The surface profile image f0(x) before loading is obtained twice, respectively as the reference surface profile image and the target surface profile image. s (x s );

[0027] The root mean square error of surface displacement is calculated by using 3D-DIC method, as the surface displacement field measurement uncertainty.

[0028] The volume image f0(x) before loading is obtained twice, respectively as the reference volume image and the target volume image.

[0029] The root mean square error of displacement is calculated by using DVC method, as the overall three-dimensional displacement field measurement uncertainty.

[0030] Preferably, the geometric model is established from the shape of the test piece, and the geometric model is meshed, specifically,

[0031] The internal microstructure of the test piece is processed by image segmentation and vectorization method based on the loading precursor image f0(x), to generate a complete microstructure grid (which can use GMESH, Mimics and other software), and all grid node displacements are obtained by DVC method in subsequent measurement;

[0032] The surface profile f of the test piece is reconstructed using 3D-DIC method s (x s ), and the surface grid of the test piece is extracted, and the surface grid node displacement is obtained by 3D-DIC method in subsequent measurement;

[0033] Preferably, the complete surface profile image g s (x s ’,t) of the test piece during loading and the volume image g(x’,T i ) loaded to the T i th stage are obtained, specifically:

[0034] The test piece is loaded, and the complete surface profile image g s (x s ’,t) of the test piece during loading is reconstructed by the images of the test piece at different angles at the same time through continuous acquisition by multiple cameras, and the surface grid node displacement of the test piece is continuously obtained by 3D-DIC method;

[0035] The test piece after loading is irradiated by a radiation source, and the radiation passing through the test piece loaded to the T i th stage is collected by an image acquisition device and converted into a post-loading projection image, and the post-loading projection image is reconstructed by a computing workstation to obtain the volume image g(x’,T i ) at the T i th stage, and the overall grid node displacement of the test piece is obtained by DVC method.

[0036] The grid, the surface profile image f s (x s ) before loading and the volume image f0(x), the complete surface profile image g s (x s ’,t) of the test piece and the volume image g(x’,T i ) after loading are calculated and simulated to obtain a digital model, and a series of virtual deformed volume images g’(x’,t) at different times are generated, specifically:

[0037] The complete microstructure grid is imported into a numerical simulation software to establish a numerical model, and the material mechanical parameters obtained from literature or conventional experiments are used as initial values (such as elastic modulus, Poisson's ratio, etc.) to assign to the numerical model

[0038] A continuous load is applied to the specimen, and multiple cameras continuously acquire and obtain a series of complete surface profile images of the specimen. s (x s The displacement of the mesh nodes on the specimen surface was continuously obtained using the 3D-DIC method.

[0039] The surface displacement results calculated by 3D-DIC are applied as boundary conditions to the outer surface mesh nodes of the numerical model for numerical simulation. The simulation is performed with the condition of minimizing the difference between the simulated load and the actual applied load. Update model parameters to enable interaction between experimental data and numerical simulation, and obtain optimal model material mechanical parameters;

[0040] Under the optimal model material mechanics parameters, the full-field three-dimensional stress-strain field of the specimen during loading and its visualization were obtained by numerical simulation;

[0041] The full-field displacement of the specimen obtained by numerical simulation is applied to the pre-loading image f0(x) to obtain the specimen deformation images g'(x',t) at different stages, thereby realizing the three-dimensional visualization of the internal structure of the specimen during loading.

[0042] Pause loading and hold the load, get the Tth... i Volumetric image g(x',T) of the stage specimen i Simultaneously, a complete surface contour image g of the specimen is acquired. s (x s ',t);

[0043] Using the pre-loaded volume image f0(x) as the reference volume image, g(x',T) i ( ) is the target volume image, and the DVC method is used to calculate the full-field displacement of the specimen;

[0044] The displacement of the surface obtained by 3D-DIC under the load is used as the boundary condition for numerical simulation, and the condition is to minimize the difference between the simulated load and the actual applied load (i.e. Furthermore, the condition of minimizing the difference between the nodal displacements obtained from DVC and the corresponding nodal displacements in numerical simulation is added (i.e. Construct the total function of constraints. , ( Using weights, optimize the mechanical parameters of the microstructure;

[0045] Numerical simulation is performed using optimized parameters. The three-dimensional stress-strain field is obtained and visualized from the numerical simulation. The displacement obtained from the numerical simulation is applied to the f0(x) image to obtain the virtually generated deformable body image g'(x',t), thus realizing the visualization of the specimen structure.

[0046] Alternatively, the virtually generated Tth...i The deformable image g'(x',T) at each stage i ) and the actual deformable image g(x',T i The grayscale residual is used as an evaluation condition for optimizing mechanical parameters, satisfying the condition of minimizing the grayscale residual (i.e., ...). The mechanical parameters at that time.

[0047] In the formula, [M] is the Hessian matrix. For the measured nodal displacements, These are the nodal displacements obtained from numerical simulation calculations. The standard deviation of the noise, expressed in grayscale. for The degrees of freedom (e.g., a displacement field has n grid nodes, each node has 3 degrees of freedom, for a total of 3n degrees of freedom);

[0048] In finite element theory, the displacement of each point within a mesh element can be expressed as nodal displacements. , For shape functions, Let represent the degrees of freedom of the i-th node, and [M] be the Hessian matrix. , The grayscale gradient of the reference image;

[0049] In the formula For the measured load, The load is obtained from numerical simulation calculation. The standard deviation of the load measurement. for The number of elements (the number of load measurements);

[0050] In the formula, VOI represents the computational region of DVC.

[0051] Optionally, by establishing a numerical model and mechanical parameters, the loading conditions can be reset, and the stress and deformation fields at different stages can be obtained through numerical simulation. The displacement field can be applied to the volume image f0(x), and virtual volume images g'(x',t) of the microstructure of the specimen at different stages can be generated virtually, thereby realizing the observation of the spatiotemporal continuity of the specimen structure under different loading conditions.

[0052] Optionally, an early warning can be set based on the stress-strain field obtained from numerical simulation, according to the strength criterion and the crack propagation process of the simulated volume image, such as when the probability of damage initiation is >60% or the loss of structural integrity is >85%.

[0053] Optionally, such as numerical model meshes, DVC calculation meshes and 3D-DIC surface meshes are divided separately, requiring spatial registration to ensure a unified physical coordinate system and consistent scale.

[0054] An experimental device for visualizing the evolution of rock structures and three-dimensional stress-strain fields based on digital twins is characterized by comprising a base, a radiation source, an image acquisition device, a loading device, multiple cameras, a guide rail, a scanning control device, a camera control device, a loading control device, and a computing workstation.

[0055] The base is equipped with the radiation source, the image acquisition device, the loading device, and multiple cameras.

[0056] The radiation source and the image acquisition device are arranged opposite to each other, the loading device is located between the radiation source and the image acquisition device, the specimen is placed on the loading device, the loading device is used to apply a load to the specimen and can drive the specimen to rotate circumferentially, the radiation source is used to emit radiation, the image acquisition device is used to acquire the radiation passing through the specimen and convert it into a projection image, and the computing workstation reconstructs the projection image into a volumetric image.

[0057] The multiple cameras are arranged around the loading device to acquire surface images of the specimen from different angles. The computing workstation then reconstructs the surface images from different angles and synthesizes the surface contour image.

[0058] This application integrates in-situ loading experiments, numerical modeling, and simulation analysis, using the internal deformation field of the DVC and the surface deformation field measurement of the 3D-DIC as a bridge to construct a digital twin system of the rock core. The model parameters are updated in real time, realizing a mirror simulation that synchronizes the simulation and experimental processes, constructing a four-dimensional dynamic volume image with spatiotemporal continuity, and visualizing the entire process of internal structural evolution. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort, and this application can be applied to other similar scenarios based on the provided drawings. Unless obvious from the linguistic context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0060] Figure 1 This is a flowchart of the first implementation of the visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins in this application;

[0061] Figure 2This is a flowchart of the second implementation of the visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins in this application (reducing the DVC calculation part);

[0062] Figure 3 This is a flowchart of the third implementation of the visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins in this application (reducing the 3D-DIC calculation part);

[0063] Figure 4 This is a schematic diagram of the loading time of the visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins, as described in this application.

[0064] Figure 5 This is a schematic diagram of the experimental setup for the visualization method of rock structure and three-dimensional stress-strain field evolution based on digital twins, as described in this application.

[0065] Figure 6 This is an example of the computational grid for the visualization method of rock structure and three-dimensional stress-strain field evolution based on digital twins in this application.

[0066] 1-Base, 2-Radiation source, 3-Image acquisition device, 4-Loading device, 5-Multiple cameras, 6-Guide rail, 7-Scanning control device, 8-Camera control device, 9-Loading control device, 10-Computing workstation; Detailed Implementation

[0067] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are merely illustrative of the application and not intended to limit it. The described embodiments are only a part of the embodiments of the present application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without inventive effort are within the scope of protection of the present application.

[0068] It should be noted that, for ease of description, only the parts relevant to the application are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in these embodiments can be arbitrarily combined, provided that the combined technical features are not contradictory. All feasible combinations of features are the technical content explicitly described herein. Any one of the multiple sub-features contained in the same statement can be applied independently, without necessarily being applied together with other sub-features.

[0069] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "a," and / or "the" are not specifically singular and may include the plural. Generally, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements. An element defined by the phrase "comprising an..." does not exclude the presence of other identical elements in the process, method, product, or apparatus that includes the element.

[0070] In the description of the embodiments of this application, unless otherwise stated, " / " means "or", for example, A / B can mean A or B; "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments of this application, "multiple" refers to two or more.

[0071] In-situ loading CT scans are used to obtain three-dimensional volumetric images of an object under different loading conditions. Combined with digital volume correlation (DVC), the deformation field inside the object can be obtained. However, because the X-ray scanning process is lengthy, the entire loading process is usually divided into several stages. In each stage, the load is maintained while X-ray scanning is performed to obtain stage images. This process leads to a lack of spatiotemporal continuity in the observation of the structure and displacement field.

[0072] Three-dimensional digital image correlation (3D-DIC) can calculate the three-dimensional deformation field of the specimen surface by acquiring a series of images with speckle characteristics on the specimen surface during the loading process. It is accompanied by a high-speed camera and the measurement has high spatiotemporal continuity. However, its measurement results can only reflect the deformation of the specimen surface.

[0073] In the field of digital rock mechanics, CT scans are used to obtain the microstructure of rock cores. After image segmentation and vectorization, a numerical model of the core is established, and then numerical simulation software is used for simulation analysis to obtain the mechanical properties of the core. The initial values ​​of the model's micromechanical parameters are obtained through macroscopic mechanical experiments, and the mechanical parameters are corrected by comparing the macroscopic stress-strain curves of the numerical simulation and the experiments. However, due to the heterogeneity and discontinuous structure of rocks, simulation results often differ from experimental results. The reason for this is the lack of data fusion between internal deformation measurements and numerical simulations.

[0074] This application discloses a visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins, including:

[0075] Obtain the surface profile image f at time T0 before loading. s (x s ) and volumetric image f0(x);

[0076] A geometric model is established based on the shape of the specimen, and the geometric model is then meshed.

[0077] One or more stages of load are applied to the specimen to obtain the loaded specimen;

[0078] Obtain a complete surface contour image g of the specimen during the loading process. s (x s ',t) and loaded to the Tth i The volumetric image of the stage g(x',T) i );

[0079] Based on the mesh, the surface contour image f before loading s (x s The specimen contains a volume image f0(x) and a complete surface contour image g. s (x s The loaded volume image g(x',T) and t) i ) Perform calculations and simulations to obtain a digital model and generate a virtual deformable image g'(x', t).

[0080] 's' represents the surface.

[0081] T i This represents the i-th loading stage, where i = 1, 2, 3, ...;

[0082] t represents loading time

[0083] g' represents a virtual volume image.

[0084] x s ' represents the coordinates (x, y) of each point on the surface of the specimen after loading. s ',y s ',z s '), x s '=x'+u s (x s );

[0085] x represents the coordinates (x, y, z) of each point inside the specimen;

[0086] The coordinates (x', y', z') of each point inside the specimen after x' is not loaded, x' = x + u(x);

[0087] u s (x s ) is the displacement vector (u) of each point on the surface of the specimen.s (x s ),v s (x s ),w s (x s ));

[0088] u(x) is the displacement vector (u(x),v(x),w(x)) of each point on the specimen.

[0089] This application discloses a visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins. It integrates digital volume image correlation (DVC), three-dimensional digital image correlation (3D-DIC), and digital twin technology to construct a digital twin model with four-dimensional spatiotemporal continuity. The system uses DVC to identify the internal microstructure of the specimen and perform full-field displacement analysis. Combined with 3D-DIC technology, it acquires real-time dynamic distribution data of the surface stress field. By identifying constraints and parameters, it establishes a digital twin model, overcoming the limitations of spatiotemporal discontinuity in traditional volumetric image acquisition.

[0090] Traditional X-ray scanning only provides static volumetric image data. DVC (internal microstructure) technology and 3D-DIC (surface stress field) are mostly independent systems. This method establishes a three-dimensional collaborative observation system integrating DVC (internal microstructure), 3D-DIC (surface stress field), and digital twin (spatiotemporal evolution) to achieve cross-dimensional data fusion. Leveraging the four-dimensional modeling capabilities of digital twins, it enables continuous dynamic reconstruction with high temporal resolution, making it possible to visualize transient processes such as crack initiation / propagation.

[0091] This application provides the following three reference implementation examples:

[0092] Please see Figure 1 process, Figure 4 Indication and Figure 6 This application discloses a visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins. The first implementation includes:

[0093] At time T0, surface profile images f of the specimen before loading are acquired using multiple cameras. s (x s The loading process involves creating a speckle pattern on the surface of the specimen, acquiring the surface contour image of the specimen before loading using a multi-camera system (5), and reconstructing the surface image before loading using a computing workstation to obtain the surface contour image f0(x). s (x s The X-ray source 2 emits X-rays to the specimen before loading, and the image acquisition device 3 collects the X-rays passing through the specimen before loading and converts them into a pre-loading projection image. The computing workstation 10 reconstructs the pre-loading projection image to obtain the pre-loading body image f0(x).

[0094] The speckle pattern can be created by at least one of spraying, dotting, and transfer printing, or by utilizing the natural texture of the specimen surface as the speckle pattern.

[0095] Before applying a load to the specimen and obtaining the loaded specimen, the method further includes obtaining the measurement uncertainty of the surface displacement field and the measurement uncertainty of the internal displacement field, specifically...

[0096] The surface contour image f before loading is acquired twice. s (x s (), respectively serving as the reference surface contour image and the target surface contour image;

[0097] The root mean square error of the surface displacement is calculated using the 3D-DIC method and used as the measurement uncertainty of the surface displacement field.

[0098] The loading pre-image f0(x) is obtained twice, and used as the reference image and the target image, respectively.

[0099] The root mean square error of displacement is calculated using the DVC method and used as the measurement uncertainty of the three-dimensional displacement field.

[0100] A geometric model is established based on the shape of the specimen, and the geometric model is then meshed. Specifically,

[0101] The internal microstructure of the specimen is processed by image segmentation and vectorization methods based on the loading pre-image f0(x) to generate a complete microstructure mesh (which can be generated using software such as GMESH and Mimics). In subsequent measurements, the displacement of all mesh nodes is obtained by the DVC method.

[0102] The surface profile of the specimen was reconstructed using the 3D-DIC method. s (x s The surface mesh of the specimen was extracted, and the displacement of the surface mesh nodes was obtained in the subsequent measurement using the 3D-DIC method.

[0103] Optionally, the numerical model mesh, the DVC calculation mesh, and the 3D-DIC surface mesh are divided separately, and spatial registration is required to ensure a unified physical coordinate system and consistent scale.

[0104] Get the surface profile image g at loading time s (x s ',t) and loaded to the Tth i The volumetric image of the stage g(x',T) i Specifically, a load is applied to the specimen, and multiple cameras 5 continuously acquire images of the specimen surface from different angles at the same time. A computing workstation 10 then reconstructs the complete surface contour image g of the specimen during loading.s (x s The displacement of the mesh nodes on the specimen surface was continuously obtained using the 3D-DIC method.

[0105] X-rays are emitted from X-ray source 2 onto the loaded specimen, and the image acquisition device 3 captures the rays passing through the specimen at the set load stage T. i The X-rays of the specimen are converted into a post-loaded projection image, and the post-loaded projection image is reconstructed by the computing workstation 10 to obtain the Tth... i After stage loading, the volumetric image g(x',T) i The DVC method was used to obtain the overall mesh node displacement of the specimen.

[0106] Based on the mesh, the surface contour image f before loading s (x s The image f0(x) and the volumetric image f0(x), wherein the complete surface contour image g of the specimen during loading. s (x s The loaded volumetric image g(x',T) and g(x',T) i ) Perform calculations and simulations to obtain a digital model and generate a virtual deformable image g'(x',t). Specifically,

[0107] The complete microstructure mesh is imported into the numerical simulation software to establish a numerical model. Constitutive relations are selected, and material mechanical parameters obtained from literature or conventional experiments are used as initial values ​​(such as elastic modulus, Poisson's ratio, etc.) to assign to the numerical model.

[0108] Optionally, an early warning can be set based on the stress-strain field obtained from numerical simulation, according to the strength criterion and the crack propagation process of the simulated volume image, such as when the probability of damage initiation is >60% or the loss of structural integrity is >85%.

[0109] A continuous load is applied to the specimen, and multiple cameras (5 in total) continuously acquire and obtain a series of complete surface profile images of the specimen. s (x s The displacement of the mesh nodes on the specimen surface was continuously obtained using the 3D-DIC method.

[0110] The surface displacement results calculated by 3D-DIC are applied as boundary conditions to the outer surface mesh nodes of the numerical model for numerical simulation. The simulation is performed with the condition of minimizing the difference between the simulated load and the actual applied load. Update model parameters to enable interaction between experimental data and numerical simulation, and obtain optimal model material mechanical parameters;

[0111] Under the optimal model material mechanics parameters, the full-field three-dimensional stress-strain field of the specimen during loading and its visualization were obtained by numerical simulation;

[0112] The full-field displacement of the specimen obtained by numerical simulation is applied to the volume image f0(x) before loading to obtain the virtual deformed body image g'(x',t) of the specimen during different loading processes, thereby realizing the three-dimensional visualization of the internal structure of the specimen in the T0-T1 stage;

[0113] Pause loading and maintain the load, acquire the volumetric image g(x',T1) of the specimen during stages T1-T2, and simultaneously obtain the complete surface contour image g of the specimen. s (x s ',t);

[0114] Using the volumetric image f0(x) at time T0 as the reference volumetric image and g(x',T1) as the target volumetric image, the DVC method is used to calculate the full-field displacement of the specimen.

[0115] The surface displacement obtained by 3D-DIC at time T1 is used as the boundary condition for numerical simulation, with the condition being the minimum difference between the simulated load and the actual applied load (i.e. Furthermore, the condition of minimizing the difference between the nodal displacements obtained from DVC and the corresponding nodal displacements simulated by the numerical simulation is added (i.e. ), construct the total function , ( Using weights, optimize the mechanical parameters of the microstructure;

[0116] Numerical simulation is performed using the optimized parameters. The displacement obtained from the numerical simulation is applied to the f0(x) image to obtain the virtually generated deformable body image g'(x',T1).

[0117] The grayscale residual between the virtually generated deformable body image g'(x, T1) and the actual deformable body image g(x', T1) is used as the evaluation condition for optimizing the mechanical parameters, satisfying the condition of minimizing the grayscale residual (i.e., ... The mechanical parameters at that time.

[0118] In the formula, [M] is the Hessian matrix. For the measured nodal displacements, These are the nodal displacements obtained from numerical simulation calculations. The standard deviation of the noise, expressed in grayscale. for The degrees of freedom (e.g., a displacement field has n grid nodes, each node has 3 degrees of freedom, for a total of 3n degrees of freedom);

[0119] In finite element theory, the displacement of each point within a mesh element can be expressed as nodal displacements. , For shape functions, Let represent the degrees of freedom of the i-th node, and [M] be the Hessian matrix. , The grayscale gradient of the reference image;

[0120] In the formula For the measured load, The load is obtained from numerical simulation calculation. The standard deviation of the load measurement. for The number of elements (the number of load measurements);

[0121] In the formula, VOI represents the computational region of DVC;

[0122] Optionally, by using the established numerical model and mechanical parameters, the loading conditions can be reset, and the stress and deformation fields at different stages can be obtained through numerical simulation. The displacement field can be applied to the volume image f0(x), and virtual volume images g'(x',t) of the microstructure of the specimen at different stages can be generated virtually, thereby realizing the observation of the spatiotemporal continuity of the specimen structure under different loading conditions.

[0123] Please see Figure 2 process, Figure 4 Indication and Figure 6 This application discloses a visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins. A second implementation method includes:

[0124] At time T0, surface profile images f of the specimen before loading are acquired using multiple cameras. s (x s The loading pre-loading image f0(x) was obtained by X-ray scanning. Specifically, speckle patterns were created on the surface of the specimen, and images of the specimen surface before loading at different angles were acquired by multiple cameras 5. The loading pre-loading surface images were then reconstructed by a computing workstation to obtain the loading pre-loading surface contour image f0(x). s (x s ).

[0125] Before applying a load to the specimen and obtaining the loaded specimen, the surface displacement field measurement uncertainty and the overall displacement field measurement uncertainty are obtained. Specifically,

[0126] The surface contour image f before loading is acquired twice. s (x s (), respectively serving as the reference surface contour image and the target surface contour image;

[0127] The root mean square error of the surface displacement is calculated using the 3D-DIC method and used as the measurement uncertainty of the surface displacement field.

[0128] A geometric model is established based on the shape of the specimen, and the geometric model is then meshed. Specifically,

[0129] The surface profile of the specimen was reconstructed using the 3D-DIC method. s (x s The surface mesh of the specimen was extracted, and the displacement of the surface mesh nodes was obtained in the subsequent measurement using the 3D-DIC method.

[0130] A continuous load is applied to the specimen, and multiple cameras (5 in total) continuously acquire and obtain complete surface contour images of the specimen. s (x s The displacement of the mesh nodes on the specimen surface was continuously obtained using the 3D-DIC method.

[0131] The surface displacement results calculated by 3D-DIC are applied as boundary conditions to the outer surface mesh nodes of the numerical model for numerical simulation. The simulation is performed with the condition of minimizing the difference between the simulated load and the actual applied load. This allows for the updating of model parameters and enables interaction between experimental data and numerical simulation.

[0132] Please see Figure 3 process, Figure 4 Indication and Figure 6 This application discloses a third implementation of a visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins, which includes:

[0133] At time T0, the pre-loading image f0(x) is obtained by X-ray scanning. Specifically, X-ray source 2 emits X-rays to the pre-loading specimen, and image acquisition device 3 collects the X-rays passing through the pre-loading specimen and converts them into a pre-loading projection image. The pre-loading projection image is reconstructed by computing workstation 10 to obtain the pre-loading image f0(x).

[0134] The loading pre-image f0(x) is obtained twice, and used as the reference image and the target image, respectively.

[0135] The root mean square error of displacement is calculated using the DVC method and used as the measurement uncertainty of the three-dimensional displacement field.

[0136] A geometric model is established based on the shape of the specimen, and the geometric model is then meshed. Specifically,

[0137] The internal microstructure of the specimen is processed by image segmentation and vectorization methods based on the loading pre-image f0(x) to generate a complete microstructure mesh (which can be generated using software such as GMESH and Mimics). In subsequent measurements, the displacement of all mesh nodes is obtained by the DVC method.

[0138] Combining the aforementioned mesh, the loaded precursor image f0(x), and the Tth... i The volumetric image g(x',T) after stage loading i) Perform calculations and simulations to obtain a digital model and generate a virtual deformable image g'(x',t), specifically,

[0139] The complete microstructure mesh is imported into the numerical simulation software to establish a numerical model. Constitutive relations are selected, and material mechanical parameters obtained from literature or conventional experiments are used as initial values ​​(such as elastic modulus, Poisson's ratio, etc.) to assign to the numerical model.

[0140] Optionally, an early warning can be set based on the stress-strain field obtained from numerical simulation, according to the strength criterion and the crack propagation process of the simulated volume image, such as when the probability of damage initiation is >60% or the loss of structural integrity is >85%.

[0141] The specimen is loaded to a preset value, loading is paused and maintained, and the specimen is scanned using X-rays to obtain a volumetric image g(x',T) of the specimen. i During this period, the model continuously updates the simulated volume image g'(x',T) based on the load values. i );

[0142] The condition is to minimize the difference between the load simulated by the model and the actual load applied (i.e.) ), and the condition that minimizes the difference between the internal nodal displacements obtained from DVC and the corresponding nodal displacements in numerical simulation (i.e. ), construct the total function , ( Using weights as parameters, different weights are used to construct functions, further refining the mechanical parameters of each microstructure, updating the simulation parameters, and updating the virtual volume image g'(x',T). i ).

[0143] Please refer to the experimental setup for a visualization method based on digital twins for rock structure and three-dimensional stress-strain field evolution disclosed in this application. Figure 5 It includes a base, a radiation source, an image acquisition device, a loading device, a multi-camera system, a guide rail, a scanning control device, a camera control device, a loading control device, a computing workstation, and a 10.

[0144] The base 1 is equipped with the radiation source, the image acquisition device 3, the loading device 4, and the multi-camera 5.

[0145] The X-ray source 2 and the image acquisition device 3 are arranged opposite to each other. The loading device 4 is located between the X-ray source 2 and the image acquisition device 3. The specimen is placed on the loading device 4. The loading device 4 is used to apply a load to the specimen and can drive the specimen to rotate circumferentially. The X-ray source 2 is used to emit X-rays. The image acquisition device 3 is used to acquire the X-rays passing through the specimen and convert them into a projection image. The computing workstation 10 reconstructs the projection image into a solid image.

[0146] The multiple cameras 5 are arranged around the loading device 4 to acquire surface images of the specimen from different angles, and the computing workstation 10 reconstructs the surface images from different angles and synthesizes the surface contour image.

[0147] The experimental setup is an improvement on the setup in patent CN119043228A, which changes the surface imaging device to a multi-camera setup surrounding the loading device 4.

[0148] The multi-camera 5 includes, but is not limited to, one or more imaging devices.

[0149] This application discloses a visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins, which can output the following structures, including but not limited to: load data (load-time curve, loading history), image data (volume images and surface images at each stage, displacement fields on the surface and inside of the specimen), virtual volume image (g'(x',t)), microstructure evolution (crack propagation path, material phase transformation process), and parameter evolution (local stiffness distribution, dynamic adjustment record of elastic modulus, etc.).

[0150] The model achieves continuity of the specimen across time and space. Experimental data (load, displacement, CT images) drive the virtual model in a time-series manner, dynamically mapping the spatiotemporal evolution of the specimen's structure, stress, and strain throughout the loading process. Continuous spatiotemporal data can accurately reproduce the entire process of the specimen from microscopic damage initiation to macroscopic fracture, enabling life prediction and failure early warning based on parameter identification and evolution patterns. It achieves full-field stress / strain distribution analysis and simulation at the macroscopic scale, and can also analyze stress concentration, crack initiation, and propagation mechanisms at the microscopic scale. By adjusting simulation parameters, the microstructural characteristics, stress distribution, and displacement field changes of the specimen can be simulated at different scales, providing theoretical support for practical engineering applications.

Claims

1. A visualization method for the evolution of rock structure and three-dimensional stress-strain field based on digital twins, characterized in that, include: Obtain the specimen surface profile image f at time T0 before loading. s (x s ) and volumetric image f0(x); A geometric model is established based on the shape of the specimen, and the geometric model is then meshed. One or more stages of load are applied to the specimen to obtain the loaded specimen; Obtain a complete surface contour image g of the specimen during the loading process. s (x s ',t) and loaded to the Tth i The volumetric image of the stage g(x',T) i ); Based on the mesh, the surface contour image f before loading s (x s The specimen contains a volume image f0(x) and a complete surface contour image g. s (x s The loaded volumetric image g(x',T) and g(x',T) i ) Perform calculations and simulations to obtain a digital model and generate virtual deformable body images g'(x',t) at different times. 's' represents the surface. T i This represents the i-th loading stage, where i = 1, 2, ...; t represents loading time g' represents a virtual volume image. x s ' represents the coordinates (x, y) of each point on the surface of the specimen after loading. s ',y s ',z s '), x s '=x'+u s (x s ); x represents the coordinates (x, y, z) of each point inside the specimen; x' is the coordinates (x', y', z') of each point inside the specimen after loading, x' = x + u(x); u s (x s ) is the displacement vector (u) of each point on the surface of the specimen. s (x s ),v s (x s ),w s (x s )); u(x) is the displacement vector (u(x), v(x), w(x)) of each point on the specimen.

2. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 1, characterized in that, Obtain the surface profile image f at time T0 before loading. s (x s ) and volumetric image f0(x), specifically A speckle pattern is created on the surface of the specimen or the natural texture of the specimen is utilized. At time T0, a surface contour image of the specimen before loading is acquired by a multi-camera system (5), and the surface image before loading is reconstructed by a computing workstation (10) to obtain the surface contour image f before loading. s (x s ) The X-ray source (2) emits X-rays to the specimen before loading, and the image acquisition device (3) acquires the X-rays passing through the specimen before loading and converts them into a pre-loading projection image. The computing workstation (10) reconstructs the pre-loading projection image to obtain the pre-loading body image f0(x).

3. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 1, characterized in that, Before applying a load to the specimen and obtaining the loaded specimen, the method further includes: obtaining the surface displacement field measurement uncertainty and the overall displacement field measurement uncertainty.

4. The full-field displacement measurement method according to claim 3, characterized in that, Obtain the measurement uncertainty of the surface displacement field and the measurement uncertainty of the overall displacement field, specifically as follows: The surface contour image f before loading is acquired twice. s (x s (), respectively serving as the reference surface contour image and the target surface contour image; The root mean square error of the surface displacement is calculated using the 3D-DIC method and used as the measurement uncertainty of the surface displacement field. The loading pre-image f0(x) is obtained twice, and used as the reference image and the target image, respectively; The root mean square error of the displacement is calculated using the DVC method and used as the measurement uncertainty of the overall three-dimensional displacement field.

5. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 1, characterized in that, A geometric model is established based on the shape of the specimen, and the geometric model is then meshed. Specifically, The internal microstructure of the specimen is processed by image segmentation and vectorization methods based on the loading pre-image f0(x) to generate a complete microstructure mesh (which can be generated using software such as GMESH and Mimics). In subsequent measurements, the displacement of all mesh nodes is obtained by the DVC method. The surface profile of the specimen was reconstructed using the 3D-DIC method. s (x s The surface mesh of the specimen was extracted, and the displacement of the surface mesh nodes was obtained in the subsequent measurement using the 3D-DIC method.

6. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 1, characterized in that, Obtain a complete surface contour image g of the specimen during the loading process. s (x s ',t) and loaded to the Tth i The volumetric image of the stage g(x',T) i Specifically, A load is applied to the specimen, and multiple cameras (5) continuously acquire images. The complete surface contour image g of the specimen under load is reconstructed by a computing workstation (10) using images of the specimen surface at different angles at the same time. s (x s The displacement of the mesh nodes on the specimen surface was continuously obtained using the 3D-DIC method. The X-ray source (2) emits X-rays onto the loaded specimen, and the image acquisition device (3) captures the X-rays passing through the loaded specimen at the set load stage T. i The X-rays of the specimen are converted into a post-loaded projection image, and the post-loaded projection image is reconstructed by the computing workstation (10) to obtain the Tth loading. i Stage volume image g(x',T) i The DVC method was used to obtain the overall mesh node displacement of the specimen.

7. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 1, characterized in that, Based on the mesh, the surface contour image f before loading s (x s The image f0(x) and the volumetric image f0(x), wherein the complete surface contour image g of the specimen during loading. s (x s The loaded volumetric image g(x',T) and g(x',T) i The calculations and simulations are performed to obtain a digital model and generate virtual deformable body images g'(x',t) at different times, specifically: The complete microstructure mesh is imported into the numerical simulation software to establish a numerical model. Constitutive relations are selected, and material mechanical parameters obtained from literature or conventional experiments are used as initial values ​​(such as elastic modulus, Poisson's ratio, etc.) to assign to the numerical model. A load is continuously applied to the specimen, and multiple cameras (5) continuously acquire and obtain a series of complete surface contour images of the specimen. s (x s The displacement of the mesh nodes on the specimen surface was continuously obtained using the 3D-DIC method. The surface displacement results calculated by 3D-DIC are applied as boundary conditions to the outer surface mesh nodes of the numerical model for numerical simulation. The simulation is performed with the condition of minimizing the difference between the simulated load and the actual applied load. Update model parameters to enable interaction between experimental data and numerical simulation, and obtain optimal model material mechanical parameters; Under the optimal model material mechanics parameters, the full-field three-dimensional stress-strain field of the specimen during loading and its visualization were obtained by numerical simulation; The full-field displacement of the specimen obtained by numerical simulation is applied to the pre-loading image f0(x) to obtain virtual deformation images g'(x',t) of the specimen at different stages, thereby realizing three-dimensional visualization of the internal structure of the specimen during loading. Pause loading and hold the load, get the Tth... i Volumetric image g(x',T) of the stage specimen i Simultaneously, a complete surface contour image g of the specimen is acquired. s (x s ',t); Using the pre-loaded volume image f0(x) as the reference volume image, g(x',T) i ( ) is the target volume image, and the DVC method is used to calculate the full-field displacement of the specimen; The displacement of the surface obtained by 3D-DIC under the load is used as the boundary condition for numerical simulation, and the condition is to minimize the difference between the simulated load and the actual applied load (i.e. Furthermore, the condition of minimizing the difference between the internal nodal displacements obtained from DVC and the corresponding nodal displacements in numerical simulation is added (i.e. ), construct the total function of constraints , ( Using weights, optimize the mechanical parameters of the microstructure; Numerical simulation is performed using optimized parameters. The three-dimensional stress-strain field is obtained and visualized using numerical simulation. The displacement obtained from numerical simulation is applied to the f0(x) image to obtain the virtually generated deformable body image g'(x',t), thus realizing the visualization of the specimen structure. Alternatively, the virtually generated Tth... i The deformable image g'(x',T) at time T i ) and the actual deformable image g(x',T i The grayscale residual is used as an evaluation condition for optimizing mechanical parameters, satisfying the condition of minimizing the grayscale residual (i.e., ...). The mechanical parameters at that time. In the formula, [M] is the Hessian matrix. For the measured nodal displacements, These are the nodal displacements obtained from numerical simulation calculations. The standard deviation of noise expressed in grayscale. for The degrees of freedom (e.g., a displacement field has n grid nodes, each node has 3 degrees of freedom, so there are 3n degrees of freedom); In finite element theory, the displacement of each point within a mesh element can be expressed as nodal displacements. In the formula For shape functions, Let represent the degrees of freedom of the i-th node, and [M] be the Hessian matrix. , The grayscale gradient of the reference image. In the formula For the measured load, The load is obtained from numerical simulation calculation. The standard deviation of the load measurement. for The number of elements (the number of load measurements); In the formula, VOI represents the computational region of DVC.

8. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 7, characterized in that, Its digital model applications also include: By establishing a numerical model and mechanical parameters, resetting the loading conditions, and obtaining the stress and deformation fields at different stages through numerical simulation, the displacement field is applied to the volume image f0(x), which can virtually generate virtual volume images g'(x',t) of the microstructure of the specimen at different times, thereby realizing the observation of the spatiotemporal continuity of the specimen structure under different loading conditions.

9. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 7, characterized in that, Its digital twin model applications also include: Based on the stress-strain field obtained from numerical simulation, and according to the strength criterion and the crack propagation process in the simulated volume image, an early warning is set when the probability of damage initiation is >60% or the loss of structural integrity is >85%.

10. The visualization method for rock structure and three-dimensional stress-strain field evolution based on digital twins according to claim 1, characterized in that, The digital model applies or reduces model building conditions based on the implementation instance.

11. The visualization method for rock force structure and three-dimensional stress-strain field evolution based on digital twins according to claim 5, characterized in that, Mesh generation also includes: for example, numerical model meshes, DVC calculation meshes and 3D-DIC surface meshes are separately generated and require spatial registration to ensure a unified physical coordinate system and consistent scale.

12. An experimental apparatus for visualizing the evolution of rock structures and three-dimensional stress-strain fields based on digital twins, characterized in that... Includes a base (1), a radiation source (2), an image acquisition device (3), a loading device (4), multiple cameras (5), a guide rail (6), a scanning control device (7), a camera control device (8), a loading control device (9), and a computing workstation (10); The base (1) is equipped with the radiation source (2), the image acquisition device (3), the loading device (4), and the multi-camera (5). The X-ray source (2) and the image acquisition device (3) are arranged opposite to each other. The loading device (4) is located between the X-ray source (2) and the image acquisition device (3). The specimen is placed on the loading device (4). The loading device (4) is used to apply a load to the specimen and can drive the specimen to rotate circumferentially. The X-ray source (2) is used to emit X-rays. The image acquisition device (3) is used to acquire the X-rays passing through the specimen and convert them into a projection image. The computing workstation (10) reconstructs the projection image into a solid image. The multiple cameras (5) are arranged around the loading device (4) to simultaneously acquire surface images of the specimen from different angles, and the computing workstation (10) reconstructs the surface images from different angles and synthesizes the surface contour image.

Images