High-precision complex conglomerate three-dimensional digital model reconstruction method

Abstract

The invention discloses a high-precision complex conglomerate three-dimensional digital model reconstruction method, which comprises the following steps of: sampling a natural conglomerate on site to obtain an enough number of loose single gravels; performing three-dimensional CT scanning on each gravel to obtain a three-dimensional digital model of a single gravel; grouping the single gravels according to shape features, and obtaining grading features of the gravels in the natural conglomerate; performing rotation and scaling operation on each gravel three-dimensional digital model to obtain gravel three-dimensional digital models with different sizes and inclination angles; establishing a three-dimensional rock mass digital model; inserting each group of gravels in the three-dimensional rock mass digital model in batches according to grading characteristics; and establishing and outputting a high-precision complex conglomerate three-dimensional digital model slice. According to the three-dimensional digital model reconstruction method, the high-precision complex conglomerate three-dimensional digital model similar to the natural complex conglomerate in feature can be accurately established, and on the basis, destructive tests of the high-precision complex conglomerate three-dimensional digital model completely consistent in structure under various loading conditions can be carried out.

Description

Technical Field

[0001] The present invention belongs to the technical field of rock mass engineering and relates to a method for reconstructing a three-dimensional digital model of high-precision complex conglomerate. Background Art

[0002] With the rapid development of the world economy, the mineral resources in the shallow part of the earth are gradually running out, and energy development is continuously moving towards the deep part of the earth. The exploration and exploitation of deep conglomerate reservoirs on the earth are currently hot topics in the global oil and gas industry. Conglomerate is a heterogeneous geological material composed of gravel, matrix material, initial fractures, etc. It exhibits strong heterogeneous characteristics due to its complex mesoscopic structure inside. The special mechanical properties and failure modes of conglomerate also have an important impact on the safety and stability assessment of related projects.

[0003] At present, the research on the mechanical properties of conglomerate at home and abroad is still in its infancy. The method of combining three-dimensional reconstruction technology and numerical simulation is mostly used to carry out related destructive test studies on conglomerate. For example, the paper titled Cross-scale analysis on the mechanical behavior of the conglomerate of Yulin Grottoes in China published by Zhang Huihui et al. in the Internation journal of rock mechanics and mining sciences in 2024. Zhang et al. roughly reconstructed the conglomerate two-dimensionally using modeling software such as CAD according to the spatial structure characteristics of natural conglomerate. However, there are large differences in the rock mineral components inside the conglomerate, obvious differences in the mechanical properties of each component, and different particle sizes. It is very difficult for the reconstructed model to accurately reflect the fine structural characteristics inside the conglomerate, and most of the reconstructed models are two-dimensional, ignoring the spatial structure characteristics of the conglomerate, resulting in a large difference between the mechanical properties and failure modes of the conglomerate obtained by numerical simulation and the conclusions observed in reality.

[0004] Therefore, the reconstruction of a three-dimensional digital model of high-precision complex conglomerate is difficult to achieve due to the complexity of its modeling process. Summary of the Invention

[0005] The present invention provides a method for reconstructing a three-dimensional digital model of high-precision complex conglomerate, solves the technical problem of difficult modeling of a three-dimensional digital model of high-precision complex conglomerate, realizes the mechanical tests of complex conglomerate under different loading conditions with different gravel gradations, and finally obtains the mechanical properties and failure characteristics of a conglomerate model with a structure highly similar to that of natural conglomerate, providing a guiding technical effect for the related exploitation of oil and gas projects.

[0006] The technical solution of the present invention:

[0007] A method for reconstructing a three-dimensional digital model of high-precision complex conglomerate, comprising the following steps:

[0008] Take on-site samples of natural conglomerate to obtain a sufficient number of loose single-grain gravels;

[0009] Perform three-dimensional CT scanning on the gravels to obtain three-dimensional digital models of individual gravels;

[0010] Group the individual gravels according to their shape characteristics to obtain the grading characteristics of each gravel in the natural conglomerate;

[0011] Perform rotation and scaling operations on the three-dimensional digital models of each gravel to obtain three-dimensional digital models of gravels with different sizes and inclinations;

[0012] Establish a three-dimensional rock mass digital model;

[0013] Insert each group of gravels into the three-dimensional rock mass digital model in batches according to the grading characteristics;

[0014] Establish and output slices of the three-dimensional digital model of high-precision complex conglomerate.

[0015] Preferably, taking on-site samples of natural conglomerate to obtain a sufficient number of loose single-grain gravels includes:

[0016] Select natural conglomerate with complete structure and take on-site samples;

[0017] Among them, the complete structure of natural conglomerate means that there are no obvious geological structures such as large faults in the conglomerate, and there are no obvious fractures in the internal gravels;

[0018] Perform a destructive loading test on the obtained natural conglomerate specimen, and collect all the single gravels that fall after the specimen undergoes final failure, ensuring that the number of single gravels collected is sufficient;

[0019] Among them, the destructive loading test refers to a mechanical test that can cause large-scale damage to the conglomerate specimen until it has no bearing capacity;

[0020] All the single gravels collected should cover the main structural characteristics of the internal gravels of the natural conglomerate.

[0021] Preferably, performing three-dimensional CT scanning on each gravel to obtain a three-dimensional digital model of an individual gravel includes:

[0022] Clean each gravel to remove surface debris and foreign objects;

[0023] Perform three-dimensional CT scanning on each cleaned gravel one by one to obtain three-dimensional digital model files of all single gravels;

[0024] Among them, the three-dimensional digital model file should be a file with the suffix of.stl or a three-dimensional point cloud data file;

[0025] Process the three-dimensional digital model files of each gravel and convert them into three-dimensional digital matrices.

[0026] Preferably, the individual gravels are grouped according to their shape characteristics to obtain the grading characteristics of each gravel in the natural conglomerate, including:

[0027] Statistical analysis of the major axis dimensions of all single gravels;

[0028] Among them, the major axis dimension of a single gravel refers to the length of the single gravel along the longest direction;

[0029] Group according to the major axis dimensions of the gravels;

[0030] Among them, the major axis lengths of the gravels within each group are all within a certain same range.

[0031] Preferably, the three-dimensional digital models of the gravels are rotated and scaled to obtain three-dimensional digital models of the gravels with different sizes and inclinations, including:

[0032] Statistical analysis of the inclination range of each gravel in the natural conglomerate;

[0033] Perform rotation and scaling operations on the three-dimensional digital models of the gravels according to the inclination range and major axis dimension range of each group of gravels;

[0034] Among them, the models generated for each group of gravels should ensure a sufficient quantity to meet the requirements of multiple sizes and multiple inclinations;

[0035] Save all the generated three-dimensional gravel models for subsequent use.

[0036] Preferably, the establishment of the three-dimensional rock mass digital model includes:

[0037] Establish a three-dimensional digital model of the complete rock mass;

[0038] Among them, the three-dimensional digital model should be stored as a three-dimensional digital matrix, and the complete rock mass refers to a rock mass without any defects and with the material properties of any internal point being completely consistent.

[0039] Preferably, the insertion of each group of gravels into the three-dimensional rock mass digital model according to the grading characteristics includes:

[0040] Determine the grading characteristics of each gravel in the to-be-built conglomerate model according to the geological exploration results;

[0041] Insert into the three-dimensional rock mass digital model step by step in descending order of the major axis dimensions of each gravel group;

[0042] Among them, in order to increase the final filling rate of the model, the gravels at all levels should be inserted in descending order of the major axis size. In addition, collision detection should be performed during insertion to prevent the gravels from overlapping and crossing each other.

[0043] Preferably, the establishment and output of the slices of the high-precision complex conglomerate three-dimensional digital model include:

[0044] Determine the number of slices that the model needs to be segmented into;

[0045] Convert the format of the filled conglomerate model and output the slices;

[0046] Among them, all operations of file format conversion should be processed by programming software and superimposed by boolean operations of matrices.

[0047] Based on the same inventive concept, the present application also provides a high-precision complex conglomerate three-dimensional digital model obtained by the above-mentioned high-precision complex conglomerate three-dimensional digital model reconstruction method.

[0048] Advantages of the present invention: The high-precision complex conglomerate three-dimensional digital model reconstruction method provided by the present invention can accurately establish a complex conglomerate three-dimensional digital model that is highly similar to the in-situ conglomerate structure. By conducting surveys, sampling, and CT scans on the conglomerate at the engineering site to be established, the three-dimensional structure and grading characteristics of each gravel in the conglomerate are obtained. Then, each group of gravels is inserted step by step according to the grading in the complete rock mass through programming-based modeling, realizing the reconstruction of a three-dimensional model of a conglomerate that is highly similar to the natural conglomerate structure at the engineering site. Therefore, on this basis, destructive tests of the high-precision complex conglomerate three-dimensional digital model with exactly the same structure can be carried out under different loading conditions, solving the problem of difficult three-dimensional digital modeling of high-precision complex conglomerates, improving the understanding of the mechanical properties, failure modes, and excavation characteristics of high-precision complex conglomerates, and providing a reference for the design and safety stability evaluation of related oil and gas engineering such as exploration and exploitation of deep-earth conglomerate reservoirs. In this way, the problem of difficult three-dimensional digital modeling of high-precision complex conglomerates is effectively solved, realizing the destructive tests of the high-precision complex conglomerate three-dimensional digital model with exactly the same structure under different loading conditions and grading conditions, and finally obtaining the mechanical properties, failure modes, and excavation characteristics of high-precision complex conglomerates, providing a technical effect of guiding the design of related oil and gas engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic flow chart of a high-precision complex conglomerate three-dimensional digital model reconstruction method provided by an embodiment of the present invention;

[0050] Figure 2 It is a natural conglomerate sample obtained from the engineering site provided by an embodiment of the present invention;

[0051] Figure 3 The failure mode of the conglomerate specimen provided by the embodiment of the present invention after the uniaxial compression test;

[0052] Figure 4 The grading grouping of single washed gravel according to the long-axis size provided by the embodiment of the present invention, where (a) is the conglomerate block collected from the engineering site, (b) is the size of each gravel, and (c) is the gravel group arranged according to the size.

[0053] Figure 5 The three-dimensional matrix data structure diagram of the storage gravel space structure provided by the embodiment of the present invention.

[0054] Figure 6 The schematic diagram of converting the gravel stl file into a three-dimensional matrix provided by the embodiment of the present invention. Among them, (a) is the shape of the gravel displayed in the stl format, (b) is the three-dimensional matrix of the gravel under a larger unit size, and (c) is the three-dimensional matrix of the gravel under a smaller unit size.

[0055] Figure 7 The schematic diagram of converting the gravel point cloud file into a three-dimensional matrix provided by the embodiment of the present invention. Among them, (a) is the shape of the gravel displayed in the point cloud format, and (b) is the three-dimensional matrix of the gravel under a smaller unit size.

[0056] Figure 8 The schematic diagram of improving the accuracy of the three-dimensional matrix provided by the embodiment of the present invention through element encryption and size reduction. Among them, (a) is the three-dimensional matrix of the gravel under a larger unit size, (b) is the three-dimensional matrix of the gravel under a smaller unit size, and (c) is the size reduction schematic diagram.

[0057] Figure 9 The schematic diagram of the rotation of the gravel three-dimensional matrix provided by the embodiment of the present invention. Among them, (a) is the original three-dimensional matrix of the gravel, and (b) is the three-dimensional matrix rotated counterclockwise by 45°.

[0058] Figure 10 The schematic diagram of the scaling of the gravel three-dimensional matrix provided by the embodiment of the present invention. Among them, (a) is the original three-dimensional matrix of the gravel, and (b) is the three-dimensional matrix of the gravel reduced to 1 / 2.

[0059] Figure 11 The three-dimensional digital models of each group of gravels under different inclinations and sizes provided by the embodiment of the present invention. Among them, (a) is the three-dimensional view of the gravel group, (b) is the front view of the gravel group, and (c) is the top view of the gravel group.

[0060] Figure 12 The schematic diagram of the process of inserting gravel into the three-dimensional matrix provided by the embodiment of the present invention.

[0061] Figure 13The diagram of the step-by-step filling process of the conglomerate model provided by the embodiment of the present invention;

[0062] Figure 14 The diagram of the high-precision complex conglomerate three-dimensional digital model provided by the embodiment of the present invention, where (a) is the three-dimensional model of the conglomerate, (b) is the top view of the model, (c) is the front view of the model, and (d) is the left view of the model. Specific embodiments

[0063] The following further describes the specific embodiments of the present invention in combination with the drawings and technical solutions.

[0064] See Figure 1 , a method for reconstructing a high-precision complex conglomerate three-dimensional digital model provided by this embodiment includes the following steps:

[0065] Step 101, conduct on-site sampling of natural conglomerate to obtain a sufficient number of loose single-grain gravels.

[0066] Select natural conglomerate with a complete structure and conduct on-site sampling. Conduct a destructive loading test on the obtained natural conglomerate specimen, and collect all the single gravels that fall after the specimen undergoes final failure.

[0067] Specifically, first select the natural conglomerate near the Mogao Grottoes in Dunhuang City, Gansu Province, China as the survey site, and the survey results of the conglomerate can be seen in Figure 2 . The inside of the conglomerate is mainly formed by stacking loose gravels of different sizes and shapes, with several pores inside. The cross-sectional structure of the gravels is mostly round irregular quasi-elliptical. In addition, the gaps between the gravels are filled with matrix components;

[0068] Then, through on-site sampling, a natural conglomerate cuboid specimen with a size of 5 cm × 5 cm × 10 cm is obtained. Use an electro-hydraulic servo loading testing machine to conduct a uniaxial compression failure test on the specimen, and end the loading when the specimen completely loses its bearing capacity. The damaged specimen can be seen in Figure 3 shown. Cracks mainly occur and propagate in the matrix, but the expansion of cracks is also found in a few gravels;

[0069] Furthermore, sort out the damaged specimen and collect the intact loose single gravels inside the specimen.

[0070] Optionally, the selection of natural conglomerate can be carried out at any engineering site, and the number of collected loose single gravels only needs to be able to reflect the main three-dimensional characteristics of the gravels in the natural conglomerate. The number can be many and there is no limit. The shape of the obtained natural conglomerate specimen can be a cube, a cuboid or a cylinder.

[0071] Specifically, in this embodiment, the structural feature of the natural conglomerate specimen is a cuboid. Considering the fragility of the natural conglomerate structure, the cutting method is adopted to obtain a complete specimen. In addition, the number of loose single gravels obtained is 21.

[0072] Step 102: Perform three-dimensional CT scans on each gravel to obtain a three-dimensional digital model of a single gravel.

[0073] Perform a cleaning operation on each gravel to remove surface debris and foreign objects. Next, perform three-dimensional CT scans on each cleaned gravel one by one to obtain three-dimensional digital model files of all single gravels.

[0074] Specifically, use a towel to wipe the obtained loose single gravels to remove the dust and debris on the surface of the gravels. Be careful enough during the wiping process because although some gravels appear intact on the outside, they have been damaged in the uniaxial compression destructive test. The loose single gravels after wiping are shown in Figure 4 as shown;

[0075] Next, use a three-dimensional CT scanning machine to scan each gravel one by one. Before scanning, place the single gravel at the center position of the scanning tray. Ensure the stability of the gravel during scanning and prevent phenomena such as tipping or moving during the scanning process;

[0076] Furthermore, obtain the three-dimensional digital model file of a single gravel through the built-in processing system of the three-dimensional CT scanning machine. The three-dimensional digital model information of 21 gravels is shown in Table 1. The three-dimensional shape of the gravels is mainly cobblestone-shaped, and some are irregular polyhedron-shaped. Among them, the cobblestone-shaped gravels are further divided into flat-shaped and round-shaped. Next, use programming software to process the three-dimensional model files of each gravel and convert them into three-dimensional digital matrices for storing the single gravel models.

[0077] Table 1: Gravel Archive Table, where X is the major axis, Y is the minor axis, and Z is the thickness, with the unit of mm

[0078]

[0079] The three-dimensional digital matrix J of the gravel represents the spatial distribution of the gravel using a three-dimensional digital matrix. The size of the entire matrix is u*v*w, where u is the number of rows of the three-dimensional matrix, v is the number of columns of the three-dimensional matrix, and w is the number of layers of the three-dimensional matrix. J(u0, v0, w0) is the value of the three-dimensional digital matrix J when the row is u0, the column is v0, and the layer is w0. The value of J(u0, v0, w0) is 0 or 1. When J takes 0, it means that there is no substance at that position, and vice versa. As Figure 5 shown, where the positions with a value of 1 are connected together to form an irregular polygon pattern.

[0080] Regardless of whether the three-dimensional data of the gravel is stored in an stl file or point cloud data, it can depict the external contour of the gravel. Using the external contour of the gravel, the three-dimensional structure of the gravel can be stored in the form of a three-dimensional matrix.

[0081] When the gravel is stored in an stl file, we first obtain the contour of the gravel through programming processing, as shown in Figure 6 Figure (a). Next, we construct a three-dimensional matrix that is slightly larger in three dimensions than the gravel, where the values of the matrix are all 0, and at this time the matrix is in a state of unoccupied material. Although the three-dimensional matrix itself has no physical size, when we use the three-dimensional matrix to represent the gravel, we endow the matrix with the concept of physical size. If the maximum value of the three-dimensional size of the gravel is about 7 cm, we then construct a three-dimensional matrix with a size of 10 cm * 10 cm * 10 cm. When the u*v*w of the matrix is set to 5, 5, 5, it means that each unit in the three-dimensional matrix represents a physical size of 2 cm * 2 cm * 2 cm. When u*v*w is set to 10, 10, 10, it means that each unit in the three-dimensional matrix represents a physical size of 1 cm * 1 cm * 1 cm.

[0082] When representing the spatial structure of the conglomerate using a three-dimensional matrix, we first place the conglomerate in the central part of the three-dimensional matrix, not necessarily the strict center, as long as the three-dimensional matrix can completely enclose the gravel. Then, we index each unit of the three-dimensional matrix. When the position of a certain unit is within the gravel, the unit at that position is assigned a value of 1 (indicating that there is gravel there), and conversely, it is assigned a value of 0 (indicating that there is no material there).

[0083] As can be seen from Figures (b) and (c), the size of the three-dimensional matrix controls the accuracy of the three-dimensional structure of the reconstructed gravel. When a larger three-dimensional matrix is used for representation, the result is better and the accuracy is higher, and the fine structural features of the gravel can be reconstructed. When a smaller three-dimensional matrix is used for representation, the local structural features of the gravel will be ignored, affecting the accuracy of the final reconstruction.

[0084] When the gravel is stored in a point cloud file, the method we use is similar to the processing method of the stl file. However, at this time, there is no need to determine the contour of the gravel, but directly process all the points that make up the gravel. Similarly, we first construct a three-dimensional matrix with a physical size slightly larger than the gravel, and then select the number of units of the three-dimensional matrix: u*v*w. Then index all the points Point that make up the gravel, determine the unit in the three-dimensional matrix where the position of the point Point(i) (i is the serial number of each point) is located, and then assign a value of 1 to this unit. Repeat this operation until all the points are indexed, as shown in Figure 7 shown. When performing this operation, a parallel processing method can be used to speed up the calculation speed.

[0085] Optionally, three-dimensional CT scanning operations can be performed on any number of gravels, and the more the better. In addition, the three-dimensional CT scanning machine should meet the accuracy requirements to ensure the accuracy of the three-dimensional electronic model of a single gravel. Any three-dimensional CT scanning machine can be arbitrarily selected on the premise of meeting the basic accuracy requirements. The three-dimensional digital model file of a single gravel should be a file with the suffix of (.stl) or a three-dimensional point cloud data file (.pts). Any programming software can be used to process the.stl file or the three-dimensional point cloud file to obtain the three-dimensional matrix storage format of a single gravel. The number of cells in the three-dimensional matrix can be selected according to the accuracy requirements. If higher accuracy is required, the number of cells should be increased.

[0086] Specifically, in this embodiment, a large three-dimensional CT scanning machine with high precision is used to perform three-dimensional scanning operations on gravels. Considering the time cost and money consumption, a total of 21 single gravels are scanned. The three-dimensional digital model file of a single gravel is saved as an.stl file, and the MATLAB software is used to process the.stl file to convert it into a three-dimensional point cloud file, and finally converted into a three-dimensional matrix for storage. Since the sizes of each gravel are different, in order to obtain higher accuracy, the size of the three-dimensional matrix cell used is 1mm * 1mm * 1mm.

[0087] Step 103: Group the single gravels according to their shape characteristics to obtain the grading characteristics of each gravel in the natural conglomerate.

[0088] Statistically analyze the spatial dimensions of all single gravels and group them according to the long-axis dimensions of the gravels.

[0089] Specifically, place each single gravel on the desktop one by one, and use a ruler to measure the lengths of the single gravel in three directions. After obtaining the three size characteristics of each gravel, take the maximum value of the three lengths of the gravel as the representative size of the single gravel;

[0090] Furthermore, according to the representative size of the single gravel, all single gravels are sorted from largest to smallest according to the representative size for grouping. After grouping is completed, the gravels in each group are placed in the same container respectively. As shown in Figure 4 After grouping is completed, the representative sizes of the gravels in each group should be close to each other.

[0091] Optionally, different methods can be adopted for grouping the gravels, as long as all single gravels can be regularly distributed.

[0092] Specifically, in this embodiment, when determining the three length dimensions of a single gravel, the long-axis dimension is selected first. The long-axis dimension of a single gravel refers to the length of the single gravel along the longest direction, and then the other two dimensions are determined. The directions of the other two dimensions are perpendicular to the long-axis direction.

[0093] Step 104: Rotate and scale each three-dimensional digital model of gravel to obtain three-dimensional digital models of gravel at different sizes and inclination angles.

[0094] Statistically analyze the inclination angle ranges of each gravel in natural conglomerate, and perform rotation and scaling operations on the three-dimensional digital models of each gravel according to the inclination angle ranges and long-axis size ranges of each group of gravel.

[0095] Specifically, through on-site investigation of natural conglomerate, obtain the rotation range {a1 - a2, b1 - b2, c1 - c2} of each gravel in the conglomerate in space, where a1, a2, b1, b2, c1, and c2 are the upper and lower thresholds of the rotatable angle ranges in three directions of the gravel in the space coordinate system respectively;

[0096] Next, according to the grading characteristics of the gravel, obtain the size ranges {s i,min , s i,max} of each group of gravel. Where i is the number of each gravel group, and the value range of i is from 1 to 6, corresponding to 6 different gravel groups respectively. s i,min , s i,max are the upper and lower thresholds of the size ranges of each group of gravel respectively;

[0097] Furthermore, use MATLAB software to perform rotation and scaling operations on each group of gravel. Among them, perform 100 operations on each of the 6 groups of gravel, for a total of 600 operations. In each operation of each gravel group, first randomly select 1 gravel from this gravel group, and randomly generate a size: s and an inclination angle group: a, b, c within the corresponding size range {s i,min , s i,max} and rotation range {a1 - a2, b1 - b2, c1 - c2} of this gravel group. Then use MATLAB software to program to achieve matrix rotation and matrix scaling operations on the three-dimensional matrix form of the gravel model, so as to achieve rotation and scaling processing of the three-dimensional model of the gravel.

[0098] For the rotation operation, first record the indices (u i , v i , w i ) of all units with a value of 1 in the three-dimensional matrix representing the gravel. i is the number of all units with a value of 1. Next, perform encryption processing on these units, as shown in Figure 8 . After the encryption processing, record the unit indices and physical sizes of the new three-dimensional matrix. Since the size of a single unit is reduced, the corresponding physical size also needs to be reduced. In order to ensure the accuracy during the subsequent three-dimensional matrix rotation, as many split units as possible should be added during the encryption processing, such as 2 3 , 3 3 , 43 times

[0099] After encryption processing, the three-dimensional matrix needs to be rotated in three directions. Assume that the rotation of the three-dimensional matrix is around a certain rotation axis, and the rotation axis is represented by the unit vector N = (N x , N y , N z ). The rotation angle is θ, the physical coordinates of the matrix element to be rotated are P(x, y, z), and after rotation, it becomes P'(x', y', z'). Regarding the physical coordinates of the elements in the three-dimensional matrix, we can calculate them according to the physical size of the elements in the matrix. Just calculate with the coordinates of the center element as the origin (0, 0, 0), and the physical coordinates of each element in the matrix are taken as the center position of the element.

[0100] The rotation calculation formula is as follows:

[0101] P' = P·cos(θ) + (u×P)·sin(θ) + u(u·P)(1 - cos(θ))

[0102] Since the rotation is around the X, Y, and Z coordinate axes, N is (1, 0, 0), (0, 1, 0), and (0, 0, 1) respectively. By calculating all the elements in the three-dimensional matrix, the physical coordinates of all the rotated elements can be obtained. Next, decode the rotated physical coordinates, that is, map these physical coordinates to the corresponding matrix elements. In this step, still use the method of converting point cloud data into a three-dimensional matrix. By traversing the rotated physical coordinates, assign the corresponding three-dimensional matrix element a value of 1. Note that in this assignment stage, the scale of the three-dimensional matrix element before rotation should be used, rather than the encrypted one. The encrypted elements are only used to ensure their accuracy during rotation. The rotated three-dimensional matrix is as Figure 9 shown.

[0103] The scaling process of the matrix is also carried out in a similar way. First, the rotated matrix is encrypted, then the physical coordinates are determined, and then the scaling function is used for scaling calculation. When scaling, the central unit of the matrix is also taken as the origin (0, 0, 0), and the physical coordinates T(x, y, z) of each encrypted unit are calculated, as well as the scaling factor S' = s / Dimension(max(u, v, w)). Here, s is the required gravel size, and max(u, v, w) is the largest value among the three scales of the current gravel three-dimensional matrix. For example, when the three-dimensional matrix is 50 rows × 78 columns × 26 layers, max(u, v, w) is taken as 78. Dimension(max(u, v, w)) is to obtain the total physical size of these 78 units. If the size of each unit is 1 mm, then Dimension(max(u, v, w)) is taken as 78 mm. Suppose the required gravel size is 25 mm, then S' is calculated as 25 mm / 78 mm = 0.3205. The new physical coordinates after scaling are T'(x', y', z') = S' · T(x, y, z). After that, the coordinate decoding work is carried out, and finally the scaled three-dimensional matrix is obtained through indexing, such as Figure 10 as shown

[0104] Finally, the matrices of 100 three-dimensional models generated by each gravel group are stored in the database respectively. The database includes 6 groups of gravel three-dimensional models, and each group contains 100 matrix format data of single gravel three-dimensional models. The 6 groups of a total of 600 gravel three-dimensional models established are shown in the appendix Figure 11 as shown

[0105] Optionally, the determination of the rotation inclination angle and the size range of each gravel group can be set artificially and does not depend on the results of on-site engineering surveys. The number of operations of each gravel group can be set arbitrarily. The more operations, the richer the established database. The rotation and scaling operations of the gravel three-dimensional model matrix format can be carried out using different programming software, such as MATLAB, Python, R, or C++ language

[0106] Specifically, in this embodiment, MATLAB software is used for the scaling and rotation operations of the single gravel three-dimensional digital model matrix format. In addition, each gravel group has carried out 100 operations, which is a choice considering the computing speed of the computer and the richness of the database

[0107] Step 105, establish a three-dimensional rock mass digital model

[0108] Establish a three-dimensional digital model of the complete rock mass

[0109] Specifically, determine the sizes d1, d2, d3 of the conglomerate model to be established, where the shape of the conglomerate model is a three-dimensional cuboid of d1 × d2 × d3

[0110] Further, a corresponding three-dimensional matrix is established in MATLAB to store the complete three-dimensional rock mass model, where the size of the three-dimensional matrix corresponds to the size of the conglomerate model, that is, the dimensions of the three-dimensional matrix are d1×d2×d3, and d1, d2, and d3 are the number of rows, columns, and layers of the three-dimensional matrix respectively, and their values are all assigned as t;

[0111] Optionally, different software can be used to establish the three-dimensional digital model of the complete rock mass, such as: MATLAB, Python, R, C++, etc.

[0112] Specifically, in this embodiment, MATLAB is used to establish the matrix of the corresponding complete three-dimensional rock mass model, where the model sizes d1, d2, and d3 of the conglomerate are 600mm, 600mm, and 1200mm respectively, and the corresponding size of the three-dimensional matrix is 600×600×1200, that is, each matrix pixel point corresponds to a length of 1mm.

[0113] Step 106: Insert each group of gravels into the three-dimensional rock mass digital model in batches according to the grading characteristics.

[0114] According to the geological exploration results, determine the grading characteristics and mineral species characteristics of each gravel in the to-be-built conglomerate model, and insert the gravels under each mineral composition into the three-dimensional rock mass digital model step by step in descending order of the long axis size of each gravel group.

[0115] Specifically, first, determine the grading of the complex high-precision conglomerate model to be established according to the results of the geological exploration of natural conglomerate at the engineering site. The gravels in the natural conglomerate are divided into 6 categories, namely A, B, C, D, E, and F, according to their size characteristics. The ranges of the long axis sizes of each category of gravels are >50mm, 23 - 40mm, 15 - 22mm, 7.5 - 11mm, 2.3 - 4mm, and 1.15 - 2mm respectively. Among them, the shapes of the gravels in groups A, E, and F include 3 kinds, namely 1, 2, and 5; the shapes of the gravels in group B include 8 kinds, namely 3, 6, 7, 8, 9, 10, 16, and 17; the shapes of the gravels in groups C and D include 9 kinds, namely 0, 4, 11, 12, 13, 14, 15, 18, and 20. This grading characteristic is the characteristic of each group of gravels in the gravel database established in step 104, as shown in Table 2.

[0116] Table 2: Information characteristics of 6 kinds of gravel groups

[0117]

[0118] Next, determine the mineral composition within the natural conglomerate. A total of 4 different minerals and pores are found in the surveyed natural conglomerate. Mineral 1 consists of gravel particles of 5 groups, namely A, B, C, D, and E. The spatial proportion of each group of gravel particles is approximately 2%, 6%, 5%, 5%, and 6%. Mineral 2 consists of gravel particles of 5 groups, namely A, B, C, D, and E. The spatial proportion of each group of gravel particles is approximately 2%, 4%, 11%, 10%, and 6%. Mineral 3 consists of gravel particles of 5 groups, namely A, B, C, D, and E. The spatial proportion of each group of gravel particles is approximately 1%, 10%, 10%, 5%, and 3%. Mineral 4 consists of gravel particles of 2 groups, namely E and F. The spatial proportion of each group of gravel particles is approximately 5% and 5%. The spatial proportion of the pores is approximately 4%. The specific gravel composition of each mineral is shown in Table 3.

[0119] Table 3: Information Characteristics of Mineral Composition

[0120]

[0121] Furthermore, use the MATLAB programming software to perform the filling operation of gravel at all levels. Since there are 4 different minerals, use k1, k2, k3, and k4 to represent each group of minerals. Among them, k1, k2, k3, and k4 are 4 integers in the range of 0 - 255, and they should be distributed relatively evenly and dispersedly to facilitate easier differentiation of different minerals. When filling, use the method of collision detection to ensure that the newly inserted gravel does not overlap or collide with the existing gravel. Take the simplest one-dimensional space as an example. The three-dimensional complete rock mass matrix is t×ones(d1, d2, d3), where t×ones(d1, d2, d3) is a three-dimensional matrix with all elements of size d1×d2×d3 being t. If the matrix data of the space to be filled is represented as A = [t tk1 k1k1k1t tttk1], and the newly inserted gravel belongs to Mineral 2, denoted as B = [k2 k20 00 0k2 k2k2k20], then perform an overlap detection on these two one-dimensional matrices. First, set all t in the A matrix to 0. Next, perform the multiplication operation on the corresponding matrix elements of A and B matrices and then sum, that is, sum = ∑(A i ×B i ), where i is the index of the matrix. In these two matrices, i takes values from 1 to 11. If sum is 0, it is considered that there is no overlap between the newly inserted gravel and the gravel in the space to be filled, and then this gravel B can be inserted into the space to be filled. If sum is a non-zero positive integer, it indicates that there is a collision and overlap between the newly inserted gravel and the space to be inserted, and then this gravel B cannot be inserted. When inserting gravel B into the space to be filled A, only need to replace the part with the value of k2 in the B matrix at the corresponding position in the A matrix.

[0122] Figure 12It is a schematic diagram of operations under a two-dimensional matrix. Here, Ω is the space where gravel is to be inserted. There are already three different types of gravel, k1, k2, and k3, inside it. Ω is extracted from the overall conglomerate matrix. Now, it is necessary to insert gravel k4 into this space. By assigning 0 to t within Ω, and then performing a dot product on the matrix to which Ω belongs and the matrix of gravel k4, if the sum of the dot product is 0, it means that gravel k4 can be inserted into the Ω space without overlapping with other gravel. Next, the value within the unit of gravel k4 in the Ω space can be modified from t to k4.

[0123] Finally, fill in according to the order of the gravel sizes from large to small. First, fill in the A particles in Mineral 1, Mineral 2, Mineral 3, and Mineral 4. Next are the B particles, C particles, D particles, E particles, and F particles in the four minerals.

[0124] Optionally, 3D digital models of the complete rock mass can be established using different software, such as: MATLAB, Python, R, C++, etc. The division of minerals can be set according to personal needs. If it is a natural conglomerate of a single mineral, one mineral can be set. k1, k2, k3, and k4 can be set arbitrarily as long as the values are different. Different methods can be used to judge the overlap of gravel. This embodiment only provides a relatively simple and computationally efficient judgment method. When filling gravel, it can be inserted in any order, such as filling in according to the order of the gravel sizes from small to large, or filling in in any order such as E, F, A, C, D, B, etc.

[0125] Specifically, in this embodiment, MATLAB software is used for the overlap detection of gravel and the host filling operation. In addition, k1, k2, k3, and k4 are taken as 50, 100, 150, and 200, so that the numerical differences of the four minerals are relatively large, which is easy to make different types of judgments. In this embodiment, in order to increase the final filling rate of the model, when inserting gravel at each level, it is inserted in the order of the major axis size from large to small. The process of gradually inserting gravel and the finally established high-precision 3D conglomerate model are shown in Figure 13 and 14 shown.

[0126] Step 107, establishment and output of slices of the high-precision complex conglomerate 3D digital model.

[0127] Determine the number of slices that the model needs to be segmented into, perform format conversion on the filled conglomerate model, and perform slice output. Among them, all operations of file format conversion should be processed by programming software and superimposed through the Boolean operation of matrices.

[0128] Before output, the accuracy of the reconstructed 3D conglomerate model can be improved again by means of unit encryption to meet the high-precision requirements.

[0129] Specifically, the three-dimensional digital model of high-precision complex conglomerate can be sliced through three-dimensional modeling software and then imported into other three-dimensional modeling software for subsequent processing, thus obtaining an editable three-dimensional digital model of high-precision complex conglomerate.

[0130] Furthermore, the established three-dimensional digital model of high-precision complex conglomerate can be used for finite element analysis or 3D printing of the model.

[0131] Optionally, the three-dimensional modeling software can be selected as the software that supports slice import.

[0132] Specifically, in this embodiment, a three-dimensional numerical model of high-precision complex conglomerate is established by importing slices using finite element analysis software, and force analysis and destructive analysis can be performed based on this three-dimensional numerical model.

[0133] As can be seen from the above embodiments, the method for reconstructing the three-dimensional digital model of high-precision complex conglomerate provided by this application first obtains a complete sample of natural conglomerate through on-site sampling of natural conglomerate, and conducts a destructive loading test on this sample to obtain a sufficient number of loose single-grain conglomerates; next, cleaning and three-dimensional CT scanning operations are performed on each single-grain conglomerate to obtain a sufficient number of three-dimensional digital models of single-grain conglomerates; the three-dimensional digital models of each conglomerate are rotated and scaled through programming software to obtain three-dimensional digital models of conglomerates with different sizes and inclinations; then a complete three-dimensional rock mass model is established using programming software, and all the conglomerates are inserted into the complete rock mass model in sequence according to the grouping of conglomerates in descending order of major axis size through overlap detection of the conglomerates; finally, the established three-dimensional digital model of high-precision complex conglomerate is sliced to obtain slices of the three-dimensional digital model of high-precision complex conglomerate, and the slices are imported into other three-dimensional modeling software for subsequent editing processing or imported into finite element analysis software for force analysis and destructive simulation.

[0134] Through the method of the present invention, destructive tests of a three-dimensional digital model of high-precision complex conglomerate with completely consistent structure under different loading conditions can be realized by using numerical simulation technology, solving the problem of difficult modeling of the three-dimensional digital model of high-precision complex conglomerate, improving the understanding of the mechanical properties and failure modes of high-precision complex conglomerate, and providing a reference for carrying out relevant oil and gas engineering design, exploitation and safety stability evaluation in complex conglomerate. Therefore, the method of the present invention can improve the understanding of the mechanical properties and failure modes of high-precision complex conglomerate and provide a reference for relevant oil and gas engineering design and safety stability evaluation.

[0135] The specific embodiments described above further elaborate on the objective, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only for the specific embodiments of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for reconstructing a three-dimensional digital model of high-precision complex conglomerate, characterized in that, It includes the following steps: Take on-site samples of natural conglomerate to obtain a sufficient number of loose single-grain gravels; Perform three-dimensional CT scans on the gravels to obtain three-dimensional digital models of individual gravels; Group the individual gravels according to their shape characteristics to obtain the grading characteristics of each gravel in the natural conglomerate; Perform rotation and scaling operations on the three-dimensional digital models of each gravel to obtain three-dimensional digital models of gravels at different sizes and inclinations; Establish a three-dimensional rock mass digital model; Insert each group of gravels into the three-dimensional rock mass digital model in batches according to the grading characteristics; Establish and output slices of the high-precision complex conglomerate three-dimensional digital model.

2. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, wherein Take on-site samples of natural conglomerate to obtain a sufficient number of loose single-grain gravels, including: Select natural conglomerate with a complete structure and take on-site samples; Among them, the complete structure of natural conglomerate means that there are no obvious large faults in the conglomerate and the internal gravels have no obvious fractures; Perform a destructive loading test on the obtained natural conglomerate, and collect all the single gravels that fall after the natural conglomerate undergoes final failure, ensuring that the number of single gravels collected is sufficient; Among them, the destructive loading test refers to a mechanical test that can cause large-scale damage to the natural conglomerate until it has no bearing capacity; All the single gravels collected should cover the main structural characteristics of the internal gravels of the natural conglomerate.

3. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, wherein Perform three-dimensional CT scans on each gravel to obtain three-dimensional digital models of individual gravels, including: Clean each gravel to remove surface debris and foreign objects; Perform three-dimensional CT scans on each cleaned gravel one by one to obtain three-dimensional digital model files of all single gravels; Among them, the three-dimensional digital model files should be files with the.stl suffix or three-dimensional point cloud data files; Process the three-dimensional digital model files of each gravel and convert them into three-dimensional digital matrices.

4. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, characterized in that Preferably, the grouping of individual gravels according to their shape characteristics to obtain the grading characteristics of each gravel in the natural conglomerate includes: Statistical analysis of the long-axis dimensions of all single gravels; Among them, the long-axis dimension of a single gravel refers to the length of the single gravel along the longest direction; Group according to the long-axis dimensions of the gravels; Among them, the long-axis lengths of the gravels within each group are within a certain same range.

5. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, wherein The rotation and scaling operations on the three-dimensional digital models of each gravel to obtain three-dimensional digital models of gravels at different sizes and inclinations include: Statistical analysis of the inclination range of each gravel in the natural conglomerate; Perform rotation and scaling operations on the three-dimensional digital models of each gravel according to the inclination range and long-axis dimension range of each group of gravels; Among them, the number of three-dimensional digital models generated for each group of gravels should be sufficient to meet the requirements of multiple sizes and multiple inclinations; Save all the generated three-dimensional digital models of gravels for subsequent use.

6. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, characterized in that The establishment of the three-dimensional rock mass digital model includes: Establish a three-dimensional digital model of a complete rock mass; Among them, the three-dimensional digital model should be stored in a three-dimensional digital matrix, and a complete rock mass refers to a rock mass without any defects and with the same material properties at any internal point.

7. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, characterized in that The insertion of each group of gravels into the three-dimensional rock mass digital model in batches according to the grading characteristics includes: Determine the grading characteristics of each gravel in the to-be-built conglomerate model according to the geological exploration results; Insert them into the three-dimensional rock mass digital model step by step in descending order according to the major axis size of each gravel group; Among them, in order to increase the final filling rate of the to-be-built conglomerate model, the gravels at each level should be inserted in descending order according to the major axis size. In addition, collision detection should be performed during insertion to prevent overlap and intersection of gravels.

8. The high-precision complex conglomerate three-dimensional digital model reconstruction method according to claim 1, wherein The establishment and output of slices of the high-precision complex conglomerate three-dimensional digital model include: Determine the number of slices that the model needs to be divided into; Convert the format of the filled conglomerate model and perform slice output; Among them, all operations of file format conversion should be processed by programming software and superimposed by Boolean operations of matrices.

9. A high-precision complex conglomerate three-dimensional digital model obtained by using the high-precision complex conglomerate three-dimensional digital model reconstruction method described in claims 1-8.

Images