Quantification method of structural plane damage characteristics and strength characteristic quantification method

Abstract

The application provides a structure surface damage feature quantification method and a strength feature quantification method, and belongs to the technical field of rock structure surface topography. The method comprises the following steps: carrying out an indoor shear test on a sawtooth structure surface, acquiring point cloud data of the structure surface before and after shearing by using a three-dimensional laser scanning technology, and collecting visual images of the structure surface; preliminarily extracting a damage area and calculating a preliminary damage degree D d of the structure surface d ; judging the overall damage condition of the structure surface according to D d ; for a seriously damaged area, quantitatively evaluating the damage range by using the three-dimensional laser scanning technology; for a structure surface with slight overall damage or a locally slightly damaged area in a seriously damaged structure surface, finely extracting the damage area by using a computer vision technology, and calculating a projection area proportion of the damage area on an XOY plane, denoted as D r ; using the damage degree D d and the damage range D r to jointly represent the damage feature of the structure surface. D d , D r can be directly used to establish a quantification model of the peak strength, and provide a quantification input with a more explicit physical meaning for strength evaluation.

Description

Technical Field

[0001] This application relates to the field of rock structural surface morphology technology, and in particular to a method for quantifying structural surface damage characteristics and a method for quantifying strength characteristics. Background Technology

[0002] With the advancement of deep earth exploration and deep resource extraction projects, such as deep-buried mines, deep geothermal development, and the construction of deep earth laboratories, rock masses are in a high-stress, high-osmotic-pressure, and complex dynamic environment. Under this extreme mechanical environment, rock structural surfaces (such as the serrated structural surfaces of granite and basalt) exhibit significant brittle-ductile composite characteristics during shear failure: some rock teeth undergo instantaneous cutting failure under high pressure stress, exhibiting brittle overall fracture; while other areas undergo slip wear, exhibiting extremely thin-layer powdering or surface polishing effects, showing a ductile-like evolution process.

[0003] To assess the stability of structural surfaces, it is essential to accurately quantify the damage characteristics before and after shearing. Current mainstream methods include geometric morphology comparison based on 3D laser scanning and visual feature recognition based on digital image processing. However, in complex shear failure scenarios driven by high ground stress, existing quantification schemes face the following insurmountable technical bottlenecks: Firstly, 3D laser scanning technology is prone to the submergence effect of minute damage signals in this scenario. Specifically, under high ground stress, a large amount of damage to the structural surface manifests as sub-millimeter or even micrometer-level surface wear or polishing. Traditional 3D laser scanners, limited by their physical resolution and environmental noise, struggle to accurately capture these minute morphological changes. When calculating volumetric damage, these critical minor damage areas are often mistakenly identified as measurement errors and discarded, leading to a severe underestimation of the damage range and an inability to accurately reflect the gradual decay of the structural surface shear strength. On the other hand, although traditional computer vision methods can delineate damaged areas by recognizing color or texture changes in images, they are essentially projections of three-dimensional shapes onto a two-dimensional plane. In severely damaged areas such as incisors, vision technology cannot provide crucial information on depth and volume loss, and therefore cannot accurately reflect the degree of loss of geometric interlocking force of the structural surface, leading to significant deviations in the assessment of the shear resistance of the structural surface.

[0004] Therefore, there is a need to provide an improved technical solution that addresses the shortcomings of the existing technology. Summary of the Invention

[0005] The purpose of this application is to provide a method for quantifying structural surface damage characteristics and a method for quantifying strength characteristics, so as to solve or alleviate the problems existing in the prior art.

[0006] To achieve the above objectives, this application provides the following technical solution: This application provides a method for quantifying structural surface damage characteristics, including: Indoor shear tests were conducted on the sawtooth structure surface. Three-dimensional laser scanning technology was used to obtain point cloud data of the structure surface before and after shearing, and visual images of the structure surface were also acquired. Preliminary extraction of the damaged area: Based on the damaged area, the volume difference of the structural surface specimen before and after shearing is calculated. This volume difference is used to measure the degree of damage to the structural surface, denoted as D. d ; According to D d To assess the overall damage to the structural surface, three-dimensional laser scanning technology is used to quantitatively evaluate the damage range for severely damaged areas. For structural surfaces with minor overall damage, or areas with minor local damage within severely damaged structural surfaces, computer vision technology is used to finely extract the damaged areas and calculate the proportion of the projected area of ​​the damaged area on the XOY plane to quantify the damage range of the structural surface, denoted as D. r ; Using damage level D d and damage range D r Commonly characterizing structural surface damage features.

[0007] In the above scheme, three-dimensional laser scanning (volume difference) and computer vision (planar projection) are organically combined to measure the degree of damage, D. d (Volume difference) and damage range D r (Projected area ratio) Two complementary quantitative indicators jointly characterize the damage features of the structural surface, D d D reflects the actual degree of damage to the structural surface. r This directly reflects the reduction in contact area and changes in the friction interface, helping to characterize the spatial distribution of damage on the overall surface. Using both together provides a more direct and comprehensive description of the post-shear damage state than existing geometric parameters (such as tilt angle and peak height). Secondly, by determining the damage identification method selection strategy based on the damage situation (damage type / damage degree), the most suitable technical path can be selectively used, reducing the risk of errors from a single identification method being propagated to the final quantitative indicators and improving the accuracy of damage identification. Furthermore, D... d D r As a damage characteristic, it can be directly used to establish regression / correlation quantitative models with peak strength and failure mode, thereby providing more direct quantitative input for strength assessment, stability evaluation and support design.

[0008] In conjunction with the first aspect, preferably, the determination of structural surface damage specifically includes: The degree of damage D d Compared with a preset damage threshold, if the damage degree D dIf the damage is less than the damage threshold, the structural surface is determined to be slightly damaged; otherwise, the structural surface is determined to be severely damaged.

[0009] In conjunction with the first aspect, preferably, the visual image of the structural surface after shearing is observed to determine the damage to the structural surface, as follows: If the structure surface only shows white band-shaped scratches and no breakage is seen in the visual image after cutting, it is considered to be slightly damaged. If the visual image after cutting shows both white band-shaped scratches and fracture areas on the structural surface, with the fracture area being dominant, it indicates severe damage to the structural surface.

[0010] The above scheme defines the specific method for judging the damage to the structural surface from two different perspectives: by assessing the degree of damage D d By comparing the damage with a preset damage threshold, the damage assessment becomes more objective and standardized. Observing the visual image of the sheared structural surface to determine the damage is an intuitive and rapid qualitative assessment method, beneficial for quick evaluation on-site. Both methods can adapt to the needs of different engineering scenarios.

[0011] In conjunction with the first aspect, preferably, three-dimensional laser scanning technology is selected to identify the damaged areas of the structural surface, including:

[0012] The cut point cloud data is subjected to boundary clipping and interpolation homogenization to obtain the interpolated point cloud data. The structural surface is discretized into spatial cubes. For each cube, the Z coordinate values ​​of the point cloud data before and after shearing at the same (X, Y) position are differentially calculated to obtain the height difference ΔZ. Based on experimental calibration, the critical value ΔZ0 for damage discrimination is determined, and the regions at the four vertices of each cube that satisfy ΔZ≥ΔZ0 are identified as the damaged regions of the structural surface. Accordingly, the volume difference of the structural surface specimen before and after shearing is calculated based on the damaged area, including: Calculate the sum of the volume differences of all cubes within all damaged regions before and after shearing, and denote this as the damage degree D. d .

[0013] In the above scheme, through systematic data processing and clear physical discrimination criteria, abstract damage characteristics are transformed into precisely calculable spatial parameters, enabling objective and accurate extraction of the damage area. Simultaneously, by discretizing the damage into tiny cubes (voxels), the continuous, macroscopic structural surface damage problem is decomposed into discrete, microscopic local deformation analysis. Height difference is used to directly reflect the degree of damage, and the damage degree D is given. d The detailed calculation method.

[0014] In conjunction with the first aspect, preferably, the method includes: Based on the results of indoor shear tests, the relationship between the sawtooth inclination angle and sawtooth height of the sawtooth structure and the degree and extent of damage to the structure is analyzed. Based on the positive correlation between sawtooth inclination angle, sawtooth height and the degree of structural surface damage, and the fact that the sawtooth inclination angle has a greater impact on the degree of damage than the sawtooth height, a quantitative model of the individual effects and interactions of sawtooth inclination angle and sawtooth height on the degree of structural surface damage is constructed. Based on the positive correlation between sawtooth inclination angle and damage range, and the lack of a regular relationship between sawtooth height and damage range, a quantitative model of the interaction between sawtooth inclination angle and sawtooth height on the damage range of structural surface is constructed. Based on the positive correlation between damage range and damage degree, a quantitative model of the relationship between damage range and damage degree is constructed.

[0015] The above scheme establishes a systematic damage prediction model framework based on experimental data. This framework not only reveals the intrinsic correlation and primary and secondary relationship between sawtooth geometry (inclination angle, height) and damage characteristics (degree, range), but also fixes these relationships in the form of quantitative models. This realizes the transformation of structural surface damage characteristic quantification from post-event description to pre-event prediction, enabling engineering practice to directly quantify damage characteristics based on these quantitative models. This enhances the engineering guidance value of damage characteristic quantification and provides a scientific basis for stability control and support design optimization in rock mass engineering.

[0016] Secondly, this embodiment provides a method for quantifying the strength characteristics of structural surfaces, including: The degree of damage D of the structural surface is calculated using the structural surface damage feature quantification method described in any of the above embodiments. d and damage range D r ; Based on the positive correlation between sawtooth inclination angle, sawtooth height and peak strength, and the fact that after removing the influence of damage degree, there is no linear relationship between damage range and peak strength, a quantitative model of the interaction between damage range and damage degree on the peak strength of the structural surface is constructed.

[0017] This scheme uses quantified damage characteristics as input parameters to reveal the mechanism by which damage affects strength, and constructs a quantitative model based on this. It overcomes the limitations of traditional methods that estimate strength based only on initial geometric parameters (such as sawtooth inclination angle and height) while ignoring the influence of damage evolution during shearing. It achieves a key leap from damage evolution analysis to macroscopic mechanical property prediction, and provides a direct theoretical basis and calculation tool for strength design in rock mass engineering.

[0018] Thirdly, this embodiment provides a computer system, including a memory, a processor, and a computer program stored in the memory, characterized in that the processor executes the computer program to implement the structural surface damage feature quantification method of any of the above embodiments, or the steps of the structural surface strength feature quantification method of any of the embodiments.

[0019] Fourthly, this embodiment provides a computer-readable storage medium storing a computer program / instructions thereon, characterized in that, when the computer program / instructions are executed by a processor, they implement the steps of the structural surface damage feature quantification method or the structural surface strength feature quantification method described in any of the above embodiments.

[0020] Fifthly, this embodiment provides a computer program product, including a computer program / instruction, characterized in that, when the computer program / instruction is executed by a processor, it implements the steps of the structural surface damage feature quantification method or the structural surface strength feature quantification method described in any of the above embodiments.

[0021] The technical effects achieved in the third, fourth, and fifth aspects of this application can be referred to the descriptions in the first and second aspects, and will not be repeated here. Attached Figure Description

[0022] Figure 1 The diagram shows the analysis of the shear test results of the first group of specimens (J1~J16), where (a) is a diagram showing the relationship between sawtooth height and peak strength, (b) is a diagram showing the relationship between sawtooth angle and peak strength, and (c) is a diagram showing the relationship between peak strength under the interaction of sawtooth height and angle.

[0023] Figure 2 The diagram shows the analysis of the shear test results of the first group of samples (J1~J16). (a) is a schematic diagram of the relationship between the sawtooth angle and the peak strength, (b) is a schematic diagram of the relationship between the sawtooth height and the peak strength, and (c) is a schematic diagram of the relationship between the peak strength under the interaction of the sawtooth angle and the height.

[0024] Figure 3 The comparison shows the serration angle and serration height in two groups of samples, where (a) is the relationship between the serration angles of the two groups of samples and (b) is the relationship between the serration heights of the two groups of samples.

[0025] Figure 4 The diagrams show the effects of sawtooth height and sawtooth tilt angle on peak strength for all samples. (a) shows the relationship between sawtooth angle and peak strength, (b) shows the relationship between sawtooth height and peak strength, and (c) shows the relationship between peak strength under the interaction of sawtooth angle and height.

[0026] Figure 5The images show the reconstructed structural planes before and after shearing, where (a) is the reconstructed image before shearing, (b) is the reconstructed image after shearing, and (c) is the reconstructed image after splicing.

[0027] Figure 6 This is a schematic diagram of the use of three-dimensional laser scanning technology to identify the damaged area, where (a) is the rough surface of the structural surface before and after shearing, and (b) is an enlarged view of the yellow circle.

[0028] Figure 7 The damage characteristics of samples H1 and H4 are shown, where (a) is sample H1 and (b) is sample H4.

[0029] Figure 8 The influence of sawtooth height and sawtooth tilt angle on the degree of damage (J1-J16) is shown in (a), where (a) is the relationship between sawtooth height and degree of damage, (b) is the relationship between sawtooth angle and degree of damage, and (c) is the relationship between degree of damage under the interaction of sawtooth angle and height.

[0030] Figure 9 The influence of sawtooth height and sawtooth inclination angle on the degree of damage for the second group of samples (H1~H16) is shown in (a) for the relationship between sawtooth angle and degree of damage, (b) for the relationship between sawtooth height and degree of damage, and (c) for the interaction between sawtooth height and sawtooth inclination angle on the degree of damage.

[0031] Figure 10 The effects of sawtooth height and sawtooth tilt angle on the degree of damage are shown in (a), where (b) is the effect of sawtooth height on the degree of damage, (c) is the effect of sawtooth angle on the degree of damage, and (d) is the effect of the interaction between sawtooth angle and height on the degree of damage.

[0032] Figure 11 The diagram illustrates the relationship between damage range, damage degree, and peak intensity. (a) shows the relationship between damage range and damage degree, (b) shows the relationship between damage degree and peak intensity, and (c) shows the effect of damage range + damage degree on peak intensity.

[0033] Figure 12 The results are from a partial regression analysis of the damage extent and peak intensity. Detailed Implementation

[0034] The embodiments of this application will now be described with reference to the accompanying drawings.

[0035] This embodiment provides a method for quantifying structural surface damage characteristics, including: Indoor shear tests were conducted on the sawtooth structure surface. Three-dimensional laser scanning technology was used to acquire point cloud data of the structure surface before and after shearing, and visual images of the structure surface were also acquired. The damage area was initially extracted. Based on the initially extracted damage area, the overall volume difference of the structure surface specimen before and after shearing was calculated. This volume difference was used to measure the degree of damage to the structure surface, denoted as D. d According to D d To assess the overall damage to the structural surface, three-dimensional laser scanning technology is used to quantitatively evaluate the damage range for severely damaged areas. For structural surfaces with minor overall damage, or areas with minor local damage within severely damaged structural surfaces, computer vision technology is used to finely extract the damaged areas and calculate the proportion of the projected area of ​​the damaged area on the XOY plane to quantify the damage range of the structural surface, denoted as D. r Damage level D d and damage range D r Commonly characterizing structural surface damage features.

[0036] Serrated surfaces, characterized by regular or irregular serrated morphology, consist of a series of raised serrations and recessed grooves. In rock mechanics, the geometric characteristics of the serrations (such as dip angle, height, and spacing) directly influence the mechanical behavior and failure mode of the surface. Serrated surfaces were chosen as the research object because they are prevalent in engineering rock masses, and their regular geometric shape facilitates the establishment of a quantitative relationship between damage characteristics and geometric parameters, providing an idealized physical model for mechanistic research.

[0037] Three-dimensional laser scanning technology refers to the use of laser ranging principle to obtain a dense set of three-dimensional coordinate points (i.e., point cloud data) on the surface of a structure by emitting a laser beam and receiving the reflected signal. By accurately recording the microscopic morphological features of the structure, it can completely capture the three-dimensional morphological changes of the structure before and after shearing, providing an accurate data basis for subsequent damage area identification and volume difference calculation.

[0038] Point cloud data refers to a collection of a large number of spatial points acquired through 3D laser scanning. Each point contains 3D coordinate information (X, Y, Z). Point cloud data can accurately represent the spatial geometry of a structural surface. Visual images refer to color or grayscale images of a structural surface acquired through imaging devices such as digital cameras, reflecting the visual characteristics of the structural surface, such as texture and color.

[0039] In this embodiment, the degree of damage D is introduced. d The driven dynamic decision-making mechanism first extracts the damage area, and then calculates the overall damage degree D based on the extracted damage area. d Then based on D dThe overall damage status of the structural surface is assessed, and an identification technology is selected. For structural surfaces with minor overall damage (such as only surface scratches), due to the minimal changes in morphology, 3D laser scanning may easily mistake the damage for noise. However, significant changes in optical characteristics (color, texture, reflectivity) are observed, and computer vision technology is used to capture the subtle damaged surfaces. For structural surfaces with severe overall damage (such as fractures), the degree of damage in each local area is reassessed before processing: for severely damaged areas, due to the large changes in morphology before and after damage, 3D laser scanning provides better stability for obtaining geometric losses, and 3D laser scanning technology is used for quantification. For some minor damaged areas within the severe overall damage of the structural surface, computer vision technology is still used for fine extraction and damage range quantification. This mechanism of dynamically selecting subsequent structural surface quantification based on the initial assessment results enables adaptive and accurate characterization of heterogeneous damage (large-scale chipping and micro-grinding). The damage level D is used. d and damage range D r By jointly characterizing the damage features of structural surfaces and establishing a dual index system that includes both the intensity dimension of damage degree and the breadth dimension of damage range, the physical mechanism of shear failure can be more completely reconstructed, providing a prerequisite for damage evolution and strength research.

[0040] In this embodiment, there are two technical approaches to determine the damage status of the structural surface.

[0041] One approach is to assess the damage to the structural surfaces, specifically by determining the degree of damage, D. d Compared with the preset damage threshold, if the damage degree D d If the damage is less than the damage threshold, the structural surface is judged to be slightly damaged; otherwise, the structural surface is judged to be severely damaged.

[0042] As mentioned earlier, the degree of damage D d The damage threshold is defined by calculating the volume difference of the structural surface within the damaged area before and after shearing. Its physical meaning is the total volume of rock material lost (or generated) due to wear, breakage, or spalling during shearing. The preset damage threshold is a critical value determined in advance through experiments or experience, serving as a benchmark to distinguish between minor and severe damage. The setting of this threshold depends on the specific lithology, the mechanical properties of the structural surface, and the engineering precision requirements.

[0043] For example, a series of indoor shear tests can be used to establish different degrees of rock sample damage, D. d The relationship with macroscopic mechanical behavior (such as peak strength reduction) is determined by selecting a D that has a significant impact on damage morphology. d The value is used as a threshold.

[0044] In this embodiment, the damage state of the structural surface is divided into two qualitative categories: minor damage and severe damage. Minor damage refers to damage degree D. dCases not exceeding a preset threshold typically correspond to minor wear or polishing of the structural surface, possibly accompanied by white scratches, but without macroscopic rock fracture or large-scale shearing of protrusions, and with minimal changes in mechanical properties. Severe damage typically corresponds to significant fracturing, erosion, or plastic deformation of the structural surface, with clear material loss and macroscopic fracture surfaces, and its mechanical properties (such as shear strength) usually show significant deterioration. These two classifications of minor and severe damage directly relate to subsequent technology selection, ensuring that the identification method matches the severity of the damage.

[0045] Another way to judge is to observe the visual image of the structural surface after shearing and judge the damage of the structural surface, as follows: if the structural surface only shows white band-shaped scratches and no fracture is seen in the visual image after shearing, it is considered to be slightly damaged; if the visual image after shearing shows that the structural surface has both white band-shaped scratches and fracture areas, and the fracture area is dominant, it is considered to be severely damaged.

[0046] The aforementioned scheme is a qualitative / semi-quantitative damage assessment method based on visual morphological features. Its core lies in establishing a rapid diagnostic logic based on typical visual characteristics of damage: when a white, band-like scratch is identified without a fracture area, it indicates that the shearing action only caused surface friction and wear, constituting minor damage; when both a white, band-like scratch and a fracture area are identified simultaneously, and the fracture area visually dominates, it indicates severe damage. This method does not require complex point cloud data processing and volume calculations; it can quickly assess the damage status at the test site or engineering site based solely on visual observation. Furthermore, its technical threshold and implementation cost are lower than 3D laser scanning technology, making it suitable for widespread application in scenarios with limited resources or requiring large-scale preliminary screening.

[0047] It should be noted that in this application, computer vision technology specifically refers to damage recognition based on two-dimensional image texture and spectral features, excluding the case of establishing three-dimensional point clouds using depth vision.

[0048] Furthermore, for areas classified as severely damaged, three-dimensional laser scanning technology is used to quantify the damage range. This includes: performing boundary clipping and interpolation homogenization on the sheared point cloud data to obtain interpolated point cloud data; discretizing the structural surface into spatial cubes, and for each cube, performing a difference operation on the Z coordinate values ​​of the point cloud data before and after shearing at the same (X, Y) position to obtain the height difference ΔZ; determining the damage discrimination critical value ΔZ0 based on experimental calibration, and identifying the areas where ΔZ≥ΔZ0 at all four vertices of each cube as the damaged areas of the structural surface; correspondingly, calculating the volume difference of the structural surface specimen before and after shearing based on the damaged areas, including: calculating the sum of the volume differences of all cubes in all damaged areas before and after shearing, denoted as the damage degree D. d .

[0049] In this embodiment, the damaged area of ​​the structural surface is identified by the region where ΔZ≥ΔZ0 is satisfied at all four vertices. The final damaged area is then precisely identified. Based on this, the sum of the volume differences of all cubes in the final damaged area before and after shearing is taken as the final damage degree. Therefore, the damage degree here is still denoted as D. d However, due to the different extent of the damage, D d The size and D calculated during the initial extraction d Different sizes.

[0050] Here, boundary clipping refers to removing the edge regions around the structural surface point cloud data in the digital model, ensuring that the analysis focuses on the effective structural surfaces themselves. Since the density of the original point cloud data may be uneven, interpolation homogenization is performed, and mathematical algorithms (such as Kriging interpolation, inverse distance weighted interpolation, etc.) are used to estimate the elevation values ​​(Z coordinates) on the grid nodes, generating a regular, continuous, and uniformly dense structural surface model, laying the foundation for subsequent difference operations.

[0051] The regularized 3D structural surface model, after interpolation, is logically divided into a series of tiny, regularly arranged spatial cubes. Each cube is defined by the coordinates of its four vertices and the projection coordinates of those vertices onto the XOY plane. This provides the basic geometric unit for subsequent precise calculations of the volume change of each micro-element before and after shearing. For each spatial cube, the height difference ΔZ between its four vertices before and after shearing is calculated. Combined with a pre-calibrated damage discrimination threshold ΔZ0, this objectively identifies whether the cube belongs to the damaged region, quantifying the extent of the damaged region. Finally, the sum of the volume differences of all cubes within all damaged regions before and after shearing is calculated to obtain the damage degree D. d The extent of damage to the structural surface is quantified by calculating the proportion of the projected area of ​​the damaged region on the XOY plane, denoted as D. r .

[0052] For structural surfaces with minor overall damage, or for structural surfaces with severe overall damage but minor damage in certain areas, computer vision techniques are used for fine extraction of the damaged areas. This includes the following sub-steps: (1) Identify the damaged area: To accurately identify damaged areas on structural surfaces, the damage images were first segmented and labeled using the image annotation tool X-Anylabeling. Damaged areas were manually selected and categorized as "damage" based on their outlines, generating label files (in .txt format). The annotation process required the participation of professional rock mechanics scholars to ensure accurate damage area determination. Especially when dealing with complex damage types, the experience of these scholars is crucial, helping to accurately identify areas with different degrees of damage. After each image was annotated, a corresponding label file was exported for subsequent training and analysis.

[0053] (2) Establish a damage area segmentation model: Based on the labeled files showing structural surface damage, a structural surface damage image segmentation model was established using the YOLOv8 algorithm. YOLOv8 is a state-of-the-art object detection and image segmentation model, capable of efficiently handling damage identification tasks in images.

[0054] (3) Predict the extent of damage and the proportion of the damaged area: After model training is complete, the YOLOv8 segmentation model is used to predict damaged regions. The model can automatically identify all damaged regions in the structural surface image and calculate the proportion of the damaged area based on the ratio of the damaged region to the total image area. The specific steps are as follows: Contour extraction: Use the "cv2.findContours" function in OpenCV to extract the outer contour of each damaged area.

[0055] Area calculation: The area is calculated by counting the number of non-zero pixels in the mask of each damaged area.

[0056] Area percentage calculation: Sum the areas of all damaged regions and calculate their proportion in the total image area. Finally, output the results in percentage format to visually represent the relative size of the damaged regions.

[0057] In this implementation, damage level D is used. d and damage range D r Commonly characterizing structural surface damage features. D d With D r The combination of these two factors constitutes the loss fingerprint of the structural surface. Different combinations of these factors have a very clear indication of the damage mechanism, which is very easy for engineers to understand. For example: if D... d With D r Both are relatively large, indicating that the overall damage has been severe and widespread; if D d Larger, D r A smaller value indicates that very severe damage has occurred locally; conversely, if D... d Smaller, D rA larger value indicates extensive minor wear or polishing; of course, if D... d With D r Both are small, indicating that the structural surface is basically intact. Engineers only need to consider the different combinations of these two indicators to quickly determine the extent of damage and failure to the structural surface and take appropriate engineering measures. Furthermore, D... d D r It can be directly used to establish a quantitative model of peak intensity, providing a more physically meaningful quantitative input for intensity assessment.

[0058] In terms of the degree of damage to the quantified structural surface D d The following steps also include: analyzing the relationship between the sawtooth inclination angle, sawtooth height, and the degree and extent of damage to the sawtooth structure based on the results of indoor shear tests; constructing a quantitative model of the individual effects and interactions of sawtooth inclination angle and sawtooth height on the degree of damage to the structure, based on the positive correlation between sawtooth inclination angle and sawtooth height and the degree of damage to the structure, and the fact that the effect of sawtooth inclination angle on the degree of damage is greater than that of sawtooth height; constructing a quantitative model of the interaction between sawtooth inclination angle and sawtooth height on the extent of damage to the structure, based on the positive correlation between sawtooth inclination angle and damage extent, and the lack of a regular relationship between sawtooth height and damage extent; and constructing a quantitative model of the relationship between damage extent and degree of damage, based on the positive correlation between damage extent and degree of damage.

[0059] The sawtooth inclination angle is the angle between the inclined surface of a single sawtooth and a horizontal reference plane (such as the XOY plane) in the sawtooth structure. The sawtooth height refers to the vertical protrusion of a single sawtooth, usually defined as the vertical distance from the bottom of the sawtooth valley to the top of the tooth.

[0060] The relationship between the sawtooth inclination angle and sawtooth height of the sawtooth structure surface and the degree and extent of damage to the structure surface is analyzed. Specifically, mathematical statistical methods (such as correlation analysis and regression analysis) are used to study the relationship between the two independent variables, sawtooth inclination angle and sawtooth height, and the degree of damage D. d Damage range D r The relationship between these two dependent variables is used to reveal the damage mechanism of structural surfaces, thereby guiding engineering practice.

[0061] The analysis results show a positive correlation between the sawtooth inclination angle, sawtooth height, and the degree of structural surface damage. Therefore, for ease of description, the quantitative models of the structural surface damage degree caused by the sawtooth inclination angle and sawtooth height acting alone are denoted as the first quantitative model and the second quantitative model, respectively. The general expressions are as follows: (1-1) (1-2) In the formula, , These are the sawtooth tilt angle and the sawtooth height. , , , These are the coefficients to be fitted for the first and second quantization models, which need to be fitted and calibrated using experimental data.

[0062] The quantitative model (denoted as the third quantitative model) of the effect of the interaction between sawtooth inclination angle and sawtooth height on the degree of damage to the structural surface is as follows: (1-3) In the formula, D represents the degree of damage after standardized treatment. d , , These are the standardized sawtooth angle and sawtooth height. , The coefficients to be fitted in the third quantization model need to be fitted and calibrated using experimental data. Since the sawtooth inclination angle has a greater impact on the degree of damage than the sawtooth height, therefore... .

[0063] It should be noted that in this embodiment, "individual effect" refers to the model considering the independent influence of the sawtooth tilt angle or sawtooth height on the degree of damage, while "interaction" refers to the model considering the influence of both tilt angle and height simultaneously, characterizing the synergistic effect on damage when the two parameters change together.

[0064] The analysis further revealed a positive correlation between the sawtooth inclination angle and the damage range, while no consistent relationship was found between the sawtooth height and the damage range. Specifically, no statistically significant and stable correlation was found between the sawtooth height and the damage range, indicating that the damage range cannot be directly predicted using the sawtooth height alone. Therefore, quantitative models (denoted as the fourth and fifth quantitative models) were constructed to assess the impact of the sawtooth inclination angle alone, as well as the interaction between the sawtooth inclination angle and the sawtooth height, on the damage range of the structural surface. The general expressions are as follows: (1-4) (1-5) In the formula, Indicates the extent of damage after standardization. , , , These are the coefficients to be fitted for the fourth and fifth quantization models, respectively, which need to be fitted and calibrated using experimental data.

[0065] It should be noted that since there is no significant relationship between the sawtooth height and the damage range, a quantitative model of the structural surface damage range based solely on the sawtooth height has not been established. In other words, this embodiment does not consider the individual effect of height, but still retains its interaction with the tilt angle to avoid constructing an incorrect or misleading model.

[0066] The analysis results also show that as the damage area increases, the damage severity also increases, and the two are approximately linearly related. Based on this, a quantitative model between damage area and damage severity can be established, denoted as the sixth quantitative model, with the following expression: (1-6) In the formula, , These are the coefficients to be fitted in the sixth quantization model, which need to be fitted and calibrated using experimental data.

[0067] In this embodiment, the analysis of quantitative results based on structural surface damage characteristics deepens the scientific understanding of the structural surface shear failure mechanism, thereby simplifying the complex rock mass mechanical behavior into a computable mathematical model. By constructing multiple interrelated quantitative models, the following progressive technical effects are achieved: In the actual engineering investigation stage, only the geometric parameters (dip angle, height) of the rock mass structural surface need to be measured. The aforementioned models can then be used to pre-quantify the damage characteristics (damage degree and damage range) that may occur under future shear loads, providing forward-looking data support for rock mass stability evaluation and strength assessment. Furthermore, the first to sixth quantitative models together form a model system, enabling the prediction of damage characteristics even when a certain geometric parameter is missing, thus improving the applicability of the entire method system.

[0068] Based on the structural surface damage feature quantification method provided in the above embodiments, this application also provides a structural surface strength feature quantification method, which includes: The degree of damage D of the structural surface is calculated using the structural surface damage feature quantification method described in any of the above embodiments. d and damage range D r Based on the linear relationship between damage range, damage degree, and peak intensity, and considering that no linear relationship exists between damage range and peak intensity after removing the influence of damage degree, quantitative models are constructed for the effect of damage degree alone, as well as the interaction between damage range and damage degree on the peak intensity of the structural surface (denoted as the seventh quantitative model and the eighth quantitative model). The expressions are as follows: (1-7) (1-8) In the formula, Indicates peak intensity. Represents the standardized peak intensity. , These are the standardized D r D d , , , , The coefficients to be fitted for the seventh and eighth quantization models need to be fitted and calibrated using experimental data.

[0069] In this embodiment, based on experimental data, the correlation between damage characteristics and strength characteristics was analyzed to be a linear relationship between sawtooth inclination angle, sawtooth height, and peak strength. Furthermore, after eliminating the influence of damage degree, no linear relationship was observed between damage range and peak strength. Based on this, a strength prediction model with damage degree as the core independent variable was established, providing a more reliable peak strength estimation method that considers damage characteristics, thus improving the scientific rigor of rock mass engineering strength assessment. By explicitly excluding the damage range, the model avoids mistakenly introducing correlation as causation, thereby improving its accuracy.

[0070] Furthermore, the quantitative model for peak surface intensity is a linear intensity model with damage degree as the single independent variable. This model is very simple, with well-defined parameters and clear physical meaning: coefficients The slope reflects the rate at which the strength decreases with increasing damage. This can be understood as the baseline strength under the initial undamaged state. It directly reflects the mechanical essence that the greater the damage, the lower the peak strength. It is easily accepted and used by engineers, and once calibrated through testing... , This model can be widely applied to the rapid strength prediction of similar rock masses and structural surfaces, providing a scientific basis for the timing of support and the optimized design of support schemes.

[0071] The technical solution of this application is illustrated below with examples.

[0072] For example, the technical solution of this application may include the following steps: preparation of serrated surface specimens, indoor shear test, analysis of shear mechanical properties of serrated surface, quantification of damage degree of serrated surface and quantification of damage range of serrated surface.

[0073] Step 1, the preparation steps for the serrated surface specimens, can be referred to Chinese patent application number 2025115573930, and will not be repeated here. This step yields two sets of serrated surface specimens, whose numbers and geometric parameters are recorded, as shown in Tables 1 and 2: Table 1. Samples with serrated surface structure in the first group

[0074] Table 2. Samples with serrated surface structure in the second group.

[0075] The structural surface specimen material is red sandstone, and its basic mechanical parameters are shown in Table 3. Table 3 Basic mechanical parameters of red sandstone samples

[0076] Step 2, indoor shear test.

[0077] Using a KYZW-100 rock weak surface direct shear apparatus, an indoor shear test was conducted on the above-mentioned serrated structural surface specimens under a normal stress of 2MPa. Shear test curves were obtained, and visual images of the structural surface specimens before and after shearing were acquired to obtain the test results.

[0078] Step 3: Analysis of the shear mechanical properties of the sawtooth structure surface.

[0079] First, analyze the failure type. The sample numbered H1 ( =1.11 mm, =17.10°) belongs to the "shear slip type" failure. Therefore, taking specimen H1 as an example, the basic laws of shear slip type failure are analyzed from multiple aspects such as shear process, damage condition and shear curve. Similarly, taking H16 ( =2.83 mm, Taking 45.7° as an example, we analyze the basic laws of "tension-type" failure.

[0080] Secondly, for the first group of samples (J1~J16), the distribution of sawtooth height and peak intensity was analyzed based on the test results, such as... Figure 1 As shown, combined with Figure 1 The analysis results are explained below: (1) Under different sawtooth inclination angles, the correlation coefficients between sawtooth height and peak intensity were 0.92, 0.88, 0.99 and 0.85, respectively, indicating that there is a significant correlation between the two; (2) Under the same sawtooth inclination angle, the peak intensity increases with the increase of sawtooth height, and the two are approximately linearly related. The fitting results are shown in equations (3-4 to 3-7). Figure 1 As shown in (a), the fitted R-values ​​and regression coefficients are both high, with R values ​​of 0.92, 0.88, 0.99, and 0.85, respectively, and regression coefficients for sawtooth height of 0.29, 0.20, 0.22, and 0.40, respectively. (3-4) in, That is the peak intensity.

[0081] (3-5) (3-6) (3-7) (3) As the sawtooth inclination angle increases, the slope of the fitting line between sawtooth height and peak strength shows an upward trend, indicating that the influence of sawtooth height on peak strength gradually increases; (4) Under different sawtooth inclination angles, the range of peak strength for different sawtooth heights is 0.26MPa, 0.30MPa, 0.36MPa and 0.92MPa, respectively, which further shows that the influence of sawtooth height on peak strength becomes more significant as the inclination angle increases.

[0082] Since the correlation between sawtooth inclination angle and peak strength for samples J1-J16 at the same sawtooth height cannot be investigated individually, the data of sawtooth inclination angle and peak strength for samples J1-J16 are summarized and analyzed. The correlation coefficient between the two is calculated to be 0.95, which is significantly higher than the correlation coefficient between sawtooth height and peak strength, indicating a close relationship between the two. As the sawtooth inclination angle increases, the peak strength shows an upward trend, and the two are approximately linearly related, as shown in equation (3-8). Figure 1 As shown in (b), the R value is 0.95.

[0083] (3-8) After standardizing the serration angle, height, and peak intensity of samples J1~J16, a quantitative relationship among the three was established through multiple regression analysis. The specific results are as follows: Figure 1 As shown in equation (c) and equation (3-9), regression analysis revealed that the regression coefficients for the serration angle and serration height were 0.77 and 0.31, respectively, with corresponding p-values ​​of 3.73e-08 and 3.19e-04, both significantly less than 0.05. This indicates a significant statistical correlation between these two variables and peak intensity. In particular, the regression coefficient for the serration angle was greater than that for the serration height, further illustrating that this result suggests the serration angle should be considered a key factor when studying the influence of serrated structures on peak intensity.

[0084] (3-9) in, , and These represent the standardized peak intensity, maximum angle, and maximum height, respectively.

[0085] For the second group of samples (H1~H16), the distribution of sawtooth inclination angle and peak intensity was analyzed based on the test results, such as... Figure 2As shown, the following pattern emerges: (1) At different sawtooth heights, the correlation coefficients between sawtooth inclination angle and peak intensity were 0.96, 0.99, 0.96 and 0.96, respectively, indicating that there is a significant correlation between the two. (2) Under the same sawtooth height, the peak intensity increases with the increase of the sawtooth inclination angle, and the two are approximately linearly related. The fitting results are shown in equations (3-10~3-13) and Figure 2 As shown in (a), the R-values ​​and regression coefficients are both high, with R values ​​of 0.998, 0.987, 0.961 and 0.960, respectively. (3-10) (3-11) (3-12) (3-13) (3) The regression coefficients of the sawtooth inclination angles changed little, at 0.10, 0.14, 0.10 and 0.10 respectively, indicating that the influence of the sawtooth inclination angle on the peak intensity remained basically consistent under different sawtooth heights.

[0086] Since the correlation between sawtooth height and peak strength for samples H1-H16 at the same sawtooth inclination angle cannot be investigated individually, the data on sawtooth height and peak strength for samples H1-H16 are summarized and analyzed. The correlation coefficient is calculated to be 0.72, indicating a certain correlation between the two, but not as strong as the relationship between sawtooth inclination angle and peak strength. Peak strength shows an increasing trend with increasing sawtooth inclination angle, but the data exhibits significant dispersion (see...). Figure 2 (b) in the middle.

[0087] Based on the study of samples J1 to J16, both the sawtooth inclination angle and the sawtooth height are closely related to the peak strength, but the influence of the sawtooth inclination angle is more significant. Figure 3 The comparison of serration inclination angle and serration height for samples J1~J16 and H1~H16 is shown. The serration inclination angles are arranged in ascending order, and each inclination angle corresponds to a serration height. As can be seen from the figure, the serration inclination angle of samples H1~H16 is significantly higher than that of samples J1~J16, while the difference in serration height between the two groups of samples is not significant. The serration inclination angle has a more pronounced impact on the test results, leading to more sample fracture, while the impact of serration height on the test results is "masked" or "weakened".

[0088] After standardizing the serration angle, height, and peak intensity of samples H1~H16, a quantitative relationship among the three was established through multiple regression analysis. The specific results are as follows: Figure 2As shown in (c) and equation (3-14).

[0089] (3-14) After summarizing the test data of all samples J1~16 and H1~H16, the correlation coefficient between the sawtooth inclination angle and the peak strength was calculated to be 0.96, and the correlation coefficient between the sawtooth height and the peak strength was 0.73. Figure 4 Figures (a), (b), and equations (3-15~3-16) show the distribution of sawtooth inclination angle and peak intensity, and sawtooth height and peak intensity.

[0090] After standardizing the sawtooth inclination angle, sawtooth height, and peak intensity, the relationship between the three was established through multiple regression analysis, such as... Figure 4 As shown in equation (c) and equation (3-17), the regression coefficients for the serration angle and height are 0.82 and 0.23, respectively, and the corresponding p-values ​​(1.53e-15 and 1.01e-04) are both much less than 0.05, indicating a significant correlation between them and peak intensity. The regression coefficient for the serration angle is greater than that for the serration height, indicating that the serration angle has a greater impact on peak intensity than the serration height.

[0091] (3-15) (3-16) (3-17) In summary, both the sawtooth inclination angle and sawtooth height are significantly positively correlated with peak intensity, and the sawtooth inclination angle has a greater impact on peak intensity than the sawtooth height. The relationship among the three can be quantitatively described by equations (3-15 to 3-17).

[0092] The above results are from the perspective of the shear mechanical properties of the structural surface from the perspective of the geometric parameters of the structural surface (sawtooth inclination angle and sawtooth height), which provide theoretical and experimental basis for the quantification of damage characteristics.

[0093] Step 4: Quantification of the damage level of the serrated surface.

[0094] The degree of damage (D) is measured by the volume difference of the structural surface sample before and after shearing using three-dimensional laser scanning technology. d The following is an introduction to the specific operating methods and application effects.

[0095] (1) Conduct scanning experiments.

[0096] Before the shear test, a three-dimensional laser scan was performed on all six surfaces of the specimen. First, the specimen was placed with its rough surface facing up on the scanner turntable, which rotated 8 times at 45° intervals to complete a 360° scan. The scanner and software effectively stitched together the data from the 8 scans, ultimately exporting point cloud data containing the rough surface and the four side surfaces. Next, the specimen was placed with its bottom surface facing up on the scanner turntable, and the above operation was repeated, ultimately exporting point cloud data containing the bottom surface and the four side surfaces. The two sets of point cloud data were imported into Geomagic software, and the common marker points of the two sets were carefully examined. Using multi-point manual registration and global registration functions, the two sets of point cloud data were stitched together, ultimately forming a complete point cloud data containing the rough surface, bottom surface, and the four side surfaces. After the shear test, the above scanning and data processing process was repeated.

[0097] (2) Reconstruct the coordinate system.

[0098] Since the scan data before and after the shear test are not in the same coordinate system, they cannot be directly compared. Therefore, it is necessary to reconstruct the coordinate system using a fixed reference. During the shear test, the failure area of ​​the specimen is mainly concentrated on the rough surface, while the structural rock wall does not show obvious cracking. Therefore, the fitting planes of the bottom surface and four sides of the specimen can be used as fixed references. There are various methods for reconstructing the coordinate system. The core of these methods is to establish a new coordinate system based on the fixed reference and calculate the coordinates of each point in the point cloud data in this coordinate system. Taking MATLAB as an example, the specific steps are as follows: 1) Select the local data of the left side, front side, and bottom surface in Geomagic software and export them as ASCII files respectively; 2) Use MATLAB to calculate the fitting planes of the left side, front side, and bottom surface respectively, and define them as the YOZ, XOZ, and XOY planes; 3) Calculate the distances from each point in the point cloud data to the YOZ, XOZ, and XOY planes, and use these distances as the new X, Y, and Z coordinates. After the shear test, repeat the above data processing process.

[0099] (3) Point cloud data splicing.

[0100] After importing the point cloud data before and after shearing into Geomagic software, the common marker points in the two sets of data were carefully examined, and the data were stitched together using the multi-point manual registration and global registration functions. At this point, the two sets of point cloud data almost completely overlapped, with only some differences at the rough surface. These differences correspond to the damage areas after shearing of the sample. By comparing the image generated by Geomagic with the actual damage image, if the two are relatively consistent, it indicates that the data processing process is reliable. Finally, the rough surface point cloud data before and after shearing were exported as ASCII files and named D1 and D2, respectively.

[0101] (4) Identify the damaged area.

[0102] Point cloud data satisfying 5mm < X < 95mm and 5mm < Y < 95mm were selected, and interpolation was performed at intervals of 0.25mm in both the X and Y directions to achieve point cloud homogenization. Next, the Z values ​​of the rough surface point cloud data before and after shearing were subtracted (see...). Figure 5 In theory, It should be greater than or equal to 0, where The region with a value of 0 represents an undamaged region. Areas with a value greater than 0 indicate damaged areas. Due to factors such as external environment, object surface characteristics, data processing algorithms, and human stitching, 3D laser scanning still faces certain difficulties in identifying minute deformations, leading to... The distribution of damage was rather chaotic. By comparing a large number of damage images generated by MATLAB with actual damage images, 0.01 mm was ultimately selected as the critical value for determining the damage range of the structural surface, i.e. Areas with a value <0.01 are identified as undamaged areas, while... Areas with a value >0.01 were identified as damaged areas. The results are as follows: Figure 6 As shown, the red grid in the left figure (a) represents the damaged area, and the right figure (b) is an enlarged view of the yellow circle in the left figure.

[0103] The rough surface is composed of a series of spatial quadrilaterals ABCD (before shearing: A1, B1, C1, D1; after shearing: A2, B2, C2, D2), whose projection onto the XOY plane is a square A'B'C'D' with dimensions of 0.25mm × 0.25mm. The area of ​​spatial quadrilateral ABCD is equal to the sum of the areas of triangles ABC and ACD. The closed space enclosed by eight points A, B, C, D, A', B', C', and D' can be simplified to a cube with a length and width of 0.25mm and a height equal to the average height of the four vertices. The damage volume is calculated as follows: traverse each point within the cube... The value of the four vertices Cubes with values ​​greater than 0.01 were marked as damaged areas; then, the volume difference of this area before and after shearing was calculated, and the sum of the volume differences of all damaged areas was the damage volume. Table 4 shows the damage degree of structural surface specimens J1~J16 and H1~H16.

[0104] Table 4 Damage degree D of all specimens d

[0105] Taking specimens H1 and H4 as examples, the damage characteristics are analyzed, and at the same time, the application effect of using a single three-dimensional laser scanning technology in identifying the damaged area of the structural plane and evaluating the damage degree is analyzed. After shear, specimen H1 shows a "shear-slip type" failure, and the surface mainly presents white banded scratches, without obvious fracture failure; while specimen H4 is a "shear + tension type" failure, with both white banded scratches and fracture areas on the surface, and the fracture area is dominant. Figure 7 For the damage characteristics of specimens H1 and H4, that is, the actual damage pictures (visual images) of the specimens and the three-dimensional reconstruction images. In the three-dimensional reconstruction images, only the damaged areas are shown, and the Z-axis represents the coordinate difference of the corresponding points on the structural plane before and after shear in the Z-axis direction.

[0106] In this embodiment, the damage condition of the structural plane is evaluated from the perspective of shear marks. The shear marks on the structural plane can be divided into two categories: one is the scratch area with less damage, smooth touch and no obvious roughness (such as Figure 7 the white scratches in (a) below), which is classified as minor damage, and the other is the fracture area with deeper damage and obvious roughness of touch (such as Figure 7 the fracture area in (b) below), which is classified as severe damage. By analyzing the test results of specimens J1~J16 and H1~H16, it is found that three-dimensional laser scanning can accurately capture the fracture area of the structural plane, but it is more difficult to depict the scratch area. The main reasons are as follows: the damage depth of the scratch area is usually less than 1 mm, and due to the influence of external environment, object surface characteristics, data processing algorithms and manual splicing and other factors, it is difficult for three-dimensional laser scanning to accurately capture the minor damage before and after specimen shear. Therefore, the following two situations usually occur: some scratch areas cannot be clearly presented in the three-dimensional point cloud data (such as Figure 7 the white scratches in (b) below); some non-scratch areas are incorrectly displayed as discrete points in the three-dimensional point cloud (such as Figure 7 the discrete point set in (a) below). To sum up, three-dimensional laser scanning can accurately identify the significant damaged area of the structural plane, but it is difficult to accurately capture the minor scratch area. Figure 7 As shown in part (a) below, the damage degree D d of specimen H1 calculated by MATLAB is 670.75 mm³, and the damaged area D r is 939.69 mm². By comparing Figure 7 (a) and (b) below, it is found that the actual damaged area of specimen H1 mainly concentrates in the following ranges: [25 < X < 38, 0.5 < Y < 98], [48 < X < 60, 0.5 < Y < 98], [70 < X < 80, 0.5 < Y < 98] and [90 < X < 97, 0.5 < Y < 63]. In these areas, the damaged volume D dThe damage area is 667.41 mm³. r The area was 849.69 mm², representing a change of -0.5% and -9.6% respectively compared to previous values. The white scratches on sample H4 were difficult to accurately represent in the 3D reconstructed image, leading to a significant deviation in the damage range. The 2D damage area ratio measured by computer vision technology was 0.42, while the 2D damage area ratio obtained by 3D laser scanning technology was 0.28, with a relative error of -34.4%. However, a detailed comparison between the sheared sample H4 and the 3D reconstructed image revealed that significant damage areas were accurately identified. Since the volume difference between the white scratch areas that were not identified by 3D scanning and those before and after shearing was minimal, the damage degree was mainly dominated by the significant damage areas. Therefore, the damage degree of sample H4 obtained using 3D laser scanning technology is accurate. Other samples showed similar results.

[0107] In summary, 3D laser scanning technology has difficulty accurately identifying minor scratch areas on structural surfaces. These errors may lead to significant deviations in the damage range, but have a minimal impact on the degree of damage. Therefore, while using 3D laser scanning technology to study the degree of damage to structural surfaces is feasible, it may introduce substantial biases when studying the damage range in scratched areas.

[0108] Based on the three-dimensional laser scanning data and visual images of the first group of samples (J1~J16), combined with Figure 8 The distribution pattern of sawtooth height and damage degree was analyzed, and the conclusions are as follows: (1) Under different saw tooth inclination angles, the correlation coefficients between saw tooth height and damage degree were 0.94, 0.96, 0.91 and 0.96, respectively, indicating that there is a strong correlation between the two.

[0109] (2) Under the same sawtooth inclination angle, the degree of damage increases with the increase of sawtooth height, and the two are approximately linearly related. The fitting results are shown in equations (3-18) to (3-21) and Figure 8 As shown, the R-values ​​and regression coefficients are both high, with R values ​​of 0.94, 0.96, 0.91 and 0.96, and regression coefficients of 744.36, 1432.34, 1633.54 and 2161.13, respectively.

[0110] (3-18) (3-19) (3-20) (3-21) (3) As the saw tooth inclination angle increases, the slope of the fitted line between saw tooth height and damage degree gradually increases, indicating that the saw tooth height has a more significant effect on the damage degree (Equations 3-18 to 3-21).

[0111] (4) Under different sawtooth inclination angles, the range of damage degree for different sawtooth heights was 593.19 mm³, 1770.57 mm³, 2975.1 mm³ and 3882.84 mm³, respectively, which further shows that the influence of sawtooth height on damage degree gradually increases with the increase of inclination angle.

[0112] Since the correlation between sawtooth inclination angle and damage degree for samples J1~J16 at the same sawtooth height cannot be investigated individually, the data on sawtooth inclination angle and damage degree for samples J1~J16 are summarized and analyzed. The correlation coefficient between the two is calculated to be 0.96, which is significantly higher than the correlation coefficient between sawtooth height and damage degree, indicating a close relationship between the two. As the sawtooth inclination angle increases, the damage degree shows an upward trend, and the two are approximately linearly related. The fitting results are shown in equation (3-22) and... Figure 8 As shown in (b), the R value is 0.96.

[0113] (3-22) After standardizing the serration angle, height, and damage degree of samples J1~J16, a quantitative relationship among the three was established through multiple regression analysis. The specific results are as follows: Figure 8 As shown in equation (c) and equation (3-23), regression analysis revealed that the regression coefficients for the serration angle and serration height were 0.81 and 0.26, respectively, with corresponding p-values ​​of 1.55E-08 and 1.3e-03, both significantly less than 0.05. This indicates a significant statistical correlation between these two variables and the degree of damage. In particular, the regression coefficient for the serration angle was greater than that for the serration height, further illustrating that the serration angle had a greater impact on the degree of damage than the serration height.

[0114] (3-23) in, To standardize the degree of damage.

[0115] Based on the three-dimensional laser scanning data and visual images of the second group of samples (H1~H16), combined with Figure 9 , Figure 10 The influence of sawtooth inclination angle on the degree of damage was analyzed, and the conclusions are as follows: (1) At different saw tooth heights, the correlation coefficients between saw tooth inclination angle and damage degree were 0.86, 0.91, 0.92 and 0.95, respectively, indicating that there is a strong correlation between the two.

[0116] (2) Under the same sawtooth height, the degree of damage increases with the increase of the sawtooth inclination angle, and the two are approximately linearly related. The fitting results are shown in equations (3-24) to (3-27) and Figure 9 As shown in (a), the R-values ​​and regression coefficients are both high, with R values ​​of 0.86, 0.91, 0.92 and 0.95, and regression coefficients of 372.38, 690.24, 1163.97 and 755.05, respectively.

[0117] (3-24) (3-25) (3-26) (3-27) (3) As the serration height increases, the slope of the fitted line between the serration inclination angle and the degree of damage gradually increases, indicating that the influence of the serration inclination angle on the degree of damage becomes more significant (see Equations 3-24 to 3-27). When the protrusion height is 2.83 mm, the sample shows more fracture failure; when the serration inclination angle is 38.51° and 45.71°, the sample almost completely undergoes shear failure, so the slope of the fitted line no longer increases.

[0118] (4) At different sawtooth heights, the range of damage degree for different sawtooth inclination angles was 3813.98 mm³, 6708 mm³, 13081.11 mm³ and 14787.13 mm³, respectively, which further shows that the influence of sawtooth inclination angle on damage degree becomes more significant with the increase of sawtooth height.

[0119] Since the correlation between sawtooth height and damage degree under the same sawtooth inclination angle could not be investigated individually for samples H1~H16, the data of sawtooth height and damage degree of samples H1~H16 were summarized and analyzed. The correlation coefficient between the two was calculated to be 0.54, indicating that the relationship between the two is not as close as that shown by the relationship between sawtooth height and damage degree. Figure 9 (As shown in (b)). According to the previous study, both the sawtooth inclination angle and the sawtooth height are closely related to the degree of damage, but the effect of the sawtooth inclination angle is more significant.

[0120] After standardizing the serration angle, height, and damage degree of samples H1~H16, a quantitative relationship among the three was established through multiple regression analysis. The specific results are as follows: Figure 9 As shown in (c) and Equation 3-28.

[0121] (3-28) After summarizing the test data of all samples (J1~16 and H1~H16), the correlation coefficient between the serration inclination angle and the degree of damage was calculated to be 0.86, and the correlation coefficient between the serration height and the degree of damage was 0.61. After standardizing the serration inclination angle, serration height, and degree of damage, a multiple regression analysis was used to establish the relationship between the three. The regression coefficients for inclination angle and height were 0.77 and 0.14, respectively, indicating a significant correlation between them and the degree of damage. The regression coefficient for serration inclination angle was greater than that for serration height, indicating that the influence of serration inclination angle on the degree of damage was greater than that of serration height.

[0122] (3-29) (3-30) (3-31) In summary, both the serration angle and serration height are significantly positively correlated with the degree of damage, and the serration angle has a greater impact on the degree of damage than the serration height. Based on the fitting results of this experiment, the relationship among the three can be quantitatively described by equations (3-29) to (3-31), where equations (3-29) and (3-30) are the fitting forms of the first and second quantitative models, respectively, and equation (3-31) is the fitting form of the third quantitative model. In practice, parameter calibration can be performed on similar rocks, and the calibration results can be extended to similar scenarios, thereby directly and quantitatively estimating the degree of damage under individual or interactive effects based on the serration angle and height.

[0123] Step 5: Quantification of the damage range of the serrated surface.

[0124] This application comprehensively employs three-dimensional laser scanning technology and computer vision technology to assess the damage extent (D) by measuring the proportion of the projected area of ​​the damaged region on the XOY plane. r Among these methods, 3D laser scanning technology is used to assess significant damage areas on the structural surface, while computer vision technology is used to identify areas with minor scratches. After shearing tests, obvious white scratches usually appear on the surface of the specimen. A specimen that had slipped was washed with water and air-dried. Before cleaning, obvious white scratches appeared on the surface of the specimen, indicating that there was a certain range of damage. However, after cleaning, these scratches became very weak, and the damage marks almost completely disappeared, indicating that the damage to the specimen was extremely minor. Although the damage from these white scratches was low, they were of great significance for a deeper understanding of the shear failure mechanism of the structural surface. Since three-dimensional laser scanning technology is difficult to accurately identify areas with minor scratches, it can be supplemented by computer vision technology. The combination of three-dimensional laser scanning and computer vision technology has the following advantages: (1) Computer vision technology can effectively identify the damage range of areas with minor scratches. Since minor scratches are white stripes and easy to identify by the human eye, the damage range can be accurately determined by manually marking and drawing the outline of the scratch area. (2) Three-dimensional laser scanning technology is suitable for identifying areas with significant damage. For these areas with more obvious damage, three-dimensional laser scanning can accurately capture the damage range on two-dimensional and three-dimensional scales, and the digital image is highly consistent with the actual damage image. Because the fractured and undamaged areas in images of structural damage are not visually different, computer vision technology performs poorly in recognizing such areas.

[0125] Among them, computer vision technology can be existing machine learning recognition models (such as the classic CNN model, the aforementioned YOLO model), and the trained model can be obtained through steps such as collecting sample data and model training. Then, the model can be used to perform recognition to obtain the damaged area. In addition, the damaged area can also be obtained by manually annotating the visual image. As an example, the annotation steps are as follows: (1) Identify the damaged area: Use the image annotation tool X-Anylabeling to perform image segmentation and annotation on the structural surface damage image. According to the damage contour, the damaged area is marked as the "damage" category, and the annotation results of each damaged image are exported as a separate TXT tag file. When selecting the damaged area, rock mechanics experts with rich experience should be invited to participate to ensure the accurate determination of the damaged area and to identify the damage of different degrees as accurately as possible. (2) Calculate the area ratio of the damaged area: By comparing the pixels of the damaged area and the whole image, the number of pixels of the damaged area is first extracted, and then compared with the total number of pixels of the whole image. According to the proportion of pixels occupied by the damaged area, the area ratio of the damaged area in the overall structural surface can be accurately calculated.

[0126] Damage range D of J1-J16 and H1-H16 specimens r As shown in Table 5, Table 5 is as follows: Table 5 Damage range D for all specimens r

[0127] Using the same analytical method as for the degree of damage, the first and second groups of samples were analyzed separately to obtain the influence of sawtooth height and sawtooth inclination angle on the damage range. The analysis conclusions for the first group of samples are as follows: (1) Under different sawtooth inclination angles, the correlation coefficients between sawtooth height and damage range were -0.93, -0.97, -0.43 and -0.94, respectively, indicating that the relationship between the two is fluctuating.

[0128] (2) Under the same sawtooth inclination angle, the damage range tends to decrease as the sawtooth height increases, and the two are approximately linearly related. The fitting results are shown in equations (3-32) to (3-35). The R values ​​are 0.93, 0.97, 0.43 and 0.94, respectively, and the absolute values ​​of the regression coefficients for sawtooth height are generally small, at 0.078, 0.076, 0.039 and 0.023, respectively.

[0129] (3-32) (3-33) (3-34) (3-35) (3) As the sawtooth inclination angle increases, the slope of the fitted straight line between the sawtooth height and the damage range gradually decreases, indicating that the influence of the sawtooth height on the damage range gradually weakens (Equations 32-35).

[0130] (4) Under different sawtooth inclination angles, the range of damage range for different sawtooth heights was 0.081, 0.11, 0.144 and 0.042, respectively, showing a low overall level.

[0131] Since the correlation between sawtooth inclination angle and damage range for samples J1~J16 at the same sawtooth height cannot be investigated individually, the data of sawtooth inclination angle and damage range for samples J1~J16 are summarized and analyzed. The correlation coefficient between the two is calculated to be 0.92, indicating a close relationship between them. As the sawtooth inclination angle increases, the damage range shows an upward trend, and the two are approximately linearly related. The fitting result is shown in equation (3-36), with an R value of 0.92.

[0132] (3-36) After standardizing the serration angle, height, and damage range of samples J1~J16, a quantitative relationship among the three was established through multiple regression analysis, as shown in equation (3-37). The regression analysis results show that the regression coefficient for the serration angle is 1.041, with a corresponding p-value of 1.28E-06, which is much less than 0.05, indicating a significant correlation between the serration angle and the damage range. The regression coefficient for the serration height is -0.206, whose absolute value is significantly lower than that of the serration angle, and the corresponding p-value is 0.104, significantly greater than 0.05, indicating no significant close relationship between the serration height and the damage range.

[0133] (37) in, It is a standardized range of damage.

[0134] The analytical conclusions for the second group of samples are as follows: (1) At different saw tooth heights, the correlation coefficients between saw tooth inclination angle and damage range were 0.83, 0.94, 0.92 and 0.95, respectively, indicating that there is a strong correlation between the two.

[0135] (2) Under the same sawtooth height, the damage range showed an increasing trend with the increase of the sawtooth inclination angle, and the two were approximately linearly related. The fitting results are shown in equations (3-38~3-41). The R values ​​and regression coefficients were both high, with R values ​​of 0.83, 0.94, 0.92 and 0.95, respectively, and regression coefficients for the sawtooth inclination angle of 0.021, 0.045, 0.043 and 0.018, respectively.

[0136] (3-38) (3-39) (3-40) (3-41) Since the correlation between sawtooth height and damage range for samples H1-H16 at the same sawtooth inclination angle could not be analyzed individually, the sawtooth height and damage range data for samples H1-H16 were summarized and analyzed, and the correlation coefficient was calculated to be 0.39. By plotting scatter plots of all sawtooth heights and damage ranges, the results showed no clear pattern between the two. Therefore, there is no significant correlation between sawtooth height and damage range.

[0137] After standardizing the serration angle, height, and damage range of samples H1~H16, a quantitative relationship among the three was established through multiple regression analysis, as shown in equation (3-42). The regression coefficient for the serration angle was 0.909, with a corresponding p-value of 6.5E-04, which is much less than 0.05, indicating a significant correlation between the serration angle and the damage range. The regression coefficient for the serration height was -0.101, whose absolute value was significantly lower than that of the serration angle, and the corresponding p-value was 0.62, which is significantly greater than 0.05, further indicating that there is no significant correlation between the serration height and the damage range.

[0138] (3-42) After summarizing the test data of all samples J1~16 and H1~H16, the correlation coefficient between the sawtooth inclination angle and the damage range was calculated to be 0.85, and the correlation coefficient between the sawtooth height and the damage range was 0.45. After standardizing the sawtooth inclination angle, sawtooth height, and damage range, the relationship between the three was established through multiple regression analysis, as shown in equation (3-44). The regression coefficients for the inclination angle and height were 0.919 and -0.112, respectively, with corresponding p-values ​​of 3.14E-08 and 0.37. The results show that the sawtooth inclination angle is significantly positively correlated with the damage range, while no obvious relationship is observed between the sawtooth height and the damage range.

[0139] (3-43) (3-44) Equations (3-43) and (3-44) are the fitting results of the fourth and fifth quantitative models, respectively. After parameter calibration, the above formulas can be directly applied to the quantitative prediction of damage range in similar rocks.

[0140] Step 6: Analyze the relationship between damage characteristics and strength characteristics.

[0141] After summarizing the test data of all samples J1~16 and H1~H16, the correlation coefficient between the damage range and the damage degree was calculated to be 0.9. As the damage range increases, the damage degree shows an increasing trend, and the two are approximately linearly related, as shown in equation (3-45) and... Figure 11 As shown in (a), the R value is 0.91.

[0142] (3-45) Equation (3-45) is the fitting result of the sixth quantization model.

[0143] After summarizing the test data of all samples J1~16 and H1~H16, the correlation coefficient between damage degree and peak strength was calculated to be 0.91. As the damage degree increases, the peak strength shows an increasing trend, and the two are approximately linearly related. The fitting results are shown in equation (3-46) and... Figure 11 As shown in (b), the R value is 0.91.

[0144] (3-46) Equation (3-46) is the fitting result of the seventh quantization model.

[0145] After summarizing the test data of all samples J1~16 and H1~H16, and standardizing the damage range, damage degree, and peak intensity, a quantitative relationship among the three was established through multiple regression analysis. The specific results are as follows: Figure 11 As shown in equation (c) and equation (3-47), the regression analysis results show that the regression coefficient for damage degree is 0.666, with a p-value of 6.61E-04, which is much smaller than 0.05, indicating a significant correlation between damage degree and peak intensity. However, the regression coefficient for damage range is 0.269, significantly lower than that for damage degree, and the corresponding p-value is 0.13, significantly larger than 0.05, indicating no significant correlation between damage range and peak intensity.

[0146] (3-47) Equation (3-47) is the fitting result of the eighth quantization model.

[0147] To investigate the independent contribution of damage extent to peak intensity, we employed partial regression analysis. The specific steps are as follows: First, all variables, including damage extent, damage severity, and peak intensity, were standardized to eliminate dimensional differences and ensure that all indicators were analyzed on a uniform scale. Next, to study the independent effect of damage extent on peak intensity, we used regression analysis to eliminate the interference of damage severity on peak intensity. Then, we calculated the partial regression coefficient of damage extent, which reflects the independent contribution of damage extent to peak intensity after controlling for the influence of damage severity. This method effectively removes the influence of other variables and reveals the true role of damage extent on peak intensity under the control of damage severity. Figure 12 Equation (3-48) shows the results of the partial regression analysis. Both the regression coefficient and correlation coefficient are low, further indicating that after eliminating the influence of damage degree, there is no significant linear relationship between damage range and peak intensity. Specifically, although damage range may affect peak intensity to some extent, this effect becomes weak and insignificant after controlling for the interference of damage degree. Therefore, a quantitative model of the peak intensity of the structural surface under the sole effect of damage range is not constructed to avoid errors.

[0148] (3-48) in, It is the result of standardizing the peak strength residual value after removing the influence of the degree of damage.

[0149] In summary, a significant positive correlation exists between the sawtooth inclination angle and the damage range, while no clear regularity is observed between the sawtooth height and the damage range. The correlation between the two gradually weakens as the sawtooth inclination angle increases. The quantitative relationship among the three can be described by equations (3-43) and (3-44). A significant positive correlation exists between the damage range and the damage degree, which can be quantitatively described by equation (3-45). A significant positive correlation also exists between the damage degree and the peak strength. After removing the influence of the damage degree, no significant linear relationship is observed between the damage range and the peak strength. The quantitative relationship among the three can be described by equations (46-47). These quantitative relationships can accurately and comprehensively capture and analyze the damage characteristics and damage evolution laws of structural surfaces, effectively guiding the strength assessment, rock mass stability evaluation, and support design of engineering projects.

[0150] By combining 3D laser scanning and computer vision technologies, high-precision capture and analysis of structural surface damage features are achieved, and a method for quantifying the damage range and degree is proposed. Damaged areas on structural surfaces can be divided into two categories: one is the scratched area, which is lightly damaged, smooth to the touch, and has no obvious roughness; the other is the fractured area, which is more deeply damaged and rough to the touch. 3D laser scanning technology can accurately identify fractured areas, but it has difficulty identifying minor scratched areas; while computer vision technology can accurately identify minor scratched areas, but its identification of significant fractured damage areas is not precise enough. By judging the damage and selecting between the two technical approaches, the damage range of the structural surface can be accurately quantified. Since the degree of damage in the fractured area is dominant in the overall structural surface, it is feasible to use 3D laser scanning technology to study the degree of damage on structural surfaces.

[0151] Both sawtooth inclination angle and sawtooth height are significantly positively correlated with the degree of structural surface damage, with the sawtooth inclination angle having a greater impact on the damage degree than the sawtooth height. Based on this, a quantitative model is constructed to assess the individual effects and interactions of sawtooth inclination angle and sawtooth height on the degree of structural surface damage. A significant positive correlation is found between sawtooth inclination angle and damage range, while no clear correlation is observed between sawtooth height and damage range. The correlation weakens with increasing sawtooth inclination angle; therefore, a quantitative model is constructed to assess the interaction between sawtooth inclination angle and sawtooth height on the damage range of the structural surface. A significant positive correlation exists between damage range and damage degree; based on this, a quantitative model is constructed to assess the interaction between the two. A significant positive correlation is also found between damage degree and peak intensity, and after removing the influence of damage degree, no significant linear relationship is observed between damage range and peak intensity. Therefore, a quantitative model is constructed to assess the interaction between damage range and damage degree on the peak intensity of the structural surface.

[0152] This embodiment also provides a computer system, including a memory, a processor, and a computer program stored in the memory, characterized in that the processor executes the computer program to implement the structural surface damage feature quantification method of any of the above embodiments, or the steps of the structural surface strength feature quantification method of any of the above embodiments.

[0153] This embodiment also provides a computer-readable storage medium storing a computer program / instruction thereon, characterized in that, when the computer program / instruction is executed by a processor, it implements the steps of any of the structural surface damage feature quantification methods described in the above embodiments, or the structural surface strength feature quantification methods described in any of the above embodiments.

[0154] This embodiment also provides a computer program product, including a computer program / instruction, characterized in that, when the computer program / instruction is executed by a processor, it implements the structural surface damage feature quantification method described in any of the above embodiments, or the steps of the structural surface strength feature quantification method described in any of the above embodiments.

[0155] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for quantifying structural surface damage characteristics, characterized in that, include: Indoor shear tests were conducted on the sawtooth structure surface. Three-dimensional laser scanning technology was used to obtain point cloud data of the structure surface before and after shearing, and visual images of the structure surface were also acquired. The damaged area is initially extracted, and the overall volume difference of the structural surface specimen before and after shearing is calculated based on the initially extracted damaged area. This volume difference is used to measure the degree of damage to the structural surface, denoted as D. d ; According to D d To assess the overall damage to the structural surface, three-dimensional laser scanning technology is used to quantitatively evaluate the damage range for severely damaged areas. For structural surfaces with minor overall damage, or areas with minor damage within severely damaged structural surfaces, computer vision technology is used to finely extract the damaged areas, and the proportion of the projected area of ​​the damaged area on the XOY plane is calculated to quantify the damage range of the structural surface, denoted as D. r ; Using damage level D d and damage range D r Commonly characterizing structural surface damage features.

2. The method according to claim 1, characterized in that, Also includes: Based on the results of the three-dimensional laser scanning test, the relationship between the sawtooth inclination angle and sawtooth height of the sawtooth structure surface and the degree and extent of damage to the structure surface is analyzed. Based on the linear relationship between sawtooth inclination angle, sawtooth height and the degree of structural surface damage, and considering that the sawtooth inclination angle has a greater impact on the degree of damage than the sawtooth height, a quantitative model is constructed to assess the degree of structural surface damage caused by the individual effects of sawtooth inclination angle and sawtooth height, as well as their interaction. Based on the linear relationship between sawtooth inclination angle and damage range, and the lack of a regular relationship between sawtooth height and damage range, a quantitative model is constructed to assess the impact of sawtooth inclination angle alone, as well as the interaction between sawtooth inclination angle and sawtooth height on the damage range of the structural surface. Based on the linear relationship between the damage range and the damage degree, a quantitative model of the relationship between the damage range and the damage degree is constructed.

3. The method according to claim 1, characterized in that, According to D d To assess the overall damage to the structural surfaces, the following steps are taken: The degree of damage D d Compared with a preset damage threshold, if the damage degree D d If the damage is less than the damage threshold, the structural surface is determined to be slightly damaged; otherwise, the structural surface is determined to be severely damaged.

4. The method according to claim 1, characterized in that, Also includes: Observe the visual image of the structural surface after shearing to determine the overall damage to the structural surface, as follows: If the structure surface only shows white band-shaped scratches and no breakage is seen in the visual image after cutting, it is considered to be slightly damaged. If the visual image after cutting shows both white band-shaped scratches and fracture areas on the structural surface, with the fracture area being dominant, it indicates severe damage to the structural surface.

5. The method according to claim 1, characterized in that, For severely damaged areas, three-dimensional laser scanning technology is used to quantitatively assess the extent of the damage, including: The cut point cloud data is subjected to boundary clipping and interpolation homogenization to obtain the interpolated point cloud data. The structural surface is discretized into spatial cubes. For each cube, the Z coordinate values ​​of the point cloud data before and after shearing at the same (X, Y) position are differentially calculated to obtain the height difference ΔZ. Based on experimental calibration, the critical value ΔZ0 for damage discrimination is determined, and the regions at the four vertices of each cube that satisfy ΔZ≥ΔZ0 are identified as the damaged regions of the structural surface. Accordingly, the volume difference of the structural surface specimen before and after shearing is calculated based on the damaged area, including: Calculate the sum of the volume differences of all cubes within all damaged regions before and after shearing, and denote this as the damage degree D. d .

6. A method for quantifying the strength characteristics of structural surfaces, characterized in that, include: The damage degree D of the structural surface is calculated using the structural surface damage feature quantification method as described in any one of claims 1 to 5. d and damage range D r ; Based on the linear relationship between damage range, damage degree and peak intensity, a quantitative model of the peak intensity of the structural surface under the sole effect of damage degree is constructed, as well as a quantitative model of the effect of the interaction between damage range and damage degree on the peak intensity of the structural surface.

7. A computer system comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the structural surface damage feature quantification method of any one of claims 1 to 5, or the structural surface strength feature quantification method of claim 6.

8. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the structural surface damage feature quantification method of any one of claims 1 to 5, or the structural surface strength feature quantification method of claim 6.

9. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the structural surface damage feature quantification method of any one of claims 1 to 5, or the structural surface strength feature quantification method of claim 6.

Images