Steel fiber grading method and system based on skeleton gap reverse adaptation

Abstract

The application discloses a steel fiber grading method and system based on skeleton gap reverse adaptation, and comprises the following steps: obtaining a tomographic image of a reference test piece; performing threshold segmentation on the tomographic image to obtain mutually distinguished aggregate area and skeleton gap area, and defining the interface between the aggregate area and the skeleton gap area as an aggregate boundary; adopting a marker-controlled watershed algorithm based on depth threshold constraint to segment the skeleton gap area into multiple independent gap units; drawing an equivalent ellipse for each gap unit, taking the minor axis length of the equivalent ellipse as the gap size of the gap unit, and counting the frequency distribution of the gap size of all the gap units; and determining the mass proportion of steel fibers with different lengths according to the frequency distribution of the gap size, so that the length grading of the steel fibers matches the size distribution of the skeleton gap. The application can start from the spatial characteristics of the coarse aggregate skeleton, realize active adaptation of fiber distribution and matrix microstructure, and significantly improve the performance of UHPC-CA.

Description

Technical Field

[0001] This invention relates to the field of ultra-high performance concrete material gradation design technology, specifically to a steel fiber gradation method and system based on reverse adaptation of skeleton gaps. Background Technology

[0002] Ultra-high performance concrete (UHPC) has broad application prospects in the engineering field due to its extremely high strength and excellent durability. UHPC-CA (UHPC-containing ultra-high performance concrete) with coarse aggregate can further reduce costs, suppress shrinkage, and increase the modulus of elasticity. However, the introduction of coarse aggregate constitutes a rigid spatial skeleton network, which significantly interferes with and restricts the distribution of steel fibers.

[0003] Currently, steel fiber design generally suffers from the following problems: In terms of gradation design, single-length fibers or simple blending based on experience are often used, lacking a scientific basis for determining the proportion of fibers of different lengths and lacking a correlation with the internal space capacity of the matrix. This leads to severe space competition between fibers and coarse aggregates, which weakens the toughening effect and degrades workability.

[0004] In the field of concrete microstructure analysis, X-ray computed tomography (X-CT) technology has been widely used. As a non-destructive testing method, X-CT can acquire two-dimensional slice images of the material's interior. Image segmentation techniques can divide the image into different regions based on density differences, enabling the identification and extraction of components such as coarse aggregate, matrix, and pores. Currently, commonly used image segmentation methods include traditional algorithms such as threshold segmentation, region growing, and edge detection, as well as deep learning-based segmentation methods. The application of these image segmentation techniques is mostly for analyzing the internal distribution patterns and pore structure characteristics of materials, or establishing the correlation between phase distribution and mechanical properties; none of these are directly related to steel fiber gradation design.

[0005] In the field of image segmentation, the watershed algorithm is a classic segmentation method. Traditional watershed algorithms are usually used in conjunction with gradient images, which is the most classic and mainstream technical approach. However, directly applying the traditional watershed algorithm to the image gradient in tomographic images of ultra-high performance concrete containing coarse aggregates faces an insurmountable contradiction: if the original gradient image is directly segmented, a large number of pseudo-minimum points generated by noise and minor undulations at aggregate edges are treated as independent segmentation starting points, leading to severe oversegmentation and numerous erroneous decompositions. These segmentation results cannot be directly used for subsequent dimensional measurements.

[0006] This technical deficiency has been confirmed by numerous studies in the field, and existing improvement schemes mostly use gradient correction methods to suppress it. Although the gradient correction method can suppress oversegmentation to a certain extent, its essence is to remedy the situation after the gradient image is generated, rather than to avoid the generation of pseudo-minimum values ​​from the source. This method is highly sensitive to morphological parameters. Overcorrection will lead to blurring of the true gap boundary or even undersegmentation, while undercorrection will leave pseudo-minimum values, resulting in severe oversegmentation. Moreover, it processes the gray-scale change rate rather than spatial geometric distance, and the segmentation result lacks a direct correspondence with the true capacity of the gap. It has always been unable to provide accurate and directly measurable gap units for steel fiber gradation design. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention provides a steel fiber gradation method and system based on reverse adaptation of skeleton gaps. It can start from the spatial characteristics of the coarse aggregate skeleton to achieve active adaptation between fiber distribution and matrix microstructure, significantly improving the performance of UHPC-CA.

[0008] Firstly, a steel fiber gradation method based on reverse adaptation of skeleton gaps includes: Obtain tomographic images of the skeleton reference specimen to be analyzed, wherein the skeleton reference specimen does not contain steel fibers; The tomographic scan image is subjected to threshold segmentation to obtain a threshold segmentation image; the threshold segmentation image contains mutually distinguishable aggregate regions and skeleton gap regions, and the interface between the aggregate regions and the skeleton gap regions is defined as the aggregate boundary; A marker-controlled watershed algorithm based on depth threshold constraints is used to divide the skeleton gap region into multiple independent gap units; For each gap element, an equivalent ellipse is drawn, and the length of the minor axis of the equivalent ellipse is used as the gap size of the gap element. The frequency distribution of the gap sizes of all gap elements is statistically analyzed, and the frequency distribution is used to determine the length gradation of the steel fibers. Based on the frequency distribution of the gap size, the mass ratio of steel fibers of different lengths is determined so that the length gradation of the steel fibers matches the size distribution of the skeleton gap.

[0009] In one possible implementation of the first aspect, the skeleton reference specimen uses continuously graded coarse aggregate with a volumetric content of 15% to 20%.

[0010] In one possible implementation of the first aspect, the threshold segmentation process is a binarization process.

[0011] In one possible implementation of the first aspect, the step of extracting foreground markers based on a minimum gap depth threshold and segmenting the skeleton gap region into multiple independent gap units using a watershed algorithm includes: A distance transformation is performed on the skeleton gap region to generate a distance map; the pixel value of each pixel in the distance map is used to represent the shortest distance from that pixel to the aggregate boundary; local maxima in the distance map are determined by pixel neighborhood comparison; Set a minimum gap depth threshold H, where the minimum gap depth threshold H is a physical length value used to characterize the radial depth from the center of the gap unit to the edge of the gap unit; Based on the spatial resolution of the tomographic scan image, the minimum gap depth threshold H is converted into a pixel unit threshold. Extract local maxima points in the distance map whose pixel values ​​are greater than or equal to the unit threshold of that pixel, and use them as foreground marker points; Extract background marker points for the aggregate area, the background marker points being used to identify the aggregate area; Modifying the distance map includes: setting the pixel value at the location of the foreground marker point to a local minimum, and eliminating all local minimums at locations other than the foreground marker point based on the locations of the aggregate boundary and the background marker points; A watershed transformation is performed on the modified distance map, with each foreground marker point serving as an injection point, spreading outwards until the spread areas of different injection points intersect. A watershed line is constructed at the intersection, and the watershed line is the edge of each gap unit. The connected skeleton gap region is divided into several independent gap units through the watershed line.

[0012] In one possible implementation of the first aspect, the step of performing distance transformation on the skeleton gap region to generate a distance map includes: Obtain the threshold segmentation image corresponding to the skeleton gap region; For each pixel in the threshold segmentation image, traverse the aggregate boundaries surrounding the skeleton gap region, calculate the distance from the pixel to each aggregate boundary, and take the minimum value among them; The calculated minimum value is assigned to the corresponding pixel to generate a distance map.

[0013] In one possible implementation of the first aspect, based on the minimum length L of the steel fiber min Set the minimum gap depth threshold H, where H ≥ L min .

[0014] In one possible implementation of the first aspect, the statistical analysis of the frequency distribution of the gap dimensions of all gap elements includes: The gap size is divided into multiple continuous size intervals. The number or proportion of gap units in each size interval is counted, and a frequency histogram is constructed.

[0015] In one possible implementation of the first aspect, determining the mass ratio of steel fibers of different lengths based on the frequency distribution of the gap size includes: Based on the frequency distribution of gap dimensions, a correspondence between size ranges and steel fiber lengths is established, so that each gap unit is classified into the corresponding steel fiber length according to its size. The number of interstitial units corresponding to each length of steel fiber was counted to determine the volumetric dosing of each length of steel fiber, and the corresponding dosing mass was calculated based on the density of each length of steel fiber.

[0016] Secondly, a system for performing the steel fiber gradation method based on skeleton gap reverse fitting includes: The image scanning module is used to acquire tomographic images of the skeleton reference specimen to be analyzed. The image processing module is used to perform threshold segmentation on the tomographic scan image to obtain mutually distinguishable aggregate regions and skeleton gap regions, and the interface between the aggregate regions and the skeleton gap regions is defined as the aggregate boundary. The gap segmentation module uses a marker-controlled watershed algorithm based on depth threshold constraints to segment the skeleton gap region into multiple independent gap units; The statistical analysis module is used to draw an equivalent ellipse for each gap unit, and use the length of the minor axis of the equivalent ellipse as the gap size of the gap unit, and to statistically analyze the frequency distribution of the gap size of all gap units. The gradation design module is used to determine the mass ratio of steel fibers of different lengths based on the frequency distribution of the gap size, so that the length gradation of the steel fibers matches the size distribution of the skeleton gap.

[0017] In one possible implementation of the second aspect, the gap segmentation module includes: A distance transformation unit is used to perform distance transformation on the skeleton gap region to generate a distance map; the pixel value of each pixel in the distance map is used to represent the shortest distance from the pixel to the aggregate boundary; local maxima points in the distance map are determined by pixel neighborhood comparison; The marker extraction unit is used to set the minimum gap depth threshold H and convert the minimum gap depth threshold H into a pixel unit threshold according to the spatial resolution of the tomographic scan image; extract local maxima points in the distance map whose pixel values ​​are greater than or equal to the pixel unit threshold as foreground marker points; and extract background marker points in the aggregate area to identify the aggregate area. The terrain modification unit is used to modify the distance map, including: setting the pixel value at the position of the foreground marker point to a local minimum, and eliminating all local minimums at non-foreground marker point positions according to the positions of the aggregate boundary and the background marker point; The watershed processing unit is used to perform watershed transformation on the modified distance map, spreading outwards from each foreground marker point as an injection point until the spread areas of different injection points intersect each other, and constructing a watershed line at the intersection. The watershed line is the edge of each gap unit; the connected skeleton gap region is divided into several independent gap units through the watershed line.

[0018] Compared with existing technologies, this invention starts from the real three-dimensional gap space of the coarse aggregate skeleton inside UHPC-CA and establishes a set of steel fiber design criteria based on tomographic scanning, a marker-controlled watershed algorithm based on depth threshold constraints, and the equivalent minor axis of an ellipse. This fills the technical gap in the field of UHPC fiber design containing coarse aggregate, which lacks microscopic geometric basis, and overcomes the blindness of traditional empirical methods. Specifically, this application has the following beneficial effects: 1. Replacing grayscale gradient with distance transformation to directly reflect the spatial geometric features of the gap: This invention first performs distance transformation on the skeleton gap area to generate a distance map. The pixel value in the distance map directly represents the shortest spatial distance from the point to the aggregate boundary, rather than the gradient value. Compared with the traditional operation of directly using grayscale gradient images, this invention directly reflects the spatial geometric features of the gap through the topographic undulation of the distance map, laying the foundation for subsequent accurate segmentation.

[0019] 2. Suppressing Oversegmentation at the Root Cause through a Two-Level Filtering Mechanism: Traditional watershed algorithms only use all marker points as the starting point for segmentation. This invention suppresses oversegmentation through a two-level filtering mechanism of depth threshold constraint and background constraint elimination. First, based on the minimum gap depth threshold, foreground marker points with sufficient radial depth to accommodate fibers are selected from the positive direction. Thus, the generation of marker points is no longer controlled by image noise, but by the true gap space scale. Second, based on the aggregate boundary and background marker points, local minima (including pseudo-minima) at all non-foreground marker point positions are eliminated from the negative direction, retaining only the minimum value depression at the marker point position. This ensures that the watershed transformation only spreads from the true gap center, suppressing oversegmentation at the root cause, ensuring that the segmentation result corresponds one-to-one with the true gap in physical space, and ensuring the accuracy of the segmentation result. Attached Figure Description

[0020] Figure 1 A flowchart illustrating a steel fiber gradation method based on reverse adaptation of skeleton gaps, provided for embodiments of this application; Figure 2 A flowchart of a marker-controlled watershed algorithm based on depth threshold constraints provided for embodiments of this application; Figure 3 A CT scan image of a UHPC-CA skeleton reference specimen provided for an embodiment of this application; Figure 4(a) is a CT scan enhanced image of a UHPC-CA skeleton reference specimen provided in an embodiment of this application; Figure 4(b) is a CT scan binarized image of a UHPC-CA skeleton reference specimen provided in an embodiment of this application; Figure 5(a) is a skeleton gap image segmented using the watershed algorithm provided in the embodiment of this application; Figure 5(b) is a magnified view of position A in Figure 5(a); Figure 6(a) is a diagram of the gap element markings obtained after dividing and numbering the watershed in Figure 5(a); Figure 6(b) is the equivalent ellipse diagram of the gap element corresponding to measurement area B in Figure 6(a); Figure 7 A frequency histogram of skeleton gap dimensions is provided for the embodiments of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the application will now be described in further detail with reference to the accompanying drawings.

[0022] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0023] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0024] like Figures 1-2 As shown, a steel fiber gradation method based on reverse adaptation of skeleton gaps includes the following steps: S100: Obtain a tomographic image of the UHPC-CA skeleton reference specimen to be analyzed, wherein the UHPC-CA skeleton reference specimen does not contain steel fibers; S200: The tomographic scan image is subjected to threshold segmentation processing to obtain a threshold segmentation image. The threshold segmentation image contains mutually distinguishable aggregate regions and skeleton gap regions. The interface between the aggregate regions and the skeleton gap regions is defined as the aggregate boundary. S300: The marker-controlled watershed algorithm based on depth threshold constraints is used to divide the skeleton gap region into multiple independent gap units; S400: Draw an equivalent ellipse for each gap unit, and use the length of the minor axis of the equivalent ellipse as the gap size of the gap unit. Statistically analyze the frequency distribution of the gap sizes of all gap units. The frequency distribution is used to determine the length gradation of the steel fibers. The equivalent ellipse refers to an ellipse with the same area and the same second moment as the target gap unit. The major axis of the equivalent ellipse represents the longest extension direction of the gap unit, and the minor axis represents the minimum width perpendicular to the major axis direction. S500: Based on the frequency distribution of the gap size, determine the mass ratio of steel fibers of different lengths so that the length gradation of the steel fibers matches the size distribution of the skeleton gap.

[0025] In one optional implementation, the skeleton reference specimen in step S100 uses continuously graded coarse aggregate, and the volumetric content of the coarse aggregate is generally 15%-20%.

[0026] In one optional implementation, step S200 employs binarization processing to perform threshold segmentation on the computed tomography image.

[0027] In one optional implementation, step S300 specifically includes the following sub-steps: S310: Perform distance transformation on the skeleton gap region to generate a distance map; the pixel value of each pixel in the distance map is used to represent the shortest distance from the pixel to the aggregate boundary; determine the local maxima in the distance map by comparing the pixel neighborhood; It should be noted that, for the entire tomographic scan image, the skeletal gap region is a single connected void composed of multiple gap units. In the distance map, a larger pixel value indicates that the pixel is farther from the aggregate boundary and has a more open local space. The pixel / set of pixels whose pixel value is the local maximum within each gap unit is defined as the local maximum point / local maximum region of that gap unit. The pixel value of a local maximum point is greater than or equal to the pixel values ​​of all its adjacent pixels, and at least one adjacent pixel has a pixel value less than that point. When multiple adjacent pixels have equal pixel values, these multiple adjacent pixels together constitute a local maximum region. The local maximum point / local maximum region represents the position farthest from the aggregate boundary within the gap unit and can be considered as the center point / central region of the gap unit. S320: Set a minimum gap depth threshold H, where the minimum gap depth threshold H is a physical length value used to characterize the radial depth from the center of the gap unit to the edge of the gap unit, i.e., the effective accommodating depth of the gap unit; S330: Based on the spatial resolution of the tomographic scan image, convert the minimum gap depth threshold H into a pixel unit threshold. S340: Extract local maxima points whose distance from the pixel value in the image is greater than or equal to the unit threshold of that pixel, and use them as foreground marker points; It should be noted that the minimum gap depth threshold H is a preset physical length value, based on the minimum length L of the target steel fiber. min To confirm, for example, you can directly set H = L min The physical meaning is: only when the actual physical distance from a local maximum point in the diagram to the nearest aggregate boundary is greater than or equal to H, is the gap element corresponding to that local maximum point considered to have sufficient depth to accommodate a length not less than L. min Steel fibers.

[0028] Before extracting foreground markers, the physical length needs to be converted into a pixel unit threshold H based on the spatial resolution of the tomographic image. pixel When the spatial resolution of the image is known, H pixel = H / r , r Spatial resolution (mm / pixel); H pixel As a pixel-level threshold, extract pixels with a distance from the image greater than or equal to H. pixel The local maxima are used as foreground markers; S350: Extract background marker points of the aggregate region to identify the aggregate region; the background marker points are used to identify the aggregate region when modifying the distance map in the future, so as to eliminate local minima caused by noise or material unevenness in the aggregate region and prevent the watershed algorithm from generating invalid segmentation in the aggregate. S360: Modify the distance map, including: setting the pixel value at the position of the foreground marker point to a local minimum, and eliminating all local minimums at non-foreground marker point positions according to the positions of the aggregate boundary and the background marker point; It should be noted that the remaining local minima mainly include local minima within the aggregate region, local minima at the aggregate boundary (true minima), and a large number of pseudo-minima points caused by noise and aggregate boundary undulations; among them, pseudo-minima points are usually distributed in the skeleton gap region; if these local minima are not eliminated, the watershed algorithm will spread from the wrong position, resulting in severe oversegmentation of the gap cells; Therefore, the present invention modifies the distance map so that the local minimum exists only at the position of the foreground marker point in the modified distance map, and there are no local minimums in the other areas. The purpose is to make the watershed transformation spread only from the position of the foreground marker point, thereby filtering out pseudo gaps that cannot accommodate steel fibers and suppressing over-segmentation. S370: Perform a watershed transformation on the modified distance map, spreading outwards from each foreground marker point as an injection point until the spread areas of different injection points intersect. Construct a watershed line at the intersection, which is the edge of each gap unit. Divide the connected skeleton gap region into several independent gap units through the watershed line.

[0029] In an optional implementation, step S310 performs a distance transformation on the skeleton gap region to generate a distance map, specifically including the following sub-steps: S311: Obtain the threshold segmentation image corresponding to the skeleton gap region; use the aggregate boundary as the reference object for distance calculation, thereby determining a spatial reference benchmark for distance transformation; S312: For each pixel in the threshold segmentation image, traverse the aggregate boundaries surrounding the skeleton gap region, calculate the distance from the pixel to each aggregate boundary, and take the minimum value among them; S313: Assign the calculated minimum value to the corresponding pixel to generate a distance map.

[0030] In an alternative implementation, step S320 is based on the minimum length L of the steel fiber. min Set the minimum gap depth threshold H, and H ≥ L min .

[0031] In one optional implementation, step S400 involves statistically analyzing the frequency distribution of the gap dimensions of all gap elements, specifically as follows: The gap size (equivalent minor axis length of an ellipse) is divided into multiple continuous size intervals. The number or proportion of gap units in each size interval is counted, and a frequency histogram is constructed.

[0032] In one optional implementation, step S500 determines the mass ratio of steel fibers of different lengths based on the frequency distribution of the gap size, including: Based on the frequency distribution of gap dimensions, a correspondence between size ranges and steel fiber lengths is established, so that each gap unit is classified into the corresponding steel fiber length according to its size. The number of interstitial units corresponding to each length of steel fiber is counted to determine the volumetric dosing of each length of steel fiber. The corresponding dosing mass is calculated based on the density of each length of steel fiber. Smaller interstitial units are suitable for shorter fibers, and larger interstitial units are suitable for longer fibers.

[0033] Based on the same inventive concept, the present invention also provides a system for performing the steel fiber gradation method based on UHPC-CA skeleton gap reverse adaptation, comprising: The image scanning module acquires tomographic images of the skeleton reference specimen to be analyzed; The image processing module is used to perform threshold segmentation on the tomographic scan image to obtain mutually distinguishable aggregate regions and skeleton gap regions, and the interface between the aggregate regions and the skeleton gap regions is defined as the aggregate boundary. The gap segmentation module uses a marker-controlled watershed algorithm based on depth threshold constraints to segment the skeleton gap region into multiple independent gap units; The statistical analysis module draws an equivalent ellipse for each gap unit, uses the length of the minor axis of the equivalent ellipse as the gap size of the gap unit, and statistically analyzes the frequency distribution of the gap sizes of all gap units. The gradation design module determines the mass ratio of steel fibers of different lengths based on the frequency distribution of gap dimensions, so that the length gradation of steel fibers matches the size distribution of the skeleton gaps.

[0034] Furthermore, the gap segmentation module includes the following processing unit: A distance transformation unit is used to perform distance transformation on the skeleton gap region to generate a distance map; the pixel value of each pixel in the distance map is used to represent the shortest distance from the pixel to the aggregate boundary; local maxima points in the distance map are determined by pixel neighborhood comparison; The marker extraction unit is used to set the minimum gap depth threshold H, convert the minimum gap depth threshold H into a pixel unit threshold, extract local maxima points in the distance map whose pixel values ​​are greater than or equal to the pixel unit threshold, and use them as foreground marker points; and extract background marker points in the aggregate area. The terrain modification unit is used to modify the distance map, including: setting the pixel value at the location of the foreground marker point to a local minimum and eliminating the remaining local minimums; The watershed processing unit is used to perform watershed transformation on the modified distance map, spreading outwards from each foreground marker point as an injection point until the spread areas of different injection points intersect each other, and constructing a watershed line at the intersection. The watershed line is the edge of each gap unit; the connected skeleton gap region is divided into several independent gap units through the watershed line.

[0035] It should be understood that the division of processing modules / units in the above system is only a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, the processing modules / units in the system can be implemented by a processor calling software; for example, the system includes a processor connected to memory, which stores instructions. The processor calls the instructions stored in memory to implement any of the above methods or to realize the functions of each processing module / unit in the system. Here, the processor is a general-purpose processor, such as a central processing unit or a microprocessor, and the memory is either internal or external to the system.

[0036] The feasibility and effectiveness of this application will be explained in detail below with an application example.

[0037] Step 1: Prepare UHPC-CA skeleton reference specimens without steel fibers, wherein the volume content of coarse aggregate is 20%, continuous gradation is adopted, and the particle size range is 2.36mm~16mm; Those skilled in the art can adjust the volumetric content of coarse aggregate within the range of 15% to 20% according to actual needs, and it is not limited to 20% in this embodiment.

[0038] Step Two: Perform tomographic scanning on the specimen using industrial X-ray CT scanning to obtain CT scan images of the internal structure of the UHPC-CA skeleton reference specimen, such as... Figure 3 As shown in the figure, Pore represents pores, UHPC matrix represents the UHPC matrix, and Coarseaggregate represents coarse aggregate.

[0039] Step 3: Enhance the contrast of the CT scan image and then perform binarization to distinguish between the coarse aggregate area and the skeleton gap area. As shown in Figure 4(a) and Figure 4(b), the enhanced CT scan image and the binarized image of the skeleton reference specimen are respectively. It can be seen that the skeleton gap after binarization is in a large-area connected state, which needs to be divided into independent gap units.

[0040] Step 4: Using a label-controlled watershed algorithm based on depth threshold constraints, the skeleton gap region shown in Figure 4(b) is divided into multiple independent gap units. The specific operation is as follows: (1) Perform distance transformation on the binarized image to obtain a distance map. The value of each pixel in the distance map represents the shortest distance from that point to the nearest aggregate boundary, forming a topographic surface that reflects the spatial morphology of the gaps. (2) Set the minimum gap depth threshold H according to the length of the target steel fiber; in this embodiment, the designed length of the target steel fiber includes 3mm, 6mm, 8mm, 10mm, 13mm, and 16mm, and the minimum fiber length is 3mm, so H is set to 3mm. If the on-site construction requires a longer minimum fiber length (such as 6mm), the value of H is increased accordingly (such as H = 6mm); convert H to a pixel unit threshold according to the spatial resolution, and extract the pixel values ​​in the distance map that are greater than or equal to H. pixel The local maxima points are used as foreground markers; at the same time, background markers of the aggregate region are extracted to eliminate local minima within the aggregate region in subsequent steps, thereby avoiding segmentation within the aggregate region. (3) Modify the distance image, keeping only the foreground markers at the skeleton gaps, resetting the local maxima at the foreground markers to local minima, and eliminating the rest of the local minima (including but not limited to local minima inside the aggregate region, local minima at the aggregate boundary, and a large number of pseudo-minima points in the skeleton gap region caused by noise and aggregate boundary undulations); after processing, the image only has depressions at the marker locations, and the terrain in the rest of the area is flat. (4) Perform watershed transformation on the modified distance map. Water spreads from each foreground marker point to the surrounding areas and watershed lines are constructed at the intersection of different areas of influence, as shown in Figure 5(a) and Figure 5(b). Thus, the connected gaps are accurately divided into several independent gap units, each unit corresponding to the real gap area surrounded by coarse aggregate in the physical space.

[0041] Step 5: Number each gap unit to form several measurement areas, as shown in Figure 6(a). Each independent gap unit is assigned a unique number (e.g., 1, 2, 3...), forming several identifiable measurement areas. Draw an equivalent ellipse for each gap unit, using the length of the minor axis of the equivalent ellipse to characterize the gap size, as shown in Figure 6(b), which is the equivalent ellipse diagram of the gap unit corresponding to measurement area B in Figure 6(a). Divide the length of the minor axis of the equivalent ellipse into multiple continuous size intervals, count the number or proportion of gap units in each size interval, construct a frequency histogram, and count its frequency distribution; Figure 7 As shown, in this embodiment, the width of the size range is 1mm, and the size range covers the range from 0mm to 18mm; in the figure, d50 and d90 represent the gap size when the cumulative frequency is 50% and 90%, respectively. d50=5.45 means that 50% of the gap sizes are below 5.45mm, and d90=10.42 means that 90% of the gap sizes are below 10.42mm; μ g (q0) represents the peak value of the fitted curve, μ g (q0)=5.71866±0.13723 indicates that the gap size of 5.71866mm is the most frequently distributed length.

[0042] Step 6: Based on the frequency distribution of the gap size, determine the mass of six steel fibers (3mm, 6mm, 8mm, 10mm, 13mm, 16mm) within the range of 3mm to 16mm, with a steel fiber volume content of 2%. The specific criteria are as follows: In this embodiment, six lengths of steel fibers—3mm, 6mm, 8mm, 10mm, 13mm, and 16mm—are used. When the gap size is less than 4.5mm, a 3mm steel fiber is used; when the gap size is between 4.5mm and 7.0mm, a 6mm steel fiber is used; when the gap size is between 7.0mm and 9.0mm, an 8mm steel fiber is used; when the gap size is between 9.0mm and 11.5mm, a 10mm steel fiber is used; when the gap size is between 11.5mm and 14.5mm, a 13mm steel fiber is used; and when the gap size is greater than 14.5mm, a 16mm steel fiber is used. After obtaining the volumetric dosage of different steel fibers, the corresponding dosage mass is calculated, as shown in Table 1.

[0043] Table 1. Mass of steel fibers of different sizes added to UHPC-CA 20%-2% specimens.

[0044] Using the steel fiber composite admixture scheme of this invention, at the same total steel fiber volume content (2.0%), compared with the traditional single-length fiber scheme, the 28-day flexural strength of UHPC-CA increased from 26.9 MPa to 31.8 MPa, an increase of 18.2%, and the splitting tensile strength increased from 17.2 MPa to 22.3 MPa, an increase of 29.7%, with a slight increase in compressive strength. Simultaneously, due to the better compatibility between the steel fibers and the matrix spatial structure designed in this invention, spatial interference between fibers and aggregates is effectively reduced, resulting in better spreadability of fresh concrete at the same admixture content compared to the traditional scheme, thus improving workability. CT scan verification shows that the steel fibers designed in this invention are more uniformly dispersed in the matrix, effectively avoiding fiber clumping and forming a synergistic reinforcing network where short fibers fill micro-gaps and long fibers bridge cracks, maximizing the toughening efficiency of the steel fibers.

[0045] The above embodiments are only for illustrating the technical concept and features of this application, and are intended to enable those skilled in the art to understand the content of this application and implement it accordingly. They should not be used to limit the scope of protection of this application. For those skilled in the art, several improvements and modifications can be made without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for steel fiber gradation based on reverse adaptation of skeleton gaps, characterized in that, include: Obtain tomographic images of the skeleton reference specimen to be analyzed, wherein the skeleton reference specimen does not contain steel fibers; The tomographic scan image is subjected to threshold segmentation to obtain a threshold segmentation image; the threshold segmentation image contains mutually distinguishable aggregate regions and skeleton gap regions, and the interface between the aggregate regions and the skeleton gap regions is defined as the aggregate boundary; A marker-controlled watershed algorithm based on depth threshold constraints is used to divide the skeleton gap region into multiple independent gap units; For each gap element, an equivalent ellipse is drawn, and the length of the minor axis of the equivalent ellipse is used as the gap size of the gap element. The frequency distribution of the gap sizes of all gap elements is statistically analyzed, and the frequency distribution is used to determine the length gradation of the steel fibers. Based on the frequency distribution of the gap size, the mass ratio of steel fibers of different lengths is determined so that the length gradation of the steel fibers matches the size distribution of the skeleton gap.

2. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 1, characterized in that, The reference specimen for the skeleton uses continuously graded coarse aggregate, with a volumetric content of 15% to 20%.

3. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 1, characterized in that, The threshold segmentation process is a binarization process.

4. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 1, characterized in that, The foreground marker extraction based on the minimum gap depth threshold, and the watershed algorithm segmenting the skeleton gap region into multiple independent gap units, include: A distance transformation is performed on the skeleton gap region to generate a distance map; the pixel value of each pixel in the distance map is used to represent the shortest distance from that pixel to the aggregate boundary; local maxima in the distance map are determined by pixel neighborhood comparison; Set a minimum gap depth threshold H, where the minimum gap depth threshold H is a physical length value used to characterize the radial depth from the center of the gap unit to the edge of the gap unit; Based on the spatial resolution of the tomographic scan image, the minimum gap depth threshold H is converted into a pixel unit threshold. Extract local maxima points in the distance map whose pixel values ​​are greater than or equal to the unit threshold of that pixel, and use them as foreground marker points; Extract background marker points for the aggregate area, the background marker points being used to identify the aggregate area; Modifying the distance map includes: setting the pixel value at the location of the foreground marker point to a local minimum, and eliminating all local minimums at locations other than the foreground marker point based on the locations of the aggregate boundary and the background marker points; A watershed transformation is performed on the modified distance map, with each foreground marker point serving as an injection point, spreading outwards until the spread areas of different injection points intersect. A watershed line is constructed at the intersection, and the watershed line is the edge of each gap unit. The connected skeleton gap region is divided into several independent gap units through the watershed line.

5. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 4, characterized in that, The step of performing distance transformation on the skeletal gap region to generate a distance map includes: Obtain the threshold segmentation image corresponding to the skeleton gap region; For each pixel in the threshold segmentation image, traverse the aggregate boundaries surrounding the skeleton gap region, calculate the distance from the pixel to each aggregate boundary, and take the minimum value among them; The calculated minimum value is assigned to the corresponding pixel to generate a distance map.

6. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 4, characterized in that, Based on the minimum length L of the steel fiber min Set the minimum gap depth threshold H, where H ≥ L min .

7. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 1, characterized in that, The frequency distribution of the gap dimensions of all gap elements is statistically analyzed, including: The gap size is divided into multiple continuous size intervals. The number or proportion of gap units in each size interval is counted, and a frequency histogram is constructed.

8. The steel fiber gradation method based on reverse adaptation of skeleton gaps according to claim 7, characterized in that, The determination of the mass ratio of steel fibers of different lengths based on the frequency distribution of the gap size includes: Based on the frequency distribution of gap dimensions, a correspondence between size ranges and steel fiber lengths is established, so that each gap unit is classified into the corresponding steel fiber length according to its size. The number of interstitial units corresponding to each length of steel fiber was counted to determine the volumetric dosing of each length of steel fiber, and the corresponding dosing mass was calculated based on the density of each length of steel fiber.

9. A system for performing the steel fiber gradation method based on reverse adaptation of skeleton gaps as described in any one of claims 1-8, characterized in that, include: The image scanning module is used to acquire tomographic images of the skeleton reference specimen to be analyzed. The image processing module is used to perform threshold segmentation on the tomographic scan image to obtain mutually distinguishable aggregate regions and skeleton gap regions, and the interface between the aggregate regions and the skeleton gap regions is defined as the aggregate boundary. The gap segmentation module uses a marker-controlled watershed algorithm based on depth threshold constraints to segment the skeleton gap region into multiple independent gap units; The statistical analysis module is used to draw an equivalent ellipse for each gap unit, and use the length of the minor axis of the equivalent ellipse as the gap size of the gap unit, and to statistically analyze the frequency distribution of the gap size of all gap units. The gradation design module is used to determine the mass ratio of steel fibers of different lengths based on the frequency distribution of the gap size, so that the length gradation of the steel fibers matches the size distribution of the skeleton gap.

10. The system according to claim 9, characterized in that, The gap segmentation module includes: A distance transformation unit is used to perform distance transformation on the skeleton gap region to generate a distance map; the pixel value of each pixel in the distance map is used to represent the shortest distance from the pixel to the aggregate boundary; local maxima points in the distance map are determined by pixel neighborhood comparison; The marker extraction unit is used to set the minimum gap depth threshold H and convert the minimum gap depth threshold H into a pixel unit threshold according to the spatial resolution of the tomographic scan image; extract local maxima points in the distance map whose pixel values ​​are greater than or equal to the pixel unit threshold as foreground marker points; and extract background marker points in the aggregate area to identify the aggregate area. The terrain modification unit is used to modify the distance map, including: setting the pixel value at the position of the foreground marker point to a local minimum, and eliminating all local minimums at non-foreground marker point positions according to the positions of the aggregate boundary and the background marker point; The watershed processing unit is used to perform watershed transformation on the modified distance map, spreading outwards from each foreground marker point as an injection point until the spread areas of different injection points intersect each other, and constructing a watershed line at the intersection. The watershed line is the edge of each gap unit; the connected skeleton gap region is divided into several independent gap units through the watershed line.

Images