Three-dimensional morphology reconstruction and adhesion performance evaluation method of inert steel slag aggregate

Abstract

The application discloses a kind of inert steel slag aggregate three-dimensional morphology reconstruction and adhesion performance evaluation method, comprising the following steps;Step (1): inert pretreatment is carried out to steel slag;Step (2): 3D modeling is used to collect aggregate three-dimensional morphology parameters, and morphology parameter quantization is realized by MATLAB platform;The aggregate is inert steel slag aggregate;Step (3): using Pearson correlation analysis method, the correlation of quantized morphology parameters is analyzed, and the influence of parameter correlation on the adhesion performance of aggregate and asphalt is explored;Step (4): based on PFC3D platform, aggregate-asphalt virtual pull-out test is carried out, and test data is collected;Step (5): using grey correlation model, the influence of morphology parameters on adhesion performance is analyzed, and sensitivity sorting is completed.The application effectively improves the adhesion performance of steel slag-asphalt interface, and promotes the transformation of steel slag from alternative aggregate to graded and designed high-performance aggregate.

Description

Technical Field

[0001] This invention belongs to the field of three-dimensional morphology control technology of inert steel slag, specifically relating to a method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate. Background Technology

[0002] Existing steel slag treatment processes mainly focus on scavenging f-CaO and f-MgO or blocking moisture contact, which have limitations such as long treatment cycles, high costs, and environmental pollution. This invention... Obtaining and quantifying the three-dimensional morphology of steel slag aggregate is crucial for studying the influence of aggregate morphology on asphalt adhesion properties. Currently, various methods and indices for measuring aggregate morphology exist both domestically and internationally. Wang et al. observed the morphology of steel slag aggregate using a 3D laser scanning confocal microscope and evaluated the aggregate based on three parameters: shape factor, waviness, and roughness. This method has high resolution and can acquire submicron-level morphological features, but the equipment cost is high, the sample needs to be flattened, and the test area is limited. Kong Dezhi used an Artificial Image Measurement System (AIMS) to measure steel slag aggregate, taking pictures of the static aggregate from top to bottom, and calculating the sphericity, angularity index, and surface texture index from an empirical model. This method can achieve automated detection of aggregate and has a high degree of standardization of indicators, but it cannot acquire multi-faceted images of the aggregate, and the description of the overall particle morphology is not complete. Li Luoke et al. used a non-contact close-range white light scanner to acquire the feature point array on the aggregate surface. They characterized the three-dimensional morphological features of aggregate particles at different levels using six feature parameters (sphericity, shape factor, ellipsoidity, maximum surface curvature, two-dimensional roughness index, and three-dimensional roughness coefficient). This method can acquire high-density point cloud data, realize particle geometry reconstruction, and has a comprehensive morphological index system; however, the point cloud processing algorithm is relatively complex.

[0003] In summary, the main technical shortcomings of existing methods for acquiring and quantifying the three-dimensional morphology of steel slag aggregates are that while the indicators are highly standardized, they cannot obtain multi-faceted images of the aggregates, resulting in an incomplete description of the overall particle morphology. The point cloud processing algorithms are also relatively complex, and CT equipment is expensive, with long scanning and calculation times. Summary of the Invention

[0004] In order to overcome the above technical problems, the purpose of this invention is to provide a method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate. This method effectively improves the adhesion performance of steel slag-asphalt interface and promotes the transformation of steel slag from a substitute aggregate to a gradeable and designable high-performance aggregate.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate, comprising the following steps; Step (1): Inertization pretreatment of steel slag; Step (2): Use a 3D modeling and digitization machine to collect the three-dimensional morphological parameters of the aggregate, and quantify the morphological parameters through the MATLAB platform; The aggregates include inertized steel slag aggregates and two types of limestone and diabase aggregates used for comparison. Step (3): Pearson correlation analysis was used to analyze the correlation of quantified morphological parameters and explore the influence of parameter correlation on the adhesion performance between aggregate and asphalt. Step (4): Conduct a virtual pull-out test of aggregate-asphalt based on the PFC3D platform and collect test data; Step (5): Use the grey relational model to analyze the influence of morphology parameters on adhesion performance and complete the sensitivity ranking.

[0006] In step (1), the specific steps of the inertization pretreatment are as follows: mechanically crushing the inert steel slag; crushing the aggregate to a particle size range of 13.2-19mm; magnetically separating the crushed inert steel slag; and obtaining aggregate with excellent stability and high-value-added mineral admixtures.

[0007] Step (2) specifically involves: using an Einstar handheld 3D modeling and digitization machine to acquire surface point cloud data of the aggregate, importing the data into the MATLAB platform for analysis and calculation, establishing an index system covering macroscopic shape, mesoscopic angularity, and microscopic texture, and obtaining three-dimensional morphological parameters including volume, surface area, equivalent sphere diameter, sphericity, elongation, flatness, angularity, convexity, and surface roughness to quantitatively characterize the morphological features of the aggregate particles. In step (2), the specific steps for quantifying the morphological parameters using the MATLAB platform are as follows: The parameters of the aggregate particles were obtained by using an algorithm built on the MATLAB platform based on the 3D point cloud and mesh data exported from the Einstar scanner. volume The volume of a particle is calculated by using the volume integration method of closed triangular facets to represent the size of the space occupied by the particle in three-dimensional space. The particle volume is obtained by dividing the particle surface into many triangular facets, each of which forms a tetrahedron with the origin of the coordinate system. The volume of the particle is obtained by summing the volumes of all tetrahedrons. The coordinates of the vertices of the triangular facets on the particle surface are given by equation (1): in, : No. The three vertices of a triangular facet, in mm; For the first The coordinates of the vertices of the triangular facet in three-dimensional space, in mm; Let the first The coordinates of the three vertices of the triangular facet are given in equations (2) to (4): in, : No. The three vertices of a triangular facet, in mm; The volume calculation formula is as follows (5): in, Particle volume, in mm 3 ; : Number of triangular facets, in units of one; : Vector magnitude; Vector cross product; vector dot product; Surface area This represents the area of ​​the particle surface. The particle surface area is obtained by summing the areas of the triangular faces on the particle surface. The formula for calculating the area of ​​a triangular facet is shown in equation (6): in, : No. Area of ​​each triangular facet, in mm² 2 ; The formula for calculating the total surface area of ​​particles is Equation (7): in, : Particle surface area, in mm 2 ; Equivalent sphere diameter This refers to the diameter of a sphere with the same volume as the particle. This invention is based on particle volume. The diameter of the equivalent sphere of the particle is calculated by reverse calculation; The volume calculation formula is as shown in equation (8): in, : Equivalent sphere diameter, in mm; Pi (π) is a dimensionless mathematical constant. sphericity To characterize the degree to which the particle shape approximates a sphere, sphericity is calculated by dividing the surface area of ​​an equal-volume sphere by the surface area of ​​the particle. The formula for calculating the surface area of ​​a sphere with the same volume as the particle is as shown in equation (9): in, Surface area of ​​a sphere of equal volume, in mm² 2 ; The formula for calculating sphericity is as shown in equation (10): in, : Sphericity, dimensionless; : Particle surface area, in mm 2 ; Substituting the formula for calculating the surface area of ​​a sphere of equal volume, the final formula for calculating sphericity is as shown in equation (11): Extensibility The elongation is calculated by the ratio of the longest direction to the shortest direction of the particle. The value is ≥1. The larger the value, the finer and longer the particle; the closer the value is to 1, the more equiaxed the particle. First, extract the dimensions of the bounding box of the particles. The formulas for calculating the dimensions of the bounding box in the three directions are as shown in equations (12) to (14): in, : x-axis length, in mm; : y-axis length, in mm; : Z-axis length, in mm; Sort the three dimensions from largest to smallest to obtain the major axis, median axis, and minor axis: in, : Longest shaft dimension, in mm; Intermediate shaft dimensions, in mm; : Shortest shaft dimension, in mm; Extensibility The calculation formula is shown in equation (16).

[0008] in, : Extension or aspect ratio, dimensionless; Flatness Flatness is used to describe the degree of flaky texture of particles, calculated by the ratio of the shortest axis dimension to the longest axis dimension. The formula for calculating flatness is as shown in equation (17): in: Flatness, dimensionless; : Shortest shaft dimension, in mm; : Longest shaft dimension, in mm; angularity It is used to describe the severity of the sharpness of the particle surface edges. The particle edge index is characterized by calculating the average angle between adjacent triangular facets. The final edge index is obtained by normalization. The formula for calculating the included angle between adjacent triangular facets is shown in equation (18): in, : Angle between the normals of adjacent facets, in rad; : No. The unit normal vector of a triangular facet is dimensionless. : the adjacent first The unit normal vector of a triangular facet is dimensionless. The formula for calculating the overall angularity of particles is as shown in equation (19): in: Angularity, dimensionless; Pi, here representing the maximum possible normal angle (180°); : The number of adjacent triangular facet pairs, in units of ; : No. The angle between the normals of adjacent facets, in rad; convexity The index used to describe the degree of concavity of the particle profile is the convexity, which is calculated by the ratio of the particle surface area to the convex hull area. The particle convex hull is the smallest convex polyhedron that can enclose all points of the particle. The line connecting any two points on the convex hull surface falls inside or on the surface of the polyhedron. The convex hull area is obtained by the triangular facet method. The calculation formula is referred to from formula (6) to formula (7). convexity The calculation formula is as shown in equation (20): in, : Convexity, dimensionless; : Area of ​​particle convex hull, in mm 2 A: Particle surface area, in mm² 2 ; Surface roughness To characterize the dispersion of particle point clouds in three-dimensional space, surface roughness is calculated using the ratio of principal component eigenvalues ​​(PCA). Principal component analysis is performed on the point cloud coordinates to obtain eigenvalues ​​along the three principal component directions. Each eigenvalue corresponds to the dispersion of the data along that axis. The following conditions should be met: in, : Eigenvalues ​​of the first principal component; : Eigenvalues ​​of the second principal component; : Eigenvalues ​​of the third principal component.

[0009] The ratio of the third eigenvalue to the first eigenvalue is used as the surface roughness index; in, Surface roughness index, dimensionless; : Eigenvalues ​​of the first principal component, in mm 2 ; : Eigenvalues ​​of the third principal component, in mm 2 .

[0010] The specific steps (3) are: Pearson correlation analysis based on the original measured data (screening the main control morphology parameters); Nine three-dimensional morphological parameters were collected from at least 30 sets of inert steel slag, limestone, and diabase aggregates as samples. The nine three-dimensional morphological parameters are volume V, surface area A, and equivalent sphere diameter. Sphericity ψ, elongation AR, flatness F, angularity AI, convexity C, surface roughness ; The Pearson correlation coefficient was used to analyze the linear correlation between the parameters. Redundant variables were removed. The formula for calculating the Pearson correlation coefficient is as follows: in, , These are sample values ​​for two morphological parameters, , The sample mean. The value range of is [-1, 1]; The criteria for judgment are: Highly correlated The correlation is moderate. The analysis results show that there is a highly significant positive correlation between volume, surface area, and equivalent sphere diameter. ), elongation and flatness are significantly negatively correlated ( The correlation between the remaining parameters is relatively weak; From the nine parameters, the following were selected: surface area (A), sphericity (ψ), flatness (F), angularity (AI), convexity (C), and surface roughness. A total of 6 features were selected as the main morphological features for subsequent geometric reconstruction and simulation experiments.

[0011] This step involves correlation analysis based on the original measured morphology data. The purpose is to simplify the parameter system and avoid variable redundancy in subsequent reconstruction and simulation.

[0012] Based on parameter-controlled particle geometry reconstruction, virtual steel slag particles are generated; based on the selected six main control parameters and their value ranges, an algorithm is constructed in the MATLAB platform to regenerate virtual inert steel slag particles with specific target parameters.

[0013] In step (4): virtual particle geometry reconstruction based on core morphology parameter control is performed through the MATLAB platform. The virtual steel slag particles generated by the reconstruction are exported as STL format files and imported into the PFC3D platform. An inert steel slag-asphalt pull-out test model is constructed based on PFC3D and a virtual pull-out test is carried out. Adhesion strength data of the steel slag-asphalt interface under different morphology parameter conditions are collected.

[0014] The specific method is as follows: Based on the measured data of 30 sets of inert steel slag, the minimum, mean and maximum values ​​of each main control parameter were determined, forming the parameter control range; The method employs a triangular patch method and a parameter-driven deformation algorithm: First, a basic spherical or ellipsoidal mesh is generated. Then, it is scaled along the principal axis according to the target flatness, the surface normal is perturbed according to the target angularity, and Gaussian noise or fractal texture is added according to the target surface roughness. Finally, a closed triangular mesh model is generated. The algorithm ensures that the generated particles are statistically consistent with the distribution of measured parameters. Thirteen groups of inert steel slag particles with different morphologies were reconstructed. Each group corresponds to a specific set of values ​​for surface area, sphericity, flatness, angularity, convexity, and surface roughness. These values ​​were exported as STL files for subsequent simulations.

[0015] The 13 groups here refer to the maxima of the 6 morphological parameters (with the mean of the other parameters), the minima of the 6 morphological parameters (with the mean of the other parameters), and the mean of all parameters, totaling 13 groups.

[0016] This step involves geometric reconstruction independent of the original particles. Based on selected parameters and their range of variation, virtual particles with specific morphologies are actively generated, rather than simply replicating the original particles. The advantage of this approach is that it allows for independent control of individual parameter variations, thereby accurately identifying the contribution of each parameter to adhesion strength in the simulation.

[0017] Step (4) specifically involves: a pull-out simulation test of inert steel slag-asphalt based on PFC3D. Import the generated reconstructed particle STL file into PFC3D (3D particle flow program), and use clump to create rigid particles to preserve the complete morphology; A pull-out model was constructed, including a bottom substrate, an asphalt film, a pull-out head, and a pull rod. The asphalt was represented by spherical particles, and a linear parallel bond model was used to describe the bonding behavior within the asphalt and at the steel slag-asphalt interface. Model parameters were inverted and calibrated based on macroscopic test results of 70# base asphalt (e.g., parallel bond modulus). Pa, tensile strength Pa, etc.); The "asphalt" here refers to the "asphalt film" that coats the surface of the steel slag particles.

[0018] A vertical, uniform pulling displacement (the rate was determined by trial and error) was applied to each reconstructed particle. The force-displacement curves during the pulling process were recorded, and the adhesion strength was calculated using the following formula: in, The tensile strength is expressed in MPa. This represents the maximum pulling force (N) that occurs during the pulling process. The effective adhesion area (mm²) between asphalt and steel slag particles is calculated as the product of the reconstructed particle surface area and the asphalt contact area coefficient. Each test was repeated three times, and the average value was taken as the adhesion strength under that combination of morphological parameters.

[0019] This step utilizes the generated reconstructed particles to perform pull-out simulations, obtaining a quantitative adhesive strength.

[0020] The sensitivity ranking in step (5) based on the grey relational model is as follows: Using 13 sets of adhesion strength as a reference sequence and 6 morphological parameters as a comparison sequence, grey relational analysis was employed to quantitatively evaluate the influence of each morphological parameter on adhesion strength. The grey relational analysis formula is as follows: in, This is the reference sequence (adhesion strength) for the k-th sample. For the comparison sequence of the i-th morphology parameter, The resolution coefficient (0.5 in this invention) is used. The correlation coefficient, This represents the grey relational degree. The higher the relational degree, the more significant the influence of the morphological parameter on the adhesion strength. According to the calculation results, the sensitivity of each morphological parameter to adhesion strength is ranked as follows: flatness (0.637) > angularity (0.614) > surface roughness (0.594) > surface area (0.576) > sphericity (0.526) > convexity (0.510). This ranking quantitatively reveals that flatness is the most critical morphological index affecting the adhesion performance of the steel slag-asphalt interface.

[0021] This step uses a grey relational model to calculate the correlation between each morphological parameter and the adhesion strength, and finally obtains the sensitivity ranking.

[0022] A three-dimensional morphology reconstruction and shaping system for inert steel slag is divided into two main modules: aggregate morphology scanning and morphology parameter extraction. It relies on the Einstar handheld 3D modeling and digitizing machine and MATLAB algorithm to achieve three-dimensional point cloud acquisition of aggregates, model reconstruction and multi-dimensional morphology parameter quantification. The accuracy of the whole system can match the CT scan results and meet the requirements of experimental analysis. A method for reconstructing the three-dimensional morphology of inert steel slag In the aggregate morphology scanning module, the inert steel slag aggregate is first pretreated, including ultrasonic cleaning and high-temperature drying. Then, titanium dioxide is thinly coated to give the steel slag surface a matte finish in order to eliminate imaging interference. Then, the accompanying EXstar software was launched, and distance and whiteness calibration were completed using a dedicated calibration board. The object scanning mode was selected, and the exposure value was adjusted to 40-60 according to the brightness of the aggregate, with the resolution set to 0.1mm to preserve microscopic textures. During scanning, the aggregate was placed in the center of the scanning stage, and the scanner moved slowly within a range of 30-40cm to collect point cloud data of the particle surface from all directions. The scanning time for a single particle was 30-60 minutes, and missing areas were scanned to ensure no blind spots. After scanning, the original point cloud data was corrected, and impurities and noise points were removed. The model was encapsulated, and ASC and STL format files were exported to provide the original 3D model for subsequent parameter extraction. In the topography parameter extraction module, a dedicated algorithm based on the MATLAB platform is used to perform calculations on the point cloud and mesh data obtained from the scan. First, import the point cloud file. The point cloud file refers to the ASC format file of the three-dimensional shape of the aggregate obtained by the handheld 3D modeling and copying machine. The system automatically removes NaN values, infinity values, and outliers to complete data cleaning. Then, it calculates nine morphological parameters across three dimensions: macroscopic shape (volume, surface area, equivalent sphere diameter, sphericity, elongation, flatness), mesoscopic angularity (angularity, convexity), and microscopic texture (surface roughness). After batch processing 30 sets of aggregate sample data, a Pearson correlation analysis algorithm is used to identify the coupling relationships between parameters, remove highly correlated redundant indicators, and ultimately select six core control parameters: surface area, sphericity, flatness, angularity, convexity, and surface roughness. The system also automatically generates aggregate point cloud maps, bounding box models, parameter radar charts, and size scale diagrams, and outputs a complete analysis report, providing quantitative basis for subsequent particle geometry reconstruction and simulated pull-out tests.

[0023] The beneficial effects of this invention are: This invention utilizes an Einstar handheld 3D modeling and digitization machine to collect point cloud data of aggregate surfaces. Combined with an algorithm built on the MATLAB platform, it can completely reconstruct the macroscopic contour, mesoscopic angularity, and microscopic texture features of particles. It also establishes a quantitative index system covering nine parameters, including volume, surface area, sphericity, flatness, angularity, convexity, and surface roughness, overcoming the shortcomings of traditional methods such as incomplete description of the overall particle morphology, expensive equipment, or complex algorithms. Secondly, through Pearson correlation analysis, six core control parameters—surface area, sphericity, flatness, angularity, convexity, and surface roughness—are selected from the nine parameters, effectively eliminating redundant variables such as volume, equivalent sphere diameter, and elongation. Furthermore, it enables controllable adjustment and geometric reconstruction of individual morphological parameters, overcoming the technical challenge of parameter interaction interference in traditional analysis. Thirdly, relying on PFC3D particle flow software, a single-particle steel slag-asphalt pull-out simulation model is constructed. Combined with grey relational analysis, the influence weight of each morphological parameter on adhesion performance is quantitatively evaluated, clarifying flatness, angularity, and surface roughness as key control factors. This provides clear morphology control indicators and process optimization suggestions for the processing and shaping of steel slag aggregates, thereby effectively improving the adhesion performance of the steel slag-asphalt interface and promoting the transformation of steel slag from a substitute aggregate to a gradeable and designable high-performance aggregate.

[0024] This invention uses the three-dimensional morphology parameters of inert steel slag aggregate to determine the main controlling morphology features through Pearson correlation analysis. By adjusting the relevant three-dimensional morphology parameters, the geometry of the particles is reconstructed on the MATLAB platform, and the reconstructed particles are imported into PFC3D software to construct an inert steel slag-asphalt pull-out test model. The adhesion strength is used as an index to analyze the influence of different morphological features of inert steel slag on the adhesion performance of asphalt. Subsequently, a grey relational model is used to rank the sensitivity of the influence of morphology parameters on adhesion performance. Attached image description. Figure 1 Images of inert steel slag, limestone, and diabase particles.

[0026] Figure 2 This is a process for extracting the three-dimensional morphology of aggregates.

[0027] Figure 3 This is a comparison image of inert steel slag particles obtained by an Einstar image scanner and the actual particles.

[0028] Figure 4 The image shows the actual particle morphology of inert steel slag particles, compared with Einstar digitizer imaging and CT scan imaging.

[0029] Figure 5 Reconstruct particle images for inert steel slag.

[0030] Figure 6 Inert steel slag-asphalt pull-out test model.

[0031] Figure 7 Results of the three-dimensional morphological indices of inert steel slag and its adhesion strength to asphalt. Detailed Implementation

[0032] The present invention will now be described in further detail with reference to the accompanying drawings.

[0033] A method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate, comprising the following steps; Step (1): Inertization pretreatment of steel slag; Step (2): Use a 3D modeling and digitization machine to collect the three-dimensional morphological parameters of the aggregate, and quantify the morphological parameters through the MATLAB platform; The aggregates include inertized steel slag aggregates and two types of limestone and diabase aggregates used for comparison. Step (3): Pearson correlation analysis was used to analyze the correlation of quantified morphological parameters and explore the influence of parameter correlation on the adhesion performance between aggregate and asphalt. Step (4): Conduct a virtual pull-out test of aggregate-asphalt based on the PFC3D platform and collect test data; Step (5): Use the grey relational model to analyze the influence of morphology parameters on adhesion performance and complete the sensitivity ranking.

[0034] Step (1) of this invention involves the inertization treatment of steel slag, resulting in an uncontrollable three-dimensional morphology of the product. The purpose of this invention is to study the influence of its morphology parameters on adhesion performance.

[0035] Extraction of three-dimensional morphology parameters of aggregates: The study used an Einstar handheld 3D modeling and digitization machine to acquire 30 sets of surface point cloud data of inert steel slag, limestone, and diabase aggregates. Then, an algorithm was constructed using the MATLAB platform to extract the three-dimensional morphology parameters of the aggregates.

[0036] Einstar Handheld 3D Modeling Reverse Engineering Machine (hereinafter referred to as Einstar Reverse Engineering Machine) Equipment Introduction: The Einstar Reverse Engineering Machine adopts the infrared structured light scanning method. The built-in VCSEL (Vertical-Cavity Surface-Emitting Laser) projects a regular infrared grating onto the surface of the object being measured. Two high-resolution cameras collect images of the grating deformation after surface reflection. The three-dimensional coordinates of each point on the object surface are reconstructed using a triangulation algorithm, and finally a high-density point cloud model is generated.

[0037] The scanning process is as follows: (1) Aggregate pretreatment: The aggregate is cleaned and dried. Since the low brightness of the steel slag surface and local metal reflection will affect the imaging effect, the inert steel slag aggregate needs to be matte treated separately (coated with titanium dioxide) before scanning. Place the pretreated aggregate in the center area of ​​the scanning stage, connect the scanner, and open the EXstar software; (2) Equipment calibration: Before starting scanning, use the calibration board provided with the equipment to perform calibration, calibrate the distance and whiteness, and reduce recognition errors; (3) Parameter settings: Use the "object" scanning mode, turn on texture scanning, adjust the exposure value to 40-60 for darker inert steel slag and diabase aggregate, and adjust the exposure value to 30-50 for brighter limestone aggregate. Set the resolution to 0.1 to preserve the morphology and texture as much as possible. (4) Morphology scanning: Press the scan start button to start scanning, and double-click to stop scanning. Keep the scanner moving slowly within a radius of 30-40cm from the aggregate particles, observe the display area, and ensure that the aggregate is captured from all angles. The scanning time for a single particle is 30-60 minutes. Keep the scanner stable during the scanning process and avoid image disturbance caused by moving too fast or shaking violently, which may lead to splicing failure. (5) Model export: After the scan is completed, observe the particle reconstruction in the display interface, scan the missing data areas, obtain the complete particle surface point cloud, click model encapsulation, and export the data.

[0038] This invention utilizes an Einstar handheld 3D modeling scanner to acquire the three-dimensional morphology of aggregates. Based on the principle of infrared structured light, this instrument projects a regular infrared grating through a VCSEL, acquires deformed grating images via dual cameras, and reconstructs the three-dimensional coordinates of the particle surface using a triangulation algorithm to generate a high-density point cloud model. Before scanning, the aggregates require pretreatment: first, they are cleaned three times (5 minutes each time) with an ultrasonic cleaner, then dried at 105±5℃ to constant weight and cooled to room temperature. To address the issue of metallic reflection and dim brightness on the surface of inert steel slag, an additional matte treatment is required, i.e., coating a thin layer of titanium dioxide with a particle size ≤1μm (thickness ≤0.1mm). After drying, excess powder is gently brushed away with a soft brush to avoid obscuring the original morphology. The pretreated individual aggregate particles are placed in the center of a white scanning stage, and if necessary, a very small area at the bottom (≤1mm²) is fixed with low-viscosity double-sided tape.

[0039] Hardware Connection and Software Preparation: Connect the scanner to the high-performance computer via USB 3.0 interface, launch the EXstar software and create a new project, naming it "Aggregate Type-Particle Size-Date", and set the project storage path to the solid-state drive. Equipment Calibration: If the scanner has not been used for a long time (>7 days), it must be recalibrated. The environment should avoid direct sunlight and electromagnetic interference. Place the calibration plate horizontally in the center of the scanning stage, and under the software guidance, complete the initial alignment and multi-position acquisition (at least 5 sets of data in each of the horizontal, 45°, and vertical directions) until the software displays that the calibration is qualified. Set the scanning parameters according to the aggregate type: Use "Object Mode" for inert steel slag and diabase, enable texture scanning, and set the exposure value to 40~60; for limestone, set the exposure value to 30~50; the resolution should be uniformly set to 0.1mm, and the "Real-time Monitoring" function should be enabled.

[0040] Scanning Operation: Click "Start Scanning," place the scanner 30-40cm away from the particle, maintaining an angle of 45°±15° with the particle surface. Acquire the overall contour from four directions: 0°, 90°, 180°, and 270°. For complex areas such as edges and depressions, reduce the distance to 20-30cm and decrease the moving speed (≤1cm / s), repeating the scan 2-3 times. Observe the software imaging in real time, promptly scanning missing areas. If data stitching fails, restart the software and rescan. Stop scanning when there are no missing areas on the particle surface and the contour and texture are complete. Finally, encapsulate the model (0.1mm accuracy, texture blending), export an STL file, and name it consistent with the project. Pay attention to controlling the titanium dioxide coating thickness; clean and verify after scanning; maintain a stable environment, avoiding movement or vibration; back up the point cloud (.ply) and mesh (.stl) data after completing each particle, and establish a log recording the number, preprocessing method, parameters, time, and operator to ensure data traceability.

[0041] Simulated pull-out test: This study uses the PFC3D (particle flow code in 3 dimensions) three-dimensional particle flow program to construct a single inert steel slag-asphalt film pull-out model, and obtains the influence of the three-dimensional morphology characteristics of the steel slag on its adhesion performance with asphalt. The core of PFC3D is to simulate the mechanical behavior between particles using the discrete element method. The discrete element method is a numerical simulation method based on discrete, independently moving particles, and is widely used to analyze the mechanical properties of asphalt mixtures.

[0042] For the aggregate-asphalt virtual pull-out test based on the PFC3D platform in step (3), the construction process of the inert steel slag-asphalt pull-out model is as follows: (1) Model building: The model consists of a base plate, inert steel slag aggregate, asphalt film particles, a cylindrical pull-out head, and a tie rod. The inert steel slag aggregate is imported as an STL file and created using the clump method to preserve particle morphology. Asphalt is represented by spherical particles. An asphalt matrix region is generated above the inert steel slag, and asphalt particles are allowed to fall naturally into contact with the inert steel slag particles. Finally, only the asphalt particles in contact with and adjacent to the steel slag surface are retained, forming a double-layer asphalt film (0.72 mm thick) covering the steel slag surface.

[0043] (2) Contact establishment: A linear parallel bond model is used to describe the bonding behavior within the asphalt film and at the steel slag-asphalt interface. This model can transmit normal, tangential, and moment forces at particle contact points; bond failure occurs when the tensile or shear stress at the contact point exceeds a predetermined strength. Based on the fundamental properties of inert steel slag and 70# base asphalt, appropriate density, stiffness, bond strength, and friction parameters are assigned to the inert steel slag particles, asphalt particles, and the interface.

[0044] The model parameters were calibrated using a macroscopic experimental inversion method. First, the basic hardness of the asphalt material was determined based on conventional performance indicators such as penetration, ductility, and softening point of 70# base asphalt. Using the results of indoor steel slag-asphalt pull-out tests as a target, the microscopic contact parameters were adjusted to ensure that the numerical simulation results were basically consistent with the experimental results. The main microscopic parameters determined at the end are as follows: the parallel bond modulus within the asphalt film is... Pa, tensile strength is Pa, cohesion is Pa, the parallel bond modulus of the steel slag-asphalt interface is Pa, interfacial tensile strength is Pa, cohesion is Pa.

[0045] (3) Rod generation and loading: A loading rod with a diameter of 10 mm was set up to apply a vertical uniform pulling displacement to the steel slag particles, and the relationship between the pulling strength and the displacement during the loading process was recorded.

[0046] The formula for calculating tensile strength is: in, Pull-out strength, in MPa; : The maximum pulling force that occurs during the drawing process, in N; Effective adhesion area between asphalt and inert steel slag particles, in mm².

[0047] Asphalt mixture preparation and determination of optimal asphalt-aggregate ratio: This invention conducts tests on AC-16 dense-graded asphalt mixtures according to the "Test Procedures for Asphalt and Asphalt Mixtures in Highway Engineering" (JTG E20-2011) and the "Technical Specifications for Construction of Asphalt Pavement on Highways" (JTG F40-2004). Specimens were prepared using inert steel slag, limestone, and diabase aggregates mixed with SBS (IC type) modified asphalt, and the optimal asphalt-aggregate ratio was determined. The aggregates were dried at 160-170℃ for 4 hours, and the asphalt was heated to 170-180℃. The aggregates were dry-mixed for 10 seconds according to the design gradation, followed by wet mixing with the corresponding asphalt-aggregate ratio for 90 seconds. Mineral powder was then added, and mixing continued for another 90 seconds. The mixture was placed in Marshall molds preheated to 145℃, compacted 75 times on each side, and demolded after cooling at room temperature for 12 hours to obtain standard specimens. Multiple sets of specimens were prepared using gradient asphalt-aggregate ratios, and the bulk density, stability, flow value, porosity, aggregate void ratio, and asphalt saturation were tested. The volumetric indices of the mixture are calculated using formulas, and the relationship curves between the indices and the asphalt-aggregate ratio are plotted. The characteristic asphalt-aggregate ratios corresponding to density, stability, porosity, and asphalt saturation are determined respectively. The final optimal asphalt-aggregate ratio is obtained by combining the OAC1 and OAC2 calculation formulas, and all indices are verified to meet the specifications.

[0048] Quantitative characterization and geometric reconstruction of three-dimensional morphology of aggregates: Three-dimensional morphology extraction of aggregates: This invention uses an Einstar reverse engineering machine to acquire 30 sets of morphological data for inert steel slag, limestone, and diabase aggregates with particle sizes ranging from 13.2 to 19 mm. The particle sizes of the inert steel slag, limestone, and diabase are shown in the figure. Figure 1 .

[0049] The process for extracting the three-dimensional morphology of aggregates is as follows: Figure 2 .

[0050] Comparison of Einstar digitization machine image and actual particle morphology of inert steel slag particles (see) Figure 3 .

[0051] Some studies have used CT scans to obtain aggregate morphology. This invention compares and analyzes the actual morphology of inert steel slag particles, 3D imaging by an Einstar digitizer, and CT scan imaging. (See aggregate imaging for details.) Figure 4 .

[0052] in accordance with Figure 4 Comparing the 3D imaging particles from the reverse engineering machine with the real particle morphology reveals that: the overall shape and size ratio of the reconstructed particles are completely consistent with the real particles, with no scaling or deformation; the rough texture and raised details on the particle surface are completely preserved, but small cracks are covered, and the texture features are completely preserved; the irregular shape (edges, curvature) of the particle edges is highly consistent with the real morphology, and the particle morphology is relatively completely restored.

[0053] Comparing the 3D imaging particles from the reverse engineering machine with those from CT scans revealed that: the overall outline and 3D morphology of the reconstructed particles completely matched the CT scan results with no morphological deviation; the surface texture and other details of the particles were highly consistent with the CT scan results, with high feature retention; and the irregular shape of the particle edges matched the real appearance without shape distortion.

[0054] In summary, the imaging results of the digitizer are highly consistent with the morphology of real particles and CT scans, and can completely preserve the morphological characteristics of inert steel slag particles, thus meeting the requirements for experimental accuracy.

[0055] Three-dimensional morphology characterization of aggregates: This invention utilizes a parameter extraction algorithm built on the MATLAB platform to calculate nine morphological parameters covering three dimensions: macroscopic contour shape, mesoscopic edges and corners, and microscopic surface texture. The morphology of inert steel slag, limestone, and diabase was quantitatively characterized, and the morphological differences among the three aggregates were compared. This provides a quantitative basis and reliable reference for subsequent morphological reconstruction of inert steel slag aggregates. The aggregate morphology parameter extraction command is provided in Appendix A.

[0056] Analysis of the effect of steel slag inertization treatment: This invention uses mechanical crushing and magnetic separation to remove the silicate mineral phase from the surface of steel slag, obtaining inert steel slag rich in RO phase. XRD and XRF analysis results show that the diffraction peak intensities of dicalcium silicate (Ca2SiO4) and feldspar (Ca2MgSi2O7) in the RO-rich inert steel slag basically disappears, while the FeO (RO phase) diffraction peak remains. The Fe2O3 content increases from 23.9% to 31.6%, and the f-CaO content decreases from 0.38% to 0.27%. The results indicate that mechanical crushing + magnetic separation can significantly reduce the f-CaO content in steel slag and effectively separate the silicate mineral phase, improving the stability of the steel slag.

[0057] Based on the three-dimensional morphological data and Pearson analysis results of 30 groups of inert steel slag, limestone, and diabase aggregates, the following findings were made: On the one hand, compared with limestone and diabase aggregates, inert steel slag has lower elongation and higher sphericity, flatness, angularity, convexity, and surface roughness. The rich angularity and surface texture features give the three-dimensional morphology of inert steel slag good interfacial bonding potential. On the other hand, the three morphological parameters of volume, surface area, and equivalent sphere diameter showed a highly significant positive correlation, while the two morphological parameters of elongation and flatness showed a highly significant negative correlation. Considering the representativeness and correlation of each parameter, six parameters, namely surface area, sphericity, flatness, angularity, convexity, and surface roughness, were finally selected as morphological control variables for subsequent steel slag and asphalt pull-out tests.

[0058] This invention further conducts correlation analysis on the three-dimensional morphology parameters of the aggregate to eliminate redundant indicators and simplify the control variable system. Specifically, based on the three-dimensional point cloud data of inert steel slag aggregate obtained by an Einstar digitizer, nine morphology parameters—volume, surface area, equivalent sphere diameter, sphericity, elongation, flatness, angularity, convexity, and surface roughness—are calculated using the MATLAB platform. Pearson correlation coefficients are then used to perform pairwise correlation analysis on each parameter. The results show that volume, surface area, and equivalent sphere diameter are significantly positively correlated (correlation coefficient ≥ 0.99), elongation and flatness are significantly negatively correlated (correlation coefficient -0.975), and the correlations among the remaining parameters are weak, independently reflecting different morphological characteristics of the particles. Considering the representativeness, independence, and potential impact of the parameters on asphalt adhesion performance, this invention selects six parameters—surface area, sphericity, flatness, angularity, convexity, and surface roughness—from the above nine parameters as morphology control variables for subsequent steel slag-asphalt interface pull-out simulation tests. This screening method avoids the interference of variable redundancy on experimental analysis while retaining the ability of key morphological features to characterize adhesion performance.

[0059] Analysis of the three-dimensional morphology of steel slag and the results of asphalt pull-out tests revealed that inert steel slag particles with high flatness, prominent edges, and rougher surfaces are more conducive to improving the adhesion strength of the steel slag-asphalt interface. The sensitivity ranking of the three-dimensional morphology parameters of inert steel slag to asphalt adhesion performance is: flatness > angularity > surface roughness > surface area > sphericity > convexity. The processing technology for steel slag aggregates should prioritize the control of flatness, angularity, and surface roughness. During processing, flat particles should be reduced, moderate edges should be retained, surface roughness should be enhanced, and a reasonable surface area range should be controlled. A processing technology combining multi-stage crushing, impact shaping, screening and grading, and surface cleaning is beneficial for improving the morphological fullness and surface adhesion activity of steel slag particles, thereby improving the adhesion performance of the asphalt-steel slag interface.

[0060] This invention first obtains key morphological parameters of inert steel slag aggregate, such as surface area, sphericity, flatness, angularity, convexity, and surface roughness, through three-dimensional morphological reconstruction. A single-particle steel slag-asphalt interface pull-out simulation model is then established using PFC3D discrete element software, and a series of numerical tests are conducted with adhesion strength as the evaluation index. Subsequently, a grey relational analysis method is introduced to quantitatively calculate the correlation between each morphological parameter and adhesion strength, resulting in the following sensitivity ranking of parameters to interfacial adhesion performance: flatness > angularity > surface roughness > surface area > sphericity > convexity. This ranking reveals that flatness has the most significant impact on asphalt adhesion performance, followed by angularity and surface roughness, while sphericity and convexity have relatively weaker effects. Based on these sensitivity patterns, this invention further proposes a processing control method for steel slag aggregate morphology: priority should be given to controlling the three core indicators of flatness, angularity, and surface roughness. Specifically, during processing, the content of flat particles should be reduced, moderate angular characteristics of the aggregate should be retained, surface roughness texture should be enhanced, and the surface area range should be reasonably controlled.

[0061] Evaluation index selection and calculation methods: To systematically characterize the three-dimensional morphological features of aggregates, this invention selects nine morphological parameters covering three levels: macroscopic shape, mesoscopic edges and corners, and microscopic texture. The parameter indices are shown in Table 1.

[0062] Table 1. Characterization system of three-dimensional morphology parameters of aggregates The parameters of each particle were calculated using an algorithm built on the MATLAB platform based on the 3D point cloud and mesh data exported from the Einstar scanner.

[0063] volume This represents the size of the space occupied by a particle in three-dimensional space. This invention uses the closed triangular facet volume integration method to calculate the particle volume. The particle surface is divided into many triangular facets, each of which forms a tetrahedron with the origin. The particle volume is obtained by summing the volumes of all tetrahedrons.

[0064] The coordinates of the vertices of the triangular facets on the particle surface are given by equation (1): in, : No. The three vertices of a triangular facet, in mm; For the first The coordinates of the vertices of a triangular facet in three-dimensional space, in mm.

[0065] Let the first The coordinates of the three vertices of the triangular facet are given in equations (2) to (4): in, : No. The three vertices of a triangular facet, in mm.

[0066] The volume calculation formula is as follows (5): in, Particle volume, in mm 3 ; : Number of triangular facets, in units of one; : Vector magnitude; Vector cross product; vector dot product.

[0067] Surface area This represents the area of ​​the particle surface. The present invention obtains the particle surface area by summing the areas of the triangular faces on the particle surface.

[0068] The formula for calculating the area of ​​a triangular facet is shown in equation (6): in, : No. Area of ​​each triangular facet, in mm² 2 .

[0069] The formula for calculating the total surface area of ​​particles is Equation (7): in, : Particle surface area, in mm 2 .

[0070] (1) Equivalent sphere diameter Equivalent sphere diameter This refers to the diameter of a sphere with the same volume as the particle. This invention is based on particle volume. The diameter of the equivalent sphere of the particle is calculated by reverse calculation.

[0071] The volume calculation formula is as shown in equation (8): in, : Equivalent sphere diameter, in mm; Pi (π) is a dimensionless mathematical constant.

[0072] sphericity This invention characterizes the degree to which a particle's shape approximates a sphere. The sphericity is calculated by dividing the surface area of ​​an equal-volume sphere by the particle's surface area. The numerical range is 0 to 1; the closer the value is to 1, the closer the particle is to a sphere; the smaller the value, the more irregular the particle shape.

[0073] The formula for calculating the surface area of ​​a sphere with the same volume as the particle is as shown in equation (9): in, Surface area of ​​a sphere of equal volume, in mm² 2 .

[0074] The formula for calculating sphericity is as shown in equation (10): in, : Sphericity, dimensionless; : Particle surface area, in mm 2 .

[0075] Substituting the formula for calculating the surface area of ​​a sphere of equal volume, the final formula for calculating sphericity is as shown in equation (11): Extensibility Also known as aspect ratio, it describes the needle-like nature of the particles. This invention calculates elongation using the ratio of the longest direction to the shortest direction of the particle. The value ranges from ≥1; the larger the value, the finer and longer the particles; the closer the value is to 1, the more equiaxed the particles.

[0076] First, extract the dimensions of the bounding box of the particles. The formulas for calculating the dimensions of the bounding box in the three directions are as shown in equations (12) to (14): in, : x-axis length, in mm; : y-axis length, in mm; Z-axis length, in mm.

[0077] Sort the three dimensions from largest to smallest to obtain the major axis, median axis, and minor axis: in, : Longest shaft dimension, in mm; Intermediate shaft dimensions, in mm; : Shortest shaft dimension, in mm.

[0078] Extensibility The calculation formula is shown in equation (16).

[0079] in, : Extension or aspect ratio, dimensionless.

[0080] Flatness Flatness is used to describe the degree of flaky texture of particles. This invention calculates flatness using the ratio of the shortest axis to the longest axis. The value ranges from 0 to 1; the smaller the value, the finer and longer the particles; the larger the value, the more isometric the particles.

[0081] The formula for calculating flatness is as shown in equation (17): in: Flatness, dimensionless; : Shortest shaft dimension, in mm; : Longest shaft dimension, in mm.

[0082] angularity This invention describes the severity of the sharp edges on the surface of particles. It characterizes the particle sharpness index by calculating the average angle between adjacent triangular facets, and obtains the final sharpness through normalization. The value ranges from 0 to 1; the larger the value, the more pronounced the particle sharpness; the smaller the value, the more rounded the particle.

[0083] The formula for calculating the included angle between adjacent triangular facets is shown in equation (18): in, : Angle between the normals of adjacent facets, in rad; : No. The unit normal vector of a triangular facet is dimensionless. : the adjacent first The unit normal vector of a triangular facet is dimensionless.

[0084] The formula for calculating the overall angularity of particles is as shown in equation (19): in: Angularity, dimensionless; Pi, here representing the maximum possible normal angle (180°); : The number of adjacent triangular facet pairs, in units of ; : No. The angle between the normals of adjacent facets, in rad.

[0085] convexity An index used to describe the degree of concavity in particle contours. This invention calculates convexity using the ratio of particle surface area to convex hull area. The particle convex hull is the smallest convex polyhedron that can enclose all points of the particle. The line connecting any two points on the convex hull surface falls inside or on the surface of the polyhedron. The convex hull area is obtained using the triangular facet method, and the calculation formulas are referenced from equations (6) to (7). The numerical range is 0~1; the larger the value, the fuller the particle contour; the smaller the value, the higher the degree of concavity in the particle.

[0086] convexity The calculation formula is as shown in equation (20): in, : Convexity, dimensionless; : Area of ​​particle convex hull, in mm 2 A: Particle surface area, in mm² 2 .

[0087] Surface roughness This invention is used to characterize the dispersion of particle point clouds in three-dimensional space. It employs PCA principal component eigenvalue ratios to calculate surface roughness, performing principal component analysis on the point cloud coordinates to obtain eigenvalues ​​along three principal component directions. Each eigenvalue corresponds to the dispersion of the data along that axis. The following conditions should be met: in, : Eigenvalues ​​of the first principal component; : Eigenvalues ​​of the second principal component; : Eigenvalues ​​of the third principal component.

[0088] The ratio of the third eigenvalue to the first eigenvalue is used as the surface roughness index.

[0089] in, Surface roughness index, dimensionless; : Eigenvalues ​​of the first principal component, in mm 2 ; : Eigenvalues ​​of the third principal component, in mm 2 .

[0090] 3D morphology quantification and analysis: (1) Extraction of three-dimensional morphology parameters of aggregates Based on the evaluation index calculation method, the three-dimensional morphology parameters of the aggregates were extracted using the MATLAB platform. The morphology parameter data for 30 sets of inert steel slag, limestone, and diabase aggregates are shown in Tables 2, 3, and 4.

[0091] Table 2 Three-dimensional morphology parameters of inert steel slag aggregate Aggregate Number Volume (mm3) Surface area (mm2) Equivalent sphere diameter (mm) sphericity Extensibility Flatness angularity convexity Surface roughness 1 2340.645 968.675 16.473 0.8801 2.0026 0.4994 0.7912 0.4295 0.1003 2 1634.390 733.071 14.615 0.9153 1.9720 0.5071 0.8752 0.4986 0.1747 3 1834.416 798.891 15.188 0.9071 1.1088 0.9019 0.9318 0.4951 0.5745 4 2487.083 965.524 16.810 0.9194 1.6176 0.6182 0.8942 0.5120 0.1579 5 1663.216 737.435 14.700 0.9206 1.7433 0.5736 0.7968 0.5227 0.2358 6 1489.917 684.856 14.171 0.9211 1.6052 0.6230 0.8336 0.4502 0.2605 7 2465.263 947.640 16.761 0.9313 1.4469 0.6911 0.7740 0.4734 0.2396 8 2037.914 883.786 15.730 0.8796 1.3735 0.7281 0.9960 0.4464 0.2735 9 2144.794 873.433 16.000 0.9208 1.5995 0.6252 0.7583 0.4471 0.1548 10 2675.769 1025.050 17.225 0.9093 1.6684 0.5994 0.7666 0.4881 0.1472 11 1872.593 809.617 15.293 0.9075 1.5424 0.6483 0.6766 0.5487 0.3659 12 1612.310 718.619 14.548 0.9253 1.3132 0.7615 0.9960 0.4577 0.4721 13 2406.017 954.310 16.625 0.9099 1.5293 0.6539 0.8252 0.4479 0.1962 14 3382.547 1190.587 18.624 0.9153 1.7095 0.5850 0.8224 0.4355 0.1657 15 1639.374 770.926 14.629 0.8721 1.8049 0.5541 0.9324 0.4492 0.1620 16 2635.231 1046.314 17.137 0.8818 2.2121 0.4521 0.6706 0.5587 0.3965 17 2054.294 857.649 15.772 0.9112 1.3909 0.7190 0.9932 0.4039 0.5541 18 1318.138 652.072 13.604 0.8916 1.5627 0.6399 0.9977 0.4483 0.1348 19 2965.657 1109.877 17.825 0.8994 1.5834 0.6315 0.6365 0.4260 0.2073 20 2791.121 1047.707 17.469 0.9150 1.5027 0.6655 0.7722 0.4457 0.2064 21 1646.070 735.129 14.649 0.9171 1.5109 0.6619 0.7592 0.4659 0.3407 22 1881.240 807.712 15.316 0.9124 1.4652 0.6825 0.9781 0.4476 0.1817 23 2039.390 878.627 15.734 0.8851 1.6589 0.6028 0.6097 0.4779 0.2857 24 2391.784 957.644 16.592 0.9032 1.5625 0.6400 0.8517 0.4331 0.1872 25 1327.837 640.228 13.637 0.9125 1.5298 0.6537 0.7688 0.4730 0.3881 26 1564.588 737.524 14.403 0.8837 1.7789 0.5621 0.7827 0.4333 0.3129 27 1572.495 731.566 14.428 0.8939 1.6546 0.6044 0.7102 0.4691 0.3249 28 2762.070 1041.037 17.408 0.9145 1.4833 0.6742 0.7642 0.4113 0.2099 29 2638.602 1032.203 17.145 0.8946 1.7898 0.5587 0.7760 0.4633 0.1367 30 1569.047 720.808 14.417 0.9059 1.2971 0.7709 0.8930 0.4346 0.4275 Table 3 Three-dimensional morphological parameters of limestone aggregate Aggregate Number Volume (mm3) Surface area (mm2) Equivalent sphere diameter (mm) sphericity Extensibility Flatness angularity convexity Surface roughness 1 3018.660 1152.691 17.931 0.8763 1.5073 0.6634 0.8603 0.4215 0.3850 2 3512.027 1297.569 18.859 0.8611 1.7705 0.5648 0.6820 0.3922 0.1962 3 2283.237 930.363 16.337 0.9013 1.5633 0.6397 0.9811 0.4879 0.2009 4 2404.855 945.436 16.623 0.9181 1.3310 0.7513 0.8448 0.4085 0.3124 5 1778.133 776.531 15.031 0.9140 1.1912 0.8395 0.9517 0.5132 0.2912 6 2425.516 966.870 16.670 0.9029 1.9533 0.5120 0.6558 0.4355 0.1261 7 2014.433 859.562 15.669 0.8974 1.7082 0.5854 0.6611 0.4316 0.2183 8 2323.464 948.561 16.433 0.8944 1.8113 0.5521 0.9186 0.4508 0.1471 9 2368.492 942.488 16.538 0.9117 1.3462 0.7429 0.8234 0.4043 0.3493 10 3385.772 1146.774 18.630 0.9508 1.2469 0.8020 0.9494 0.4159 0.3958 11 2593.277 1007.578 17.046 0.9060 1.4049 0.7118 0.7583 0.4624 0.2261 12 2827.283 1061.076 17.544 0.9113 1.6692 0.5991 0.7708 0.4749 0.1743 13 2168.293 881.997 16.059 0.9185 1.4548 0.6874 0.8700 0.4139 0.2545 14 2096.587 858.692 15.880 0.9225 1.5498 0.6453 0.7592 0.4504 0.2324 15 2632.603 1117.519 17.132 0.8251 2.3900 0.4184 0.6668 0.4826 0.0736 16 3090.313 1089.335 18.072 0.9419 1.3697 0.7301 0.7813 0.4320 0.3081 17 2497.437 969.980 16.833 0.9177 1.3971 0.7158 0.7479 0.4913 0.2230 18 1565.392 761.195 14.406 0.8565 2.0671 0.4838 0.7873 0.4501 0.1103 19 2423.341 956.506 16.665 0.9122 1.4497 0.6898 0.7297 0.4036 0.2571 20 2720.604 1027.793 17.320 0.9170 1.6641 0.6009 0.7176 0.4465 0.1904 21 2645.007 1057.819 17.158 0.8744 1.6552 0.6041 0.6588 0.3948 0.2008 22 2400.078 1007.798 16.612 0.8602 1.7200 0.5814 0.8099 0.3986 0.1295 23 3470.090 1228.423 18.784 0.9023 1.6515 0.6055 0.7357 0.4465 0.1669 24 2585.814 1049.799 17.029 0.8678 2.3166 0.4317 0.6503 0.4447 0.1132 25 2752.524 1107.019 17.388 0.8580 2.4301 0.4115 0.7923 0.4345 0.0907 26 1236.341 611.013 13.316 0.9117 1.8748 0.5334 0.9868 0.4351 0.1691 27 2298.630 927.188 16.374 0.9084 1.3644 0.7329 0.9186 0.5020 0.3555 28 2227.147 920.234 16.203 0.8962 1.3658 0.7322 0.9113 0.4544 0.4041 29 1855.838 839.308 15.247 0.8701 2.0971 0.4769 0.9026 0.4762 0.1058 30 2535.718 1037.457 16.919 0.8668 1.5830 0.6317 0.9005 0.3624 0.1795 Table 4 Three-dimensional morphological parameters of diabase aggregate Aggregate Number Volume (mm3) Surface area (mm2) Equivalent sphere diameter (mm) sphericity Extensibility Flatness angularity convexity Surface roughness 1 2836.481 1038.741 17.563 0.9329 1.3856 0.7217 0.7428 0.5246 0.3203 2 2212.118 921.908 16.166 0.8906 1.7960 0.5568 0.5813 0.4252 0.1454 3 3491.906 1232.732 18.823 0.9029 1.7115 0.5843 0.7875 0.5168 0.1693 4 2917.314 1109.383 17.728 0.8900 1.5352 0.6514 0.9355 0.3790 0.2112 5 2315.896 997.763 16.415 0.8484 2.2378 0.4469 0.6610 0.4061 0.1120 6 2556.177 1003.332 16.964 0.9011 1.4125 0.7079 0.8760 0.4267 0.4108 7 2553.921 1031.385 16.959 0.8761 1.3043 0.7667 0.9426 0.5020 0.3770 8 2343.915 958.616 16.481 0.8902 1.1533 0.8671 0.9376 0.4426 0.6188 9 3262.586 1144.737 18.402 0.9293 0.9293 0.6883 0.8714 0.4541 0.2627 10 2339.989 967.242 16.472 0.8812 1.4395 0.6947 0.7230 0.4084 0.2758 11 2685.507 982.425 17.246 0.9510 1.5255 0.6555 0.7556 0.4771 0.2759 12 1770.030 761.865 15.008 0.9288 1.5221 0.6570 0.8711 0.5514 0.2983 13 1986.191 866.261 15.596 0.8821 2.1268 0.4702 0.5233 0.5574 0.1634 14 1319.916 648.564 13.610 0.8972 1.5749 0.6350 0.9095 0.4081 0.2265 15 1861.357 879.199 15.262 0.8323 1.3845 0.7223 0.9854 0.3794 0.3824 16 2196.437 898.852 16.128 0.9091 1.5831 0.6317 0.9378 0.4573 0.2808 17 1725.881 807.041 14.882 0.8622 1.7361 0.5760 0.6094 0.4441 0.2140 18 2873.611 1105.998 17.639 0.8838 1.5041 0.6648 0.7769 0.4664 0.4223 19 2798.317 1054.219 17.484 0.9109 1.4665 0.6819 0.9431 0.4995 0.1765 20 2680.868 1047.882 17.236 0.8906 1.6953 0.5899 0.7953 0.4347 0.1802 21 2051.942 888.676 15.766 0.8787 1.8279 0.5471 0.9877 0.4635 0.1550 22 1939.612 896.047 15.473 0.8394 2.6167 0.3822 0.6873 0.4073 0.0912 23 1837.140 843.802 15.195 0.8597 1.9410 0.5152 0.6614 0.4441 0.1509 24 3590.177 1254.237 18.998 0.9040 1.4440 0.6925 0.7078 0.4093 0.2307 25 2222.735 939.701 16.192 0.8765 2.2922 0.4363 0.6897 0.4399 0.1067 26 1321.938 750.659 13.617 0.7760 2.8994 0.3449 0.9926 0.4671 0.0503 27 2747.223 1057.989 17.377 0.8966 1.2253 0.8161 0.9919 0.5528 0.4590 28 2162.951 896.664 16.045 0.9020 1.4135 0.7075 0.8262 0.4113 0.2564 29 1777.607 823.905 15.030 0.8613 2.3065 0.4336 0.8535 0.5137 0.1222 30 1698.958 774.528 14.805 0.8890 1.7811 0.5615 0.7417 0.4092 0.1745 Table 5 summarizes the morphological parameters of inert steel slag, limestone and diabase aggregates.

[0092] Table 5 Summary of three-dimensional morphology of inert steel slag, limestone, and diabase aggregates Aggregate type sphericity Extensibility Flatness angularity convexity Surface roughness Inert steel slag 0.9052 1.6007 0.6363 0.8211 0.4631 0.2658 limestone 0.8958 1.6635 0.6226 0.8061 0.4406 0.2196 diabase 0.8858 1.6924 0.6136 0.8102 0.4560 0.2440 Based on Table 6, the three-dimensional morphological parameters of three aggregates—inert steel slag, limestone, and diabase—were analyzed by integrating 30 sets of morphological parameter data. The results show that inert steel slag exhibits significant advantages over limestone and diabase aggregates in several key morphological indicators.

[0093] The sphericity of inert steel slag is 0.9052, slightly higher than that of limestone (0.8958) and diabase (0.8858), indicating that its particle shape is closer to spheres, which is beneficial for maintaining moderate fluidity in the mixture while also ensuring adhesion stability. The elongation is 1.6007, lower than that of limestone (1.6635) and diabase (1.6924); the flatness is 0.6363, higher than that of limestone (0.6226) and diabase (0.6136). The lower elongation and higher flatness indicate that the steel slag particles are closer to an equiaxed state, which helps improve particle packing and reduce the adverse effects of needle-like and flaky particles. The angularity is 0.8211, higher than that of limestone (0.8061) and diabase (0.8102). The higher angularity results in stronger interparticle mechanical interlocking ability of inert steel slag, which helps improve the shear strength and adhesion stability of asphalt mixtures. The convexity is 0.4631, higher than that of limestone (0.4406) and diabase (0.4526). This indicates that the overall profile of inert steel slag particles is fuller, with a relatively lower degree of concavity. Higher convexity facilitates the formation of a continuous coating of asphalt on the particle surface, thereby improving interfacial adhesion stability. The surface roughness is 0.2658, higher than that of limestone (0.2196) and diabase (0.2440). This indicates that the surface texture of inert steel slag is richer, providing a larger contact area and friction, which is beneficial for enhancing the bonding ability between asphalt and aggregate.

[0094] In summary, inert steel slag exhibits lower elongation and higher sphericity, flatness, angularity, convexity, and surface roughness compared to limestone and diabase aggregates. The higher angularity, convexity, and surface roughness contribute to enhanced intergranular interlocking and interfacial adhesion, thus inert steel slag possesses better skeletal support and interfacial bonding potential in terms of three-dimensional morphology.

[0095] (2) Pearson correlation analysis There are correlations among the various three-dimensional morphological parameters of the aggregate. To avoid redundancy of variables in subsequent simulation pull-out tests, this invention uses the Pearson correlation coefficient to analyze the degree of correlation between the morphological parameters, thus simplifying the parameter index system.

[0096] The formula for calculating the Pearson correlation coefficient is shown in Equation 23: Where r is the correlation coefficient, with a value between -1 and 1; |r|≥0.8 indicates high correlation, 0.5≤|r|<0.8 indicates moderate correlation, 0.3≤|r|<0.5 indicates low correlation, and |r|<0.3 indicates weak correlation. The results of Pearson correlation analysis of the morphological parameters of inert steel slag are shown in Table 6.

[0097] Table 6. Pearson Correlation Analysis of Morphological Parameters of Inert Steel Slag Where ** indicates a significant correlation at the 0.01 level; * indicates a significant correlation at the 0.05 level.

[0098] Table 6 shows that there are varying degrees of correlation among the morphological parameters of inert steel slag aggregate. Volume, surface area, and equivalent sphere diameter show a highly significant positive correlation, with correlation coefficients all greater than 0.99. Elongation and flatness show a highly significant negative correlation, with a correlation coefficient of -0.975. Sphericity, angularity, convexity, and surface roughness parameters show weak correlations and can reflect particle morphological characteristics relatively independently. Since aggregate surface area and flatness directly affect the contact area with asphalt, significantly influencing the adhesion performance between aggregate and asphalt, considering the representativeness and correlation of each parameter, surface area, sphericity, flatness, angularity, convexity, and surface roughness were ultimately selected as the morphological control variables for subsequent steel slag-asphalt pull-out tests.

[0099] Geometric reconstruction of inert steel slag aggregate: Based on aggregate morphology parameter data and calculation methods, this invention utilizes a MATLAB platform to construct an algorithm for geometric reconstruction of inert steel slag particles, obtaining inert steel slag particles with specific morphologies for subsequent simulated pull-out test studies. The aggregate geometric reconstruction command is detailed in Appendix B. The reconstructed morphology of the inert steel slag particles is shown below. Figure 5 .

[0100] Based on 30 sets of inert steel slag morphology data and Pearson correlation analysis results, six morphology parameters were finally selected. Each parameter was taken within three ranges: minimum, mean, and maximum. Thirteen inert steel slag particles were reconstructed for subsequent asphalt pull-out performance studies. The target parameter settings for the 13 inert steel slag particles are shown in Table 7.

[0101] Table 7 Target morphology parameters of inert steel slag Steel slag-asphalt simulated pull-out test: The adhesion properties between asphalt and aggregates are crucial factors affecting the strength and durability of asphalt mixtures. The three-dimensional morphology of the aggregates directly influences the encapsulation, interlocking, and interfacial adhesion of asphalt on the aggregate surface. Based on PFC3D software, an inert steel slag-asphalt pull-out test model was established. Adhesion strength was used as the evaluation index to analyze the influence of different morphological characteristics on interfacial adhesion strength. The pull-out test command is provided in Appendix C.

[0102] Inert steel slag-asphalt pull-out model, such as Figure 6 .

[0103] During the pull-out process, the interface failure manifests in three stages. In the initial loading stage, the pull-out force increases rapidly with the increase of displacement, indicating that the asphalt and steel slag particles still maintain a good bond. As the pull-out displacement continues to increase, local contact at the interface begins to be damaged, and the rate of increase of the pull-out force gradually slows down, resulting in a plateau. When the pull-out force reaches its peak, the interface bond failure develops rapidly, the pull-out force decreases, and the steel slag particles gradually detach from the asphalt matrix.

[0104] The results of the pull-out tests of inert steel slag particles and asphalt are shown in Table 8. Table 8 Adhesion strength of asphalt-inert steel slag particles Inert steel slag aggregate number Surface area (mm2) sphericity Flatness angularity Surface roughness convexity Adhesion strength (MPa) 1 868.617 0.9052 0.6363 0.8211 0.2658 0.4631 12.4292 2 1190.587 0.9052 0.6363 0.8211 0.2658 0.4631 12.9013 3 640.228 0.9052 0.6363 0.8211 0.2658 0.4631 10.1616 4 868.617 0.9313 0.6363 0.8211 0.2658 0.4631 12.3907 5 868.617 0.8721 0.6363 0.8211 0.2658 0.4631 12.6175 6 868.617 0.9052 0.9019 0.8211 0.2658 0.4631 10.2848 7 868.617 0.9052 0.4521 0.8211 0.2658 0.4631 12.7647 8 868.617 0.9052 0.6363 0.9977 0.2658 0.4631 12.8718 9 868.617 0.9052 0.6363 0.6097 0.2658 0.4631 10.3796 10 868.617 0.9052 0.6363 0.8211 0.5745 0.4631 13.3016 11 868.617 0.9052 0.6363 0.8211 0.1003 0.4631 10.0072 12 868.617 0.9052 0.6363 0.8211 0.2658 0.5587 12.6225 13 868.617 0.9052 0.6363 0.8211 0.2658 0.4039 12.0014 The results of the three-dimensional morphology indices of inert steel slag and the adhesion strength to asphalt are as follows: Figure 7 .

[0105] Depend on Figure 7 It can be seen that the surface area, flatness, angularity, convexity and surface roughness of inert steel slag are generally positively correlated with the adhesion strength of asphalt, while sphericity is weakly negatively correlated with the adhesion strength.

[0106] In terms of specific values, the surface area is approximately 640.228 mm². 2 Increased to approximately 1190.587 mm 2 At that time, the adhesion strength increased from 10.616 MPa to 12.9013 MPa, an increase of about 21.5%, indicating that a larger aggregate surface area is beneficial to increasing the contact interface between asphalt and steel slag, thereby improving the interfacial adhesion strength.

[0107] When the sphericity increased from approximately 0.8721 to 0.9313, the adhesive strength decreased from 12.6175 MPa to 12.3907 MPa, a decrease of approximately 1.8%, indicating that the effect of sphericity on adhesive strength is relatively small. Higher sphericity makes the particle shape closer to a regular sphere, which may weaken the mechanical interlocking effect between asphalt and aggregate surfaces; while a moderately irregular particle shape is more conducive to improving the interfacial bonding effect.

[0108] In this invention, a smaller flatness parameter value indicates flatter particles, while a larger value indicates fuller particle shape. As shown in the figure, as the flatness parameter increases from 0.4521 to 0.9019, the tensile strength increases from 10.284 MPa to 12.764 MPa, an increase of approximately 24.1%. This indicates that flat particles may lead to stress concentration at the interface during stress application, reducing the stable bonding between the asphalt film and the aggregate surface.

[0109] When the angularity increased from approximately 0.6097 to approximately 0.9977, the adhesion strength increased from 10.3796 MPa to 12.878 MPa, an increase of approximately 24.1%, indicating that increased angularity can significantly improve the adhesion strength of asphalt. Steel slag particles with more prominent edges can enhance the interlocking and mechanical bonding between asphalt and aggregates, thereby improving interfacial adhesion performance.

[0110] When the surface roughness increased from approximately 0.1003 to approximately 0.5745, the adhesion strength increased from 10.0072 MPa to 13.3016 MPa, an increase of approximately 32.9%, indicating that surface roughness significantly affects the adhesion strength between inert steel slag and asphalt. A rougher aggregate surface provides more micro-adhesion points, enhancing the spreading, embedding, and mechanical bonding of the asphalt film, thus contributing to improved interfacial strength between steel slag and asphalt.

[0111] When the convexity increased from approximately 0.4039 to approximately 0.5587, the adhesive strength increased from 12.0014 MPa to 12.6225 MPa, an increase of approximately 5.2%, indicating that the increase in convexity has a certain promoting effect on adhesive strength, but the degree of influence is relatively limited. Appropriately improving the particle shape integrity and surface protrusion characteristics helps to improve the contact state between asphalt and aggregate, but its effect is weaker than that of surface roughness, angularity, and surface area.

[0112] Comprehensive analysis shows that surface roughness, angularity, and surface area have a significant effect on improving the adhesion strength of the asphalt-steel slag interface; excessive flatness weakens adhesion performance; and sphericity and convexity have relatively weak effects. Therefore, in controlling the morphology of steel slag aggregates, priority should be given to selecting particles with rougher surfaces, moderate angularity, and larger specific surface areas, while avoiding excessive content of flaky and flat particles.

[0113] Analysis of Simulation Pull-Out Test Results and Engineering Recommendations: Three-dimensional morphology parameters and adhesion performance analysis This invention employs grey relational analysis to analyze the sensitivity of various three-dimensional morphology indicators of inert steel slag to adhesion strength. Using adhesion strength as the reference sequence and each particle morphology indicator as the comparison sequence, the influence of each indicator on adhesion strength is evaluated based on the magnitude of the correlation. The resolution coefficient is set as follows: .

[0114] The formula for calculating grey relational analysis is as follows: In the formula, For the first Grey relational degree between morphological parameters and adhesion strength. The larger the value, the stronger the correlation between the morphological parameter and the adhesion strength. The results of the grey relational analysis of the morphological parameters and adhesion strength of inert steel slag are shown in Table 9.

[0115] Table 9. Grey Relational Analysis of Morphology Parameters and Adhesion Strength of Inert Steel Slag Morphological parameters Related directions Grey relational degree Range of tensile strength (MPa) Correlation judgment Flatness Positive correlation 0.637 2.480 Strong angularity Positive correlation 0.614 2.498 Strong Surface roughness Positive correlation 0.594 3.294 Strong Surface area Positive correlation 0.576 2.740 Strong sphericity weak negative correlation 0.526 0.227 Weak convexity Positive correlation 0.510 0.621 Weak Based on the results in Table 9, there is a certain correlation between the three-dimensional morphology parameters of inert steel slag aggregate and the adhesion strength of asphalt, but the degree of influence varies. The sensitivity of each parameter is ranked as follows: Among them, the gray correlation coefficients of flatness, angularity, surface roughness, and surface area are all greater than 0.57, indicating that they have a strong correlation with adhesion strength; the gray correlation coefficients of sphericity and convexity are 0.526 and 0.510, respectively, indicating a relatively weak correlation.

[0116] The gray correlation coefficient between flatness and adhesion strength was the highest, at 0.637, with a pull-out strength variation range of 2.480 MPa, indicating that flatness is a crucial factor affecting interfacial adhesion performance. The flatness used in this invention is defined as the short axis / long axis ratio; a higher value indicates a fuller particle morphology, while a lower value indicates a flatter particle shape. Therefore, it is evident that excessively flat steel slag particles are not conducive to adhesion with asphalt, while equiaxed particles are more beneficial for enhancing the adhesion strength between steel slag and asphalt.

[0117] The gray correlation coefficient for angularity is 0.614, and the pull-out strength ranges from 2.498 MPa, indicating a strong correlation. Angularity is positively correlated with adhesion strength; the more prominent the particle edges, the higher the adhesion strength between steel slag and asphalt. Angularity enhances the interlocking effect between steel slag and asphalt surfaces, thereby improving interfacial adhesion.

[0118] The gray correlation coefficient for surface roughness is 0.594, and the pull-out strength varies by 3.294 MPa, making it the indicator with the largest strength variation among all morphological parameters. Although its gray correlation coefficient ranking is lower than that of flatness and angularity, surface roughness has the most significant effect on improving adhesion strength in terms of the range of strength variation. Higher surface roughness can increase the micro-undulations and adhesion points on the aggregate surface, making it easier for asphalt to embed into the pores and uneven structures of the steel slag surface, thereby enhancing interfacial adhesion performance.

[0119] The grey relational degree of surface area is 0.576, and the pull-out strength ranges from 2.740 MPa. This indicates a strong correlation between surface area and adhesion strength. As the aggregate surface area increases, the effective contact area between asphalt and steel slag increases, which is beneficial for improving interfacial bond strength.

[0120] Sphericity is negatively correlated with adhesion strength, with a grey relational degree of 0.526 and a minimum range of 0.227 MPa for pull-out strength. This indicates that sphericity has a weak effect on adhesion strength. Higher sphericity results in more spherical particles, potentially weakening the mechanical interlocking between asphalt and aggregate; however, due to the small range of sphericity variation in this data set, its impact on adhesion strength is not significant.

[0121] The gray correlation coefficient for convexity is 0.510, indicating the lowest correlation. The pull-out strength ranges from 0.621 MPa, suggesting that convexity has a relatively limited impact on adhesion strength. Increasing convexity can improve particle shape integrity and interfacial contact stability to some extent, but its effect is weaker than that of flatness, angularity, surface roughness, and surface area.

[0122] In summary, flatness, angularity, surface roughness, and surface area are the main controlling factors; sphericity and convexity have little effect on the steel slag-asphalt interface. Equiaxed, angular, and rough steel slag particles are more conducive to improving the adhesion strength of the steel slag-asphalt interface.

[0123] Aggregate morphology control indicators and processing optimization suggestions: Based on the results of steel slag-asphalt pull-out tests and grey relational analysis, it is recommended that flatness, angularity, and surface roughness be included in the quality evaluation system for steel slag aggregates. Sphericity and convexity can be used as auxiliary control indicators, but should not be used as the main evaluation criteria.

[0124] In the processing of steel slag aggregates, a morphology control approach of "reducing flat particles, retaining angular features, enhancing surface roughness, and optimizing surface area" should be adopted to improve the interfacial adhesion between steel slag and asphalt. A combined process of "jaw crushing or cone crushing for primary crushing + impact crushing or vertical shaft impact crushing for shaping" is recommended for aggregate processing and shaping.

[0125] (1) Primary crushing stage: Use a jaw crusher or cone crusher to crush the aggregate. During the compression process, steel slag will crack along the weak internal surface to form a new fracture surface. These fracture surfaces are usually relatively rough and have certain angularity, which is conducive to preserving the basic angular characteristics of the aggregate. However, simple jaw crushing is prone to producing a certain proportion of flaky and strip-shaped particles, so a shaping process is required later.

[0126] (2) Shaping stage: Impact crushing using impact crushers or vertical shaft impact crushers improves the morphology of aggregates. First, flat particles are more prone to secondary crushing due to their small thickness and weak structure, and are crushed into smaller and fuller particles, which can effectively reduce the content of flat particles. At the same time, it can avoid the weakening of the angularity and texture of aggregates by excessive abrasion or polishing, and preserve the angularity and surface roughness of aggregates.

[0127] (3) Screening control: Add a screening and morphology inspection step for flat particles during the screening stage. Particles with too low flatness parameters, too high sphericity, or too smooth surface should be rejected or returned to the shaping process.

Claims

1. A method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate, characterized in that, Includes the following steps; Step (1): Inertization pretreatment of steel slag; Step (2): Use a 3D modeling and digitization machine to collect the three-dimensional morphological parameters of the aggregate, and quantify the morphological parameters through the MATLAB platform; The aggregate is inertized steel slag aggregate; Step (3): Pearson correlation analysis was used to analyze the correlation of quantified morphological parameters and explore the influence of parameter correlation on the adhesion performance between aggregate and asphalt. Step (4): Conduct a virtual pull-out test of aggregate-asphalt based on the PFC3D platform and collect test data; Step (5): Use the grey relational model to analyze the influence of morphology parameters on adhesion performance and complete the sensitivity ranking.

2. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 1, characterized in that, In step (1), the specific steps of the inertization pretreatment are as follows: mechanically crushing the inert steel slag; crushing the aggregate to a particle size range of 13.2-19mm; magnetically separating the crushed inert steel slag; and obtaining aggregate with excellent stability and high-value-added mineral admixtures.

3. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 1, characterized in that, The specific steps (2) are as follows: using an Einstar handheld 3D modeling and copying machine to obtain surface point cloud data of the aggregate, importing the data information into the MATLAB platform for analysis and calculation, establishing an index system covering macroscopic shape, mesoscopic edges and corners and microscopic texture, obtaining three-dimensional morphological parameters including volume, surface area, equivalent sphere diameter, sphericity, elongation, flatness, angularity, convexity and surface roughness, and quantitatively characterizing the morphological features of the aggregate particles.

4. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 3, characterized in that, In step (2), the specific steps for quantifying the morphological parameters using the MATLAB platform are as follows: The parameters of the aggregate particles were obtained by using an algorithm built on the MATLAB platform based on the 3D point cloud and mesh data exported from the Einstar scanner. volume The volume of a particle is calculated by using the volume integration method of closed triangular facets to represent the size of the space occupied by the particle in three-dimensional space. The particle volume is obtained by dividing the particle surface into many triangular facets, each of which forms a tetrahedron with the origin of the coordinate system. The volume of the particle is obtained by summing the volumes of all tetrahedrons. The coordinates of the vertices of the triangular facets on the particle surface are given by equation (1): in, : No. The three vertices of a triangular facet, in mm; For the first The coordinates of the vertices of the triangular facet in three-dimensional space, in mm; Let the first The coordinates of the three vertices of the triangular facet are given in equations (2) to (4): in, : No. The three vertices of a triangular facet, in mm; The volume calculation formula is as follows (5): in, Particle volume, in mm 3 ; : Number of triangular facets, in units of one; : Vector magnitude; Vector cross product; vector dot product; surface area This represents the area of ​​the particle surface. The particle surface area is obtained by summing the areas of the triangular faces on the particle surface. The formula for calculating the area of ​​a triangular facet is shown in equation (6): in, : No. Area of ​​each triangular facet, in mm² 2 ; The formula for calculating the total surface area of ​​particles is equation (7): in, : Particle surface area, in mm 2 ; Equivalent sphere diameter This refers to the diameter of a sphere with the same volume as the particle, based on the particle volume. The diameter of the equivalent sphere of the particle is calculated by reverse calculation; The volume calculation formula is as shown in equation (8): in, : Equivalent sphere diameter, in mm; Pi (π) is a dimensionless mathematical constant. sphericity To characterize the degree to which the particle shape approximates a sphere, sphericity is calculated by dividing the surface area of ​​an equal-volume sphere by the surface area of ​​the particle. The formula for calculating the surface area of ​​a sphere with the same volume as the particle is as shown in equation (9): in, Surface area of ​​a sphere of equal volume, in mm² 2 ; The formula for calculating sphericity is as shown in equation (10): in, : Sphericity, dimensionless; : Particle surface area, in mm 2 ; Substituting the formula for calculating the surface area of ​​a sphere of equal volume, the final formula for calculating sphericity is as shown in equation (11): Extensibility The elongation is calculated by the ratio of the longest direction to the shortest direction of the particle; First, extract the dimensions of the bounding box of the particles. The formulas for calculating the dimensions of the bounding box in the three directions are as shown in equations (12) to (14): in, : x-axis length, in mm; : y-axis length, in mm; Z-axis length, in mm; Sort the three dimensions from largest to smallest to obtain the major axis, median axis, and minor axis: in, : Longest shaft dimension, in mm; Intermediate shaft dimensions, in mm; : Shortest shaft dimension, in mm; Extensibility The calculation formula is shown in equation (16); in, : Extension or aspect ratio, dimensionless; Flatness Flatness is used to describe the degree of flaky texture of particles, calculated by the ratio of the shortest axis dimension to the longest axis dimension. The formula for calculating flatness is as shown in equation (17): in: Flatness, dimensionless; : Shortest shaft dimension, in mm; : Longest shaft dimension, in mm; angularity It is used to describe the severity of the sharpness of the particle surface edges. The particle edge index is characterized by calculating the average angle between adjacent triangular facets. The final edge index is obtained by normalization. The formula for calculating the included angle between adjacent triangular facets is shown in equation (18): in, : Angle between the normals of adjacent facets, in rad; : No. The unit normal vector of a triangular facet is dimensionless. : the adjacent first The unit normal vector of a triangular facet is dimensionless. The formula for calculating the overall angularity of particles is as shown in equation (19): in: Angularity, dimensionless; Pi, here representing the maximum possible normal angle of 180°; : The number of adjacent triangular facet pairs, in units of ; : No. The angle between the normals of adjacent facets, in rad; convexity The index used to describe the degree of concavity of the particle profile is the convexity calculated by the ratio of the particle surface area to the convex hull area. The particle convex hull is the smallest convex polyhedron that can enclose all points of the particle. The line connecting any two points on the convex hull surface falls inside or on the surface of the polyhedron. The convex hull area is obtained by the triangular facet method. convexity The calculation formula is as shown in equation (20): in, : Convexity, dimensionless; : Area of ​​particle convex hull, in mm 2 A: Particle surface area, in mm² 2 ; Surface roughness To characterize the dispersion of particle point clouds in three-dimensional space, surface roughness is calculated using the ratio of principal component eigenvalues ​​(PCA). Principal component analysis is performed on the point cloud coordinates to obtain eigenvalues ​​along the three principal component directions. Each eigenvalue corresponds to the dispersion of the data along that axis. The following conditions should be met: in, : Eigenvalues ​​of the first principal component; : Eigenvalues ​​of the second principal component; : Eigenvalues ​​of the third principal component; The ratio of the third eigenvalue to the first eigenvalue is used as the surface roughness index; in, Surface roughness index, dimensionless; : Eigenvalues ​​of the first principal component, in mm 2 ; : Eigenvalues ​​of the third principal component, in mm 2 .

5. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 1, characterized in that, The specific step (3) is: Pearson correlation analysis based on the original measured data; The three-dimensional morphological parameters of inert steel slag were collected as samples, including volume V, surface area A, and equivalent sphere diameter. Sphericity ψ, elongation AR, flatness F, angularity AI, convexity C, surface roughness ; The Pearson correlation coefficient was used to analyze the linear correlation between the parameters. Redundant variables were removed. The formula for calculating the Pearson correlation coefficient is as follows: in, , These are sample values ​​for two morphological parameters, , The sample mean. The value range of is [-1, 1]; The criteria for judgment are: Highly correlated The correlation is moderate. Volume, surface area, and equivalent sphere diameter showed a highly significant positive correlation, elongation and flatness showed a highly significant negative correlation, and the other parameters showed weak correlations. Filter the parameters as follows: surface area A, sphericity ψ, flatness F, angularity AI, convexity C, and surface roughness. A total of 6 features were selected as the main control morphological features for subsequent geometric reconstruction and simulation experiments; Virtual steel slag particles are generated based on particle geometry reconstruction with parameter control. Based on the selected master control parameters and their value ranges, an algorithm is constructed in the MATLAB platform to regenerate virtual inert steel slag particles with specific target parameters.

6. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 5, characterized in that, In step (4): Virtual particle geometry reconstruction based on core morphology parameter control is carried out through the MATLAB platform. The virtual steel slag particles generated by reconstruction are exported in STL format and imported into the PFC3D platform. An inert steel slag-asphalt pull-out test model is constructed based on PFC3D and a virtual pull-out test is carried out. Adhesion strength data of steel slag-asphalt interface under different morphology parameter conditions are collected. The specific method is as follows: Based on the measured data of inert steel slag, the minimum, mean and maximum values ​​of each main control parameter are determined, which constitute the parameter control range; The method employs triangular facets and a parameter-driven deformation algorithm: First, a basic sphere or ellipsoid mesh is generated. Then, it is scaled along the principal axis according to the target flatness, the surface normal is perturbed according to the target angularity, and Gaussian noise or fractal texture is added according to the target surface roughness. Finally, a closed triangular mesh model is generated. The reconstructed inert steel slag particles with different morphologies, each group corresponding to a specific set of surface area, sphericity, flatness, angularity, convexity, and surface roughness values, are exported as STL files for subsequent simulation use.

7. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 6, characterized in that, Step (4) specifically involves: a pull-out simulation test of inert steel slag-asphalt based on PFC3D. Import the generated reconstructed particle STL file into PFC3D, and use clump to create rigid particles to preserve the complete morphology. A pull-out model was constructed, including a bottom substrate, an asphalt film, a pull-out head, and a pull rod. The asphalt was represented by spherical particles, and a linear parallel bonding model was used to describe the bonding behavior inside the asphalt and at the steel slag-asphalt interface. The model parameters were inverted and calibrated based on the macroscopic test results of the base asphalt. A vertical, uniform pulling displacement is applied to each reconstructed particle, the force-displacement curve during the pulling process is recorded, and the adhesion strength is calculated according to the following formula: in, The tensile strength is expressed in MPa. This represents the maximum pulling force (N) that occurs during the pulling process. The effective adhesion area (mm²) between asphalt and steel slag particles is the product of the reconstructed particle surface area and the asphalt contact area coefficient.

8. The method for three-dimensional morphology reconstruction and adhesion performance evaluation of inert steel slag aggregate according to claim 7, characterized in that, The sensitivity ranking in step (5) based on the grey relational model is as follows: Using adhesion strength as the reference sequence and morphological parameters as the comparison sequence, grey relational analysis was employed to quantitatively evaluate the influence of each morphological parameter on adhesion strength. The formula for calculating grey relational degree is as follows: in, This is the reference sequence for the k-th sample. For the comparison sequence of the i-th morphology parameter, The resolution coefficient, The correlation coefficient, This represents the grey relational degree.

9. A three-dimensional morphology reconstruction and shaping system for inert steel slag based on the method of any one of claims 1-8, characterized in that, It is divided into two main modules: aggregate morphology scanning and morphology parameter extraction. Relying on the Einstar handheld 3D modeling and digitizing machine and MATLAB algorithm, it realizes the acquisition of three-dimensional point cloud of aggregate, model reconstruction and multi-dimensional morphology parameter quantification. The accuracy of the whole system can match the CT scan results and meet the requirements of experimental analysis. In the aggregate morphology scanning module, the inert steel slag aggregate is first pretreated, including ultrasonic cleaning and high-temperature drying. Then, titanium dioxide is thinly coated to give the steel slag surface a matte finish in order to eliminate imaging interference. Then, the accompanying EXstar software was launched, and distance and whiteness calibration were completed using a dedicated calibration board. The object scanning mode was selected, and the exposure value was adjusted to 40-60 according to the brightness of the aggregate, with the resolution set to 0.1mm to preserve microscopic textures. During scanning, the aggregate was placed in the center of the scanning stage, and the scanner moved slowly within a range of 30-40cm to collect point cloud data of the particle surface from all directions. The scanning time for a single particle was 30-60 minutes, and missing areas were scanned to ensure no blind spots. After scanning, the original point cloud data was corrected, and impurities and noise points were removed. The model was encapsulated, and ASC and STL format files were exported to provide the original 3D model for subsequent parameter extraction. In the topography parameter extraction module, a dedicated algorithm based on the MATLAB platform is used to perform calculations on the point cloud and mesh data obtained from the scan. First, import the point cloud file. The point cloud file refers to the ASC format file of the three-dimensional shape of the aggregate obtained by the handheld 3D modeling and copying machine. The system automatically removes NaN values, infinity values, and outliers to clean the data. Then, it calculates morphological parameters from three dimensions: macroscopic shape, mesoscopic angularity, and microscopic texture. After batch processing of aggregate sample data, it uses the Pearson correlation analysis algorithm to identify the coupling relationship between parameters, remove highly correlated redundant indicators, and finally select control parameters for surface area, sphericity, flatness, angularity, convexity, and surface roughness. The system also automatically generates aggregate point cloud maps, bounding box models, parameter radar charts, and size scale diagrams, and outputs a complete analysis report, providing quantitative basis for subsequent particle geometry reconstruction and simulated pull-out tests.

Images