Non-reactive asphalt regenerant diffusion effect evaluation method

Abstract

This invention belongs to the field of intelligent sensors, specifically relating to a method for evaluating the diffusion effect of non-reactive asphalt rejuvenator. The method includes: uniformly dispersing nanoscale high atomic number tracer particles into the rejuvenator to prepare a tracer rejuvenator; constructing a diffusion sample of the rejuvenator in contact with aged asphalt; acquiring three-dimensional density distribution data at different times using industrial micro-CT; reconstructing the spatial morphology of the rejuvenator within the aged asphalt through three-dimensional reconstruction and generating a three-dimensional grayscale cloud map; extracting diffusion interface information, calculating the fractal dimension and interface complexity factor to evaluate penetration uniformity; and fitting cross-time-segment sequence data based on the principle of material diffusion to generate a diffusion activation energy spectrum reflecting the diffusion barrier. This invention accurately evaluates the penetration effect through transparent-level microscopic quantitative indicators, guiding the design of rejuvenation processes in practical engineering.

Description

Technical Field

[0001] This invention belongs to the field of intelligent sensors, specifically relating to a method for evaluating the diffusion effect of non-reactive asphalt rejuvenator. Background Technology

[0002] With the continuous growth in demand for highway infrastructure maintenance, asphalt pavement recycling technology has become a key means to achieve resource recycling and reduce maintenance energy consumption. Non-reactive recyclers, through physical miscibility and penetration with aged asphalt matrix, can restore the rheological properties of aged asphalt, which is of great significance for extending the service life of pavements. Especially in high-performance pavement recycling projects, the microscopic diffusion depth and uniformity of the recycler in aged asphalt directly determine the stability and durability of the recycled material, which places high demands on the accuracy, real-time performance, and multi-dimensional quantification capabilities of diffusion effect evaluation methods.

[0003] Traditional evaluation methods primarily rely on macroscopic physical indicators such as penetration or softening point, which suffer from a severe black box effect, making it difficult to reveal the spatiotemporal evolution of rejuvenators within aged asphalt. This results in an inability to accurately identify pseudo-softening phenomena that only occur on the surface. Existing diffusion coefficient measurements often employ destructive testing techniques such as slicing or infrared spectroscopy, creating discontinuities in the temporal dimension of the test samples and failing to capture the nonlinear dynamic characteristics of the initial diffusion stage. Limited by the opacity and high-viscosity multiphase system characteristics of asphalt materials, conventional microscopic methods struggle to track the microscopic migration paths of rejuvenators in real-world conditions. This leads to technical challenges in the quantitative analysis of diffusion fronts and concentration gradients, hindering accurate guidance for dosage design and process optimization in practical engineering projects. Summary of the Invention

[0004] The purpose of this invention is to provide a method for evaluating the diffusion effect of non-reactive asphalt rejuvenator, which can effectively solve the problems in the background art.

[0005] To achieve the above objectives, the technical solution adopted by this invention is as follows: a method for evaluating the diffusion effect of a non-reactive asphalt recycling agent, comprising the following specific steps: Step 1: Prepare a regenerator containing tracer labels. Add a specific proportion of nanoscale high atomic number tracer particles to the non-reactive asphalt regenerator. Through mechanical shearing, the nanoscale high atomic number tracer particles are uniformly distributed within the non-reactive asphalt regenerator, ensuring that the migration behavior of the nanoscale high atomic number tracer particles is highly consistent with the molecular motion characteristics of the non-reactive asphalt regenerator. Step 2: Construct a diffusion test sample. Inject a preset amount of aged asphalt into a sample container and perform a leveling process. After the aged asphalt cools to a predetermined temperature, coat its upper surface with the recycling agent containing tracer markers to form a contact interface between the recycling agent and the aged asphalt. Step 3: Perform non-destructive imaging acquisition. Place the diffusion test sample in the detection chamber of the industrial micro-computed tomography equipment. Utilize the difference in X-ray absorption coefficient between the nanoscale high atomic number tracer particles and aged asphalt to obtain the three-dimensional density distribution data of the diffusion test sample at different diffusion time points. Step 4: Construct a digital twin 3D model. Based on the 3D density distribution data, use a 3D reconstruction algorithm to restore the spatial distribution of the non-reactive asphalt rejuvenator inside the aged asphalt and generate a 3D grayscale cloud map to intuitively represent the propagation process of the diffusion front. Step 5: Quantify diffusion morphology characteristics, extract diffusion interface information from the three-dimensional grayscale cloud map using image processing algorithms, calculate the fractal dimension of the diffusion interface, and evaluate the penetration uniformity of the non-reactive asphalt rejuvenator inside the aged asphalt through the interface complexity factor. Step 6: Analyze the diffusion kinetics, perform a time-series scan on the same diffusion test sample, and fit the concentration evolution trend of different depth layers over time in combination with the principle of material diffusion to generate a diffusion activation energy spectrum that reflects the diffusion barrier.

[0006] Preferably, in step 1, the nanoscale high atomic number tracer particles are selected as nano-bismuth oxide or nano-barium agents. The atomic number of the nanoscale high atomic number tracer particles is in a specific high range to ensure sufficient contrast increment during micro-computed tomography. The average particle size of the nanoscale high atomic number tracer particles is controlled within a very small preset range to avoid sedimentation caused by particle weight and to ensure its Brownian motion characteristics in viscous asphalt media.

[0007] Preferably, in step 1, the mechanical shearing process involves placing the mixed system in a preset high-temperature environment, applying intense hydrodynamic force to the mixed liquid using a high-speed rotating shear head, setting the shearing speed to a first preset value, and the duration reaching a predetermined cycle. Subsequently, the dispersion coefficient of the nanoscale high atomic number tracer particles is detected using the dynamic light scattering principle to ensure that its dispersion uniformity is higher than a preset mass threshold.

[0008] Preferably, in step 2, the preparation of the aged asphalt is carried out by a rotating thin film oven heating method or a pressure aging container aging method to simulate different aging stages after the road surface has been in service. The penetration, softening point and ductility of the aged asphalt need to be adjusted to meet the experimental requirements of a specific degree of aging. The smoothing treatment is achieved by standing in a preset high temperature constant temperature chamber to eliminate bubbles and uneven structures on the sample surface.

[0009] Preferably, in step 3, the resolution of the industrial micro-computed tomography scanning device is set at the micrometer level, and the scanning parameters include tube voltage, tube current and exposure time. The parameters are optimized according to the overall density of the diffusion test sample and the mass fraction of tracer particles. During the scanning process, the temperature inside the detection chamber is maintained at the preset mixing temperature or service temperature by a constant temperature control system to simulate the real engineering working environment.

[0010] Preferably, in step 4, the construction process of the digital twin 3D model involves performing filtering and deconvolution operations on the original projection data, establishing a spatial coordinate system through 3D voxel reconstruction technology, and the gray value of each pixel in the 3D grayscale cloud map represents the local density feature of that site. By normalizing the gray values, a linear mapping of the regenerant concentration is achieved.

[0011] Preferably, in step 5, the fractal dimension is calculated using the cube covering method. The specific steps are as follows: the extracted diffusion interface is fully covered by cube units of different scale lengths, the number of cubes containing the interface at each scale length is recorded, and linear regression analysis is performed in a double logarithmic coordinate system with the reciprocal of the scale length as the abscissa and the number of cubes containing the interface as the ordinate. The absolute value of the slope of the obtained straight line is the fractal dimension.

[0012] Preferably, the logic for determining the interface complexity factor is as follows: by comparing the difference between the actually measured fractal dimension and the fractal dimension of the ideal smooth interface, when the difference is lower than a first preset threshold, the diffusion process is determined to be uniform penetration; when the difference is higher than a second preset threshold and the diffusion interface exhibits finger-shaped or root-like protrusions, it is determined that there is a false diffusion phenomenon, that is, the regenerator only migrates in local channels and does not achieve uniform regeneration throughout the entire area.

[0013] Preferably, in step 6, the setting of the specific time series covers the initial diffusion stage, the intermediate stabilization stage, and the late saturation stage. The time interval is adaptively adjusted according to the viscosity grade of the asphalt. For high-viscosity aged asphalt, the total monitoring time is increased and the interval between single monitoring is appropriately extended.

[0014] Preferably, the principle of material diffusion is expressed using a modified Fick's second law, that is, the first partial derivative of the diffusion concentration with time is equal to the product of the diffusion coefficient and the second partial derivative of the diffusion concentration with spatial displacement. The diffusion coefficient under different temperature conditions is solved by integral fitting of the concentration distribution curves at different scanning times.

[0015] Preferably, the generation of the diffusion activation energy spectrum is based on the evolution of the Arrhenius equation. By calculating the linear relationship between the natural logarithm of the diffusion coefficient and the reciprocal of the absolute temperature, the activation energy required for the diffusion of regenerator molecules is determined. The smaller the activation energy value, the easier it is for the non-reactive asphalt regenerator to overcome the cohesive forces between aged asphalt molecules, and the higher its regeneration efficiency.

[0016] Preferably, the processing of the three-dimensional grayscale cloud image further includes median filtering noise reduction of the image, using a filter window of preset size to traverse the three-dimensional voxels to remove salt-and-pepper noise generated by ray scattering, and using the Laplacian operator to sharpen and enhance the diffusion boundary to improve the geometric accuracy of fractal dimension calculation.

[0017] Preferably, the evaluation method also involves establishing a multi-dimensional evaluation index system, which includes not only the average diffusion depth, but also the standard deviation of the bulk density distribution, the interface roughness coefficient, and the attenuation constant of the concentration gradient. By weighting and summarizing the various indicators, a comprehensive performance score for a specific non-reactive asphalt recycling agent is obtained.

[0018] Preferably, the mass fraction of the specific proportion of nanoscale high atomic number tracer particles in the regenerator is set within a specific range. The design principle of this range is to ensure sufficient grayscale recognition in the microscopic scanning image while ensuring that the dynamic viscosity and chemical activity of the regenerator are not changed due to excessive solid particle content.

[0019] Preferably, during the non-destructive radiographic imaging acquisition process, a precision displacement compensation device is installed in the detection chamber to ensure that the physical center of the sample under test and the scanning rotation axis remain coincident during the sequential scanning process that lasts for tens of hours, thereby eliminating motion artifacts caused by minute creep of the sample.

[0020] Preferably, the fractal dimension analysis also incorporates pore structure characteristics. By extracting the micropore distribution pattern inside aged asphalt, the contribution ratio of the capillary driving effect and molecular diffusion effect of the non-reactive asphalt rejuvenator in the porous medium is analyzed.

[0021] Preferably, the evaluation method is also applied to optimize construction process parameters. By comparing diffusion kinetic parameters under different preheating temperatures, different spraying pressures, and different settling times, the construction technology scheme that achieves the best regeneration effect is determined.

[0022] Preferably, the three-dimensional reconstruction algorithm adopts an algebraic reconstruction technique based on iterative reconstruction. When processing low signal-to-noise ratio projection data, it retains the edge details of the diffusion front by introducing a total variational regularization constraint, thus preventing diffusion depth assessment errors caused by excessive smoothing.

[0023] Preferably, the diffusion effect evaluation method further includes conducting microhardness tests on the aged asphalt after the rejuvenator diffusion, verifying the difference in mechanical modulus between the diffused and non-diffused areas through nanoindentation tests, and performing spatial coupling analysis of the micromechanical data and three-dimensional grayscale distribution data to construct a mapping model between mechanical properties and diffusion concentration.

[0024] Preferably, the temperature control system achieves a predetermined high-precision level of temperature control accuracy. It monitors the temperature gradient inside the sample in real time through multi-point distributed thermocouple sensors, ensuring that the diffusion process is under a single-variable thermodynamic state and eliminating the interference of temperature fluctuations on the diffusion rate.

[0025] Preferably, the quantification of the interface complexity factor also takes into account the spatial heterogeneity of the fractal dimension. By performing regional slicing on the three-dimensional diffusion interface, the local fractal dimension in different normal directions is calculated to evaluate the anisotropic diffusion characteristics of the regenerator in the horizontal and vertical directions.

[0026] Preferably, in step 3, the scanning process adopts a continuous rotation mode, and a predetermined number of projection images are acquired within a single rotation cycle. The overlap of the projection images is set above a certain ratio to ensure that the reconstructed three-dimensional voxels have high spatial continuity and isotropic resolution.

[0027] Preferably, the regenerant containing tracer markers needs to be degassed in a pre-set vacuum degassing chamber before contacting aged asphalt to remove tiny air bubbles introduced during the stirring process, preventing air bubbles from forming a physical barrier at the diffusion interface and affecting the accurate observation of diffusion behavior.

[0028] Preferably, the evaluation method is automated through a computer software platform. The software automatically performs image segmentation, fractal calculation, and dynamic fitting, and automatically generates an evaluation report containing diffusion cloud maps, concentration curves, and activation energy levels according to preset judgment criteria.

[0029] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention overcomes the damage to samples caused by traditional destructive testing methods such as slicing and infrared spectroscopy. It enables continuous tracking of the complete evolution process of the non-reactive asphalt rejuvenator from the contact surface to its complete penetration into the aged asphalt using the same specimen. This continuous dynamic monitoring improves the consistency and logical rigor of the evaluation data and eliminates statistical errors caused by sample differences.

[0030] 2. This invention provides microscopic quantitative indicators at the level of transparency. By combining nanoscale high atomic number tracer particles with industrial micro-computed tomography (CT) technology, the previously physically invisible internal diffusion behavior is successfully made explicit and digitized. This invention can accurately calculate the average diffusion depth of the regenerator and, more importantly, quantify the roughness and uniformity of the diffusion interface from a geometric and topological perspective using fractal dimension as a mathematical tool. This depth evaluation mechanism can identify and eliminate inferior regenerators that only produce a softening effect on the surface while the interior remains in an aged state, filling the gap in the industry's standards for evaluating the penetration quality of regenerators.

[0031] 3. This invention improves the correlation between evaluation results and engineering practice. The method has the capability to conduct in-situ tests under real-world conditions such as simulated high-temperature mixing or pavement service. By analyzing the diffusion kinetics of the same sample at different temperatures and time points, the obtained core parameters, such as diffusion activation energy, can more accurately guide the design of recycling agent dosage and the setting of key process parameters such as mixing temperature and time in actual highway recycling projects. This solves the long-standing problem of the disconnect between conventional laboratory indicators and actual pavement recycling effects, providing a scientific basis for precise asphalt pavement recycling. Attached Figure Description

[0032] Figure 1 This is a flowchart illustrating the overall technical solution according to the present invention; Figure 2 This is a schematic diagram of data flow according to the present invention; Figure 3 A logical flow diagram illustrating the construction of a regenerant containing tracer markers for preparation and diffusion testing samples according to the present invention; Figure 4 This is a flowchart illustrating the process of generating a three-dimensional grayscale cloud map characterizing a diffusion front based on three-dimensional density distribution data, according to the present invention. Figure 5 This is a flowchart illustrating the evaluation of penetration uniformity based on the fractal dimension and interface complexity factor of the diffusion interface according to the present invention. Figure 6 This is a flowchart illustrating how to fit concentration evolution trends and generate diffusion activation energy maps based on cross-time-series scan data according to the present invention. Detailed Implementation

[0033] Example 1: Please refer to the appendix Figure 1 To be continued Figure 6 This embodiment provides a method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator, and its specific implementation process strictly follows the following steps: In the evaluation method, step 1 is performed: preparing a regenerator containing tracer labels. Specifically, in this step, a non-reactive asphalt regenerator is first selected as the matrix material according to experimental requirements. This non-reactive asphalt regenerator typically consists of light oil components, lubricating oil components, and penetration enhancers. To ensure the non-reactive asphalt regenerator is recognizable in subsequent non-destructive imaging, a specific proportion of nanoscale high atomic number tracer particles is added to the non-reactive asphalt regenerator.

[0034] The nanoscale high atomic number tracer particles are preferably nano-bismuth oxide or nano-barium. Nano-bismuth oxide has an atomic number of 83, which is in the very high range and can generate extremely strong X-ray attenuation capabilities. The mass fraction of the nanoscale high atomic number tracer particles is set between 0.5% and 2.5%. This design principle ensures sufficient grayscale contrast in the microscopic scanning image while preventing changes in the dynamic viscosity and chemical activity of the regenerator due to excessive solid particle content. The average particle size of the nanoscale high atomic number tracer particles is controlled within a preset range of 20 to 50 nanometers. This size is much smaller than the gaps between large molecular clusters within asphalt, avoiding sedimentation caused by particle weight and ensuring its Brownian motion characteristics in the viscous asphalt medium.

[0035] During the mechanical shearing process in step 1 above, the mixture is placed in a preset high-temperature environment of 150 to 160 degrees Celsius to reduce the surface tension of the regenerant. A high-speed rotating shear head applies intense hydrodynamic force to the mixture, with the shearing speed set to a first preset value, i.e., 3500 to 4500 revolutions per minute, for a duration of 60 minutes.

[0036] During the shearing process, the initial aggregation state of the tracer particles is broken by the velocity gradient generated by the high-speed flow field. The dispersion coefficient of the nanoscale high atomic number tracer particles is detected using the dynamic light scattering principle. Specifically, a small amount of the mixture is extracted and placed in a cuvette, irradiated with a laser and the scattered signal is received. The particle size distribution curve is calculated using the autocorrelation function to ensure that its dispersion uniformity is higher than a preset mass threshold of 0.95. Before contacting aged asphalt, the prepared regenerant containing the tracer label is placed in a preset vacuum degassing chamber for degassing treatment. The vacuum degree is controlled below -0.09 MPa and maintained for 15 minutes to remove microbubbles introduced during stirring and prevent bubbles from forming a physical barrier at the diffusion interface.

[0037] In the evaluation method, step 2 is performed: constructing diffusion test samples. In the specific implementation of this step, aged asphalt is first prepared. The aged asphalt is prepared using a rotating thin-film oven heating method, where the base asphalt is continuously rotated and heated in an oven at 163 degrees Celsius for 85 minutes to simulate the short-term aging during the production mixing stage; or using a pressure aging vessel aging method, where the asphalt is held at 2.1 MPa and 100 degrees Celsius for 20 hours to simulate the long-term aging stage of the pavement after 5 to 10 years of service.

[0038] The penetration, softening point, and ductility of the aged asphalt need to be adjusted to meet the experimental requirements for a specific degree of aging, for example, the penetration should be in the range of 15 to 30 (0.1 mm). A predetermined amount (e.g., 20 grams) of aged asphalt is injected into a cylindrical aluminum sample container with an inner diameter of 30 mm. A leveling process is then performed by placing the container containing the aged asphalt in a pre-set high-temperature constant temperature chamber at 160 degrees Celsius for 30 minutes, allowing the asphalt's own gravity leveling effect to eliminate air bubbles and uneven structures on the sample surface. After the aged asphalt cools to a predetermined temperature of 25 to 60 degrees Celsius, the regenerator containing tracer markers is uniformly coated on its smooth upper surface, with a coating thickness controlled between 0.5 mm and 1.0 mm, forming a clear contact interface between the regenerator and the aged asphalt.

[0039] In the evaluation method, step 3 is performed: non-destructive radiographic imaging acquisition is carried out. The diffusion test sample is placed in the detection chamber of an industrial micro-computed tomography (CT) scanner. The resolution of this scanner is set to the micrometer range of 2 to 5 micrometers. The scanning parameters are optimized based on the overall density of the diffusion test sample and the mass fraction of tracer particles, with the tube voltage set to 80 kV to 120 kV, the tube current set to 100 μA to 200 μA, and the exposure time set to 500 ms per frame.

[0040] During the scanning process, the temperature inside the detection chamber is maintained at a preset mixing or service temperature by a constant temperature control system to simulate real engineering operating conditions. The constant temperature control system achieves a high precision of 0.1 degrees Celsius, using multi-point distributed thermocouple sensors to monitor the temperature gradient inside the sample in real time. The scanning process employs a continuous rotation mode, acquiring 1440 to 2880 projected images per rotation cycle, with an overlap of over 20%. A precision displacement compensation device installed inside the detection chamber monitors the displacement of the sample due to thermal expansion or creep in real time and feeds it back to the rotary stage driver for micron-level compensation, ensuring that the physical center of the sample remains aligned with the scanning rotation axis throughout the sequential scanning process.

[0041] In the evaluation method, step 4 is performed: constructing a digital twin 3D model. Based on the original projection data obtained in step 3, preprocessing is first performed. Median filtering is used to denoise the image using computer software, setting a 3×3×3 filtering window to traverse the three-dimensional voxels and remove salt-and-pepper noise generated by ray scattering. Subsequently, a 3D reconstruction algorithm is used to reconstruct the spatial distribution morphology of the non-reactive asphalt rejuvenator within the aged asphalt. The 3D reconstruction algorithm employs algebraic reconstruction technology based on iterative reconstruction. When processing low signal-to-noise ratio projection data, it retains the edge details of the diffusion front by introducing a total variational regularization constraint. Its specific implementation logic is as follows: First, define the projection data obtained in step 3 as a vector. The three-dimensional voxel space to be reconstructed is defined as a vector. The reconstruction process involves solving a system of linear equations. The optimal solution, where The system matrix represents the geometric relationship of the projection.

[0042] To improve reconstruction accuracy and suppress noise, this invention employs the following iterative update formula for solving: ; in, Indicates the number of iterations. The relaxation factor (ranging from 0 to 1, with a value of 0.2 in this embodiment) is used to control the convergence speed. and They represent the first The projection values ​​and system matrix row vectors corresponding to each projection ray.

[0043] To preserve edge details of the diffusion front and suppress jagged artifacts caused by noise during iteration, this invention introduces a total variational regularization constraint after each iteration. This total variational regularization is achieved by solving the following optimization problem: ; in, This is a regularization parameter (ranging from 0.0001 to 0.01, and in this embodiment, it is adaptively adjusted according to the noise level). For three-dimensional total variation, its discretized form is defined as: ; in, Spatial index for voxels, It is an extremely small positive number (e.g.) ), used to ensure the differentiability of a function at the point where the gradient is zero.

[0044] The original projection data can be obtained through the iterative algorithm described above. Reconstructed into a three-dimensional voxel matrix Each element in the matrix The magnitude of the value is proportional to the linear attenuation coefficient at the corresponding position. After normalization, this voxel matrix constitutes the three-dimensional grayscale cloud map described in claim 1.

[0045] The construction process involves filtering and deconvolution operations on the original projection data, and establishing a three-dimensional spatial coordinate system using three-dimensional voxel reconstruction technology. Each pixel in the generated three-dimensional grayscale cloud map represents the local density feature of that location. By normalizing the grayscale values, the grayscale range is mapped to between 0 and 1. A linear mapping model is used to convert the grayscale values ​​into the concentration of the regenerant, where the highest grayscale value corresponds to 100% regenerant purity, and the lowest grayscale value corresponds to the aged asphalt matrix. The Laplacian operator is used to sharpen and enhance the diffusion boundary, and the geometric position of the diffusion front is precisely located by calculating the zero point of the second derivative of the local grayscale values.

[0046] In the evaluation method, step 5 is performed: quantifying diffusion morphology features. Diffusion interface information in the 3D grayscale cloud image is extracted using image processing algorithms. Specifically, the fractal dimension is calculated using the cube cover method.

[0047] In a preferred embodiment, the "fractal dimension" mentioned in step 5 is calculated using the cube cover method. First, the diffusion interface within the 3D grayscale cloud image from step 4 is extracted and defined as a 3D point set. This point set This represents the geometry of the diffusion front of the rejuvenator in aged asphalt.

[0048] The calculation process is as follows: First, a series of scale lengths are defined. Its value ranges from the minimum resolution Up to maximum coverage size Increasing in a geometric progression. Among them, Typically, one or two voxel side lengths are chosen. Get point set The projection length is 1 / 5 to 1 / 10 in any dimension.

[0049] For each given scale length Divide the entire three-dimensional space into sections with sides of length . A cubic grid. The statistics contain at least one set of points. The total number of points in the cube grid is denoted as . .

[0050] According to fractal theory, and There exists a power-law relationship: ; Taking the natural logarithm of both sides of the above relation, we obtain a linear equation: ; in, That is, the fractal dimension we are looking for. It is a constant. The least squares method is used to analyze the data points. Perform linear regression to obtain the slope of the regression line. Then the fractal dimension .

[0051] Furthermore, to evaluate the uniformity of penetration, this invention introduces an interface complexity factor. The calculation formula is as follows: ; in, The fractal dimension of an ideal smooth interface is theoretically 2.0 for a two-dimensional plane and between 2.0 and 2.5 for a three-dimensional curved surface. In this embodiment, the ideal value for the three-dimensional diffusion interface is... The default value is 2.1. When The smaller the value, the closer the interface is to an ideal smooth state, and the more uniform the penetration; when If the value exceeds a preset threshold (e.g., 0.2) and a "finger-shaped" protrusion is observed in the 3D cloud map, then a false diffusion phenomenon is determined to exist.

[0052] To evaluate spatial homogeneity, this invention also calculates the spatial heterogeneity of the fractal dimension. Specifically, it calculates the spatial heterogeneity of the diffusion interface point set. In cylindrical coordinates, the circumferential direction is divided into... For each sector, calculate the local fractal dimension of each sector. ( Then, through calculation coefficients of dispersion To evaluate spatial uniformity: ; in, For all The average value, For all The standard deviation of the coefficient of variation. The larger the value, the worse the uniformity of the regenerant flow in the horizontal direction.

[0053] First, a series of scale lengths L are defined, ranging from the scan resolution width to 1 / 10 of the total sample diameter. The extracted diffusion interface is fully covered using cubic units of different scale lengths, and the number N of cubes containing the interface at each scale length is recorded. Linear regression analysis is performed in a double logarithmic coordinate system, with the natural logarithm of the reciprocal of the scale length L as the x-axis and the natural logarithm of the number of cubes containing the interface N as the y-axis. The absolute value of the slope of the resulting regression line is the fractal dimension. Furthermore, the penetration uniformity is evaluated using an interface complexity factor.

[0054] The logic for determining the interface complexity factor is as follows: calculate the difference between the actually measured fractal dimension and the fractal dimension of the ideal smooth interface (value 2.0). When the difference is lower than a first preset threshold (e.g., 0.15), the diffusion process is determined to be uniform penetration; when the difference is higher than a second preset threshold (e.g., 0.45) and the diffusion interface shows finger-shaped or root-like protrusions in the cloud map, a pseudo-diffusion phenomenon is determined, that is, the regenerator only undergoes rapid migration in local large pore channels and does not achieve uniform regeneration throughout the entire area. The analysis also incorporates pore structure characteristics, extracting the micropore distribution pattern inside aged asphalt, analyzing the correlation between the diffusion interface and pore orientation, and determining the contribution ratio of capillary driving effect and molecular diffusion effect.

[0055] In the evaluation method, step 6 is performed: analyzing the diffusion kinetics.

[0056] In a preferred embodiment, the "material diffusion principle" described in step 6 is described using Wiefick's second law. Considering the diffusion process of the rejuvenator in aged asphalt, this embodiment uses the following mathematical model for fitting: ; in, Indicates depth place, time The concentration of the regenerant was obtained by averaging the three-dimensional grayscale cloud map in the depth direction in step 4. Let be the diffusion coefficient, which is assumed to be constant.

[0057] For the finite-thickness diffusion sample constructed in step 2 (thickness is...) The initial conditions are:

[0058] The boundary conditions are: ( (for constant concentration boundary) ( The boundary is adiabatic, meaning the concentration gradient is 0.

[0059] The solution to the above partial differential equation is: ; in, This represents the initial concentration of the surface regenerator. This was determined by monitoring different scanning times. Acquired concentration distribution curve The diffusion coefficient at the current temperature can be obtained by performing a nonlinear least squares fitting. .

[0060] The "diffusion activation energy map" described in step 6 is generated based on the Arrhenius equation. The Arrhenius equation describes the diffusion coefficient. With absolute temperature The relationship between them: ; in, Pre-exponential factor, For diffusion activation energy, It is the ideal gas constant (8.314 J / (mol·K)).

[0061] Taking the natural logarithm of both sides of the above equation, we obtain the linear form: ; Obtain at least three different temperatures through step 6. Diffusion coefficient at the following levels ,by Using the vertical axis as the ordinate, with Plot a scatter plot on the x-axis and perform linear regression. The slope of the regression line is... Therefore, the diffusion activation energy The calculation formula is: ; in, The slope of the regression line. The calculated... Value and temperature By performing correlation, a diffusion activation energy spectrum reflecting the diffusion barrier can be generated. This spectrum can visually demonstrate the magnitude of the energy barrier that regenerator molecules need to overcome for diffusion under different temperature conditions.

[0062] A time series of scans was performed on the same diffusion test sample, covering the initial diffusion stage (first 2 hours), the intermediate stabilization stage (2 to 12 hours), and the late saturation stage (after 12 hours). The time intervals were adaptively adjusted according to the viscosity grade of the asphalt; for high-viscosity aged asphalt, the total monitoring time was increased and the interval between single scans was extended to 4 hours.

[0063] The concentration evolution trend over time at different depths was fitted using the principle of mass diffusion. This principle is expressed using a modified Fick's second law, whereby the first partial derivative of the diffusion concentration with time is equal to the product of the diffusion coefficient and the second partial derivative of the diffusion concentration with spatial displacement. The diffusion coefficient D under different temperature conditions was obtained by integral fitting of the concentration distribution curves at different scanning times.

[0064] Furthermore, a diffusion activation energy spectrum reflecting the diffusion barrier is generated. This spectrum is generated based on the evolution of the Arrhenius equation, determining the activation energy required for the diffusion of rejuvenator molecules by calculating the linear relationship between the natural logarithm of the diffusion coefficient D and the reciprocal of the absolute temperature. The smaller the activation energy value, the easier it is for the non-reactive asphalt rejuvenator to overcome the van der Waals forces and cohesive forces between aged asphalt molecules, resulting in higher rejuvenation efficiency.

[0065] As a supplementary verification method, the evaluation method also includes microhardness testing of the aged asphalt after the rejuvenator diffusion. Through nanoindentation testing, test points are selected in both the reconstructed diffusion and non-diffusion regions to measure their mechanical modulus and hardness. The micromechanical data and three-dimensional grayscale distribution data are spatially coupled and analyzed to construct a mapping model between mechanical properties and rejuvenator concentration, verifying the physical consistency between the softening effect and the diffusion depth.

[0066] The evaluation method establishes a multi-dimensional evaluation index system, which includes average diffusion depth, standard deviation of bulk density distribution, interface roughness coefficient, and concentration gradient attenuation constant. By assigning different weight coefficients to each index and summing them, a comprehensive performance score for a specific non-reactive asphalt recycling agent is obtained. This comprehensive performance score is directly applied to optimize construction process parameters. By comparing the kinetic parameters under different preheating temperatures and spraying pressures, recommended values ​​for mixing time and temperature to achieve the best recycling effect are determined.

[0067] Example 2: Based on Example 1, this example adjusts and applies specific parameters to evaluate the diffusion effect of a high-viscosity modified aged asphalt, in order to further demonstrate the applicability and reliability of this evaluation method.

[0068] In step 1, for high-viscosity media, the addition ratio of nanoscale high atomic number tracer particles is finely adjusted to 1.8% by mass. Nanoscale bismuth oxide, surface-modified with a silane coupling agent, is selected to enhance the affinity of the tracer particles for light oil components in the non-reactive regenerator, preventing localized agglomeration in high-viscosity environments. The mechanical shearing speed is increased, and the shearing time is extended.

[0069] In step 2, the aged asphalt is high-viscosity modified asphalt that has undergone long-term aging with PAV. To ensure the smoothness of the interface, the leveling treatment temperature is increased and the settling time is extended. Before applying the recycling agent, the temperature distribution on the surface of the aged asphalt is monitored using an infrared thermometer to ensure that the surface temperature difference is less than 0.5 degrees Celsius, in order to eliminate the uneven diffusion caused by the Marangoni effect due to temperature difference.

[0070] In the non-destructive imaging acquisition stage of step 3, considering the scattering effect of polymer modifiers in high-viscosity modified asphalt on X-rays, the tube voltage of the industrial micro-computed tomography (CT) equipment was increased, and a copper filter was added to harden the X-ray energy spectrum and reduce energy spectrum artifacts. The scanning time sequence was set to 0 hours, 1 hour, 4 hours, 8 hours, 16 hours, 32 hours, and 64 hours. The constant temperature control system simulated the high-temperature service conditions of the road surface in summer, keeping the temperature of the testing chamber constant at 70 degrees Celsius.

[0071] In step 4, during the construction of the digital twin 3D model, a denoising module based on nonlocal means was added to the 3D reconstruction algorithm to address the complex phase structure of the modified asphalt. Before voxel reconstruction, this denoising module performs a weighted average by searching for similar structural blocks in the image, which can suppress structural background noise generated by polymer particles while preserving the diffusion front edge. Gray-level normalization processing employs piecewise linear mapping, mapping the aged asphalt matrix, the rejuvenator saturation region, and the diffusion transition region to different gray-level threshold ranges.

[0072] In the quantitative evaluation of step 5, spatial heterogeneity analysis was introduced based on fractal dimension calculation to address the potential filamentous diffusion characteristics of high-viscosity modified asphalt. By slicing the three-dimensional diffusion interface into regions, the local fractal dimension was calculated every 100 micrometers along the axis perpendicular to the diffusion direction. The calculations revealed that in the early stages of diffusion, the local fractal dimension exhibited significant anisotropy in both the horizontal and vertical directions, indicating that the rejuvenator preferentially penetrates along the interface between the modifier and the asphaltene. The spatial distribution cloud map of the interface complexity factor clearly identified the preferred diffusion path behavior in the high-viscosity system.

[0073] In the kinetic analysis of step 6, a viscosity-diffusion coupling model was established using the fitted diffusion coefficient and the rheological parameters of the high-viscosity system. Analysis revealed that the diffusion activation energy spectrum exhibits a stepped distribution at different depths. This indicates that in the complex network structure of modified asphalt, the rejuvenator molecules need to overcome a higher energy barrier. This embodiment, through evaluation, suggests that for this type of high-viscosity aged asphalt, it is recommended to extend the mixing time or increase the spraying pressure during construction to ensure that the rejuvenator can penetrate the polymer network and enter the asphalt core.

[0074] Example 3: This example describes the integrated application of this evaluation method on an automated testing platform, demonstrating the fully automated processing logic from data acquisition to report generation.

[0075] In the evaluation method, the raw projection data acquired in step 3 is transmitted in real time to the back-end high-performance computing workstation via 10 Gigabit Ethernet. The automated software platform first retrieves calibration parameters stored in a preset database to automatically correct the ring artifacts caused by ray hardening during the scanning process. The correction logic employs a mean balancing algorithm under polar coordinate transformation, which restores the true contrast of the image by identifying and subtracting abnormal pixel stripes that are symmetrical about the rotation axis.

[0076] In the automated reconstruction process of step 4, the software employs a graphics processor-based algebraic reconstruction acceleration engine. During voxel reconstruction, the system automatically identifies the boundary coordinates of the sample container and automatically removes interfering data from the container wall using a cylindrical mask algorithm, reconstructing only the internal asphalt-recycling system. After reconstruction, the system automatically performs grayscale histogram analysis and uses the Otsu method to automatically find the grayscale segmentation threshold between aged asphalt and recycling agent, achieving automatic extraction of the diffusion region.

[0077] In the feature quantization stage of step 5, the software automatically performs multi-scale cube coverage operations. To improve computational efficiency, an octree data structure is used to recursively partition the 3D space, and only nodes containing interface pixels are refined during scanning. The linear regression analysis process is completed automatically by the algorithm, and the validity of the fractal dimension is determined based on the goodness-of-fit coefficient. The system automatically compares the calculated interface complexity factor with a pre-stored industry standard threshold; if it is determined to be a false diffusion, a red warning is issued in the evaluation report.

[0078] In the dynamic evaluation of step 6, the automated platform performs spatiotemporal registration on the multi-time-segment scan results. Before performing the cross-time-segment sequence analysis in step 6, this invention employs a rigid registration algorithm based on mutual information for spatiotemporal registration. Let the reference image... Initial time 3D grayscale cloud map, image to be registered For a moment The goal of registration is to find the spatial transformation. , making After transformation and To maximize mutual information. The mutual information... Defined as: ; in, The information entropy of an image, Represents the joint entropy. The spatial transformation... It is a rigid transformation, and its parameters include translations along the three coordinate axes. and rotation angle The parameters are iteratively optimized using gradient descent, such that... To reach the maximum value, spatial alignment of 3D images at different times is achieved.

[0079] In the image preprocessing stage, three-dimensional median filtering is used to denoise the original projection data. For any voxel point in the three-dimensional image... Its filtered grayscale value Defined as the median of the gray values ​​of all voxels within the surrounding window of the given voxel, the mathematical expression is: ; in, The original grayscale value. In this embodiment, the radius of the filtering window is... The value can be 1 or 2, corresponding to a window size of 3×3×3 or 5×5×5. This process can effectively remove salt-and-pepper noise generated by ray scattering while preserving the edge information of the diffusion interface.

[0080] Because slight thermal expansion of the samples during long-term scanning may cause coordinate shifts, the system employs a rigid registration algorithm based on mutual information. Using the geometric features of the aged asphalt surface as a reference, it aligns the 3D cloud images at different times in 3D space. After alignment, concentration-time curve clusters are automatically extracted along the depth direction. The diffusion coefficient is then calculated by automatically fitting the modified Fick's second law equation using the nonlinear least squares method.

[0081] Based on a comprehensive performance scoring model, the system automatically summarizes average diffusion depth, diffusion rate, fractal uniformity, and activation energy level, generating an evaluation report that includes a 3D dynamic evolution video, a full data table, and engineering recommendations. This evaluation report can be directly exported as a portable document and sent to the engineering quality management terminal in real time via a network interface. By comparing the diffusion kinetics spectra of different batches of regenerants, real-time monitoring and automated incoming inspection of non-reactive regenerant products can be achieved.

[0082] Example 4: This example details the specific software implementation logic of fractal dimension calculation in the evaluation method and the spatial evolution analysis of interface complexity factor, further refining the technical details of step 5.

[0083] In step 5, the image processing module first receives the 3D voxel matrix reconstructed in step 4. This 3D voxel matrix has a resolution of 1024×1024×1024. First, the algorithm uses the 3D Sobel operator to calculate the spatial gradient field, accurately extracting the 3D point cloud data of the diffusion interface by finding the maximum gradient magnitude. To remove background noise points, a gradient threshold is set, retaining only points with a gradient intensity greater than 20% of the maximum intensity value.

[0084] The parallel computation process for fractal dimension is then initiated. The cube sequence scale L is defined, with values ​​of 2 pixels, 4 pixels, 8 pixels, 16 pixels, 32 pixels, 64 pixels, and 128 pixels. During computation, multiple processing threads are activated. For each given value of L, the entire 3D scan space is divided into a non-overlapping cube mesh with side length L. The statistical logic is as follows: for each mesh, if it contains point cloud data of at least one diffusion interface, the counter N is incremented by 1.

[0085] After completing the statistical analysis at all scales, the algorithm enters the regression analysis phase. The system constructs two sets of sequences in memory: the first set is the natural logarithm of the reciprocal of the scale (logarithm); the second set is the natural logarithm of the corresponding count value (logarithm N). The least squares regression algorithm is used to calculate the linear correlation between these two sets of sequences. The slope of the regression equation is the fractal dimension of the diffusion interface. To ensure the scientific validity of the results, the system automatically calculates the square of the regression coefficients. If this square value is lower than 0.98, it indicates that the diffusion interface does not possess typical fractal characteristics. In this case, the algorithm automatically adjusts the range of values ​​for scale L and restarts the calculation.

[0086] The quantification of the interface complexity factor also considers the spatial heterogeneity of fractal dimension. The software radially segments the three-dimensional diffusion interface, dividing it into sectors every 30 degrees with the container's central axis as the origin. The local fractal dimension within each sector is calculated. By calculating the dispersion coefficients of these local fractal dimensions, a spatial uniformity evaluation index is obtained. If the dispersion coefficient is greater than a first preset fluctuation threshold, even if the overall fractal dimension is low, it is still determined that the recycler exhibits uneven flow in the horizontal direction. This refined quantitative analysis can capture the flow behavior of the recycler within the asphalt caused by the local aggregation of aggregates or modifiers, providing more accurate data support for evaluating full-domain recycling.

[0087] Example 5: This example focuses on the textual description of the physical transformation logic and the principle of matter diffusion in step 6 regarding the generation of the diffusion activation energy spectrum.

[0088] During the kinetic analysis in step 6, the system first acquires scan data sequences under at least three different isothermal environments, such as scan sequences performed at 45°C, 60°C, and 75°C. For each temperature point, the system extracts data on the change in diffusion depth over time. The application of the mass diffusion principle follows a modified Fick's second law, which, in written terms, states that within any given small spatial unit of aged asphalt, the rate of change of rejuvenator concentration depends on the spatial rate of change of the concentration gradient at that site.

[0089] In the specific calculation process, the algorithm performs lateral averaging on the 3D grayscale cloud image at each time step to obtain the concentration distribution function that varies with depth Z. By fitting the evolution of this concentration distribution function over time with a partial differential equation, the effective diffusion coefficient D at the corresponding temperature is solved. After obtaining the diffusion coefficients at different temperatures, the system initiates the energy evaluation logic based on Arrhenius's law.

[0090] The activation energy spectrum is generated as follows: The system uses the reciprocal of the Kelvin temperature as the independent variable and the natural logarithm of the diffusion coefficient D as the dependent variable, performing a linear fit in a Cartesian coordinate system. According to the theory of molecular thermal motion, the slope of this linear relationship is directly related to the diffusion activation energy. The system multiplies the absolute value of the slope by the ideal gas constant to obtain the final diffusion activation energy Ea. This value represents the minimum energy required for regenerator molecules to overcome intermolecular adsorption energy and undergo positional transitions from the coating layer into the aged asphalt.

[0091] In the evaluation report, the diffusion activation energy spectrum is visualized as a thermodynamic contour map. The horizontal axis of the contour map represents the aging time of the aged asphalt, and the vertical axis represents the component ratio of the rejuvenator. By observing the energy troughs in the spectrum, engineers can directly identify the optimal applicable temperature range for the non-reactive rejuvenator. For example, if the spectrum shows a sudden change in the slope of the activation energy curve in the 60°C to 80°C range, it indicates that the asphalt has undergone a viscosity-temperature transition within this temperature range, representing the optimal window for rejuvenator penetration. This evaluation method, based on kinetic principles, avoids the problem of traditional macroscopic indicators being affected by environmental fluctuations.

[0092] Example 6: This example describes a targeted treatment scheme for the evaluation method of the present invention in a complex multiphase asphalt system.

[0093] In step 1, considering that the X-ray absorption characteristics of rubber powder particles may overlap with those of the tracer particles, barium nanoparticles with higher atomic numbers and specific K-edge absorption characteristics were selected as the tracer label. The mass fraction of the tracer particles was increased to enhance the signal intensity against a high scattering background.

[0094] In step 3, a dual-energy CT scanning mode was used to distinguish between rubber particles and the diffusion front. This involved scanning at the same location using both 80 kV and 140 kV tube voltages. By utilizing the differences in the attenuation coefficients of different substances under different energy spectra, a dual-energy decomposition algorithm was used to extract the density distribution map of pure tracer particles, thus eliminating the interference of rubber particles on the diffusion depth assessment.

[0095] In the quantitative evaluation of step 5, considering the obstructive effect of rubber particles on the diffusion path, a tortuosity coefficient was introduced into the fractal dimension calculation model. The ratio of the actual diffusion path length to the projected straight-line length was calculated by extracting the diffusion interface morphology around the rubber particles. The interface complexity factor was modified to be the product of the fractal dimension and the tortuosity coefficient, used to characterize the flow-around ability of the regenerator in the particulate-reinforced phase medium. This method can accurately reflect the actual penetration performance of non-reactive regenerators in modern high-performance pavement materials.

[0096] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the diffusion effect of a non-reactive asphalt recycling agent, characterized in that, Includes the following steps: Step 1: Prepare a regenerator containing tracer labels. Add nano-sized high atomic number tracer particles to a non-reactive asphalt regenerator. Through mechanical shearing, the nano-sized high atomic number tracer particles are uniformly distributed within the non-reactive asphalt regenerator, ensuring that the migration behavior of the nano-sized high atomic number tracer particles is consistent with the molecular motion characteristics of the non-reactive asphalt regenerator. Step 2: Construct a diffusion test sample. Pour aged asphalt into a sample container and level it. After the aged asphalt cools to a predetermined temperature, coat its upper surface with the regenerator containing the tracer marker to form a contact interface between the regenerator and the aged asphalt. Step 3: Perform non-destructive imaging acquisition. Place the diffusion test sample in the detection chamber of the industrial micro-computed tomography equipment. Utilize the difference in X-ray absorption coefficient between the nanoscale high atomic number tracer particles and aged asphalt to obtain the three-dimensional density distribution data of the diffusion test sample at different diffusion time points. Step 4: Construct a digital twin 3D model. Based on the 3D density distribution data, use a 3D reconstruction algorithm to restore the spatial distribution of the non-reactive asphalt rejuvenator inside the aged asphalt and generate a 3D grayscale cloud map to characterize the propagation process of the diffusion front. Step 5: Quantify diffusion morphology characteristics, extract diffusion interface information from the three-dimensional grayscale cloud map using image processing algorithms, calculate the fractal dimension of the diffusion interface, and evaluate the penetration uniformity of the non-reactive asphalt rejuvenator inside the aged asphalt through the interface complexity factor. Step 6: Analyze the diffusion kinetics, perform a time-series scan on the same diffusion test sample, and fit the concentration evolution trend of different depth layers over time in combination with the principle of material diffusion to generate a diffusion activation energy spectrum that reflects the diffusion barrier.

2. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, In the process of preparing the regenerator containing tracer labels, the nanoscale high atomic number tracer particles are selected as nano-bismuth oxide or nano-barium agents. The atomic number of the nanoscale high atomic number tracer particles is in the high range, which is used to generate contrast increment during micro-computed tomography. The average particle size of the nanoscale high atomic number tracer particles is controlled within a preset range to avoid sedimentation caused by particle weight and to ensure its Brownian motion characteristics in the viscous asphalt medium. The mass fraction of the nanoscale high atomic number tracer particles in the non-reactive asphalt rejuvenator is set within a preset range to ensure grayscale recognition in the microscopic scanning image without changing the dynamic viscosity and chemical activity of the non-reactive asphalt rejuvenator.

3. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, The mechanical shearing process includes: placing the mixed system in a preset high-temperature environment, applying hydrodynamic force to the mixed liquid using a high-speed rotating shear head, setting the shearing speed to a first preset value, and using the velocity gradient generated by the high-speed flow field to break the initial aggregation state of the nanoscale high atomic number tracer particles. Subsequently, the dispersion coefficient of the nanoscale high atomic number tracer particles was detected using the dynamic light scattering principle, and the particle size distribution curve was calculated using the autocorrelation function to ensure that its dispersion uniformity is higher than the preset mass threshold. Before the regenerant containing the tracer mark comes into contact with the aged asphalt, it is placed in a preset vacuum degassing chamber for degassing treatment. The vacuum level is controlled below the preset vacuum level to remove the tiny air bubbles introduced during the stirring process and prevent the air bubbles from forming a physical barrier at the contact interface.

4. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, In the process of constructing diffusion test samples, the aged asphalt is prepared by rotating thin film oven heating method or pressure aging container aging method to simulate different aging stages after the road surface has been in service. The penetration, softening point and ductility of the aged asphalt are adjusted to meet the experimental requirements of the degree of aging. The leveling process is achieved by placing a container filled with aged asphalt in a pre-set high-temperature constant temperature chamber, utilizing the gravity leveling effect of the aged asphalt itself to eliminate air bubbles and uneven structures on the sample surface. Before applying the regenerator containing tracer markers, the temperature distribution on the surface of the aged asphalt is monitored using an infrared thermometer to ensure that the surface temperature difference is less than a preset fluctuation threshold, thereby eliminating the Marangoni effect caused by temperature difference.

5. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, During the non-destructive imaging acquisition process, the resolution of the industrial micro-computed tomography device is set at the micrometer level, and the scanning parameters include tube voltage, tube current and exposure time. These parameters are configured according to the overall density of the diffusion test sample and the mass fraction of the nanoscale high atomic number tracer particles. During the scanning process, the temperature inside the detection chamber is maintained at a preset mixing temperature or service temperature by a constant temperature control system. The temperature control accuracy of the constant temperature control system reaches a preset accuracy level, and the temperature gradient inside the diffusion test sample is monitored in real time by thermocouple sensors distributed at multiple points. The scanning process adopts a continuous rotation mode, and a predetermined number of projected images are acquired within a single rotation cycle. The overlap of the projected images is set to be above a preset ratio. By installing a precision displacement compensation device in the detection chamber, the displacement of the diffusion test sample caused by thermal expansion or creep is monitored and fed back to the rotary stage driver for compensation, ensuring that the physical center of the diffusion test sample coincides with the scanning rotation axis during the sequential scanning process.

6. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, In the process of constructing a digital twin 3D model, the original projection data is first processed by median filtering to remove noise using image processing algorithms. Then, the 3D voxels are traversed using a filter window of a preset size to remove salt-and-pepper noise generated by ray scattering. The three-dimensional reconstruction algorithm adopts an algebraic reconstruction technique based on iterative reconstruction. When processing low signal-to-noise ratio projection data, it introduces a total variational regularization constraint to preserve the edge details of the diffusion front. The reconstruction process involves filtering and deconvolution operations on the original projection data, and establishing a spatial coordinate system through three-dimensional voxel reconstruction technology; Each pixel in the three-dimensional grayscale cloud map represents the local density feature of that coordinate point. By normalizing the grayscale values, the grayscale range is mapped to a preset numerical range, and the grayscale values ​​are converted into the concentration of the non-reactive asphalt recycling agent using a linear mapping model. The diffusion boundary is sharpened and enhanced using the Laplacian operator, and the geometric position of the diffusion front is locked by calculating the zero point of the second derivative of the local gray value.

7. The method for evaluating the diffusion effect of a non-reactive asphalt recycling agent according to claim 1, characterized in that, The fractal dimension is calculated using a cube covering method, the specific steps of which include: setting a series of scale lengths and using cube units of different scale lengths to fully cover the extracted diffusion interface; During the calculation process, the three-dimensional space is recursively partitioned using an octree data structure, and the number of cubes containing the diffusion interface is counted at each scale length. In a double logarithmic coordinate system, a linear regression analysis is performed with the natural logarithm of the reciprocal of the scale length as the abscissa and the natural logarithm of the number of cubes containing the diffusion interface as the ordinate. The absolute value of the slope of the resulting regression line is the fractal dimension. The system automatically calculates the squared value of the regression coefficient. If the squared value is lower than the preset correlation threshold, the system automatically adjusts the range of the scale length and recalculates.

8. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, The logic for determining the interface complexity factor includes: calculating the difference between the actually measured fractal dimension and the reference value of the fractal dimension of the ideal smooth interface; When the difference is lower than the first preset threshold, the diffusion process is determined to be uniform penetration; When the difference is higher than the second preset threshold and the diffusion interface presents finger-shaped or root-like protrusions in the three-dimensional grayscale cloud image, it is determined that there is a false diffusion phenomenon. The quantitative diffusion morphology features also include calculating the spatial heterogeneity of the fractal dimension, dividing the diffusion interface into sectors, calculating the local fractal dimension in each sector, and obtaining the spatial uniformity evaluation index by calculating the discrete coefficient of the local fractal dimension. If the dispersion coefficient is greater than the preset fluctuation threshold, it is determined that the non-reactive asphalt rejuvenator has uneven flow in the horizontal direction. The evaluation method also incorporates pore structure characteristics, extracting the micropore distribution pattern inside the aged asphalt to analyze the contribution ratio of capillary driving effect and molecular diffusion effect of the non-reactive asphalt rejuvenator in the porous medium.

9. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, In the process of analyzing the diffusion dynamics, the time series of the sequence scan covers the initial stage, the middle stable stage and the late saturation stage of diffusion, and the time interval is adaptively adjusted according to the viscosity grade of the asphalt. The automated platform performs spatiotemporal registration on the scanning results from multiple time periods. It adopts a rigid registration algorithm based on mutual information and uses the geometric features of the aging asphalt bottom surface as a reference to spatially align the three-dimensional grayscale cloud images at different times. The principle of material diffusion is expressed using a modified Fick's second law, which states that the first partial derivative of the concentration of the non-reactive asphalt rejuvenator with respect to time is equal to the product of the diffusion coefficient and the second partial derivative of the concentration with respect to spatial displacement. The diffusion coefficient under different temperature conditions was obtained by integral fitting of the concentration distribution curves at different scanning times.

10. The method for evaluating the diffusion effect of a non-reactive asphalt rejuvenator according to claim 1, characterized in that, The generation of the diffusion activation energy spectrum is based on the evolution of the Arrhenius equation. The activation energy required for the diffusion of the non-reactive asphalt rejuvenator molecules is determined by calculating the linear relationship between the natural logarithm of the diffusion coefficient and the reciprocal of the absolute temperature. The activation energy is obtained by multiplying the absolute value of the slope of the linear relationship by the ideal gas constant. The smaller the activation energy value, the easier it is for the non-reactive asphalt rejuvenator to overcome the cohesive forces between the aged asphalt molecules. The evaluation method also includes conducting microhardness tests on the recycled aged asphalt, measuring its mechanical modulus and hardness index through nanoindentation tests, and performing spatial coupling analysis between the micromechanical data and the three-dimensional density distribution data to construct a mapping model between mechanical properties and rejuvenator concentration. A multi-dimensional evaluation index system is established, which includes the average diffusion depth, the standard deviation of the volume density distribution, the interface roughness coefficient, and the attenuation constant of the concentration gradient. The comprehensive performance score is obtained by weighting and summing the various indicators, and the construction process parameters are optimized based on the score.

Images