A multi-scale evaluation method for diffusion behavior of fusion interface based on multiple regeneration processes of new and old asphalt

Abstract

The application discloses a kind of fusion interface diffusion behavior multiscale evaluation method based on new and old asphalt multiple regeneration process, it is related to bituminous regeneration technical field, comprising: to the fusion interface diffusion behavior data is carried out feature screening and constructs fusion interface diffusion behavior evaluation base;Based on fusion interface diffusion behavior evaluation base fitting fusion interface diffusion behavior deep feature factor layer, obtain deep feature evaluation factor group and calculate multi-factor comprehensive weight and the trustworthiness of each evaluation factor in deep feature evaluation factor group, and based on trustworthiness constructs trustworthiness matrix;Multi-factor comprehensive weight and trustworthiness matrix are carried out random forest model regression analysis, obtain final evaluation model, according to final evaluation model carries out fusion interface diffusion behavior multiscale evaluation.The application is quantitatively characterized by multiple scale test method, the diffusion behavior of new asphalt and different degree of aging asphalt, to lay theoretical and experimental foundation for subsequent multiple regeneration asphalt mixture design and its performance optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of asphalt regeneration, and more particularly to a multi-scale evaluation method for diffusion behavior of a fusion interface based on multiple regeneration processes of new and old asphalt. Background Art

[0002] Multiple asphalt pavement recycling refers to a paving technology that recycles an already recycled pavement a second or even more times after it reaches the end of its service life. In recent years, as recycled asphalt pavements constructed earlier in my country have reached, or are nearing, the end of their service life, the ability to recycle them multiple times has become a key scientific and technical challenge facing the development of recycling technology.

[0003] Compared to single-stage recycling, multiple asphalt pavement recycling differs in that the asphalt coated on the old aggregate is aged to varying degrees. This means that diffusion occurs between the new and aged asphalt during the recycling process. Under certain conditions, the diffusion behavior at the interface between the new and old asphalt directly impacts the performance and service life of the recycled asphalt pavement. Furthermore, as the asphalt ages, the diffusion problem caused by partial mixing becomes more pronounced. Therefore, it is necessary to study the more complex diffusion behavior during multiple asphalt recycling. This will lay a theoretical and experimental foundation for the design and performance optimization of multi-recycled asphalt mixtures.

[0004] Current research on multiple asphalt pavement recycling primarily focuses on evaluating the overall macroscopic performance of asphalt and its mixture, analyzing how the properties of the recycled asphalt and its mixture change with increasing recycling cycles. While asphalt recovers to the performance level of the original asphalt after each aging and recycling cycle, asphalt mixtures experience a rapid decline in low-temperature performance and a logarithmic increase in high-temperature performance with increasing recycling cycles. This macroscopic performance evaluation system demonstrates the feasibility of multiple recycling technology, but it focuses primarily on analyzing test results and fails to explain the fundamental causes of performance evolution. It also overlooks the unique characteristics of multiple recycling, namely, the varying aging cycles of the asphalt attached to the old aggregate, which results in diffusion between new and aged asphalt during the recycling process. Under certain conditions, the diffusion behavior at the interface between new and old asphalt directly affects the performance and service life of the recycled asphalt pavement. Furthermore, as the asphalt ages, the diffusion problem caused by partial mixing becomes more pronounced. Therefore, it is necessary to investigate the more complex diffusion behavior of asphalt during multiple recycling processes. This will provide a theoretical and experimental foundation for the design and performance optimization of multiple recycled asphalt mixtures.

[0005] Therefore, how to propose a multi-scale evaluation method for the diffusion behavior of the fusion interface based on the multiple regeneration process of new and old asphalt, and use multi-scale experimental methods to quantitatively characterize the diffusion behavior of the fusion interface between new asphalt and asphalt of different degrees of aging during multiple regeneration processes, so as to lay a theoretical and experimental foundation for the subsequent design of multiple recycled asphalt mixtures and their performance optimization is an urgent problem that technical personnel in this field need to solve. Summary of the Invention

[0006] In light of this, the present invention provides a multi-scale evaluation method for the diffusion behavior of the fusion interface between new and old asphalt during multiple regeneration processes. This multi-scale test method quantitatively characterizes the diffusion behavior of the fusion interface between new asphalt and asphalt of varying degrees of aging during multiple regeneration processes, laying a theoretical and experimental foundation for the subsequent design and performance optimization of multiple-regeneration asphalt mixtures. To achieve the above objectives, the present invention employs the following technical solutions:

[0007] A multi-scale evaluation method for diffusion behavior of the fusion interface based on multiple regeneration processes of new and old asphalt, including:

[0008] Perform feature screening on the diffusion behavior data of the fusion interface, and construct a fusion interface diffusion behavior evaluation basis based on the screened features;

[0009] Based on the fusion interface diffusion behavior evaluation base, the deep characteristic factor layer of the fusion interface diffusion behavior is fitted to obtain the deep characteristic evaluation factor group;

[0010] Calculate the comprehensive weight of multiple factors based on the deep feature evaluation factor group, calculate the trustworthiness of each evaluation factor in the deep feature evaluation factor group, and construct a trust matrix based on the trustworthiness;

[0011] The multi-factor comprehensive weights and trust matrix were subjected to random forest model regression analysis to obtain the final evaluation model, and a multi-scale evaluation of the diffusion behavior of the fusion interface was performed based on the final evaluation model.

[0012] Optionally, the feature screening of the fusion interface diffusion behavior data includes: using the maximum information coefficient MIC to perform feature screening and remove irrelevant factors.

[0013] Optionally, the fusion interface diffusion behavior evaluation substrate includes substrate factors, and the substrate factors are macro-scale evaluation, meso-scale evaluation, micro-scale evaluation and molecular-scale evaluation.

[0014] Optionally, the deep characteristic factor layer includes beam body factors, and the beam body factors are diffusion depth and average diffusion rate.

[0015] Optionally, the macro-scale evaluation includes: evaluating the macro-performance of the fusion of new and old asphalt by measuring the change values ​​of the three major indicators of asphalt, evaluating the stability of the asphalt structure by turbidity titration, and dynamic shear rheology testing to obtain viscosity, complex shear modulus, storage modulus, loss modulus and phase angle test results.

[0016] Optionally, the diffusion depth and average diffusion rate include:

[0017] Fluorescence microscopy was used to obtain images, and Image Pro Plus image processing software was used to quantitatively calculate the area and grayscale value of the gray shift region in the image to obtain the diffusion depth.

[0018] Tracer elements are added to the new asphalt, and the distribution map of the tracer elements is obtained by scanning with an energy spectrometer. The diffusion depth and the affected area of ​​the new asphalt during each fusion process with the old asphalt are dynamically tracked.

[0019] A scanning electron microscope combined with an energy dispersive spectrometer was used to analyze the composition of different intervals of the fusion of new and old asphalt;

[0020] Using the peak force tapping mode of an atomic force microscope, the changes in the two-dimensional and three-dimensional micromorphology of different sections during the fusion process of new and old asphalt were observed. The roughness, Young's modulus, and adhesion micromechanical parameters were collected. The Hertz contact model and the DMT model were used to fit the parameters and the effective modulus of the interface between the new and old asphalt was calculated for each fusion.

[0021] By setting a specific time interval, the average diffusion rate is obtained and the speed of diffusion under different conditions is evaluated.

[0022] Optionally, the method of fitting the fused interface diffusion behavior deep feature factor layer based on the fused interface diffusion behavior evaluation base to obtain the deep feature evaluation factor group includes: dividing the base factors of the fused interface diffusion behavior evaluation base into multiple beam body factors according to the K-means clustering method to construct a fused interface diffusion behavior deep factor layer; and sequentially composing the multiple beam body factors included in the fused interface diffusion behavior deep factor layer into a deep feature evaluation factor group.

[0023] Optionally, the multi-factor comprehensive weights and trust matrix are subjected to random forest model regression analysis to obtain a final evaluation model including: the random forest is composed of M decision trees, and during training, the M decision trees are trained in a cycle, and the training samples of each decision tree are obtained by Bootstrap sampling from the original training set, and the features used in training each node of the decision tree are also obtained by random sampling from the new feature space N, and each decision tree is recursively split according to the judgment criteria. After the training of the M decision trees is completed, the mean of each leaf node is the evaluation result of the final regression.

[0024] Optionally, each decision tree is recursively split according to the judgment criteria, including: the extracted sample set constitutes the root node, and according to the judgment criteria, the sample set is split into a first sub-sample and a second sub-sample; the left subtree is recursively established with the first sub-sample, and the right subtree is established with the second sub-sample; and the condition for stopping the splitting is set. When the splitting cannot continue, the node is marked as a leaf node and a value is assigned at the same time.

[0025] It can be seen from the above technical solution that, compared with the prior art, the present invention discloses a multi-scale evaluation method for diffusion behavior of the fusion interface based on the multiple regeneration process of new and old asphalt, which has the following beneficial effects:

[0026] This invention proposes a multi-scale evaluation method for the diffusion behavior of the fusion interface based on the multiple recycling process of new and old asphalt. The method comprises the following steps: screening the diffusion behavior data of the fusion interface, constructing a diffusion behavior evaluation base based on the screened features; fitting a deep characteristic factor layer of the fusion interface based on the diffusion behavior evaluation base to obtain a deep characteristic evaluation factor cluster; calculating the multi-factor comprehensive weights based on the deep characteristic evaluation factor cluster, calculating the confidence of each evaluation factor in the deep characteristic evaluation factor cluster, and constructing a confidence matrix based on the confidence; subjecting the multi-factor comprehensive weights and the confidence matrix to random forest regression analysis to obtain a final evaluation model, and performing a multi-scale evaluation of the diffusion behavior of the fusion interface based on the final evaluation model. This invention conducts preliminary performance evaluation and partial micro-scale experimental research on multiple recycled asphalt. Focusing on the more complex diffusion behavior of the new and old asphalt fusion interface during multiple recycling processes, the method quantitatively characterizes the diffusion behavior of new asphalt and asphalt of varying degrees of aging. Using a multi-scale experimental method, the method quantitatively characterizes the diffusion behavior of the fusion interface between new asphalt and asphalt of varying degrees of aging during multiple recycling processes, laying a theoretical and experimental foundation for the subsequent design and performance optimization of multiple recycled asphalt mixtures. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0028] Figure 1 A schematic flow chart of a multi-scale evaluation method for diffusion behavior of a fusion interface based on multiple regeneration processes of new and old asphalt provided by the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0030] The embodiment of the present invention discloses a multi-scale evaluation method for diffusion behavior of the fusion interface based on the multiple regeneration process of new and old asphalt, such as Figure 1 Shown, including:

[0031] Perform feature screening on the diffusion behavior data of the fusion interface, and construct a fusion interface diffusion behavior evaluation basis based on the screened features;

[0032] Based on the fusion interface diffusion behavior evaluation base, the deep characteristic factor layer of the fusion interface diffusion behavior is fitted to obtain the deep characteristic evaluation factor group;

[0033] Calculate the comprehensive weight of multiple factors based on the deep feature evaluation factor group, calculate the trustworthiness of each evaluation factor in the deep feature evaluation factor group, and construct a trust matrix based on the trustworthiness;

[0034] The multi-factor comprehensive weights and trust matrix were subjected to random forest model regression analysis to obtain the final evaluation model, and a multi-scale evaluation of the diffusion behavior of the fusion interface was performed based on the final evaluation model.

[0035] Furthermore, the feature screening of the fusion interface diffusion behavior data includes: using the maximum information coefficient MIC to perform feature screening and remove irrelevant factors.

[0036] In a specific embodiment, the feature screening using the maximum information coefficient MIC to remove irrelevant factors includes: calculating the correlation coefficient Ω=(Ω1, Ω2, ..., Ω n ), the maximum value of the correlation coefficient Ω max The corresponding feature is selected as the first feature, assuming q1 = Y k ; Then calculate the MIC value M between feature q1 and other features = (m1, ..., m k-1 , m k+1 ,...,m n ), let P=0.5*Ω(i≠k)+0.5*(1-M), and change P maxThe corresponding feature is selected as the second feature f2; the first feature is removed and the above steps are repeated until a sufficient number of features are obtained or the maximum value of the current MIC is less than a certain threshold. Based on the macro-scale evaluation, meso-scale evaluation, micro-scale evaluation and molecular-scale evaluation, the feature vector of each fusion interface diffusion behavior is expressed as A i =(α1, α2, α 3， α 4， α5), the characteristic space of all fusion interface diffusion behaviors is expressed as A = (A1, A2, ..., A n ) T , where n is the number of diffusion behaviors at the fusion interface, and the corresponding diffusion behavior at the fusion interface is expressed as γ = (γ1, γ2, ..., γ n ) T ,First, all the features are normalized for variance, and then the relationship between each factor and the evaluation results is analyzed.

[0037] Furthermore, the evaluation base of the fusion interface diffusion behavior includes base factors, and the base factors are macro-scale evaluation, meso-scale evaluation, micro-scale evaluation and molecular-scale evaluation.

[0038] In a specific embodiment, mesoscopic evaluation: Fluorescence images of old asphalt containing tracers are analyzed using a fluorescence microscope. Image-ProPlus image analysis software is used to quantitatively analyze the fluorescence images using three optical parameters: mean optical density (MOD), mean gray value (MGV), and integrated optical density (IOD) to assess the degree of fusion between the new and old asphalt. Microscopic evaluation: Fourier transform infrared spectroscopy (FTIR), gel permeation chromatography (GPC), and atomic force microscopy (AFM) are used to study the fusion behavior of the new and old asphalt interfaces from a microstructural perspective, including parameters such as the carboxyl index (I(S=O)), sulfoxide index (I(C=O)), macromolecular size (LMS), and roughness index. Molecular evaluation: Molecular dynamics (MD) and dissipative particle dynamics (DPD) simulations are used to investigate the interfacial interactions between the rejuvenator and old asphalt, as well as between the rejuvenator and old asphalt and aggregate, under warm mix conditions, to assess the fusion between the new and old asphalt at the molecular level.

[0039] Furthermore, the deep characteristic factor layer includes beam body factors, and the beam body factors are diffusion depth and average diffusion rate.

[0040] Furthermore, the macro-scale evaluation includes: evaluating the macro-performance of the fusion of new and old asphalt by measuring the change values ​​of the three major indicators of asphalt, evaluating the stability of the asphalt structure by turbidity titration, and dynamic shear rheology testing, and obtaining the viscosity, complex shear modulus, storage modulus, loss modulus and phase angle test results.

[0041] Furthermore, the diffusion depth and average diffusion rate include:

[0042] Fluorescence microscopy was used to obtain images, and Image Pro Plus image processing software was used to quantitatively calculate the area and grayscale value of the gray shift region in the image to obtain the diffusion depth.

[0043] Tracer elements are added to the new asphalt, and the distribution map of the tracer elements is obtained by scanning with an energy spectrometer. The diffusion depth and the affected area of ​​the new asphalt during each fusion process with the old asphalt are dynamically tracked.

[0044] A scanning electron microscope combined with an energy dispersive spectrometer was used to analyze the composition of different intervals of the fusion of new and old asphalt;

[0045] Using the peak force tapping mode of an atomic force microscope, the changes in the two-dimensional and three-dimensional micromorphology of different sections during the fusion process of new and old asphalt were observed. The roughness, Young's modulus, and adhesion micromechanical parameters were collected. The Hertz contact model and the DMT model were used to fit the parameters and the effective modulus of the interface between the new and old asphalt was calculated for each fusion.

[0046] By setting a specific time interval, the average diffusion rate is obtained and the speed of diffusion under different conditions is evaluated.

[0047] Furthermore, the method of fitting the fused interface diffusion behavior deep feature factor layer based on the fused interface diffusion behavior evaluation base to obtain the deep feature evaluation factor group includes: dividing the base factors of the fused interface diffusion behavior evaluation base into multiple beam body factors according to the K-means clustering method to construct a fused interface diffusion behavior deep factor layer; and sequentially composing the multiple beam body factors included in the fused interface diffusion behavior deep factor layer into a deep feature evaluation factor group.

[0048] Furthermore, it also includes: evaluating factor clusters based on deep features and calculating the comprehensive weights of multiple factors based on game theory, specifically including:

[0049] A total of K different weight determination methods, both subjective and objective, are used to calculate the basic weight vector set of all evaluation factors in the deep feature evaluation factor group; based on the basic weight vector set, a linear combination of weights is obtained by introducing weight coefficients; based on the linear combination of weights and the basic weight vector set, a multi-objective game set model is constructed; based on the multi-objective game set model, the multi-factor comprehensive weight is calculated; based on the differential properties of the matrix, the optimal first-order derivative condition and matrix expression of the multi-objective game set model are obtained; the matrix expression is obtained, the linear combination coefficient is obtained and normalized to obtain the normalized linear combination coefficient; the basic weight vector set and the weight coefficient obtained after the normalization of its linear combination coefficient are multiplied and summed to obtain the final multi-factor comprehensive weight.

[0050] Furthermore, the method further includes: calculating the trustworthiness of each beam factor in the deep feature evaluation factor group, and constructing a fuzzy relationship matrix based on the trustworthiness, specifically including:

[0051] The diffusion behavior of the fusion interface is divided into four evaluation levels: normal, basically normal, subnormal and abnormal. The degree of trust of each beam factor in the deep feature evaluation factor group in the evaluation level is calculated. The size or property of each factor is divided according to the four normality evaluation levels to clarify the trust of each factor. The trust is the degree to which each factor trusts each target state. The qualitative factors are obtained by expert scoring method and empirical method based on confidence index. The quantitative factors are divided into four ranges according to the numerical size and evaluation level. The trapezoidal trust function is selected for calculation. The factors with larger values ​​are better, and the ascending semi-trapezoidal distribution function is used; the factors with smaller values ​​are better, and the descending semi-trapezoidal distribution function is used. The monitoring values ​​of each factor are substituted into the trapezoidal trust function to calculate the degree of trust of each beam factor in the deep feature evaluation factor group in the evaluation level. A fuzzy relationship matrix is ​​constructed based on the trust.

[0052] Furthermore, the random forest model regression analysis of the multi-factor comprehensive weights and the trust matrix is ​​performed to obtain the final evaluation model, which includes: the random forest is composed of M decision trees. During training, the M decision trees are trained in a loop. The training samples of each decision tree are obtained by bootstrap sampling from the original training set. The features used in training each node of the decision tree are also randomly sampled from the new feature space N. Each decision tree is recursively split according to the judgment criteria. After the training of the M decision trees is completed, the mean of each leaf node is the final regression evaluation result. Specifically, bootstrap sampling is to extract n samples with replacement from a set of n samples to form a data set.

[0053] Furthermore, each decision tree is recursively split according to the judgment criteria, including: the extracted sample set constitutes the root node, and according to the judgment criteria, the sample set is split into a first sub-sample H1 and a second sub-sample H2; the first sub-sample H1 is used to recursively establish a left subtree, and the second sub-sample H2 is used to establish a right subtree; and the condition for stopping the splitting is set. When the splitting cannot continue, the node is marked as a leaf node and assigned a value at the same time.

[0054] In a specific embodiment, the implementation method of the decision criterion is as follows: calculate the regression error of the root node, that is, the mean square error of the label value and the regression value of all samples, which is defined as: Based on the multi-factor comprehensive weight and trust matrix, the characteristic vector of the diffusion behavior of each fusion interface is expressed as B i=(β1, β2), the characteristic space of all fusion interface diffusion behaviors is expressed as B = (B1, B2, ..., B j ) T , where j is the number of diffusion behaviors on the fusion interface. A new feature space B is constructed by multi-factor comprehensive weight and trust matrix, and features B are randomly extracted from the new feature space B. t , sort the extracted training samples H according to the value of the feature from small to large; use the evaluation results of the diffusion behavior of each fusion interface as the threshold in turn, divide the samples into left and right parts, and then calculate the mean square error of the left and right subtrees; the error index of the split is defined as the regression error before the split minus the regression error of the left and right subtrees after the split: G = G(H) - G(H1) - G(H2); continue splitting when this index is maximized; when the depth of the decision tree is reached or the calculated regression error is greater than the artificially set threshold, stop splitting and set the node as a leaf node, and the value of the leaf node is the mean of the label value of the sample set of this node; at this point, the training of a single decision tree is completed.

[0055] In a specific embodiment, a multi-scale evaluation method for the diffusion behavior of the fusion interface of new and old asphalt during multiple regeneration processes is developed. For artificially aged asphalt and naturally aged asphalt extracted from on-site milled waste materials, an orthogonal experimental design is used to conduct dynamic shear viscosity, multi-stress creep recovery (MSCR), frequency sweep, time sweep, and temperature sweep tests at the macro level to evaluate the rheological properties of the repeatedly regenerated asphalt during the fusion process. At the micro level, fluorescence microscopy, energy dispersive spectrometer (EDS), scanning electron microscopy (SEM), atomic force microscopy (AFM) and other technical means are used to observe and analyze the micromorphology, diffusion depth and average diffusion rate at the asphalt fusion diffusion interface. The specific steps include:

[0056] (1) Comparison and evaluation analysis of different rheological test indicators;

[0057] (2) Quantitative study of the interface fusion diffusion depth and average diffusion rate between new asphalt and asphalt of different aging degrees under different influencing factors;

[0058] (3) The intrinsic relationship between diffusion depth, average diffusion rate and macroscopic performance indicators is to select the macroscopic rheological test and microscopic technical means with the best correlation.

[0059] Specifically, 1) Pre-preparation of Multiple Recycled Asphalt: In this example, multiple asphalt regeneration processes include single, double, and triple regenerations. The old asphalt in both the primary and secondary regenerations was derived from both artificial aging and natural pavement aging. During the experiment, the results of the fusion of the two different sources of aged asphalt were compared and analyzed. The selection of new and aged asphalt materials is shown in Table 1.

[0060] Table 1 Raw material selection

[0061]

[0062] The control variables in the process of preparing recycled asphalt are shown in Table 2. The second-aged asphalt is obtained by aging the first-aged asphalt after fusion, and the same is true for the third-aged asphalt.

[0063] Table 2 Summary of control variables

[0064]

[0065]

[0066] 2) Comparison and evaluation of different rheological test indicators: The rheological parameters of asphalt materials can reflect their fusion-diffusion behavior, and are mainly used to dynamically monitor and analyze the overall rheological properties of multiple fusions of new and old asphalt.

[0067] This section uses double-layer specimens composed of new and old asphalt (old asphalt on the bottom). A Brookfield viscometer was used to measure the dynamic changes in viscosity during the fusion process of the new and old asphalt at different test temperatures (same mixing temperature), recording viscosity results every 5 seconds until the results stabilized. A dynamic shear rheometer (DSR) was used to perform frequency sweep, time sweep, and temperature sweep tests. Under stress (variable) control mode, rheological parameters such as the complex shear modulus, storage modulus, loss modulus, and phase angle were recorded. The differences in test results during each asphalt fusion process were analyzed.

[0068] 3) Quantitative study of diffusion depth and average diffusion rate: This study mainly defines the diffusion influence area of ​​the new and old asphalt fusion interface during multiple regeneration processes, and calculates the average diffusion rate based on the evolution of the diffusion area during different mixing times, thereby intuitively characterizing the fusion-diffusion behavior.

[0069] A fluorescence microscope was used to observe the migration of the ash area at the interface between the new and old asphalt. Multiple regenerated asphalt samples with a specific stirring time were frozen and then cut at low temperature. The cross-section at the interface between the new and old asphalt was taken as the observation object, and the image under the fluorescence microscope was obtained. Image Pro Plus image processing software was used to quantitatively calculate the area and grayscale value of the ash migration area to obtain the diffusion depth. Element tracing technology was used, and elements that were not present in the asphalt were selected as tracers. This embodiment intends to use cobalt, lead, and titanium as three tracer elements. The tracers are cobalt cyclohexane, lead isooctanoate, and nano-TiO2, which are added to the new asphalt. The distribution map of the tracer elements was obtained by scanning with an energy dispersive spectrometer (EDS), and the diffusion depth of the new asphalt during each fusion process with the old asphalt and the diffusion-affected area were dynamically tracked. A scanning electron microscope (SEM) was used in conjunction with EDS to perform component analysis in different intervals of the fusion of new and old asphalt. With the help of the peak force tapping mode (Peak Force QNM) of atomic force microscopy (AFM), the changes in the two-dimensional and three-dimensional micromorphologies of different sections during the fusion process of new and old asphalt were observed, and micromechanical parameters such as roughness, Young's modulus, and adhesion were collected. The Hertz contact model and the Derjaguin-Muller-Toporov (DM T) model were used for fitting, and the effective modulus of the new and old asphalt interface was calculated for each fusion.

[0070] Specifically, the calculation of the effective modulus includes: The Hertz contact model is suitable for describing the mechanical behavior of two materials when they are in contact, especially in the case of elastic contact. For the interface between new and old asphalt, the formula of the Hertz model can be used: E eff =4 / 3*F / δ 3 *R, where: E eff is the effective modulus, F is the contact force, δ is the contact deformation, and R is the contact radius. The effective modulus is calculated using the Hertz model by experimentally measuring the contact force and contact deformation. The DMT model provides a more accurate description when considering the effects of contact force and surface energy and is suitable for smaller contact areas. The effective modulus calculation formula of the DMT model is: E DMT =3 / 2*(F+2πRγ) / δ 3 , where γ is the surface energy. The remaining variables are the same as above. By comparing the results of the Hertz model and the DMT model, we can gain a more comprehensive understanding of the effective modulus of the new-to-old asphalt interface.

[0071] Using a variety of advanced experimental techniques, the internal microstructure of the fusion of new and old asphalt is studied from the perspectives of components, molecules, and nanometers. The average diffusion rate is obtained at specific time intervals to evaluate the speed of diffusion under different conditions.

[0072] Specifically, in the diffusion experiment design, appropriate experimental methods (such as thin film diffusion method, column diffusion method, etc.) are selected to measure the diffusion characteristics at the interface between new and old asphalt. At different time intervals, the concentration changes of the diffusing substances are recorded; the average diffusion rate is calculated. According to Fick's law, the average diffusion rate can be expressed as: D avg =ΔC / (Δt*A), where ΔC is the concentration change, Δt is the time interval, and A is the diffusion area. This method systematically studies the effective modulus and microstructure of the fused asphalt, while also evaluating the diffusion characteristics under different conditions. This will provide important theoretical basis and experimental support for the optimization and application of asphalt materials.

[0073] 4) The intrinsic relationship between diffusion depth, average diffusion rate and macro-performance indicators: Through macro-performance tests, test results such as viscosity, complex shear modulus, storage modulus, loss modulus, and phase angle are obtained, which are respectively associated with diffusion depth and average diffusion rate. The correlation between these indicators is analyzed, and macro-performance tests with greater correlation with diffusion behavior characterization parameters (diffusion depth, average diffusion rate) are selected to characterize the diffusion behavior of multiple regeneration interfaces. Specifically, the fusion interface diffusion behavior data are subjected to feature screening, and a fusion interface diffusion behavior evaluation base is constructed based on the screened features; based on the fusion interface diffusion behavior evaluation base, a deep feature factor layer of the fusion interface diffusion behavior is fitted to obtain a deep feature evaluation factor group; based on the deep feature evaluation factor group, the multi-factor comprehensive weight is calculated, and the trust of each evaluation factor in the deep feature evaluation factor group is calculated, and a trust matrix is ​​constructed based on the trust; the multi-factor comprehensive weight and the trust matrix are subjected to random forest model regression analysis to obtain the final evaluation model, and a multi-scale evaluation of the fusion interface diffusion behavior is performed based on the final evaluation model.

[0074] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0075] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-scale evaluation method for diffusion behavior of fusion interface based on multiple regeneration processes of new and old asphalt, characterized by: include: Perform feature screening on the diffusion behavior data of the fusion interface, and construct a fusion interface diffusion behavior evaluation basis based on the screened features; The evaluation base of the fusion interface diffusion behavior includes base factors, and the base factors are macro-scale evaluation, meso-scale evaluation, micro-scale evaluation and molecular-scale evaluation; Based on the fusion interface diffusion behavior evaluation base, the deep characteristic factor layer of the fusion interface diffusion behavior is fitted to obtain the deep characteristic evaluation factor group; The deep characteristic factor layer includes beam body factors, and the beam body factors are diffusion depth and average diffusion rate; The method of fitting the fused interface diffusion behavior deep feature factor layer based on the fused interface diffusion behavior evaluation base to obtain the deep feature evaluation factor group includes: dividing the base factors of the fused interface diffusion behavior evaluation base into multiple beam body factors according to the K-means clustering method to construct the fused interface diffusion behavior deep factor layer; and sequentially composing the multiple beam body factors included in the fused interface diffusion behavior deep factor layer into a deep feature evaluation factor group; Calculate the comprehensive weight of multiple factors based on the deep feature evaluation factor group, calculate the trustworthiness of each evaluation factor in the deep feature evaluation factor group, and construct a trust matrix based on the trustworthiness; Calculate the trustworthiness of each beam factor in the deep feature evaluation factor group, and construct a trustworthiness matrix based on the trustworthiness, specifically including: The diffusion behavior of the fusion interface is divided into four evaluation levels: normal, basically normal, subnormal and abnormal by the four-point evaluation level method; the degree of trust of each beam factor in the deep feature evaluation factor group in the evaluation level is calculated; the size or nature of each factor is divided according to the four normality evaluation levels, and the trust of each factor is clarified. The trust is the degree to which each factor trusts each target state; the qualitative factors are obtained by the expert scoring method based on the confidence index and the empirical method; the quantitative factors are divided into four ranges according to the numerical value and the evaluation level, and the trapezoidal trust function is selected for calculation. The factors with larger values ​​are better, and the ascending semi-trapezoidal distribution function is used; the factors with smaller values ​​are better, and the descending semi-trapezoidal distribution function is used; the monitoring values ​​of each factor are substituted into the trapezoidal trust function to calculate the degree of trust of each beam factor in the deep feature evaluation factor group in the evaluation level; and a trust matrix is ​​constructed based on the trust. The multi-factor comprehensive weights and trust matrix are subjected to random forest model regression analysis to obtain the final evaluation model, and the multi-scale evaluation of the diffusion behavior of the fusion interface is carried out based on the final evaluation model; The random forest model regression analysis of the multi-factor comprehensive weights and the trust matrix is ​​performed to obtain the final evaluation model, which includes: the random forest is composed of M decision trees. During training, the M decision trees are cyclically trained, and the training samples of each decision tree are obtained by bootstrap sampling from the original training set. The features used in training each node of the decision tree are also randomly sampled from the new feature space N. Each decision tree is recursively split according to the judgment criteria. After the training of the M decision trees is completed, the mean of each leaf node is the final regression evaluation result.

2. The multi-scale evaluation method for diffusion behavior of the fusion interface based on multiple regeneration processes of new and old asphalt according to claim 1 is characterized in that: The feature screening of the fusion interface diffusion behavior data includes: using the maximum information coefficient MIC to perform feature screening and remove irrelevant factors.

3. The multi-scale evaluation method for diffusion behavior of the fusion interface based on multiple regeneration processes of new and old asphalt according to claim 1 is characterized in that: The macro-scale evaluation includes: evaluating the macro-performance of the fusion of new and old asphalt by measuring the change values ​​of the three major indicators of asphalt, evaluating the stability of the asphalt structure by turbidity titration, and dynamic shear rheology testing to obtain viscosity, complex shear modulus, storage modulus, loss modulus and phase angle test results.

4. The multi-scale evaluation method for diffusion behavior of the fusion interface based on multiple regeneration processes of new and old asphalt according to claim 1 is characterized in that: The diffusion depth and average diffusion rate include: Fluorescence microscopy was used to obtain images, and Image Pro Plus image processing software was used to quantitatively calculate the area and grayscale value of the gray shift region in the image to obtain the diffusion depth. Tracer elements are added to the new asphalt, and the distribution map of the tracer elements is obtained by scanning with an energy spectrometer. The diffusion depth and the affected area of ​​the new asphalt during each fusion process with the old asphalt are dynamically tracked. A scanning electron microscope combined with an energy dispersive spectrometer was used to analyze the composition of different intervals of the fusion of new and old asphalt; Using the peak force tapping mode of an atomic force microscope, the changes in the two-dimensional and three-dimensional micromorphology of different sections during the fusion process of new and old asphalt were observed. The roughness, Young's modulus, and adhesion micromechanical parameters were collected. The Hertz contact model and the DMT model were used to fit the parameters and the effective modulus of the interface between the new and old asphalt was calculated for each fusion. By setting a specific time interval, the average diffusion rate is obtained and the speed of diffusion under different conditions is evaluated.

5. The multi-scale evaluation method for diffusion behavior of the fusion interface based on multiple regeneration processes of new and old asphalt according to claim 1 is characterized in that: Each decision tree is recursively split according to the judgment criteria, including: the extracted sample set constitutes a root node, and according to the judgment criteria, the sample set is split into a first sub-sample and a second sub-sample; the left subtree is recursively established using the first sub-sample, and the right subtree is established using the second sub-sample; and the condition for stopping the splitting is set. When the splitting cannot continue, the node is marked as a leaf node and a value is assigned at the same time.

Images