Performance detection method of self-repairing concrete for civil engineering

Abstract

The invention relates to the technical field of civil engineering material detection, and particularly discloses a performance detection method of self-repairing concrete for civil engineering, which comprises the following steps: acquiring micro-deformation field data of a crack area in real time in a crack induction process, and constructing a joint simulation model; predicting an initial contact state between a crack and a repair material based on the model, extracting interface pressure distribution and effective bonding area change information, and calculating an interface bonding failure risk characteristic value through an improved fuzzy C-means clustering algorithm; a periodic load is applied to a preset crack area, a closed response signal is collected, and a multi-scale crack expansion coupling degree characteristic value is constructed to quantify coordination between microscopic filling and macroscopic recovery; fusing the two feature values into a composite repair quality feature vector, inputting the composite repair quality feature vector into a recognition model based on an improved deep residual network, recognizing and repairing a blind area, and generating a spatial positioning tag map; and finally, dynamically adjusting the monitoring path and the loading frequency according to an identification result.

Description

Technical Field

[0001] This invention relates to the field of civil engineering material testing technology, specifically to a method for testing the performance of self-healing concrete for civil engineering. Background Technology

[0002] With the rapid development of infrastructure construction, concrete, as one of the most widely used structural materials in civil engineering, inevitably suffers from the effects of loads, environment, and construction defects during long-term service, leading to the generation and propagation of cracks. These cracks not only affect the durability and load-bearing capacity of the structure but can also become channels for the intrusion of corrosive media such as moisture and chloride ions, thereby accelerating steel corrosion and structural deterioration. To address this issue, self-healing concrete has emerged and is gradually becoming an important direction in the research of high-performance concrete materials. Self-healing concrete introduces intelligent components such as microcapsules, microorganisms, or mineral-based repair agents into the matrix. After cracks appear, it can spontaneously or under external stimulation trigger a repair reaction, achieving the sealing of cracks and partial restoration of mechanical properties, thereby effectively extending the service life of the structure and reducing maintenance costs.

[0003] In the field of self-healing concrete technology, the following technical problems exist: when self-healing materials fill cracks, the interfacial bond strength between them and the original concrete matrix is ​​insufficient. This leads to a situation where, although macroscopic inspections show the cracks closed, the structural load-bearing capacity and durability are not substantially restored, resulting in a "false repair" phenomenon. Especially after repeated freeze-thaw cycles, wet-dry cycles, or repeated external loads, the repaired area is prone to secondary cracking, interfacial peeling, or even localized spalling, severely weakening the overall stability and service life of the structure and affecting the long-term effectiveness of the self-healing function and the reliability of its engineering applications. Therefore, how to improve the interfacial bond strength between the self-healing material and the concrete matrix, and achieve a coordinated match of the mechanical properties of the repaired area, has become a key technical challenge that urgently needs to be solved. Summary of the Invention

[0004] The purpose of this invention is to provide a performance testing method for self-healing concrete used in civil engineering, so as to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions: A method for testing the performance of self-healing concrete for civil engineering includes the following steps: S1: Real-time acquisition of micro-deformation field data in the crack region during crack induction, and acquisition of chemical activity and mechanical recovery parameters of self-healing materials; construction of a joint simulation model including hydration reaction, crack opening displacement and local stress and strain by combining a multi-physics field coupled simulation system; S2: Based on the joint simulation model, the initial contact state between the crack and the repair material is predicted, the information on the change of interfacial pressure distribution and effective bonding area is extracted, and the characteristic value of interfacial bonding failure risk is calculated to evaluate the bonding stability under different crack widths. S3: Apply periodic load to the preset crack area and collect the closure response signal. Combine the crack geometry and diffusion rate to calculate the local healing efficiency. Construct multi-scale crack propagation coupling characteristic value based on the healing difference between adjacent segments to quantify the coordination between micro-filling and macro-restoration. S4: The interface bonding failure risk feature value and the multi-scale crack propagation coupling degree feature value are fused into a composite repair quality feature vector, which is then input into the identification repair model. Based on the model output, the repair blind zone is identified. S5: Based on the identification and repair model, identify repair blind spots and dynamically adjust the monitoring path and loading frequency.

[0006] As a further aspect of the present invention: the construction process of the co-simulation model is as follows: During the crack induction process, digital image correlation technology is used to collect micro-deformation field data of the crack area surface in real time to obtain information on crack opening displacement and local strain distribution. Embedded fiber optic sensors were used to synchronously collect data on temperature and humidity changes inside cracks, and the chemical activity parameters of self-healing materials were analyzed by combining the hydration kinetics model of the materials. Based on real-time acquired surface micro-deformation field data of the crack region, a joint simulation model is constructed, which includes the hydration reaction process, crack opening displacement field, and local stress-strain field. In this model, a nonlinear interface transition zone is introduced to simulate the initial bonding state between the self-healing material and the substrate. The joint simulation model is dynamically corrected by a finite element inversion algorithm.

[0007] As a further aspect of the present invention: the prediction of the initial contact state between the crack and the repair material based on the co-simulation model specifically includes: The surface morphology of cracks and the particle distribution of repair materials were simulated using a constructed multiphysics coupled simulation model. The non-smooth Newton iteration method from contact mechanics is introduced to determine the local contact at the interface of crack repair materials. By combining the microstructural parameters of the interface transition zone, the local embedding and bridging behavior of the repair material in the early stage of crack opening is predicted; Output the contact coverage and effective contact point density between the repair material and the substrate for different crack widths.

[0008] As a further aspect of the present invention: the extraction of interfacial pressure distribution and effective bonding area change information specifically includes: Normal and tangential stress cloud diagrams at the crack-repair interface were extracted based on the finite element analysis results. The pressure distribution in the interface region is spatially discretized using the sliding window integration method. Define the bond strength threshold, identify and count the area of ​​the effective bonded region that meets the bond conditions; The effective bond area change rate under different loading stages was calculated as the basic data for bond stability assessment.

[0009] As a further aspect of the present invention: the process for obtaining the interface bonding failure risk characteristic value is as follows: Based on the contact coverage and effective contact point density, and combined with the crack width and repair material particle size parameters, a local contact quality index is constructed. The average shear strength and bond degradation rate of the interface region are calculated using the normal and tangential stress distribution cloud maps and the effective bond area change rate. Contact quality indicators and mechanical degradation parameters are input into a risk identification model based on an improved fuzzy C-means clustering algorithm. The model introduces a weighted distance function and a dynamic fuzzy factor to classify and identify the bonding failure trends in different crack regions. For each clustering result, a membership degree-failure risk mapping function is established to output a continuous interface bonding failure risk score. The score is then normalized to form an interface bonding failure risk feature value in the [0,1] interval, which is used for data fusion and path optimization decision-making in subsequent identification and repair models.

[0010] As a further aspect of the present invention: the application of a periodic load to the preset crack area and the acquisition of the closure response signal, combined with the crack geometry and diffusion rate to calculate the local healing efficiency, specifically includes: After crack induction is completed, a servo-controlled loading device is used to apply a sinusoidal or step-type periodic load to the preset crack area. The strain-time response curves of the crack region under load were acquired in real time using a distributed fiber Bragg grating sensor. By combining digital image correlation technology, the evolution process of the displacement field on the crack surface is acquired simultaneously, and the crack opening and closing displacement sequences are extracted. Time-frequency analysis and filtering were performed on the acquired signals to separate the displacement change components caused by elastic deformation, viscoelastic recovery and self-healing filling; A reaction model was established based on the crack width distribution and the diffusion coefficient of the repair material to predict the spatial concentration evolution of the repair agent inside the crack; The measured crack closure displacement is compared with the theoretical maximum closure amount, and the local healing efficiency is defined as the ratio of the measured displacement to the theoretical displacement.

[0011] As a further aspect of the present invention: the process for obtaining the multi-scale crack propagation coupling degree characteristic value is as follows: Extract the curves of the filling rate of the repair material versus the growth of the contact area of ​​the crack inner wall at the microscale. To obtain the differences in healing efficiency of crack segments and the structural stiffness recovery rate at a macroscopic scale; A multi-scale correlation analysis model based on Granger causality was used to identify the driving effect of micro-repair behavior on macro-performance recovery. Output multi-scale crack propagation coupling degree feature values ​​to quantify the synergistic consistency between micro-repair and macro-structural response, and normalize them to the [0,1] interval.

[0012] As a further aspect of the present invention: the construction process of the identification and repair model is as follows: Construct a multidimensional composite repair quality feature vector, which includes normalized interfacial bonding failure risk feature values ​​and multi-scale crack propagation coupling degree feature values; The feature vector is input into a repair blind zone identification model built on an improved deep residual network. This model integrates an attention mechanism and a time-series memory module to capture the nonlinear evolution of the repair state of the crack area. The model outputs the repair confidence score for each crack sampling point and, in conjunction with a preset threshold, classifies the areas into high-risk, medium-risk, and low-risk zones. Areas with repair confidence levels below a set threshold are marked as potential repair blind spots, and spatial location label maps are generated as input for subsequent dynamic path adjustment and enhanced detection strategies.

[0013] As a further aspect of the present invention: the dynamic adjustment of the monitoring path specifically includes: An initial detection path optimization scheme is generated based on the spatial positioning label map of the repair blind zone, prioritizing coverage of high-risk and medium-risk areas; An improved A* path planning algorithm is introduced, combined with a local path density factor, to perform global obstacle avoidance and local encryption processing on the probe's movement path; The weights of path nodes are dynamically updated, and the sampling density in key areas is increased based on the real-time collected crack closure response signals. The optimized adaptive monitoring path instruction set is output to drive automated detection equipment to perform high-precision scanning tasks.

[0014] As a further aspect of the present invention: the process of adjusting the loading frequency is as follows: The loading excitation level is set according to the confidence score of the repair blind zone and the difference in local healing efficiency, and the low confidence area corresponds to the high frequency excitation strategy; A fuzzy logic-based feedback control system is adopted, which automatically adjusts the frequency and amplitude of the next loading cycle by combining the trend of crack closure response curve changes during the previous loading cycle. A fatigue damage threshold determination mechanism is introduced during high-frequency loading to prevent secondary crack propagation caused by overloading. The adjusted loading parameters are synchronized to the closed-loop feedback system to enable continuous tracking and dynamic intervention of the repair process.

[0015] The beneficial effects of this invention are: (1) This invention constructs a refined detection and simulation analysis system for the behavior of self-healing concrete cracks by integrating high-precision experimental observation and multi-physics coupling modeling. Specifically, digital image correlation (DIC) and embedded fiber optic grating (FBG) sensors are introduced at the crack induction stage to achieve high spatiotemporal resolution dynamic acquisition of key parameters such as micro-deformation field, temperature, and humidity in the crack area, providing high-quality input data for subsequent modeling. On this basis, a three-dimensional joint simulation model is constructed by combining hydration reaction kinetics, nonlinear interface transition zone (ITZ) modeling, and finite element inversion algorithm, which includes crack opening displacement and local stress-strain evolution process. This model can accurately simulate the initial contact state and mechanical response characteristics between the crack and the repair material. This model can not only predict the embedding and bridging behavior of the repair material under different crack widths, but also quantitatively evaluate micro-parameters such as interface contact coverage and effective contact point density, thereby revealing the bonding failure mechanism and identifying potential risk areas. The interface bonding failure risk characteristic values ​​obtained thus provide a scientific basis for optimizing the component design, construction process, and application strategy of self-healing materials, improving the reliability and durability of crack repair, and has important engineering application value and promotion prospects.

[0016] (2) This invention innovatively constructs a repair blind zone identification model based on an improved deep residual network. Combining an attention mechanism and a time-series memory module, it achieves high-precision intelligent identification of the repair status of self-healing concrete crack areas. The model takes the interface bond failure risk feature value and the multi-scale crack propagation coupling degree feature value as the core input, and integrates multi-dimensional information such as crack geometric parameters, environmental temperature and humidity fluctuations, and local healing efficiency to form a multi-dimensional composite repair quality feature vector, thereby comprehensively depicting the co-evolution law between microscopic bond behavior and macroscopic mechanical response in the crack repair process. By training with a sample dataset containing historical loading cycles and manually labeled repair effects, the model can output the repair confidence score of each crack sampling point, and classify the risk level accordingly. It can accurately identify potential repair blind zones with "false repair" or bond degradation risk, and then generate a pixel-level spatial positioning label map, providing a visual basis for subsequent dynamic monitoring strategies. Building upon this foundation, the system further introduces a fuzzy logic-based feedback control mechanism and a fatigue damage threshold determination strategy. Based on the identification results, it adjusts the probe monitoring path and loading excitation frequency in real time: on one hand, an improved A* path planning algorithm optimizes scanning priority, implementing path encryption and repeated observation in high-risk areas to increase data acquisition density in key areas; on the other hand, it sets differentiated loading strategies based on repair confidence scores, enhancing the ability to capture dynamic changes in repair behavior while ensuring structural safety. It also automatically triggers a loading protection mechanism when abnormal strain accumulation or a sudden increase in crack opening displacement is detected, preventing secondary crack propagation caused by high-frequency excitation. All control commands are linked with the simulation modeling and risk assessment modules through a closed-loop feedback system, forming a complete intelligent control chain from data acquisition, feature extraction, risk identification to path-loading collaborative optimization. This method not only significantly improves the automation and intelligence level of self-healing concrete performance evaluation but also provides scientific and efficient technical support for the dynamic adjustment of repair strategies and structural health maintenance in engineering practice, greatly enhancing the adaptability and engineering feasibility of self-healing technology in complex service environments. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a flowchart of a performance testing method for self-healing concrete for civil engineering according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 As shown, this invention provides a performance testing method for self-healing concrete used in civil engineering, comprising the following steps: S1: Real-time acquisition of micro-deformation field data in the crack region during crack induction, and acquisition of chemical activity and mechanical recovery parameters of self-healing materials; construction of a joint simulation model including hydration reaction, crack opening displacement and local stress and strain by combining a multi-physics field coupled simulation system; S2: Based on the joint simulation model, the initial contact state between the crack and the repair material is predicted, the information on the change of interfacial pressure distribution and effective bonding area is extracted, and the characteristic value of interfacial bonding failure risk is calculated to evaluate the bonding stability under different crack widths. S3: Apply periodic load to the preset crack area and collect the closure response signal. Combine the crack geometry and diffusion rate to calculate the local healing efficiency. Construct multi-scale crack propagation coupling characteristic value based on the healing difference between adjacent segments to quantify the coordination between micro-filling and macro-restoration. S4: The interface bonding failure risk feature value and the multi-scale crack propagation coupling degree feature value are fused into a composite repair quality feature vector, which is then input into the identification repair model. Based on the model output, the repair blind zone is identified. S5: Based on the identification and repair model, identify repair blind spots and dynamically adjust the monitoring path and loading frequency.

[0021] In S1, micro-deformation field data of the crack region are acquired in real time during crack induction, and the chemical activity and mechanical recovery parameters of the self-healing material are obtained. A joint simulation model is constructed using a multiphysics coupled simulation system, incorporating hydration reaction, crack opening displacement, and local stress-strain, specifically including: During the crack induction process, a binocular stereo vision system combined with a high-speed CCD camera was used to continuously photograph the surface of the concrete specimen. The image acquisition frequency was no less than 50Hz, and the spatial resolution reached 0.01mm / pixel. Digital image correlation (DIC) technology was used to perform feature point matching and displacement field calculation on the acquired image sequence, thereby obtaining real-time micro-deformation field data of the crack area, including key parameters such as crack opening displacement (COD), local strain distribution, and crack propagation rate. During the concrete pouring stage, embedded fiber Bragg grating (FBG) sensors are pre-embedded and arranged in a linear array along the crack induction direction with a spacing of 10-20 mm. These sensors are used to simultaneously acquire temperature and humidity change curves in the crack interior area. Combined with the hydration kinetic model of the self-healing material (such as the Kamnev type or Avrami equation), the least squares fitting algorithm is used to invert its chemical activity parameters under different temperature and humidity conditions, including the reaction rate constant k, activation energy Ea, diffusion coefficient D, and induction period time t0. Based on the above measured data, a three-dimensional finite element co-simulation model was constructed, which includes the hydration reaction process, crack opening displacement field, and local stress-strain field. The model was built using the ABAQUS / CAE platform, in which the concrete matrix was modeled using C3D8R elements, and the crack region was simulated using cohesive elements (COH3D8) to simulate its nonlinear fracture behavior. In particular, a nonlinear interface transition zone (ITZ) element was introduced at the interface between the self-healing material and the concrete matrix. The thickness of the ITZ layer was set to 10-50 μm, and its mechanical response characteristics were described using a bilinear bond-slip constitutive relation. To improve the model's prediction accuracy, a finite element inversion algorithm was used to dynamically correct the co-simulation model. Measured strain data from the DIC model were input as boundary conditions into the simulation model, and key material parameters such as elastic modulus, Poisson's ratio, and bond strength were iteratively updated using an optimized solver (such as the Isight integrated Optimization module). This ensured that the relative error between the simulated crack opening displacement and the experimental measurement was controlled within 5%. The final co-simulation model accurately reflects the real-time response behavior of the self-healing material during crack evolution, providing a high-precision modeling foundation for subsequent interfacial bond failure risk assessment and multi-scale repair quality analysis.

[0022] In S2, the initial contact state between the crack and the repair material is predicted based on a co-simulation model. Information on interfacial pressure distribution and effective bond area changes is extracted, and interfacial bond failure risk characteristic values ​​are calculated to evaluate bond stability under different crack widths. Specifically, this includes: In a preferred embodiment of the present invention, after crack induction is completed, the initial contact state between the self-healing concrete crack and the repair material is simulated and predicted based on the aforementioned multiphysics joint simulation model. Specifically, firstly, using the established three-dimensional finite element model, combined with crack surface morphology scanning data and statistical results of repair material particle size distribution, the discrete element-finite element coupling method is employed to simulate the random filling behavior of repair material particles within the crack. A microscopic contact distribution map at the crack-repair material interface is then output through a visualization post-processing module. Subsequently, the non-smooth Newton method from contact mechanics is introduced to determine the local contact state between the crack and the repair material, identifying the main contact areas and their corresponding contact types (e.g., point contact, line contact, or surface contact). Further, by combining the microstructural parameters of the interface transition zone (ITZ), including porosity, roughness, and mineral composition distribution, the local embedding depth of the repair material and the load-bearing capacity of bridging fibers / particles in the early stages of crack opening are predicted, thus simulating its initial bonding mechanism. Finally, the contact coverage and effective contact point density between the repair material and the matrix are output for different crack widths (increasing in 0.1 mm increments within the range of 0.1–1.5 mm) for subsequent bonding performance evaluation.

[0023] Extracting information on interfacial pressure distribution and effective bond area variation: Based on the finite element analysis results, normal and tangential stress cloud maps at the crack-repair material interface were exported from ABAQUS / CAE software. The pressure distribution in the interfacial region was spatially discretized using the sliding window integration method, with the sliding window size set to 10 μm × 10 μm to capture local stress concentration effects. Subsequently, a bond strength threshold was defined (e.g., a shear strength greater than 0.5 MPa is considered effective bonding), and the effective bond area that meets the bonding conditions was identified and statistically analyzed accordingly. Finally, the rate of change of effective bond area under multiple loading stages (preloading, unloading, and reloading) was calculated as the basic data for bond stability assessment.

[0024] To quantify the risk of interfacial bonding failure, the following characteristic value calculation process is constructed: First, based on the contact coverage and effective contact point density, combined with the crack width and repair material particle size parameters (such as average particle size d), the following is used to calculate the characteristic value: 50=10–100 μm), construct the local contact quality index CQM = f(coverage, point density, crack width, particle size); secondly, use the normal and tangential stress distribution cloud map and the effective bond area change rate to calculate the average shear strength τ_avg and bond degradation rate k_degradation of the interface region; input the above contact quality index and mechanical degradation parameters into the risk identification model based on the improved fuzzy C-Means (IFCM) clustering algorithm, which introduces a weighted Euclidean distance function and a dynamic fuzzy factor α (within the range of 0.6 to 1.0) to improve the clustering stability under small sample data; for each clustering result (high risk, medium risk, low risk), establish the membership degree μ and failure risk mapping function R = g(μ) is used to output a continuous interface bonding failure risk score. Finally, the score is normalized to form an interface bonding failure risk feature value in the interval [0,1], which is used for data fusion and path optimization decision-making of the subsequent composite repair quality assessment model. This significantly improves the identification accuracy and engineering practicality of the "false repair" phenomenon.

[0025] In S3, a periodic load is applied to the preset crack region and the closure response signal is collected. The local healing efficiency is calculated by combining the crack geometry and diffusion rate. Furthermore, a multi-scale crack propagation coupling characteristic value is constructed based on the healing differences between adjacent segments to quantify the coordination between microscopic filling and macroscopic restoration. Specifically, this includes: In a preferred embodiment of the present invention, after crack induction is completed, a periodic load is applied to a preset crack area and its closure response signal is acquired. Specifically, a high-precision servo-controlled loading device is used to apply a sinusoidal or step-type periodic load to the preset crack area on the concrete specimen. The loading frequency is set to 0.1 to 2 cycles per second, and the loading amplitude is dynamically adjusted according to 20% to 50% of the concrete compressive strength to ensure that the loading process does not cause new structural damage. At the same time, multiple distributed fiber optic grating sensors are arranged along the crack direction in the crack area. The spacing between these sensors is set to 10 to 20 mm to acquire strain change data of the crack under periodic load in real time. The sampling frequency is not less than 1,000 times per second, and the spatial resolution reaches a level where every micro-strain unit can be accurately identified.

[0026] By combining digital image correlation techniques, random speckle patterns were sprayed onto the surface of the specimen, and the displacement changes of the crack surface during loading were simultaneously recorded using a high-speed camera, thereby extracting the specific displacement values ​​for crack opening and closing. Subsequently, the acquired raw signals underwent time-frequency analysis and filtering, including using wavelet transform to remove noise interference and using a bandpass filter to effectively separate the displacement change components caused by elastic deformation, viscoelastic recovery, and self-healing filling, thus obtaining the contribution ratio of different mechanisms to crack closure behavior.

[0027] Further calculations were performed on the local healing efficiency. The specific method is as follows: Based on crack width distribution data and the diffusion capacity of the repair material, a one-dimensional diffusion-reaction model was established to predict the spatial concentration evolution of the repair agent within the crack. The diffusion rate of the repair agent was determined experimentally, typically varying between 10^-12 and 10^-10 per square meter, while the reaction rate was estimated based on hydration kinetics. The theoretical maximum crack closure displacement was obtained through finite element simulation, representing the maximum closure effect achievable assuming the crack is completely filled with the repair material. Then, the actual measured crack closure displacement was compared with this theoretical value. The local healing efficiency was defined as the ratio of the measured crack closure displacement to the theoretical maximum closure displacement, ranging from 0 to 1. A higher value indicates a better repair effect for that crack segment.

[0028] To assess the coordination between microscopic repair behavior and macroscopic structural restoration, a multi-scale crack propagation coupling degree feature value is constructed. The specific implementation is as follows: First, at the microscopic scale, the velocity of repair material entering the crack and the change curve of the contact area between the repair material and the crack inner wall over time are extracted. Second, at the macroscopic scale, the differences in healing efficiency between adjacent crack segments and the proportion of overall structural stiffness restoration are obtained. Next, a causal-based multi-scale correlation analysis method is used to identify the degree of influence of microscopic repair parameters on the restoration of macroscopic mechanical properties, i.e., to determine whether the microscopic repair behavior of a certain crack segment significantly promotes the improvement of the overall structural performance of that area. Finally, a multi-scale crack propagation coupling degree feature value is output, which reflects the synergistic consistency between microscopic repair and macroscopic restoration. Its numerical range is compressed to between 0 and 1 using a range normalization method, so that it can be used as a unified dimension input feature in the data fusion and path optimization decision-making of the composite repair quality assessment model.

[0029] In S4, the characteristic values ​​of interfacial bonding failure risk and multi-scale crack propagation coupling degree are fused into a composite repair quality feature vector, which is input into the identification and repair model. Based on the model output, repair blind spots are identified, specifically including: In a preferred embodiment of the present invention, after extracting the characteristic values ​​of interface bonding failure risk and multi-scale crack propagation coupling degree, the two are further integrated into a comprehensive composite repair quality feature vector for quantitatively evaluating the overall repair status of the self-healing concrete crack area. Specifically, the two aforementioned feature values ​​are first normalized to unify their numerical range to the [0,1] interval, thereby eliminating the influence of different physical dimensions. Then, these two normalized feature values ​​are used as the main input variables, and combined with auxiliary information such as crack width, local healing efficiency standard deviation, and environmental temperature and humidity fluctuation coefficient, a multi-dimensional composite repair quality feature vector is constructed. This vector has a dimension of not less than 5 dimensions, which can comprehensively reflect the microscopic bonding performance, macroscopic recovery ability, and spatial distribution consistency of the crack area.

[0030] The composite repair quality feature vector is input into a repair blind zone identification model built on an improved deep residual network. This model structure includes multiple residual learning modules and integrates an attention mechanism at key layers to enhance the model's focus on high-risk areas. Simultaneously, a time-series memory module is introduced, enabling the model to capture the evolution of crack repair status over time and identify phenomena such as "pseudo-repair" where early adhesion is good but later degradation occurs. During training, a labeled sample dataset is used, where each sample includes crack location, historical loading cycle count, measured healing effect, and manually labeled repair effectiveness level. The model's hyperparameters are optimized using cross-validation to ensure its generalization ability in unknown crack regions.

[0031] After the model runs, it outputs the repair confidence score for each sampling point within the crack area. This score reflects the effectiveness level of the self-healing behavior in the current area, and the numerical range is also normalized to the [0,1] interval, where a score close to 1 indicates a good repair effect, and a score close to 0 indicates a poor repair effect or a potential failure risk. Subsequently, all sampling points are classified according to a preset threshold (e.g., 0.6). Areas with a confidence score higher than the threshold are classified as low-risk areas, areas between 0.4 and 0.6 are classified as medium-risk areas, and areas below 0.4 are classified as high-risk areas. Finally, the system automatically identifies all areas with a repair confidence score lower than the set threshold and marks them as potential repair blind spots. At the same time, it generates a spatial positioning label map corresponding to the crack geometry. This label map marks the specific location and coverage of the repair blind spots with pixel-level precision, serving as an important input basis for subsequent dynamic path adjustment and enhanced detection strategies. This enables accurate identification and targeted intervention of weak repair areas in self-healing concrete structures.

[0032] In S5, based on the identification and repair model, repair blind spots are identified, and the monitoring path and loading frequency are dynamically adjusted, specifically including: In a preferred embodiment of the present invention, after completing the risk assessment of the repair status of the crack area, an initial detection path optimization scheme is generated based on the spatial positioning label map of the repair blind zone. Specifically, the system first reads the spatial positioning label map output by the aforementioned repair blind zone identification model, which marks the specific locations of high-risk, medium-risk, and low-risk areas with pixel-level precision; then, the probe scanning priority is set according to the risk level, prioritizing the coverage of high-risk areas with confidence scores below a set threshold, and appropriately increasing the sampling point density of medium-risk areas to improve the data acquisition accuracy of key parts; based on this, an improved A* path planning algorithm is introduced, which combines a local path density factor to perform global obstacle avoidance and local optimization of the probe movement path, wherein the local path density factor is used to control the smoothness of the path transition and the probe movement efficiency between areas of different risk levels, avoiding frequent start-stop operations that affect the continuity of detection.

[0033] During the detection process, the path node weights are dynamically updated: the system determines whether the repair behavior of the current area has significant changes based on the real-time collected crack closure response signals. If an area shows obvious healing fluctuations or stress concentration trends in multiple consecutive loading cycles, the priority weight of the path nodes in that area is automatically increased, driving the probe to perform multiple revisit scans in that area, thereby enhancing the sampling density and data reliability of key areas. Finally, the optimized adaptive monitoring path instruction set is output, including control parameters such as probe movement sequence, dwell time, and scanning frequency, which are used to drive the automated detection equipment to perform high-precision scanning tasks and achieve continuous focused observation of the repaired weak areas.

[0034] The system also dynamically adjusts the loading frequency and excitation strategy based on the confidence score of the repair blind zone and the difference in local healing efficiency between adjacent crack segments. Specifically, for areas with low repair confidence scores, the system sets a high-frequency loading strategy with a loading frequency of 1 to 2 cycles per second to accelerate the capture of their repair behavior; while for low-risk areas with high healing efficiency and good stability, a low-frequency loading strategy is adopted with a loading frequency of 0.1 to 0.5 cycles per second to reduce unnecessary mechanical fatigue damage. In addition, the system integrates a feedback control system based on fuzzy logic, which can automatically adjust the frequency and amplitude of the next round of loading according to the trend of crack closure response curve changes in the previous loading cycle. For example, when the crack closure rate suddenly drops, the system will appropriately reduce the loading amplitude to avoid aggravating structural damage.

[0035] During high-frequency loading, the system also introduces a fatigue damage threshold determination mechanism: by pre-setting the material fatigue life curve and cumulative strain threshold, the strain accumulation in the crack area is monitored in real time during loading. Once the strain value exceeds the safety threshold or the crack opening displacement shows an abnormal growth trend, the loading protection mechanism is immediately triggered, switching to low-frequency loading mode or pausing the loading action to prevent secondary crack expansion caused by overloading. All adjusted loading parameters are synchronously uploaded to the closed-loop feedback system, forming a closed-loop control link with the repair blind zone identification model and the composite repair quality feature vector calculation module, thereby realizing continuous tracking, intelligent intervention and dynamic optimization of the self-healing concrete repair process.

[0036] The working principle of this invention: This invention aims to achieve high-precision, multi-scale, closed-loop intelligent assessment and dynamic intervention of the crack repair process. The technical solution includes the following key steps: First, during the crack induction process, digital image correlation (DIC) technology and embedded fiber optic grating (FBG) sensors are used to collect real-time data on micro-deformation fields, temperature, and humidity in the crack area. A joint simulation model including hydration reaction, crack opening displacement, and local stress and strain is constructed by combining a hydration kinetics model and a finite element inversion algorithm. Subsequently, based on this model, the initial contact state between the crack and the repair material is predicted, and information on the interface pressure distribution and effective bonding area changes is extracted. An improved fuzzy C-means clustering algorithm is used to calculate the interface bonding failure risk characteristic value, which is used to quantify the bonding stability under different crack widths. Further, by applying periodic loads to a preset crack area and collecting closure response signals, the local healing efficiency is calculated by combining crack geometry and diffusion rate. A multi-scale correlation analysis method based on causality is used to construct a multi-scale crack repair mechanism. The crack propagation coupling feature value is used to assess the coordination between micro-filling and macro-restoration. Then, the two feature values ​​are fused into a composite repair quality feature vector, which is input into an identification model built based on an improved deep residual network. This model outputs a repair confidence score for each crack sampling point, identifying high, medium, and low-risk areas and marking potential repair blind spots, generating a spatial positioning label map. Finally, the monitoring path and loading frequency are dynamically adjusted based on the identification results. On one hand, an improved A* algorithm optimizes the probe scanning path, prioritizing coverage of high-risk areas and increasing sampling density in key areas. On the other hand, the loading excitation level is set based on the difference between repair confidence and healing efficiency. Combined with a fuzzy logic feedback control system, loading parameters are automatically adjusted, and a fatigue damage threshold judgment mechanism is introduced to prevent secondary crack propagation. All adjustment results are synchronized to the closed-loop feedback system, achieving continuous tracking and intelligent control of the entire repair process. This method has significant innovations in data acquisition, modeling analysis, risk identification, and path-loading co-optimization, effectively improving the accuracy and engineering applicability of performance evaluation for self-healing concrete structures.

[0037] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0038] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0039] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0040] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for testing the performance of self-healing concrete for civil engineering, characterized in that, Includes the following steps: S1: Real-time acquisition of micro-deformation field data in the crack region during crack induction, and acquisition of chemical activity and mechanical recovery parameters of self-healing materials; construction of a joint simulation model including hydration reaction, crack opening displacement and local stress and strain by combining a multi-physics field coupled simulation system; S2: Based on the joint simulation model, the initial contact state between the crack and the repair material is predicted, the information on the change of interfacial pressure distribution and effective bonding area is extracted, and the characteristic value of interfacial bonding failure risk is calculated to evaluate the bonding stability under different crack widths. S3: Apply periodic load to the preset crack area and collect the closure response signal. Combine the crack geometry and diffusion rate to calculate the local healing efficiency. Construct multi-scale crack propagation coupling characteristic value based on the healing difference between adjacent segments to quantify the coordination between micro-filling and macro-restoration. S4: The interface bonding failure risk feature value and the multi-scale crack propagation coupling degree feature value are fused into a composite repair quality feature vector, which is then input into the identification repair model. Based on the model output, the repair blind zone is identified. S5: Based on the identification and repair model, identify repair blind spots and dynamically adjust the monitoring path and loading frequency.

2. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The construction process of the co-simulation model is as follows: During the crack induction process, digital image correlation technology is used to collect micro-deformation field data of the crack area surface in real time to obtain information on crack opening displacement and local strain distribution. Embedded fiber optic sensors were used to synchronously collect data on temperature and humidity changes inside cracks, and the chemical activity parameters of self-healing materials were analyzed by combining the hydration kinetics model of the materials. Based on real-time acquired surface micro-deformation field data of the crack region, a joint simulation model is constructed, which includes the hydration reaction process, crack opening displacement field, and local stress-strain field. In this model, a nonlinear interface transition zone is introduced to simulate the initial bonding state between the self-healing material and the substrate. The joint simulation model is dynamically corrected by a finite element inversion algorithm.

3. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The prediction of the initial contact state between the crack and the repair material based on the co-simulation model specifically includes: The surface morphology of cracks and the particle distribution of repair materials were simulated using a constructed multiphysics coupled simulation model. The non-smooth Newton iteration method from contact mechanics is introduced to determine the local contact at the interface of crack repair materials. By combining the microstructural parameters of the interface transition zone, the local embedding and bridging behavior of the repair material in the early stage of crack opening is predicted; Output the contact coverage and effective contact point density between the repair material and the substrate for different crack widths.

4. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The extraction of information on interfacial pressure distribution and effective bonding area changes specifically includes: Normal and tangential stress cloud diagrams at the crack-repair interface were extracted based on the finite element analysis results. The pressure distribution in the interface region is spatially discretized using the sliding window integration method. Define the bond strength threshold, identify and count the area of ​​the effective bonded region that meets the bond conditions; The effective bond area change rate under different loading stages was calculated as the basic data for bond stability assessment.

5. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The process for obtaining the characteristic value of the interface bonding failure risk is as follows: Based on the contact coverage and effective contact point density, and combined with the crack width and repair material particle size parameters, a local contact quality index is constructed. The average shear strength and bond degradation rate of the interface region are calculated using the normal and tangential stress distribution cloud maps and the effective bond area change rate. Contact quality indicators and mechanical degradation parameters are input into a risk identification model based on an improved fuzzy C-means clustering algorithm. The model introduces a weighted distance function and a dynamic fuzzy factor to classify and identify the bonding failure trends in different crack regions. For each clustering result, a membership degree-failure risk mapping function is established to output a continuous interface bonding failure risk score. The score is then normalized to form an interface bonding failure risk feature value in the [0,1] interval, which is used for data fusion and path optimization decision-making in subsequent identification and repair models.

6. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The process of applying a periodic load to the preset crack area and acquiring the closure response signal, combined with the crack geometry and diffusion rate to calculate the local healing efficiency, specifically includes: After crack induction is completed, a servo-controlled loading device is used to apply a sinusoidal or step-type periodic load to the preset crack area. The strain-time response curves of the crack region under load were acquired in real time using a distributed fiber Bragg grating sensor. By combining digital image correlation technology, the evolution process of the displacement field on the crack surface is acquired simultaneously, and the crack opening and closing displacement sequences are extracted. Time-frequency analysis and filtering were performed on the acquired signals to separate the displacement change components caused by elastic deformation, viscoelastic recovery and self-healing filling; A reaction model was established based on the crack width distribution and the diffusion coefficient of the repair material to predict the spatial concentration evolution of the repair agent inside the crack; The measured crack closure displacement is compared with the theoretical maximum closure amount, and the local healing efficiency is defined as the ratio of the measured displacement to the theoretical displacement.

7. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The process for obtaining the multi-scale crack propagation coupling degree feature value is as follows: Extract the curves of the filling rate of the repair material versus the growth of the contact area of ​​the crack inner wall at the microscale. To obtain the differences in healing efficiency of crack segments and the structural stiffness recovery rate at a macroscopic scale; A multi-scale correlation analysis model based on Granger causality was used to identify the driving effect of micro-repair behavior on macro-performance recovery. Output multi-scale crack propagation coupling degree feature values ​​to quantify the synergistic consistency between micro-repair and macro-structural response, and normalize them to the [0,1] interval.

8. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The process of constructing the identification and repair model is as follows: Construct a multidimensional composite repair quality feature vector, which includes normalized interfacial bonding failure risk feature values ​​and multi-scale crack propagation coupling degree feature values; The feature vector is input into a repair blind zone identification model built on an improved deep residual network. This model integrates an attention mechanism and a time-series memory module to capture the nonlinear evolution of the repair state of the crack area. The model outputs the repair confidence score for each crack sampling point and, in conjunction with a preset threshold, classifies the areas into high-risk, medium-risk, and low-risk zones. Areas with repair confidence levels below a set threshold are marked as potential repair blind spots, and spatial location label maps are generated as input for subsequent dynamic path adjustment and enhanced detection strategies.

9. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The dynamic adjustment of the monitoring path specifically includes: An initial detection path optimization scheme is generated based on the spatial positioning label map of the repair blind zone, prioritizing coverage of high-risk and medium-risk areas; An improved A* path planning algorithm is introduced, combined with a local path density factor, to perform global obstacle avoidance and local encryption processing on the probe's movement path; The weights of path nodes are dynamically updated, and the sampling density in key areas is increased based on the real-time collected crack closure response signals. The optimized adaptive monitoring path instruction set is output to drive automated detection equipment to perform high-precision scanning tasks.

10. The performance testing method for self-healing concrete for civil engineering according to claim 1, characterized in that, The process of adjusting the loading frequency is as follows: The loading excitation level is set according to the confidence score of the repair blind zone and the difference in local healing efficiency, and the low confidence area corresponds to the high frequency excitation strategy; A fuzzy logic-based feedback control system is adopted, which automatically adjusts the frequency and amplitude of the next loading cycle by combining the trend of crack closure response curve changes during the previous loading cycle. A fatigue damage threshold determination mechanism is introduced during high-frequency loading to prevent secondary crack propagation caused by overloading. The adjusted loading parameters are synchronized to the closed-loop feedback system to enable continuous tracking and dynamic intervention of the repair process.

Images