Modeling and prediction method of performance degradation of sc-uhpc based on corrosion degree

Abstract

The application discloses a SC-UHPC performance degradation modeling and prediction method based on corrosion degree, and belongs to the technical field of material performance evaluation and prediction. The application aims to build a SC-UHPC degradation modeling method taking the corrosion degree of steel fibers as a main control factor, and fusing interface damage identification, pull-out behavior quantification and bridging energy modeling. The application realizes accurate prediction from micro-damage to macro-performance degradation by building a quantitative mapping relationship among the corrosion grade, bridging performance and macro-mechanical response, and provides a reliable tool for performance evaluation and durability design of the SC-UHPC component in service.

Description

Technical Field

[0001] This invention relates to the field of material performance evaluation and prediction, and in particular to a method for modeling and predicting the performance degradation of SC-UHPC based on the degree of corrosion. Background Technology

[0002] Self-compacting ultra-high performance concrete (SC-UHPC) is a high-performance cement-based material that combines high strength, high toughness, high density, and good flowability. It has been widely used in engineering applications with high durability requirements, such as marine engineering components, bridge structures, high-speed railways, tunnel linings, and ultra-long span structures. To achieve its excellent ductility and crack control performance, SC-UHPC typically incorporates a certain volume fraction of chopped steel fibers to enhance bridging and resist crack initiation and propagation. However, during service, the structure inevitably faces chloride corrosion, wet-dry cycles, and load coupling. This causes the steel fibers to corrode, significantly reducing interfacial bonding performance and leading to a degradation of bridging capacity. Consequently, the flexural properties, ductility, and fracture toughness of SC-UHPC components decrease across the board, ultimately threatening component safety and service life.

[0003] Existing component performance modeling methods are mostly based on the assumption of overall material homogeneity or use simplified corrosion field variables, simulating the component's mechanical response by setting uniform constitutive parameters. These methods often neglect the spatial variability of interfacial bonding performance degradation caused by steel fiber corrosion. Especially under conditions of multiple corrosion levels, non-uniform fiber distribution, and complex environmental effects in actual components, a single constitutive model cannot accurately describe the mapping relationship between corrosion, bridging, and performance. Furthermore, existing research largely focuses on the durability evolution of the matrix material itself; a clear method is lacking for a systematic approach to quantitatively characterize the degree of steel fiber corrosion and drive component performance degradation modeling, and the convertibility between experimental and model parameters is weak.

[0004] Based on this, a method for modeling and predicting the performance degradation of SC-UHPC based on the degree of corrosion is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a corrosion-based modeling and prediction method for SC-UHPC performance degradation. The aim is to construct a SC-UHPC degradation modeling method that uses the degree of steel fiber corrosion as the main controlling factor, integrating interface damage identification, pull-out behavior quantification, and bridging capability modeling. It establishes a quantitative mapping relationship between "corrosion level - interface characteristics - bridging performance - macroscopic mechanical response," enabling accurate prediction from microscopic damage to macroscopic performance degradation, and providing a reliable tool for performance evaluation and durability design of in-service SC-UHPC components.

[0006] To achieve the above objectives, this invention provides a method for modeling and predicting SC-UHPC performance degradation based on corrosion degree, comprising the following steps:

[0007] S1. Prepare self-compacting ultra-high performance concrete specimens according to a unified mix proportion and curing system, and carry out gradient corrosion treatment to simulate different service environments to obtain specimen groups with different steel fiber corrosion degrees.

[0008] S2. Extract steel fiber samples from the specimen group obtained in S1, conduct microscopic morphology observation and macroscopic performance testing, and establish a multi-dimensional corrosion level classification standard for steel fibers based on the observation and test results.

[0009] S3. Single fiber pull-out tests were conducted on steel fibers under different corrosion levels to obtain interface bridging performance parameters under different corrosion levels, and a degradation function of interface bridging performance as corrosion level evolved was constructed.

[0010] S4. Perform a bending test on the corroded specimen to obtain macroscopic mechanical property indicators;

[0011] S5. Using corrosion level and interface bridging performance parameters as input variables and macroscopic mechanical performance indicators as output targets, a nonlinear prediction model is constructed and trained to predict the macroscopic mechanical performance degradation and service life assessment of SC-UHPC components.

[0012] Preferably, in step S1, during the preparation of self-compacting ultra-high performance concrete specimens, the rheological properties are controlled by adjusting the yield stress and plastic viscosity of the self-compacting ultra-high performance concrete paste, achieving a balance between fluidity and stability in the self-compacted state. This ensures the uniform distribution of steel fibers in the freshly mixed state, thereby improving the consistency of the initial interface state and the comparability of bridging performance tests. Specifically, the yield stress is 15~30 Pa, and the plastic viscosity is 15~25 Pa. s.

[0013] Preferably, in step S1, the gradient corrosion treatment is achieved by setting multiple sets of dry and wet cycle corrosion conditions of sodium chloride solution. By adjusting the cycle period, solution concentration and ambient temperature parameters, different levels of steel fiber corrosion from mild to severe are simulated.

[0014] Preferably, in S2, the characteristics on which the multi-dimensional corrosion level classification standard is based include: the change in element content of corrosion products on the steel fiber surface obtained by energy dispersive spectroscopy analysis, the attenuation trend of single fiber pull-out force, and the determination of corrosion level by combining the overall change trend of fiber surface corrosion morphology under different corrosion states.

[0015] Preferably, in step S3, the single fiber pull-out test uses displacement-controlled loading with a loading rate of 0.1 mm / min; load-displacement data are obtained through the single fiber pull-out test, and interface bridging performance parameters are extracted to establish a mapping relationship between corrosion level and interface performance.

[0016] Preferably, in S3, the interface bridging performance parameters are one or more of the following: ultimate pull-out force, pull-out energy consumption, and other interface mechanical characteristic parameters extracted from the single fiber pull-out curve.

[0017] Preferably, in step S3, the degradation function adopts an exponential or power function form and is fitted using the weighted least squares method. Confidence intervals and residual evaluation are introduced, and confidence interval analysis and residual distribution evaluation can be combined to ensure the physical rationality and numerical stability of the model.

[0018] Preferably, in S4, the bending test is a four-point bending test, and the load-deflection response curve is obtained through the test, thereby obtaining the macroscopic mechanical performance index;

[0019] The obtained macroscopic mechanical performance indicators are at least one of ultimate bending strength, peak deflection, or residual bearing capacity; and during the bending test, digital image correlation technology is used to monitor crack propagation behavior and obtain characteristic parameters during crack propagation, which are peak deflection, crack propagation rate, or residual bearing capacity.

[0020] Preferably, in step S5, the nonlinear mapping model is established using multivariate nonlinear regression, support vector regression, BP neural network method, or other machine learning methods that meet the prediction accuracy requirements, and the prediction accuracy is evaluated and improved through cross-validation of the training set and the validation set.

[0021] Preferably, in step S5, the obtained nonlinear mapping model is used to assess the service life bearing capacity and lifespan of the actual SC-UHPC component.

[0022] Therefore, the SC-UHPC performance degradation modeling and prediction method based on corrosion degree of the present invention has the following beneficial effects:

[0023] (1) This invention systematically establishes a complete and quantitative degradation path from “micro-corrosion analysis of steel fibers” to “micro-mechanical properties of fiber-matrix interface” and then to “macro-bending properties of components”. The model has a solid physical basis.

[0024] (2) Using the interface bridging performance parameters obtained by precise experiments as the core input, the blindness of traditional empirical models is avoided, and complex nonlinear relationships are captured through machine learning algorithms, which significantly improves the prediction accuracy and reliability.

[0025] (3) The diversity of actual service environment was simulated through controlled gradient corrosion test. The established model can be directly used to evaluate the residual bearing capacity and remaining life of SC-UHPC components under different corrosion conditions, providing a scientific basis for the safe operation and maintenance and economical repair and reinforcement of the structure.

[0026] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0027] Figure 1 This is a flowchart of an embodiment of the present invention;

[0028] Figure 2 This is a load-deflection curve according to an embodiment of the present invention. Detailed Implementation

[0029] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, but not all embodiments.

[0031] Example

[0032] This embodiment takes the SC-UHPC used in coastal bridge structures as the research object and implements the modeling and prediction method in detail.

[0033] like Figure 1 As shown, the specific steps are as follows:

[0034] S1. Specimen preparation and gradient corrosion:

[0035] An SC-UHPC slurry was prepared using a water-to-binder ratio of 0.19, incorporating 2.0% (volume fraction) of straight steel fibers with a length of 13 mm, and employing a polycarboxylate superplasticizer (1.5%). Rotational rheometer testing showed a yield stress of 22 Pa and a plastic viscosity of 20 Pa·s, meeting the requirements for self-compactness and uniform fiber dispersion. Further microscopic observation indicated a uniform fiber distribution within the rheological window, without agglomeration, providing a unified initial interface state for corrosion and bridging performance testing.

[0036] The grout was poured into a 100mm×100mm×400mm beam-type component mold and cured for 28 days under standard conditions. The specimens cured for 28 days were then subjected to gradient corrosion treatment in a chloride salt environment.

[0037] This embodiment uses a 3.5% NaCl solution for wet-dry cycling, with each cycle consisting of two stages: drying for 24 hours and soaking for 24 hours at 50°C. To obtain steel fiber states with different corrosion levels, different corrosion cycles are set, such as 4 weeks, 10 weeks, and 20 weeks, to create a corrosion environment ranging from light to severe.

[0038] S2. Corrosion Level Characterization and Grading: The corrosion level of the fiber is determined by the changes in the composition of EDS corrosion products and the trend of pull-out force. The specimen group that meets the slight corrosion criterion is classified as corrosion level D=0, the specimen group that meets the moderate corrosion criterion is classified as corrosion level D=1, and the specimen group that meets the severe corrosion criterion is classified as corrosion level D=2. The correspondence between "corrosion cycle - corrosion level" is given.

[0039] Specifically, steel fiber samples were carefully extracted from the corroded specimens, and energy dispersive spectroscopy (EDS) analysis confirmed that the corrosion products were mainly FeOOH and Fe3O4. With prolonged corrosion time, the corrosion product peaks gradually intensified, indicating that a corrosion process progressed from mild to severe on the steel fiber surface. This embodiment quantitatively graded the degree of corrosion based on the evolution of elemental content in EDS and the attenuation trend of single fiber pull-out force.

[0040] EDS analysis showed that after 4 weeks of wet-dry cycling, the steel fiber only underwent localized oxidation in the fracture zone (oxygen content 47.25%→27.42%, iron content 46.85%→64.46%). At 10 weeks, the oxygen content in the fracture zone increased to 47.44%, with significant chlorine enrichment. At 20 weeks, the oxygen content in all three regions exceeded 35%, while the iron content decreased by 15-20% compared to the initial value, indicating that the degree of corrosion gradually increased with each cycle. Simultaneously, the single-fiber fracture load decreased from approximately 90 N at 4 weeks to approximately 50 N at 10 weeks, and further decreased at 20 weeks.

[0041] Based on the above elemental evolution trends (O, Fe, Cl) and the characteristics of pull-out force attenuation, the corrosion levels are classified as follows:

[0042] D=0 (slight corrosion, 4 weeks), D=1 (moderate corrosion, 10 weeks), D=2 (severe corrosion, 20 weeks). The specific corrosion levels are shown in Table 1 below.

[0043] Table 1: Correlation Cycle - Correlation Level Correlation Relationship Table

[0044]

[0045] Based on the above information, the corrosion level D is used as the driving variable for interface degradation modeling.

[0046] S3, Interface Bridging Performance Degradation Modeling:

[0047] A displacement-controlled single-fiber pull-out tester was used to conduct pull-out tests at a rate of 0.1 mm / min. Load-displacement curves of specimens with different corrosion levels were obtained, the ultimate pull-out force was directly read, and the pull-out work was obtained by integrating the area under the curve.

[0048] The experimental results obtained in this embodiment are shown in Table 2, which provide a parameter basis for subsequent modeling of the evolution of interface bridging performance with corrosion level D.

[0049] Table 2: Results of steel fiber single fiber pull-out test under different corrosion levels

[0050]

[0051] Ultimate pull-out force and pull-out work are used as key parameters of interface bridging performance, and a degradation function is established based on their variation with corrosion level D, followed by linear fitting: with corrosion level D as the independent variable and ultimate pull-out force as the limiting parameter. and pulling out Using the performance function as the dependent variable, least squares regression yields the following results:

[0052] ;

[0053] ;

[0054] in, The coefficients of determination of the above two equations The values ​​were 0.94 and 0.99 respectively, and the residual test was passed, indicating that the interface bridging performance degradation / evolution function structure based on corrosion level is reasonable and can be used for subsequent interface degradation modeling and mechanical property prediction.

[0055] S4. Macroscopic mechanical property testing:

[0056] Four-point bending tests were conducted on SC-UHPC beam specimens after different cyclic corrosion treatments. The span was set to 300 mm, the loading span to 100 mm, and the loading rate to 0.5 mm / min. The load-deflection curves were recorded. Figure 2 As shown in Table 3 below, DIC image data processing was performed, and macroscopic mechanical properties such as ultimate bending strength and fracture toughness of specimens under each corrosion level were calculated. A "corrosion level-macroscopic performance" test database was constructed to provide calibration data for the output of subsequent multi-scale mapping models. The energy dissipated during cracking This is the bending toughness coefficient.

[0057] Table 3: Corrosion Grade - Macroscopic Performance Test Results

[0058]

[0059] S5. Construction of multi-scale nonlinear prediction model:

[0060] Based on the interface bridging performance index obtained in step S3 and the macroscopic bending performance index measured in step S4, a support vector regression (SVR) model was adopted, with the radial basis function (RBF) selected as the kernel function. 80% of the sample data was used as the training set, and 20% as the validation set. For example, 12 sets of data are shown in Table 4 below. After training and cross-validation, the prediction error of the model on the validation set was controlled within... Within 5%, the predicted values ​​and experimental values ​​are in high agreement, proving the effectiveness of the model.

[0061] Table 4: SVR Prediction Results

[0062]

[0063] The trained SVR model is used to predict the mechanical properties of SC-UHPC components under unknown corrosion conditions, thereby evaluating their service life load-bearing capacity and remaining life.

[0064] Therefore, this invention provides a corrosion-based modeling and prediction method for SC-UHPC performance degradation. First, it achieves self-compacting properties of the slurry and uniform dispersion of steel fibers by controlling rheological properties, ensuring the consistency and controllability of the initial state of the component interface. Based on this, it simulates the corrosion level of steel fibers under different service environments by controlling the corrosion cycle. Scanning electron microscopy and energy dispersive spectroscopy are used to characterize the corrosion morphology and interface damage features of the steel fibers. Furthermore, pull-out tests are conducted to obtain bridging performance parameters under different corrosion levels. Based on the above, a degradation model between the corrosion level of steel fibers and bridging characteristics is constructed, thereby establishing a multi-scale mapping relationship between corrosion level and SC-UHPC bending performance. This enables the prediction of component mechanical performance degradation driven by microscopic corrosion characteristics. This method is characterized by strong operability, complete modeling closed-loop, and high experimental-prediction coupling. It can be widely applied to the performance evaluation and durability design of SC-UHPC components in service, such as marine engineering components and coastal bridges, and has high engineering applicability and promising prospects for promotion.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for modeling and predicting the performance degradation of self-compacting ultra-high performance concrete (SC-UHPC) based on corrosion degree, characterized in that, Includes the following steps: S1. Specimen Preparation and Gradient Corrosion: Self-compacting ultra-high performance concrete specimens were prepared using a water-cement ratio of 0.19, incorporating 2.0% by volume and 13mm in length straight steel fibers, and SC-UHPC slurry with 1.5% polycarboxylate superplasticizer. The resulting specimens had a yield stress of 22 Pa and a plastic viscosity of 20 Pa·s. After 28 days of standard curing, SC-UHPC beam specimens were subjected to wet-dry cyclic corrosion using a 3.5% NaCl solution. Each cycle included 24 hours of drying at 50℃ and 24 hours of soaking. Three corrosion states were established after 4, 10, and 20 weeks. S2. Corrosion level characterization and classification: Based on the evolution law of O, Fe and Cl element content and the attenuation trend of single fiber pull-out force in energy spectrum analysis, the corrosion state is divided into D=0, D=1 and D=2, where D=0 represents slight corrosion after 4 weeks of corrosion, D=1 represents moderate corrosion after 10 weeks of corrosion and D=2 represents severe corrosion after 20 weeks of corrosion. S3. Interface Bridging Performance Degradation Modeling: A displacement-controlled single-fiber pull-out testing machine was used to conduct pull-out tests at a rate of 0.1 mm / min. Load-displacement curves of specimens with different corrosion levels were obtained. The ultimate pull-out force Fmax was directly read, and the pull-out work Wp was obtained by integrating the area under the curve. The ultimate pull-out force and pull-out work were used as key parameters of interface bridging performance, and a linear degradation function of the two as a function of corrosion level D was established, specifically: ； ； in, The coefficients of determination of the above two equations The values ​​were 0.94 and 0.99 respectively, and the residual test passed. S4. Macroscopic Mechanical Performance Testing: Four-point bending tests were conducted on SC-UHPC beam specimens after different corrosion treatment cycles. Load-deflection curves were recorded, and bending performance indicators of SC-UHPC beam specimens were obtained through the four-point bending tests. A "corrosion level-macroscopic performance" test database was constructed to provide calibration data for the output of subsequent multi-scale nonlinear prediction models. The bending performance indicators are the energy dissipated during cracking and the bending toughness coefficient. S5. Construction of multi-scale nonlinear prediction model: Using corrosion level D, ultimate pull-out force Fmax, and pull-out work Wp as inputs, a bending performance prediction model is established using radial basis function and support vector regression model. Based on the established bending performance prediction model, the service life bearing capacity and service life of SC-UHPC components under unknown corrosion conditions are evaluated.

2. The method for modeling and predicting the performance degradation of self-compacting ultra-high performance concrete (SC-UHPC) based on corrosion degree as described in claim 1, characterized in that: In S1, the dimensions of the prepared SC-UHPC beam specimen are 100mm×100mm×400mm.

3. The method for modeling and predicting the performance degradation of self-compacting ultra-high performance concrete (SC-UHPC) based on corrosion degree as described in claim 1, characterized in that: In S2, the corrosion products are FeOOH and Fe3O4.

4. The method for modeling and predicting the performance degradation of self-compacting ultra-high performance concrete (SC-UHPC) based on corrosion degree as described in claim 1, characterized in that: In S4, the span of the four-point bending test is 300 mm, the loading span is 100 mm, the loading rate is 0.5 mm / min, and during the bending test, digital image correlation technology is used to monitor the crack propagation behavior and obtain the characteristic parameters of the crack propagation process.

Images