Method and system for judging initial crack strength of UHPC (Ultra High Performance Concrete) bending test based on image method

Abstract

The invention discloses a UHPC bending test initial crack strength judgment method and system based on an image method, and belongs to the technical field of mechanical property tests.The UHPC bending test initial crack strength judgment method comprises the steps that random speckles are sprayed on the surface of a UHPC test piece, and a fiber bragg grating sensor array is pre-embedded in the UHPC test piece; performing X-ray computed tomography on the same batch of test pieces to establish a fiber-matrix interface transition region damage evolution model; when three-point anti-bending loading is carried out on the test piece, surface speckle images are collected in real time, full-field strain distribution is calculated through a DIC algorithm, and the surface initial cracking time and the coordinates of a strain concentration area are determined in combination with criteria; meanwhile, monitoring an optical fiber wavelength drift distance, and calculating a variable coefficient to determine internal initial cracking time; calculating the interface debonding rate and the interface initial crack point time according to the model and the real-time interface stress; and constructing a multi-source initial crack feature data set containing surface strain, internal strain anomaly and interface debonding rate, inputting the data set into the neural network model to obtain a weight coefficient, and calculating the initial crack breaking strength through weighted summation.

Description

Technical Field

[0001] This invention belongs to the field of mechanical property testing technology, specifically a method and system for determining the initial crack strength of UHPC flexural test based on image method. Background Technology

[0002] In flexural tests of Ultra-High Performance Concrete (UHPC), accurately determining the initial crack strength is crucial for evaluating UHPC material properties. Currently used methods for determining initial crack strength, such as the load-displacement curve inflection point method, strain gauge / extensometer method, and acoustic emission technology, all have certain limitations. The load-displacement curve inflection point method is affected by the bridging effect of steel fibers, resulting in an inflection point that is not prominent, making it difficult to accurately identify the initial cracking time. The strain gauge / extensometer method suffers from problems such as contact slippage leading to data distortion and large gauge lengths causing missed microcrack initiation points. While acoustic emission technology can capture microcracks, the signal is easily affected by environmental interference and cannot directly reflect crack morphology. Furthermore, existing methods often rely on single data indicators, lacking multi-dimensional information fusion, resulting in insufficient accuracy and reliability, and failing to meet the testing requirements of the complex mechanical properties of UHPC. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention proposes a method and system for determining the initial crack strength of UHPC flexural tests based on image processing. First, random speckle patterns are sprayed onto the surface of the UHPC specimen, and an array of fiber Bragg grating sensors is embedded internally. X-ray computed tomography (CT) scans are performed on specimens from the same batch to establish a damage evolution model of the fiber-matrix interface transition zone. During three-point flexural loading of the specimen, surface speckle images are acquired in real time. The DIC algorithm is used to calculate the full-field strain distribution, and the initial crack time and coordinates of the strain concentration zone are determined based on the criteria. Simultaneously, the fiber wavelength drift is monitored, and the coefficient of variation is calculated to determine the internal initial crack time. The interface debonding rate and the interface initial crack point time are calculated based on the model and real-time interface stress. A multi-source initial crack feature dataset containing surface strain, internal strain anomalies, and interface debonding rate is constructed, input into a neural network model to obtain weighting coefficients, and the initial crack flexural strength is calculated through weighted summation.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for determining the initial crack strength of UHPC flexural tests based on image analysis includes:

[0006] Random speckle was sprayed on the surface of the UHPC specimen, and a fiber Bragg grating sensor array was pre-embedded inside the specimen. At the same time, X-ray computed tomography was performed on the same batch of UHPC specimens to establish a damage evolution model of the fiber-matrix interface transition zone.

[0007] Three-point flexural loading tests were conducted on the specimens, and speckle images of the specimen surface were acquired in real time. The strain distribution across the entire field was calculated using the DIC algorithm. Combined with the pre-set surface initial cracking criterion, the surface initial cracking time and the coordinates of the local strain concentration area at the time of initial cracking were determined. At the same time, the fiber Bragg grating sensor array was used to monitor the fiber wavelength drift in real time, and the wavelength drift variation coefficient was calculated to determine the internal initial cracking time.

[0008] Based on the fiber-matrix interface transition zone damage evolution model and combined with the interface stress measured in real time during the loading test, the interface debonding rate is calculated, and the interface initial cracking point time is determined according to the preset interface initial cracking judgment criteria.

[0009] The earliest of the following times was taken as the overall initial crack time: the initial crack time of the surface, the initial crack time of the interior, and the initial crack point time of the interface.

[0010] The surface strain, internal strain anomaly, and interface debonding rate corresponding to the initial cracking time are integrated to form a multi-source initial cracking feature dataset. This dataset is then substituted into a pre-trained neural network model and combined with the weight coefficients obtained during training. The initial cracking flexural strength of UHPC is calculated by weighted summation.

[0011] Specifically, the process of performing X-ray computed tomography on UHPC specimens from the same batch to establish a damage evolution model of the fiber-matrix interface transition zone includes:

[0012] Select an X-ray computed tomography (CT) scanner and adjust the scanning parameters according to the size and density of the UHPC specimen; the scanning parameters include X-ray energy, scanning speed, and slice thickness.

[0013] The UHPC specimens of the same batch were placed on the scanning table of the scanning equipment for all-round scanning. The scanning equipment converted the acquired transmission signals into a three-dimensional structural image of the specimen through the data processing system.

[0014] Edge detection is performed on the three-dimensional structural image of the specimen obtained by scanning to identify the information of the fiber-matrix interface transition zone;

[0015] Based on the identified fiber-matrix interface transition zone information, and combined with the mechanical property parameters of UHPC material, a finite element model of the interface transition zone is established.

[0016] In the finite element model of the interface transition zone, the material nonlinearity, geometric nonlinearity and the interaction between the fiber and the matrix in the interface transition zone are considered. Boundary conditions and loading methods are set. By calculating the established finite element model of the interface transition zone, the stress-strain relationship of the UHPC specimen during the flexural loading test is simulated.

[0017] Based on the finite element analysis results, the stress distribution, strain distribution, and damage evolution of the interface transition zone under different loading stages were obtained, and a damage evolution model of the fiber-matrix interface transition zone was established. The damage evolution was described by defining damage variables, which included plastic strain and energy dissipation indices.

[0018] Specifically, the calculation of the full-field strain distribution using the DIC algorithm includes:

[0019] The collected speckle images of the specimen surface are preprocessed;

[0020] Select an initial reference image and a deformed speckle image from the pre-processed speckle images on the specimen surface; the initial reference image is a random speckle image of the specimen surface sprayed before the loading test; the deformed speckle image is a speckle image of the specimen surface after deformation, collected at set time intervals during the loading test.

[0021] A predetermined number of sub-regions are selected on the initial reference image and the deformed speckle image, respectively;

[0022] The DIC algorithm is used to match sub-regions on the initial reference image and the deformed speckle image. By calculating the correlation coefficient between the sub-regions, the displacement vector of each sub-region in the deformed speckle image relative to the corresponding sub-region in the initial reference image is determined.

[0023] Based on the displacement vector, the full-field strain distribution on the specimen surface is calculated using the least squares method, thus obtaining strain information of the specimen under different loading stages.

[0024] Specifically, the process of determining the initial cracking time and the coordinates of the local strain concentration area at the time of initial cracking by combining a pre-set surface cracking criterion includes:

[0025] Based on the characteristics of UHPC materials and experimental requirements, a surface initial cracking criterion is pre-set based on the strain threshold.

[0026] In the calculated full-field strain distribution, the strain value and strain gradient change at each location are monitored in real time;

[0027] When the strain value at any location exceeds the preset surface initial cracking criterion, the time at this moment is recorded as the surface initial cracking time. At the same time, the coordinates of the location that exceeds the preset surface initial cracking criterion are determined as the coordinates of the local strain concentration area at the time of initial cracking and stored.

[0028] Specifically, the step of using a fiber Bragg grating sensor array to monitor the fiber wavelength drift in real time, calculate the wavelength drift variation coefficient, and determine the internal initial cracking time includes:

[0029] Connect the fiber Bragg grating sensor array to the fiber Bragg grating demodulator via optical fiber;

[0030] Before the loading test begins, the fiber Bragg grating sensors are initially calibrated, and the initial wavelength value of each fiber Bragg grating sensor is recorded under stress-free conditions.

[0031] During the bending load test, the fiber Bragg grating demodulator acquires the wavelength data of the fiber Bragg grating sensors in real time, processes the acquired wavelength data to obtain the current wavelength value, and calculates the wavelength drift of each fiber Bragg grating sensor; the wavelength drift is equal to the current wavelength value minus the initial wavelength value.

[0032] Statistical analysis was performed on the wavelength drift of all fiber Bragg grating sensors in the fiber Bragg grating sensor array, and the average value and standard deviation of the wavelength drift were calculated.

[0033] The wavelength drift coefficient of variation is calculated based on the average value and standard deviation of the wavelength drift; the wavelength drift coefficient of variation is equal to the standard deviation of the wavelength drift divided by the average value of the wavelength drift.

[0034] During the flexural loading test, the changes in the wavelength drift variation coefficient were monitored in real time.

[0035] When the wavelength drift variation coefficient exceeds the preset internal initial crack determination threshold, the time at this moment is recorded as the internal initial crack time.

[0036] Specifically, the process of calculating the interface debonding rate based on the fiber-matrix interface transition zone damage evolution model and the interfacial stress measured in real time during the loading test includes:

[0037] During the flexural loading test, stress sensors arranged on or inside the specimen are used to measure the stress value in the interface transition zone in real time.

[0038] The stress values ​​of the interface transition zone measured in real time are input into the pre-established fiber-matrix interface transition zone damage evolution model.

[0039] Based on the damage variables and stress-strain relationship defined in the fiber-matrix interface transition zone damage evolution model, the damage degree of the interface transition zone at different loading times is calculated.

[0040] Based on the calculated degree of damage, combined with the geometry and size of the interface transition zone, the interface debonding rate is calculated; the interface debonding rate is the ratio of the debonded area of ​​the interface transition zone to the total area.

[0041] Specifically, determining the initial interface cracking point time according to the preset interface initial cracking judgment criteria includes:

[0042] Based on the characteristics and test requirements of UHPC materials, a pre-defined standard for judging initial interface cracking is established; the standard for judging initial interface cracking is set based on the interface debonding rate threshold index.

[0043] The changes in interface debonding rate and interface stress are monitored in real time from the calculated interface debonding rate and the real-time measured interface stress data.

[0044] When the interface debonding rate exceeds the preset interface debonding rate threshold, the time at this moment is recorded as the interface initial cracking point time.

[0045] Specifically, the calculation of the initial crack flexural strength of UHPC by weighted summation includes:

[0046] Obtain the constructed multi-source initial fracture feature dataset and the obtained weight coefficients;

[0047] Based on the weighting coefficients obtained from the neural network model, the surface strain, internal strain anomaly, and interface debonding rate in the multi-source initial crack feature dataset are weighted and summed to obtain the initial crack flexural strength of UHPC; the surface strain is obtained by the DIC algorithm; the internal strain anomaly is determined by the wavelength drift variation coefficient.

[0048] The initial crack flexural strength of UHPC is output to the test report or display interface and stored in the database.

[0049] The image-based UHPC flexural strength determination system includes: a data acquisition module, an initial crack characteristic determination module, an initial crack time determination module, and a strength calculation module.

[0050] The data acquisition module is used to collect data of the UHPC specimen during the flexural loading test;

[0051] The initial crack feature determination module is used to determine the initial crack features of the surface, interior and interface based on the collected data.

[0052] The initial cracking time determination module is used to synchronize the surface initial cracking time, the internal initial cracking time, and the interface initial cracking point time to determine the overall initial cracking time.

[0053] The strength calculation module is used to construct a multi-source initial crack feature dataset and to calculate the initial crack flexural strength of UHPC using a neural network model.

[0054] Specifically, the initial crack feature determination module includes: a surface initial crack feature determination unit, an internal initial crack feature determination unit, and an interface initial crack feature determination unit;

[0055] The surface initial crack feature determination unit is used to process the collected speckle image of the specimen surface using the DIC algorithm, calculate the strain distribution in the whole field, and determine the surface initial crack time and the coordinates of the local strain concentration area at the time of initial cracking.

[0056] The internal initial crack feature determination unit is used to process the fiber wavelength drift data collected by the fiber Bragg grating sensor array, calculate the wavelength drift variation coefficient, and determine the internal initial crack time.

[0057] The interface initial crack feature determination unit is used to calculate the interface debonding rate and determine the time of the interface initial cracking point based on the fiber-matrix interface transition zone damage evolution model and the interface stress measured in real time during the loading test.

[0058] Compared with the prior art, the beneficial effects of the present invention are:

[0059] 1. This invention proposes an image-based system for determining the initial crack strength of the UHPC flexural test, and optimizes and improves its architecture, operation steps, and process. The system has the advantages of simple process, low investment and operating costs, and low production costs.

[0060] 2. This invention proposes an image-based method for determining the initial crack strength in UHPC flexural tests. By spraying speckle patterns onto the surface of the UHPC specimen, embedding sensors internally, and performing X-ray scanning modeling, it achieves accurate acquisition of multi-dimensional damage information on the surface, interior, and interface of the specimen. During the three-point flexural loading test, it can monitor and determine the initial crack time and location on the surface and the initial crack time inside the specimen in real time. It can also determine the initial crack point time at the interface based on the model, comprehensively and meticulously capturing the initial crack characteristics of the UHPC specimen during the loading test.

[0061] 3. This invention proposes an image-based method for determining the initial crack strength of UHPC flexural tests. By synchronizing the initial crack times of multiple sources and taking the earliest as the comprehensive initial crack time, a multi-source initial crack feature dataset including surface strain, internal strain anomaly, and interface debonding rate is constructed. This dataset is then input into a neural network model to calculate the initial crack flexural strength. This method comprehensively considers the damage conditions of different parts of the specimen, effectively improving the accuracy and reliability of the initial crack flexural strength determination. This comprehensive determination method can more realistically reflect the initial crack characteristics of UHPC materials under actual stress conditions. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the image-based method for determining the initial crack strength of UHPC flexural test according to the present invention;

[0063] Figure 2 This is a flowchart illustrating the principle of the image-based method for determining the initial crack strength of UHPC flexural tests according to the present invention.

[0064] Figure 3 This is a diagram of the system architecture for determining the initial crack strength of the UHPC flexural test based on the image method of the present invention. Detailed Implementation

[0065] Example 1:

[0066] Please see Figure 1 and Figure 2 The present invention provides an embodiment of a method for determining the initial crack strength of a UHPC flexural test based on an image method, the method comprising steps S1 to S5, including the following steps:

[0067] S1: Spray random speckle on the surface of UHPC specimens, embed fiber Bragg grating sensor array inside the specimens, and simultaneously perform X-ray computed tomography on UHPC specimens of the same batch to establish a damage evolution model of the fiber-matrix interface transition zone.

[0068] Furthermore, the process of spraying random speckle patterns onto the surface of the UHPC specimen includes:

[0069] First, the surface of the UHPC specimen is cleaned to remove oil, dust, and other impurities, ensuring a clean and smooth surface. Then, a spraying material is selected and evenly sprayed onto the specimen surface using a spray gun to form a randomly distributed speckle pattern. The spraying material should have good adhesion and contrast. During the spraying process, the spray gun pressure, spraying distance, and spraying speed are controlled to ensure uniform speckle distribution and appropriate density. In addition, the density and size of the speckles should meet the requirements of subsequent Digital Image Correlation (DIC) algorithm processing to ensure accurate capture of deformation information on the specimen surface.

[0070] Furthermore, the process of pre-embedding a fiber Bragg grating sensor array inside the specimen includes:

[0071] Based on the dimensions of the UHPC specimen and the expected stress distribution, the layout of the fiber Bragg grating sensor array was designed, and the position and spacing of each sensor were determined. Before casting the UHPC specimen, the fiber Bragg grating sensors were fixed in the pre-embedded mold according to the design layout to ensure that the sensor positions were accurate and that they would not shift during the casting process. When casting the UHPC material, care was taken to protect the fiber Bragg grating sensors to avoid damage, while ensuring that the optical fiber and the UHPC material were in full contact to ensure that the sensors could accurately sense the strain changes inside the specimen.

[0072] S2: Three-point bending load is applied to the specimen, speckle images of the specimen surface are acquired in real time, the strain distribution in the whole field is calculated using the DIC algorithm, and the surface initial cracking time and the coordinates of the local strain concentration area at the time of initial cracking are determined by combining the pre-set surface initial cracking criterion. At the same time, the fiber Bragg grating sensor array is used to monitor the fiber wavelength drift in real time, calculate the wavelength drift variation coefficient, and determine the internal initial cracking time.

[0073] Among them, the initial surface cracking time reflects the moment when the specimen surface first fails; the coordinates of the local strain concentration area indicate the specific location where the failure occurs; and the initial internal cracking time reflects the moment when damage first occurs inside the specimen, reflecting the changes in the material's internal properties.

[0074] Furthermore, a three-point flexural loading test was conducted on the specimen, including:

[0075] (1) Place the UHPC specimen with random speckle spraying and pre-embedded fiber Bragg grating sensor array on the support of the three-point bending tester, and adjust the position of the specimen so that the center line of the specimen coincides with the loading axis.

[0076] (2) According to the test requirements, set the loading rate. The loading rate should be able to simulate the stress situation of UHPC components in actual engineering.

[0077] (3) Start the testing machine and perform a three-point bending load test on the specimen. During the loading process, keep the loading rate stable and collect speckle images of the specimen surface and monitoring data of the fiber Bragg grating sensor array in real time.

[0078] S3: Based on the fiber-matrix interface transition zone damage evolution model and combined with the interface stress measured in real time during the loading test, calculate the interface debonding rate and determine the interface initial cracking point time according to the preset interface initial cracking judgment criteria.

[0079] Among them, the initial cracking point time at the interface reflects the moment when the bonding performance between the fiber and the matrix begins to fail, which has an important impact on the overall performance of the material.

[0080] S4: Synchronize the initial cracking time of the surface, the initial cracking time of the interior, and the initial cracking point time of the interface, and take the earliest time as the comprehensive initial cracking time;

[0081] Furthermore, the specific steps of S4 include:

[0082] (1) Establish a unified time reference to ensure that the measurement and recording of the surface initial cracking time, internal initial cracking time and interface initial cracking point time are based on the same time coordinate system;

[0083] (2) During the data processing, a time synchronization algorithm is used to calibrate and synchronize the time data obtained by different monitoring methods;

[0084] (3) By comparing the values ​​at the three time points, the earliest time is determined as the overall initial cracking time;

[0085] (4) Store and manage the synchronized time data for subsequent analysis and verification.

[0086] S5: Integrate the surface strain, internal strain anomaly, and interface debonding rate corresponding to the initial cracking time to form a multi-source initial cracking feature dataset. Substitute this dataset into the pre-trained neural network model and combine it with the weight coefficients obtained during training. Calculate the initial cracking flexural strength of UHPC by weighted summation.

[0087] The surface strain is obtained by the DIC algorithm; the internal strain anomaly is determined by the wavelength drift variation coefficient.

[0088] Furthermore, the surface strain, internal strain anomaly, and interfacial debonding rate corresponding to the initial cracking time are integrated, including:

[0089] (1) Extract the surface strain distribution data, internal strain anomaly data and interface debonding rate data corresponding to the comprehensive initial cracking time from the data obtained by real-time acquisition and calculation. Among them, the internal strain anomaly data is such as the strain abrupt change data monitored by the fiber Bragg grating sensor.

[0090] (2) Standardize the format of the extracted data to ensure that different types of data can be integrated and analyzed in the same dataset;

[0091] (3) Correlate and integrate the surface strain, internal strain anomaly and interface debonding rate data to form a complete multi-source initial crack feature dataset;

[0092] (4) Perform quality checks on the integrated multi-source primary fracture feature dataset to ensure the integrity and accuracy of the data.

[0093] Furthermore, the process of extracting the surface strain distribution data corresponding to the overall initial cracking time from the data acquired and calculated in real time includes:

[0094] (1) Associate the real-time acquired surface speckle image data with the time record of the loading test. Since the image acquisition device usually records the acquisition timestamp of each image, and the loading test also records the timeline of the entire loading process, the loading time corresponding to each image can be determined by matching the timestamps;

[0095] For example, if the image acquisition frequency is 100 frames per second and the loading test starts from 0 seconds, then the time corresponding to the first frame is 0.01 seconds, the second frame corresponds to 0.02 seconds, and so on.

[0096] (2) Based on the determined initial cracking time, find the image frame that is closest to it. Since image acquisition is discrete, the initial cracking time will not exactly coincide with the acquisition time of a certain frame. Therefore, it is necessary to select the closest frame. For example, if the initial cracking time is 5.23 seconds and the image acquisition timestamps are 5.20 seconds and 5.30 seconds, then select the image frame corresponding to 5.20 seconds as the target frame image.

[0097] (3) Preprocess the target frame image, including noise reduction and contrast enhancement, to improve image quality and facilitate subsequent strain calculation;

[0098] (4) Based on the region of interest set before the loading test, determine the region in the image where strain calculation needs to be performed. It should be noted that the selection of ROI should avoid parts such as the edge of the specimen and the fixture that may affect the accuracy of strain calculation.

[0099] (5) Within the region of interest, the DIC algorithm is used to calculate the surface strain distribution. The basic principle of the DIC algorithm is to calculate the displacement field by comparing the gray-level distribution of subsets in the images before and after deformation, and then obtain the strain field. Specifically, it includes:

[0100] A series of subsets are selected in the image before deformation, each subset containing a certain number of pixels; then, the region in the image after deformation that is most similar to the gray-level distribution of the subset before deformation is searched to determine the displacement of the subset; the distribution of each component of the strain tensor in the region of interest is calculated based on the displacement field using numerical differentiation methods.

[0101] (6) Extract the distribution data of each component of the calculated strain tensor in the region of interest to form a surface strain distribution dataset. The surface strain distribution dataset can be stored in the form of a matrix or an array, with each element corresponding to the strain value at a specific location in the region of interest.

[0102] Furthermore, the process of extracting internal strain anomaly data corresponding to the overall initial cracking time from the data acquired and calculated in real time includes:

[0103] (1) Associate the wavelength drift data collected in real time by the fiber Bragg grating sensor with the time record of the loading test. The data acquisition system of the fiber Bragg grating sensor usually records the timestamp of each data point. By matching it with the time axis of the loading test, the loading time corresponding to each wavelength drift data is determined.

[0104] (2) Based on the overall initial cracking time, find the fiber Bragg grating sensor data point that is closest to the time. Since the data acquisition is also discrete, it is also necessary to select the data point that is closest to the overall initial cracking time.

[0105] (3) Based on the wavelength-strain relationship formula of the fiber Bragg grating sensor, the selected wavelength drift data is converted into strain values, as shown in the formula: ,in, Indicates the strain value. Indicates the wavelength shift. Indicates the strain sensitivity coefficient. Indicates the center wavelength of the fiber Bragg grating;

[0106] (4) Determine whether the strain value is abnormal based on the pre-set strain anomaly judgment criteria. The judgment criteria are determined based on material properties or engineering experience. For example, a strain threshold is set. When the strain value exceeds the strain threshold, it is judged as strain anomaly.

[0107] If the strain is determined to be abnormal, the strain value and the corresponding fiber Bragg grating sensor position information are extracted as internal strain abnormality data; if it is not determined to be abnormal, the data of other fiber Bragg grating sensors near the composite initial crack time are checked.

[0108] Furthermore, the input is fed into the trained neural network model to obtain weight coefficients, including:

[0109] (1) Based on the characteristics of the multi-source initial fracture feature dataset and the experimental requirements, a convolutional neural network model was selected;

[0110] (2) Use known UHPC flexural strength test data to train the convolutional neural network model. During the training process, use appropriate optimization algorithms and loss functions to adjust the model's weights and bias parameters so that the convolutional neural network model can accurately learn the mapping relationship between multi-source data and initial crack flexural strength.

[0111] (3) After training, the real-time multi-source initial crack feature dataset is input into the trained convolutional neural network model. The convolutional neural network model calculates the weight coefficients corresponding to each data source based on the input multi-source data. The weight coefficients reflect the degree of influence of different data sources on the initial crack flexural strength.

[0112] Furthermore, the image-based method for determining the initial crack strength of UHPC in flexural tests includes, after calculating the initial crack strength of UHPC, a process of further analysis and application of the test results:

[0113] (1) Statistical analysis was conducted on the initial crack flexural strength test results of UHPC specimens of different batches and different mix proportions to identify the main factors affecting the initial crack flexural strength;

[0114] (2) Based on the test results and analysis results, optimize and improve the mix design and production process of UHPC material to improve the initial crack flexural strength performance of UHPC;

[0115] (3) Provide the test results and analysis reports to relevant engineering design and construction personnel to provide a reference for the application of UHPC in engineering structures and ensure the safety and reliability of engineering structures.

[0116] The process of performing X-ray computed tomography on UHPC specimens from the same batch to establish a damage evolution model of the fiber-matrix interface transition zone includes:

[0117] S1.1: Select an X-ray computed tomography (CT) scanner and adjust the scanning parameters according to the size and density of the UHPC specimen; the scanning parameters include X-ray energy, scanning speed, and slice thickness;

[0118] S1.2: Place the UHPC specimens of the same batch on the scanning table of the scanning equipment for all-round scanning. The scanning equipment converts the acquired transmission signals into a three-dimensional structural image of the specimen through the data processing system.

[0119] S1.3: Perform edge detection on the three-dimensional structural image inside the specimen obtained by scanning to identify the fiber-matrix interface transition zone information. Edge detection is a prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0120] S1.4: Based on the identified fiber-matrix interface transition zone information and combined with the mechanical property parameters of UHPC material, a finite element model of the interface transition zone is established;

[0121] Furthermore, the specific steps in S1.4 include:

[0122] (1) Obtain the structural features of the identified fiber-matrix interface transition zone, including the thickness and shape of the transition zone;

[0123] (2) Based on the structural feature information of the identified interface transition area, use CAD software to establish the geometric model of the interface transition area;

[0124] (3) Determine the material parameters of the interface transition zone. It should be noted that the material properties of the interface transition zone are different from those of the fiber and the matrix. They need to be determined by experimental testing. For example, the elastic modulus of the interface transition zone can be measured by nanoindentation experiment.

[0125] (4) Assign the material parameters of the interface transition zone obtained by the experiment to the geometric model so that it becomes a finite element model with actual material properties. At the same time, it is also necessary to define the material properties of the fiber and the matrix.

[0126] (5) Select the hexahedral element type for mesh generation based on the geometry and stress characteristics of the interface transition zone;

[0127] (6) Discretize the geometric model into a finite number of elements;

[0128] (7) Determine the boundary conditions of the interface transition zone model according to the actual working conditions. For example, if the interface transition zone is in a fixed support structure, the corresponding fixed constraints need to be set in the model.

[0129] (8) Apply loads to the model according to the actual stress conditions. The loads can be force, pressure, or temperature. For example, if the interface transition zone is subjected to external tensile force, the corresponding tensile load needs to be applied to the model.

[0130] (9) Select dynamic analysis based on the nature of the problem;

[0131] (10) Use the solver to perform calculations and output the stress, strain, and displacement results of the interface transition zone;

[0132] (11) Compare the calculation results with the theoretical analysis or experimental results to verify the rationality of the model. If the calculation results differ greatly from the theoretical analysis or experimental results, it is necessary to check whether there are any problems with the geometry, material properties, boundary conditions, mesh generation, etc. of the model, and make corresponding adjustments.

[0133] S1.5: In the finite element model of the interface transition zone, the material nonlinearity, geometric nonlinearity and the interaction between the fiber and the matrix in the interface transition zone are considered. Boundary conditions and loading methods are set. The stress-strain relationship of the UHPC specimen during the flexural loading test is simulated by calculating the established finite element model of the interface transition zone.

[0134] Furthermore, the specific steps in S1.5 include:

[0135] (1) Open the finite element software, import the established interface transition zone finite element model, define the material parameters of the interface transition zone, fiber and matrix in the software, including basic parameters such as elastic modulus, Poisson's ratio, density, as well as material nonlinearity and geometric nonlinearity related parameters. At the same time, set the contact parameters and friction coefficient between the fiber and the matrix.

[0136] (2) Based on the actual test conditions, set symmetric constraint boundary conditions in the model, and set loading curves and loading parameters. For example, for rectangular cross-section specimens, set symmetric constraints on the symmetric plane perpendicular to the loading direction, and only calculate half of the model.

[0137] (3) Select the solver and set the convergence criterion, number of iterations, number of substeps and other solution parameters;

[0138] (4) Start the solver to perform the calculation, and monitor the iteration process and real-time changes of the results during the calculation;

[0139] (5) After the calculation is completed, extract the stress-strain data and stress distribution cloud map from the result file;

[0140] (6) Plot the stress-strain curve, analyze the nonlinear characteristics of the curve and the stress distribution, and conduct in-depth analysis in combination with model parameters and loading conditions.

[0141] S1.6: Based on the finite element analysis results, the stress distribution, strain distribution, and damage evolution of the interface transition zone under different loading stages are obtained. A damage evolution model of the fiber-matrix interface transition zone is established. This model can describe the variation law of interface damage with loading time and loading stress. The damage evolution is described by defining damage variables. The damage variables include plastic strain and energy dissipation index.

[0142] Furthermore, the specific steps in S1.6 include:

[0143] (1) Run the finite element analysis program to obtain stress, strain and damage data of the interface transition zone under different loading stages;

[0144] (2) Extract stress, strain and damage data, draw stress cloud map, strain cloud map and damage cloud map, analyze the damage initiation point and propagation path, and summarize the damage evolution law;

[0145] (3) Select the damage evolution model form and determine the model parameters;

[0146] (4) Compare the model calculation results with the finite element analysis results and experimental data to verify the accuracy of the model. If the model is inaccurate, adjust the model parameters or model form and verify it again;

[0147] (5) Apply the established damage evolution model to the analysis and prediction of actual engineering problems and evaluate the damage status of the interface transition zone.

[0148] The calculation of the full-field strain distribution using the DIC algorithm includes:

[0149] A1: Preprocess the collected speckle images on the surface of the specimen;

[0150] A2: Select an initial reference image and a deformed speckle image from the pre-processed speckle images on the specimen surface; the initial reference image is a random speckle image of the specimen surface sprayed before the loading test; the deformed speckle image is a speckle image of the specimen surface after deformation, collected at set time intervals during the loading test.

[0151] A3: Select a predetermined number of sub-regions on the initial reference image and the deformed speckle image respectively. The size and shape of the sub-regions are determined according to the speckle characteristics and deformation of the specimen surface.

[0152] A4: The DIC algorithm is used to match sub-regions on the initial reference image and the deformed speckle image. By calculating the correlation coefficient between the sub-regions, the displacement vector of each sub-region in the deformed speckle image relative to the corresponding sub-region in the initial reference image is determined. The DIC algorithm is prior art in this field and is not an inventive solution of this application. It will not be described in detail here.

[0153] A5: Based on the displacement vector, the strain distribution of the specimen surface is calculated using the least squares method to obtain the strain information of the specimen under different loading stages. The least squares method is prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0154] The process of determining the initial cracking time and the coordinates of the local strain concentration area at the time of initial cracking by combining a pre-set surface cracking criterion includes:

[0155] B1: Based on the characteristics of UHPC material and test requirements, the surface initial cracking criterion is pre-set based on the strain threshold;

[0156] B2: In the calculated full-field strain distribution, monitor the strain value and strain gradient change at each location in real time;

[0157] B3: When the strain value at any location exceeds the preset surface initial cracking criterion, record the time at this moment as the surface initial cracking time. At the same time, determine the coordinates of the location that exceeds the preset surface initial cracking criterion as the coordinates of the local strain concentration area at the time of initial cracking, and store them.

[0158] The method of using a fiber Bragg grating sensor array to monitor the fiber wavelength drift in real time, calculate the wavelength drift variation coefficient, and determine the internal initial cracking time includes:

[0159] C1: Connect the fiber Bragg grating sensor array to the fiber Bragg grating demodulator via optical fiber;

[0160] C2: Before the loading test begins, the fiber Bragg grating sensor is initially calibrated, and the initial wavelength value of each fiber Bragg grating sensor under stress-free conditions is recorded.

[0161] C3: During the bending load test, the fiber Bragg grating demodulator acquires the wavelength data of the fiber Bragg grating sensor in real time, processes the acquired wavelength data to obtain the current wavelength value, and calculates the wavelength drift of each fiber Bragg grating sensor; the wavelength drift is equal to the current wavelength value minus the initial wavelength value.

[0162] C4: Perform statistical analysis on the wavelength drift of all fiber Bragg grating sensors in the fiber Bragg grating sensor array, and calculate the average and standard deviation of the wavelength drift.

[0163] C5: Calculate the wavelength drift coefficient of variation based on the average and standard deviation of the wavelength drift; the wavelength drift coefficient of variation is equal to the standard deviation of the wavelength drift divided by the average of the wavelength drift.

[0164] C6: During the flexural loading test, monitor the changes in the wavelength drift variation coefficient in real time;

[0165] C7: When the wavelength drift variation coefficient exceeds the preset internal initial cracking threshold, record the time at this time as the internal initial cracking time; the internal initial cracking threshold is determined based on experimental experience and the characteristics of UHPC material.

[0166] The process of calculating the interface debonding rate based on the fiber-matrix interface transition zone damage evolution model and the interfacial stress measured in real time during the loading test includes:

[0167] D1: During the flexural loading test, stress values ​​in the interface transition zone are measured in real time using stress sensors arranged on or inside the specimen.

[0168] D2: Input the stress value of the interface transition zone measured in real time into the pre-established fiber-matrix interface transition zone damage evolution model;

[0169] D3: Calculate the degree of damage in the interface transition zone at different loading times based on the damage variables and stress-strain relationship defined in the fiber-matrix interface transition zone damage evolution model.

[0170] D4: Based on the calculated degree of damage, combined with the geometry and size of the interface transition zone, calculate the interface debonding rate; the interface debonding rate is the ratio of the debonded area of ​​the interface transition zone to the total area.

[0171] The step of determining the initial interface cracking point time according to the preset interface initial cracking judgment criteria includes:

[0172] E1: Based on the characteristics and test requirements of UHPC materials, a pre-defined standard for judging initial interface cracking is established; the standard for judging initial interface cracking is set based on the interface debonding rate threshold index.

[0173] E2: Monitor the changes in interface debonding rate and interface stress in real time from the calculated interface debonding rate and the real-time measured interface stress data.

[0174] E3: When the interface debonding rate exceeds the preset interface debonding rate threshold, record the time at this moment as the interface initial cracking point time.

[0175] The initial crack flexural strength of UHPC is calculated by weighted summation, including:

[0176] S5.1: Obtain the constructed multi-source initial fracture feature dataset and the obtained weight coefficients;

[0177] S5.2: Based on the weight coefficients obtained from the neural network model, the surface strain, internal strain anomaly, and interface debonding rate in the multi-source initial crack feature dataset are weighted and summed to obtain the initial crack flexural strength of UHPC; the surface strain is obtained by the DIC algorithm; the internal strain anomaly is determined by the wavelength drift variation coefficient; the initial crack flexural strength of UHPC comprehensively considers the damage information of the specimen surface, interior, and interface, and can more accurately reflect the initial crack performance of UHPC material during flexural loading. The neural network model is prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0178] S5.3: Output the initial crack flexural strength of UHPC to the test report or display interface and store it in the database.

[0179] Example 2:

[0180] Please see Figure 3 Another embodiment of the present invention provides: an image-based UHPC flexural strength initial crack determination system, comprising:

[0181] Data acquisition module, initial crack characteristic determination module, initial crack time determination module, strength calculation module;

[0182] The data acquisition module is used to collect various data of UHPC specimens during the flexural loading test.

[0183] The initial crack feature determination module is used to determine the initial crack features of the surface, interior and interface based on the collected data.

[0184] The initial cracking time determination module is used to synchronize the surface initial cracking time, the internal initial cracking time, and the interface initial cracking point time to determine the overall initial cracking time.

[0185] The strength calculation module is used to construct a multi-source initial crack feature dataset and use a neural network model to calculate the initial crack flexural strength of UHPC.

[0186] The data acquisition module includes: an image acquisition unit and a data acquisition unit;

[0187] The image acquisition unit is used to acquire speckle images of the UHPC specimen surface in real time at set time intervals during the three-point flexural loading test after spraying random speckle patterns onto the specimen surface. These images are then used to calculate the full-field strain distribution using the DIC algorithm.

[0188] The data acquisition unit is used to monitor the fiber wavelength drift of the fiber Bragg grating sensor array embedded inside the specimen in real time, providing data support for determining the internal initial cracking time.

[0189] The initial crack feature determination module includes: a surface initial crack feature determination unit, an internal initial crack feature determination unit, and an interface initial crack feature determination unit;

[0190] The surface initial crack feature determination unit is used to process the acquired speckle image of the specimen surface using the DIC algorithm, calculate the strain distribution across the entire field, and determine the surface initial crack time and the coordinates of the local strain concentration area at the time of initial cracking.

[0191] The internal initial crack feature determination unit is used to process the fiber wavelength drift data collected by the fiber Bragg grating sensor array, calculate the wavelength drift variation coefficient, and determine the internal initial crack time.

[0192] The interface initial crack feature determination unit is used to calculate the interface debonding rate and determine the time of the interface initial cracking point based on the damage evolution model of the fiber-matrix interface transition zone and the interface stress measured in real time during the loading test.

[0193] The initial cracking time determination module includes: a time synchronization unit and a time determination unit;

[0194] The time synchronization unit is used to establish a time synchronization mechanism to ensure that the records of surface initial cracking time, internal initial cracking time, and interface initial cracking point time have a unified time reference.

[0195] The time determination unit is used to compare the recorded surface initial crack time, internal initial crack time, and interface initial crack point time, and select the earliest time as the comprehensive initial crack time.

[0196] The intensity calculation module includes: a dataset construction unit and a model calculation unit;

[0197] Dataset construction unit, used to construct a multi-source initial fracture feature dataset containing multi-source information;

[0198] The model calculation unit is used to input the constructed multi-source initial crack feature dataset into the trained neural network model to obtain weight coefficients, and then calculate the initial crack flexural strength of UHPC by weighted summation.

[0199] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the present invention. All of these variations are within the protection scope of the present invention.

Claims

1. A method for determining the initial crack strength of UHPC flexural test based on image method, characterized in that, include: Random speckle was sprayed on the surface of the UHPC specimen, and a fiber Bragg grating sensor array was pre-embedded inside the specimen. At the same time, X-ray computed tomography was performed on the same batch of UHPC specimens to establish a damage evolution model of the fiber-matrix interface transition zone. Three-point flexural loading tests were conducted on the specimens, and speckle images of the specimen surface were acquired in real time. The strain distribution across the entire field was calculated using the DIC algorithm. Combined with the pre-set surface initial cracking criterion, the surface initial cracking time and the coordinates of the local strain concentration area at the time of initial cracking were determined. At the same time, the fiber Bragg grating sensor array was used to monitor the fiber wavelength drift in real time, and the wavelength drift variation coefficient was calculated to determine the internal initial cracking time. Based on the fiber-matrix interface transition zone damage evolution model and combined with the interface stress measured in real time during the loading test, the interface debonding rate is calculated, and the interface initial cracking point time is determined according to the preset interface initial cracking judgment criteria. The earliest of the following times was taken as the overall initial crack time: the initial crack time of the surface, the initial crack time of the interior, and the initial crack point time of the interface. The surface strain, internal strain anomaly, and interface debonding rate corresponding to the initial cracking time are integrated to form a multi-source initial cracking feature dataset. This dataset is then substituted into a pre-trained neural network model and combined with the weight coefficients obtained during training. The initial cracking flexural strength of UHPC is calculated by weighted summation.

2. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 1, characterized in that, The process of performing X-ray computed tomography on UHPC specimens from the same batch to establish a damage evolution model of the fiber-matrix interface transition zone includes: Select an X-ray computed tomography (CT) scanner and adjust the scanning parameters according to the size and density of the UHPC specimen; the scanning parameters include X-ray energy, scanning speed, and slice thickness. The UHPC specimens of the same batch were placed on the scanning table of the scanning equipment for all-round scanning. The scanning equipment converted the acquired transmission signals into a three-dimensional structural image of the specimen through the data processing system. Edge detection is performed on the three-dimensional structural image of the specimen obtained by scanning to identify the information of the fiber-matrix interface transition zone; Based on the identified fiber-matrix interface transition zone information, and combined with the mechanical property parameters of UHPC material, a finite element model of the interface transition zone is established. In the finite element model of the interface transition zone, the material nonlinearity, geometric nonlinearity and the interaction between the fiber and the matrix in the interface transition zone are considered. Boundary conditions and loading methods are set. By calculating the established finite element model of the interface transition zone, the stress-strain relationship of the UHPC specimen during the flexural loading test is simulated. Based on the finite element analysis results, the stress distribution, strain distribution, and damage evolution of the interface transition zone under different loading stages were obtained, and a damage evolution model of the fiber-matrix interface transition zone was established. The damage evolution was described by defining damage variables, which included plastic strain and energy dissipation indices.

3. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 2, characterized in that, The calculation of the full-field strain distribution using the DIC algorithm includes: The collected speckle images of the specimen surface are preprocessed; Select an initial reference image and a deformed speckle image from the pre-processed speckle images on the specimen surface; the initial reference image is a random speckle image of the specimen surface sprayed before the loading test; the deformed speckle image is a speckle image of the specimen surface after deformation, collected at set time intervals during the loading test. A predetermined number of sub-regions are selected on the initial reference image and the deformed speckle image, respectively; The DIC algorithm is used to match sub-regions on the initial reference image and the deformed speckle image. By calculating the correlation coefficient between the sub-regions, the displacement vector of each sub-region in the deformed speckle image relative to the corresponding sub-region in the initial reference image is determined. Based on the displacement vector, the full-field strain distribution on the specimen surface is calculated using the least squares method, thus obtaining strain information of the specimen under different loading stages.

4. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 3, characterized in that, The process of determining the initial cracking time and the coordinates of the local strain concentration area at the time of initial cracking by combining a pre-set surface cracking criterion includes: Based on the characteristics of UHPC materials and experimental requirements, a surface initial cracking criterion is pre-set based on the strain threshold. In the calculated full-field strain distribution, the strain value and strain gradient change at each location are monitored in real time; When the strain value at any location exceeds the preset surface initial cracking criterion, the time at this moment is recorded as the surface initial cracking time. At the same time, the coordinates of the location that exceeds the preset surface initial cracking criterion are determined as the coordinates of the local strain concentration area at the time of initial cracking and stored.

5. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 4, characterized in that, The method of using a fiber Bragg grating sensor array to monitor the fiber wavelength drift in real time, calculate the wavelength drift variation coefficient, and determine the internal initial cracking time includes: Connect the fiber Bragg grating sensor array to the fiber Bragg grating demodulator via optical fiber; Before the loading test begins, the fiber Bragg grating sensors are initially calibrated, and the initial wavelength value of each fiber Bragg grating sensor is recorded under stress-free conditions. During the bending load test, the fiber Bragg grating demodulator acquires the wavelength data of the fiber Bragg grating sensors in real time, processes the acquired wavelength data to obtain the current wavelength value, and calculates the wavelength drift of each fiber Bragg grating sensor; the wavelength drift is equal to the current wavelength value minus the initial wavelength value. Statistical analysis was performed on the wavelength drift of all fiber Bragg grating sensors in the fiber Bragg grating sensor array, and the average value and standard deviation of the wavelength drift were calculated. The wavelength drift coefficient of variation is calculated based on the average value and standard deviation of the wavelength drift; the wavelength drift coefficient of variation is equal to the standard deviation of the wavelength drift divided by the average value of the wavelength drift. During the flexural loading test, the changes in the wavelength drift variation coefficient were monitored in real time. When the wavelength drift variation coefficient exceeds the preset internal initial crack determination threshold, the time at this moment is recorded as the internal initial crack time.

6. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 5, characterized in that, The process of calculating the interface debonding rate based on the fiber-matrix interface transition zone damage evolution model and the interfacial stress measured in real time during the loading test includes: During the flexural loading test, stress sensors arranged on or inside the specimen are used to measure the stress value in the interface transition zone in real time. The stress values ​​of the interface transition zone measured in real time are input into the pre-established fiber-matrix interface transition zone damage evolution model. Based on the damage variables and stress-strain relationship defined in the fiber-matrix interface transition zone damage evolution model, the damage degree of the interface transition zone at different loading times is calculated. Based on the calculated degree of damage, combined with the geometry and size of the interface transition zone, the interface debonding rate is calculated; the interface debonding rate is the ratio of the debonded area of ​​the interface transition zone to the total area.

7. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 6, characterized in that, The step of determining the initial interface cracking point time according to the preset interface initial cracking judgment criteria includes: Based on the characteristics and test requirements of UHPC materials, a pre-defined standard for judging initial interface cracking is established; the standard for judging initial interface cracking is set based on the interface debonding rate threshold index. The changes in interface debonding rate and interface stress are monitored in real time from the calculated interface debonding rate and the real-time measured interface stress data. When the interface debonding rate exceeds the preset interface debonding rate threshold, the time at this moment is recorded as the interface initial cracking point time.

8. The method for determining the initial crack strength of UHPC flexural test based on image method as described in claim 7, characterized in that, The initial crack flexural strength of UHPC is calculated by weighted summation, including: Obtain the constructed multi-source initial fracture feature dataset and the obtained weight coefficients; Based on the weighting coefficients obtained from the neural network model, the surface strain, internal strain anomaly, and interface debonding rate in the multi-source initial crack feature dataset are weighted and summed to obtain the initial crack flexural strength of UHPC; the surface strain is obtained by the DIC algorithm; the internal strain anomaly is determined by the wavelength drift variation coefficient. The initial crack flexural strength of UHPC is output to the test report or display interface and stored in the database.

9. A system for determining the initial crack strength of a UHPC flexural test based on an image-based method, used to implement the method for determining the initial crack strength of a UHPC flexural test based on an image-based method as described in any one of claims 1-8, characterized in that, include: Data acquisition module, initial crack characteristic determination module, initial crack time determination module, strength calculation module; The data acquisition module is used to collect data of the UHPC specimen during the flexural loading test; The initial crack feature determination module is used to determine the initial crack features of the surface, interior and interface based on the collected data. The initial cracking time determination module is used to synchronize the surface initial cracking time, the internal initial cracking time, and the interface initial cracking point time to determine the overall initial cracking time. The strength calculation module is used to construct a multi-source initial crack feature dataset and to calculate the initial crack flexural strength of UHPC using a neural network model.

10. The image-based UHPC flexural strength determination system as described in claim 9, characterized in that, The initial crack feature determination module includes: a surface initial crack feature determination unit, an internal initial crack feature determination unit, and an interface initial crack feature determination unit; The surface initial crack feature determination unit is used to process the collected speckle image of the specimen surface using the DIC algorithm, calculate the strain distribution in the whole field, and determine the surface initial crack time and the coordinates of the local strain concentration area at the time of initial cracking. The internal initial crack feature determination unit is used to process the fiber wavelength drift data collected by the fiber Bragg grating sensor array, calculate the wavelength drift variation coefficient, and determine the internal initial crack time. The interface initial crack feature determination unit is used to calculate the interface debonding rate and determine the time of the interface initial cracking point based on the fiber-matrix interface transition zone damage evolution model and the interface stress measured in real time during the loading test.

Images