Method for determining the fracture toughness of concrete based on the identification of porosity by bse-ia

Abstract

The present application relates to a method for determining the fracture toughness of concrete based on BSE-IA recognition porosity, comprising the steps of: obtaining a microstructure SEM image of the concrete under scanning electron microscope by scanning imaging through backscattering test after slicing treatment of the concrete sample; performing median filter denoising and histogram equalization enhancement on the microstructure SEM image of the concrete to obtain a pretreatment image; reading the gray information of the pretreatment image and drawing a gray image; further drawing a cumulative gray histogram based on the gray information of the gray image; performing second derivative processing on the cumulative gray histogram and using the first peak point of the derivative curve as the pore segmentation threshold; performing pore segmentation according to the segmentation threshold to identify all the pores in the pretreatment image and calculate the corresponding porosity; calculating the cement content according to the mix proportion of the concrete; establishing a fracture toughness model; and calculating the fracture toughness of the concrete through the fracture toughness model using the cement content and the porosity.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mechanical properties and microscopic parameter detection of building materials, and particularly relates to a method for determining concrete fracture toughness based on BSE-IA (backscattered electron image) recognition porosity. BACKGROUND

[0002] Fracture toughness is a key indicator for evaluating whether a building structure with cracks can work stably for a long time. The accurate determination of the fracture toughness value is directly related to the reasonable setting of the engineering safety factor and is an important basis for ensuring the safety and durability of building structures.

[0003] In the prior art, technicians generally calculate the fracture toughness by measuring the macroscopic peak load of the test piece during the stress process. This method has high detection cost and a complex process. The cracking of concrete cracks does not occur directly on the macroscopic level. The cause of this process is the gradual development and expansion of micro defects such as pores and micro cracks in the material, which eventually leads to the fracture phenomenon. If the cracking instability process of the test piece is accurately reflected from the mechanism level, the fracture toughness can also be accurately calculated.

[0004] Therefore, based on the above problems, there is an urgent need for a fracture toughness calculation method with high precision, convenience and low cost. SUMMARY

[0005] To solve the problem of high detection cost and complex process of the existing method, a method for determining the fracture toughness of concrete based on BSE-IA recognition porosity is proposed. The porosity is obtained by backscattered electron image analysis technology and image recognition method, and the fracture toughness is calculated by combining the known cement content with the fracture toughness model. The effective prediction of the fracture toughness of concrete based on the micro level is realized, and the method has high accuracy and convenience.

[0006] To achieve the above purpose, the present application proposes a method for determining the fracture toughness of concrete based on BSE-IA recognition porosity, which comprises:

[0007] Step 1: After slicing the concrete test piece, scanning imaging is performed by backscattering test to obtain a concrete micro-SEM image under a scanning electron microscope;

[0008] Step 2: The concrete micro-SEM image is subjected to median filter denoising and histogram equalization enhancement to obtain a preprocessed image;

[0009] Step 3: The gray scale information of the preprocessed image is read, and a gray scale image is drawn;

[0010] Step 4: Based on the gray scale information of the gray scale image, a cumulative gray scale histogram is further drawn;

[0011] Step 5: A second derivative is applied to the cumulative gray scale histogram, and the first peak point of the derivative curve is used as the pore segmentation threshold;

[0012] Step 6: Pore segmentation is performed according to the segmentation threshold, all pores in the preprocessed image are identified, and the corresponding porosity is calculated;

[0013] Step 7: The cement content is calculated according to the mix proportion of the concrete;

[0014] Step 8: A fracture toughness model is established, and the fracture toughness of the concrete is calculated by the fracture toughness model using the cement content and the porosity.

[0015] The gray scale information in step 3 is a property of the image, and different phases in the picture have different gray scale values. In cement-based materials, the gray scale of pores is the smallest, followed by various hydrated substances and some unreacted cement clinker.

[0016] In step 5, the peak point of the second derivative is used as the basis for evaluation, which is mainly to reduce the influence of human subjective factors in selecting the threshold.

[0017] Further, the fracture toughness model in step 8 is represented by formula (6):

[0018]

[0019] where K IC represents the actual predicted fracture toughness, K IC,0 is the theoretical fracture toughness value of the material in an ideal pore-free state; C represents the cement content ratio (0 < C ≤ 1); P is the porosity, B S is the sensitive pore size distribution rate, representing the proportion of sensitive pore size in all pores, and η is the sensitivity coefficient.

[0020] The theoretical fracture toughness K IC,0 of the material in an ideal pore-free state, the cement content ratio C, the porosity P, the sensitive pore size distribution rate B S and the sensitivity coefficient B S are key parameters included in the model, realizing the quantitative calculation of the fracture toughness of the concrete. The model comprehensively considers the ideal performance, composition, microstructure defects and pore characteristic factors of the material itself, making the prediction result more in line with the actual situation. Compared with the prior art, it can more comprehensively reflect the complex factors affecting the fracture toughness of the concrete, and improve the applicability and prediction accuracy of the model.

[0021] Further, B S is calculated by gray correlation entropy, as shown in formulas (1)-(5):

[0022] First, the known fracture toughness constitutes a reference sequence X0(k), and the comparison data sequence X i (k)(k=1,2,…,m;i=1,2,…,n) is obtained by formula (1) for X0(k) and X i (k) are mean value processed, and Y0(k) and Y i (k) sequences are obtained:

[0023]

[0024] The correlation coefficient R i (k) is shown in formula (2):

[0025]

[0026] Wherein, min and max are the minimum and maximum of the difference; ρ is the resolution coefficient, ρ is 0.5;

[0027] The grey correlation entropy of the correlation coefficient sequence R i (k) is shown in formula (3):

[0028]

[0029] Wherein, P h is the density value of the mapping distribution of the correlation coefficient sequence R i (k), which is shown in formula (4):

[0030]

[0031] Wherein, h=1,2,...,m;

[0032] The grey entropy correlation degree is shown in formula (5):

[0033] D(x i )=H(R i ) / H max (5);

[0034] Wherein, D(x i ) is the grey entropy correlation degree, H max is the maximum value of the grey correlation entropy, H max =ln(m), m is the sample quantity of the data in the reference sequence and the comparison sequence.

[0035] The grey correlation entropy analysis can effectively deal with the relationship between the pore diameter and the fracture toughness with uncertainty and fuzziness. Through mean value processing, correlation coefficient calculation, grey correlation entropy and grey entropy correlation degree derivation, the sensitive pore diameter which has significant influence on the fracture toughness can be accurately identified, the deviation of subjective judgment is avoided, the determination of the sensitive pore diameter distribution rate is more reasonable and reliable, and then accurate parameter support is provided for the fracture toughness model.

[0036] Further, the step 7 is represented as formula (7):

[0037] C = (cement mass / total mass) % (7).

[0038] Further, the step 6 porosity is represented as formula (8):

[0039] P = (S P / S A ) % (8);

[0040] Wherein, S P is the total area of all pore regions in the binary image, and S A is the actual volume of the concrete micro area.

[0041] The porosity is determined by the ratio of the total area of the pore region in the binary image to the actual volume of the concrete micro area. This calculation method is based on the objective data obtained by image recognition, which can accurately reflect the proportion of pores in the concrete microstructure. Compared with the traditional porosity determination method (such as mercury intrusion method), it is more direct, convenient and low cost, and reduces the error caused by experimental operation, and provides accurate microstructure parameters for the fracture toughness model.

[0042] The K IC,0 is specifically:

[0043] For different cement aggregate concretes, multiple groups of test pieces with different porosities are prepared, the porosities of the test pieces are obtained by BSE-IA method, and the fracture toughnesses of the corresponding test pieces are measured by standard fracture test;

[0044] The porosity and the fracture toughness are regressed and analyzed, and the K IC,0 value corresponding to different cement aggregate is fitted, which is used as a reference parameter in the fracture toughness prediction model.

[0045] This method matches the value with the specific cement aggregate type, which is used as a reference parameter in the fracture toughness prediction model, can significantly improve the adaptability of the model to different types of concrete, ensure that the fracture toughness prediction result is still accurate and reliable when facing concrete with different aggregate composition, and enhance the universality of the model.

[0046] Further, the median filter denoising of step 2 comprises:

[0047] The concrete microscopic SEM image is subjected to adaptive median filtering, the window size is dynamically adjusted according to the local region noise density, the pore edge details are preferentially retained, and the isolated noise points are removed;

[0048] The image is smoothed by using a Gaussian kernel with sigma = 1.5-2.5, the mineral crystal surface reflection noise is weakened, and the gray difference between the pores and the matrix is highlighted.

[0049] The median filtering denoising can effectively improve the quality of the concrete microscopic SEM image, reduce the interference of noise on subsequent pore identification and analysis, ensure the clear presentation of the pore features in the image, lay a high-quality image foundation for accurately determining the pore segmentation threshold and calculating the porosity, and improve the accuracy of the whole method.

[0050] Through the above technical solutions, the present application has the following advantages:

[0051] (1) The present application has high accuracy. The concrete test piece is subjected to slicing treatment, and then backscattering test scanning imaging is used to obtain the concrete microscopic SEM image, which provides accurate original data for subsequent analysis. Subsequently, the image is subjected to median filtering denoising and histogram equalization enhancement pretreatment operations, effectively removes noise interference, significantly improves the image contrast, ensures the clear presentation of key microscopic features such as pores in the image, lays a solid foundation for subsequent accurate identification and analysis of pores, and greatly improves the calculation accuracy of key parameters such as porosity. When calculating the porosity, the calculation formula given in the present application determines the ratio of the total area of the pore region in the binary image to the actual volume of the concrete microscopic region. This calculation method is based on objective data obtained by image recognition, is more direct, convenient and low-cost compared with the traditional porosity determination method (such as the mercury intrusion method), significantly reduces the error caused by experimental operation, provides accurate microscopic structure parameters for the fracture toughness model, and further ensures the detection accuracy.

[0052] (2) The present application is relatively simple to operate. The porosity and fracture toughness can be obtained by briefly sampling and testing in actual engineering, which saves the tedious steps of designing experiments for verification. It is helpful to quickly evaluate the safety and stability of concrete in engineering practice, and the porosity can also be designed according to the model to achieve a better microscopic structure, which provides strong support for concrete mix proportion design and quality control. DETAILED DESCRIPTION

[0053] Figure 1 The step flow chart of the method for determining the concrete fracture toughness based on BSE-IA identification of porosity of the present application;

[0054] Figure 2A theoretical analysis diagram of the relationship between porosity and fracture toughness for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0055] Figure 3 A fracture toughness theoretical analysis diagram considering the influence of sensitive pore size for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0056] Figure 4 A concrete micro-SEM image for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0057] Figure 5 A pre-processed image for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0058] Figure 6 A gray level histogram for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0059] Figure 7 A cumulative gray level histogram for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0060] Figure 8 A second derivative curve of the cumulative gray level histogram for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity;

[0061] Figure 9 A pore identification binary image for the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity. DETAILED DESCRIPTION

[0062] Embodiment 1

[0063] As shown in Figure 1 , the method for determining the fracture toughness of concrete based on BSE-IA identification of porosity comprises:

[0064] Step 1: After slicing the concrete test piece, a backscattering test is performed to scan and image, and a concrete micro-SEM image under a scanning electron microscope is obtained (as shown in Figure 4 );

[0065] Step 2: The concrete micro-SEM image is subjected to median filtering denoising and histogram equalization enhancement to obtain a pre-processed image (as shown in Figure 5 );

[0066] Step 3: The gray scale information of the pre-processed image is read, and a gray scale image is drawn (as shown in Figure 6 );

[0067] Step 4: Based on the gray information of the gray image, further draw an accumulated gray histogram (as shown in FIG. 4B) ; Figure 7

[0068] Step 5: Adopt a second derivative processing to the accumulated gray histogram, and use the first peak point of the derivative curve as a pore segmentation threshold (as shown in FIG. 4C) ; Figure 8

[0069] Step 6: According to the segmentation threshold, perform pore segmentation to identify all pores in the preprocessed image, and calculate the corresponding porosity (as shown in FIG. 4D) ; Figure 9

[0070] Step 7: Calculate the cement content according to the mix proportion of the concrete;

[0071] Step 8: Establish a fracture toughness model, and calculate the fracture toughness of the concrete by the fracture toughness model using the cement content and the porosity.

[0072] The fracture toughness model in Step 8 is represented by formula (6) :

[0073]

[0074] wherein K IC represents the actually predicted fracture toughness, K IC,0 is the theoretical fracture toughness value of the material in an ideal non-porous state; C represents the cement content ratio (0 < C ≤ 1); P is the porosity, B S is the sensitive pore size distribution rate, representing the ratio of the sensitive pore size in all pores, and η is the sensitive coefficient.

[0075] B S is calculated by gray correlation entropy as shown in formulas (1)-(5) :

[0076] First, obtain a known fracture toughness to form a reference sequence X0(k), and use different pore sizes to form comparison data sequences X i (k) (k = 1, 2, …, m; i = 1, 2, …, n), and perform mean value processing on X0(k) and X i (k) by formula (1) to obtain Y0(k) and Y i (k) sequences:

[0077]

[0078] The correlation coefficient R i (k) is shown in formula (2) :

[0079]

[0080] ​​​Wherein, min and max are the minimum and maximum of the difference value; p is the resolution coefficient, p is 0.5;

[0081] Correlation coefficient sequence R i The grey correlation entropy of (k) is shown in formula (3):

[0082]

[0083] Wherein, P h is the correlation coefficient sequence R i (k) mapping distribution density value, as shown in formula (4):

[0084]

[0085] Wherein, h=1, 2,..., m;

[0086] The grey entropy correlation degree is shown in formula (5):

[0087] D(x i ) = H(R i ) / H max (5);

[0088] Wherein, D(x i ) is the grey entropy correlation degree, H max is the maximum value of grey correlation entropy, H max = ln(m), m is the sample quantity of data in reference sequence and comparison sequence.

[0089] The step 7 is shown in formula (7):

[0090] C = (cement mass / total mass) % (7).

[0091] The step 6 porosity is shown in formula (8):

[0092] P = (S P / S A ) % (8);

[0093] Wherein, S P is the total area of all pore regions in the binary image, S A is the actual volume of concrete micro area.

[0094] The K IC,0 is specifically:

[0095] For different cement aggregate concrete, a plurality of test pieces with different porosities are prepared, the porosities of the test pieces are obtained by BSE-IA method, and the fracture toughness of the corresponding test pieces is measured by standard fracture test;

[0096] The porosity and the fracture toughness are subjected to regression analysis, and K IC,0 values corresponding to different cement aggregates are fitted, which are used as reference parameters in the fracture toughness prediction model.

[0097] The median filtering denoising in step 2 comprises:

[0098] The concrete microscopic SEM image is subjected to adaptive median filtering, the window size is dynamically adjusted according to the local region noise density, the pore edge details are preferentially retained, and the isolated noise points are removed.

[0099] The image is smoothed by using a Gaussian kernel with sigma = 1.5-2.5, the mineral crystal surface reflection noise is weakened, and the gray difference between the pores and the matrix is highlighted.

[0100] In the embodiment, the K IC,0 values under different aggregates are shown in Table 1:

[0101] Table 1 K IC,0 (MPa·m 1 / 2 )

[0102]

[0103] The sensitivity coefficients η under different aggregates are shown in Table 2:

[0104] Table 2 η under different aggregates

[0105]

[0106]

[0107] The B S under different aggregates is calculated and shown in Table 3:

[0108] Table 3 B S

[0109]

[0110] Example 2

[0111] In order to verify that the method of identifying the porosity based on BSE-IA to determine the fracture toughness of concrete in Example 1 has good effect, a control experiment is set up by taking coal gangue concrete as an example to verify the effect of the present application.

[0112] Materials and specimen preparation:

[0113] The 42.5 ordinary portland cement used in the experiment, the coarse aggregate is coal gangue, the particle size range is controlled in 5-10mm, the fine aggregate is river sand, the fineness modulus is 2.6, and the water reducing agent is poly carboxylic acid ether high efficiency water reducing agent.

[0114] According to different mixing ratio design, a plurality of groups of coal gangue concrete test pieces are prepared, each group of test pieces is cured under the same curing condition to a specified age period, so as to ensure the accuracy and comparability of experimental data. In order to verify the effect of the present application, the following three groups of experiments are carried out:

[0115] ① The first group is based on the three-point bending test of the boundary effect model to determine the fracture toughness;

[0116] ② The second group is based on the formula proposed in the present application and the conventional pore test method (mercury intrusion method) to calculate the fracture toughness;

[0117] ③ The third group is determined by the formula proposed in the present application and the empirical coefficient value obtained from a large amount of data in combination with the backscattering image recognition to determine the fracture toughness.

[0118] The experimental results of experiment ① are shown in Table 4:

[0119] Table 4 Three-point bending test results of coal gangue concrete

[0120]

[0121] In experiment ②, only the relationship between porosity and fracture toughness is considered, and the formula K Figure 2 The formula K IC = K IC,0 C(1-P) is used in combination with K IC ( Table 4) obtained from the three-point bending test and the pore test results ( Table 5) to fit K IC,0 The results are shown in Table 6:

[0122] Table 5 Microscopic test results of coal gangue concrete

[0123]

[0125] Table 6 K IC,0 value of coal gangue concrete

[0126]

[0127] In experiment ②, the influence of pore size distribution is comprehensively considered, and the following analysis is carried out based on the grey correlation entropy, and the influence results of each pore size are shown in Table 7:

[0128] Table 7 Grey entropy correlation degree of factors such as pore size distribution on fracture toughness

[0129]

[0130] From Table 7, the grey entropy correlation degree values of each factor are close, proving that these parameters are sensitive parameters affecting the fracture toughness of concrete, ranked from large to small: D(r<20nm)>D(r=20-100nm)=D(C)>D(P)>D(r>500nm)>D(r=100-500nm). Among them, cement content C and porosity P are important influencing factors. In addition to porosity P and cement content C, among the pore size distribution factors, the proportion of harmless pores (r<20nm) is also an important factor affecting the fracture toughness of coal gangue concrete, so B S is the proportion of this pore size in the total pore size. Based on the above research, the fracture toughness of coal gangue concrete considering the influence of sensitive pore size is shown in FIG. 8, Figure 3 Figure 3 (a) P=0, (b) P≠0 and (c) P≠0 (B S ). Regression analysis of test results is performed using the formula , as shown in Table 8:

[0131] Table 8 Regression results of fracture toughness-pore size model of coal gangue concrete

[0132]

[0133]

[0134] Finally, based on the calculated K IC,0 , B S and η, the fracture toughness is obtained by the prediction formula as shown in Table 9:

[0135] Table 9 Fracture toughness results predicted based on BSE-IA

[0136]

[0137] Experiment ③ calculates the fracture toughness based on backscattered image recognition and empirical value of formula coefficient. The porosity acquisition method of experiment ③ is more simple than that of experiment ②. By combining the porosity obtained by image recognition with the empirical value of the prediction formula parameter, the prediction model is brought in, and the results are compared with the results obtained by three-point bending test, as shown in Table 10.

[0138] Table 10 Comparison of fracture toughness prediction results and test results

[0139]

[0140]

[0141] Experiment ① derives the fracture toughness based on the macroscopic response of the specimen under stress failure. The load-displacement curve is obtained by three-point bending test, and the boundary effect model is combined to finally determine the fracture toughness.

[0142] ​In experiment 3, porosity is accurately extracted from the backscattered image. Compared with experiment 1, experiment 3 is simple, low in cost and has better accuracy. The difference between the calculated values of fracture toughness in the two experiments is small, so the method has reliability in calculating fracture toughness.

[0143] The above-described embodiments are only preferred embodiments of the present application, and are not intended to limit the scope of the application. Any equivalent changes or modifications made in accordance with the structure, features and principles described in the patent scope of the present application should be included in the patent scope of the present application.

Claims

1. A method for determining the fracture toughness of concrete based on the identification of porosity by BSE-IA, characterized in that, The application relates to a method for calculating the fracture toughness of concrete, and belongs to the technical field of concrete. Step 1: after slicing treatment of a concrete test piece, a back scattering test is used to scan and image, and a concrete microscopic SEM image under a scanning electron microscope is obtained; Step 2: the concrete microscopic SEM image is subjected to median filtering denoising and histogram equalization enhancement, and a pretreatment image is obtained; Step 3: the gray information of the pretreatment image is read, and a gray image is drawn; Step 4: based on the gray information of the gray image, a cumulative gray histogram is further drawn; Step 5: the cumulative gray histogram is subjected to second derivative processing, and the first peak point of the derivative curve is used as a pore segmentation threshold value; Step 6: pores in the pretreatment image are all identified according to the segmentation threshold value, and the corresponding porosity is calculated; Step 7: the cement content is calculated according to the mixing proportion of the concrete; Step 8: a fracture toughness model is established, and the cement content and the porosity are used to calculate the fracture toughness of the concrete through the fracture toughness model.

2. The method for determining the fracture toughness of concrete based on BSE-IA recognition of porosity according to claim 1, characterized in that, The fracture toughness model in step 8 is expressed as formula (6): where K IC represents the actual predicted fracture toughness, K IC,0 is the theoretical fracture toughness value of the material in an ideal pore-free state; C represents the cement content ratio (0 < C ≤ 1); P is the porosity, B S is the sensitive pore size distribution rate, representing the proportion of sensitive pore size in all pores, and η is the sensitivity coefficient.

3. The method for determining the fracture toughness of concrete based on BSE-IA recognition of porosity according to claim 2, characterized in that, B is calculated by grey correlation entropy S As shown in equations (1)-(5): First, the known fracture toughness constitutes a reference series X0(k), and the comparative data series X i (k)(k = 1, 2, …, m; i = 1, 2, …, n) are obtained by formula (1) for X0(k) and X i (k) are mean value processed, and Y0(k) and Y i (k) series are obtained. Correlation coefficient R i (k) as shown in equation (2): Wherein, min and max are the minimum and maximum values of the difference; and rho is a resolution coefficient, and rho is 0.5; The correlation coefficient sequence R i The grey correlation entropy of (k) is shown in formula (3): where P h is the correlation coefficient sequence R i (k) is the density value of the mapping distribution, as expressed in equation (4): Wherein, h=1, 2,..., m; The gray entropy correlation degree is shown in formula (5): D(x i ) = H(R i ) / H max (5); Wherein, D(x i ) is the gray entropy correlation degree, H max is the maximum value of the gray correlation entropy, H max = ln(m), m is the sample quantity of data in the reference series and the comparison series.

4. The method for determining the fracture toughness of concrete based on BSE-IA recognition of porosity according to claim 2, characterized in that, The step 7 is shown in formula (7): C=(cement mass / total mass) % (7).

5. The method for determining the fracture toughness of concrete based on BSE-IA recognition of porosity according to claim 2, characterized in that, The porosity in step 6 is shown in formula (8): P = (S P / S A ) % (8); where S P is the total area of all pore regions in the binary image, S A is the actual volume of the concrete micro-region.

6. The method for determining the fracture toughness of concrete based on BSE-IA recognition of porosity as claimed in claim 2, wherein, The K IC,0 Specifically: For different cement aggregates of the concrete, a plurality of test pieces with different porosities are prepared, the porosities of the test pieces are obtained by using the BSE-IA method, and the fracture toughnesses of the corresponding test pieces are measured through a standard fracture test; The porosity and the fracture toughness were regressed and analyzed, and the K IC,0 value corresponding to different cement aggregates was fitted, which was used as a reference parameter in the fracture toughness prediction model.

7. The method for determining the fracture toughness of concrete based on BSE-IA recognition of porosity according to claim 1, characterized in that, The median filtering denoising in step 2 comprises: The concrete microscopic SEM image is subjected to adaptive median filtering, the window size is dynamically adjusted according to the local region noise density, the pore edge details are preferentially reserved, and isolated noise points are removed; A Gaussian kernel with sigma=1.5-2.5 is used to smooth the image, the mineral crystal surface reflection noise is weakened, and the gray difference between the pores and the matrix is highlighted.

Images