A method for characterizing the instability behavior of uniaxial tensile simulation curves

Abstract

The present invention relates to a method for characterizing the instability behavior of a uniaxial tensile simulation curve, comprising obtaining the true stress and true strain curves of the material, fitting the selected constitutive model test data, and correcting the goodness of fit. The corrected constitutive model formula is A_R 2 =1-[(1-R 2 )×(n+n / R 2 ) / (n‑p‑1)], the two constitutive models with the highest goodness of fit after correction are selected and integrated into a new constitutive model C. The integration formula is C=P×A+(1‑P)×B. The cross-sectional load and gauge length change process of the central area of ​​the specimen in the simulation test is obtained under the new constitutive model. The force and displacement curves after the instability behavior in the simulation test are compared with the corresponding tensile test results. The correction coefficient in the new constitutive model is adjusted, and the above two steps are repeated until the coincidence degree of the simulation instability curve and the experimental instability curve is the highest. The disadvantage that the instability prediction result of the simulation curve is quite different from that of the test due to curve fitting is overcome, and the instability behavior of the simulation curve is accurately and reliably predicted.

Description

Technical Field

[0001] The invention relates to the technical field of material performance testing, and in particular to a method for characterizing the instability behavior of a uniaxial tensile simulation curve. Background Art

[0002] In the uniaxial tensile simulation test, it is necessary to define material parameters for the established simulation model. Among the material parameter definitions, the material's mechanical property curve has the most significant impact on the simulation accuracy. However, the mechanical property curve measured by the tensile testing machine can only predict the mechanical behavior before the simulation test becomes unstable, and is powerless to characterize the stress mutation and deformation characteristics that occur after the material becomes unstable, resulting in the instability behavior in the simulation test being far from the results measured by the testing machine. Therefore, the test personnel are very eager to explore an accurate and reliable method for characterizing the instability behavior of the simulation curve to improve the prediction accuracy of the simulation model, which is crucial to the promotion and application of materials.

[0003] At present, simulation experiments usually use the test curve before the material becomes unstable to predict the stress-strain response of the material, ignoring the influence of the sudden stress change after the material becomes unstable on the simulation results. In addition, there are also cases where the test curve is fitted with a single constitutive model and then used, but the accuracy of the fitting curve is low and it is impossible to accurately characterize the instability behavior of the material. It can be seen that whether it is the test curve or the fitting curve, it can only guarantee the prediction accuracy before the material becomes unstable. The characterization of the mechanical behavior of the material after instability is often very different from the results measured by the testing machine. In addition, when judging the failure of the material, the instability behavior characterization of the simulation curve has a greater impact on the damage accumulation of the material. Therefore, it is very important to determine an accurate and reliable method for characterizing the instability behavior of the simulation curve.

[0004] A Chinese patent with publication number CN109632484B discloses a method and system for determining the true stress of an unstable section of a specimen under unidirectional tension. In the process of determining the true stress of an unstable section under unidirectional tension of a specimen, the method takes into account the fact that the calculated true stress has a large error due to the uneven deformation of the plastic deformation zone and the long unstable section during the unidirectional tension of the specimen. By adopting image acquisition and measurement means, the effective bearing area of ​​the specimen can be accurately measured and calculated, thereby obtaining the true stress of the specimen material; however, the patent has no effect on the instability behavior of the unidirectional tension simulation curve, and because image acquisition is easily affected by the external environment, errors in the measured true stress occur.

[0005] The Chinese patent with publication number CN108573099A discloses a method for obtaining a critical compressive stress instability curve of a metal plate, which includes: 1) making a plurality of specimens of different sizes; 2) performing a tensile test on each of the specimens based on a DIC strain measurement system, and calculating the primary and secondary stress distribution of each specimen in combination with a hardening / yield model; 3) drawing a cross-sectional profile curve corresponding to any frame after the surface distortion is formed in the sampling frame number of the DIC strain measurement system along the cross section of the middle area of ​​the specimen, and taking the separation point of the straight segment and the curved segment in the cross-sectional profile curve as the instability boundary position of the surface distortion zone; 4) for each specimen: extracting the surface distortion The secondary stress value at the instability boundary position is plotted, and a secondary stress value-shooting frame number curve is drawn to find the lowest point of the secondary stress, which is the critical compressive stress of surface distortion instability; 5) The critical compressive stress of surface distortion instability corresponding to the instability boundary position of the surface distortion of each specimen is plotted to obtain the critical compressive stress instability curve. The invention obtains the critical compressive stress instability curve through an experimental method; however, the patent requires the use of multiple specimens and multiple tests. The experimental process is complicated and time-consuming, and there are artificial differences in determining the surface distortion zone by observing the surfaces of each specimen after stretching, which leads to inaccurate critical compressive stress instability curve and has no effect on the characterization of the simulated instability curve. Summary of the invention

[0006] The present invention provides a method for characterizing the instability behavior of a uniaxial tensile simulation curve, in which the true stress and true strain curves of a sample are fitted with a constitutive model, a more accurate new constitutive model is determined using the corrected goodness of fit, and the instability behavior of the sample in the uniaxial tensile simulation curve is accurately characterized by optimizing the correction coefficient in the new constitutive model.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A method for characterizing the instability behavior of a uniaxial tensile simulation curve comprises the following steps:

[0009] 1) Obtain the true stress and true strain curves of the material;

[0010] 2) Fit the experimental data corresponding to different constitutive models and calculate the goodness of fit of each constitutive model after correction:

[0011] A_R 2 =1-[(1-R 2 )×(n+n / R 2 ) / (np-1)](1)

[0012] Among them, A_R 2 is the goodness of fit of the modified constitutive model, R 2 is the goodness of fit of the constitutive model before modification, n is the sample size, and p is the number of fitting parameters;

[0013] 3) Select the two constitutive models whose goodness of fit is closest to 1 after correction and integrate them into the new constitutive model C. The integration formula is:

[0014] C=P×A+(1-P)×B(2)

[0015] Among them, P is the correction coefficient, the correction range of P is 0 to 1, and A and B are the two constitutive models whose goodness of fit is closest to 1 after correction;

[0016] 4) The new constitutive model is used to obtain the cross-sectional load and gauge length variation process in the central area of ​​the specimen in the simulation test;

[0017] 5) Compare the cross-sectional load and gauge length variation curves after the instability behavior occurs in the simulation test with the corresponding tensile test results;

[0018] 6) Adjust the correction coefficient P in the new constitutive model, repeat steps 4) and 5) to find the simulation curve with the highest coincidence between the simulation instability curve and the experimental instability curve.

[0019] Furthermore, at least three pairs of arbitrary plastic deformation zones before instability occurs are selected as test data according to the true stress and true strain curves obtained from the test, and the constitutive models of at least two materials are selected according to the constitutive model in the prior art, and the obtained test data are fitted.

[0020] Furthermore, for materials with a yield strength below 220 MPa, the initialization correction factor P is 0.15.

[0021] Furthermore, for materials with a yield strength of 220 to 420 MPa, the initialization correction factor P is 0.25.

[0022] Furthermore, for materials with a yield strength of 420 to 720 MPa, the initialization correction factor P is 0.3.

[0023] Furthermore, for materials with a yield strength of 720 MPa or more, the initial correction factor P is 0.6.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1) Fit the true stress and true strain curves of the specimen to the constitutive model, and use the corrected goodness of fit to determine a more accurate new constitutive model;

[0026] 2) By optimizing the correction coefficient in the new constitutive model, a simulated instability curve that is closer to the experimental instability curve is found.

[0027] 3) It overcomes the shortcoming that the prediction results of simulation curve instability caused by curve fitting are quite different from the test, and can accurately and reliably predict the instability behavior of the simulation curve. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram comparing the simulation instability curve of the existing constitutive model and the experimental curve.

[0029] Figure 2 It is a schematic diagram comparing the simulated instability curve and the test curve with the correction coefficient adjusted under the new constitutive model of the present invention. DETAILED DESCRIPTION

[0030] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings:

[0031] The present invention provides a method for characterizing the instability behavior of a uniaxial tensile simulation curve, comprising the following steps:

[0032] 1) Perform uniaxial tensile test according to the tensile test standard led by the government or independently formulated by the market, select tensile specimens and perform uniaxial tensile test on them to obtain the true stress and true strain curves.

[0033] 2) According to the true stress and true strain curves obtained in the test, at least 3 pairs of arbitrary plastic deformation zones before instability occurs are selected as test data, and at least 2 constitutive models of materials are selected according to the constitutive model in the prior art. The obtained test data are fitted, and the goodness of fit of each constitutive model after correction is calculated. The formula is as follows:

[0034] A_R 2 =1-(1-R 2 )×(n+n / R 2 ) / (np-1)(3)

[0035] The closer the corrected goodness of fit is to 1, the better the constitutive model fits the experimental data. 2 is the goodness of fit of the modified constitutive model, R 2 is the goodness of fit of the constitutive model before correction, n is the sample size, and p is the number of fitting parameters.

[0036] 3) The two constitutive models with the highest goodness of fit after correction are positioned as A and B, and the new constitutive model after integration is C. The integration formula is as follows:

[0037] C=P×A+(1-P)×B(4)

[0038] Among them, P is the correction coefficient, and the P correction range is 0 to 1.

[0039] 4) Lock the fitting parameters of constitutive models A and B, use constitutive model C to fit the test data, plastically expand the fitting curve to the ultimate strain of the material, and define it as the material property in the simulation model that is completely consistent with the tensile test condition standard. In the simulation model consistent with the tensile test condition, the cross-section and change process of the central area of ​​the specimen are obtained.

[0040] 5) Constitutive models A and B are fitted with data before instability and have a high goodness of fit. Therefore, the simulation accuracy of the new constitutive model C before the instability behavior occurs is sufficient. Therefore, the cross-sectional load-gauge length change process curve after the instability behavior occurs in the simulation test is compared with the corresponding tensile test results;

[0041] 6) According to the comparison results between the simulation results and the test results, adjust the correction coefficient P, repeat steps 4) and 5). For materials with a yield strength below 220 MPa, the initial correction coefficient P is 0.15, for materials with a yield strength of 220-420 MPa, the initial correction coefficient P is 0.25, for materials with a yield strength of 420-720 MPa, the initial correction coefficient P is 0.3, for materials with a yield strength of more than 720 MPa, the initial correction coefficient P is 0.6, start adjusting from the initial correction coefficient P until the curve with the highest overlap between the simulation instability curve and the test instability curve is found, and select the simulation curve with the correction coefficient P at this time.

[0042] The following examples are implemented on the premise of the technical solution of the present invention, and provide detailed implementation methods and specific operation processes, but the protection scope of the present invention is not limited to the following examples. The methods used in the following examples are conventional methods unless otherwise specified.

[0043] [Example 1]

[0044] like Figure 1 As shown in the figure, a characterization method for the instability behavior of a uniaxial tensile simulation curve is presented. According to the national standard GB / T228, a uniaxial tensile test is carried out to obtain the true stress-true strain curve of the dual-phase steel DP780, and the ultimate strain of the dual-phase steel DP780 is 17.6%. According to the true stress-true strain curve obtained in the test, the plastic deformation zone with a plastic strain of 0.5% to 10% is selected as the test data, and the Hollomon, Ludwik, and Voce constitutive models are selected to fit the test data. According to the formula A_R 2 =1-[(1-R 2 )×(n+n / R 2 ) / (np-1)], and calculate the goodness of fit of each constitutive model after correction. In Hollomon's constitutive model, R 2=0.9812, n = 1168, p = 2, the corrected goodness of fit was calculated to be 0.962; R 2 =0.9691, n = 1168, p = 3, goodness of fit calculated as 0.937; R 2 =0.9991, n=1168, p=3, goodness of fit is calculated to be 0.998; the Hollomon constitutive model and Voce constitutive model with the highest modified goodness of fit are selected, the modified Hollomon constitutive model is defined as A, the modified Voce constitutive model is defined as B, and a new constitutive model C is integrated using C=P×A+(1-P)×B. The fitting curve is plastically extended to the ultimate strain of 17.6% of the material, and the cross-section and gauge length change process of the central area of ​​the specimen is obtained in the finite element simulation model consistent with the tensile test condition. The cross-section load-gauge length change process curve after the instability behavior occurs in the simulation test is compared with the corresponding tensile test results, and the correction coefficient P is manually adjusted until the coincidence degree of the simulation instability curve and the experimental instability curve is the highest.

[0045] [Example 2]

[0046] like Figure 2 As shown, the basic steps of a method for characterizing the instability behavior of a uniaxial tensile simulation curve are the same as those of Example 1, with the addition of an initialization value for setting the correction parameter to improve the correction speed. The yield strength of the selected dual-phase steel DP780 is 420-720 MPa, and the initialization correction coefficient P is set to 0.3, which is manually adjusted from 0.3. When the correction coefficient P is 0.51, the curve comparison results show that the simulated instability curve and the experimental instability curve have the highest coincidence.

[0047] [Example 3]

[0048] The basic steps of a method for characterizing the instability behavior of a unidirectional tensile simulation curve are the same as those of Example 1, with the addition of an initialization value for setting the correction parameter to increase the correction speed. The yield strength of the selected DC01 steel is below 220 MPa, and the initialization correction coefficient P is set to 0.15. Manual adjustment is started from 0.15. When the correction coefficient P is 0.11, the curve comparison results show that the simulated instability curve and the experimental instability curve have the highest degree of overlap.

[0049] [Example 4]

[0050] The basic steps of a method for characterizing the instability behavior of a unidirectional tensile simulation curve are the same as those of Example 1, with the addition of an initialization value for setting the correction parameter to improve the correction speed. The yield strength of the selected dual-phase steel DP590 is 220-420 MPa, and the initialization correction coefficient P is set to 0.25. Manual adjustment is started from 0.25. When the correction coefficient P is 0.19, the curve comparison results show that the simulated instability curve and the experimental instability curve have the highest overlap.

[0051] [Example 5]

[0052] The basic steps of a method for characterizing the instability behavior of a unidirectional tensile simulation curve are the same as those of Example 1, with the addition of an initialization value for setting the correction parameter to improve the correction speed. The yield strength of the selected dual-phase steel DP1180 is above 720 MPa, and the initialization correction coefficient P is set to 0.6. Manual adjustment is started from 0.6. When the correction coefficient P is 0.78, the curve comparison results show that the simulated instability curve and the experimental instability curve have the highest degree of overlap.

Claims

1. A method for characterizing the instability behavior of a uniaxial tensile simulation curve, characterized in that: The steps include: 1) Obtain the true stress and true strain curves of the material; 2) Fit the experimental data corresponding to different constitutive models and calculate the goodness of fit of each constitutive model after correction: A_R 2 =1-[(1-R 2 )×(n+n / R 2 ) / (n-p-1)] (1) Among them, A_R 2 is the goodness of fit of the modified constitutive model, R 2 is the goodness of fit of the constitutive model before modification, n is the sample size, and p is the number of fitting parameters; 3) Select the two constitutive models whose goodness of fit is closest to 1 after correction and integrate them into the new constitutive model C. The integration formula is: C=P×A+(1-P)×B (2) Among them, P is the correction coefficient, the correction range of P is 0 to 1, and A and B are the two constitutive models whose goodness of fit is closest to 1 after correction; 4) Using the new constitutive model to obtain the cross-sectional load and gauge length variation process in the central area of ​​the specimen in the simulation test: Lock the fitting parameters of constitutive models A and B, use constitutive model C to fit the test data, plastically extend the fitting curve to the ultimate strain of the material, and define it as the material property in the simulation model that is completely consistent with the tensile test condition standard, and obtain the cross-sectional area and variation process of the central area of ​​the specimen in the simulation model that is consistent with the tensile test condition; 5) Compare the cross-sectional load and gauge length variation curves after the instability behavior occurs in the simulation test with the corresponding tensile test results; 6) Adjust the correction coefficient P in the new constitutive model, repeat steps 4) and 5) to find the simulation curve with the highest coincidence between the simulation instability curve and the experimental instability curve.

2. The method for characterizing the instability behavior of a uniaxial tensile simulation curve according to claim 1, characterized in that: In the above step 2), at least 3 pairs of arbitrary plastic deformation zones before instability occurs are selected as test data according to the true stress and true strain curves obtained in the test, and the constitutive models of at least 2 materials are selected according to the constitutive model in the prior art, and the obtained test data are fitted.

3. The method for characterizing the instability behavior of a uniaxial tensile simulation curve according to claim 1, characterized in that: For materials with a yield strength below 220 MPa, the initial correction factor P is 0.

15.

4. The method for characterizing the instability behavior of a uniaxial tensile simulation curve according to claim 1, characterized in that: For materials with a yield strength of 220 to 420 MPa, the initial correction factor P is 0.

25.

5. The method for characterizing the instability behavior of a uniaxial tensile simulation curve according to claim 1, characterized in that: For materials with a yield strength of 420 to 720 MPa, the initial correction factor P is 0.

3.

6. The method for characterizing the instability behavior of a uniaxial tensile simulation curve according to claim 1, characterized in that: For materials with a yield strength of 720 MPa or more, the initial correction factor P is 0.6.

Images