Material local damage quantitative evaluation method based on multi-scale mechanical parameters

Abstract

The invention relates to a material local damage quantitative evaluation method based on multi-scale mechanical parameters. Comprising the following steps: acquiring an original three-dimensional grayscale image in a free state; applying a micro-amplitude load to obtain a loaded three-dimensional grayscale image; calculating a local strain field after applying the micro-amplitude load by adopting DVC, and constructing an initial Poisson's ratio field: introducing macroscopic damage to the sample, and repeating the steps of scanning in a free state and under the micro-amplitude load and constructing the Poisson's ratio field to obtain a Poisson's ratio field after damage; carrying out registration on the image; in the registered image, calculating a gray scale entropy; determining a minimum volume unit through a sliding window method; determining a minimum analysis unit, and constructing a quantitative damage index model; and generating a three-dimensional damage distribution diagram according to the damage index. The resolution bottleneck of traditional nondestructive testing is broken through, precise positioning and quantitative evaluation of submicron damage are achieved, and a new means is provided for research of a material damage evolution mechanism.

Description

Technical Field

[0001] The present invention belongs to the field of non-destructive testing, and in particular relates to a method for quantitatively evaluating local damage of materials based on multi-scale mechanical parameters. Background Art

[0002] In materials science and engineering, nondestructive testing (NDT) is a core tool for assessing material damage and predicting structural lifespan. However, as materials evolve toward heterogeneity and micro- and nanoscale structures (e.g., porous metal composites), traditional testing methods face the following technical bottlenecks: First, insufficient resolution: Conventional NDT methods (e.g., ultrasonic testing and infrared thermal imaging) are limited by physical principles, with spatial resolution typically on the millimeter to centimeter scale. This makes it difficult to capture early damage, such as micron-scale cracks and pores, resulting in a high rate of missed detections and an inability to meet the needs of high-precision material health monitoring. Second, local damage characterization is lacking: Traditional methods focus primarily on overall performance degradation (e.g., loss of macroscopic strength and stiffness), but failure in heterogeneous materials often begins with the initiation and propagation of localized microdamage. Third, existing technologies lack the ability to quantitatively analyze local mechanical parameters (e.g., subvoxel-level Poisson's ratio), making it impossible to accurately locate damage sources or predict their evolutionary paths. Finally, statistical representativeness is insufficient: For multiphase heterogeneous materials, small-scale analysis areas are susceptible to random microstructural interference (e.g., aggregate distribution and pore aggregation), leading to large fluctuations in mechanical parameters and unreliable results.

[0003] To address these issues, recent research has attempted to combine digital image correlation (DIC) and CT technology to improve detection sensitivity, but the following limitations persist. First, multi-phase data alignment is difficult: CT images before and after damage produce geometric deviations due to specimen displacement or deformation, making it difficult to achieve sub-pixel alignment using traditional registration algorithms (such as rigid transformations), which affects the accuracy of local parameter calculations. Second, dynamic RVE definition is lacking: Existing RVE definitions are often based on statistical assumptions about undamaged materials, failing to consider the enhanced local heterogeneity caused by damage, leading to inappropriate selection of analysis regions. Summary of the Invention

[0004] The purpose of the present invention is to provide a quantitative assessment method for local damage based on multi-scale mechanical parameters. Through dual-condition CT scanning, dynamic RVE definition and dual-parameter fusion model, it breaks through the traditional resolution limitation, realizes the precise positioning and reliable assessment of submicron damage, and provides an innovative solution for the structural health management of heterogeneous materials.

[0005] The technical solution to achieve the purpose of the present invention is: a method for quantitatively evaluating local damage of materials based on multi-scale mechanical parameters, comprising the following steps:

[0006] Step (1): performing tomographic imaging of the undamaged sample to be tested in a free state to obtain an original three-dimensional grayscale image;

[0007] Step (2): Apply a slight load, keep the load stable, and then perform a second tomographic imaging with the same scanning parameters as step (1) to obtain a three-dimensional grayscale image after loading;

[0008] Step (3): Use digital volume correlation (DVC) technology to calculate the local strain field after applying a small load and construct the initial Poisson's ratio field v0:

[0009] Step (4): Introduce macro damage to the specimen, repeat the CT scanning steps in the free state and under micro-load and the Poisson's ratio field construction steps of steps (1)-(3), and obtain the Poisson's ratio field after damage v d ;

[0010] Step (5): Using a registration algorithm, spatially match the CT images before and after injury to eliminate the position deviation of the specimen under different test conditions;

[0011] Step (6): In the registered CT image, extract the grayscale distribution of the local area and calculate the grayscale entropy; through the sliding window method, when the grayscale entropy change rate within the window is ≤5%, determine the current window as the minimum volume element RVE;

[0012] Step (7): Take the scale of 3 times the minimum volume unit RVE as the minimum analysis unit of local damage and construct a quantitative damage index model:

[0013]

[0014] Step (8): Generate a three-dimensional damage distribution map based on the damage index to achieve quantitative assessment of the damage.

[0015] Furthermore, the slight load in step (2) is a load that is less than 10% of the ultimate strength of the sample.

[0016] Furthermore, the initial Poisson's ratio field v0 in step (3) is calculated as follows:

[0017]

[0018] Where, ε xx is the strain in the X direction, ε yy is the strain in the Y direction and ε zz Z-direction strain

[0019] Furthermore, the macro damage in step (4) is external damage, specifically load or temperature stress.

[0020] Furthermore, the registration algorithm used in step (5) is a registration algorithm based on SIFT feature points.

[0021] Furthermore, the ray source of the tomography imaging device is X-ray, neutron source or gamma ray.

[0022] Furthermore, the grayscale entropy in step (6) is calculated by the following formula:

[0023] H=-∑p i logp i

[0024] Among them, p i is the gray value probability distribution.

[0025] Furthermore, the above method is applicable to the service status monitoring of metal structural parts.

[0026] Compared with the prior art, the present invention has the following significant advantages:

[0027] 1. High accuracy: Compared with conventional non-destructive testing methods (such as ultrasonic testing and infrared thermal imaging), this method can obtain sub-pixel level damage and is more suitable for analyzing local damage.

[0028] 2. Strong applicability: Compared with the previous DVC method, this method uses image registration technology to obtain the position movement information of the sample during multiple scanning processes, thereby allowing the sample to move during multiple tests, and is therefore suitable for the detection of local strain before and after damage.

[0029] 3. Non-destructive: This method uses non-contact tomography technology, which does not damage the sample; and the applied loads are all micro-loads, which can ensure that the microstructure of the sample does not change during the loading process.

[0030] 4. Conversion from strain to damage: Through the correlation formula between local Poisson's ratio and damage, mechanical information is converted into damage information of the sample, providing a new method and idea for the study of local damage of materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a typical cross section of the nanosilver sample in its initial state.

[0032] Figure 2 The typical cross section ε obtained by applying a small load to the nanosilver sample before damage xx field.

[0033] Figure 3 The typical cross section ε obtained by applying a small load to the nanosilver sample before damage yy field.

[0034] Figure 4 The typical cross section ε obtained by applying a small load to the nanosilver sample before damage zz field.

[0035] Figure 5 The typical cross-sectional Poisson's ratio field obtained when a slight load is applied to the nanosilver sample before damage.

[0036] Figure 6 This is a typical cross section of the nanosilver sample in a damaged state.

[0037] Figure 7 The typical cross section ε obtained by applying a micro-load to the nanosilver sample after damage xx field.

[0038] Figure 8 The typical cross section ε obtained by applying a micro-load to the nanosilver sample after damage yy field.

[0039] Figure 9 The typical cross section ε obtained by applying a micro-load to the nanosilver sample after damage zz field.

[0040] Figure 10 The typical cross-sectional Poisson's ratio field obtained when a micro-load is applied to the nanosilver sample after damage.

[0041] Figure 11 This is the calculated typical transverse damage degree diagram of the nanosilver sample. DETAILED DESCRIPTION

[0042] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments, but the scope of protection claimed in the present invention is not limited thereto.

[0043] Example 1

[0044] The present invention discloses a method for quantitatively assessing local damage of materials based on multi-scale mechanical parameters. The method specifically comprises the following steps:

[0045] Step 1: Synthesize nanosilver powder by chemical reduction: 0.1 mol / L silver nitrate and 0.2 mol / L ascorbic acid are mixed in a volume ratio of 1:1, 1 wt% polyvinyl pyrrolidone is added as a dispersant, and the mixture is stirred in a 50°C water bath for 2 h to obtain 50 nm silver particles. After centrifugal cleaning at 8000 rpm and vacuum drying, the mixture is cold isostatically pressed through a cemented carbide mold (inner diameter 2 mm) and maintained at a pressure of 200 MPa for 5 min to obtain a cylindrical body. Subsequently, a two-stage sintering is performed in a vacuum tube furnace (vacuum degree ≤ 10-3 Pa): first, the temperature is increased to 300°C at 5°C / min and maintained for 1 h to promote particle bonding, and then the temperature is increased to 400°C at 2°C / min and maintained for 2 h to form a porous structure with a porosity of 30±2%, thereby obtaining a cylindrical nanosilver sintered sample with a diameter of 2 mm and a height of 4 mm.

[0046] Step 2: First, perform a free-state CT scan on the undamaged specimen. The CT scanner model used is Xradia Versa 615, the voltage used is 100KV, the current used is 0.3mA, and the effective resolution is 3μm. A typical slice of the original 3D grayscale image is obtained as follows: Figure 1 shown.

[0047] Step 3: Use the 5000N in-situ loading device produced by Suzhou Huachuang Company to apply a uniaxial compressive load of 20MPa to the sample. While maintaining the load, perform CT scanning again with the same scanning parameters as the first time to obtain the original three-dimensional grayscale image. Use digital volume image correlation (DVC) technology to calculate the local strain field before and after the application of the uniaxial compressive load, and obtain the cross-sectional strain field image of the sample under micro-vibration loading before damage, including ε xx Field, ε yy Field and ε zz Field, respectively Figure 2 、 Figure 3 、 Figure 4 shown.

[0048] Step 4: Construct the initial Poisson's ratio field according to the following formula:

[0049]

[0050] A typical cross-sectional view of the obtained Poisson's ratio field is shown in Figure 5 shown.

[0051] Step 5: Apply a compressive load of 100 MPa to the nanosilver sample to introduce macroscopic damage. Subsequently, the above CT scanning steps under the free state and 20 MPa micro-load were repeated to obtain typical cross-sectional images of the sample under damage, as shown in Figure 5. Figure 6 shown.

[0052] Step 6: Obtain the cross-sectional strain field image of the damaged sample under a micro-load, including ε xx Field, ε yy Field and ε zz Field, respectively Figure 7 、 Figure 8 、 Figure 9 As shown. The Poisson's ratio distribution of the typical cross section is further obtained as follows Figure 10 shown.

[0053] Step 7: To ensure the comparability of the data before and after the injury, a registration algorithm is used to spatially match the CT images before and after the injury to eliminate the position deviation of the specimen under different test conditions. The representative volume element (RVE) is defined using grayscale entropy, and its calculation formula is:

[0054] H=-∑p i logp i

[0055] Among them, p i is the gray value probability distribution.

[0056] In the registered CT image, the grayscale distribution of the local area is extracted and the grayscale entropy is calculated. The initial window size is set to 50×50 pixels (corresponding to the physical size of 150×150μm) by the sliding window method. 2 ), the sliding step is 10 pixels (corresponding to 30μm), and the grayscale entropy in each window is calculated. When the grayscale entropy change rate after sliding the adjacent windows is ≤5% (i.e. ), determine that the current window is the minimum RVE.

[0057] Step 8: Using a scale of 3 times RVE or more as the minimum analysis unit for local damage, a quantitative damage index evaluation model was constructed to quantitatively evaluate the damage degree of the nanosilver sample. The quantitative damage index calculation formula is as follows:

[0058]

[0059] Step 9: Generate a three-dimensional damage distribution map of the nanosilver sample based on the quantitative damage index evaluation model. The typical damage index distribution map is as follows: Figure 11 As shown in the figure, the damage area and damage degree of the specimen are intuitively displayed, providing strong support for the study of the material damage evolution mechanism.

Claims

1. A quantitative assessment method for local damage of materials based on multi-scale mechanical parameters, characterized in that: The steps include: Step (1): performing tomographic imaging of the undamaged sample to be tested in a free state to obtain an original three-dimensional grayscale image; Step (2): Apply a slight load, keep the load stable, and then perform a second tomographic imaging with the same scanning parameters as step (1) to obtain a three-dimensional grayscale image after loading; Step (3): Use digital volume correlation (DVC) technology to calculate the local strain field after applying a small load and construct the initial Poisson's ratio field v0: Step (4): Introduce macro damage to the specimen, repeat the CT scanning steps in the free state and under micro-load and the Poisson's ratio field construction steps of steps (1)-(3), and obtain the Poisson's ratio field after damage v d ; Step (5): Using a registration algorithm, spatially match the CT images before and after injury to eliminate the position deviation of the specimen under different test conditions; Step (6): In the registered CT image, extract the grayscale distribution of the local area and calculate the grayscale entropy; through the sliding window method, when the grayscale entropy change rate within the window is ≤5%, determine the current window as the minimum volume element RVE; Step (7): Take the scale of 3 times the minimum volume unit RVE as the minimum analysis unit of local damage and construct a quantitative damage index model: Step (8): Generate a three-dimensional damage distribution map based on the damage index to achieve quantitative assessment of the damage.

2. The method according to claim 1, characterized in that The slight load in step (2) is a load that is less than 10% of the ultimate strength of the specimen.

3. The method according to claim 2, characterized in that The initial Poisson's ratio field v0 in step (3) is calculated as follows: Where, ε xx is the strain in the X direction, ε yy is the strain in the Y direction and ε zz Z-direction strain.

4. The method according to claim 3, characterized in that The macro damage in step (4) is external damage, specifically load or temperature stress.

5. The method according to claim 4, characterized in that The registration algorithm used in step (5) is a registration algorithm based on SIFT feature points.

6. The method according to claim 5, characterized in that The radiation source of the tomography imaging device is X-ray, neutron source or gamma ray.

7. The method according to claim 6, characterized in that The grayscale entropy in step (6) is calculated by the following formula: H=-∑p i logp i Among them, p i is the gray value probability distribution.

8. The method according to any one of claims 1 to 7, characterized in that Suitable for service condition monitoring of metal structures.

Images