MD-rve-fe cross-scale in-situ sem tensile analysis and experimental methods

Abstract

The present application relates to the technical field of material mechanics and computational materials science, and specifically relates to a MD-RVE-FE cross-scale in-situ SEM tensile analysis and experimental method, comprising: performing molecular dynamics loading analysis on the interface of a two-phase material to obtain peak interface strength data and fracture energy data; performing constitutive construction and calibration of a cohesive zone model to obtain interface constitutive parameters; taking a microscopic image of a tensile sample, establishing a finite element model of the tensile sample based on representative volume elements in a region of interest image and the interface constitutive parameters, performing displacement-controlled tensile loading, and obtaining simulation results; performing in-situ tensile testing on the tensile sample in an in-situ scanning electron microscope chamber, performing micro-digital image correlation processing based on the post-tensile image taken by the scanning electron microscope, and obtaining experimental results; comparing the simulation results and the experimental results to obtain comparison and error evaluation results; and the present application can achieve consistency verification of SEM and mu-DIC.

Description

Technical Field

[0001] This invention relates to the fields of materials mechanics and computational materials science, specifically to a cross-scale in-situ SEM tensile analysis and experimental method using MD-RVE-FE. Background Technology

[0002] Interfacial debonding, crack initiation, and propagation in multiphase materials are key mechanisms affecting macroscopic mechanical response and failure evolution. Existing technologies typically suffer from the following problems: interfacial parameters are difficult to reliably obtain under multiple temperature conditions; atomic-scale information is difficult to traceably transfer to continuum models; and verification often focuses on macroscopic stress-strain curve fitting, lacking discriminative verification of local strain fields and damage location / evolution.

[0003] Specifically, while molecular dynamics (MD) can reveal interfacial bonding, debonding, and crack initiation mechanisms at the atomic scale, its results are often difficult to directly apply to macroscopic structural analysis. Representative volume elements (RVEs) based on real microstructures can describe the microstructural characteristics of materials, but often fail to accurately reflect local damage evolution due to simplified parameter sources and boundary conditions. Macroscopic finite element methods (FEs) have strong engineering applicability at the macroscopic scale, but their accuracy is highly dependent on the reasonable input of microscopic parameters and damage criteria. Meanwhile, in-situ scanning electron microscopy (SEM) stretching can observe crack initiation, propagation, and interfacial failure processes in real time, but lacks a unified parameter mapping and verification chain with the aforementioned multi-scale models.

[0004] Therefore, a method is needed that can achieve the connection between atomic, microscopic and macroscopic parameters and can verify consistency through in-situ scanning electron microscopy (SEM) and microscopic digital image correlation (μ-DIC). Summary of the Invention

[0005] In view of the above problems, the present invention provides a cross-scale in-situ SEM tensile analysis and experimental method of MD-RVE-FE, which solves the technical problems in the prior art that it is difficult to quantitatively obtain key parameters of interface debonding, lack a unified and reproducible link for cross-scale parameter transfer, and lack an experimental verification path that simultaneously covers macroscopic response and microscopic field quantity.

[0006] This invention provides a cross-scale in-situ SEM tensile analysis and experimental method for MD-RVE-FE, comprising the following steps: Step S1: Perform molecular dynamics loading analysis on the interface of the two-phase materials, including: applying cracking loading in multiple directions to the interface of the two-phase materials at different temperatures, and obtaining peak interface strength data and fracture energy data; Step S2: Based on the predetermined cohesive region model form, peak interface strength data and fracture energy data, construct and calibrate the constitutive model of the cohesive region to obtain interface constitutive parameters; Step S3: Prepare a tensile specimen, take a microscopic image of the tensile specimen, and obtain the region of interest in the microscopic image; establish a finite element model of the tensile specimen based on the representative volume elements in the region of interest image and the interface constitutive parameters. The finite element model was subjected to displacement-controlled tensile loading, and simulation results were obtained. Step S4: The tensile specimen is subjected to in-situ tensile testing inside the in-situ scanning electron microscope chamber. Microscopic digital image correlation processing is performed based on the tensile images captured by the scanning electron microscope to obtain experimental results. Step S5: Compare the load-displacement / stress-strain response curves, local strain field distribution, interface debonding and damage evolution data in the simulation results and experimental results to obtain comparison and error assessment results.

[0007] Preferably, step S1 specifically includes: Step S1-1: Establish a molecular dynamics model that includes the interface between two phase materials. The molecular dynamics model includes the atomic structure of the first phase, the atomic structure of the second phase, and the interface region formed by their contact. Minimize and thermally balance the molecular dynamics model. Step S1-2: Apply cracking loading in multiple directions to the interface of the two-phase material, including normal, tangential and mixed-mode loading; Steps S1-3: During the loading process, extract the interface traction-separation curve, and determine the normal peak interface strength and tangential peak interface strength based on the maximum traction value in the interface traction-separation curve. The area enclosed by the traction-separation curve is taken as the interfacial fracture energy.

[0008] Preferably, in step S2, the predetermined cohesive region model is a bilinear model; The interface constitutive parameters include normal stiffness, tangential stiffness, normal fracture energy, tangential fracture energy, and hybrid mode evolution parameters.

[0009] Preferably, step S3 specifically includes: Step S3-1: Prepare tensile specimens and perform surface polishing. Step S3-2: Obtain a microscopic image of the sample surface and select a statistically representative region as the region of interest; Step S3-3: Perform image segmentation and geometric reconstruction on the region of interest to generate representative voxels containing enhanced phase morphology and spatial distribution; A finite element model of the tensile specimen is constructed, and representative volume elements are embedded at positions corresponding to the region of interest based on the interface constitutive parameters. Step S3-4: Apply displacement-controlled tensile loading to the finite element model and obtain simulation results.

[0010] Preferably, step S3-3 specifically includes: Preprocessing of the original microscopic images includes denoising, contrast enhancement, and brightness equalization; Image segmentation technology is used to identify the boundaries of different phases and extract the boundary contours of each phase; Based on the segmentation results of the boundary contours of each phase, the geometric model of the representative volume element is reconstructed. The geometric model of the representative volume element is embedded into the region of interest.

[0011] Preferably, step S4 specifically includes: Step S4-1: Prepare surface markers suitable for microscopic digital image correlation in the region of interest; Step S4-2: The tensile specimen is subjected to in-situ tensile testing inside the in-situ scanning electron microscope cavity, and image sequences are acquired using a segmented load-stop-imaging-reload strategy. Step S4-3: Perform microscopic digital image correlation processing on the image sequence to obtain experimental results.

[0012] Preferably, step S4-1 specifically includes: Square grid markers are formed in the region of interest by spraying speckle or etching a grid. In step S4-3, the step of performing microscopic digital image correlation processing on the image sequence specifically includes: Using the initial unloaded image as a reference frame, sub-region matching and displacement tracking are performed on the images after each loading stage, thereby calculating the displacement field, principal strain field, shear strain field, and strain concentration zone distribution within the region of interest.

[0013] Preferably, step S5 specifically includes: Step S5-1: Compare the load-displacement / stress-strain response curves of the simulation and the experiment, and obtain the root mean square error, mean absolute percentage error and correlation coefficient of the load-displacement / stress-strain response curves of the simulation and the experiment. Step S5-2: Compare the local strain field distribution obtained from the simulation of the region of interest with that obtained from the microscopic digital image correlation processing, and obtain the structural similarity index, normalized cross-correlation coefficient and root mean square error of the local strain field distribution between the simulation and the experiment. Step S5-3: Compare the debonding initiation position, expansion path, quantity evolution and local strain hotspots at the interface to obtain the debonding initiation position deviation, debonding quantity error and evolution rate error, as well as the spatial consistency index between the debonding area and the local strain concentration area.

[0014] Compared with the prior art, the present invention has at least the following beneficial effects: (1) This invention directly obtains key parameters such as peak interfacial strength and fracture energy by conducting molecular dynamics simulations of the interface of two-phase materials under different temperatures and multiple cracking loading modes. Compared with methods that rely solely on empirical assumptions or macroscopic inversion, this method can more realistically reflect the failure characteristics of the interface under the coupled effects of temperature and loading direction, thereby improving the accuracy and traceability of interface parameters from the source.

[0015] (2) This invention is based on a pre-determined cohesive region model and combines the peak interfacial strength and fracture energy obtained by MD to complete constitutive construction and calibration, thereby obtaining the interfacial constitutive parameters and realizing parameter transfer from the molecular scale to the continuous medium scale. This cross-scale modeling method unifies the microscopic interfacial damage mechanism with the macroscopic cohesive region response, avoiding the problem of traditional CZM parameters relying heavily on trial and error fitting, and can significantly improve the physical consistency, prediction stability and reproducibility of the model.

[0016] (3) This invention establishes a finite element model by extracting representative volume elements from microscopic images and obtains experimental results by combining in-situ scanning electron microscopy tensile and microscopic digital image correlation techniques. This not only verifies the overall load-displacement or stress-strain response, but also simultaneously obtains the local strain field distribution, interface debonding location, and damage evolution process. This "simulation-experiment" two-way comparison method makes model verification no longer limited to the macroscopic curve level, but elevates it to the level of microscopic deformation and failure mechanism, thereby significantly enhancing the credibility and engineering applicability of the method.

[0017] (4) This invention systematically compares and evaluates the simulation results with the experimental results, which can comprehensively verify the rationality of the established interface constitutive model and finite element analysis framework, and provide a reliable basis for subsequent material design, interface optimization and failure prediction. Since this invention does not limit the specific material name, but is aimed at multiphase material systems with significant interface damage mechanisms, it has strong versatility and promotion value, and can be applied to the interface strength evaluation and fracture behavior study of various composite materials and metal-based / ceramic-based / polymer-based multiphase systems. Attached Figure Description

[0018] The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of the invention.

[0019] Figure 1 The flowchart shows the MD-RVE-FE cross-scale in-situ SEM tensile analysis and experimental method provided by this invention.

[0020] Figure 2 This is a schematic diagram of the overall process of the MD-RVE-FE cross-scale in-situ SEM tensile analysis and experimental method provided by the present invention.

[0021] Figure 3The embedded RVE macro FE model, interface cohesive unit insertion, and boundary coupling diagram provided by this invention. Detailed Implementation

[0022] To better understand the above-described objectives, features, and advantages of the present invention, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other. Furthermore, the present invention can be implemented in other ways different from those described herein; therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0023] To illustrate the effectiveness of the method proposed in this invention, the following detailed description of the above technical solution is provided through a specific embodiment, such as... Figure 1 , Figure 2 As shown, a cross-scale in-situ SEM tensile analysis and experimental method of MD-RVE-FE is disclosed. In this embodiment, a two-phase material with a clear phase interface is used as the object of description. The two-phase material can be a composite material system composed of a matrix phase and a reinforcing phase, or any two-phase interface in any two-phase or multiphase material with interface debonding, crack initiation and propagation behavior. This invention establishes a parameter transfer and result comparison link of "molecular dynamics - cohesive region model - representative volume element - macroscopic finite element - in-situ experimental verification". The specific implementation steps are as follows: Step S1: Perform molecular dynamics loading analysis on the interface of the two-phase materials, including: applying cracking loading in multiple directions to the interface of the two-phase materials at different temperatures, and obtaining peak interface strength data and fracture energy data; In this step, the present invention first establishes a molecular dynamics model that includes the interface between two phase materials. The model includes the atomic structure of the first phase, the atomic structure of the second phase, and the interface region formed by their contact, ensuring that the interface region can accurately reflect the atomic arrangement and bonding characteristics of the material.

[0024] After the model is established, energy minimization is first performed to eliminate unreasonable high-energy configurations in the initial structure; then, thermal equilibrium is applied under a preset temperature set to bring the system to a stable thermodynamic state. The temperature set can be room temperature and several high- or low-temperature conditions, thereby obtaining the law of interfacial performance change with temperature.

[0025] After thermal equilibrium is achieved, cracking loading in multiple directions is applied to the interface of the two-phase materials. This cracking loading can include interface normal cracking loading and interface tangential slip loading, corresponding to Mode I cracking conditions and Mode II shearing conditions, respectively. In some embodiments, a mixed-mode loading with both normal and tangential displacement components can be further applied to characterize the separation behavior of the interface under complex stress conditions. Specifically, displacement control or velocity control can be applied to the atomic layers on both sides of the interface to gradually separate the interface along a predetermined direction, while simultaneously recording data such as interface reaction force, atomic displacement, and interface separation amount during the loading process.

[0026] During the loading process, traction-separation (TS) curve data is extracted. The traction-separation curve is plotted with interface separation displacement on the horizontal axis and interface traction stress on the vertical axis. For each temperature and loading mode, complete curve data is recorded from initial loading, reaching peak stress, stress softening, to complete separation. In a preferred embodiment, the data acquisition frequency should ensure that key feature points of the curve are accurately represented, and each operating condition is simulated 3-5 times independently to obtain the statistical average and standard deviation, thereby improving data reliability.

[0027] For the normal loading condition, the maximum traction value in the normal traction-separation curve is extracted as the normal peak interfacial strength; for the tangential loading condition, the maximum traction value in the tangential traction-separation curve is extracted as the tangential peak interfacial strength. Correspondingly, the area enclosed by the traction-separation curve is taken as the interfacial fracture energy, which is used to characterize the energy required for the interface to completely separate.

[0028] Furthermore, the evolution of interface strength and fracture energy under mixed-modal loading was analyzed. By changing the ratio of normal and tangential loading components, the relationship between interface strength and fracture energy and modal mixing ratio was obtained, providing data support for the mixed-modal criteria of the subsequent cohesive region model.

[0029] The final output includes data on normal peak interface strength, tangential peak interface strength, normal fracture energy, tangential fracture energy, and mixed mode coupling law at different temperatures, which serve as the basis for subsequent cohesive region model construction and calibration.

[0030] Step S2: Based on the predetermined cohesive region model form, peak interface strength data and fracture energy data, construct and calibrate the constitutive model of the cohesive region to obtain interface constitutive parameters; In this step, the present invention maps the interface parameters obtained at the atomic scale to the cohesive region model at the continuum scale to achieve cross-scale parameter transfer, as described in detail below.

[0031] First, select the appropriate constitutive form of the cohesive zone model (CZM). Commonly used cohesive zone model forms include bilinear models, exponential models, trapezoidal models, and power models.

[0032] In a preferred embodiment, the present invention employs a bilinear cohesive region model, which includes the elastic response stage of the interface, the damage initiation stage, and the damage evolution stage until complete failure, and has clear physical meaning and numerical stability.

[0033] This invention constructs and calibrates the constitutive model of the cohesive region based on peak interface strength data and fracture energy data. Specifically, during calibration, the normal peak interface strength and tangential peak interface strength are mapped to the normal and tangential strengths in the cohesive region model, respectively; the normal fracture energy and tangential fracture energy are mapped to the normal damage evolution energy parameter and the tangential damage evolution energy parameter. For the bilinear model, the critical separation displacement corresponding to the interface can also be deduced from the relationship between peak strength and fracture energy, thereby completely determining the interface constitutive parameters in the cohesive constitutive model.

[0034] Furthermore, for mixed-modal interface behavior, appropriate modal coupling evolution criteria can be selected based on mixed-modal cracking data, such as power-law mixed-modal criteria or BK-type mixed-modal criteria, and the corresponding coupling parameters can be determined by fitting the fracture energy evolution law under different modal ratios. For data obtained at different temperatures, the above interface constitutive parameters are calibrated at different temperatures to form a temperature-dependent interface parameter table or parameter function library.

[0035] The interface constitutive parameters finally obtained in this step may include: normal stiffness, tangential stiffness, normal fracture energy, tangential fracture energy, and mixed mode evolution parameters. The interface constitutive parameters are then organized into an interface constitutive parameter table or function library that can be directly called by the finite element solver.

[0036] Step S3: Prepare a tensile specimen, take a microscopic image of the tensile specimen, and obtain the region of interest in the microscopic image; establish a finite element model of the tensile specimen based on the representative volume elements in the region of interest image and the interface constitutive parameters. The finite element model was subjected to displacement-controlled tensile loading, and simulation results were obtained. In this step, the present invention constructs representative volume elements based on real microscopic morphology and embeds them into a macroscopic finite element model to achieve micro-macroscopic coupled finite element simulation, which is described in detail below.

[0037] Step S3-1: Prepare tensile specimens and perform surface polishing. In this step, a tensile specimen for the experiment is first prepared. The specimen surface is then finely ground and polished. In some embodiments, the grinding process may use progressively decreasing sandpaper, and the polishing may employ an alumina or diamond suspension to ensure the clarity and contrast of the microscopic images. For certain materials, further chemical etching or ion thinning treatments may be performed to enhance the microscopic contrast of the phase interfaces.

[0038] Step S3-2: Obtain a microscopic image of the sample surface and select a statistically representative region as the region of interest; Microscopic images of the sample surface are captured using a scanning electron microscope (SEM) or an optical microscope. In a preferred embodiment, imaging is performed using the backscattered electron (BSE) mode of a scanning electron microscope, which provides clear microscopic images based on differences in atomic number.

[0039] Subsequently, a Region of Interest (ROI) is selected from the captured microscopic images. The ROI should meet the basic requirement of a Representative Volume Element (RVE), meaning its microstructure characteristics represent the statistical homogeneity of the material. In practice, the selection of the ROI should comprehensively consider the following factors: the volume fraction of the reinforcing phase within the region should be similar to the overall volume fraction of the material; the size distribution, shape characteristics, and spatial distribution of the reinforcing phase should be typical; and the region size should be large enough to eliminate boundary effects. The ROI can be rectangular or square. When selecting an ROI, obvious pores, cracks, or other macroscopic defects should be avoided. Image analysis software can be used to quantitatively characterize multiple candidate ROIs, selecting the most representative region.

[0040] Step S3-3: Perform image segmentation and geometric reconstruction on the region of interest to generate representative voxels containing enhanced phase morphology and spatial distribution; A finite element model of the tensile specimen is constructed, and representative volume elements are embedded at positions corresponding to the region of interest based on the interface constitutive parameters. First, the original microscopic images are preprocessed, including denoising, contrast enhancement, and brightness equalization, to improve the accuracy of subsequent segmentation. Then, image segmentation techniques are used to identify the boundaries of different phases. Preferred methods include thresholding, edge detection, watershed algorithms, or machine learning segmentation algorithms. For two-phase materials, adaptive thresholding based on grayscale differences can be used; for unclear phase boundaries, morphological operations (such as erosion, dilation, opening, and closing operations) can be combined to optimize the segmentation results. After segmentation, the binarized image is converted into vectorized geometry, and the boundary contours of each phase are extracted.

[0041] Based on the segmentation results of the boundary contours of each phase, a geometric model of representative volume element (RVE) is reconstructed. The RVE model truly preserves the shape, size, orientation, and distribution characteristics of the reinforcing phase or second phase particles.

[0042] A macroscopic finite element model is established based on the actual size and shape of the macroscopic tensile specimen, and the RVE sub-model reconstructed from the microscopic image is embedded into the macroscopic region corresponding to the ROI, or the macroscopic boundary conditions are passed to the RVE region using the sub-model method.

[0043] Figure 3 The diagram illustrates the embedded RVE macro FE model, interface cohesive unit insertion, and boundary coupling.

[0044] This invention inserts cohesive units at phase interfaces. These cohesive units are used to simulate the debonding and damage evolution processes at the interfaces. In the finite element model, zero-thickness cohesive units are automatically or manually inserted along all contact interfaces between the matrix phase and the reinforcing phase, and the constitutive parameters of the interfaces are assigned to the cohesive units.

[0045] Through the above steps, the present invention completes the finite element modeling and obtains the finite element model of the tensile specimen.

[0046] Step S3-4: Apply displacement-controlled tensile loading to the finite element model and obtain simulation results.

[0047] In this step, displacement-controlled tensile loading is applied to the finite element model of the tensile specimen by setting boundary conditions and loading methods. The loading direction is consistent with the tensile direction of the actual specimen, and the loading rate or displacement increment is consistent with or similar to that in subsequent experiments. As the displacement gradually increases, the load-displacement / stress-strain of the macroscopic specimen is obtained, resulting in the load-displacement / stress-strain response curve.

[0048] Simultaneously, the local strain field distribution, interface debonding, and damage evolution data within the ROI region are output. These output results collectively constitute the simulation results in step S3, providing a basis for subsequent experimental comparisons.

[0049] Step S4: The tensile specimen is subjected to in-situ tensile testing inside the in-situ scanning electron microscope chamber. Microscopic digital image correlation processing is performed based on the tensile images captured by the scanning electron microscope to obtain experimental results. This step involves obtaining experimental data on macroscopic mechanical response, microscopic strain field, and interface damage evolution through in-situ scanning electron microscopy tensile experiments, specifically including: Step S4-1: Prepare surface markers suitable for micro-Digital ImageCorrelation (μ-DIC) in the ROI region.

[0050] μ-DIC technology calculates strain fields by tracking the displacement field of surface features, thus requiring the preparation of randomly distributed, high-contrast speckle patterns or regular grids on the sample surface. Preparation methods include: spraying speckles, etching grids, etc., to form square grid markings in the ROI region.

[0051] Step S4-2: The tensile specimen is subjected to in-situ tensile testing inside the in-situ scanning electron microscope cavity, and image sequences are acquired using a segmented load-stop-imaging-reload strategy. After mounting the macroscopic tensile specimen on the in-situ scanning electron microscope (SEM) stage, an in-situ tensile experiment is performed inside the SEM chamber. Preferably, a segmented loading strategy is adopted during the experiment, that is, the tensile displacement is increased step by step in the manner of "loading-stopping-imaging-continuing loading". At each predetermined displacement level, the load and displacement are recorded, and high-resolution continuous imaging is performed on the corresponding region of the region of interest (ROI) to obtain an image sequence.

[0052] For implementation methods that require studying the effects of temperature, a temperature control module can be introduced into the in-situ stretching platform to make the experimental temperature consistent with the target temperature used in the aforementioned steps, so as to achieve cross-scale verification under the same temperature conditions.

[0053] Step S4-3: Perform microscopic digital image correlation processing on the image sequence to obtain experimental results.

[0054] Microscopic digital image correlation processing was performed on the acquired scanning electron microscope (SEM) image sequence. Specifically, using the initial unloaded image as a reference frame, sub-region matching and displacement tracking were performed on the images after each loading stage to calculate the displacement field, principal strain field, shear strain field, and strain concentration zone distribution within the ROI region. Simultaneously, by combining the interface morphology changes, local gray-scale abrupt changes, or crack opening changes in the image sequence, the initiation location, propagation path, and evolution sequence of interface debonding and microcracks were identified.

[0055] By organizing the experimental results dataset, the following data were summarized: load-displacement / stress-strain response curves, i.e., the relationship between the load measured by the force sensor and the displacement of the tension table, which can be converted into engineering stress-strain curves; local strain field distribution, organized in the form of strain cloud map matrices at different loading stages; strain statistical characteristics, such as the evolution of average strain, maximum strain, and strain standard deviation with load; interface debonding and crack evolution data, including initial load, initial position coordinates, propagation path, and damage quantity-load curves; and the spatial correspondence between damage events and the strain field, i.e., the degree of overlap between the damage location and the strain concentration area. The above data were saved in a standardized format to provide an experimental benchmark for the comparative verification in step S5.

[0056] Step S5: Compare the load-displacement / stress-strain response curves, local strain field distribution, interface debonding and damage evolution data in the simulation results and experimental results to obtain comparison and error assessment results.

[0057] This step evaluates the accuracy and reliability of cross-scale modeling methods through multi-level and multi-dimensional comparative verification, specifically including: Step S5-1: Compare the load-displacement / stress-strain response curves from simulation and experiment.

[0058] In this step, the root mean square error, mean absolute percentage error, or correlation coefficient of the load-displacement / stress-strain response curves of the simulation and experiment can be calculated.

[0059] Step S5-2: Compare the local strain field distribution obtained from the ROI region simulation with that obtained from μ-DIC.

[0060] In this step, the structural similarity index of the local strain field distribution between the simulation and the experiment, the root mean square error of the normalized cross-correlation coefficient, etc. can be calculated.

[0061] Step S5-3: Compare the interface debonding starting position, expansion path, quantity evolution and its relationship with local strain hotspots.

[0062] In this step, the simulation and experimental results of the interface debonding initiation location, propagation path, quantity evolution, and their correspondence with local strain hotspots can be compared and analyzed. Preferably, under the same ROI region and the same load step, displacement step, or strain stage, the interface damage region in the simulation and experiment is registered, and the deviation of the debonding initiation location, the error of the debonding quantity and the error of the evolution rate, as well as the spatial consistency index between the debonding region and the local strain concentration area are calculated to characterize the simulation's ability to predict the interface debonding and damage evolution process.

[0063] Through the above steps, this invention realizes a complete technical process from atomic-scale interface parameter extraction, continuum interface constitutive construction, RVE modeling based on real microscopic morphology, macroscopic finite element simulation, and in-situ scanning electron microscopy experimental verification. This enables traceable, quantifiable, and experimentally verifiable cross-scale analysis of interface debonding, local strain concentration, and damage propagation behavior in two-phase materials.

[0064] While the specific embodiments of the present invention depict actions or steps in a particular order, this should be understood as requiring such actions or steps to be performed in the shown specific order or sequential order, or requiring all illustrated actions or steps to be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations. The above descriptions are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention.

[0065] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-scale in-situ SEM tensile analysis and experimental method using MD-RVE-FE, characterized in that, The steps include the following: Step S1: Perform molecular dynamics loading analysis on the interface of the two-phase materials, including: applying cracking loading in multiple directions to the interface of the two-phase materials at different temperatures, and obtaining peak interface strength data and fracture energy data; Step S2: Based on the predetermined cohesive region model form, peak interface strength data and fracture energy data, construct and calibrate the constitutive model of the cohesive region to obtain interface constitutive parameters; Step S3: Prepare a tensile specimen, take a microscopic image of the tensile specimen, and obtain the region of interest in the microscopic image; establish a finite element model of the tensile specimen based on the representative volume elements in the region of interest image and the interface constitutive parameters. The finite element model was subjected to displacement-controlled tensile loading, and simulation results were obtained. Step S4: The tensile specimen is subjected to in-situ tensile testing inside the in-situ scanning electron microscope chamber. Microscopic digital image correlation processing is performed based on the tensile images captured by the scanning electron microscope to obtain experimental results. Step S5: Compare the load-displacement / stress-strain response curves, local strain field distribution, interface debonding and damage evolution data in the simulation results and experimental results to obtain comparison and error assessment results.

2. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 1, characterized in that, Step S1 specifically includes: Step S1-1: Establish a molecular dynamics model that includes the interface between two phase materials. The molecular dynamics model includes the atomic structure of the first phase, the atomic structure of the second phase, and the interface region formed by their contact. Minimize and thermally balance the molecular dynamics model. Step S1-2: Apply cracking loading in multiple directions to the interface of the two-phase material, including normal, tangential and mixed-mode loading; Steps S1-3: During the loading process, extract the interface traction-separation curve, and determine the normal peak interface strength and tangential peak interface strength based on the maximum traction value in the interface traction-separation curve. The area enclosed by the traction-separation curve is taken as the interfacial fracture energy.

3. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 2, characterized in that, In step S2, the predetermined cohesive region model is a bilinear model; The interface constitutive parameters include normal stiffness, tangential stiffness, normal fracture energy, tangential fracture energy, and hybrid mode evolution parameters.

4. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 3, characterized in that, Step S3 specifically includes: Step S3-1: Prepare tensile specimens and perform surface polishing. Step S3-2: Obtain a microscopic image of the sample surface and select a statistically representative region as the region of interest; Step S3-3: Perform image segmentation and geometric reconstruction on the region of interest to generate representative voxels containing enhanced phase morphology and spatial distribution; A finite element model of the tensile specimen is constructed, and representative volume elements are embedded at positions corresponding to the region of interest based on the interface constitutive parameters. Step S3-4: Apply displacement-controlled tensile loading to the finite element model and obtain simulation results.

5. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 4, characterized in that, Step S3-3 specifically includes: Preprocessing of the original microscopic images includes denoising, contrast enhancement, and brightness equalization; Image segmentation technology is used to identify the boundaries of different phases and extract the boundary contours of each phase; Based on the segmentation results of the boundary contours of each phase, the geometric model of the representative volume element is reconstructed. The geometric model of the representative volume element is embedded into the region of interest.

6. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 5, characterized in that, Step S4 specifically includes: Step S4-1: Prepare surface markers suitable for microscopic digital image correlation in the region of interest; Step S4-2: The tensile specimen is subjected to in-situ tensile testing inside the in-situ scanning electron microscope cavity, and image sequences are acquired using a segmented load-stop-imaging-reload strategy. Step S4-3: Perform microscopic digital image correlation processing on the image sequence to obtain experimental results.

7. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 6, characterized in that, Step S4-1 specifically includes: Square grid markers are formed in the region of interest by spraying speckle or etching a grid. In step S4-3, the step of performing microscopic digital image correlation processing on the image sequence specifically includes: Using the initial unloaded image as a reference frame, sub-region matching and displacement tracking are performed on the images after each loading stage, thereby calculating the displacement field, principal strain field, shear strain field, and strain concentration zone distribution within the region of interest.

8. The MD-RVE-FE multi-scale in-situ SEM tensile analysis and experimental method according to claim 7, characterized in that, Step S5 specifically includes: Step S5-1: Compare the load-displacement / stress-strain response curves of the simulation and the experiment, and obtain the root mean square error, mean absolute percentage error and correlation coefficient of the load-displacement / stress-strain response curves of the simulation and the experiment. Step S5-2: Compare the local strain field distribution obtained from the simulation of the region of interest with that obtained from the microscopic digital image correlation processing, and obtain the structural similarity index, normalized cross-correlation coefficient and root mean square error of the local strain field distribution between the simulation and the experiment. Step S5-3: Compare the debonding initiation position, expansion path, quantity evolution and local strain hotspots at the interface to obtain the debonding initiation position deviation, debonding quantity error and evolution rate error, as well as the spatial consistency index between the debonding area and the local strain concentration area.

Images