Metal material surface microcrack positioning method

Abstract

The invention provides a metal material surface microcrack positioning method. The method comprises the following steps: applying a torsional load to a numerical model, calculating stress-strain distribution in a material, describing a crystal deformation behavior by adopting a crystal plasticity constitutive equation, and obtaining dislocation density distribution and a slip system activation condition in each crystal grain; according to the dislocation density distribution and the slip system activation condition, the position where microcracks are easy to form in the material is judged, and when the dislocation density exceeds a preset critical value and multiple activation slip systems exist, the position is determined to be the position where the microcracks are easy to form; according to the fracture mechanical parameters and the fracture toughness of the material, whether the microcracks are expanded or not is judged, when the stress intensity factor exceeds the fracture toughness, it is determined that the microcracks are expanded, and the crack expansion path is determined by calculating the direction of the maximum release rate.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for locating microcracks on the surface of metallic materials. Background Technology

[0002] In the field of metallic materials science and engineering, the precise localization of microcracks on the surface of metallic materials has always been a crucial aspect of research and application, especially under complex conditions of torsional loading, where this problem becomes even more challenging. When materials are subjected to torsion, the complex changes in the internal stress-strain state, such as dislocation activation, slip surface activation, and plastic flow at grain boundaries, have a profound impact on the formation and propagation trajectory of microcracks. However, our current understanding of the interaction between these microscopic mechanisms and macroscopic crack behavior is still in its early stages. Particularly in multiphase materials, the stress concentration phenomenon at different phase interfaces and the control of crack propagation paths by material anisotropy make the analysis of the problem even more complex. Furthermore, under high-temperature environments, the dynamic recrystallization and microstructural evolution of materials further exacerbate the uncertainty of microcrack localization. Faced with these challenges, how to construct a mechanical model spanning from the microscopic to the macroscopic scale, closely linking the microscopic mechanisms of torsional deformation with macroscopic crack behavior to achieve high-precision microcrack localization has become a major research topic in materials science. In summary, in the methods for locating microcracks on the surface of metallic materials, it is an urgent technical problem to explore the intrinsic relationship between the micro-mechanism under torsional deformation and the macro-crack behavior, and to construct a multi-scale mechanical model that can accurately predict the behavior of microcracks. Summary of the Invention

[0003] This invention provides a method for locating microcracks on the surface of metallic materials, mainly comprising:

[0004] Based on the microstructure characteristics and chemical composition of metallic materials, a three-dimensional digital model of the material is established. The finite element method is used to mesh the three-dimensional digital model to obtain a numerical model containing microstructure information of grains, grain boundaries, and first-phase particles.

[0005] A torsional load is applied to the numerical model to calculate the stress and strain distribution inside the material. The crystal deformation behavior is described by the crystal plastic constitutive equation, and the dislocation density distribution and slip system activation status within each grain are obtained.

[0006] Based on the dislocation density distribution and the activation status of slip systems, the location in the material that is prone to microcrack formation is determined. When the dislocation density exceeds the preset critical value and there are multiple activated slip systems, the location is determined to be a location that is prone to microcrack formation.

[0007] For the identified locations where microcracks are likely to form, a micromechanical model containing the cracks is established. The micromechanical model combines grain orientation, grain boundary characteristics and first phase distribution. The extended finite element method is used to simulate the crack propagation process to obtain the stress intensity factor at the crack tip and the J-integral fracture mechanics parameters.

[0008] Based on fracture mechanics parameters and material fracture toughness, it is determined whether microcracks will propagate. When the stress intensity factor exceeds the fracture toughness, microcrack propagation is confirmed. The crack propagation path is determined by calculating the direction of the maximum release rate of the J integral.

[0009] Acoustic emission detection method is used to monitor the material surface in real time, and the amplitude, energy and frequency characteristic parameters of the acoustic emission signal are obtained. Wavelet transform is used to perform time and frequency analysis on the signal to obtain the acoustic characteristics of microcrack initiation and propagation. Based on the arrival time difference of the acoustic emission signal, the location of the acoustic emission source is preliminarily estimated.

[0010] Based on the time-frequency characteristics of the acoustic emission signal, the preliminary estimated sound source location, and the microcrack propagation path obtained from numerical simulation, a triangulation algorithm is used to determine the specific location of the microcrack. The time difference of the acoustic emission signals received by multiple sensors is used to determine the sound source coordinates through iterative calculation. By comparing multiple positioning results, the distribution of the microcrack on the material surface is obtained.

[0011] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0012] This invention discloses a method for locating microcracks on the surface of metallic materials. The method first establishes a three-dimensional digital model of the material, and calculates the internal stress-strain distribution and dislocation density through finite element analysis and crystal plasticity constitutive equations to determine the locations where microcracks are likely to form. Then, a micromechanical model containing the crack is established, and the extended finite element method is used to simulate crack propagation to obtain fracture mechanics parameters. Simultaneously, acoustic emission detection technology is used to monitor the material surface in real time to obtain the acoustic characteristics of microcrack initiation and propagation. Finally, combining the numerical simulation results and acoustic emission signals, a triangulation algorithm is used to determine the specific location and distribution of the microcracks. This invention achieves accurate prediction and location of microcracks in metallic materials at the microscale, providing a new method for material failure analysis and life assessment. Attached Figure Description

[0013] Figure 1 This is a flowchart of a method for locating microcracks on the surface of a metallic material according to the present invention.

[0014] Figure 2 This is a schematic diagram of a method for locating microcracks on the surface of a metallic material according to the present invention.

[0015] Figure 3 This is another schematic diagram of a method for locating microcracks on the surface of a metallic material according to the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0017] like Figure 1 -3, This embodiment of a method for locating microcracks on the surface of a metallic material may specifically include:

[0018] S101. Based on the microstructure characteristics and chemical composition of metallic materials, a three-dimensional digital model of the material is established. The finite element method is used to mesh the three-dimensional digital model to obtain a numerical model containing microstructure information of grains, grain boundaries, and first-phase particles.

[0019] Scanning electron microscope (SEM) images of the microstructure and chemical composition data of metallic materials are acquired. Edge detection and region growth are performed based on the SEM images to obtain the position, shape, and size information of grains, grain boundaries, and first-phase particles. Mechanical property parameters corresponding to different element contents are retrieved from a material database based on the chemical composition data to determine the material property parameters of each phase. A 3D reconstruction method based on cross-section interpolation is used to construct a 3D digital model of the metallic material based on its position, shape, and size information. The 3D digital model is discretized into a digital representation composed of tetrahedral elements using a tetrahedral element partitioning method, and corresponding material property parameters are assigned to each tetrahedral element. The digital representation is meshed using an adaptive mesh refinement algorithm based on stress gradients, increasing the mesh density in regions with large stress gradients to obtain a finite element mesh model containing microstructure information. Boundary conditions and load cases are set based on the finite element mesh model, and constitutive equations describing the mechanical behavior of the metallic material are established using the Johnson-Cook model. The microscopic stress-strain distribution of the metallic material is calculated using a nonlinear finite element solution process based on the Newton-Raphson iterative method to determine the deformation and fracture behavior of the metallic material.

[0020] For example, scanning electron microscope (SEM) images of the microstructure and chemical composition data of metallic materials are acquired. Edge detection and region growing algorithms are used to segment and extract features from the microstructure images, identifying the position, shape, and size of grains, grain boundaries, and first-phase particles. Based on the chemical composition data, mechanical property parameters corresponding to different element contents are queried from a material database to determine the material property parameters of each phase. Based on the extracted microstructure features, a three-dimensional digital model of the metallic material is constructed using a cross-section interpolation-based 3D reconstruction method. The continuous geometric model is discretized into a digital representation composed of a large number of tetrahedral elements using a tetrahedral element partitioning method, with each tetrahedral element assigned corresponding material properties. The constructed 3D digital model is then meshed using an adaptive mesh refinement algorithm based on stress gradients. Mesh density is increased in regions with large stress gradients, and mesh refinement is performed in stress concentration regions such as grain boundaries and second-phase particles, generating a finite element mesh model containing microstructure information. Based on the generated finite element mesh model, boundary conditions and load cases were set. The Johnson-Cook model was used to describe the plastic deformation behavior of the material, and constitutive equations were established to describe the mechanical behavior of the material. A nonlinear finite element solution process based on the Newton-Raphson iterative method was used to calculate the microscopic stress-strain distribution and predict the deformation and fracture behavior of the material. Scanning electron microscope images of the microstructure of the metallic material were acquired, with a resolution of 2048x2048 pixels and 256 gray levels. The Canny edge detection algorithm was used to extract grain boundary contours, with thresholds set to a low threshold of 50 and a high threshold of 150. A region growth algorithm was used to identify grains, with the seed point selected as the local gray-level minimum point, and the growth threshold set to 10. Morphological operations were used to extract the first-phase particles, with an opening kernel size of 3x3 pixels. Chemical composition data was obtained, including the content of elements such as C 0.4%, Mn 0.7%, Si 0.3%, and Cr 1.0%. Mechanical property parameters corresponding to different element contents were queried using a material database, such as an elastic modulus of 210 GPa and a yield strength of 500 MPa. Based on the extracted microstructural features, 3D reconstruction was performed using cross-sectional interpolation. The cross-sectional spacing was set to 0.1 μm, and 100 layers were reconstructed, generating a 10 μm × 10 μm × 10 μm 3D digital model. A tetrahedral element meshing method was used, with an element size of 0.05 μm, generating approximately 8 million tetrahedral elements. Each tetrahedral element was assigned corresponding attribute values ​​based on material property parameters. The 3D digital model was meshed using an adaptive mesh refinement algorithm based on stress gradients. The initial mesh size was set to 0.1 μm, and the stress gradient threshold was set to 100 MPa / μm. When the local stress gradient exceeded the threshold, the mesh size was reduced by half. Forced mesh refinement was applied within a 5-element range around grain boundaries and first-phase particles, with a minimum mesh size limit of 0.01 μm. The final finite element mesh model contained approximately 12 million elements was generated.Based on the generated finite element mesh model, the boundary conditions were set as follows: the bottom surface was fixed, and a uniform pressure of 100 MPa was applied to the top surface. The Johnson-Cook model was used to describe the plastic deformation behavior of the material. The model parameters were set as follows: initial yield stress A = 500 MPa, B = 600 MPa; strain hardening exponent n = 0.4; temperature softening coefficient C = 0.015; temperature softening exponent m = 1.0; and B is a parameter related to strain hardening, reflecting the degree of hardening of the material with increasing plastic strain. The Newton-Raphson iterative method was used to solve the nonlinear equations, with the convergence criterion set to a residual less than 1e⁻⁶. The micro-stress-strain distribution was calculated through 100 loading steps, with each step limited to 20 iterations. Based on the von Mises equivalent stress distribution, it was determined whether the element underwent plastic deformation. When the equivalent stress exceeded 700 MPa, fracture was considered to have occurred in that region. Finally, the predicted deformation and fracture behavior of the material under load were obtained.

[0021] S102. Apply torsional load to the numerical model, calculate the stress and strain distribution inside the material, use the crystal plastic constitutive equation to describe the crystal deformation behavior, and obtain the dislocation density distribution and slip system activation status within each grain.

[0022] The continuum is discretized into a single-element mesh using the finite element method. A global stiffness matrix and load vector are then established based on this mesh. The nonlinear equations are solved using the Newton-Raphson iteration method. If the norm of the residual vector is less than a preset threshold, the nodal displacement field within the material is obtained. The strain tensor and stress tensor of each element are calculated based on this nodal displacement field. For each grain, the decomposed shear stress on the slip surface is calculated. If the decomposed shear stress exceeds the critical decomposed shear stress, the corresponding slip system is activated. The dislocation density evolution on each slip system is calculated using the Orowan equation. The rate of change of dislocation density is determined by the dislocation multiplication rate and annihilation rate. Microstructural evolution information during material deformation is obtained based on this dislocation density evolution, including the activation status of slip systems and the dislocation density distribution.

[0023] For example, a torsional load is set in the numerical model by applying torques in opposite directions at both ends of a cylindrical specimen. The continuum is discretized into a mesh using the finite element method, establishing a global stiffness matrix and load vector. The nonlinear equations are solved using the Newton-Raphson iteration, calculating the residual vector for each iteration. Convergence is considered achieved when the residual norm is less than a preset threshold, yielding the nodal displacement field within the material. The strain tensor for each element is calculated based on the nodal displacement field, and the stress tensor is calculated using the applicable rate-dependent crystal plasticity constitutive equation. Considering the anisotropy of the crystal, a crystal orientation matrix is ​​constructed using Euler angles, transforming the stress and strain in the single-crystal coordinate system to the global coordinate system. The crystal plasticity constitutive equation considers hardening effects, including self-hardening and cross-hardening. For each grain, the decomposed shear stress on the slip surface is calculated. When the decomposed shear stress exceeds the critical decomposed shear stress, the corresponding slip system is activated. The critical decomposed shear stress is correlated with the dislocation density through a hardening relation. The plastic shear strain increment on the activated slip system is calculated, and the crystal orientation is updated. The dislocation density evolution on each slip system was calculated using the Orowan equation, with the rate of change determined by the dislocation multiplication rate and annihilation rate. The Frank-Read source multiplication and dislocation loop propagation mechanisms, as well as dislocation crossing and annihilation processes, were considered. The dislocation density distribution within each grain was updated to obtain microstructural evolution information during material deformation, including the activation status of slip systems and dislocation density distribution. A torque of 1000 N·m was applied to a numerical model of a cylindrical specimen with a diameter of 10 mm and a height of 30 mm. A mesh of 50,000 elements was generated using 20-node hexahedral elements. The Newton-Raphson iterative method was used to solve the nonlinear equations, with a maximum iteration count of 50 and a convergence threshold of 1e⁻⁶. In each iteration, the residual vector was calculated and the stiffness matrix was updated until convergence or the maximum iteration count was reached. After obtaining the nodal displacement field, the strain tensor of each element was calculated. The rate-dependent crystal plastic constitutive equation was used, considering the hardening effect, with the hardening modulus set to 1000 MPa. A crystal orientation matrix is ​​constructed using Euler angles (φ1, Φ, φ2), where φ1∈[0, 2π], Φ∈[0, π], and φ2∈[0, 2π], uniformly and randomly distributed. For face-centered cubic metals, 12 {111}<110> slip systems are considered, i.e., the {111} family of planes, with the slip direction being the <110> direction. The decomposed shear stress on each slip system is calculated, with the initial critical decomposed shear stress set to 100 MPa. When the decomposed shear stress exceeds the critical value, the slip system is activated and the plastic shear strain increment is calculated. The critical decomposed shear stress is updated using the Taylor hardening relation, with the Taylor factor set to 1.1. The dislocation density evolution is calculated using the Orowan equation, with the initial dislocation density set to 1e12m^-2, considering Frank-Read source multiplication and dislocation loop expansion, with a multiplication factor of 0.1 and an annihilation factor of 10.

[0024] S103. Based on the dislocation density distribution and the activation status of slip systems, determine the locations within the material where microcracks are likely to form. When the dislocation density exceeds a preset critical value and multiple active slip systems exist, this location is determined to be a location where microcracks are likely to form.

[0025] The process involves: acquiring dislocation density distribution data for each region of the material, storing the dislocation density value of each grid cell using a three-dimensional array; calculating a critical dislocation density value based on the data; determining whether the dislocation density value of each grid cell exceeds the critical value, and marking the grid cell as a potential microcrack formation region if it does; reading the slip system activation status of each grain, recording the activated slip system number in each grain using a two-dimensional array; counting the number of activated slip systems in each grain; and determining if the number of activated slip systems is greater than or equal to a preset threshold. If the grain is identified as a multi-slip system activation region, then a Boolean array representing the locations prone to microcrack formation is generated based on the labeling results of the potential microcrack formation regions and the multi-slip system activation regions. Local stress-strain states are calculated for the locations prone to microcrack formation, including dislocation density gradient, grain boundary mismatch, and local stress triaxiality. A microcrack formation probability model is established based on the local stress-strain states, and this model is used to calculate the microcrack formation index. A three-dimensional distribution map of microcrack-prone regions is generated by combining grain boundary characteristics and grain orientation data.

[0026] For example, dislocation density distribution data for each region of the material is acquired, and the dislocation density value of each grid cell is stored in a three-dimensional array. Based on the material's yield strength and shear modulus, the critical dislocation density value is calculated using the Taylor relation. If the dislocation density exceeds the calculated critical value, the cell is marked as a potential microcrack formation region. The slip system activation status of each grain is read, and the activated slip system number in each grain is recorded using a two-dimensional array. The number of activated slip systems is counted, and a threshold of 3 is set. If the number of activated slip systems is greater than or equal to 3, the grain is marked as a multi-slip system activation region. Based on the dislocation density distribution and the marking results of slip system activation, a logical AND operation is used to compare the corresponding positions of the elements in the dislocation density supercritical array and the multi-slip system activation array. A new Boolean array is generated to represent the locations prone to microcrack formation, and these regions are identified as locations prone to microcrack formation. For the identified locations prone to microcrack formation, the local stress-strain state is calculated. A microcrack formation probability model is established considering factors such as dislocation density gradient, grain boundary mismatch, and local stress triaxiality. Weighting coefficients are set, and the microcrack formation index is calculated by comprehensively considering multiple factors. A three-dimensional distribution map of microcrack-prone regions is generated by combining grain boundary characteristics and grain orientation data. Dislocation density values ​​for each region of the material, ranging from 1e10 to 1e15m⁻², are stored in a 100×100×100 three-dimensional array. Assuming the material's yield strength σy is 300 MPa and shear modulus G is 80 GPa, the critical dislocation density value is calculated according to the Taylor relation ρc=(σy / (αGb))², where the constant α=0.5 and the modulus b of the Burgers vector is 2.5e⁻¹⁰ m, yielding a critical value of 1.4e14m⁻². A 3×3×3 three-dimensional convolution kernel is used to scan the dislocation density array, marking regions exceeding the critical value. Simultaneously, the activation status of 12 possible slip systems is recorded in a 50×50×50 two-dimensional array, represented by 0 and 1. A sliding window method is used to count the number of active slip systems in each grain, with a window size of 5×5×5. When the number of active slip systems within the window is ≥3, the central element is marked as a multi-slip system activation region. An element-wise logical AND operation is performed on the two marked arrays to generate a 100×100×100 Boolean array representing locations prone to microcrack formation. The local stress-strain state at these locations is calculated, considering the dislocation density gradient (threshold set to 1e13m⁻² / μm), grain boundary mismatch (critical angle 5°), and local stress triaxiality (critical value 0.6). Weights of 0.4, 0.3, and 0.3 are assigned to these three factors, respectively, to calculate the comprehensive microcrack formation index. Finally, based on the calculation results, the probability of microcrack formation is marked with different colors in three-dimensional space, generating a three-dimensional distribution heatmap of microcrack-prone regions.

[0027] S104. For the identified locations where microcracks are prone to form, a micromechanical model incorporating the cracks is established. This micromechanical model, combined with grain orientation, grain boundary characteristics, and first-phase distribution, is used to simulate the crack propagation process using the extended finite element method, yielding the crack tip stress intensity factor and J-integral fracture mechanics parameters.

[0028] A grain structure was constructed using the Voronoi segmentation method to obtain a micromechanical model containing cracks. Mesh generation was performed based on the micromechanical model, with an adaptive refinement strategy (h) applied to the crack tip region. Based on the mesh generation results, a constitutive equation incorporating crystal anisotropy was established, using the generalized Hooke's law to describe the crystal's elastic behavior. The displacement field was calculated based on the constitutive equation, and the stress intensity factor at the crack tip was obtained from the displacement field. If the stress intensity factor was greater than a preset threshold, the J integral value was calculated using the domain integration method. The crack propagation direction was determined based on the J integral value, which was determined using the maximum circumferential stress criterion. The crack propagation rate was calculated using the Paris formula to obtain the crack propagation path.

[0029] For example, based on the identified locations where microcracks are likely to form, a micromechanical model incorporating the cracks is constructed. The grain structure is built using the Voronoi segmentation method, and grain boundary characteristics are described using the boundary element method. A 3D reconstruction algorithm generates a geometric model containing grains, grain boundaries, and the first phase, assigning grain orientation, grain boundary characteristics, and first phase distribution information to the corresponding geometric entities. Tetrahedral elements are used for mesh generation. An adaptive refinement strategy (h-adaptive refinement) is applied to the crack tip region to mesh the micromechanical model, generating a finite element mesh model containing the crack. Extended degrees of freedom are introduced into the elements containing the crack. A constitutive equation considering crystal anisotropy is established, the generalized Hooke's law is used to describe the elastic behavior of the crystal, and the Hill yield criterion is used to characterize plastic deformation. Grain orientation information is converted into a material stiffness matrix. Boundary conditions and loads are set, and the displacement field is solved using the extended finite element method to calculate the stress-strain distribution. The stress intensity factor at the crack tip is calculated based on the displacement field, and the J-integral value is calculated using the domain integration method, considering path independence. The crack propagation direction was determined based on the maximum circumferential stress criterion, and the crack propagation rate was calculated using the Paris formula. The crack morphology was updated, and the calculation process was repeated to obtain the crack propagation path. Within a 100×100×100μm³ cubic region, 500 grains were generated using the Voronoi method, with an average grain size of 20μm. Grain boundary characteristics were described using the boundary element method, with a grain boundary thickness of 0.5μm and a grain boundary strength of 80% of the intragranular strength. Fifty spherical first-phase particles with diameters ranging from 2 to 5μm were randomly distributed in the model. The initial crack was set as an ellipse with a length of 10μm and a width of 0.1μm. Tetrahedral elements were used for mesh generation, with an initial element size of 2μm. An h-type adaptive refinement strategy was applied within a 5μm range at the crack tip, with a minimum element size of 0.1μm, generating approximately 1 million elements in total. The elastic behavior of the crystal is described by the generalized Hooke's law. Taking a face-centered cubic structure as an example, the elastic constants are set as C11 = 204 GPa, C12 = 137 GPa, and C44 = 126 GPa. The Hill yield criterion is used, and the initial yield strength is set to 200 MPa. A tensile load of 100 MPa is applied to the top of the model, while the bottom is fixed. The displacement field is solved using the extended finite element method, with an iterative convergence accuracy of 1e⁻⁶. The calculated crack tip stress intensity factor KI = 5 MPa·m⁰.⁵ is obtained. The J integral value is calculated using an 8-point Gaussian integral, and five integration paths are used to verify path independence; the J integral value is 0.1 kJ / m³. The crack propagation direction is determined based on the maximum circumferential stress criterion, and the crack propagation rate is calculated using the Paris formula da / dN = C(ΔK)⁻¹m, where the material constants C = 1e⁻¹¹ and m = 3. a is the crack length, N is the number of stress cycles, and ΔK is the range of stress intensity factors, which is a measure of the stress field near the crack tip. The crack grows by 0.1 μm per cycle, the crack morphology is updated, and the calculation is repeated 50 times to obtain the crack propagation path.

[0030] S105. Based on fracture mechanics parameters and the fracture toughness of the material, determine whether microcracks will propagate. When the stress intensity factor exceeds the fracture toughness, microcrack propagation is confirmed, and the crack propagation path is determined by calculating the direction of the maximum release rate of the J integral.

[0031] The fracture mechanics parameters and fracture toughness values ​​of the material are obtained. The stress intensity factor is calculated using finite element post-processing based on these parameters. The J-integral value is calculated using the virtual displacement method, and the energy release rate is obtained by numerical integration along multiple integration paths around the crack tip. If the crack propagation direction is determined, a crack propagation increment is set, and the crack extends according to the direction of the maximum energy release rate. The crack morphology is updated using the crack propagation increment method, and a new mesh is generated based on the updated crack morphology, and the stress field is calculated. An experimental observation scheme is set up to measure the strain field on the sample surface, and the accuracy of the numerical simulation results is verified based on the strain field data.

[0032] For example, fracture mechanics parameters and material fracture toughness values ​​are obtained. The stress intensity factor is calculated using finite element post-processing. The Irwin fracture criterion is applied, and microcrack propagation is determined when the stress intensity factor KI at the crack tip exceeds the material's fracture toughness KIC. The J-integral value is calculated using the virtual displacement method. A unit virtual displacement is applied to the crack tip, and the virtual power is calculated to obtain the J-integral value. Multiple integration paths are selected around the crack tip, and the energy release rate is solved through numerical integration. The J-integral value is calculated every 5° within a 360° range at the crack tip, and the direction corresponding to the maximum value is identified. The crack propagation direction is determined using the maximum energy release rate criterion. The J-integral is used in fracture mechanics to describe the energy release rate at the crack tip. Based on the determined crack propagation direction, the crack propagation increment is set to 0.1 mm, and the crack extends along the direction of the maximum energy release rate. The crack morphology is updated using the crack propagation increment method, the mesh is regenerated, and the stress field is calculated. The first two steps are repeated to obtain the complete crack propagation path. An experimental observation scheme is set up, and the strain field on the sample surface is measured using in-situ tensile testing and digital image correlation technology. Deformation images were acquired using a high-resolution CCD camera, and the strain field was calculated using digital image correlation software. A piezoelectric sensor array was deployed to acquire acoustic emission signals and locate the crack initiation point. The crack propagation process was recorded using acoustic emission monitoring. Experimental results were compared with numerical simulation results to verify the accuracy of the simulation. For a surface crack with a length of 10 mm, the stress intensity factor KI = 35 MPa·m⁰.⁵ was calculated using finite element post-processing. The fracture toughness KIC of the material was 40 MPa·m⁰.⁵, and the Irwin fracture criterion was used to determine crack propagation. A virtual displacement of 1 μm was applied at the crack tip, and the calculated J integral value was 0.15 kJ / m². Five integration paths with radii of 0.1 mm, 0.2 mm, 0.3 mm, 0.4 mm, and 0.5 mm were selected, and the energy release rate was solved using the Simpson numerical integration method. The J integral value was calculated every 5° within a 360° range from the crack tip, for a total of 72 directions. The direction corresponding to the maximum J integral value of 0.18 kJ / m² was identified at 70°. The crack propagation increment was set to 0.1 mm, and the crack was extended along the 70° direction. After updating the crack morphology, a mesh containing approximately 500,000 tetrahedral elements was regenerated, and the calculation process was repeated 10 times to obtain a crack propagation path with a total length of 1 mm. In the experiment, a CCD camera with a resolution of 4096×3072 pixels was used, with a sampling frequency of 1 Hz, to capture 500 deformation images. Using digital image correlation software, the sub-region size was set to 21×21 pixels, and the step size to 5 pixels, resulting in a strain field resolution of 0.01%. Eight piezoelectric sensors were deployed, with a sampling frequency of 5 MHz and a preset threshold of 40 dB, recording 153 acoustic emission signals. The crack initiation location was determined using a triangulation algorithm, with an average deviation of 0.08 mm from the crack path predicted by numerical simulation.

[0033] S106. The material surface is monitored in real time using acoustic emission detection methods to obtain the amplitude, energy, and frequency characteristic parameters of the acoustic emission signal. Time-frequency analysis of the signal is performed using wavelet transform to obtain the acoustic characteristics of microcrack initiation and propagation. The location of the acoustic emission source is preliminarily estimated based on the arrival time difference of the acoustic emission signal.

[0034] The process involves acquiring raw acoustic emission signals from an array of acoustic emission sensors arranged on a material surface; filtering the raw acoustic emission signals using a bandpass filter to obtain a first signal; performing wavelet threshold denoising on the first signal to obtain a second signal; calculating signal parameters based on the second signal, including peak amplitude, rise time, duration, count, energy, average frequency, and peak frequency; performing wavelet transform on the second signal to obtain wavelet coefficients; plotting a time-frequency diagram based on the wavelet coefficients to identify characteristic frequency bands and energy distributions during microcrack initiation and propagation; calculating the coordinates of the acoustic emission source based on the time difference between the arrival of the acoustic emission signals at multiple sensors; solving the overdetermined equations using the least squares method if the time difference meets a preset condition; and plotting a crack evolution diagram based on the acoustic emission source coordinates, which is used to track the initiation location and propagation path of microcracks.

[0035] For example, an array of acoustic emission sensors is arranged on the material surface, and acoustic emission signals are continuously acquired via a data acquisition card. The sampling frequency and threshold are set, and the acquired raw signals are preprocessed. A bandpass filter (50kHz-1MHz) and wavelet threshold denoising are used to process the raw signals, resulting in clear acoustic emission waveforms. Feature extraction is performed on the preprocessed acoustic emission signals, calculating the signal amplitude, energy, and frequency domain parameters, including peak amplitude, rise time, duration, count, energy, average frequency, and peak frequency. The signal spectrum is calculated using Fast Fourier Transform, and the dominant frequency and frequency centroid are extracted to form the feature vector of the acoustic emission event. Wavelet transform is used to perform time-frequency analysis on the acoustic emission signals. Daubechies wavelets are selected based on signal characteristics, and a 5-level decomposition is determined according to the signal frequency range. The wavelet coefficients of the signal are calculated, and a time-frequency diagram is plotted to identify the characteristic frequency bands and energy distribution during microcrack initiation and propagation. Based on the time difference of the acoustic emission signals arriving at multiple sensors, the overdetermined equations are solved using the least squares method to calculate the coordinates of the acoustic emission source. By combining the time series of acoustic emission events, the initiation location and propagation path of microcracks were tracked, and a crack evolution diagram was plotted. Monte Carlo simulation was used to evaluate the positioning accuracy; within a 95% confidence interval, the positioning error was less than 5 mm. Four piezoelectric acoustic emission sensors were arranged on a 100×100 mm³ material surface, using a 16-bit A / D converter, a sampling frequency of 5 MHz, and a threshold of 40 dB. The acquired raw signal was processed by a 50 kHz–1 MHz bandpass filter, decomposed into four levels using a db4 wavelet, and denoised with soft thresholding to reconstruct a clear waveform. Features were extracted from the processed signal, including peak amplitude of 85 dB, rise time of 25 μs, duration of 150 μs, count of 56, and energy of 3.2 × 10⁻⁷ J. A 2048-point FFT was used to calculate the spectrum, yielding a dominant frequency of 320 kHz and a centroid of 280 kHz. A 5-level decomposition using the db5 wavelet was employed. Time-frequency analysis revealed that the energy concentration during microcrack initiation was concentrated in the 200-300kHz frequency band, while the propagation process was mainly concentrated in the 150-250kHz band. Using the time difference of signals received from four sensors, an overdetermined system of equations was constructed and solved using the least squares method to obtain the sound source coordinates (35.2mm, 42.8mm). 1000 Monte Carlo simulations were performed, with a localization error of 3.8mm within the 95% confidence interval. Based on the location and time information of 100 consecutive acoustic emission events, a crack evolution diagram was plotted, showing that the crack initiated at coordinates (30mm, 40mm) and propagated 15mm along a 70° direction.

[0036] S107. Based on the time-frequency characteristics of the acoustic emission signal, the preliminary estimated sound source location, and the microcrack propagation path obtained from numerical simulation, the specific location of the microcrack is determined using a triangulation algorithm; the sound source coordinates are determined by iterative calculation using the time difference of the acoustic emission signals received by multiple sensors; and the distribution of the microcrack on the material surface is obtained by comparing multiple positioning results.

[0037] The time-frequency characteristics of the acoustic emission signal are obtained. The main frequency and energy distribution of the acoustic emission signal are extracted using short-time Fourier transform to obtain the characteristic data of the acoustic emission signal. Based on the acoustic emission signals received by multiple sensors, the time difference of arrival (TDO) is calculated using the generalized cross-correlation function GCC-PHAT to determine the accuracy of the time difference estimation. A system of nonlinear equations is constructed to describe the relationship between the sound source location and the time difference. The system of nonlinear equations is solved using Newton's iteration method to obtain the sound source coordinates. Statistical analysis is performed on multiple positioning results. If a data point deviates from the mean by more than three standard deviations, the 3σ criterion is used to remove the data point. The confidence interval of the positioning result is calculated using Monte Carlo simulation to determine whether the positioning result is within a preset confidence interval.

[0038] For example, the time-frequency characteristics of the acoustic emission signal are obtained, and the dominant frequency and energy distribution of the signal are extracted using short-time Fourier transform. The preliminary estimated sound source location and the microcrack propagation path obtained from numerical simulation are extracted from the signal database, and these data are used as input parameters for the triangulation algorithm.

[0039] Based on the acoustic emission signals received from multiple sensors, the time difference of signal arrival is calculated. The generalized cross-correlation function GCC-PHAT is used to improve the accuracy of time difference estimation. A set of nonlinear equations is constructed to describe the relationship between the sound source location and the time difference.

[0040] The nonlinear equations were solved using Newton's iterative method, with the initial value set as a preliminary estimated location, an iteration step size of 0.1 mm, and a convergence threshold of 1e-6. The sound source coordinates were obtained through multiple iterations, and the intermediate results and final converged coordinates of each iteration were recorded.

[0041] Statistical analysis was performed on multiple location results, and data points deviating more than three standard deviations from the mean were removed using the 3σ criterion. The mean and standard deviation were calculated, and a heat map of the microcrack distribution on the material surface was plotted, marking high-probability areas. The 95% confidence interval of the location results was calculated using Monte Carlo simulation to evaluate the location accuracy. The location results were compared and verified with the numerical simulation path.

[0042] For a 100×100 mm³ material surface, four acoustic emission sensors were arranged. The acquired signals underwent a 1024-point short-time Fourier transform with a time window length of 0.2 ms and an overlap rate of 50%. The dominant frequency range was extracted to be 200–300 kHz, with a peak energy of 3.5 × 10⁻⁷ J. The preliminary estimated sound source location (45 mm, 55 mm) and the numerically simulated crack path were obtained from the database. The GCC-PHAT algorithm was used to calculate the signal time difference, with a sampling frequency of 5 MHz and a cross-correlation window length of 1024 points, yielding a time difference accuracy of 0.2 μs. Four nonlinear equations were constructed and solved using Newton's iteration method, with initial values ​​set at (45 mm, 55 mm), an iteration step size of 0.1 mm, a convergence threshold of 1e⁻⁶, and a maximum number of iterations of 100. After 50 iterations, the sound source coordinates were obtained as (46.3 mm, 54.8 mm). The localization process was repeated 100 times. Five outliers were removed using the 3σ criterion, leaving 95 valid results with an average of (46.2 mm, 54.9 mm) and standard deviations of (0.3 mm, 0.2 mm). Heatmaps showed that 90% of the localization results were concentrated within a 2 mm × 2 mm area. Through 10,000 Monte Carlo simulations, a 95% confidence interval of ±0.6 mm was calculated. Comparing the localization results with the numerical simulation path, the average deviation was 0.4 mm, and the maximum deviation was 0.9 mm.

[0043] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

Claims

1. A method for locating microcracks on the surface of a metal material, characterized in that: The method comprises: According to the microstructure characteristics and chemical composition of the metal material, a three-dimensional digital model of the material is established, and the three-dimensional digital model is meshed using the finite element method to obtain a numerical model containing the microstructure information of grains, grain boundaries, and first-phase particles; Apply torsional load to the numerical model, calculate the stress and strain distribution inside the material, use the crystal plasticity constitutive equation to describe the crystal deformation behavior, and obtain the dislocation density distribution and slip system activation in each grain; According to the dislocation density distribution and the slip system activation, the location where microcracks are likely to form inside the material is determined. When the dislocation density exceeds the preset critical value and there are multiple activated slip systems, the location is determined to be a location where microcracks are likely to form. For the locations that are prone to microcracks, a micromechanical model including cracks is established. The micromechanical model combines grain orientation, grain boundary characteristics and first phase distribution, and uses the extended finite element method to simulate the crack propagation process to obtain the stress intensity factor at the crack tip and the J-integral fracture mechanics parameters. According to the fracture mechanics parameters and the fracture toughness of the material, it is judged whether the microcracks extend. When the stress intensity factor exceeds the fracture toughness, the microcrack extension is determined, and the crack extension path is determined by calculating the direction of the maximum release rate of the J integral. The acoustic emission detection method is used to monitor the material surface in real time, obtain the amplitude, energy and frequency characteristic parameters of the acoustic emission signal, and perform time-frequency analysis on the signal through wavelet transform to obtain the acoustic characteristics of microcrack initiation and expansion. According to the arrival time difference of the acoustic emission signal, the location of the acoustic emission source is preliminarily estimated; According to the time-frequency characteristics of the acoustic emission signal, the preliminary estimated sound source position and the microcrack propagation path obtained by numerical simulation, the triangulation positioning algorithm is used to determine the specific position of the microcrack. The time difference of the acoustic emission signals received by multiple sensors is used to determine the coordinates of the sound source through iterative calculation. By comparing multiple positioning results, the distribution of microcracks on the material surface is obtained.

2. The method according to claim 1, wherein: The method of establishing a three-dimensional digital model of the material according to the microstructure characteristics and chemical composition of the metal material, meshing the three-dimensional digital model using the finite element method, and obtaining a numerical model containing microstructure information of grains, grain boundaries, and first-phase particles includes: Obtain the microstructure scanning electron microscope image and chemical composition data of the metal material, perform edge detection and regional growth based on the microstructure scanning electron microscope image, and obtain the position, shape and size information of the grains, grain boundaries and first phase particles; According to the chemical composition data, the mechanical properties parameters corresponding to different element contents are queried from the material database to determine the material property parameters of each phase; A 3D reconstruction method based on cross-section interpolation is used to construct a 3D digital model of metal materials according to position, shape and size information; Discretizing the three-dimensional digital model into a digital representation composed of tetrahedral units by a tetrahedral unit partitioning method, and assigning corresponding material property parameters to the tetrahedral units; The digital representation is meshed, and an adaptive mesh encryption algorithm based on stress gradient is used to increase the mesh density in the area with large stress gradient to obtain a finite element mesh model containing microstructural information; Boundary conditions and load conditions are set according to the finite element mesh model, and a constitutive equation describing the mechanical behavior of metal materials is established using the Johnson-Cook model; Based on the nonlinear finite element solution process of the Newton-Raphson iterative method, the microscopic stress and strain distribution of metal materials is calculated to determine the deformation and fracture behavior of metal materials.

3. The method according to claim 1, wherein: The method applies a torsional load to the numerical model, calculates the stress and strain distribution inside the material, uses the crystal plasticity constitutive equation to describe the crystal deformation behavior, and obtains the dislocation density distribution and slip system activation status in each grain, including: The continuum is discretized into unit grids by using a finite element discretization method, and a global stiffness matrix and a load vector are established according to the unit grids; The nonlinear equations are solved by Newton-Raphson iteration. If the norm of the residual vector is less than the preset threshold, the node displacement field inside the material is obtained. Calculating the strain tensor of each unit according to the node displacement field and simultaneously calculating the stress tensor; For each grain, the decomposed shear stress of the applied stress on the slip plane is calculated, and if the decomposed shear stress exceeds the critical decomposed shear stress, the corresponding slip system is activated; The Orowan equation is used to calculate the dislocation density evolution on each slip system, wherein the dislocation density change rate is determined by the dislocation multiplication rate and the annihilation rate; The microstructure evolution information during the material deformation process is obtained according to the dislocation density evolution, and the microstructure evolution information includes the activation status of the slip system and the dislocation density distribution.

4. The method according to claim 1, wherein: The method comprises: judging the location where microcracks are likely to form inside the material according to the dislocation density distribution and the slip system activation; and determining that the location is a location where microcracks are likely to form when the dislocation density exceeds a preset critical value and there are multiple activated slip systems, including: Obtaining dislocation density distribution data of each region of the material, wherein the dislocation density distribution data uses a three-dimensional array to store the dislocation density value of each grid unit; Calculating a dislocation density critical value according to the dislocation density distribution data; Determine whether the dislocation density value of each grid unit exceeds the dislocation density critical value, and if so, mark the grid unit as a potential microcrack formation area; Reading the slip system activation status of each grain, wherein the slip system activation status uses a two-dimensional array to record the slip system number activated in each grain; Counting the number of activated slip systems in each grain, and if the number of activated slip systems is greater than or equal to a preset threshold, marking the grain as a multi-slip system activation region; Generate a Boolean array representing a location where microcracks are easily formed according to the marking results of the potential microcrack formation area and the multi-slip system activation area; Calculating the local stress-strain state for the position prone to microcrack formation, wherein the local stress-strain state includes dislocation density gradient, grain boundary mismatch and local stress triaxiality; Establishing a microcrack formation probability model according to the local stress-strain state, wherein the microcrack formation probability model is used to calculate a microcrack formation index; Combining grain boundary characteristics and grain orientation data, a three-dimensional distribution map of microcrack-prone areas is generated.

5. The method according to claim 1, wherein: The method establishes a micromechanical model including cracks for the determined position where microcracks are likely to form; the micromechanical model combines grain orientation, grain boundary characteristics and first phase distribution, and uses an extended finite element method to simulate the crack propagation process to obtain the crack tip stress intensity factor and J-integral fracture mechanics parameters, including: The grain structure is constructed using the Voronoi segmentation method, and a micromechanical model including cracks is obtained; Meshing is performed according to the micromechanical model, wherein an h adaptive refinement strategy is adopted in the crack tip region; Based on the meshing results, a constitutive equation combining crystal anisotropy is established, wherein the constitutive equation adopts the generalized Hooke's law to describe the elastic behavior of the crystal; Calculating a displacement field based on the constitutive equation, and obtaining a crack tip stress intensity factor from the displacement field; If the stress intensity factor is greater than a preset threshold, the J integral value is calculated using the domain integration method; Determining a crack propagation direction according to the J-integral value, wherein the crack propagation direction is determined based on a maximum circumferential stress criterion; The crack growth rate is calculated by the Paris formula, and the crack growth path is obtained.

6. The method according to claim 1, wherein: Determining whether the microcracks extend according to the fracture mechanics parameters and the fracture toughness of the material; When the stress intensity factor exceeds the fracture toughness, the microcrack extension is determined, and the crack extension path is determined by calculating the direction of the maximum release rate of the J integral, including: Obtaining fracture mechanics parameters and fracture toughness values ​​of materials, and calculating stress intensity factors through finite element post-processing according to the fracture mechanics parameters and the fracture toughness values; The virtual displacement method is used to calculate the J integral value, and the energy release rate is obtained by numerically integrating multiple integral paths around the crack tip. If the crack extension direction is determined, the crack extension increment is set to extend the crack in the direction of the maximum energy release rate; The crack morphology is updated by using the crack extension increment method, and the mesh is regenerated and the stress field is calculated based on the updated crack morphology; An experimental observation plan was set up to measure the strain field on the sample surface, and the accuracy of the numerical simulation results was verified based on the strain field data.

7. The method according to claim 1, wherein: The acoustic emission detection method is used to monitor the material surface in real time, obtain the amplitude, energy and frequency characteristic parameters of the acoustic emission signal, and perform time-frequency analysis on the signal through wavelet transform to obtain the acoustic characteristics of microcrack initiation and expansion; According to the arrival time difference of the acoustic emission signal, the location of the acoustic emission source is preliminarily estimated, including: Acquiring an original acoustic emission signal collected by an acoustic emission sensor array, wherein the acoustic emission sensor array is arranged on the surface of the material; Using a bandpass filter to filter the original acoustic emission signal to obtain a first signal; Performing wavelet threshold denoising processing on the first signal to obtain a second signal; Calculating signal parameters based on the second signal, the signal parameters including peak amplitude, rise time, duration, count, energy, average frequency and peak frequency; Performing wavelet transform on the second signal to obtain wavelet coefficients; Drawing a time-frequency diagram based on the wavelet coefficients to identify characteristic frequency bands and energy distribution during microcrack initiation and propagation; Calculating the coordinates of the acoustic emission source according to the time difference between the acoustic emission signals arriving at the multiple sensors; If the time difference satisfies a preset condition, the least square method is used to solve the overdetermined system of equations; A crack evolution diagram is drawn according to the acoustic emission source coordinates, and the crack evolution diagram is used to track the initiation position and expansion path of microcracks.

8. The method according to claim 1, wherein: The specific location of the microcrack is determined by using a triangulation positioning algorithm based on the time-frequency characteristics of the acoustic emission signal, the preliminary estimated sound source location and the microcrack extension path obtained by numerical simulation; the sound source coordinates are determined by iterative calculation using the time difference of the acoustic emission signals received by multiple sensors; By comparing multiple positioning results, the distribution of microcracks on the material surface is obtained, including: Acquiring the time-frequency characteristics of the acoustic emission signal, extracting the main frequency and energy distribution of the acoustic emission signal by short-time Fourier transform, and obtaining characteristic data of the acoustic emission signal; According to the acoustic emission signals received by multiple sensors, the generalized cross-correlation function GCC-PHAT is used to calculate the signal arrival time difference and determine the time difference estimation accuracy; Constructing a nonlinear equation group to describe the relationship between the sound source position and the time difference, and solving the nonlinear equation group using Newton iteration method to obtain the coordinates of the sound source; Perform statistical analysis on multiple positioning results. If a data point deviates from the mean by more than three standard deviations, the data point is eliminated using the 3σ criterion. The confidence interval of the positioning result is calculated through Monte Carlo simulation to determine whether the positioning result is within a preset confidence interval.

Images