A surface flaw analysis method combined with metal material fatigue life prediction

Abstract

The application provides a surface flaw analysis method combined with metal material fatigue life prediction, comprising: performing mesh optimization and refinement processing on an initial three-dimensional model, performing fine reconstruction on a local area, and obtaining a three-dimensional geometric model of a metal material surface, the three-dimensional geometric model containing surface flaw information of a microscale; performing regional division on the three-dimensional geometric model, identifying and extracting a surface flaw region, performing edge detection on the surface flaw region, and determining accurate geometric contours and spatial position information of the flaw; using finite element analysis software, importing the three-dimensional geometric model of the metal material and a flaw mathematical description model, setting material properties, boundary conditions and load working conditions, and establishing a finite element analysis model containing micro flaws; performing stress and strain analysis on the finite element analysis model of the micro flaw, calculating stress distribution and stress concentration coefficients around the flaw, and using fracture mechanics theory to evaluate the influence of the flaw on fatigue crack initiation and propagation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to a surface flaw analysis method combined with metal material fatigue life prediction. BACKGROUND

[0002] In the field of metal material fatigue life prediction, the accurate characterization of surface flaws has always been a thorny problem. Traditional two-dimensional surface detection methods are difficult to fully capture the spatial geometric characteristics of flaws, resulting in limited accuracy of fatigue life prediction models. Although reverse engineering provides the possibility for three-dimensional reconstruction of flaws, it still faces many challenges in practical applications. For micron-level surface defects, existing three-dimensional reconstruction algorithms often struggle to accurately capture their fine structures. Three-dimensional point cloud data processing is time-consuming and difficult to meet the demand for rapid detection. In addition, how to seamlessly integrate the reconstructed three-dimensional flaw model into finite element analysis to achieve efficient stress concentration calculation is also a problem that needs to be solved. Finally, how to introduce a dynamic flaw evolution mechanism into the fatigue life prediction model to more accurately reflect the actual service conditions of the material is an important direction of current research. The solution of these technical problems will directly affect the accuracy and reliability of fatigue life prediction, and has important significance for improving the safety and service life of metal components. SUMMARY

[0003] The present application provides a surface flaw analysis method combined with metal material fatigue life prediction, mainly including:

[0004] A three-dimensional laser scanner is used to scan the surface of the metal material to obtain surface topography point cloud data. According to the density and distribution characteristics of the surface topography point cloud data, an initial three-dimensional model of the metal material surface is constructed by a triangular mesh reconstruction algorithm;

[0005] The initial three-dimensional model is subjected to grid optimization and refinement processing, and the local area is subjected to fine reconstruction to obtain a three-dimensional geometric model of the metal material surface. The three-dimensional geometric model contains micro-scale surface flaw information;

[0006] The three-dimensional geometric model is regionally divided, and the surface flaw region is identified and extracted. The surface flaw region is subjected to edge detection to determine the accurate geometric profile and spatial position information of the flaw;

[0007] According to the geometric profile and spatial position information of the flaw, a parameterized flaw model is established, including the key size parameters of the depth, width, and length of the flaw. A mathematical description model of the flaw is obtained by a least squares fitting method;

[0008] The three-dimensional geometric model and the mathematical description model of the metal material are imported into the finite element analysis software, material properties, boundary conditions and load conditions are set, and the finite element analysis model containing micro defects is established;

[0009] The stress-strain analysis is performed on the finite element analysis model of the micro defect, the stress distribution and the stress concentration coefficient around the defect are calculated, the fracture mechanics theory is used to evaluate the influence of the defect on the fatigue crack initiation and propagation;

[0010] According to the evaluation result, based on the Paris formula and the cumulative damage theory, the stress analysis result and the material fatigue parameters are combined to establish a fatigue life prediction model considering the influence of the defect to calculate the crack propagation life, and surface defect analysis for metal material fatigue life prediction is realized.

[0011] The technical scheme provided by the embodiment of the present application can include the following beneficial effects:

[0012] The present application discloses a kind of surface defect analysis method in combination with metal material fatigue life prediction.The method obtains metal surface topography point cloud data by three-dimensional laser scanning, constructs three-dimensional geometric model containing micro defect information.Model is divided into regions and edge detection is carried out, the accurate geometric profile and spatial position information of defect are extracted, and parameterized defect model is established.Geometric model and defect model are imported into finite element analysis software, stress-strain analysis is carried out, and the influence of defect on fatigue crack is evaluated.Based on the analysis result, Paris formula and cumulative damage theory are combined to establish fatigue life prediction model considering the influence of defect.The present application realizes the accurate characterization and quantitative analysis of metal material surface micro defect, improves the accuracy of fatigue life prediction, and provides a new technical means for safety evaluation and life prediction of metal component. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 It is a flow chart of the surface defect analysis method in combination with metal material fatigue life prediction of the present application.

[0014] Figure 2 It is a schematic diagram of the surface defect analysis method in combination with metal material fatigue life prediction of the present application.

[0015] Figure 3 It is still another schematic diagram of the surface defect analysis method in combination with metal material fatigue life prediction of the present application. DETAILED DESCRIPTION

[0016] For further understanding of the present application, the application will be described in detail with reference to the drawings and examples. The present application will be further described in detail with reference to the drawings and examples. It can be understood that the specific examples described herein are only used to explain the related application, and not to limit the application. In addition, it should be noted that, for the convenience of description, only the parts related to the application are shown in the drawings.

[0017] As Figure 1 -3, the surface flaw analysis method combined with metal material fatigue life prediction can specifically include:

[0018] S101, a three-dimensional laser scanner is used to scan the surface of the metal material to obtain surface topography point cloud data, and an initial three-dimensional model of the metal material surface is constructed by a triangular mesh reconstruction algorithm according to the surface topography point cloud data density and distribution characteristics.

[0019] A high-precision three-dimensional laser scanner is used to scan the surface of the metal material at high density to obtain high-resolution surface topography point cloud data, and the point cloud data contains three-dimensional coordinate information of the metal material surface. For the point cloud data, the effective point cloud data set is screened out, and the abnormal points and noise points are removed to obtain the optimized point cloud data set. According to the optimized point cloud data set, the principal component analysis method is used to calculate the normal vector of each point, the differential geometry method is used to calculate the curvature, the point cloud is segmented according to the curvature threshold, and the plane area, edge area and curved surface area are identified. Based on the feature segmentation result, an initial triangular mesh is constructed, the mesh quality is optimized by edge flipping and vertex insertion, the model precision is improved by using the local quadratic surface fitting method, and the surface three-dimensional model is generated. The surface three-dimensional model is dimensioned to ensure that the model is consistent with the actual metal material surface size, the surface area and volume of the model are calculated, and the standardized metal material surface three-dimensional model data is obtained.

[0020] Specifically, a high-precision three-dimensional laser scanner is used to scan the surface of the metal material at high density, appropriate scanning resolution and scanning step length are set, high-resolution surface topography point cloud data is obtained, effective point cloud data set is selected according to point cloud data density and spatial distribution characteristics, abnormal points and noise points are removed, and optimized point cloud data set is obtained. For the optimized point cloud data set, the KD tree algorithm is used for spatial indexing to improve the retrieval efficiency of the point cloud data, the principal component analysis method is used to calculate the normal vector of each point, the differential geometry method is used to calculate the curvature, and the point cloud is segmented according to the curvature threshold to identify the plane area, edge area and curved surface area. Based on the point cloud data, an initial triangular mesh is constructed, the mesh quality is optimized by edge flipping and vertex insertion, the local quadratic surface fitting method is used to improve the model accuracy, the mesh density is adjusted to balance the model accuracy and data volume, and a surface three-dimensional model with uniform mesh distribution is generated. The size of the generated surface three-dimensional model is calibrated to ensure that the model is consistent with the actual metal material surface size, the surface area and volume of the model are calculated, and the standardized metal material surface three-dimensional model data is obtained. When scanning the surface of the metal material using a high-precision three-dimensional laser scanner, the scanning resolution is set to 0.1 mm and the scanning step length is set to 0.05 mm to ensure that high-quality point cloud data is obtained. During the scanning process, the laser beam is emitted at a frequency of 50 kHz, and the reflected signal is processed to form a point cloud data set containing spatial coordinates and reflection intensity information. The original point cloud data is preprocessed, the statistical outlier filtering algorithm is used to remove abnormal points, the neighborhood radius is set to 1 mm and the standard deviation threshold is set to 2.5, and about 3% of the total data volume is removed. The KD tree algorithm is used to index the optimized point cloud data in space, and the three-dimensional space is divided into multiple cubic grids, each grid has a side length of 2 mm, and the data retrieval efficiency is improved. When calculating the normal vector of each point, a point set with a neighborhood radius of 1.5 mm is selected, and the principal component analysis method is used to solve the eigenvectors of the covariance matrix. The differential geometry method is used to calculate the curvature, the neighborhood radius is set to 1 mm, and the Gaussian curvature and mean curvature are calculated. The point cloud is segmented according to the curvature threshold of 0.01 mm^(-1) to identify the plane area, edge area and curved surface area. Based on the segmentation results, an initial triangular mesh is constructed, an improved Ball-Pivoting algorithm is used, a sphere radius of 0.5 mm is set, and an initial mesh is generated. The mesh quality is optimized by edge flipping and vertex insertion, the edge flipping angle threshold is set to 15°, and the vertex insertion interval threshold is set to 0.3 mm. The local quadratic surface fitting method is used to improve the model accuracy, the least squares fitting is performed on each triangular patch and its adjacent patches, and the fitting error threshold is set to 0.05 mm. The mesh density is adjusted to control the triangular patch side length between 0.2 mm and 1 mm, and a uniformly distributed surface three-dimensional model is generated. The generated model is calibrated in size, the feature points on the model are compared with the actual measurement values, and the error is controlled within ±0.1 mm.

[0021] S102, grid optimization and refinement processing is performed on the initial three-dimensional model, fine reconstruction is performed on the local area, and a three-dimensional geometric model of the metal material surface is obtained, and the three-dimensional geometric model contains surface defect information of a micro scale.

[0022] An initial three-dimensional model is obtained, and a grid topology, geometric features and quality indicators of the initial three-dimensional model are calculated according to a preset threshold. The quality indicators include a grid length ratio, a dihedral angle, and an area ratio. It is judged whether the quality indicators meet a preset condition. If not, the grid unit that needs to be optimized is selected. The model surface is divided according to the curvature value of the initial three-dimensional model, and a low-curvature area and a high-curvature area are obtained. The area with a curvature less than a preset first threshold is the low-curvature area, and the area with a curvature greater than a preset second threshold is the high-curvature area. A quadratic error measurement algorithm is used to simplify the grid of the low-curvature area, and the number of redundant triangular facets is reduced through edge collapse operation. The Loop subdivision algorithm is used to locally encrypt the high-curvature area and the feature boundary to obtain an optimized global grid. Fine reconstruction is performed on the identified surface defect area, and the grid density is dynamically adjusted according to the local curvature value. The adaptive grid refinement method is used to increase the local grid density, and the non-uniform rational B-spline surface fitting algorithm is used to accurately describe the defect edge to obtain a high-precision local model containing micro-scale surface defect information. The optimized global grid and the high-precision local model are fused, and the regional growth algorithm is used to realize seamless connection of the local fine model and the global model. Progressive grid density adjustment is used in the transition area to adjust the grid density and topology of the transition area, and a three-dimensional geometric model of the metal material surface is obtained, which retains the overall appearance and contains micro-defect information.

[0023] Specifically, the initial three-dimensional model is subjected to mesh quality evaluation, and the topological structure, geometric characteristics and quality indicators of the mesh are calculated, including mesh edge length ratio, dihedral angle, area ratio, and the mesh units that need to be optimized are screened according to a preset threshold value, and the mesh edge length ratio is set to be controlled within 0.5-2, and the dihedral angle is within 30°-150°. According to the evaluation results, the model surface is divided into a low-curvature region with a curvature less than 0.1 mm<->1 and a high-curvature region with a curvature greater than 1 mm<->1. A quadratic error metric algorithm is used to simplify the mesh in the low-curvature region, and the edge collapse operation is used to reduce redundant triangular facets while maintaining the overall shape of the model unchanged, and the Loop subdivision algorithm is used for local encryption for the high-curvature region and the feature boundary. The identified surface defect region is subjected to fine reconstruction, the mesh density is dynamically adjusted according to the local curvature value, the greater the curvature, the denser the mesh, the local mesh density is increased by using an adaptive mesh refinement method, and the non-uniform rational B-spline surface fitting algorithm is used to accurately describe the defect edge to generate a high-precision local model containing micro-scale surface defect information. The optimized global mesh and the fine reconstruction local model are fused, and the regional growth algorithm is used to realize seamless connection of the local fine model and the global model, and the mesh density and topological structure of the transition region are adjusted by using a gradual mesh density adjustment in the transition region, so that a three-dimensional geometric model of the metal material surface is obtained, which not only retains the overall appearance but also contains micro-defect information. When the initial three-dimensional model is subjected to mesh quality evaluation, the Hausdorff distance is used to calculate the mesh edge length ratio, and the threshold value is set to be 0.5-2; the cosine law is used to calculate the dihedral angle, and the dihedral angle is controlled within 30°-150°; the area ratio is calculated by the area ratio of adjacent triangular facets, and the area ratio is limited within 0.2-5. According to the evaluation results, the region with a curvature less than 0.1 mm<->1 is divided into a low-curvature region, and the region with a curvature greater than 1 mm<->1 is divided into a high-curvature region. The quadratic error metric algorithm is applied to simplify the mesh in the low-curvature region, and the maximum edge collapse error is set to be 0.01 mm, and the number of triangular facets is reduced by 5% in each iteration. For the high-curvature region and the feature boundary, the Loop subdivision algorithm is used for local encryption, and the subdivision threshold value is set to be 0.5 mm of edge length, and 2 iterations of subdivision are performed. When fine reconstruction is performed on the surface defect region, the adaptive mesh refinement method based on curvature is used, and when the local curvature is greater than 2 mm<->1, the mesh density is doubled. The 3rd-order non-uniform rational B-spline surface fitting algorithm is used to describe the defect edge, and the number of control points is set to be 16x16, and the fitting accuracy is 0.001 mm. When the global mesh and the local fine model are fused, the regional growth algorithm is applied, the boundary of the local model is taken as the seed point, and 5-10 layers of triangular facets are expanded outward, and the mesh density is adjusted in the transition region by using a linear interpolation method, so that the mesh edge length gradually transitions from 0.05 mm of the local model to 0.5 mm of the global model.

[0024] Based on the obtained initial three-dimensional model, grid quality analysis is performed on the three-dimensional model, and the triangular facets are optimized according to the obtained grid quality evaluation results.

[0025] Grid data of the three-dimensional model is obtained, for each triangular facet in the grid data, the edge length ratio, dihedral angle, and area ratio of the triangular facet are calculated to obtain geometric characteristic parameters of the triangular facet; according to the geometric characteristic parameters, the Gaussian curvature and the mean curvature of each vertex are calculated by using a discrete differential geometry method to obtain a grid quality evaluation index; a comprehensive quality index is calculated in combination with the edge length ratio, the dihedral angle, and the area ratio, and if the comprehensive quality index is less than a preset multi-level threshold value, it is determined that it is a low-quality grid unit; for the low-quality grid unit, a Laplace smoothing algorithm is used for local optimization to improve the shape of the triangular facet by adjusting the position of the vertex; the grid quality evaluation index is recalculated for the three-dimensional model after the local optimization, and if the grid quality evaluation index is greater than a preset threshold value, it is determined that the local optimization is effective; a grid quality analysis report including a grid quality distribution histogram, a quality improvement percentage, and a comparison chart before and after local optimization is generated.

[0026] Specifically, according to the mesh data of the initial three-dimensional model, the geometric feature parameters of each triangular facet are calculated, including the edge length ratio, dihedral angle, area ratio, and the mesh density is counted, the Gaussian curvature and mean curvature of each vertex are calculated by using the discrete differential geometry method, and the mesh quality evaluation index is obtained. The comprehensive quality index is calculated combined with the edge length ratio, dihedral angle, and area ratio, and a multi-level threshold is set to grade the mesh quality. The triangular facets with an edge length ratio less than 0.2 or greater than 5, a dihedral angle less than 30 degrees or greater than 150 degrees, and an area ratio less than 0.1 or greater than 10 are marked as low-quality mesh elements. The Laplace smoothing algorithm is used for local optimization of the low-quality mesh elements, the shape of the triangular facet is improved by adjusting the position of the vertex, the boundary constraint and volume preservation condition are set to ensure that the overall shape of the model is not changed during the optimization process, and the iterative optimization is performed until the mesh quality index meets the preset threshold, and the optimization result is preliminarily evaluated. The mesh quality evaluation index is recalculated for the optimized three-dimensional model, compared with the initial model, the mesh optimization effect is verified, the mesh quality analysis report including the mesh quality distribution histogram, quality improvement percentage, and comparison chart before and after local optimization is generated, and the data includes the mesh quality statistical data and distribution chart before and after optimization. Based on the mesh data of the initial three-dimensional model, the geometric feature analysis is performed on each triangular facet, the ratio of the longest side to the shortest side, the angle between adjacent facets, and the area ratio of adjacent facets are calculated. The k-nearest neighbor algorithm is used to count the mesh density, k=6 is taken, and the average distance of the 6 adjacent vertices around each vertex is calculated. The curvature is calculated using the discrete differential geometry method, the 1-ring neighborhood of each vertex is taken, the local coordinate system is constructed, the quadratic surface is fitted, and the Gaussian curvature and mean curvature are solved. The comprehensive quality index Q is calculated according to the formula Q=w1ER+w2EA+w3EV, wherein ER, EA, and EV are the normalized errors of the edge length ratio, dihedral angle, and area ratio, the weights w1=0.4, w2=0.4, and w3=0.2. Three-level quality thresholds are set: Q<0.6 is low quality, 0.6≤Q<0.8 is medium quality, and Q≥0.8 is high quality. The Laplace smoothing algorithm is applied to the low-quality mesh elements, and the iterative formula is Vi'=Vi+λΣ(Vj-Vi) / n, wherein Vi is the vertex to be optimized, Vj is the adjacent vertex, n is the number of adjacent vertices, λ is the smoothing factor, and the value is 0.5. The boundary vertex position is fixed and the volume change rate is less than 0.1%. The quality is evaluated every 100 iterations during the optimization process, and the iteration is stopped when the change rate of the evaluation result is less than 1% for three consecutive times. Finally, the mesh quality analysis report is generated, including the quality distribution histogram before and after optimization, the quality improvement percentage, and the comparison chart of five typical local areas before and after optimization.

[0027] The micro-defect area of the metal material surface is identified and located, the refinement range and the subdivision level of the local defect area are determined according to the size and the distribution characteristics of the micro-defect area, the grid density is dynamically adjusted according to the refinement range and the subdivision level of the local defect area to finely reconstruct the local defect area.

[0028] The Gaussian curvature and the average curvature of each point on the metal material surface are calculated, the maximum value of the Gaussian curvature and the average curvature is taken as a local curvature value, a curvature threshold value is set according to the local curvature value, and an area exceeding the curvature threshold value is marked as a potential defect area, a connected domain analysis is performed on the potential defect area by using an 8-neighborhood seed growing algorithm to obtain a boundary and an area of the defect, geometric features of each defect are calculated according to the boundary and the area of the defect, the geometric features include an area, a perimeter, an aspect ratio and a direction, the aspect ratio and the direction of the defect are calculated by using a rectangular fitting algorithm, the defect is classified and sorted, a priority of a defect area needing fine reconstruction is determined according to a weighted sum of the defect area and the curvature value, a multi-level refinement range is set for the defect area needing fine reconstruction, the refinement range is expanded from a defect center to the outside, and the defect area is divided into a core area, a transition area and an edge area, a subdivision level of each area is determined according to the size and the complexity of the defect, adaptive grid refinement is performed on the local defect area based on the refinement range and the subdivision level, the grid density is dynamically adjusted by using an octree algorithm, the top-level octree is started, whether to continue to subdivide is judged according to the curvature value in each sub-area until a preset maximum subdivision level is reached.

[0029] Specifically, the Gaussian curvature and the mean curvature of each point on the surface of the metal material are calculated by using the discrete differential geometry method, the maximum value thereof is taken as a local curvature value, a curvature threshold value is set, the area exceeding the threshold value is marked as a potential defect area, a connected domain analysis is performed by using an 8-neighborhood seed growing algorithm, and the boundary and the area of the defect are determined. According to the identified defect area, the geometric features of each defect are calculated, including the area, the perimeter, the aspect ratio, and the direction. The rectangular fitting algorithm is used to calculate the aspect ratio and the direction of the defect, the defect is classified and sorted, the priority is sorted according to the weighted sum of the defect area and the curvature value, and the priority of the defect area needing refined reconstruction is determined. For each defect area to be reconstructed, a multi-level refinement range is set, which is expanded outward from the defect center and divided into a core area, a transition area, and an edge area. The radius of the core area is 1.2 times the equivalent diameter of the defect, the transition area is 2 times, and the edge area is 3 times. The subdivision level of each area is determined according to the size and complexity of the defect. Based on the set refinement range and subdivision level, the local area of the defect is adaptively refined by using an octree algorithm to dynamically adjust the grid density. Starting from the top octree, it is determined whether to continue to subdivide according to the curvature value in each sub-area until the preset maximum subdivision level is reached. The grid density of the core area is the highest, that of the transition area gradually decreases, and that of the edge area smoothly transitions with the surrounding grid.

[0030] S103, regionally divide the three-dimensional geometric model, identify and extract a surface defect area, perform edge detection on the surface defect area, and determine accurate geometric contour and spatial position information of the defect.

[0031] The vertex Gaussian curvature and average curvature of the grid of the three-dimensional geometric model are calculated by using a discrete Laplace-Beltrami operator to obtain vertex curvature data; the model surface is segmented by using a region growing algorithm according to the vertex curvature data, and if the difference between the curvatures of adjacent vertices is less than a preset threshold, the adjacent vertices are merged to obtain a region set with similar geometric characteristics; for the region set, a local curvature histogram and a shape index are extracted as a feature vector, and a support vector machine algorithm is used for feature extraction and classification of each region to determine whether it is a potential defect region; for the identified surface defect region, a three-dimensional Canny edge detection algorithm is used for edge extraction, and the boundary point set of the defect region is obtained by calculating the gradient of the normal vector and the curvature; for the boundary point set, a cubic B-spline interpolation method is used for curve fitting to construct an accurate geometric contour of the defect, and the spatial coordinates and geometric parameters of the contour are calculated to obtain the accurate geometric contour and spatial position information of the defect.

[0032] Specifically, the mesh curvature analysis is performed on the three-dimensional geometric model, the Gaussian curvature and mean curvature of each vertex are calculated by using the discrete Laplace-Beltrami operator, the model surface is preliminarily segmented by using a region growing algorithm according to the distribution characteristics of the curvature values, and the growing criterion is that the adjacent vertex curvature difference is less than a preset threshold value, thereby obtaining a region set with similar geometric characteristics. Based on the segmentation result, a local curvature histogram and a shape index are extracted as a feature vector, a support vector machine algorithm is used for feature extraction and classification of each region, a region with abnormal geometric characteristics is marked as a potential defect region, a final surface defect region is screened out by setting a threshold value, and the gravity center coordinates and the main direction of the defect region are calculated. For the identified surface defect region, a three-dimensional Canny edge detection algorithm is used for edge extraction, specific steps include Gaussian filtering, gradient calculation, non-maximum suppression, double threshold detection and edge connection, the boundary point set of the defect region is identified by calculating the gradient of the normal vector and the curvature, and a preliminary defect contour is formed. Curve fitting is performed on the obtained defect boundary point set, a cubic B-spline interpolation method is used to construct the accurate geometric contour of the defect, the control point interval is set to 1 / 3 of the average distance of the boundary point set, the spatial coordinates and geometric parameters of the contour are calculated, including the circumference, the area, the maximum diameter, and the parameter equation of the defect contour on the model surface is calculated, thereby obtaining the accurate geometric contour and spatial position information of the defect. In the mesh curvature analysis of the three-dimensional geometric model, the vertex curvature is calculated by using the discrete Laplace-Beltrami operator, the 1-ring neighborhood of the vertex is selected, the covariance matrix is constructed, and the principal curvature is obtained by solving the eigenvalue. The curvature threshold is set to 0.5 mm-1, the region growing algorithm is used for segmentation, and the growing criterion is that the curvature difference of adjacent vertices is less than 0.1 mm-1. The features of the segmented region are extracted, a 20-dimensional local curvature histogram and a 10-dimensional shape index are calculated, and a 30-dimensional feature vector is constructed. The support vector machine with an RBF kernel function is used for classification, the number of training samples is 1000, the optimal parameters C=10 and gamma=0.1 are selected by cross-validation. The defect region determination threshold is set to 0.8, and the gravity center coordinates and the main direction feature vector are calculated. In the edge detection, the three-dimensional Canny algorithm is used, the Gaussian filtering kernel size is 5*5*5, sigma=1.4, the gradient threshold is low=0.1 and high=0.3. The boundary point set is interpolated by using the cubic B-spline, the control point interval is 1 / 3 of the average distance, about 0.1 mm. The fitting accuracy is controlled within 0.01 mm, and the contour circumference, area and maximum diameter are calculated. Finally, the least square method is used to fit the parameter equation of the defect contour on the model surface, which is u(t)=a0+a1t+a2t^2, v(t)=b0+b1t+b2t^2, the parameter t∈[0, 1], and the geometric contour and spatial position of the defect are accurately described.

[0033] S104, according to the geometric profile and spatial position information of the flaw, a parameterized flaw model is established, including the depth, width and length key size parameters of the flaw, and a mathematical description model of the flaw is obtained by a least square fitting method.

[0034] Key feature point sets of the flaw are obtained by using a curvature extreme point detection algorithm, and the key feature point sets include contour boundary points, deepest points, widest points and end points. Principal component analysis is performed according to the key feature point sets to determine a main direction of the flaw. A local coordinate system is established with the deepest point as an origin and the main direction as a z-axis. The flaw contour is uniformly sampled by using the local coordinate system to obtain dense three-dimensional point cloud data. A surface equation of the flaw is fitted by using a weighted least square method for the three-dimensional point cloud data, and the weight is inversely proportional to the distance from the point to the boundary. Based on the surface equation, the depth, width and length of the flaw are calculated, the parameter values are optimized by a gradient descent method, and an accurate parameterized flaw model is obtained. The parameterized flaw model is registered with an original three-dimensional geometric model, accurate registration is performed by using an iterative closest point algorithm, the position and direction of the flaw in a global coordinate system are calculated, the pose of the flaw in the global coordinate system is described by using a 4x4 transformation matrix according to the position and direction, and a complete mathematical description model of the flaw is constructed in combination with the key size parameters.

[0035] Specifically, according to the geometric profile and spatial position information of the defect, the curvature extreme point detection algorithm is used to extract the key feature point set of the defect, including the contour boundary point, the deepest point, the widest point and the end point, the principal direction of the defect is determined by principal component analysis, the deepest point is taken as the origin and the principal direction is taken as the z axis, and a right-handed coordinate system is established as a local coordinate system. In the local coordinate system, the defect contour is uniformly sampled to obtain dense three-dimensional point cloud data, the weighted least squares method is used to fit the surface equation of the defect, the weight is inversely proportional to the distance from the point to the boundary, a suitable basis function such as a quadratic surface or a B-spline surface is selected, the fitting accuracy is evaluated using the root mean square error, and a preliminary mathematical description of the defect is obtained. Based on the fitted surface equation, the depth, width and length of the defect are calculated, the bounding box size of the contour is taken as the initial value, the gradient descent method is used to optimize the parameter value, the stop condition is set as that the error change is less than 0.1% for 5 consecutive iterations or the maximum iteration number is reached 100 times, the fitting error is minimized, and an accurate parameterized defect model is obtained. The parameterized defect model is registered with the original three-dimensional geometric model, the iterative closest point algorithm is used for accurate registration, the position and direction of the defect in the global coordinate system are calculated, the pose of the defect in the global coordinate system is described using a 4x4 transformation matrix, and the complete defect mathematical description model is constructed by combining the key size parameters, including the position, direction and shape parameters. In the process of parameterizing the defect model, the product of the Gaussian curvature and the average curvature is used as the curvature index, the threshold is set to 0.5 mm ^ -2, and the curvature extreme points are extracted as the key feature points. The eigenvalues and eigenvectors are calculated by principal component analysis, and the eigenvector corresponding to the maximum eigenvalue is selected as the principal direction. The deepest point is taken as the origin (0, 0, 0) and the principal direction is taken as the z axis (0, 0, 1) to establish a right-handed coordinate system. In the local coordinate system, the x-y plane is uniformly sampled at an interval of 0.1 mm to obtain dense point clouds. The weighted least squares method is used to fit the quadratic surface z = ax^2 + by^2 + cxy + dx + ey + f, and the weight w = 1 / (1+d^2), where d is the distance from the point to the boundary. The Levenberg-Marquardt algorithm is used to solve the optimal parameters, the maximum iteration number is set to 100, and the convergence threshold is set to 1e-6. The root mean square error is calculated to evaluate the fitting accuracy, and if it is greater than 0.05 mm, the sampling density is increased to 0.05 mm interval for re-fitting. Based on the fitted equation, the depth is obtained by solving the extreme points of z using numerical methods, and the length and width are obtained by the maximum distance on the x-y plane. The gradient descent method is used to optimize these parameters, the learning rate is set to 0.01, and the stop condition is set as that the error change is less than 0.1% for 5 consecutive iterations or the iteration number reaches 100. Finally, the iterative closest point algorithm is used for global registration, the maximum iteration number is set to 50, and the convergence threshold is set to 0.01 mm. A 4x4 transformation matrix T is obtained, where the 3x3 sub-matrix R represents rotation and the 3x1 vector t represents translation, and a complete defect mathematical description model is constructed.

[0036] S105, import the three-dimensional geometric model of the metal material and the mathematical description model of the flaw into the finite element analysis software, set material properties, boundary conditions and load conditions, and establish a finite element analysis model containing micro flaws.

[0037] A three-dimensional geometric model of a metal material is obtained and imported into a finite element analysis software. A feature recognition algorithm is used to identify micro-faces and edges in the three-dimensional geometric model. If the size of the identified micro-faces and edges is less than a preset threshold, the micro-faces and edges are removed to obtain an optimized model topology. According to the optimized model topology and a preset mathematical description model of the flaw, micro flaws are positioned on the three-dimensional geometric model. The CSG difference set operation is used to subtract the micro flaws from the three-dimensional geometric model to obtain a complete geometric model containing micro flaws. The complete geometric model containing micro flaws is meshed. A local mesh densification technique is used in the micro flaw area. The local mesh densification technique includes setting a minimum mesh size at the flaw boundary, increasing the mesh size by a specified multiple outwardly until reaching a global mesh size. Preset material property parameters including elastic modulus and Poisson's ratio are obtained. The material property parameters are assigned to the meshed geometric model to obtain a finite element model with material properties. For the finite element model with material properties, a fixed constraint is applied at the bottom of the model. A uniform pressure is applied at the top of the model to obtain a complete finite element analysis model containing micro flaws.

[0038] Specifically, a three-dimensional geometric model of a metal material is imported into a finite element analysis software. A feature recognition algorithm is used to identify and remove micro-faces and edges smaller than 0.1 mm, optimize the model topology, eliminate repeated boundaries, small faces and sharp corners, and improve the geometric quality of the model. According to the mathematical description model of the flaw, micro flaws are accurately positioned and constructed on the three-dimensional geometric model. The CSG difference set operation is used to subtract the flaw geometry from the original model to generate a complete geometric model containing micro flaws, and record the flaw boundary curvature and depth distribution. The geometric model containing micro flaws is meshed. A local mesh densification technique is used in the flaw area. The minimum mesh size is set to 0.01 mm at the flaw boundary, and the mesh size is increased by 1.2 times outwardly until reaching a global mesh size. The material properties, boundary conditions and load conditions are set. The elastic modulus of the material is defined as 200 GPa, and the Poisson's ratio is 0.3. These parameters are assigned to the mesh elements. A fixed constraint is applied at the bottom of the model, and a uniform pressure of 100 MPa is applied at the top. The load type and size are defined to complete the establishment of the finite element analysis model containing micro flaws. During the establishment of the finite element analysis model, first, a three-dimensional geometric model in STL format is imported. A feature recognition algorithm is used to identify micro-faces and edges. The threshold is set to 0.1 mm, and 18 micro-faces and 32 edges are removed. Then, according to the mathematical description model of the flaw

[0039] z = 0.05 * (x^2 + y^2) - 0.2, unit: mm, microflaws are constructed at the position of model surface (50, 50, 0), and the CSG difference operation is used to subtract the flaw body from the original model. The maximum curvature of the flaw boundary is recorded as 0.2 mm^(-1), and the maximum depth is 0.2 mm. When meshing, the minimum mesh size is set to 0.01 mm at the flaw boundary, and the mesh size increases by 1.2 times outwardly until the global mesh size of 0.5 mm is reached, generating about 1 million tetrahedral elements. When setting material properties, the elastic modulus is defined as 200 GPa, the Poisson's ratio is 0.3, and the density is 7850 kg / m^3. In terms of boundary conditions, fixed constraints are applied at the bottom of the model to restrict all degrees of freedom; a uniform pressure of 100 MPa is applied at the top. The load case is set to static force analysis, considering large deformation effect, and the Newton-Raphson iterative method is used to solve the nonlinear equations, with the convergence criterion set to less than 0.1% of the force residual. The final finite element analysis model contains 104,568 nodes and 573,921 elements, ready for stress analysis and fatigue life assessment.

[0040] For the nonlinear behavior of metal materials, the stress and strain data of the material are obtained, the constitutive relationship model of the material is established, and the material properties are obtained. According to the three-dimensional geometric model containing flaw information and the material properties,

[0041] Set the boundary conditions, if the complexity of the boundary conditions is higher than the threshold, use the multi-point constraint method to process, get the complete boundary conditions.

[0042] Obtain the stress and strain experimental data of metal materials in the material database, and use the Savitzky-Golay filtering algorithm to smooth the experimental data to obtain the filtered stress-strain data. According to the filtered stress-strain data, use the Levenberg-Marquardt algorithm for nonlinear regression fitting to determine the constitutive relationship model parameters of the material. Combine the constitutive relationship model with the three-dimensional geometric model containing flaw information, set the initial boundary conditions, and obtain the boundary condition description. Perform complexity evaluation on the boundary condition description, calculate the ratio of boundary node number to total node number and the diversity index of constraint type, and determine whether the complexity is higher than the preset threshold. If the complexity is higher than the preset threshold, use the multi-point constraint method to connect the complex boundary nodes with the reference nodes using RBE2 elements to obtain the equivalent simplified constraints.

[0043] Specifically, the stress and strain experimental data of the metal material are obtained from the material database, the original data are smoothed by using a Savitzky-Golay filtering algorithm, outliers and noises are removed, and a reliable stress-strain curve is obtained. Based on the processed stress-strain data, a Levenberg-Marquardt algorithm is used to perform nonlinear regression fitting on the constitutive relation model of the material, a suitable nonlinear model function such as a Ramberg-Osgood model or a Johnson-Cook model is selected, model parameters are optimized by using a genetic algorithm, and a constitutive relation expression and specific parameter values of the material are obtained. The obtained material constitutive relation model is combined with a three-dimensional geometric model containing flaw information, symmetric boundary conditions are set according to the symmetry of the model, a uniform pressure is applied to the loading surface, and a displacement constraint is set at the fixed end, so that initial boundary condition setting is completed. The complexity of the set boundary condition is evaluated, a ratio of the number of boundary nodes to the total number of nodes and a diversity index of the constraint type are calculated, if the complexity is higher than a preset threshold, a multi-point constraint method is used for processing, an RBE2 unit is used to connect the complex boundary node with a reference node, the complex boundary condition is converted into an equivalent simplified constraint, and finally a complete boundary condition description is obtained. In the implementation process, 1000 groups of stress-strain experimental data points are extracted from the material database, are smoothed by using a Savitzky-Golay filtering algorithm, a window size is set to 15, a polynomial order is 3, and after the processing, 950 groups of effective data points are obtained. The Levenberg-Marquardt algorithm is used to fit the Ramberg-Osgood model, initial parameters are set as follows: an elastic modulus E is 200 GPa, a yield strength σ y is 300 MPa, and a hardening index n is 10. The genetic algorithm is used for optimization, a population size is set to 100, and iteration is performed for 50 generations, and finally the optimized parameters are obtained as follows: E is 205 GPa, σ y is 315 MPa, and n is 8.5. The constitutive model is combined with a three-dimensional geometric model, the model size is 100 mm*50 mm*20 mm, the flaw is located at the center position, the diameter is 2 mm, and the depth is 1 mm. Symmetric boundary conditions are set in the XY plane, the bottom surface is fixed, and the top surface is subjected to a 100 MPa uniform pressure. In the boundary condition complexity evaluation, the boundary node ratio is 0.15, and the constraint type diversity index is 0.8, which is higher than the preset threshold 0.6. The RBE2 unit is used for multi-point constraint processing, one master node is selected for every 10 nodes in the complex boundary area, a total of 50 RBE2 units are set, the number of boundary nodes is reduced from 15000 to 1500, and the boundary condition description is simplified.

[0044] S106, stress and strain analysis is performed on the finite element analysis model of the micro flaw, stress distribution and stress concentration factor around the flaw are calculated, and the influence of the flaw on fatigue crack initiation and propagation is evaluated by using fracture mechanics theory.

[0045] The finite element analysis model of the micro flaw is meshed, and an adaptive mesh refinement technique based on stress gradient is adopted to refine the mesh when the stress gradient of an element exceeds a preset threshold. According to the meshing result, a stress distribution field around the flaw is calculated by a Newton-Raphson iterative solver. From the stress distribution field, the stress values of the flaw tip and the surrounding area are extracted, and a stress concentration factor K_t is calculated, where K_t is equal to σ_max divided by σ_nom, σ_max is the maximum local stress, and σ_nom is the nominal stress. The stress intensity factor of the flaw tip is calculated using the J-integral method, and if the stress intensity factor exceeds the crack propagation threshold ΔK_th of the material, the crack propagation rate is calculated using the Paris formula. Based on the stress intensity factor and the crack propagation rate, combined with the fracture toughness and fatigue limit of the material, the fatigue crack initiation life is predicted using the Morrow equation. According to the fatigue crack initiation life, combined with the S-N curve and the Miners linear cumulative damage theory, the crack propagation life is calculated, and the evaluation result of the influence of the flaw on the fatigue performance of the component is obtained.

[0046] Specifically, the finite element analysis model of the micro flaw is meshed and solved, the adaptive mesh refinement technique based on stress gradient is adopted, the mesh is refined when the element stress gradient exceeds the threshold value, the mesh is encrypted around the flaw, and the stress distribution field around the flaw is calculated by the Newton-Raphson iterative solver. According to the calculated stress distribution field, the stress values of the flaw tip and the surrounding area are extracted, the stress concentration coefficient K_t = σ_max / σ_nom is calculated, where σ_max is the maximum local stress, σ_nom is the nominal stress, and the maximum principal stress criterion or Von Mises stress criterion is used to determine the stress concentration position. The stress intensity factor of the flaw tip is calculated by the J integral method, and the maximum circumferential stress criterion is used to determine the crack propagation direction. When the stress intensity factor exceeds the material crack propagation threshold ΔK_th, it is considered that the crack starts to propagate, and the Paris formula is used to calculate the crack propagation rate. Based on the calculated stress intensity factor and crack propagation rate, combined with the material fracture toughness and fatigue limit, the Morrow equation is used to predict the fatigue crack initiation life, combined with the S-N curve and Miners linear cumulative damage theory to calculate the crack propagation life, and the influence of the flaw on the fatigue performance of the component is evaluated. In the finite element analysis of the micro flaw, the model is first meshed, and the initial mesh size is set to 0.5 mm. The adaptive refinement based on stress gradient is adopted, and the element stress gradient is refined when it exceeds 200 MPa / mm, and the minimum mesh size is limited to 0.01 mm. The Newton-Raphson iterative solver is used, and the convergence threshold is set to 0.1%, and the maximum iteration number is 100. After calculating the stress distribution around the flaw, the stress values within 50 μm of the flaw tip are extracted, the stress concentration coefficient K_t is calculated, the maximum stress at the flaw is 450 MPa, the far-field nominal stress is 150 MPa, and K_t = 3 is obtained. The Von Mises stress criterion is used to determine the stress concentration position. The virtual displacement method is used to calculate the J integral, and the integral path is selected as three concentric circular paths around the flaw tip with radii of 0.05 mm, 0.1 mm and 0.15 mm. The stress intensity factor KI = 25 MPa√m is calculated, which exceeds the material crack propagation threshold ΔK_th = 15 MPa√m, and it is determined that the crack starts to propagate. The Paris formula da / dN = C(ΔK)^m is used to calculate the crack propagation rate, where C = 1e-11 and m = 3. Based on the Morrow equation and the material S-N curve, the fatigue crack initiation life is predicted to be 5e5 times. Combined with the Miners linear cumulative damage theory, the total fatigue life is calculated to be 7.2e5 times, which is 40% lower than the life of the flaw-free component.

[0047] S107, according to the evaluation result, based on Paris formula and cumulative damage theory, combining stress analysis result and material fatigue parameter, a fatigue life prediction model combining flaw influence is established to calculate crack propagation life, and surface flaw analysis of metal material fatigue life prediction is realized.

[0048] According to the stress distribution data around the flaw, the J integral value is extracted, the stress intensity factor K is obtained through the plane strain relationship, if the J integral value is greater than a preset threshold, the flaw tip stress intensity factor range ΔK is calculated, the material Paris formula parameters C and m are adopted to establish a crack propagation rate model, the adaptive Simpson integral method is used to calculate the crack propagation life from the initial length a0 to the critical length ac, the propagation life is determined by the crack propagation rate model, and when the stress intensity factor range ΔK is less than the material crack propagation threshold ΔKth, the integral calculation is stopped; combined with the material S-N curve and the Miner linear cumulative damage criterion, the crack initiation life is obtained, the crack initiation life is determined by the stress amplitude and the actual cycle number; the crack initiation life and the propagation life are added to obtain the total fatigue life; a fatigue life prediction model combining surface flaw influence is established, the prediction model takes flaw size, shape factor and stress level as independent variables, and logarithmic life as dependent variable; the prediction model parameters are determined through regression analysis; and the prediction model precision is evaluated by K-fold cross validation.

[0049] Specifically, according to the stress analysis results, the stress distribution data around the flaw is extracted, the J integral value is extracted using finite element post-processing, the stress intensity factor K is converted through the plane strain relationship, the stress intensity factor range ΔK of the flaw tip is calculated, the crack propagation rate model is established combined with the Paris formula parameters C and m of the material. The adaptive Simpson integral method is used to calculate the crack propagation life from the initial length a0 to the critical length ac based on the crack propagation rate model, considering the influence of stress amplitude and stress ratio in the load spectrum on the crack propagation rate, when ΔK is less than the crack propagation threshold ΔKth of the material, the integral calculation is stopped. Using the cumulative damage theory, combined with the S-N curve and Miner linear cumulative damage criterion of the material, the stress amplitude is used to query the S-N curve to obtain the corresponding fatigue life, combined with the actual cycle number to calculate the damage amount, and the crack initiation life is obtained. The total fatigue life is obtained by adding the crack initiation life and the propagation life, the fatigue life prediction model considering the influence of surface flaws is established, a multiple nonlinear regression model is established, the independent variables include flaw size, shape factor and stress level, and the dependent variable is logarithmic life. Through regression analysis, the model parameters are determined, the prediction accuracy of the model is evaluated using K-fold cross-validation, and the quantitative prediction of the fatigue life of metal materials is realized. In the process of establishing the fatigue life prediction model, first, the stress distribution data within 50 μm around the flaw is extracted from the finite element analysis results, the J integral value is calculated using the virtual displacement method, three integral paths are selected with radii of 0.05 mm, 0.1 mm and 0.15 mm, and the average value is converted into stress intensity factor K through the plane strain relationship K = √(EJ / (1-ν^2)), wherein E is the elastic modulus 210 GPa, and v is the Poisson's ratio 0.3. The stress intensity factor range ΔK = 25 MPa√m is calculated, combined with the Paris formula parameters C = 1e-11 and m = 3 of the material, the crack propagation rate model da / dN = 1e-11(ΔK)^3 is established. The adaptive Simpson integral method is used, the initial step is set to 0.1 mm, and the error threshold is 1e-6. The life of the crack from the initial length 0.5 mm to the critical length 5 mm is calculated, considering the influence of stress ratio R = 0.1. When ΔK is less than the crack propagation threshold ΔKth = 5 MPa√m, the integral is stopped, and the propagation life is 1.5e5 times. Using the material S-N curve, the life corresponding to the stress amplitude of 350 MPa is 2e5 times, combined with the Miner criterion to calculate the damage amount 0.75, the crack initiation life is 1.5e5 times. The total fatigue life is 3e5 times. A multiple nonlinear regression model log(Nf) = β0+β1a+β2F+β3σ+β4(a*σ) is established, wherein a is the flaw size, F is the shape factor, and σ is the stress level. The Levenberg-Marquardt algorithm is used to optimize the parameters, 10-fold cross-validation is used, and the average prediction error is 15%.

[0050] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A surface defect analysis method combining fatigue life prediction of metallic materials, characterized in that, The method includes: scanning the surface of a metal material using a 3D laser scanner to acquire surface topography point cloud data; constructing an initial 3D model of the metal material surface using a triangular mesh reconstruction algorithm based on the density and distribution characteristics of the surface topography point cloud data; performing mesh optimization and refinement on the initial 3D model, and performing fine reconstruction on local areas to obtain a 3D geometric model of the metal material surface, which contains microscale surface defect information; dividing the 3D geometric model into regions, identifying and extracting surface defect regions, performing edge detection on the surface defect regions, and determining the precise geometric contour and spatial location information of the defects; and establishing a parametric defect model based on the geometric contour and spatial location information of the defects, including the depth, width, and length of the defects. By using the least squares fitting method, a mathematical description model of the defect is obtained based on the key size parameters. Using finite element analysis software, the three-dimensional geometric model of the metallic material and the mathematical description model of the defect are imported, and material properties, boundary conditions, and load conditions are set to establish a finite element analysis model including the micro-defects. Stress-strain analysis is performed on the finite element analysis model of the micro-defects to calculate the stress distribution and stress concentration factor around the defect. Fracture mechanics theory is used to evaluate the influence of the defect on fatigue crack initiation and propagation. Based on the evaluation results, and using the Paris formula and cumulative damage theory, combined with stress analysis results and material fatigue parameters, a fatigue life prediction model incorporating the influence of the defect is established to calculate crack propagation life, thus realizing the surface defect analysis for predicting the fatigue life of metallic materials.

2. The method according to claim 1, wherein, The process involves scanning the surface of a metal material using a 3D laser scanner to acquire surface topography point cloud data. Based on the density and distribution characteristics of the point cloud data, an initial 3D model of the metal material surface is constructed using a triangular mesh reconstruction algorithm. This includes: performing high-density scanning of the metal material surface using a high-precision 3D laser scanner to acquire high-resolution surface topography point cloud data, which contains the 3D coordinate information of the metal material surface; selecting effective point cloud datasets from the point cloud data, removing outliers and noise points to obtain an optimized point cloud dataset; calculating the normal vector of each point using principal component analysis and calculating the curvature using differential geometry, performing feature segmentation on the point cloud based on a curvature threshold to identify planar regions, edge regions, and curved surface regions; constructing an initial triangular mesh based on the feature segmentation results, optimizing the mesh quality through edge flipping and vertex insertion, improving model accuracy using a local quadratic surface fitting method, and generating a 3D surface model; calibrating the dimensions of the 3D surface model to ensure consistency with the actual metal material surface dimensions, calculating the surface area and volume of the model to obtain standardized 3D model data of the metal material surface.

3. The method according to claim 1, wherein, The process involves mesh optimization and refinement of the initial 3D model, followed by refined reconstruction of local areas to obtain a 3D geometric model of the metal material surface. This 3D geometric model contains microscopic surface defect information. The process includes: acquiring the initial 3D model; calculating the mesh topology, geometric features, and quality indicators of the initial 3D model based on preset thresholds, where the quality indicators include mesh side-length ratio, dihedral angle, and area ratio; determining whether the quality indicators meet preset conditions; if not, selecting mesh elements that need optimization; and dividing the model surface according to the curvature value of the initial 3D model to obtain low-curvature and high-curvature regions. Regions with curvature less than a preset first threshold are considered low-curvature regions, and regions with curvature greater than a preset second threshold are considered high-curvature regions. Curvature regions; a quadratic error metric algorithm is used to simplify the mesh in the low-curvature region, and redundant triangular faces are reduced through edge collapse operation; the high-curvature region and feature boundaries are locally refined using the Loop subdivision algorithm to obtain an optimized global mesh; fine reconstruction is performed on the identified surface defect regions, and the mesh density is dynamically adjusted according to the local curvature value; an adaptive mesh refinement method is used to increase the local mesh density, and a non-uniform rational B-spline surface fitting algorithm is combined to accurately describe the defect edges, resulting in a high-precision local model containing microscale surface defect information; the optimized global mesh is fused with the high-precision local model, and a region growing algorithm is used to achieve a seamless connection between the local refined model and the global model; in the process The transition region employs a progressive mesh density adjustment, modifying the mesh density and topology of the transition region to obtain a 3D geometric model of the metal material surface that retains both the overall morphology and microscopic defect information. This also includes: based on the acquired initial 3D model, performing mesh quality analysis on the 3D model, and optimizing triangular facets based on the obtained mesh quality evaluation results; identifying and locating microscopic defect regions on the metal material surface, determining the refinement range and subdivision level of the defect local area based on the scale and distribution characteristics of the defect region, and dynamically adjusting the mesh density based on the refinement range and subdivision level of the defect local area to perform refined reconstruction of the defect local area; the step of performing mesh quality analysis on the 3D model based on the acquired initial 3D model, and optimizing triangular facets based on the obtained mesh quality evaluation results. The mesh quality assessment results are used to optimize triangular facets, specifically including: acquiring mesh data of the 3D model; for each triangular facet in the mesh data, calculating the side length ratio, dihedral angle, and area ratio of the triangular facet to obtain the geometric feature parameters of the triangular facet; based on the geometric feature parameters, using discrete differential geometry methods to calculate the Gaussian curvature and average curvature of each vertex to obtain a mesh quality assessment index; combining the side length ratio, dihedral angle, and area ratio to calculate a comprehensive quality index; if the comprehensive quality index is less than a preset multi-level threshold, it is determined to be a low-quality mesh cell; for the low-quality mesh cell, a Laplace smoothing algorithm is used for local optimization, improving the shape of the triangular facet by adjusting the vertex position;The mesh quality evaluation index is recalculated on the locally optimized 3D model. If the mesh quality evaluation index is greater than a preset threshold, the local optimization is deemed effective. A mesh quality analysis report is generated, including a mesh quality distribution histogram, quality improvement percentage, and a comparison image before and after local optimization. The process of identifying and locating microscopic defect areas on the metal material surface involves determining the refinement range and subdivision level of the defective local area based on its scale and distribution characteristics. The mesh density is dynamically adjusted according to the refinement range and subdivision level of the defective local area to perform refined reconstruction. Specifically, this includes: calculating the Gaussian curvature and average curvature of each point on the metal material surface, taking the maximum value of the Gaussian curvature and average curvature as the local curvature value; setting a curvature threshold based on the local curvature value, and marking areas exceeding the curvature threshold as potential defective areas; and using an 8-neighborhood seed growth algorithm to further refine the potential defective areas. Connected component analysis is performed to obtain the boundaries and areas of defects. For each defect's boundary and area, its geometric features are calculated, including area, perimeter, aspect ratio, and orientation. A rectangle fitting algorithm is used to calculate the aspect ratio and orientation of the defects, classifying and sorting them. Priority is determined based on the weighted sum of the defect's area and curvature values ​​to identify defect regions requiring refined reconstruction. For these regions, multi-level refinement ranges are set, expanding outwards from the defect center to divide them into core, transition, and edge regions. The subdivision level of each region is determined based on the defect's scale and complexity. Based on the refinement ranges and subdivision levels, adaptive mesh refinement is performed on local defect regions. An octree algorithm is used to dynamically adjust the mesh density, starting from the top-level octree and determining whether to continue subdividing based on the curvature value within each sub-region, until the preset maximum subdivision level is reached.

4. The method according to claim 1, wherein, The process of dividing the 3D geometric model into regions, identifying and extracting surface defect regions, performing edge detection on these defect regions, and determining the precise geometric contours and spatial location information of the defects includes: calculating the vertex Gaussian curvature and average curvature of the 3D geometric model mesh using the discrete Laplacian-Beltrammian operator to obtain vertex curvature data; segmenting the model surface using a region growing algorithm based on the vertex curvature data, merging adjacent vertices if the curvature difference is less than a preset threshold to obtain a set of regions with similar geometric features; extracting local curvature histograms and shape indices as feature vectors for the region set, and using a support vector machine algorithm to extract features and classify each region to determine whether it is a potential defect region; extracting edges from the identified surface defect regions using a 3D Canny edge detection algorithm, obtaining the boundary point set of the defect region by calculating the gradient of the normal vector and curvature; and performing curve fitting using a cubic B-spline interpolation method for the boundary point set to construct the precise geometric contour of the defect, calculating the spatial coordinates and geometric parameters of the contour to obtain the precise geometric contour and spatial location information of the defect.

5. The method according to claim 1, wherein, The process involves establishing a parametric defect model based on the geometric contour and spatial location information of the defect, including key dimensional parameters such as depth, width, and length. A mathematical description model of the defect is obtained using a least-squares fitting method, including: acquiring a set of key feature points of the defect using a curvature extremum point detection algorithm, the set of key feature points including contour boundary points, deepest point, widest point, and endpoints; performing principal component analysis based on the key feature point set to determine the principal direction of the defect, establishing a local coordinate system with the deepest point as the origin and the principal direction as the z-axis; and uniformly sampling the defect contour using the local coordinate system to obtain dense three-dimensional point cloud data. For the 3D point cloud data, a weighted least squares method is used to fit the surface equation of the defect, where the weights are inversely proportional to the distance from the point to the boundary. Based on the surface equation, the depth, width, and length of the defect are calculated, and the parameter values ​​are optimized using gradient descent to obtain an accurate parameterized defect model. The parameterized defect model is registered with the original 3D geometric model using an iterative nearest-point algorithm for accurate registration, and the position and orientation of the defect in the global coordinate system are calculated. Based on the position and orientation, a 4x4 transformation matrix is ​​used to describe the pose of the defect in the global coordinate system, and combined with key dimension parameters, a complete mathematical description model of the defect is constructed.

6. The method according to claim 1, wherein, The process involves importing a three-dimensional geometric model and a mathematical description model of defects from a metallic material using finite element analysis software, setting material properties, boundary conditions, and load conditions, and establishing a finite element analysis model containing micro-defects. This includes: acquiring the three-dimensional geometric model of the metallic material and importing it into the finite element analysis software; using a feature recognition algorithm to identify micro-faces and edges in the three-dimensional geometric model; removing micro-faces and edges if their dimensions are smaller than a preset threshold to obtain an optimized model topology; locating micro-defects on the three-dimensional geometric model based on the optimized model topology and the preset mathematical description model of defects; and performing CSG difference operations to... The microscopic defects are subtracted from the three-dimensional geometric model to obtain a complete geometric model containing the microscopic defects; the complete geometric model containing the microscopic defects is meshed; a local mesh refinement technique is applied to the microscopic defect region, the local mesh refinement technique including: setting a minimum mesh size at the defect boundary, increasing the mesh size of each layer outward by a specified multiple until the global mesh size is reached; obtaining preset material property parameters, including elastic modulus and Poisson's ratio; assigning the material property parameters to the meshed geometric model to obtain a finite element model with material properties; applying fixed constraints to the bottom of the finite element model with material properties; and in the model... A uniform pressure is applied to the top of the model to obtain a complete finite element analysis model containing microscopic defects; it also includes: acquiring stress and strain data of the material to address its nonlinear behavior, establishing a constitutive model of the material, and obtaining material properties; setting boundary conditions based on the three-dimensional geometric model containing defect information and material properties; if the complexity of the boundary conditions exceeds a threshold, a multi-point constraint method is used to process them to obtain complete boundary conditions, specifically including: acquiring stress and strain experimental data of metallic materials from a material database, smoothing the experimental data using the Savitzky-Golay filtering algorithm to obtain filtered stress-strain data; and based on the... The filtered stress-strain data are fitted using the Levenberg-Marquardt algorithm for nonlinear regression to determine the constitutive relation model parameters of the material. The constitutive relation model is then combined with a three-dimensional geometric model containing defect information to set initial boundary conditions and obtain a boundary condition description. The complexity of the boundary condition description is evaluated by calculating the ratio of the number of boundary nodes to the total number of nodes and the diversity index of constraint types to determine whether the complexity exceeds a preset threshold. If the complexity exceeds the preset threshold, a multi-point constraint method is adopted, using RBE2 elements to connect complex boundary nodes with reference nodes to obtain equivalent simplified constraints.

7. The method according to claim 1, wherein, The stress-strain analysis of the finite element analysis model of the micro-defects is performed to calculate the stress distribution and stress concentration factor around the defect. Fracture mechanics theory is used to evaluate the impact of the defect on fatigue crack initiation and propagation. This includes: meshing the finite element analysis model of the micro-defects; employing an adaptive mesh refinement technique based on stress gradients, refining the mesh when the element stress gradient exceeds a preset threshold; calculating the stress distribution field around the defect using a Newton-Raphson iterative solver based on the meshing results; extracting the stress values ​​at the defect tip and surrounding area from the stress distribution field; and calculating the stress concentration factor K_t, where K_t equals... Divide σ_max by σ_nom, where σ_max is the maximum local stress and σ_nom is the nominal stress. Calculate the stress intensity factor at the defect tip using the J-integral method. If the stress intensity factor exceeds the material's crack propagation threshold ΔK_th, calculate the crack propagation rate using the Paris formula. Based on the stress intensity factor and crack propagation rate, combined with the material's fracture toughness and fatigue limit, predict the fatigue crack initiation life using the Morrow equation. Based on the fatigue crack initiation life, calculate the crack propagation life using the S-N curve and Miners' linear cumulative damage theory to obtain the evaluation results of the defect's impact on the component's fatigue performance.

8. The method according to claim 1, wherein, Based on the evaluation results, and using the Paris formula and cumulative damage theory, combined with stress analysis results and material fatigue parameters, a fatigue life prediction model incorporating the influence of defects is established to calculate crack propagation life. This achieves surface defect analysis for predicting the fatigue life of metallic materials, including: extracting the J integral value based on the stress distribution data around the defect, and obtaining the stress intensity factor K through plane strain relations; if the J integral value is greater than a preset threshold, calculating the stress intensity factor range ΔK at the defect tip; establishing a crack propagation rate model using the material Paris formula parameters C and m; and calculating the crack propagation life from the initial length a0 to the critical length ac using the adaptive Simpson integral method. The extended life is determined by the crack propagation rate model; the integration calculation stops when the stress intensity factor range ΔK is less than the material crack propagation threshold ΔKth; the crack initiation life is obtained by combining the material S-N curve and Miner's linear cumulative damage criterion, and the crack initiation life is determined by the stress amplitude and the actual number of cycles; the crack initiation life is added to the extended life to obtain the total fatigue life; a fatigue life prediction model incorporating the influence of surface defects is established, with defect size, shape factor, and stress level as independent variables and logarithmic life as the dependent variable; the parameters of the prediction model are determined by regression analysis; and the accuracy of the prediction model is evaluated using K-fold cross-validation.

Images