Integrated design method for three-dimensional simulation model of spare and accessory parts

Abstract

The invention relates to the technical field of computer-aided engineering, in particular to an integrated design method for a three-dimensional simulation model of spare and accessory parts. The method comprises the following steps: obtaining a part structure drawing, and carrying out part interference analysis to obtain part interference data; reconstructing an assembly path according to the part interference data to obtain assembly path data; constructing a part three-dimensional simulation model according to the assembly path data; performing fatigue simulation according to the three-dimensional simulation model of the spare part to obtain fatigue life data; performing fracture area analysis based on the fatigue life data to obtain part fracture area data; performing grain boundary cracking detection according to the part fracture area data to obtain grain boundary cracking data; and performing grain boundary hydrogen permeation detection according to the grain boundary cracking data to obtain grain boundary hydrogen permeation data. Based on the computer-aided engineering technology, the accuracy of part structure design and the reliability of service life prediction are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer-aided engineering, and particularly relates to an integrated design method for a three-dimensional simulation model of a spare part. BACKGROUND

[0002] Traditional integrated design for a three-dimensional simulation model of a spare part mainly depends on a single physical field or static structure analysis, and is difficult to comprehensively reflect the multi-factor coupling influence of the spare part in a complex service environment, resulting in a large deviation between the simulation result and the actual performance; lacks systematic integrated analysis of key microscopic damage mechanisms such as material microstructure evolution, fatigue fracture and grain boundary hydrogen permeation, and is difficult to accurately predict the service life and failure mode of the spare part; adopts preset empirical parameters and simplified assumptions, and ignores the dynamic change of material performance degradation with service time, resulting in limited design optimization effect; interference and collision detection in the assembly process are usually limited to the geometric level, lack feedback adjustment based on fatigue life and fracture region data, and are difficult to realize the collaborative optimization of the assembly path and the structure performance; the update and structure strengthening of the simulation model lack a closed-loop feedback mechanism, resulting in a long design improvement cycle and low efficiency, and it is difficult to meet the design requirements of modern high-strength and high-reliability spare parts. SUMMARY

[0003] Therefore, it is necessary to provide an integrated design method for a three-dimensional simulation model of a spare part to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, an integrated design method for a three-dimensional simulation model of a spare part comprises the following steps:

[0005] Step S1: Obtain a spare part structure drawing, and perform spare part interference analysis to obtain spare part interference data; reconstruct an assembly path according to the spare part interference data to obtain assembly path data; and construct a three-dimensional simulation model of the spare part according to the assembly path data;

[0006] Step S2: Perform fatigue simulation according to the three-dimensional simulation model of the spare part to obtain fatigue life data; perform fracture region analysis based on the fatigue life data to obtain spare part fracture region data; and perform grain boundary cracking detection according to the spare part fracture region data to obtain grain boundary cracking data;

[0007] Step S3: Perform grain boundary hydrogen permeation detection according to the grain boundary cracking data to obtain grain boundary hydrogen permeation data; perform grain boundary cavity detection based on the grain boundary hydrogen permeation data to obtain grain boundary cavity data; and perform creep fracture analysis according to the grain boundary cavity data to obtain creep fracture data;

[0008] Step S4: performing precipitate evolution analysis according to the creep rupture data to obtain precipitate evolution data; evaluating the material performance degradation degree according to the precipitate evolution data; performing component structure reinforcement according to the material performance degradation degree to obtain component structure reinforcement data; and integrating the component structure reinforcement data into the component three-dimensional simulation model to obtain a component three-dimensional optimized simulation model.

[0009] The present application unifies the feasibility of the assembly path and the rationality of the structure by interference analysis on the structural drawing and feedback to the assembly path reconstruction stage, so that the assembly path no longer depends on the geometric arrangement, but combines the actual interference behavior. Further, the assembly path is used for the construction of the three-dimensional simulation model, and the mechanical boundary conditions and spatial constraint logic are introduced in the construction process, so that the model has higher physical reality. Then, fatigue simulation is carried out on the basis of the model to obtain fatigue life data and identify the fracture area, ensuring that the analysis is not limited to the overall stress field, but can deeply identify the local failure risk of the material. On the basis of the fracture area, grain boundary cracking detection is introduced, and through image analysis and microstructure feature extraction, the crack source mechanism is accurately identified. Further, hydrogen penetration and cavity detection of the grain boundary are carried out, and hydrogen content distribution, micro-pore merging behavior, grain boundary slip and other parameters are introduced as inputs to form a multi-scale damage analysis path, and the micro-evolution mechanism is quantitatively tracked. On this basis, creep rupture and precipitate evolution simulation are introduced, and the temperature range is controlled in the range of 400℃-700℃, reflecting the organizational change rule of the material under high temperature and long time service; and based on the relationship between precipitate behavior and stress response, the material toughness and performance degradation degree are dynamically evaluated to avoid evaluation errors caused by static parameter assumptions. For the material performance degradation part, the micro-analysis results are fed back to the macro-structure adjustment by reinforcing the structure such as wall thickness, connection structure and interface reconstruction, which opens up the structure-performance two-way channel; finally, the reinforced structure parameters are integrated into the simulation model to realize the adaptive update of the three-dimensional simulation model in the design cycle, form a complete optimization closed loop, and greatly improve the design accuracy, feedback response speed and service reliability. BRIEF DESCRIPTION OF DRAWINGS

[0010] Other features, objects, and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments made with reference to the drawings:

[0011] Fig. 1 The step flowchart for the integrated design method of the component three-dimensional simulation model of the present application is shown in the figure.

[0012] Fig. 2 The detailed step flowchart of step S1 in the present application is shown in the figure.

[0013] Fig. 3 The detailed step flowchart of step S2 in the present application is shown in the figure.

[0014] The objectives, functional characteristics and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0015] The technical method of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all the other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0016] In addition, the accompanying drawings are only schematic illustrations of the present application, and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus repeated description thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities, which do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0017] It should be understood that although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element without departing from the scope of the example embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0018] To achieve the above-mentioned object, please refer to Figs. 1 to 3 The present application provides an integrated design method for a three-dimensional simulation model of spare parts, which comprises the following steps:

[0019] Step S1: Obtain a spare part structure drawing, and perform a spare part interference analysis to obtain spare part interference data; reconstruct an assembly path according to the spare part interference data to obtain assembly path data; and construct a three-dimensional simulation model of the spare parts according to the assembly path data;

[0020] In this embodiment, by acquiring the two-dimensional structural drawing of the spare part, the digital processing technology based on CAD (Computer Aided Design) software is used to convert the two-dimensional structural drawing into an accurate three-dimensional geometric entity model. Using the three-dimensional geometric model, the interference detection module in the finite element analysis (FEA) software is used to calculate the assembly relationship of the spare part point by point. The mesh division for interference detection uses triangular mesh with a size of 0.1 mm to ensure geometric accuracy, and the interference distance threshold is set to 0.01 mm to detect small contacts or overlaps. Based on the calculated interference point and area data, combined with the assembly sequence planning algorithm, the assembly path is reconstructed by the graph theory path search method to ensure that the path meets the spatial constraints and the motion degree of freedom limit. The assembly path data includes the spatial position coordinates and motion vectors of each assembly step, with a precision control within 0.05 mm. Finally, based on the reconstructed assembly path, combined with the CAD modeling software, each spare part is gradually assembled according to the assembly sequence to form a complete three-dimensional simulation model of the spare part. The three-dimensional simulation model uses a voxel resolution of not less than 0.1 mm, and the geometric accuracy is controlled within 0.2 mm by the minimum curvature radius parameter to ensure the accuracy of subsequent simulation analysis.

[0021] Step S2: fatigue simulation based on the three-dimensional simulation model of the spare part to obtain fatigue life data; fracture region analysis based on the fatigue life data to obtain spare part fracture region data; grain boundary cracking detection based on the spare part fracture region data to obtain grain boundary cracking data;

[0022] In this embodiment, the three-dimensional simulation model of the spare part is imported into a special software platform based on finite element fatigue analysis, and the load and boundary conditions under actual working conditions are applied. The load cycle is set based on the national standard GB / T 3075-2014, the cycle number range is set to 10^4 to 10^7 times, and the loading frequency is fixed at 5 Hz. The fatigue material parameters are the standard laboratory measured values, including the fatigue limit σ_e set to 300 MPa and the fracture toughness K_IC set to 45 MPa·m^0.5. The local fatigue life is calculated through the stress-life curve under multi-axial stress state, and the fatigue life data is sampled at the surface and internal key stress concentration areas of the spare part with a spatial resolution of 0.5 mm. Based on the fatigue life data, the fracture mechanics method is used to simulate the fracture propagation in the high stress area, the threshold crack length is set to 0.1 mm, and the fracture propagation path is calculated combined with the fracture propagation rate model to obtain the spatial distribution data of the spare part fracture region. Subsequently, high-resolution scanning electron microscope (SEM) image analysis combined with fracture morphology recognition technology is used to automatically identify the grain boundary cracking characteristics of the fracture region, the fracture recognition threshold is set to 0.8 (normalized), and the crack width measurement accuracy is 0.01 μm to form a grain boundary cracking data set.

[0023] Step S3: performing grain boundary hydrogen permeation detection according to the grain boundary cracking data to obtain grain boundary hydrogen permeation data; performing grain boundary cavity detection based on the grain boundary hydrogen permeation data to obtain grain boundary cavity data; performing creep rupture analysis according to the grain boundary cavity data to obtain creep rupture data;

[0024] In this embodiment, based on the grain boundary cracking data, the grain boundary hydrogen permeation detection is performed by using thermal desorption mass spectrometry (TDS) technology, the experimental temperature is set in the range of 25-600℃, and the temperature rising rate is 10℃ / min. The hydrogen content release amount is determined by the gas desorption curve, the sampling time interval is set to 1 second, and the detection sensitivity is not less than 10^-9 mol / s. According to the hydrogen content release rate curve, the hydrogen desorption rate is calculated by using the first-order kinetic model, and the threshold value is set to 0.05 mol / (m^2·s) to identify the significant hydrogen release peak. Combined with the hydrogen trap type identification method, the desorption peak is deconvoluted by using computer-aided analysis software to identify different hydrogen trap types and their concentration gradients, and the gradient threshold is set to 0.1 mol / m^3. Further based on the hydrogen concentration gradient data, the high-concentration hydrogen enrichment area is identified by using high-resolution microscopic imaging technology, and the spatial resolution is controlled within 1 μm. The grain network structure is analyzed by using electron backscatter diffraction (EBSD) technology to obtain the grain boundary connectivity data. Combined with the grain boundary connectivity data and the hydrogen enrichment area, the finite element diffusion model is applied to simulate the diffusion behavior of hydrogen atoms in the grain boundary, and the grain boundary hydrogen permeation data is output.

[0025] Step S4: performing precipitate evolution analysis according to the creep rupture data to obtain precipitate evolution data; evaluating the material performance degradation degree according to the precipitate evolution data; performing component structure reinforcement according to the material performance degradation degree to obtain component structure reinforcement data; integrating the component structure reinforcement data into the component three-dimensional simulation model to obtain a component three-dimensional optimized simulation model.

[0026] In this embodiment, according to the grain boundary cavity data, the precipitate evolution analysis is carried out by using the high temperature creep experiment data, the experimental temperature range is set to 400-700 DEG C, and the loading stress is controlled to be 0.6-0.8 times of the yield strength of the material. The transmission electron microscope (TEM) is used to quantitatively measure the precipitate morphology and size, and the precipitate particle size distribution is analyzed, and the particle size measurement accuracy is controlled to be 0.5 nm. The evolution rate of the precipitate is calculated by statistical method, and the evolution trend of the precipitate is predicted by combining the material diffusion coefficient (about 10^-13 m^2 / s) and the precipitation kinetics model. According to the precipitate distribution and morphology data, the material toughness parameters are obtained by using the nano-hardness test, and the hardness measurement error is less than 5%. Based on the material toughness data, the structural mechanics analysis software is used to calculate the performance degradation degree of the material, and the performance indicators include the elastic modulus reduction rate and the fracture toughness reduction rate, and the threshold values are set to 15% and 10% respectively. According to the material performance degradation result, the topology optimization algorithm is applied to the strengthening design of the part structure, and the optimization target is to maximize the strength and minimize the mass, and the constraint conditions include that the maximum stress does not exceed 90% of the yield strength of the material. Finally, the strengthening structure data is integrated into the three-dimensional simulation model of the part through the CAD interface, the integrity of the model geometry and mechanical parameters is ensured, and the three-dimensional optimization simulation model construction of the part is realized.

[0027] Preferably, step S1 comprises the following steps:

[0028] Step S11: performing surface meshing on the part according to the part structure drawing to obtain part mesh data;

[0029] In this embodiment, the two-dimensional or three-dimensional geometric data is imported into a special geometric meshing tool for surface meshing processing. The meshing adopts triangular elements, and the element edge length is strictly controlled between 0.05 mm and 0.1 mm to ensure fine capture of the surface geometric details of the part. The Delaunay triangulation algorithm is used to generate the mesh to ensure the mesh quality and avoid distortion. The meshing process includes boundary reservation and curvature control, and the mesh size in the area with large curvature is automatically reduced to 0.02 mm to improve the local precision. In the mesh quality standard, the minimum angle of the internal angle is limited to more than 30 degrees, and the maximum angle is not more than 120 degrees to avoid the appearance of extremely flat triangles. After the meshing is completed, the mesh vertex coordinates, element topology relationship and other data are output to form a part mesh data file in the standard STL or OBJ file format, which is convenient for subsequent data processing and analysis.

[0030] Step S12: defining the assembly relationship of the part based on the part mesh data to obtain assembly constraint data;

[0031] In this embodiment, the assembly relationship is established by constraint conditions, including three types of position constraint, direction constraint and contact constraint. The position constraint specifically refers to the normal distance of the assembly surface of two parts must be maintained between 0mm to 0.05mm; the direction constraint limits the assembly surface angle error to be less than 0.5 degrees; the contact constraint specifies that the allowable gap range is 0.01mm to 0.02mm. The relative motion freedom of each part is established by using the assembly constraint matrix, and the assembly constraint data structure is constructed. The data structure includes constraint type, applied part number, constraint parameter (such as distance, angle threshold), constraint effective area grid index and other information. By calculating the constraint compatibility check, the constraint combination that does not meet the assembly logic is screened out, and the integrity and physical reasonableness of the assembly relationship are ensured. The finally generated assembly constraint data is stored in XML format, which is convenient for subsequent automatic simulation processing.

[0032] Step S13: Perform assembly motion simulation based on the assembly constraint data to obtain the part interference data;

[0033] In this embodiment, the simulation process is realized by a multi-body dynamics simulation engine, and rigid body motion simulation is performed on each part. The motion parameter settings include that the maximum assembly speed is limited to 0.05m / s, the maximum acceleration is not more than 0.1m / s 2 , and the time step is fixed at 0.001 seconds to ensure the continuity and calculation accuracy of the motion trajectory. The simulation initial conditions use the initial position and attitude data of each part before assembly. In the simulation process, the rigid body collision detection algorithm is used to judge the contact and interference between parts in real time, and the interference judgment threshold is set to less than 0.01mm gap as interference. The simulation output includes the part position, attitude data at each time step, and the three-dimensional space coordinates and interference volume of the interference point. All motion trajectories and interference data are recorded in the time sequence data table, which is convenient for subsequent interference analysis.

[0034] Step S14: Perform collision detection according to the part interference data to obtain the part collision data;

[0035] In this embodiment, collision detection adopts an acceleration algorithm based on the Bounding Volume Hierarchies (BVH), with AABB (axis-aligned bounding box) as the basic unit, to quickly eliminate part pairs that are unlikely to collide and reduce the amount of calculation. The specific collision judgment condition is that a fine grid collision judgment is performed when the AABBs of two parts overlap and the grid spacing is less than 0.005mm. The fine judgment is based on the point-to-surface distance between the triangular meshes, and the distance threshold is 0.001mm. If the threshold is exceeded, it is judged as no collision. The collision detection results include the coordinates of the collision point, the collision grid unit index, the direction and magnitude of the collision force (calculated based on the contact stiffness model), and the contact stiffness is set to 10^7N / m. The collision data file is stored in the standard JSON format, containing collision pairs, collision timestamps and related parameters for subsequent path reconstruction.

[0036] Step S15: reconstructing the assembly path according to the parts collision data to obtain assembly path data;

[0037] In this embodiment, the path reconstruction adopts an inverse kinematics algorithm based on the collision point data, and avoids the collision point by treating it as an obstacle area for path planning. The path planning algorithm adopts RRT* (rapid random tree optimization algorithm), the path point spacing is set to 0.01m, and the search space is limited to the allowable motion range of the parts. The planning process limits the degrees of freedom of assembly movement to ensure that all assembly actions are completed within 3 translational degrees of freedom and 3 rotational degrees of freedom. The algorithm outputs a sequence of assembly path points, including the spatial coordinates, posture angles and timestamps of the path points. The path smoothing process uses cubic spline interpolation, and the maximum curvature is limited to 0.1rad / m to avoid excessive acceleration of the robotic arm. The assembly path data is saved in CSV format, including time, spatial position and posture, for use in 3D model construction.

[0038] It is particularly important that step S15 includes the following steps:

[0039] Step S151: identifying high-frequency collision areas based on the parts collision data;

[0040] In this embodiment, the existing collision data of spare parts is structured. The collision data is from previous assembly motion simulation or dynamic simulation results, and is usually stored in three-dimensional coordinate format (for example, contact point coordinates are represented in the form of.x,.y,.z), while the number of collisions of each contact point per unit time is recorded. Using three-dimensional discrete point cloud statistical analysis method, the frequency of each collision point located in the spatial voxel is counted, and the spatial partition is performed by using the voxel grid edge length of 0.5mm*0.5mm*0.5mm. On this basis, the collision frequency threshold is set to twice the global average collision frequency (for example, when the global average is 3 times, the threshold is set to 6 times), and the spatial voxel with a frequency higher than the threshold is identified as a collision high-frequency area. The processing further eliminates discrete isolated points by clustering analysis method (such as DBSCAN density clustering), and outputs the format of three-dimensional space bounding box set, each bounding box contains collision frequency, spatial center point coordinates and spatial boundary range, forming structured collision high-frequency area data.

[0041] Step S152: constructing a three-dimensional obstacle avoidance space based on the collision high-frequency area to obtain three-dimensional obstacle avoidance space data;

[0042] In this embodiment, the three-dimensional morphological dilation processing is performed on the collision high-frequency area data, and each collision area is expanded outward by a set thickness d=2mm, which is used to form the forbidden area of the assembly path. This process adopts a three-dimensional grid model processing method based on morphological dilation operation, and uses a regular cubic grid (edge length of 0.5mm) to mark the space occupation. Then, based on the shape boundary of the CAD model, a complete three-dimensional closed space structure is established, and the collision area after dilation is removed from the space, and the remaining part constitutes the obstacle avoidance passable space. In the construction process, the geometric center of the spare part, the mounting reference surface and the assembly direction limitation condition also need to be considered, and the three-dimensional obstacle avoidance space data contains the following fields: the effective boundary point set of the space passable path, the obstacle avoidance space logical Boolean graph structure, the part initial attitude and the end target position information. All data are recorded in three-dimensional Cartesian coordinate format, which is convenient for subsequent assembly path planning calling.

[0043] Step S153: defining the assembly target constraint based on the three-dimensional obstacle avoidance space data to obtain assembly target constraint data;

[0044] In this embodiment, the geometric inverse solution of the target assembly point is performed to extract its corresponding spatial coordinates and normal vector direction. Through the target constraint inversion processing, the attitude range of the spare part entering the assembly position is determined, and the attitude change function θ(t) in the assembly process is defined, where t is the path time, θ is the assembly attitude angle, and the angle change range θ∈[0°, 30°] is set to avoid sharp attitude jump in the obstacle avoidance path. Further, the assembly speed v is set to be in the range of 1-5mm / s, and the path continuity is in the range of maximum acceleration a≤20mm / s2 The limit condition is that the positioning tolerance of the parts near the assembly position is ±0.1 mm, and the initial value of the contact surface compression force is not more than 50 N. Finally, the above-mentioned angle, speed, tolerance and mechanical parameters are packaged into the assembly target constraint data structure for subsequent path calculation module calling.

[0045] Step S154: performing assembly path analysis according to the assembly target constraint data and the three-dimensional obstacle avoidance space data to obtain assembly path data.

[0046] In this embodiment, the assembly target constraint data and the three-dimensional obstacle avoidance space data are used to jointly construct a path planning solution space. In the obstacle avoidance space, the initial path planning is performed, the starting point is set as the initial position of the parts assembly, and the end point is set as the assembly constraint target attitude and position. In order to ensure the continuity of the path and the effectiveness of the obstacle avoidance, a penalty function F(x, y, z, θ) is introduced in the search process to limit the path nodes near the obstacle boundary or discontinuous corners. The iteration search radius in the path search process is set to 1 mm, and the attitude change of the path node cannot exceed 5°. After the initial path planning is completed, the B-spline path smoothing algorithm is used for interpolation processing of the path, the interpolation interval is not greater than 0.2 mm, and the fifth-order derivative is suppressed to control the curvature change and prevent the path from shaking. Finally, the output path format is the attitude-position data pair arranged in time sequence, including the three-dimensional coordinates (x, y, z) and the attitude angle (roll, pitch, yaw) of each node, which constitutes complete assembly path data.

[0047] Step S16: constructing a three-dimensional simulation model of the parts according to the assembly path data.

[0048] In this embodiment, the simulation model is constructed based on CAD software, and after the assembly path data is imported, the positions and attitudes of the parts corresponding to the path points are updated synchronously. A dynamics simulation module based on a physical engine is used to simulate the movement of the parts in the assembly process, so as to ensure that the position relationship between the parts is consistent with the path data. The geometric data of the parts in the model is imported in STEP format, and the geometric accuracy is maintained within 0.1 mm. The distance between the parts is detected in real time during the assembly process, the collision tolerance is set to 0.01 mm, and the physical rationality of the simulation movement is ensured. After the simulation of all path points is completed, a complete three-dimensional assembly animation and model data are generated, and the output format is FBX or GLTF, which includes the geometric shape, movement trajectory and constraint information of the parts, and meets the subsequent structure analysis and dynamic simulation requirements.

[0049] Preferably, step S2 comprises the following steps:

[0050] Step S21: performing fatigue simulation according to the three-dimensional simulation model of the parts to obtain fatigue life data;

[0051] In this embodiment, the three-dimensional part model is imported into the finite element analysis software, high-quality grids are divided, and the cell size is controlled between 0.1 mm and 0.2 mm to finely capture the stress concentration area. The load conditions include static load and cyclic load, the cyclic load amplitude range is set to 5000N for the maximum load and 500N for the minimum load, and the load frequency is fixed at 10Hz. The material fatigue performance parameters are obtained by actual material test data, including S-N curve (stress amplitude-life curve) and fracture toughness. In the fatigue analysis, the fatigue life of each unit is calculated based on the stress-life theory, the damage value is accumulated by the Miner linear cumulative damage criterion, and the damage threshold is set to 1. The simulation results are output as fatigue life data of each grid unit, with a life unit of 1000 cycles, forming a fatigue life data set.

[0052] Step S22: Calculate the number of load cycles based on the fatigue life data to obtain the load cycle data;

[0053] In this embodiment, the number of load cycles is defined as the number of complete load cycles experienced by each key area of the part, and is specifically calculated by the cumulative damage model. The fatigue life data is mapped according to the grid element distribution, and the fatigue life data is converted into actual load cycles using the stress time history information. The load time sequence is collected, the time interval is set to 0.1 seconds, and the load waveform accuracy requirement is ±1%. The number of cycles for each region is counted by numerical integration method, and a structured load cycle data file is output, including grid element number and corresponding cycle number, with cycle number unit as "times".

[0054] Step S23: Extract the weld root load data from the load cycle data; and perform weld root crack detection based on the weld root load data to obtain weld root crack data;

[0055] In this embodiment, the corresponding load cycle number and stress amplitude data are extracted through the weld root grid index in the finite element model. The load threshold is set to more than 350MPa as the key load. The ultrasonic phased array detection technology is used to detect the crack at the weld root, the crack size is collected using a probe frequency of 10MHz, the spatial resolution is 0.1mm, and the crack depth detection range is 0.1mm to 5mm. Combined with the load cycle information, the crack formation position and expansion trend are analyzed, and the weld root crack data is output, including crack position coordinates, length, depth and crack morphology parameters.

[0056] Step S24: Extract the bolt hole load data from the load cycle data; and perform bolt hole deformation detection based on the bolt hole load data to obtain bolt hole deformation data;

[0057] In this embodiment, the bolt hole perimeter grid unit is locked by grid index, and the cyclic stress and load amplitude thereof are obtained. The bolt hole deformation detection adopts a three-dimensional laser scanning technology, the scanning accuracy reaches 0.01 mm, the scanning frequency is once per minute, and the deformation data in the assembly cycle is continuously collected for 10 times. The deformation threshold is set to be greater than 0.05 mm as abnormal deformation. The deformation data and the load cycle data are time-aligned and analyzed, and the bolt hole deformation data are output, including the deformation amplitude, the deformation trend and the corresponding load cycle number.

[0058] Step S25: integrating the weld root crack data and the bolt hole deformation data to obtain the abnormal data of the spare part; performing fracture prediction according to the abnormal data of the spare part to obtain the fracture data of the spare part; and calibrating the fracture area according to the fracture data of the spare part to obtain the fracture area data of the spare part;

[0059] In this embodiment, the spatial coordinates of the two types of data are registered to determine the abnormal overlap area, and a weighted superposition method is used to calculate a comprehensive abnormality index. In the abnormality index calculation, the crack length weight is set to 0.6, the deformation amplitude weight is set to 0.4, and the abnormal threshold is set to 0.5. Based on the abnormality index, the fracture mechanics theory is applied to perform fracture prediction on the key area of the spare part. The fracture prediction adopts finite element fracture propagation analysis, and the fracture propagation speed range is set to 0.01 mm / cycle to 0.1 mm / cycle. The fracture prediction result is converted into a spatial fracture risk distribution map, the fracture risk threshold is set to 70%, the high-risk area is calibrated as the fracture area, the fracture area data of the spare part is generated, and the data format is three-dimensional spatial grid weighted distribution data.

[0060] Step S26: performing grain boundary cracking detection according to the fracture area data of the spare part to obtain grain boundary cracking data.

[0061] In this embodiment, the electron backscatter diffraction (EBSD) technology is used for microstructure analysis of the fracture area, the scanning resolution is set to 50 nm, and the scanning area is determined according to the size of the fracture area, with the minimum area being 100 μm 2 . Through grain orientation analysis, the type of grain boundary and the cracking position are determined. According to the Misorientation angle of the grain boundary, the threshold is set to be greater than 15° as a high-energy grain boundary. Combined with the cracking morphology of the fracture area and the characteristics of the grain boundary, the grain boundary cracking data are output, including the cracking position, the cracking length, the grain boundary type and the related grain orientation information, and the data format is a microstructure mapping and a structured database.

[0062] Preferably, step S26 includes the following steps:

[0063] Step S261: performing stone-like fracture detection according to the fracture area data of the spare part to obtain stone-like fracture data;

[0064] In this embodiment, a high-resolution scanning electron microscope (SEM) is used to collect the three-dimensional morphology of the fracture surface. The scanning resolution is controlled at 50 nanometers, and the scanning area is set according to the size of the fracture area. The minimum collection area is not less than 100 microns × 100 microns. The surface roughness data of the fracture is collected by a surface roughness measuring instrument. The roughness parameter Ra ranges from 0.5μm to 5μm. The SEM image is segmented using an image processing algorithm to extract the fracture contour, and the large-sized grain blocks in the fracture are identified by combining grayscale gradient analysis. In the judgment standard, a stone-like fracture is defined as a grain size greater than 30μm in the fracture area, and the edge of the block presents a clear polyhedral morphology with an edge acute angle less than 60°. Based on the fracture roughness and grain size, combined with the fracture reflection intensity, the identification of the stone-like fracture is completed, and the stone-like fracture data is output. The data includes grain size distribution, three-dimensional coordinates of the fracture morphology and surface roughness parameters. The data format adopts a three-dimensional point cloud file (PLY format) and a structured database format.

[0065] Step S262: performing crystalline fracture detection based on the component fracture area data to obtain crystalline fracture data;

[0066] In this embodiment, based on the data of the fracture area of ​​spare parts, a field emission scanning electron microscope (FE-SEM) is used to capture the fracture morphology with high precision, the scanning resolution is maintained at 20 nanometers, and the scanning voltage is set to 15kV to optimize the imaging contrast. The reflection intensity value is measured by a fracture reflection intensity measuring device, the average value is taken and the spatial distribution of the fracture reflection intensity is recorded, and the reflection intensity threshold is set to 0.3 to 0.6 (normalized value range 0-1). Based on the reflection intensity distribution, the fracture details are extracted using a texture fineness analysis algorithm. The fineness index is based on the local grayscale gradient variance, and the threshold is set to be greater than 0.05. Further, the grain boundary recognition algorithm is used to determine the grain uniformity index, calculate the grain size variance, and set the grain uniformity threshold to a grain size variance of less than 100μm. 2 Combined with the above parameters, the crystalline fracture recognition is completed and the crystalline fracture data are output, including the fracture reflection intensity distribution diagram, fineness parameter matrix and grain uniformity statistics. The data format uses a multi-dimensional matrix and image file (TIFF format).

[0067] Step S263: Integrate the stone fracture data and the crystalline fracture data to obtain the crystalline fracture data of the spare part;

[0068] In this embodiment, the stone block fracture data and the crystalline fracture data are integrated. First, the spatial coordinates of the two types of fracture data are registered. The coordinate alignment is realized by using a rigid transformation algorithm based on feature points, and the transformation error is controlled within 5 nanometers. After alignment, a unified fracture morphology database is constructed, and the data structure includes fracture type identification, spatial position, size parameters, surface roughness, and reflection intensity, etc. A multi-layer classification fusion algorithm is applied to fuse the stone block and crystalline fracture features according to the spatial distribution, and the composite fracture area is divided. In the fusion rule, the spatial proximity distance threshold is set to 10 μm, and the fracture type weight is assigned to stone block 0.6 and crystalline 0.4, respectively, to ensure the integrity and consistency of the fracture data. Finally, the zero-part crystalline fracture data is generated, including spatial position coordinate set, fracture morphology label, and comprehensive feature parameter file, in the form of structured XML and point cloud data.

[0069] Step S264: According to the zero-part crystalline fracture data, the grain boundary cracking severity is evaluated to obtain the grain boundary cracking data.

[0070] In this embodiment, the morphology feature parameters of the crystalline fracture are used to construct a grain boundary cracking severity model. The input parameters of the model include grain size distribution, fracture roughness, reflection intensity, fineness, and grain uniformity. The severity quantification index is set to a percentage of 0 to 100. The index is calculated by multivariate linear weighting, in which the grain size weight is 0.35, the roughness weight is 0.25, the reflection intensity weight is 0.2, the fineness weight is 0.1, and the uniformity weight is 0.1. Based on the fracture spatial coordinates and the grain boundary network structure, combined with the fracture propagation path prediction algorithm, the spatial distribution map of the severity is determined. The area with severity exceeding 70 points is defined as a high-risk area. The output grain boundary cracking data includes severity distribution matrix, spatial coordinate mapping, and fracture risk level label. The data is stored in the geographic information system (GIS) format, which is convenient for three-dimensional visualization analysis and subsequent structure strengthening design.

[0071] Preferably, step S261 comprises:

[0072] According to the zero-part fracture area data, three-dimensional morphology acquisition is performed to obtain zero-part fracture three-dimensional morphology data;

[0073] In this embodiment, a scanning electron microscope (SEM) is used in combination with a three-dimensional laser scanner to collect high-resolution fracture surface topography. The three-dimensional laser scanning resolution is set to 0.1 microns, and the scanning range covers the entire surface of the fracture area, with a minimum scanning area of 200 microns x 200 microns. During the collection process, the laser scanner obtains the height information of the fracture by point-by-point ranging, while the SEM collects microscopic detail images of the fracture to assist in the reconstruction of the topography. The collected height point cloud data is filtered to remove noise, with the filter parameter set to Gaussian filtering and the standard deviation set to 0.05 microns to ensure the smoothness of the collected data. Finally, the point cloud data and SEM image data are combined to generate a three-dimensional fracture topography model using a three-dimensional reconstruction algorithm, with the output format being a standard three-dimensional point cloud file (such as PLY format) and the surface height matrix data, with a data accuracy of 0.1 microns.

[0074] Based on the three-dimensional fracture topography data of the spare parts, the fracture profile is identified, and the fracture profile data is obtained.

[0075] In this embodiment, an edge detection algorithm is used to extract the fracture edge from the three-dimensional topography data. The Canny operator is used, with the low threshold set to 30 and the high threshold set to 90, applied to the gradient calculation of the three-dimensional height matrix to extract the contour line. The spatial position of the fracture profile is achieved through multi-layer contour extraction technology, including the main fracture boundary and internal secondary fracture line, to ensure the integrity and continuity of the contour. The edge point set is processed through curve fitting, using spline curve fitting with a fitting error controlled within 0.05 microns to obtain a smooth and continuous fracture profile curve. The fracture profile data is saved in the form of three-dimensional coordinate point set and fitted curve parameters, with the format being XYZ point cloud and curve equation data.

[0076] Based on the fracture profile data, grain arrangement analysis is performed to obtain grain arrangement data.

[0077] In this embodiment, the grain boundary image within the fracture profile area is used for grain division, using a morphological-based image segmentation algorithm, specifically a method combining morphological gradient and region growing, with the region growing threshold set to a gray scale change greater than 15 (0-255 gray scale range) to distinguish grain boundaries. The grain arrangement direction is calculated for the segmented grain region, using Fourier transform to analyze the main direction of the grain boundary with a direction angle resolution set to 1 degree. By statistical analysis of the distribution frequency of the main direction of the grain, a grain arrangement direction histogram is generated to provide a quantitative description of the grain arrangement. The grain arrangement data includes grain position, shape parameter, and arrangement direction angle information, with the data format being structured vector data.

[0078] According to the grain arrangement data, the grain aggregation degree is determined.

[0079] In this embodiment, the grain spacing and direction consistency index in the grain arrangement data are calculated, and the neighborhood statistical method is adopted. The neighborhood radius is set to 10 microns, and the number of grains with consistent main directions and their distribution are counted. Grain aggregation is defined as the proportion of grains in the neighborhood whose main directions differ by no more than 10 degrees. A grain aggregation map is generated by calculating the spatial distribution of the aggregation value over the entire fracture area. The data is represented in the form of a two-dimensional matrix, and each matrix unit corresponds to an aggregation value within a 10 micron × 10 micron area, ranging from 0 to 1. The larger the value, the higher the degree of aggregation.

[0080] Identify coarse-grained areas based on grain aggregation;

[0081] In this embodiment, the aggregation threshold method is used, and the threshold is set to 0.8. All areas with aggregation higher than the threshold are judged as coarse grain areas. Combined with the grain area parameter, grains with an area less than 900 μm are eliminated. 2 The coarse grain region is identified accurately. A spatial connectivity analysis is performed on the coarse grain region, using an 8-neighborhood connectivity algorithm to determine the spatial extent and boundaries of the continuous coarse grain region. The coarse grain region data is output, including the 3D coordinate range and area statistics of the region, in the form of spatial polygon vector data.

[0082] Calculate the fracture roughness based on the coarse grain area;

[0083] In this embodiment, the three-dimensional topography data of the coarse grain area is selected to calculate the surface roughness parameters Ra and Rz of this area. Ra is defined as the arithmetic mean of the surface height deviation within the area. The sampling point spacing is set to 0.1 micron, and the total number of sampling points is not less than 10,000. Rz is the average value of the maximum peak-to-valley height difference within the area. The calculation takes the average value of five groups of peak-to-valley height differences. The roughness parameter calculation is completed using the calculation module of the three-dimensional topography instrument. The final fracture roughness data contains the Ra and Rz values, and the data format is a structured text file.

[0084] The stone-like fracture is detected based on the fracture roughness to obtain the stone-like fracture data.

[0085] In this embodiment, based on the threshold setting of the fracture roughness, the area where the fracture roughness Ra exceeds 2.5μm and Rz exceeds 15μm is determined to be a stone-like fracture. Combined with the three-dimensional morphological characteristics, the polyhedral structural characteristics of the fracture surface are analyzed to determine whether there are polyhedral grain blocks with a size greater than 30μm in the area. The fracture morphology texture analysis algorithm is used to extract the number, size and distribution of the surface polyhedral structure. The stone-like fracture data includes the three-dimensional spatial coordinates, grain size distribution and roughness parameters of the detected fracture area. The data is stored in the form of a combination of point cloud files and structured databases to facilitate subsequent fracture property analysis and fracture mechanism research.

[0086] Preferably, step S262 comprises:

[0087] According to the fracture area data, the fracture reflection intensity is calculated;

[0088] In this embodiment, a scanning electron microscope (SEM) is used to perform high-resolution imaging on the surface of the fracture area. The imaging acceleration voltage is set to 15 kV, and the working distance is maintained at 10 mm to ensure clear and detailed images. The gray-scale image data obtained by the SEM is used to extract the fracture surface reflection intensity information. The reflection intensity is quantified by pixel gray value, with a gray value range of 0 to 255, and a higher value indicating stronger reflection. Median filtering is used for image preprocessing, with a filter window size of 3x3 pixels. After removing noise, the gray scale of all pixels in the fracture area is counted, and the mean and standard deviation of the fracture reflection intensity are calculated. The threshold is set to 0.75 times the mean reflection intensity, and pixels below this threshold are determined to be weak reflection areas, providing basic data for subsequent fracture color dim area identification.

[0089] According to the fracture reflection intensity, the fracture color dim area is identified;

[0090] In this embodiment, the fracture gray-scale image is binarized, and pixel regions below the reflection intensity threshold set in step one (0.75 times the mean) are marked as color dim areas. The Otsu algorithm is used to automatically determine the threshold for binarization. The dim area connected domain in the binary image is identified by the 8-neighbor connected algorithm, and isolated small areas with an area less than 50 μm 2 are removed to exclude image noise. The boundary of the identified color dim area is obtained by the contour extraction algorithm, and the findContours function in OpenCV is used with a precision error controlled within 0.05 microns. The final fracture color dim area is saved in the form of two-dimensional coordinates and area data, providing spatial positioning for subsequent fineness detection.

[0091] Based on the fracture color dim area, the fracture fineness is detected;

[0092] In this embodiment, the SEM image in the identified color dull area is subjected to texture analysis, and the fracture texture fineness feature is calculated by using the gray level co-occurrence matrix (GLCM) method. The parameter settings include a calculation distance of 1 pixel, directions of 0°, 45°, 90° and 135°, and extraction of four texture characteristic values of contrast, energy, homogeneity and entropy. The fineness of the pixel points in the area is defined as the weighted combination of the above texture features, and the weights are contrast 0.3, energy 0.25, homogeneity 0.25 and entropy 0.2. The fineness value ranges from 0 to 1, and the larger the value, the finer the fracture texture. After the calculation is completed, the fineness distribution map is output for the color dull area, and the spatial variation of the fineness is displayed in the form of a heat map, and the data format is a two-dimensional matrix, and the matrix elements correspond to the local fineness value.

[0093] According to the fracture fineness, the grain uniformity is evaluated;

[0094] In this embodiment, the fracture fineness data is combined with the electron backscatter diffraction (EBSD) image to analyze the size and distribution uniformity of the fracture grains. The grain orientation data of the fracture area is collected by using the EBSD image with a resolution of 0.1 μm, the grain boundaries are divided by Voronoi diagram, and the area and aspect ratio of each grain are calculated. The grain uniformity is defined as the ratio of the standard deviation of the grain area to the average area, and the threshold is set to 0.3, which is lower than the value of the grain distribution. Combined with the fracture fineness heat map, the spatial corresponding relationship between the fracture fineness and the grain uniformity is analyzed by using the correlation coefficient, and the grain uniformity spatial distribution map is obtained. The grain uniformity data is saved in the form of a structured table, including the grain position, area and uniformity value.

[0095] According to the grain uniformity, the crystalline fracture is detected, and the crystalline fracture data is obtained.

[0096] In this embodiment, based on the grain uniformity spatial distribution map, the clustering algorithm is used to classify the grain uniformity values, the uniformity threshold is set to 0.3, and the fracture is divided into uniform and non-uniform areas. The uniform area is combined with the area with a fracture fineness greater than 0.7 to determine the crystalline fracture. The grain size in the crystalline fracture area is generally in the range of 5 μm to 30 μm, and the grain morphology is mainly polyhedron. The total area, grain number and average grain size of the crystalline fracture area are calculated by using the statistical method. The crystalline fracture data is saved in the form of three-dimensional spatial coordinates and attribute data, which supports subsequent fracture property analysis and material fracture mechanism research.

[0097] Preferably, the hydrogen permeation detection of the grain boundary according to the grain boundary cracking data in step S3 comprises:

[0098] According to the grain boundary cracking data, the hydrogen content release detection is performed, and the hydrogen content release data is obtained.

[0099] In this embodiment, the Thermal Desorption Spectroscopy (TDS) technique is used to detect the hydrogen release of the part sample with known grain boundary cracking area. During sample pretreatment, the size is controlled at 10 mm x 10 mm x 2 mm to ensure the inclusion of the complete grain boundary cracking area. The heating rate of the TDS test equipment is strictly set at 5 ℃ / min, from room temperature to 900 ℃, and the desorption hydrogen concentration is monitored in real time. The hydrogen release amount is detected by a mass spectrometer, the hydrogen ion signal intensity ranges from 10^-12 to 10^-6 Torr, and the hydrogen content release data is saved in the form of a corresponding curve of hydrogen release amount (unit: ppm) and temperature (unit: ℃). This data is used to characterize the hydrogen content and its release characteristics in the material.

[0100] Based on the hydrogen content release data, the desorption rate is calculated;

[0101] In this embodiment, the hydrogen content release curve obtained by TDS test is subjected to first-order differential processing to calculate the desorption rate (unit: ppm / ℃) corresponding to the temperature. By using numerical differentiation method, the local desorption rate is obtained by dividing the difference of hydrogen release amount of adjacent temperature points by the temperature difference, and the differential temperature interval is fixed at 1 ℃. The desorption rate data is used to analyze the migration dynamics of hydrogen from the inside of the material to the surface, and important parameters include peak desorption rate and peak temperature (usually between 400 ℃ and 700 ℃) to reflect the activation energy and trap strength of hydrogen release. The desorption rate data is stored as a two-dimensional array of temperature-rate, which is convenient for subsequent hydrogen trap type analysis.

[0102] According to the desorption rate, the hydrogen trap type data is obtained;

[0103] In this embodiment, the peak position and shape of the desorption rate curve are used to identify the hydrogen trap type. According to the classical hydrogen diffusion theory, different hydrogen traps correspond to different desorption peak temperature zones: weak traps correspond to a peak temperature of about 200 ℃-400 ℃, medium traps correspond to a peak temperature of 400 ℃-600 ℃, and strong traps correspond to a peak temperature of 600 ℃-800 ℃. By using the peak fitting method, the desorption rate curve is decomposed into multiple Gaussian peaks, and the fitting residual is set to be less than 0.01 during the fitting process to ensure the fitting accuracy. According to the peak area and peak position, the number and concentration of different hydrogen traps are determined to obtain the hydrogen trap type data, which is in the format of trap type (weak, medium, strong) and its corresponding concentration (unit: ppm).

[0104] According to the hydrogen trap type data, the concentration gradient is calculated to obtain the concentration gradient data;

[0105] In this embodiment, based on the hydrogen trap concentration and spatial distribution, the hydrogen content at the grain boundary and the adjacent region is grid divided, and the grid size is set to 1 μm x 1 μm x 1 μm. The spatial gradient of hydrogen concentration is calculated by using the finite difference method, and the unit of the gradient is ppm / μm. The gradient calculation formula is the modulus of the gradient vector. In the specific calculation, the concentration change rates of the grid nodes in the x, y and z directions are calculated respectively, and the concentration gradient is calculated by the formula The concentration gradient data is stored as a three-dimensional grid matrix, and each element corresponds to a local hydrogen concentration gradient value, which is used for subsequent high-concentration hydrogen enrichment area identification.

[0106] Based on the concentration gradient data, the high-concentration hydrogen enrichment area is identified;

[0107] In this embodiment, by setting a concentration gradient threshold, the threshold is fixed at 1.5 times the average concentration gradient, all grid elements are screened, and the area with a gradient value higher than the threshold is extracted as the high-concentration hydrogen enrichment area. Connected component analysis is used to connect adjacent high-concentration points in space using a 6-neighborhood algorithm, and isolated areas with a volume less than 10 μm 3 The extracted hydrogen enrichment area is saved in the form of three-dimensional volume data, including spatial coordinates, volume size and local average concentration gradient, which provides a basis for area positioning for grain boundary hydrogen permeation analysis.

[0108] According to the grain boundary cracking data, the grain network structure is identified, and the grain boundary connectivity data is obtained;

[0109] In this embodiment, the electron backscattering diffraction (EBSD) data of the fracture area is used to collect grain orientation information, and the image resolution is set to 0.1 μm. Voronoi segmentation method is used to divide the EBSD image and calculate the grain boundary coordinates. The connectivity algorithm in graph theory is used to construct the grain adjacency matrix, and the nodes represent the grains and the edges represent the grain boundary connection relationship. The grain boundary connectivity is quantified by two indicators of clustering coefficient and average path length. The connectivity data is saved in the form of adjacency matrix and grain boundary coordinate set, which lays a foundation for grain boundary hydrogen permeation path analysis.

[0110] According to the grain boundary connectivity data and the high-concentration hydrogen enrichment area, the grain boundary hydrogen permeation detection is performed, and the grain boundary hydrogen permeation data is obtained.

[0111] In this embodiment, the three-dimensional coordinates of the grain boundary connectivity data and the high-concentration hydrogen enrichment area are analyzed by spatial overlap, and the hydrogen enrichment volume on the grain boundary network path is screened. The hydrogen diffusion path along the grain boundary is calculated by fluid mechanics simulation, the simulation temperature is set to 300 K, and the diffusion coefficient is taken as 1 x 10^-11 m 2The hydrogen concentration distribution and the grain boundary connectivity are combined to calculate the hydrogen permeability and the cumulative concentration in the grain boundary. The hydrogen permeation data are saved in the form of a hydrogen concentration space-time distribution matrix in the grain boundary path, which supports subsequent hydrogen embrittlement and failure analysis of the material.

[0112] Preferably, the grain boundary cavity detection based on the grain boundary hydrogen permeation data in step S3 comprises:

[0113] Grain boundary micro-pore detection is performed based on the grain boundary hydrogen permeation data to obtain grain boundary micro-pore data.

[0114] In this embodiment, a scanning electron microscope (SEM) is used in combination with a three-dimensional reconstruction technique to perform high-resolution imaging of the surface and fracture surface of the part sample containing the grain boundary hydrogen permeation region. The image resolution is not less than 10 nm. Image processing software is used to perform binary processing on the SEM image, with a threshold value of 90 (0-255 scale) for gray value, and the hole outline features are extracted. The hole size is counted, and micro-pores with a pore size range of 50 nm to 500 nm are screened, and their spatial distribution coordinates and morphological features are recorded. The micro-pore data are saved in the form of a pore size, position coordinate, porosity, and pore volume distribution matrix as input parameters for subsequent high-temperature simulation.

[0115] High-temperature simulation is performed according to the grain boundary micro-pore data to obtain grain boundary high-temperature data.

[0116] In this embodiment, a finite element heat conduction analysis method is used, and an actual working condition temperature curve of the part is used, with a heating rate of 10 ℃ / min and a temperature range from room temperature 25 ℃ to 800 ℃. During the simulation, the spatial position and size of the grain boundary micro-pores are mapped to the thermal field model, and corresponding thermal physical parameters are assigned, such as thermal conductivity 0.5 W / (m·K), specific heat capacity 500 J / (kg·K), and density 7800 kg / m 3 The high-temperature simulation uses an implicit integration method with a time step of 1 second to calculate the temperature field distribution, focusing on obtaining the temperature gradient, thermal stress distribution, and their evolution over time in the micro-pore region. The simulation results are recorded as temperature-time-space three-dimensional data, referred to as grain boundary high-temperature data.

[0117] Grain boundary slip analysis is performed according to the grain boundary high-temperature data to obtain grain boundary slip data.

[0118] In this embodiment, the atomic-scale slip behavior of the grain boundary region is analyzed using molecular dynamics simulation methods. The input parameters include grain boundary structure parameters, thermal stress values (based on high-temperature simulation results, the maximum thermal stress is set to be within the range of 300 MPa), and temperature field data. The simulation duration is set to 100 ns, the time step is 1 fs, and the dislocation movement in the grain boundary surface, the atomic dislocation density, and the dislocation expansion in the grain boundary surface are recorded. By calculating the displacement vector field, the relative slip amount and slip rate of the grain boundary surface are extracted, and the grain boundary slip data, including the slip distance (unit: nm), slip direction, and distribution range, are obtained.

[0119] According to the grain boundary slip data, the micro-pore merging analysis is performed on the grain boundary micro-pore data to obtain micro-pore merging data;

[0120] In this embodiment, the slip analysis results are superimposed into the grain boundary micro-pore spatial distribution model, and the spatial proximity threshold is used to judge the possibility of micro-pore merging, and the proximity threshold is set to be 1.5 times the diameter of the micro-pore. The pores with a distance less than the proximity threshold are merged by a clustering algorithm, and the pore diameter is calculated by adding the volume and calculating the equivalent diameter after merging. The merging process is repeated until no further merging occurs. The micro-pore merging data, including the size distribution, spatial coordinates, and volume information of the merged micro-pores, are output to form a new pore network structure.

[0121] Based on the micro-pore merging data, load cycle simulation is performed to obtain micro-pore merging load data;

[0122] In this embodiment, a finite element structure analysis method is used to construct a mechanical model of the grain boundary region of the spare part. The load type is cyclic tensile-compressive load, the load amplitude is set to be the maximum stress of 300 MPa and the minimum stress of 0 MPa, the cycle frequency is 1 Hz, and the total cycle number is 10^5 times. The micro-pore merging data in the model are mapped to the stress concentration area corresponding to the pore geometry. The simulation uses an elastic-plastic material constitutive relationship to calculate the stress and strain distribution around the micro-pore in each load cycle, especially focusing on the strain accumulation at the pore boundary. The maximum strain and stress cycle variation curve of the micro-pore boundary are recorded to obtain the micro-pore merging load data, including the local stress and strain values corresponding to the cycle number.

[0123] Based on the micro-pore merging load data, cavity detection is performed to obtain grain boundary cavity data.

[0124] In this embodiment, by setting the strain threshold to 0.05, the regions exceeding the threshold are identified as potential cavity generation regions according to the local strain values in the micro-pore merging load data. Using three-dimensional image reconstruction technology, combined with stress and strain data, a cavity morphology and distribution model is generated to calculate the cavity volume and number. Through spatial connectivity analysis, continuous cavity clusters are selected, and cavities with a volume less than 1 μm 3The grain boundary hole data is stored in terms of hole position, volume, shape parameter and corresponding load cycle state, thereby providing a basis for subsequent fracture analysis.

[0125] Preferably, the creep fracture analysis according to the grain boundary hole data in step S3 comprises:

[0126] The hole area is calculated according to the grain boundary hole data to obtain hole area data;

[0127] In this embodiment, the grain boundary hole data is collected by using a focused ion beam scanning electron microscope (FIB-SEM) to perform layer-by-layer cross-section slicing scanning on the fracture area of the zero parts under a magnification of ≥10000x, and the single layer thickness is not more than 50 nanometers. After the fault image stack is constructed by Z-direction slicing, the voxel splicing is performed by using a three-dimensional image reconstruction algorithm. The grain boundary hole region is extracted by using a threshold segmentation method on the three-dimensional data set obtained after image reconstruction, and the gray threshold is set by the lowest point of the second main peak in the image gray histogram, and the threshold value range is set between 180-220. After the extraction region is subjected to binary processing, the three-dimensional boundary surface of the hole is generated by using the isosurface reconstruction tool, the volume is calculated by spatial voxel integration, and the area change of each hole on each section is determined by the single-axis projection area segmentation method, and finally the average cross-sectional area (unit: μm 2 ) and the spatial position of a single hole are generated to form a hole area data set.

[0128] The hole evolution trend fitting is performed based on the hole area data, wherein the fitting allowable error is set to be ≥0.95 to obtain the hole growth rate;

[0129] In this embodiment, the hole area data is obtained by repeatedly performing FIB-SEM scanning on the same observation region at different time points, and it is recommended to perform the scanning once every 100 hours, and the total observation period is set to 600 hours. After six groups of hole area sequences are obtained, the time sequence of the holes with the same position number is established. The curve is fitted by using the scipy library in Python, and the cubic spline interpolation algorithm is adopted to ensure that the fitting error is controlled within 5% at all nodes. The error evaluation standard is the determination coefficient R 2 between the fitted curve and the actual area sequence, and the value is set to be not less than 0.95. The growth rate is extracted by using the local slope approximation method, the area growth rate is calculated in each time interval, which is expressed by the growth percentage per hour, and the hole growth rate data of the region is obtained and arranged into a csv format document for export for use in the next stage.

[0130] The creep crack high-risk region is calibrated according to the hole growth rate;

[0131] In this embodiment, the cavity growth rate data is aligned and mapped with the three-dimensional simulation geometry of the spare parts, using spatial coordinate unification processing, so that the growth rate of each cavity corresponds to its actual position in the CAD model. The growth rate threshold is set to 0.015% / h (i.e. the cavity area grows more than 0.015% per hour). Based on the three-dimensional clustering algorithm (DBSCAN), the spatial distance threshold is set to 15 μm, and the minimum number of clustered cavities is 5. The cavity dense block with a growth rate higher than the threshold is extracted as a high-risk area. These areas need to be divided into grids in the structure diagram, with the minimum grid unit size being 10 μm x 10 μm x 10 μm, and marked with numbers. The average growth rate, maximum growth rate and the number of cavities contained in each numbered area must be recorded to generate a complete data table of high-risk areas of creep cracks.

[0132] Based on the high-risk areas of creep cracks, material creep damage detection is performed to obtain creep damage data;

[0133] In this embodiment, a creep experiment is performed on the high-risk area sample, the experimental temperature is set to 600°C, the constant load is set to 300 MPa, and the duration is 100 hours. Laser scanning confocal microscope is used for surface micro-crack propagation monitoring, and the scanning frequency is once every 10 hours. The crack density (total crack length / observation area) in different time periods is compared, and the internal micro-crystal grain slip characteristics are combined to quantify the material state using damage variable representation method. In the damage variable calculation, the weighted function of cavity density, grain boundary slip degree and micro-crack propagation rate is used, and the damage degree classification threshold is set to: mild damage <0.3, moderate damage 0.3-0.7, and severe damage >0.7, referring to the standard material state library. All damage variable results are output in three-dimensional grid form corresponding to the number, forming a creep damage data atlas.

[0134] According to the creep damage data, fracture prediction is performed to obtain creep fracture data.

[0135] In this embodiment, the maximum size of the grid unit is set to 1 mm, the minimum size is 0.1 mm, and the loading boundary is the real load and temperature conditions set in the creep experiment. The material model uses elastic-plastic constitutive + creep damage coupled constitutive, and the strength properties are dynamically adjusted according to the distribution of damage variables in each unit. The critical fracture stress is set to the measured fracture strength of the material at 600°C, which is 80 MPa. When the equivalent stress of the grid unit exceeds this value and its damage variable is ≥0.7, it is determined to be a crack source point. The crack path is automatically generated according to the maximum direction of the local stress gradient, and the parameters such as fracture occurrence time, crack initiation position and propagation length are output to complete the acquisition of fracture prediction results.

[0136] Preferably, step S4 comprises the following steps:

[0137] Step S41: Precipitate evolution analysis is performed according to the creep rupture data, wherein the temperature is set to 400-700 DEG C, and precipitate evolution data is obtained;

[0138] In this embodiment, the micro material region involved in the creep rupture data is subjected to heat treatment simulation, the temperature interval is set to 400-700 DEG C, the temperature step is set to 50 DEG C, and the holding time at each temperature is 50 hours. At each temperature point, the morphology, number density and size distribution of intracrystalline precipitated phase are observed by transmission electron microscopy (TEM). Quantitative statistics are performed by image processing tools, the average particle size of the precipitated phase is measured by Feret diameter measurement method, the number density is expressed by the number of unit volume μm -3 The classical Kampmann-Wagner numerical model (KWN model) is introduced to simulate the evolution process of the precipitates over time, the initial dislocation density (10 13 m -2 ), diffusion coefficient (temperature-dependent), solute element initial concentration (mass fraction 2.5%) and other parameters are input, the precipitate size evolution curve is simulated, and the precipitate volume fraction, particle size distribution range, average precipitation spacing and other data at each stage are output to form the precipitate evolution data.

[0139] Step S42: Material toughness detection is performed according to the precipitate evolution data, and material toughness data is obtained;

[0140] In this embodiment, according to the precipitate distribution characteristics under different heat treatment conditions in step S41, 3 groups of standard micro impact samples are prepared at each temperature, and the Charpy impact test device (impact energy 300J, the initial height of the pendulum is set to the conventional 75cm) is used to test the absorbed energy value at-20 DEG C, 25 DEG C and 100 DEG C. After testing, the absorbed energy curve, fracture morphology (observed by scanning electron microscope) and ductility index are recorded. The material toughness is quantitatively evaluated in combination with the data such as the toughness pit density (number per unit area) and depth (μm). All the measured impact toughness data are matched with the corresponding precipitate data, the influence of the precipitate particle size and volume fraction on the toughness is analyzed, and finally the material toughness data is output in the form of a numerical matrix.

[0141] Step S43: Material performance degradation degree is evaluated based on the material toughness data;

[0142] In this embodiment, the material toughness data input in step S42 and the precipitate evolution data in step S41 are input into the regression model, and a multiple linear regression method is used to establish a corresponding relationship between toughness and precipitate parameters (particle size, density, volume fraction). The material performance degradation degree is quantified by the toughness decay ratio, and the degradation degree index η is defined as: η = 1 - (E_current / E_reference), where E_current is the current material absorption energy, and E_reference is the reference absorption energy before thermal aging. If η≥0.4, it is considered that the material enters the high degradation stage; if η is between 0.2 and 0.4, it is considered to be moderately degraded; and when η<0.2, it is considered to be lowly degraded. The degradation classification results are matched with the corresponding position coordinates to output the performance degradation spatial distribution map and numerical data set.

[0143] Step S44: strengthening the structure of the parts according to the degree of material performance degradation, to obtain the structure strengthening data of the parts;

[0144] In this embodiment, according to the performance degradation distribution map in step S43, all high degradation regions with η≥0.4 are extracted, and their functional bearing roles in the structure of the parts are analyzed. If the high degradation region is located in the main bearing path, a structure reinforcement strategy is used, and the reinforcement methods include: 1) increasing the geometric thickness, and the increase is set to be 20%-30% of the original wall thickness; 2) locally replacing the material with a grain refinement alloy (average grain size ≤10 μm); 3) heat treatment strengthening, locally applying a two-phase zone annealing treatment to the region, and the treatment temperature is set to Ac1+20℃, and the temperature is maintained for 30 minutes and then quickly cooled. All the process parameters of the strengthening treatment are recorded in the "structure strengthening data table", and the spatial coordinates, treatment method, physical size change value and treatment process path of each treatment region are clearly defined.

[0145] Especially important is that step S44 includes the following steps:

[0146] Step S441: analyzing the structure degradation position according to the degree of material performance degradation, to obtain the structure degradation position data;

[0147] In this embodiment, the lower limit of fatigue strength is set to 175 MPa, and the lower limit of ductility is set to 12%. Once the strain-stress response of a region is lower than any of the indicators, it is marked as a potential degradation region. Then, by mapping these degradation points to the CAD three-dimensional structure of the parts, combining the coordinate distribution, using a volume partitioning algorithm to perform voxel grid processing on the degradation region, and using a uniform voxel unit with each side of 1 mm 3 , the local subdivision is performed, so as to form the structure degradation position data, which is output in the form of a point cloud coordinate set.

[0148] Step S442: Measure the part wall thickness based on the structural degradation position data to obtain part wall thickness data;

[0149] In this embodiment, based on the structural degradation position coordinates, the scanning sections are projected bidirectionally along the normal direction with each degradation point as the center, and the wall thickness section imaging is performed using an industrial X-ray tomography instrument (resolution 0.02 mm). The maximum and minimum boundary line segments are extracted for each section, and the actual wall thickness is calculated using a geometric distance measurement algorithm (based on the Euclidean distance formula). After all the measured wall thickness values are collected, they are stored according to the spatial position coding to form the part wall thickness data, with the unit being mm, and output in the data structure of three-dimensional coordinate points + wall thickness values, which are used for subsequent non-uniformity analysis.

[0150] Step S443: Perform part wall thickness non-uniformity detection according to the part wall thickness data to obtain part wall thickness non-uniformity data;

[0151] In this embodiment, the part wall thickness data is used to set a wall thickness fluctuation threshold σ th = ±0.3 mm as the judgment standard for wall thickness uniformity. The wall thickness difference is calculated between adjacent point pairs (distance < 5 mm), and if the wall thickness difference between any point pairs exceeds σ th, the point is marked as a wall thickness non-uniformity abnormal point. The wall thickness non-uniformity area is identified through a spatial clustering algorithm (such as DBSCAN, ε is set to 2 mm, and the minimum number of points is 10), and the corresponding spatial distribution and thickness change value are output to constitute the part wall thickness non-uniformity data set. This data is used to guide the range and direction of the connection area structure reinforcement.

[0152] Step S444: Perform connection structure reinforcement according to the part wall thickness non-uniformity data to obtain connection structure reinforcement data;

[0153] In this embodiment, a welding reinforcement band design (thickness 1.5 mm, width 1.2 times the width of the stress surface) is used to cover the area with the lowest wall thickness. Through finite element contact analysis, the connection load is set to 1.5 times the rated load, the curved surface shape and contact angle of the welding area are adjusted to ensure that the welding area is ≥ 90 mm 2 . The geometric information of the reinforced connection structure, the material reinforcement parameters (such as changing the reinforced area to ASTM A514 steel with a yield strength of 690 MPa), and the welding path coordinates are output to constitute the connection structure reinforcement data set.

[0154] Step S445: Perform part interface geometric reconstruction based on the connection structure reinforcement data to obtain part interface reconstruction data;

[0155] In this embodiment, first, the enhanced region boundary coordinates are extracted and Bezier surface fitting control points are established, with the fitting accuracy set to an error of ≤0.05 mm, forming a continuous circumscribed contour. Then, according to the interface design standard, the hole diameter change is set to not more than ±2%, the interface gap is controlled within the range of 0.1-0.3 mm, and the hole edge, buckle groove, and weld groove of the interface are reconstructed. The reconstruction process outputs the three-dimensional interface feature parameters (such as hole diameter, curvature, groove depth, etc.) and their positioning information in the part coordinate system, forming the part interface reconstruction data.

[0156] Step S446: According to the part interface reconstruction data, the part interface is reinforced to obtain part structure strengthening data.

[0157] In this embodiment, the reinforcement scheme adopts a multi-region composite reinforcement structure, that is, a metal filling layer (thickness of 0.8 mm) is superimposed at the original interface position, and additional structural ribs are introduced, with the rib length being 40% of the joint surface length and the thickness being 1.2 mm. The rib material is set to TC4 titanium alloy. The ribs are connected by spot welding, with a spot welding node arranged every 5 mm, and the welding spot diameter is 2 mm. These reinforcement structures are fused with the original interface and three-dimensional fitting calibration is performed. Finally, the CAD parameters of all geometric shapes after reinforcement, material properties, and connection mode parameters are output to form complete part structure strengthening data for integration into the three-dimensional simulation model.

[0158] Step S45: The part structure strengthening data is integrated into the three-dimensional simulation model of the part, to obtain a three-dimensional optimized simulation model of the part.

[0159] In this embodiment, the structure strengthening data generated in step S44 is mapped and integrated in the data platform for building the three-dimensional simulation model of the part (Parasolid or STEP standard format is recommended). First, the strengthening region coordinates are matched with the grid nodes in the three-dimensional model according to the part coordinate system, and the strengthening region is marked into the model using the nearest neighbor interpolation principle. Then, according to the strengthening method, the geometry of the affected area is reconstructed, and the finite element grid of the area is re-divided, with the grid element size controlled within 90% of the original element size. The material property parameters of the area after strengthening or strengthening process are updated, including Young's modulus, yield strength, fracture toughness, etc. (such as the yield strength after strengthening is improved to 540 MPa). After integration, three-dimensional stress verification analysis is performed using ABAQUS or ANSYS to verify the deformation consistency and stress distribution of the integrated model under the rated load condition, and finally the three-dimensional optimized simulation model of the part is output.

[0160] Therefore, the embodiments should be regarded, at any point, as being exemplary and not limiting, the scope of the application being defined by the appended claims and not by the above description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

[0161] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and changes will readily occur to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An integrated design method for a three-dimensional simulation model of spare parts, characterized in that: The following steps are involved: Step S1: Obtaining a component structure drawing, and performing component interference analysis to obtain component interference data; Reconstruct the assembly path based on the interference data of the parts and components to obtain the assembly path data; and construct a three-dimensional simulation model of the parts and components based on the assembly path data; Step S2: performing fatigue simulation based on the three-dimensional simulation model of the spare part to obtain fatigue life data; performing fracture area analysis based on the fatigue life data to obtain fracture area data of the spare part; performing grain boundary cracking detection based on the fracture area data of the spare part to obtain grain boundary cracking data; Step S3: performing grain boundary hydrogen permeation detection based on the grain boundary cracking data to obtain grain boundary hydrogen permeation data; performing grain boundary void detection based on the grain boundary hydrogen permeation data to obtain grain boundary void data; performing creep rupture analysis based on the grain boundary void data to obtain creep rupture data; Step S4: Perform precipitate evolution analysis based on creep rupture data to obtain precipitate evolution data; evaluate the degree of material performance degradation based on the precipitate evolution data; strengthen the component structure based on the degree of material performance degradation to obtain component structure strengthening data; integrate the component structure strengthening data into a three-dimensional simulation model of the component to obtain a three-dimensional optimized simulation model of the component.

2. The integrated design method for a three-dimensional simulation model of spare parts according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: performing mesh division on the surface of the component according to the component structure drawing to obtain the component mesh data; Step S12: defining the assembly relationship of parts based on the parts mesh data to obtain assembly constraint data; Step S13: performing assembly motion simulation based on the assembly constraint data to obtain component interference data; Step S14: performing collision detection based on the parts interference data to obtain parts collision data; Step S15: reconstructing the assembly path according to the parts collision data to obtain assembly path data; Step S16: constructing a three-dimensional simulation model of the spare part according to the assembly path data.

3. The integrated design method for a three-dimensional simulation model of a spare part according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing fatigue simulation based on the three-dimensional simulation model of the spare part to obtain fatigue life data; Step S22: Calculating the number of load cycles based on the fatigue life data to obtain load cycle data; Step S23: extracting weld root load data according to the load cycle data; performing weld root crack detection according to the weld root load data to obtain weld root crack data; Step S24: extracting bolt hole load data according to the load cycle data; performing bolt hole deformation detection according to the bolt hole load data to obtain bolt hole deformation data; Step S25: Integrate the weld root crack data and the bolt hole deformation data to obtain component abnormality data; perform fracture prediction based on the component abnormality data to obtain component fracture data; calibrate the fracture area based on the component fracture data to obtain component fracture area data; Step S26: performing grain boundary crack detection based on the component fracture area data to obtain grain boundary crack data.

4. The integrated design method for a three-dimensional simulation model of a spare part according to claim 3, characterized in that: Step S26 includes the following steps: Step S261: performing stone-like fracture detection based on the component fracture area data to obtain stone-like fracture data; Step S262: performing crystalline fracture detection based on the component fracture area data to obtain crystalline fracture data; Step S263: Integrate the stone fracture data and the crystalline fracture data to obtain the crystalline fracture data of the spare part; Step S264: evaluating the severity of grain boundary cracking based on the crystalline fracture data of the spare part to obtain grain boundary cracking data.

5. The integrated design method for a three-dimensional simulation model of spare parts according to claim 4, characterized in that: Step S261 includes: Perform three-dimensional morphology acquisition based on the fracture area data of the spare parts to obtain the three-dimensional morphology data of the fractured parts; Identify the fracture contour based on the three-dimensional morphology data of the fractured parts and obtain the fracture contour data; Perform grain arrangement analysis based on fracture profile data to obtain grain arrangement data; Determination of grain aggregation based on grain arrangement data; Identify coarse-grained areas based on grain aggregation; Calculate the fracture roughness based on the coarse grain area; The stone-like fracture is detected based on the fracture roughness to obtain the stone-like fracture data.

6. The integrated design method for a three-dimensional simulation model of a spare part according to claim 4, characterized in that: Step S262 includes: Calculate the reflective intensity of the fracture according to the fracture area data of the spare parts; Identify the dark area of ​​the fracture surface based on the reflection intensity of the fracture surface; Detect the fineness of the fracture surface based on the dark color area of ​​the fracture surface; Evaluate grain uniformity based on fracture fineness; The crystalline fracture surface is tested according to the grain uniformity to obtain the crystalline fracture surface data.

7. The integrated design method for a three-dimensional simulation model of a spare part according to claim 1, characterized in that: In step S3, performing grain boundary hydrogen permeation detection based on grain boundary cracking data includes: Perform hydrogen content release testing based on grain boundary cracking data to obtain hydrogen content release data; Calculate the desorption rate based on the hydrogen content release data; Determine the type of hydrogen trap according to the desorption rate and obtain hydrogen trap type data; Calculate the concentration gradient according to the hydrogen trap type data to obtain concentration gradient data; Identify high-concentration hydrogen-enriched areas based on concentration gradient data; Identify the grain network structure based on the grain boundary cracking data and obtain the grain boundary connectivity data; Grain boundary hydrogen permeation detection is performed based on grain boundary connectivity data and high-concentration hydrogen enrichment areas to obtain grain boundary hydrogen permeation data.

8. The integrated design method for a three-dimensional simulation model of a spare part according to claim 1, characterized in that: In step S3, performing grain boundary void detection based on grain boundary hydrogen permeation data includes: Grain boundary micropore detection is performed based on grain boundary hydrogen permeation data to obtain grain boundary micropore data; Perform high temperature simulation based on grain boundary micropore data to obtain grain boundary high temperature data; Grain boundary sliding analysis is performed based on the high temperature data of grain boundaries to obtain grain boundary sliding data; Perform micropore merging analysis on grain boundary micropore data based on grain boundary sliding data to obtain micropore merging data; Perform load cycle simulation based on micropore merged data to obtain micropore merged load data; Void detection is performed based on the micropore merged load data to obtain grain boundary void data.

9. The integrated design method for a three-dimensional simulation model of a spare part according to claim 1, characterized in that: The creep fracture analysis based on the grain boundary void data in step S3 includes: Calculate the void area based on the grain boundary void data to obtain the void area data; The void evolution trend is fitted based on the void area data, where the fitting tolerance is set to ≥ 0.95, and the void growth rate is obtained; Demarcate high-risk areas for creep cracking based on void growth rate; Conduct material creep damage detection based on high-risk areas of creep cracks to obtain creep damage data; Fracture prediction is performed based on creep damage data to obtain creep fracture data.

10. The integrated design method for a three-dimensional simulation model of a spare part according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing precipitate evolution analysis based on creep rupture data, wherein the temperature is set to 400° C.-700° C. to obtain precipitate evolution data; Step S42: performing material toughness testing based on the precipitate evolution data to obtain material toughness data; Step S43: evaluating the degree of material performance degradation based on the material toughness data; Step S44: Strengthening the component structure according to the degree of material performance degradation to obtain component structure strengthening data; Step S45: Integrate the component structure reinforcement data into the three-dimensional simulation model of the component to obtain a three-dimensional optimized simulation model of the component.

Images