Part three-dimensional model retrieval method and system based on feature extraction

Abstract

The invention relates to the technical field of information retrieval, in particular to a part three-dimensional model retrieval method and system based on feature extraction. The part three-dimensional model retrieval method comprises the following steps that a manufacturing industry part design drawing is obtained; geometric structure feature extraction and material feature extraction are carried out according to a manufacturing industry part design drawing, and geometric structure data and material data are obtained; constructing a part three-dimensional model based on the geometric structure data and the material data; performing process processing feature extraction according to a manufacturing industry part design drawing to obtain part process processing data; and performing process treatment, including surface treatment, heat treatment and welding treatment, on the part three-dimensional model according to the part process treatment data to obtain surface treatment data, heat treatment data and welding treatment data. Based on the information retrieval technology, the part three-dimensional model retrieval precision and the assembly reliability are improved, and the part design optimization rate and the production quality control efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information retrieval, and in particular to a method and system for retrieval of three-dimensional part models based on feature extraction. Background Art

[0002] Traditional part three-dimensional model retrieval methods rely solely on geometric features for modeling, ignoring multidimensional factors such as material properties and process characteristics, resulting in low retrieval accuracy and difficulty in meeting the matching requirements of comprehensive part performance in complex engineering scenarios; they fail to effectively integrate manufacturing process data such as surface treatment, heat treatment and welding, resulting in insufficient prediction capabilities for the service status of parts and an inability to assess structural risks caused by problems such as thermal stress, cracks or coating peeling; they lack quantitative analysis and feedback mechanisms for part assembly deformation behavior (such as warping), resulting in insufficient basis for subsequent design optimization and difficulty in supporting retrieval optimization for assembly process reliability; static matching strategies are usually adopted, ignoring the dynamic evolution process of parts in structure, thermal and processing stress under complex working conditions, making it difficult to accurately identify and visualize high-stress and high-risk areas; retrieval processes are often divorced from production site applications, lack scalability and intelligent feedback mechanisms, and are difficult to deeply integrate with digital manufacturing systems, restricting their promotion and application in intelligent manufacturing scenarios. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a method and system for retrieval of three-dimensional part models based on feature extraction to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a method for retrieval of a three-dimensional part model based on feature extraction includes the following steps:

[0005] Step S1: Obtain a manufacturing part design drawing; perform geometric structure feature extraction and material feature extraction based on the manufacturing part design drawing to obtain geometric structure data and material data; and construct a three-dimensional model of the part based on the geometric structure data and material data;

[0006] Step S2: Processing feature extraction is performed based on the manufacturing part design drawing to obtain part processing data; process processing is performed on the part 3D model based on the part processing data, including surface treatment, heat treatment, and welding processing, to obtain surface treatment data, heat treatment data, and welding processing data;

[0007] Step S3: performing coating peeling detection on the surface treatment data based on the heat treatment data to obtain coating peeling data; predicting high-risk peeling areas on the coating peeling data based on the welding process data to obtain high-risk peeling areas; detecting the warpage rate of the injection molded parts based on the high-risk peeling areas; identifying high-warpage parts based on the warpage rate of the injection molded parts, and performing assembly simulation of the high-warpage parts to obtain high-warpage part assembly data;

[0008] Step S4: Structural design optimization is performed based on the assembly data of high-warpage parts to obtain part structural optimization data; the part structural optimization data is uploaded to the part 3D model, and shape matching retrieval is performed to generate retrieval data.

[0009] The present invention not only obtains the geometric shape information of the parts, but also deeply understands the material properties of the parts by extracting the geometric structure features and material features of the manufacturing parts design drawings, providing comprehensive data support for subsequent design and optimization. These data help to accurately construct three-dimensional models and provide design data that is more in line with actual needs. Secondly, by extracting and applying the process characteristics of the parts (such as surface treatment, heat treatment and welding treatment), the manufacturing process can be effectively combined with the design model to obtain part performance data that is more in line with the actual application environment. This link not only strengthens the service status assessment of the parts, but also improves the ability to predict the impact of complex processes, especially when dealing with problems such as thermal stress, cracks and coating peeling, it can identify potential structural risks in advance. When considering the warping deformation behavior of the parts, by performing assembly simulation and structural optimization on high-warping parts, assembly problems can be identified and corrected in the part design stage, reducing the risk of failure due to poor assembly in the later production process. This optimization not only improves the assembly accuracy of the parts, but also enhances the reliability and performance of the parts in the actual production process. Combining 3D models with shape matching and retrieval technology enables more accurate matching and retrieval of parts in actual production environments, especially in complex engineering scenarios, leveraging multidimensional data for dynamic evolution feedback and optimization. This approach not only improves the ability to identify high-stress, high-risk areas, but also enables precise prediction and optimization of parts under complex working conditions, thus avoiding the limitations of traditional static matching strategies and enhancing the adaptability and feasibility of parts in actual production.

[0010] Preferably, this specification also provides a part three-dimensional model retrieval system based on feature extraction, which is used to execute the part three-dimensional model retrieval method based on feature extraction as described above. The part three-dimensional model retrieval system based on feature extraction includes:

[0011] The part 3D model building module is used to obtain the manufacturing part design drawings; extract the geometric structure features and material features based on the manufacturing part design drawings to obtain the geometric structure data and material data; and build the part 3D model based on the geometric structure data and material data;

[0012] The process processing module is used to extract process processing features based on the manufacturing part design drawings to obtain part process processing data; based on the part process processing data, the module performs process processing on the part 3D model, including surface treatment, heat treatment and welding treatment, to obtain surface treatment data, heat treatment data and welding data;

[0013] The assembly simulation module is used to perform coating peeling detection on surface treatment data based on heat treatment data to obtain coating peeling data; predict high-risk peeling areas on coating peeling data based on welding process data to obtain high-risk peeling areas; detect the warpage rate of injection molded parts based on the high-risk peeling areas; identify high-warpage parts based on the warpage rate of injection molded parts, and perform assembly simulation of high-warpage parts to obtain high-warpage part assembly data;

[0014] The shape matching retrieval module is used to perform structural design optimization based on the assembly data of high-warpage parts to obtain part structure optimization data; upload the part structure optimization data to the part 3D model, and perform shape matching retrieval to generate retrieval data.

[0015] The feature extraction-based three-dimensional model retrieval system of the present invention can implement any one of the feature extraction-based three-dimensional model retrieval methods of the present invention, and is used to combine the operations and signal transmission media between various modules to complete the feature extraction-based three-dimensional model retrieval method of the part. The internal modules of the system cooperate with each other to improve the accuracy of the three-dimensional model retrieval of the part and the assembly reliability, and significantly improve the component design optimization rate and production quality control efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0017] Figure 1 A schematic flow chart of the steps of a method for retrieving a three-dimensional part model based on feature extraction according to the present invention;

[0018] Figure 2 Detailed step flow diagram of step S1 in the present invention;

[0019] Figure 3 Detailed flowchart of step S13 in the present invention;

[0020] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0021] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0022] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0023] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0024] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for retrieving a three-dimensional model of a part based on feature extraction, the method comprising the following steps:

[0025] Step S1: Obtain a manufacturing part design drawing; perform geometric structure feature extraction and material feature extraction based on the manufacturing part design drawing to obtain geometric structure data and material data; and construct a three-dimensional model of the part based on the geometric structure data and material data;

[0026] In this embodiment, a part design drawing with a specific code of D-1050A is imported through the manufacturing enterprise intranet database system. The drawing is in CAD format and includes a main view, a top view, a left view, and a section view. Autodesk AutoCAD Mechanical 2023 is used with the built-in geometric feature recognition plug-in to perform boundary contour extraction. The specific extraction operation includes: executing the "Region" command to construct a two-dimensional boundary area based on the closed contours in the graphics of each perspective in the drawing, and then applying the "Boundary Detection" algorithm to identify the outer contour of the part with a length × width × height of 85.5mm × 63.2mm × 45.0mm respectively. The extracted geometric structural features include the notch depth (12.8mm), the screw hole diameter (6.5mm), the chamfer angle (45°) and the hole spacing distribution matrix. Then, the material library recognition module integrated in the PLM (product life cycle management) platform is used to retrieve the Young's modulus (210GPa), Poisson's ratio (0.3), density (7.85g / cm 3 ) and yield strength (835MPa), and other material parameters were transferred along with the extracted geometric structure data as input to PTC CreoParametric 9.0 software. Leveraging the software's 3D solid modeling capabilities, a parameter-driven 3D part solid model was rebuilt. Using a parametric modeling language, all geometric features were used as modeling constraints, and material data was bound to the solid model as physical properties, completing the 3D part model.

[0027] Step S2: Processing feature extraction is performed based on the manufacturing part design drawing to obtain part processing data; process processing is performed on the part 3D model based on the part processing data, including surface treatment, heat treatment, and welding processing, to obtain surface treatment data, heat treatment data, and welding processing data;

[0028] In this example, the constructed 3D part model was loaded using Siemens NX 2306 software, and the process treatment areas marked in the drawings were input into the process extraction module. The designated areas in the drawings were annotated with the heat treatment type "quenching and tempering treatment: heating at 860°C, oil cooling, tempering temperature at 580°C" and the surface treatment type "nickel plating, thickness 12μm." The weld location was marked in a cross-sectional view at the junction of the connecting shaft and the main body groove. The weld was a fillet weld with a length of 18mm and a leg size of 6mm. First, the heat treatment data was converted into a thermal load input. The heat treatment curve was set as follows: linear heating to 860°C (heating rate 15°C / min), holding for 45 minutes, oil cooling to 60°C, followed by tempering at 580°C for 120 minutes, and finally air cooling to room temperature. Using the thermal stress analysis module within the ANSYS Workbench platform, a heat treatment stress field was applied to the 3D part model based on the heat treatment process to obtain the residual stress distribution results after heat treatment (the maximum residual stress was found around the shaft hole, approximately 320MPa). Subsequently, in the surface treatment simulation, the surface coating parameters were defined using the “nickel plating” material layer properties in the material database, including the adhesion threshold (28 MPa), thermal expansion coefficient (13.4×10 -6 / K) and a coating thickness of 12μm. A thermo-mechanical coupling simulation was performed using ABAQUS, and the coating was applied to all exposed surface areas to generate surface treatment data. For welding, Simufact Welding software was used. TIG welding process parameters (current 110A, voltage 15V, welding speed 4mm / s) were input to establish a heat source model and perform welding simulation on the weld area. The weld residual stress field and penetration structural deformation data were extracted to complete the welding process data extraction.

[0029] Step S3: performing coating peeling detection on the surface treatment data based on the heat treatment data to obtain coating peeling data; predicting high-risk peeling areas on the coating peeling data based on the welding process data to obtain high-risk peeling areas; detecting the warpage rate of the injection molded parts based on the high-risk peeling areas; identifying high-warpage parts based on the warpage rate of the injection molded parts, and performing assembly simulation of the high-warpage parts to obtain high-warpage part assembly data;

[0030] In this embodiment, the MATLAB 2023b integrated image processing toolbox is used to perform coating peeling detection on the generated surface treatment data and heat treatment residual stress distribution results. First, the grayscale value method is used to partition the stress value of the model surface after heat treatment, and the stress gradient threshold is selected as 70MPa / mm. If the stress change rate between adjacent grids exceeds the threshold, it is marked as a stress mutation area; then it is compared with the nickel plating adhesion strength of 28MPa to determine whether it is lower than the coating adhesion lower limit. If so, it is marked as a potential peeling area; finally, a coating peeling data layout is formed. On this basis, the high-risk peeling area prediction operation calls the welding treatment residual stress results, extracts the maximum shear stress area within 10mm around the weld, and performs Boolean overlap judgment by setting the shear stress threshold of 48MPa. If the shear stress in the welding area is greater than this value and the overlapping area with the potential peeling area is greater than 25mm 2 , then the overlapping area is determined to be a high-risk area for peeling. The high-risk area is imported into the SolidWorks Plastics module to simulate the deformation behavior of the injection molding process. The injection molding material is set to PA66, the melting temperature is 265℃, the mold temperature is 85℃, and the filling time is 2.8s. The injection molding filling and cooling simulation is performed, and the warpage deformation vector field of the part after cooling is extracted. The warpage threshold is set to 0.004mm / mm. If the overall warpage of the component is greater than this value, it is determined to be a high-warpage component. Finally, the high-warpage component is imported into the DELMIA simulation system for assembly path simulation. The assembly fixture degrees of freedom are set to 3 and the degree of freedom error tolerance is 0.15mm. The assembly process analysis is performed, and the key assembly deformation displacement data and fixture interference rate are exported to form the assembly data of the high-warpage component.

[0031] Step S4: Structural design optimization is performed based on the assembly data of high-warpage parts to obtain part structural optimization data; the part structural optimization data is uploaded to the part 3D model, and shape matching retrieval is performed to generate retrieval data.

[0032] In this embodiment, the Altair Inspire Structure software is used to load the assembly data of high warpage parts, and the optimization goal is set in the structural optimization module to minimize the assembly interference rate and control the total mass to no more than ±3% of the initial design. The structural displacement inversion result is used as the input condition, and the topology optimization algorithm is applied for iterative solution. The upper limit of the material removal rate is set to 25%, and the shape change range is limited to the outer contour ±5mm. New geometric data is generated according to the optimization results, and the part structure optimization data file (STL format) is exported. The structural optimization data is uploaded to the previously constructed three-dimensional part model, and the three-dimensional shape matching algorithm based on voxel slicing and Hausdorff distance calculation is called. The slice spacing is set to 1.5mm, the shape similarity threshold is set to 95%, and the three-dimensional model is compared using Open3D, and matched with the 4980 standard part models stored in the database for retrieval. All part information with a shape overlap greater than 95% and completely consistent material categories will be output as retrieval data, including fields such as part code, application condition number, manufacturer number, etc., and finally saved as a structured retrieval data record for subsequent calls

[0033] Preferably, step S1 is specifically as follows:

[0034] Step S11: Obtaining manufacturing parts design drawings;

[0035] In this embodiment, data collection is performed on the part design drawings using a method based on optical character recognition (OCR) combined with computer-aided drawing recognition (CAD parsing). In the specific operation, first, an industrial-grade drawing scanning device (such as a Canon DR-M260 scanner) with a resolution of not less than 600dpi is used to perform high-precision digital collection of the physical paper drawings to generate a scanned image in TIFF format. Then, an image preprocessing module based on the Tesseract-OCR engine and embedded with a plug-in that supports multi-semantic drawing recognition is called to perform image region segmentation on the marked area, numbered area, dimension line, and annotation text in the drawings. Image preprocessing includes image binarization (using a fixed threshold value of 180), morphological corrosion expansion (kernel size of 3×3), and Hough line transform to remove the tilt angle (angle range set at ±15°). Subsequently, the structural lines in the drawings are subjected to structural data recognition by a CAD drawing parsing algorithm, which reads the geometric structure of the DWG file or the parsed SVG format based on the AutoLISP parsing rules, extracts the view projection angle (front view, side view, top view) and its corresponding projection size information. After completing the above processing, the drawing data is output into a unified standard intermediate representation format (such as XML / STEP format) for subsequent geometry and material feature extraction.

[0036] Step S12: Extracting geometric structure features and material features based on the manufacturing part design drawings to obtain geometric structure data and material data;

[0037] In this embodiment, geometric structure feature extraction is based on vector graphics recognition and shape semantic analysis, and material feature extraction is based on text semantic recognition and standard material library matching. Specifically, geometric structure feature extraction uses the Open Cascade Technology (OCCT) geometry engine to perform a three-dimensional boundary representation (B-rep) analysis on the STEP drawing generated in the previous step, extract the contour line coordinate value of the closed contour (in mm, with an accuracy of 0.01 mm), obtain boundary point cloud data, and construct boundary topology information. The extraction parameters include but are not limited to: the number of apertures (measured in diameters greater than 3 mm), the chamfer angle (extracting all chamfer information in the range of 0° to 60°), the critical edge length (edge ​​segment greater than 5 mm), and the curvature radius (surface radius area less than 30 mm). Material feature extraction uses a natural language processing engine (based on BERT semantic vectors) to analyze the marked area of ​​the drawing, identify description sentences that co-occur with keywords such as "material", "material", and "heat treatment requirements", and extract material type (such as SUS304 stainless steel, A7075 aluminum alloy), heat treatment method (such as T6 state), material density (lookup table value, such as SUS304 density is 7.93g / cm 3 ), yield strength (look-up table value, such as 503MPa for A7075), etc., and finally the geometric structure characteristics and material characteristics are uniformly encoded and output in JSON structured format.

[0038] Step S13: Modeling the part skeleton based on the geometric structure data to obtain part skeleton data;

[0039] In this embodiment, the skeleton modeling method based on feature drive (Feature-Based SkeletonModeling, FBSM) is used to construct the skeleton of the extracted geometric structure data. In implementation, the STEP format geometric structure data is first subjected to preliminary dimensionality reduction processing, and all closed contour projection boundaries are selected for two-dimensional sketch reconstruction. The ShapeAnalysis_FreeBounds tool in OCCT is used to identify free boundaries, and the centerline skeleton is generated in combination with the G1 continuity constraint. For multi-cavity parts, Boolean operation (Boolean cut) is used to obtain the main configuration section, and then a skeleton wireframe is constructed in the center of the section. After the skeleton wireframe is constructed, the spatial coordinate relationship of the nodes is connected by the wireframe (unit is mm, and the numerical accuracy is controlled to three decimal places) to generate a three-dimensional space skeleton graph structure. The distance between nodes shall not be less than 0.5mm, and the skeleton edge must maintain a minimum angle greater than 15° with the contour line or configuration axis. During the skeleton modeling process, a constraint-driven method is used to fix the origin, symmetry center line, and maximum axial direction, and labels are added to each skeleton edge (such as "spindle edge", "support edge", and "connection edge"). These label data are retained in the skeleton data as the basis of the part structure and output. The format uses XML structure description, and the output content includes skeleton point coordinate sets, edge sets, edge weight parameters, and edge label attribute sets.

[0040] Step S14: configuring the part attributes of the part skeleton data based on the material data to generate a three-dimensional model of the part.

[0041] In this embodiment, a parametric modeling platform (such as Siemens NX or CATIA V5) is used to load the skeleton data, and the physical properties are configured according to the extracted material data to generate the final three-dimensional model of the part. In the specific implementation process, the XML format skeleton data generated in step S13 is first imported into the modeling engine, and the skeleton is used as a geometric support to perform feature stretching, rotation, lofting and other operations to build a solid model. The stretching length and rotation angle are calculated based on the skeleton side length and the material ductility parameters. For example, when the material is A7075 aluminum alloy, the maximum stretching ratio is set to 1:5, and the stretching angle shall not exceed 35°. After the solid model is built, the material library is called to assign physical parameters to the three-dimensional structure, and the material properties are set according to the fields in the material name matching table. Specific properties include: density (in kg / m 3All parameters, including the material properties (in GPa, e.g., 193 GPa for SUS304), thermal conductivity (in W / m·K, e.g., 130 W / m·K for A7075), and Poisson's ratio (e.g., 0.30 for SUS304), are retrieved from a pre-set material database. These physical parameters are then added to the attribute layer of the 3D solid model. The resulting 3D part model is then exported in STEP AP214 or Parasolid format for subsequent processing and performance analysis.

[0042] Preferably, step S13 is specifically as follows:

[0043] Step S131: constructing a part sketch outline based on the geometric structure data;

[0044] In the present embodiment, sketch outline is constructed according to the geometric structure data extracted from the design drawings. Sketch is created using a CAD system (such as AutoCAD or Siemens NX), and first the geometric data (such as plane projection and orthogonal view) extracted from the manufacturing part design drawings is loaded. Through drawing data analysis, the contour boundary of the part is determined, including all necessary geometric features such as line segments, arcs and straight lines, to form the two-dimensional sketch outline of the part. The coordinate axis (unit is mm) with an accuracy of 0.01mm is used to generate sketch line segments to ensure that all point coordinate data meet the geometric structure constraints. In order to ensure that the contour is closed, the minimum error tolerance is set to 0.1mm to avoid the open area caused by design errors. By utilizing a graphic recognition algorithm (such as RANSAC straight line fitting or least squares arc fitting), the extracted line segments and arcs are optimized and all contour elements are connected into a closed area. This sketch outline data will serve as the basis for subsequent three-dimensional modeling.

[0045] Step S132: Filling the three-dimensional boundary according to the sketch outline of the part to obtain the three-dimensional boundary of the part;

[0046] In this embodiment, based on the sketch outline in step S131, an appropriate operation command (such as "area fill" or "stretch") is selected in the CAD system. When setting the boundary fill operation, a fixed thickness method is used for three-dimensional extension. The thickness value is set according to the part design requirements, usually 1mm to 100mm, and the specific value depends on the function and structural requirements of the part. The parameters in the stretching operation are set as follows: the stretching direction is the Z axis (that is, the direction perpendicular to the sketch plane), the stretching height is the design value, usually not more than 500mm, and the accuracy is controlled within 0.1mm. In the three-dimensional modeling process, by selecting the "solid modeling" operation, the two-dimensional sketch outline is converted into a three-dimensional solid boundary, and it is ensured that all boundary connections have a smooth transition and there are no staggered or overlapping geometric problems. This three-dimensional boundary defines the shape of the part and serves as the basis for the part structure in subsequent operations.

[0047] Step S133: Extending the structure path based on the three-dimensional boundary of the part to obtain a three-dimensional extended structure of the part;

[0048] In this embodiment, a three-dimensional modeling software (such as CATIA, SolidWorks or Siemens NX) is used to perform a path extension operation on the aforementioned three-dimensional boundary. First, the key edge or contour of the part is selected and the extension path is set. The path is selected based on the results of geometric analysis, and is usually extended along the main load-bearing direction or streamline direction of the part. The width, depth and direction of the extension path are strictly controlled according to the design requirements of the part. For example, for a part that needs to add a support structure, the width of the extension path can be set to 5mm, and the depth can be set to 10mm according to the structural strength requirements. During the extension operation, the path is stretched along the preset direction to generate a continuous three-dimensional structure. For parts with irregular surfaces, a segmented path extension method is used to decompose the irregular area into multiple controllable paths, and the angle between each path does not exceed 10 degrees to ensure the smoothness of the extension process. The final three-dimensional extension structure of the part provides a basis for the functional or assembly design of the part.

[0049] Step S134: locating the skeleton installation point according to the three-dimensional extension structure of the part;

[0050] In this embodiment, by performing stress analysis on the three-dimensional extension structure of the part, the stress concentration area is determined, and mechanical analysis software (such as ANSYS or ABAQUS) is usually selected to perform static analysis on the three-dimensional extension structure to calculate the stress distribution in each area. According to the stress concentration area and the geometric characteristics of the structure, a suitable skeleton installation point is selected, usually at the symmetry center of the part and in an area with high mechanical stability. For large-sized or multifunctional parts, the skeleton installation points can be distributed in multiple areas, and the precise position of each installation point is calculated by numerical methods. The size requirements of the installation point are set to 3mm to 20mm in diameter with an accuracy of 0.05mm to ensure that the skeleton connection part has sufficient stability. Finally, the installation point coordinates are generated according to the positioning results and output as three-dimensional space coordinate values ​​(for example, point 1: X = 20.1mm, Y = 15.3mm, Z = 10.5mm).

[0051] Step S135: Connect the part skeleton based on the skeleton installation points to obtain the part skeleton data.

[0052] In this embodiment, a three-dimensional modeling software (such as CATIA, Siemens NX, etc.) is used to perform skeleton connection at the designated installation points of the parts. First, the skeleton connection method is set: for simple parts, a straight line or curve connection can be used; for complex parts, a multi-point connection method is adopted. When the skeleton is connected, according to the design requirements of the parts, the skeleton material selection and connection method will affect the final connection strength and stability. For example, a connecting rod with a circular cross-section is used, the cross-sectional diameter is set to 5mm, the material is a high-strength aluminum alloy, the connection length is controlled within 100mm, and the connection accuracy is 0.1mm. Between the skeleton installation points, assembly is performed through connecting rods, supports and other components to ensure that the mechanical transmission path between each connection point is unobstructed and can meet the subsequent assembly and use requirements. Finally, the skeleton data is generated, including the connection point coordinates, the size of the connecting rod, the material type, the mechanical properties, etc. This data will serve as the basis for subsequent part assembly and optimization design, and will be output as a three-dimensional model file in STEP or Parasolid format for subsequent application.

[0053] Preferably, the surface treatment in step S2 is specifically:

[0054] Provide sandblasting process data based on parts process data;

[0055] In the present embodiment, the process data of the part is obtained, which is usually from the material properties, surface roughness requirements, wear resistance and other information contained in the manufacturing process design document or CAD model. The process data includes the smoothness and hardness requirements of the part surface, and the specific processing requirements of sandblasting (such as roughness Ra value, surface hardness, etc.). In a CAD system (such as AutoCAD or Siemens NX), a tool is used to extract the surface information of the part, and the specific parameters of the sandblasting process are extracted according to the requirements for surface roughness and physical properties in the design drawings. For the material of the part, the particle size and pressure range required for sandblasting can be set according to the alloy composition or heat treatment state. The input of sandblasting process data usually includes parameters such as the particle size range of the sandblasting particles, the injection pressure, the injection angle, and the sandblasting time. All process data will be stored in the form of an electronic spreadsheet or a database to ensure that the sandblasting parameters of each part at different process stages can be accurately transmitted and managed.

[0056] Identify target blasting areas based on the 3D model of the part;

[0057] In this embodiment, the part is three-dimensionally scanned or an existing CAD three-dimensional model is loaded using three-dimensional modeling software (such as SolidWorks, Siemens NX). By analyzing the surface geometry of the part, the areas that need to be sandblasted are identified. These areas usually include parts with rough surfaces, areas that need to increase surface hardness, and areas with improved contact quality. First, the surface of the part is segmented using a geometric feature extraction algorithm to identify areas with special surface features (such as uneven areas and areas with heavy wear). Then, a standard surface smoothness analysis method (such as Ra value or Rz value) is used to evaluate the entire surface of the part. For areas with higher surface roughness (Ra value greater than 3.2μm), they are calibrated as target sandblasting areas and displayed in the three-dimensional model. By setting a geometric morphology analysis algorithm, appropriate areas (such as curved surfaces, corners and edges) are selected for processing, and the spatial coordinates of these areas are recorded. These target sandblasting areas provide the basis for subsequent sandblasting process parameter settings.

[0058] Based on the sandblasting process data, 60-80 mesh coarse sandblasting particles were selected, and the sandblasting pressure was set to 0.6-0.8 MPa and the spray angle was set to 60°-75°. The target sandblasting area was subjected to primary sandblasting to obtain the rough blasting surface data.

[0059] In this embodiment, according to the target area identified in the above steps, the particle size of the sandblasting particles is determined to be 60-80 mesh, which is commonly used to remove larger surface defects and coarse oxide layers. At this time, the sandblasting pressure is selected between 0.6-0.8MPa, and the spray angle is set to 60°-75°. This angle range can ensure that the sandblasting particles hit the target surface at a suitable angle for preliminary surface treatment. In order to accurately control the spray parameters during the sandblasting process, these parameters are set using the automatic control system of the sandblasting machine, and the air flow pressure of the sandblasting machine is automatically adjusted according to the spray speed and pressure of the sandblasting machine to ensure uniform sandblasting. The sandblasting time is set to 5 to 10 minutes, depending on the roughness requirements of the part surface. Through this process, the rough surface data after preliminary sandblasting is obtained, mainly including surface roughness data and surface wear layer data. Finally, the measured surface roughness (Ra value) is usually in the range of 6-8μm, forming rough blasting surface data, and the data is saved as a three-dimensional scanning model or coordinate data.

[0060] According to the rough blasting surface data, the rough blasting surface area is identified, and 120-150 mesh medium-sized sandblasting particles are selected. The blasting pressure is adjusted to 0.4-0.6 MPa and the blasting angle is adjusted to 45°-60°. Intermediate sandblasting treatment is performed to obtain the medium blasting surface data.

[0061] In the present embodiment, the surface data after the rough blasting treatment is analyzed, and the rough blasting surface is scanned using surface analysis software (such as GOM Inspect) to identify areas where there is still a certain degree of unevenness after the rough blasting. Then, by selecting medium-sized sandblasting particles with a particle size of 120-150 mesh, setting the sandblasting pressure between 0.4-0.6MPa, and the spraying angle of 45°-60°, intermediate sandblasting is performed. This sandblasting parameter helps to further refine the surface, improve surface roughness and hardness, and remove residual particles or oxides on the surface. The pressure and angle adjustment during the sandblasting process can ensure that the impact force of the sandblasting particles is evenly distributed and optimize the surface quality. The sandblasting time is set to 3 to 5 minutes based on the data after the rough blasting treatment. During this process, the control system of the sandblasting machine automatically adjusts the airflow and particle flow according to the set pressure and angle to ensure the efficiency and consistency of the sandblasting process. Finally, the obtained medium-blasting surface data records the roughness (Ra value is usually between 3-5μm) and the change of surface morphology after surface improvement.

[0062] According to the surface data of the medium blasting, the surface area of ​​the medium blasting is identified, and 200-240 mesh fine-grained sandblasting particles are used, the spraying pressure is set to 0.2-0.4MPa, and the spraying angle is 30°-45°, and fine sandblasting treatment is performed to obtain the fine blasting surface data;

[0063] In this embodiment, the surface is made to achieve a higher degree of smoothness. Based on the medium-blasted surface data, a surface analysis tool is used to identify small-scale uneven areas that still exist on the surface. At this time, fine-grained sandblasting particles of 200-240 mesh are selected, the spraying pressure is adjusted to between 0.2-0.4MPa, and the spraying angle is set to 30°-45° for fine sandblasting. This sandblasting process helps to further remove tiny surface defects, improve surface finish, and ensure that the surface hardness meets the design requirements. The sandblasting time is set to 1 to 3 minutes. The spraying angle and pressure adjustment during the sandblasting process can ensure the precise spraying of sandblasting particles in small areas to remove tiny defects. After fine sandblasting, the surface roughness can be reduced to an Ra value between 1-2μm. Finally, the fine-blasted surface data is obtained, including the surface roughness data, hardness distribution and morphological changes after fine sandblasting.

[0064] The surface treatment data is obtained by integrating the rough blasting surface data, the medium blasting surface data and the fine sandblasting surface data.

[0065] In this embodiment, the surface data (including roughness, hardness, morphology, etc.) of each sandblasting stage are classified according to different sandblasting results. Then, data processing software (such as MATLAB or Excel) is used to merge and compare the surface data of each stage to ensure that the data of each stage are correctly recorded and analyzed. Finally, the data of the rough blasting, medium blasting and fine blasting stages are combined to generate a complete surface treatment data set containing all sandblasting process parameters, surface treatment effects, roughness distribution and hardness information. This data set can be used for subsequent part quality assessment, production process optimization and process improvement.

[0066] Preferably, the heat treatment in step S2 is specifically:

[0067] Extract part quenching parameters based on part process data;

[0068] In this embodiment, the process of extracting the quenching parameters of the parts needs to obtain the heat treatment requirements of the materials used in the parts by analyzing the design documents or process design specifications of the parts. These process data include the required heating temperature, cooling medium, heating time, cooling rate, etc. By consulting the material description of the parts (such as AISI, DIN, GB and other material standards), the material composition of the parts (such as carbon content, alloying elements, etc.) is obtained. Using this composition information, the relevant heat treatment atlas (such as heat treatment phase diagram) is consulted and the corresponding quenching temperature range is determined. In common steel quenching treatment, depending on the material, the conventional quenching temperature range is 800°C to 900°C. The optimal quenching parameters within this temperature range are obtained through a material property database or process design software (such as Aspen Plus, Thermo-Calc). In combination with the process requirements, the final quenching temperature suitable for the part is set. Ultimately, these quenching parameters will be provided to subsequent process steps in the form of process documents, databases or parameter tables.

[0069] Identify the steel area of ​​the part based on the 3D model of the part;

[0070] In this embodiment, it is necessary to load the three-dimensional model of the part through three-dimensional modeling software (such as SolidWorks, Siemens NX), and scan and geometrically analyze the model. Through geometric feature extraction technology, the material information of the part model is analyzed and its steel area is identified. This process can be achieved by setting an automated recognition algorithm to analyze the material properties of the part (usually specified as the steel type in the model) and locate the corresponding steel area. Specifically, using the material selection module in the CAD software, the steel area can be directly marked according to the material number or type. If the material information is not directly marked in the part model, it can be identified by combining the geometric features of the part (such as wall thickness, strength requirements, etc.) with manual settings or database queries. Finally, by partitioning the geometric shape of the part, it is determined which areas are steel areas, and these areas will become the target areas for subsequent heating and cooling treatments.

[0071] The steel area of ​​the part is heated and quenched according to the part quenching parameters, wherein the part is heated to the austenite region of 800°C-900°C to fully austenitize the structure and obtain the austenite structure state data of the steel part;

[0072] In this embodiment, the parts are placed in a resistance furnace or an induction heating furnace, and the heating temperature range is set to 800°C to 900°C. Within this temperature range, the steel will enter the austenitization zone and form an austenitic structure. In order to ensure that the steel parts can be fully austenitized, the heating rate needs to be controlled during the heating process. It is generally recommended that the heating rate be 50°C / min to 100°C / min to avoid excessive stress in the steel parts caused by too fast heating. Temperature control is monitored in real time by a temperature sensor (such as a thermocouple) to ensure that the temperature in the furnace is uniform and that the steel area of ​​the parts can be heated evenly. When the steel reaches the target temperature (such as 850°C), the temperature is maintained for 15 to 30 minutes to ensure that the austenitic structure is fully formed. The temperature change data, heating time, final temperature value, etc. during the heating process will be recorded as the austenitic structure state data of the steel parts. These data can be monitored and adjusted in real time by a temperature control system (such as a PID controller) to ensure that the heating process meets the preset standards.

[0073] According to the austenite structure state data of the steel part, rapid cooling treatment is carried out and oil quenching process is adopted to transform the structure into martensite to obtain the initial hardened structure data;

[0074] In this embodiment, after the austenitization of the steel part is completed, the oil quenching cooling treatment is quickly carried out. After the steel part is taken out of the furnace, it is immediately placed in a preheated oil tank for rapid cooling, and the oil temperature is usually controlled at 40°C to 60°C. During the oil quenching cooling process, the austenite on the surface of the part is transformed into martensite structure, thereby increasing the hardness of the steel part. During the oil quenching process, it is necessary to control the uniformity of the cooling to prevent the part from warping or cracking due to uneven cooling. In the oil tank, multiple sensors can be set to monitor the cooling rate in real time. The cooling rate is usually required to be 10°C / s to 30°C / s to ensure the formation of martensite structure. During the cooling process, the oil temperature needs to be continuously monitored by the temperature control system, and the final hardened structure state is recorded based on the temperature change data during the cooling process. The initial hardened structure data of the steel part can be obtained through the temperature change data during the cooling process (such as cooling time, oil temperature, final temperature, etc.), including information such as hardening depth and hardness distribution.

[0075] Tempering treatment is performed according to the initial hardened structure data, wherein the heating temperature is set to 200℃-600℃, and the tempering time and cooling rate are controlled to obtain the quenched and tempered structure data of the steel part;

[0076] In this embodiment, based on the initial hardened structure data, a suitable tempering temperature range is selected, which is usually set to 200°C to 600°C, and the tempering time is 30 minutes to 2 hours. The specific time is determined according to the thickness and hardness requirements of the steel part. The tempering process uses an electric furnace for heating, and the heating temperature is precisely controlled by a temperature sensor (such as an RTD probe). The tempering process is divided into preheating and heating stages. The preheating temperature is set to 150°C to 200°C to prevent the temperature from rising too quickly and causing thermal shock. In the tempering stage, the steel part is heated to between 200°C and 600°C, kept at a constant temperature for a period of time, and then naturally cooled or forced air cooled. The cooling rate is usually 1°C / s to 5°C / s to ensure uniform tempering of the steel part. By precisely controlling the temperature, time and cooling rate during the tempering process, the hardness and toughness of the steel part can be optimized, and finally the tempered structure data of the steel part can be obtained. These data include information such as hardness value, toughness, and internal stress distribution after tempering.

[0077] Based on the quenched and tempered structure data of steel parts, the heat treatment results of parts are modeled to obtain heat treatment data.

[0078] In the present embodiment, the hardness data, organizational state, internal stress distribution and other information after tempering are input into heat treatment simulation software (such as Thermo-Calc, DEFORM, etc.) to perform numerical simulation of the heat treatment process. This simulation process can predict the performance of steel parts under different tempering temperatures, times and cooling rates. Through modeling, the heat treatment process parameters can be further optimized to avoid the quality problems occurring in actual operation, such as cracks or dimensional changes. The heat treatment data generated by the model include final organizational state, hardness distribution, correlation between tempering temperature and time, etc. These data will provide a basis for the optimization of the heat treatment process parameters in subsequent production, and ultimately form a complete heat treatment data report.

[0079] Preferably, the welding process in step S2 is specifically as follows:

[0080] Extract part welding parameters based on part process data;

[0081] In the present embodiment, the material information (such as steel, aluminum alloy, etc.) and material thickness of the parts are obtained, and this information is usually provided by a design engineer or process design documents. By consulting material standards (such as GB / T 11345, ISO 24394, etc.), the welding process requirements of the material are obtained. These requirements include recommended welding current, voltage, welding speed, welding heat input, welding method, etc. Further, in combination with the structural characteristics of the parts (such as joint type, welding position, etc.), the preliminary range of welding parameters is determined. The welding current is usually set between 100A and 300A, the welding voltage is between 15V and 30V, and the welding speed is generally between 1mm / s and 5mm / s. Finally, these process parameters are input into the welding process simulation software (such as DEFORM, Simufact Welding) for preliminary verification, and the welding parameters are adjusted according to the simulation results to ensure the stability of the welding process and the welding quality.

[0082] Identify structural interface areas based on the three-dimensional model of the part;

[0083] In this embodiment, a complete three-dimensional model of the part is loaded, and the geometry and design intent of the part are analyzed in the CAD software. The interface areas where the components in the part are connected are determined by geometric analysis algorithms, such as algorithms based on boundary and contact surface recognition. Specifically, all joints, splices and welds on the surface of the part are identified. These areas usually include splicing edges, rib connection points or overlap areas of parts in process design. Appropriate thresholds are set in the model to distinguish between joint areas and non-joint areas. For example, the thickness threshold of the interface area is set to 5mm or more as a welded joint. The boundary processing function of the three-dimensional model is used to determine the precise position and shape of each joint area to ensure that the target area of ​​the subsequent welding process is clear.

[0084] Extract the laser reflectivity of the material based on the three-dimensional model of the part; determine the laser energy absorption efficiency based on the laser reflectivity of the material;

[0085] In this embodiment, a three-dimensional model of the part is loaded and its surface information, including properties such as surface material and roughness, is extracted. By combining the optical properties of the material (such as laser reflectivity), the optical properties database or literature (such as ASTM standards) of the material is consulted. For metal materials, common laser reflectivity values ​​are between 0.2 and 0.4, and the specific value is adjusted according to the material type and surface condition. The accurate value of the laser reflectivity is obtained by texture analysis and calculation of the part surface in CAD software. According to the characteristics of different materials (such as the laser reflectivity of stainless steel is 0.3), combined with factors such as laser irradiation angle, power and laser wavelength, the reflectivity value can be further refined and used for subsequent calculation of laser energy absorption efficiency. Laser energy absorption efficiency generally depends on the laser reflectivity of the material, the light absorption coefficient of the material and the characteristics of the laser beam. Assuming that the laser reflectivity is R, the absorption rate A can be calculated by the formula A=1-R. For example, for a material with a laser reflectivity of 0.3, its absorption rate A is 0.7. This absorption rate will directly affect the transmission efficiency of the laser energy during welding. Therefore, according to the different material reflectivities, the laser power is adjusted to ensure that sufficient energy can be delivered to the welding area. In practice, the laser power is set to an appropriate multiple of the material's energy absorption rate (for example, 1.5 to 2 times the absorption rate). Depending on the material's absorption rate, the laser beam's intensity and focus also need to be adjusted to achieve efficient energy transfer.

[0086] Select arc welding method based on laser energy absorption efficiency;

[0087] In this embodiment, for materials with higher laser energy absorption efficiency, laser welding or electron beam welding is selected, while for materials with lower energy absorption efficiency, arc welding is selected. By analyzing the welding characteristics of different welding methods, if the laser energy absorption efficiency of the material is lower than 0.5 (that is, the reflectivity is higher than 50%), arc welding is selected. Arc welding has strong welding capabilities and can weld thicker materials at higher power. In this step, based on the aforementioned laser absorption rate data, the type and power of the welding equipment are selected. The current setting of arc welding is generally 200A to 400A, and the arc voltage is 20V to 40V. The specific welding method selected will be determined based on the thickness, material and structural complexity of the welding area.

[0088] Determine welding accessibility based on arc welding method;

[0089] In this embodiment, the spatial information in the three-dimensional model is used to analyze the geometry of the welding area and the working range of the welding equipment. The operating range of the welding robot is simulated in the CAD software to ensure that the welding head can reach all welding interface areas. By setting the welding accessibility parameters (such as the maximum arm span of the welding robot, the shape of the welding head, etc.), the welding path is reasonably planned. If arc welding cannot be performed in certain areas due to geometric limitations or insufficient equipment space, it is necessary to adjust the welding parameters or change the welding method. The key operation of this step is to perform collision detection and path optimization through three-dimensional modeling tools (such as AutoCAD, SolidWorks, RobotStudio, etc.) to ensure the smooth progress of the welding process.

[0090] Planning welding trajectories based on welding accessibility;

[0091] In this embodiment, the geometric shape of the welding area and the starting and ending points of the welding path are analyzed. In combination with the size of the welding head and the shape of the welding trajectory, a trajectory planning algorithm (such as the shortest path algorithm or the Bezier curve algorithm) is used to plan the optimal welding trajectory. In the path planning, it is ensured that the welding head can evenly cover the welding area to avoid welding defects (such as uneven welds, pores, etc.) caused by unreasonable trajectories. Factors such as welding angle, welding sequence, and welding speed need to be considered in the trajectory planning process, and path simulation is performed through trajectory planning software (such as PathPilot, CAMWorks, etc.) to ensure the rationality of the trajectory.

[0092] The welding process is simulated according to the welding trajectory to obtain welding process data.

[0093] In this embodiment, the welding trajectory data is input into welding simulation software (such as Simufact Welding, ANSYS, etc.) to perform simulation analysis of the welding process. The simulation process requires setting parameters such as the thermophysical properties of the welding material, welding speed, and heat input, and the heat-affected zone, temperature field distribution, and stress field distribution generated during the welding process are simulated using the finite element analysis (FEA) method. The simulation results will output data such as temperature changes, thermal stress distribution, and weld quality during the welding process. This data is used to verify the stability of the welding process and optimize welding parameters. Ultimately, the welding process data obtained through simulation forms a complete welding process document, providing a basis for actual welding operations.

[0094] Preferably, step S3 is specifically as follows:

[0095] Step S31: drawing a thermal stress distribution map according to the heat treatment data; identifying high thermal stress areas based on the thermal stress distribution map;

[0096] In the present embodiment, the heat treatment process data of the part is obtained, and these data can be obtained by temperature sensors or thermal simulation software (such as ANSYS, ABAQUS, etc.). During the heat treatment process, the temperature changes at different positions on the surface and inside of the part are measured. These temperature changes can affect the stress distribution inside the part, so it is necessary to calculate it through the thermal stress calculation formula. Specifically, physical parameters such as the thermal expansion coefficient and Young's modulus of the material are used to calculate the stress value at different temperatures. According to the temperature field distribution and the mechanical properties of the material, the stress field is calculated using the finite element analysis method (FEA). After obtaining the stress field data, it is visualized using software to draw a thermal stress distribution diagram. This figure shows the stress magnitude through a color gradient, and a stress threshold is usually set (for example: stress greater than 300MPa is a high thermal stress area). Then, through graphic analysis technology, areas with higher stress are identified. These areas are usually located at the edge of the part, the joint area, or the part where the temperature changes faster. In the thermal stress distribution diagram, the areas with higher stress are marked and further analyzed.

[0097] Step S32: identifying surface cracks based on the surface processing data to obtain surface crack data;

[0098] In this embodiment, it is usually obtained through non-destructive testing techniques such as X-ray, ultrasonic, magnetic particle inspection and other methods. These methods can detect cracks and defects on the surface or inside of parts. The specific operation is to use ultrasonic flaw detection technology to generate sound waves on the surface of the material and receive echoes, and determine the existence and position of the cracks by the change of the echo. The size and depth of the crack are calculated according to the time delay and intensity of the echo signal. Combined with surface processing information such as surface roughness and hardness data, crack identification is performed in the data processing software. Set the standard for crack detection, such as the crack depth threshold is 0.1mm and the width threshold is 0.05mm, and mark the area that meets the conditions as the crack area. Through the image processing algorithm, the crack data is extracted from the surface scanning data to generate surface crack data.

[0099] Step S33: predicting the crack propagation path based on the surface crack data;

[0100] In this embodiment, a crack growth model (such as Paris'Law or Rice's Crack Growth Model) is applied to predict the crack propagation path. First, the position and shape of the crack are input into the crack growth model, and the crack expansion rate under different load conditions is obtained by mechanical calculation. The crack propagation path prediction takes into account the fatigue strength of the material, the stress intensity factor at the crack tip and the influence of the load cycle. By finite element analysis, the process of crack expansion on the material surface is simulated, and the time and path required for crack expansion are calculated. During the simulation process, a critical stress intensity factor threshold value for crack expansion is set, typically 5-15MPa√m. Using this threshold value, the direction and depth of crack propagation are predicted. The crack propagation path data finally obtained can be used for the coating thickness and risk assessment in the next step.

[0101] Step S34: Calculate the coating thickness based on the crack propagation path; and identify the coating mutation area according to the coating thickness;

[0102] In the present embodiment, the thickness of the coating on the surface of the part is measured by a coating thickness gauge (such as eddy current method or ultrasonic method). These data are used to analyze whether the coating can effectively prevent the expansion of the crack. The coating thickness data can be collected in real time by a sensor, or can be pre-set by the coating process specification of the coating. The thickness of the coating in different areas is counted to identify the area where the coating thickness changes greatly. Usually, the area where the coating thickness suddenly changes is the area where the coating performance is uneven or defective. By setting the standard threshold value (such as 3mm) of the coating thickness and the coating thickness change rate (such as more than 0.5mm per millimeter of change), the sudden change area is determined. The identification of the sudden change area depends on the spatial distribution of the coating thickness, and is statistically analyzed and processed by data analysis software (such as MATLAB, Excel, etc.).

[0103] Step S35: performing a region intersection operation based on the coating mutation region and the high thermal stress region to obtain coating peeling data;

[0104] In the present embodiment, utilize geometric overlap analysis technology, the coating mutation area and high thermal stress area are spatially overlapped.This process can use the spatial operation tool in 3D CAD modeling software (such as SolidWorks, AutoCAD etc.) to realize.By setting the standard of intersection operation (such as thermal stress is greater than 300MPa and coating thickness is less than 3mm), identify the area where coating peeling occurs.These intersection areas are generally the area where thermal stress is concentrated and coating mutation occurs, and represent that coating peeling occurs at these positions.Calculate the risk degree of coating peeling, obtain peeling data, and these data provide basis for the prediction of high risk area of ​​coating peeling in the next step.

[0105] Step S36: predicting high-risk areas for spalling based on the coating spalling data based on the welding process data to obtain high-risk areas for spalling;

[0106] In this embodiment, the influence of thermal stress generated during welding on the coating is simulated by calculating the welding heat input and the heat-affected zone. The temperature gradient and stress field generated during the welding process are used to calculate the difference between the thermal expansion of the coating and the thermal expansion of the substrate, and the peeling area of ​​the coating is obtained. Combined with the coating peeling data obtained in the previous steps, the stress analysis and thermal analysis results are used to predict the peeling risk of the coating in a high temperature environment. Usually, by setting a stress threshold (such as above 200MPa is high risk) and a temperature threshold (such as an area above 300°C is high risk), combined with crack data, the location of the high-risk area is determined. Ultimately, the high-risk area for peeling is the main influencing area for coating peeling.

[0107] Step S37: detecting the warpage rate of the injection molded part based on the high-risk area for peeling;

[0108] In this embodiment, the warpage of the injection molded parts is measured using a three-dimensional measuring instrument (such as a laser scanner or coordinate measuring machine) to detect deformation in the weld and coating peeling areas. Warpage is typically determined by measuring the height change at different points on the part surface. A warpage threshold is set. For example, when the warpage exceeds 2mm, the warpage in that area is considered high risk. During the measurement process, geometric deformation analysis methods are applied to compare the actual shape of the injection molded part with the original design shape to obtain warpage data. This data provides a basis for the next step of identifying high-warpage parts.

[0109] Step S38: identifying high-warpage components based on the warpage rate of the injection molded parts, and performing assembly simulation of the high-warpage components to obtain assembly data of the high-warpage components.

[0110] In this embodiment, by analyzing the warpage data, high-warpage parts whose warpage exceeds the set standard are identified. These parts usually produce large deformations due to factors such as uneven heating and coating peeling. Use three-dimensional CAD software (such as SolidWorks, CATIA, etc.) to perform assembly simulation on these high-warpage parts. The simulation process verifies whether the high-warpage parts can be assembled smoothly by setting assembly tolerances and position tolerances. During the simulation process, the assembly simulation tool is used to perform mechanical analysis on the high-warpage parts to ensure that the parts will not produce excessive stress concentration or interference due to warping during the assembly process. The final assembly data includes assembly position, deformation amount and assembly difficulty, providing a technical basis for actual assembly.

[0111] Preferably, step S37 is specifically as follows:

[0112] Step S371: performing a 3D scan of the injection molded part according to the high-risk area for peeling to obtain a 3D model of the injection molded part;

[0113] In this embodiment, after identifying the high-risk areas for peeling, a 3D scanner (e.g., a Creaform HandySCAN 3D or FARO Focus device) is used to perform a high-precision 3D scan of the injection molded part. During the scanning process, ensure that the resolution and accuracy of the scanner are sufficient to meet the requirements. A resolution of 0.05 mm is typically used, and the scanning point density is at least 500,000 points per second. The scanner needs to perform a comprehensive scan based on the geometry of the injection molded part, including all surfaces and their high-risk areas. During the scanning process, the injection molded part is placed on a stable stand to avoid any external vibration that affects the scanning accuracy. The goal of the scan is to generate complete 3D point cloud data of the injection molded part and ensure that the scan areas overlap to ensure data integrity. During the scanning process, the device automatically captures the surface of the injection molded part from different angles and generates point cloud data, ultimately synthesizing a 3D model of the entire injection molded part. This model is described as a set of discrete points in three-dimensional space through point cloud data. Finally, multiple scanning perspectives are merged using scanning software (such as Geomagic or CloudCompare) to obtain a complete 3D model of the injection molded part.

[0114] Step S372: collecting data based on the point cloud recognition of the injection molded part 3D model;

[0115] In this embodiment, the original point cloud data is processed using point cloud data processing software (such as CloudCompare, MeshLab or Geomagic). First, the original point cloud data obtained by scanning is imported, and the data is filtered for noise and irrelevant points are removed. The data is aligned to ensure that the point cloud data at different scanning angles can be correctly connected, and finally a high-precision surface model of the injection molded part is generated. Next, the point cloud data on the surface of the injection molded part is subjected to surface reconstruction processing. By performing surface fitting on the point cloud data, a surface model of the injection molded part is obtained, ensuring that each scanned point cloud data has a corresponding matching position in the model. In order to identify the accuracy of the point cloud acquisition data, the spatial coordinates (x, y, z) of each point cloud point need to be calibrated and optimized to ensure that the point cloud data can reflect the actual shape of the injection molded part. At this point, the obtained 3D model of the injection molded part contains a large amount of discrete point cloud data, and the coordinate accuracy of each point should be controlled within 0.1mm to ensure the high precision of the model.

[0116] Step S373: performing point cloud matching on the collected point cloud data according to the preset injection molded part point cloud standard data to obtain point cloud deviation data;

[0117] In this embodiment, the 3D model of the injection molded part is compared with a preset standard model through a point cloud matching algorithm. The standard model can be three-dimensional data generated by a design file (such as a CAD file), or an ideal geometric shape predefined based on process requirements. When performing point cloud matching, the point cloud data and the standard data are first preliminarily aligned using a coarse alignment technique to ensure that the approximate positions of the two are aligned. In order to further improve the accuracy, the iterative least squares (ICP) algorithm is used to finely align the point cloud data. The ICP algorithm minimizes the error by comparing the best matching point of each point cloud point on the standard model. After completing the matching of the point cloud and the standard data, the point cloud deviation data is obtained by calculating the spatial coordinate difference of each point. The deviation data reflects the geometric difference between the actual injection molded part and the standard model, and is usually expressed in millimeters. The threshold value of the point cloud deviation is usually set to 0.1mm, and any area where the deviation exceeds this value requires further inspection to evaluate the deformation or manufacturing error of the area.

[0118] Step S374: calculating the deformation of the injection molded part based on the point cloud deviation data;

[0119] In this embodiment, when calculating the deformation of an injection molded part based on point cloud deviation data, the spatial distribution of the point cloud deviations must first be analyzed. The deviation of each point refers to the difference in distance between that point in the actual injection molded part and the corresponding point in the standard model. To calculate these deviations, the actual coordinates of each point must first be compared with the coordinates of the corresponding point in the standard model. By calculating the deviations of all points, the deformation of each point is obtained. These deviations can be quantified by the difference in spatial coordinates. The deformation refers to the change in spatial position between the actual model and the standard model at a certain point, representing the degree of deformation at that point in the injection molded part. By summarizing the deviation values ​​of all points, the overall deformation of the injection molded part can be obtained. During the calculation process, a deformation threshold is typically set, such as a maximum allowable deformation of 0.5 mm. Any component exceeding this threshold requires further inspection and correction. To ensure calculation accuracy, computing tools such as MATLAB can be used to process the point cloud data using efficient algorithms. This ensures that the degree of deformation of the injection molded part can be accurately identified during deformation analysis, providing a basis for subsequent quality assessment and adjustment.

[0120] Step S375: Evaluate the warpage of the injection molded part based on the deformation of the injection molded part.

[0121] In this embodiment, warpage is an indicator that describes the degree of surface deformation of a molded part. It is typically expressed as the ratio of the maximum deformation of the molded part to its designed dimensions. This step first requires obtaining the molded part's geometric dimensions, including key dimensions such as length, width, and height. These dimensional data are typically obtained using precise measuring tools. Next, based on the calculated deformation data, the maximum deformation on the molded part's surface is determined. The point or area with the largest deformation is typically selected as the maximum deformation. The maximum geometric dimension of the molded part in a specific direction is then measured. This dimension is typically the widest or longest dimension defined during the design process. Based on these two parameters, the maximum deformation and maximum dimension, the warpage value is calculated. To ensure the accuracy of the warpage calculation, a threshold range for the warpage is typically set, typically between 0.5% and 2%. If the warpage exceeds this range, it indicates that the molded part is severely deformed and requires further inspection and correction. This process provides a detailed understanding of the actual deformation of injection molded parts during production, providing clear data support for quality control, subsequent production adjustments, and corrections. In practical applications, the evaluation process can utilize precise measuring equipment and calculation tools, such as 3D scanners and specialized software, to ensure the accuracy of deformation and dimensional data.

[0122] Preferably, step S4 is specifically as follows:

[0123] Step S41: calculating the warpage deformation variable according to the assembly data of the high-warpage component;

[0124] In this embodiment, first, the assembly data of high-warpage parts that have been measured and calibrated are obtained. These data include the size, shape, assembly position information and stress conditions of the parts. By obtaining this data, numerical calculation methods, especially based on finite element analysis (FEA) or other similar technologies, are used to calculate the deformation of the parts caused by warping during the assembly process. The deformation can be obtained by comparing the position changes of the parts before and after assembly. For example, during the assembly process, there are deformations caused by factors such as material properties, temperature changes or assembly pressure. By inputting the assembly data into the finite element model and combining it with mechanical simulation, the displacement changes between the parts before and after deformation are calculated. This process can be implemented using simulation software such as ANSYS, Abaqus, etc., in which the parameters that need to be set include the elastic modulus of the material, Poisson's ratio, stress conditions and assembly displacement. By setting these parameters, the warping deformation of the parts during the assembly process is finally obtained.

[0125] Step S42: evaluating component assembly clearance based on the warpage deformation variable;

[0126] In this embodiment, the size of the assembly gap directly affects the assembly accuracy and the fit of the parts. First, based on the warpage deformation variable, the spatial changes between the parts after assembly, especially the gaps between the assembly surfaces, are calculated. For example, warping causes the distance between certain contact surfaces of the parts to become larger or smaller. Based on this deformation information, the gaps between the assembly surfaces are measured using geometric analysis methods. This process usually relies on three-dimensional measurement tools, such as three-dimensional scanners or laser scanners, to obtain detailed point cloud data of the assembly data, and uses point cloud processing software (such as CloudCompare) to analyze the gap distribution between the parts. The gap data can be further used to analyze whether the assembly fit meets the design requirements. If the gap exceeds the set threshold, adjustments are required.

[0127] Step S43: identifying assembly stress concentration areas based on component assembly clearances;

[0128] In the present embodiment, usually, during the assembly process, uneven assembly gaps can lead to excessive stress concentration in certain areas, thereby causing potential damage or failure. In this step, first, based on the gap data obtained in the previous step and in combination with the mechanical properties of the material, the stress distribution of each part of the component is calculated by stress analysis. Usually, a finite element analysis (FEA) model is used to comprehensively analyze the gap data and the assembly state. Assembly stress concentration areas usually appear in areas where the assembly gap is small or the assembly position changes significantly, especially near the edge of the component or the contact surface. Through numerical simulation, the specific location of the stress concentration area is obtained, and the stress concentration area is marked. These areas often need to be focused on, especially in subsequent structural optimization.

[0129] Step S44: identifying structural stress weak points based on the assembly stress concentration area;

[0130] In this embodiment, the structural weak points are usually areas where stress concentration causes material fatigue, crack generation or excessive deformation. First, based on the distribution of stress concentration areas, a detailed local mechanical analysis is performed to evaluate the material strength and durability of these areas. Using a finite element analysis tool, the stress values ​​of the stress-concentrated areas are input into the model, and it is analyzed whether the stress in the area exceeds the yield limit or fatigue strength threshold of the material. Based on the analysis results, the weak points on the component structure are determined. This process requires setting some key parameters, such as the yield strength and fatigue limit of the material, and combining factors such as the working environment and load conditions of the components to comprehensively evaluate the weak points.

[0131] Step S45: performing structural redundancy enhancement according to the structural weak points to obtain part structure optimization data;

[0132] In this embodiment, after identifying the weak points on the component structure, structural redundancy enhancement is performed, that is, the load-bearing capacity of the structure is increased or the stress is dispersed through optimized design. According to the location of the weak points, enhancement measures are taken, such as adding reinforcement bars, improving material distribution, adjusting the force path, etc. Through structural optimization algorithms (such as topology optimization or size optimization), the shape or thickness of the component is adjusted according to the stress conditions to achieve the purpose of improving strength and durability. During the optimization process, a series of parameters need to be input, such as material strength, structural dimensions, optimization goals and constraints. After the optimization is completed, the optimized component structure data is obtained, which can be used for subsequent production and assembly processes.

[0133] It is particularly important that step S45 includes the following steps:

[0134] Step S451: Detecting stress conduction paths based on structural weak points;

[0135] In this embodiment, before performing structural optimization on the three-dimensional model of the part, it is first necessary to identify the stress concentration area of ​​the part under actual load conditions. Using the finite element analysis software ANSYS Mechanical 2022R2, the load boundary conditions are applied to the three-dimensional model and the static stress field distribution is solved. The load input parameters are set as follows: the fixed boundary is on the lower surface of the part, the applied load is in the center area of ​​the top of the part, the load size is 850N, and the direction is vertically downward. The simulation grid type uses a 10-node tetrahedron unit (SOLID187), and the grid size is controlled at 1.5mm to ensure a balance between accuracy and computational efficiency. The maximum stress gradient area is identified by the Von Mises stress distribution diagram, and the connection path of the equal stress area is traced according to the Euler path calculation method to form a continuous stress conduction path. The equal gradient increasing algorithm is used to control the stress change of adjacent grids in the path within 5MPa. The starting point of the path is fixed at the maximum stress point, and the end point extends to the fixed surface at the load outlet.

[0136] Step S452: Evaluate the load amount according to the stress conduction path;

[0137] In this embodiment, based on the above-mentioned stress conduction path, the piecewise integration method is used to evaluate the total load transferred in the path. First, a calculation node is set every 1.5 mm on the stress path, and the corresponding stress value σ and cross-sectional area A are extracted at each node. The cross-sectional area is extracted by the cross-sectional extraction module, which is perpendicular to the node normal direction. The transfer load of each segment is calculated by the formula F=σ*A, and finally all load segments are integrated to obtain the total transfer load on the path. To improve the accuracy, the Gaussian integration method is used for integration, and the number of sampling points is set to 5. All node data is stored as a structure array, which contains node number, three-dimensional coordinates, stress value, load value and adjacent connection node index information, which serves as the input basis for the subsequent generation of structural support units.

[0138] Step S453: generating a structural support unit according to the load amount and obtaining structural support unit data;

[0139] In this embodiment, based on the path load assessment, a support structure is constructed for the path segment with a transferred load exceeding 400N. The support unit is in the form of a closed triangular structure, and the vertices are anchored at the path nodes and the surrounding low-stress areas to form a stable support. The thickness of the support unit is set to 3mm, and the material is AISI 4340 steel with a yield strength of 470MPa. The geometry generation tool is SolidWorks 2023, and the support geometry is automatically generated according to the node coordinates through the API script. The construction rule is: the high-stress path segment node and two low-stress adjacent points (stress less than 50MPa) are combined into a stable triangular unit to generate a shell structure, and merged into the main part three-dimensional structure in Boolean Union mode. The generated structural support unit data is stored in XML format, recording the support number, node coordinates, support volume, material information and attachment node index.

[0140] Step S454: Designing a geometric connection structure based on the structural support unit data;

[0141] In this embodiment, the support unit data from the previous step is used to construct a transition connection with the main structure through spatial topological analysis. First, the node coordinates in the XML are parsed, the distance and angle at the junction of the support unit and the main structure are calculated, and the transition connection surface is generated using a three-dimensional Bezier surface interpolation algorithm. The thickness of the transition structure is set to 2.5 mm, and a three-dimensional entity is generated in SolidWorks using a surface stretching method. In order to ensure geometric continuity, the number of connection boundary interpolation control points does not exceed 6, and the interpolation curvature change is limited to 0.2. The connection structure is exported to STEP format according to the ISO / TS10303-203 standard as a geometric input file in the structural fusion processing.

[0142] Step S455: Reinforce the material according to the geometric connection structure to obtain material reinforcement data;

[0143] In this embodiment, based on the generated geometric connection structure, the material reinforcement process design is performed. The material reinforcement process uses a high-density particle reinforced composite material with a density of 2.3g / cm 3, with a Young's modulus of 120 GPa. The thickness of the reinforcement area was increased to 5 mm, with five layers of continuous glass fiber embedded inside, each 0.5 mm thick. The angle between the fiber laying direction and the principal stress direction did not exceed 15°, and the laying direction was calibrated using the optical image analysis software Image-Pro. The material reinforcement process was modeled in COMSOL Multiphysics, using a material layering model to simulate stress distribution changes. Finally, a material reinforcement data file was generated, including the fiber distribution map, reinforcement area coordinates, layer thickness, material properties, and mechanical response prediction results.

[0144] Step S456: Perform structural enhancement based on the material enhancement data to obtain part structure optimization data.

[0145] In this embodiment, the material enhancement data is imported into the ANSYS Workbench platform to perform overall stress simulation verification of the enhanced structure. The simulation model mesh type uses SOLID186 units, and the mesh size of the enhanced area is refined to 0.8 mm. The simulation boundary conditions are consistent with step S451, and the overall stress redistribution is obtained after the load is applied. The structural shape tuning module is then called to add a buffer edge with a transition chamfer radius of 1.5 mm to the edge of the enhanced area to relieve stress concentration. The enhanced structure is Boolean merged with the original part structure and output as a new three-dimensional structural model. The final part structure optimization data is stored in a structured JSON file format, and the record fields include part number, volume change before and after optimization, maximum stress difference, enhanced area geometric parameters, material ratio and load path change map.

[0146] Step S46: Upload the part structure optimization data to the part three-dimensional model, and perform shape matching retrieval to generate retrieval data.

[0147] In this embodiment, after completing the structural optimization of the component, the optimized structural data is integrated with the original three-dimensional model to update the three-dimensional design data of the component. Using CAD or CAE tools, the optimized structural features are uploaded to the three-dimensional model of the part, and shape matching retrieval is performed. When performing shape matching, the precise geometric data of the component is first obtained through a three-dimensional scanning tool, and the shape features of the component are analyzed using a feature extraction algorithm. By comparing with the standard model in the database, relevant retrieval data is extracted. The retrieval process is based on a similarity algorithm of geometric features, calculates the degree of matching between the part models, and finally generates retrieval data. This data is not only used for subsequent production monitoring, but can also be used for inventory management of components or further design modifications.

[0148] It is particularly important that step S46 includes the following steps:

[0149] Step S461: uploading the part structure optimization data to the part 3D model to obtain the part 3D optimization model;

[0150] In the present embodiment, it is first necessary to dock the structural optimization data of the part with the three-dimensional model. This operation includes uploading data such as geometric features, stress distribution, support structure design, material enhancement information generated during the part structure optimization process to the three-dimensional digital model of the part. To this end, the original three-dimensional model of the part is first opened by dedicated engineering software (such as CATIA, SolidWorks or NX). Then, the material enhancement data, stress distribution and other data information obtained during the structural optimization process are imported through an optimization design platform (such as ANSYS or Altair HyperWorks), and data is attached at the corresponding position. A unified format, such as STEP or IGES, must be used when uploading data to ensure data compatibility and seamless docking with the three-dimensional model. After this step is completed, a three-dimensional optimization model of the part containing optimized structural features is obtained, which can accurately reflect the geometric shape and material properties of the part during the optimization process.

[0151] Step S462: extracting structural features based on the three-dimensional optimized model of the part to obtain structural feature data;

[0152] In the present embodiment, based on the three-dimensional optimization model of the part, a geometric analysis tool is used to extract the structural features of the model comprehensively. This process first analyzes the geometric features of the model by CAD software (such as AutoCAD, SolidWorks), such as surface curvature, concave-convex part, edge profile, hole distribution, etc., extracts these geometric data and generates a feature matrix. In addition, the structural feature data such as the force distribution, stress concentration area, deformation behavior of the model are extracted by finite element analysis (FEA) software (such as ABAQUS or ANSYS). These data can help further analyze the performance of the part under stress. In the process, feature extraction relies on accurate geometric description, such as the shape complexity (curved surface, boundary, connection point, etc.) of the part, size and ratio, and requires each parameter (such as stress maximum, maximum deformation) in the process to be accurate to three decimal places, to ensure the accuracy and validity of the data.

[0153] Step S463: Calculating a geometric descriptor based on the structural feature data to obtain geometric descriptor data;

[0154] In this embodiment, geometric descriptors are calculated for the structural feature data extracted in the previous stage in order to perform subsequent shape matching. In this process, computational geometry technology is used, especially geometric processing methods based on boundary representation (B-rep) or volume representation. Geometric descriptors such as shape index (Shape Index), surface normal direction, curvature distribution, etc. are used to calculate the local geometric features of each part of the part. For example, the distribution of surface curvature in different parts can be calculated by a curvature analysis tool. Based on these calculations, one or more geometric descriptors are extracted to represent the morphological features of the part. In all calculation processes, the threshold value for curvature calculation is generally set between 0.1 and 1.0, and the difference range of the surface normal direction should be ±5°. These values ​​ensure the stability and operability of the geometric features.

[0155] Step S464: constructing a three-dimensional shape feature vector based on the geometric descriptor data;

[0156] In this embodiment, the calculated geometric descriptors (such as curvature, normal direction, volume, etc.) are first standardized into a numerical vector form. Each geometric descriptor represents a dimension, and the combination of all descriptors constitutes a feature vector. For example, the curvature values, normal directions, etc. of different areas on the surface of a part will be converted into individual components of the vector. In order to ensure the accuracy of the matching, the length and scale of the feature vector need to be unified, and each eigenvalue is usually normalized to ensure that the contribution of all feature dimensions is relatively balanced. In this process, the dimension of the vector is usually set to between 50 and 100 to ensure high computational efficiency and sufficient description accuracy.

[0157] Step S465: performing feature similarity matching based on the three-dimensional shape feature vector to obtain matching candidate vector data;

[0158] In this embodiment, the three-dimensional shape feature vectors of different parts are similarly calculated by a shape matching algorithm, with the aim of determining the candidate part that is most similar to the input part. This process is mainly achieved by calculating the similarity between the feature vector of each part and the feature vectors of other parts in the database. The cosine similarity method is often used for similarity calculation, which measures the angular difference between two vectors. The closer the value is to 1, the more similar the directions of the two vectors are. In order to perform similarity calculation, the feature vector of each part is first compared with the feature vectors of other parts one by one, and the similarity is measured by calculating their similarity scores. In this process, a threshold value (such as 0.8) is selected to ensure that the returned candidate vector has a high similarity with the input model, which means that only parts with a similarity higher than this threshold value will be selected as candidate data. The core goal of this step is to filter out the candidate part data that is closest to the shape of the query part for subsequent further analysis and sorting.

[0159] Step S466: sorting the search results according to the matching candidate vector data, thereby obtaining structural similarity ranking data;

[0160] In this embodiment, all candidate vectors are sorted in descending order according to the calculated similarity scores to generate a similarity-ranked list. During the sorting process, the similarity score must be above a preset threshold (e.g., 0.85), and a maximum number of candidates can be set based on actual needs, for example, returning only the top 10 or 20 most matching parts. This sorting ensures that the final search results efficiently and accurately return the most similar parts.

[0161] Step S467: Filter the optimal result according to the structure similarity sorting data to obtain the search data.

[0162] In this embodiment, a final screening is typically performed based on the first few parts in the similarity-ranked list (e.g., the first five or ten parts) to ensure that the returned parts are the most closely matched in terms of structure, function, and optimization. Depending on the needs of the actual application, other constraints, such as part size ranges and material properties, can be added to the screening process to further improve retrieval accuracy. During the screening process, techniques such as threshold screening and weighted matching can be utilized to optimize the final search data, ultimately obtaining part information that is most similar to the input model.

[0163] Preferably, this specification also provides a part three-dimensional model retrieval system based on feature extraction, which is used to execute the part three-dimensional model retrieval method based on feature extraction as described above. The part three-dimensional model retrieval system based on feature extraction includes:

[0164] The part 3D model building module is used to obtain the manufacturing part design drawings; extract the geometric structure features and material features based on the manufacturing part design drawings to obtain the geometric structure data and material data; and build the part 3D model based on the geometric structure data and material data;

[0165] The process processing module is used to extract process processing features based on the manufacturing part design drawings to obtain part process processing data; based on the part process processing data, the module performs process processing on the part 3D model, including surface treatment, heat treatment and welding treatment, to obtain surface treatment data, heat treatment data and welding data;

[0166] The assembly simulation module is used to perform coating peeling detection on surface treatment data based on heat treatment data to obtain coating peeling data; predict high-risk peeling areas on coating peeling data based on welding process data to obtain high-risk peeling areas; detect the warpage rate of injection molded parts based on the high-risk peeling areas; identify high-warpage parts based on the warpage rate of injection molded parts, and perform assembly simulation of high-warpage parts to obtain high-warpage part assembly data;

[0167] The shape matching retrieval module is used to perform structural design optimization based on the assembly data of high-warpage parts to obtain part structure optimization data; upload the part structure optimization data to the part 3D model, and perform shape matching retrieval to generate retrieval data.

[0168] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0169] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for retrieval of three-dimensional part models based on feature extraction, characterized in that: The following steps are involved: Step S1: Obtain a manufacturing part design drawing; perform geometric structure feature extraction and material feature extraction based on the manufacturing part design drawing to obtain geometric structure data and material data; Build a three-dimensional model of the part based on geometric structure data and material data; Step S2: Extract process characteristics based on the manufacturing part design drawings to obtain part process data; Performing process processing on the three-dimensional model of the part according to the process processing data of the part, including surface treatment, heat treatment and welding treatment, to obtain surface treatment data, heat treatment data and welding treatment data; Step S3: performing coating peeling detection on the surface treatment data according to the heat treatment data to obtain coating peeling data; Based on the welding process data, the coating peeling data is used to predict the high-risk area of ​​peeling and obtain the high-risk area of ​​peeling; Detect the warpage of injection molded parts based on high-risk areas for peeling; identify high-warpage parts based on the warpage of injection molded parts, and perform assembly simulation of high-warpage parts to obtain assembly data for high-warpage parts; Step S4: Optimizing the structural design based on the assembly data of the high-warpage component to obtain optimized component structure data; Upload the part structure optimization data to the part 3D model, perform shape matching retrieval, and generate retrieval data.

2. The method for retrieving a three-dimensional part model based on feature extraction according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: Obtaining manufacturing parts design drawings; Step S12: Extracting geometric structure features and material features based on the manufacturing part design drawings to obtain geometric structure data and material data; Step S13: Modeling the part skeleton based on the geometric structure data to obtain part skeleton data; Step S14: configuring the part attributes of the part skeleton data based on the material data to generate a three-dimensional model of the part.

3. The method for retrieving a three-dimensional part model based on feature extraction according to claim 2, characterized in that: Step S13 is specifically as follows: Step S131: constructing a part sketch outline based on the geometric structure data; Step S132: Filling the three-dimensional boundary according to the sketch outline of the part to obtain the three-dimensional boundary of the part; Step S133: Extending the structure path based on the three-dimensional boundary of the part to obtain a three-dimensional extended structure of the part; Step S134: locating the skeleton installation point according to the three-dimensional extension structure of the part; Step S135: Connect the part skeleton based on the skeleton installation points to obtain the part skeleton data.

4. The method for retrieving a three-dimensional part model based on feature extraction according to claim 1, wherein: The surface treatment in step S2 is specifically as follows: Provide sandblasting process data based on parts process data; Identify target blasting areas based on the 3D model of the part; Based on the sandblasting process data, 60-80 mesh coarse sandblasting particles were selected, and the sandblasting pressure was set to 0.6-0.8 MPa and the spray angle was set to 60°-75°. The target sandblasting area was subjected to primary sandblasting to obtain the rough blasting surface data. According to the rough blasting surface data, the rough blasting surface area is identified, and 120-150 mesh medium-sized sandblasting particles are selected. The blasting pressure is adjusted to 0.4-0.6 MPa and the blasting angle is adjusted to 45°-60°. Intermediate sandblasting treatment is performed to obtain the medium blasting surface data. According to the surface data of the medium blasting, the surface area of ​​the medium blasting is identified, and 200-240 mesh fine-grained sandblasting particles are used, the spraying pressure is set to 0.2-0.4MPa, and the spraying angle is 30°-45°, and fine sandblasting treatment is performed to obtain the fine blasting surface data; The surface treatment data is obtained by integrating the rough blasting surface data, the medium blasting surface data and the fine sandblasting surface data.

5. The method for retrieving a three-dimensional part model based on feature extraction according to claim 1, wherein: The heat treatment in step S2 is specifically as follows: Extract part quenching parameters based on part process data; Identify the steel area of ​​the part based on the 3D model of the part; The steel area of ​​the part is heated and quenched according to the part quenching parameters, wherein the part is heated to the austenite region of 800°C-900°C to fully austenitize the structure and obtain the austenite structure state data of the steel part; According to the austenite structure state data of the steel part, rapid cooling treatment is carried out and oil quenching process is adopted to transform the structure into martensite to obtain the initial hardened structure data; Tempering treatment is performed according to the initial hardened structure data, wherein the heating temperature is set to 200℃-600℃, and the tempering time and cooling rate are controlled to obtain the quenched and tempered structure data of the steel part; Based on the quenched and tempered structure data of steel parts, the heat treatment results of parts are modeled to obtain heat treatment data.

6. The method for retrieving a three-dimensional part model based on feature extraction according to claim 1, wherein: The welding process in step S2 is specifically as follows: Extract part welding parameters based on part process data; Identify structural interface areas based on the three-dimensional model of the part; Extract the laser reflectivity of the material based on the three-dimensional model of the part; determine the laser energy absorption efficiency based on the laser reflectivity of the material; Select arc welding method based on laser energy absorption efficiency; Determine welding accessibility based on arc welding method; Planning welding trajectories based on welding accessibility; The welding process is simulated according to the welding trajectory to obtain welding process data.

7. The method for retrieving a three-dimensional part model based on feature extraction according to claim 1, wherein: Step S3 is specifically as follows: Step S31: drawing a thermal stress distribution map according to the heat treatment data; identifying high thermal stress areas based on the thermal stress distribution map; Step S32: identifying surface cracks based on the surface processing data to obtain surface crack data; Step S33: predicting the crack propagation path based on the surface crack data; Step S34: Calculating coating thickness based on the crack propagation path; Identify coating abrupt change areas based on coating thickness; Step S35: performing a region intersection operation based on the coating mutation region and the high thermal stress region to obtain coating peeling data; Step S36: predicting high-risk areas for spalling based on the coating spalling data based on the welding process data to obtain high-risk areas for spalling; Step S37: detecting the warpage rate of the injection molded part based on the high-risk area for peeling; Step S38: identifying high-warpage components based on the warpage rate of the injection molded parts, and performing assembly simulation of the high-warpage components to obtain assembly data of the high-warpage components.

8. The method for retrieving a three-dimensional part model based on feature extraction according to claim 7, characterized in that: Step S37 is specifically as follows: Step S371: performing a 3D scan of the injection molded part according to the high-risk area for peeling to obtain a 3D model of the injection molded part; Step S372: collecting data based on the point cloud recognition of the injection molded part 3D model; Step S373: performing point cloud matching on the collected point cloud data according to the preset injection molded part point cloud standard data to obtain point cloud deviation data; Step S374: calculating the deformation of the injection molded part based on the point cloud deviation data; Step S375: Evaluate the warpage of the injection molded part based on the deformation of the injection molded part.

9. The method for retrieving a three-dimensional part model based on feature extraction according to claim 1, wherein: Step S4 is specifically as follows: Step S41: calculating the warpage deformation variable according to the assembly data of the high-warpage component; Step S42: evaluating component assembly clearance based on the warpage deformation variable; Step S43: identifying assembly stress concentration areas based on component assembly clearances; Step S44: identifying structural stress weak points based on the assembly stress concentration area; Step S45: performing structural redundancy enhancement according to the structural weak points to obtain part structure optimization data; Step S46: Upload the part structure optimization data to the part three-dimensional model, and perform shape matching retrieval to generate retrieval data.

10. A three-dimensional part model retrieval system based on feature extraction, characterized in that: For executing the method for retrieval of a three-dimensional part model based on feature extraction according to claim 1, the three-dimensional part model retrieval system based on feature extraction comprises: The part 3D model building module is used to obtain the manufacturing part design drawings; extract the geometric structure features and material features based on the manufacturing part design drawings to obtain the geometric structure data and material data; and build the part 3D model based on the geometric structure data and material data; The process processing module is used to extract process processing features based on the manufacturing part design drawings to obtain part process processing data; based on the part process processing data, the module performs process processing on the part 3D model, including surface treatment, heat treatment and welding treatment, to obtain surface treatment data, heat treatment data and welding data; The assembly simulation module is used to perform coating peeling detection on surface treatment data based on heat treatment data to obtain coating peeling data; predict high-risk peeling areas on coating peeling data based on welding process data to obtain high-risk peeling areas; detect the warpage rate of injection molded parts based on the high-risk peeling areas; identify high-warpage parts based on the warpage rate of injection molded parts, and perform assembly simulation of high-warpage parts to obtain high-warpage part assembly data; The shape matching retrieval module is used to perform structural design optimization based on the assembly data of high-warpage parts to obtain part structure optimization data; upload the part structure optimization data to the part 3D model, and perform shape matching retrieval to generate retrieval data.

Images