Finite element mesh generation method and device for target housing

Abstract

The application relates to a target shell finite element mesh generation method and device, the method comprising: acquiring an outline contour shape and multiple outline contour characteristics of a target shell; determining two corresponding skeleton reference lines of two opposite main surfaces of the target shell, and adaptively generating an internal shell skeleton of the target shell according to the two skeleton reference lines; respectively matching the internal shell skeleton and a corresponding finite element characteristic model of each outline contour characteristic from a preset database; performing finite element segmentation by using the corresponding finite element characteristic model, so as to obtain a shell skeleton discrete model corresponding to the internal shell skeleton and an outline contour discrete model corresponding to each outline contour characteristic; combining the shell skeleton discrete model and each outline contour discrete model; and outputting the combined discrete model as a standard finite element mesh. The application improves the accuracy of shell finite element mesh generation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric digital data processing, and in particular to a finite element mesh generation method and device for a target shell. BACKGROUND

[0002] The shell of a cast part is widely used in high-end equipment fields such as aerospace, automobile manufacturing, engineering machinery, etc. due to its flexible forming process and strong structural integrity, such as engine cylinder shell, spacecraft load-bearing shell, hydraulic valve group shell, etc. As the core load-bearing or protective component of equipment, the mechanical properties of such components directly determine the operation safety and reliability of the equipment, so accurate evaluation of the mechanical properties of the shell of the cast part through finite element analysis in the product design stage is a key link in the development process of the shell of the cast part.

[0003] The core premise of finite element analysis is to build a high-precision discretization model, that is, to divide the continuous shell structure of the cast part into a countable element set through finite element mesh division that meets the analysis requirements. The accuracy of mesh generation directly determines the reliability of the finite element analysis results. If the mesh has distortion, uneven size or poor boundary fitting, it will lead to stress calculation deviation, strain distribution distortion, and even analysis non-convergence, thereby misleading design decisions.

[0004] Currently, the finite element modeling of the shell of the cast part in the industry generally adopts the traditional method of directly dividing the overall geometric model. The core process is: based on the three-dimensional design drawing of the cast part, directly calling the automatic mesh division function of the finite element analysis software, and performing a one-time mesh generation operation on the overall structure of the shell. However, for cast parts with complex structural characteristics, the inherent defects are more prominent, directly leading to insufficient accuracy of finite element mesh generation. SUMMARY

[0005] The present application provides a finite element mesh generation method and device for a target shell to solve the problem of insufficient accuracy of finite element mesh generation.

[0006] In a first aspect, the present application provides a finite element mesh generation method for a target shell, comprising:

[0007] Obtaining the shape contour form and multiple shape contour features of the target shell;

[0008] determining a respective skeleton reference line for each of two opposite main surfaces of the target shell, and adaptively generating an internal shell skeleton of the target shell according to the two skeleton reference lines, wherein the two opposite main surfaces include a first main surface and a second main surface, a distance between the first main surface and the second main surface refers to a maximum thickness of the target shell in a vertical direction of the first main surface, and the internal shell skeleton is a regular structure formed by splicing at least one cuboid, and the internal shell skeleton maximally fits an external contour morphology of the target shell;

[0009] matching a finite element feature model corresponding to each of the internal shell skeleton and the contour feature from a preset database;

[0010] performing finite element segmentation by using the corresponding finite element feature model to obtain a shell skeleton discrete model corresponding to the internal shell skeleton and a contour discrete model corresponding to each of the contour features;

[0011] combining the shell skeleton discrete model and each of the contour discrete models, and outputting the combined discrete model as a standard finite element mesh.

[0012] Optionally, determining a respective skeleton reference line for each of two opposite main surfaces of the target shell, and adaptively generating an internal shell skeleton of the target shell according to the two skeleton reference lines includes:

[0013] determining two opposite main surfaces of the target shell, the two opposite main surfaces being a pair of opposite surfaces constituting a thickness direction of the target shell;

[0014] determining a skeleton reference line for each of the main surfaces inside the target shell, the skeleton reference line being a straight line arranged along an extension direction of the corresponding main surface, and the skeleton reference line being located within a contour boundary of the target shell;

[0015] respectively extending along a first direction and a second direction inside the target shell with each of the skeleton reference lines as a reference, wherein the first direction is the extension direction of the corresponding main surface, an inflection point of an extension path in the first direction coincides with a thickness change position of the corresponding main surface, the second direction is perpendicular to the first direction, and an end point of an extension path in the second direction is a shallowest position of a vertical side main surface;

[0016] generating a plurality of cuboids according to the extended boundary, the plurality of cuboids being spliced at the inflection point to obtain the internal shell skeleton.

[0017] Optionally, determining a skeleton reference line for each of the main surfaces includes:

[0018] obtaining a minimum mesh wall thickness and a wall thickness coefficient in finite element analysis, wherein the wall thickness coefficient is used to indicate a number of mesh layers of the minimum mesh wall thickness accommodated between the skeleton reference line and the corresponding main surface;

[0019] taking the minimum distance between the two opposite main surfaces as a local shell wall thickness;

[0020] determining a target spacing between the skeleton reference line and the corresponding main surface according to the minimum mesh wall thickness, the wall thickness coefficient, and the local shell wall thickness;

[0021] drawing a straight line as the skeleton reference line along the extension direction of the corresponding main surface at a target internal position of the target shell maintaining the target spacing with the main surface.

[0022] Optionally, generating a plurality of cuboids according to the extended boundary includes:

[0023] mapping the distance between the two adjacent inflection points in the first direction as the length of a cuboid;

[0024] mapping the extension distance in the second direction as the width of the cuboid;

[0025] mapping the perpendicular distance between the two skeleton reference lines as the height of the cuboid;

[0026] forming a cuboid according to the length, width and height of each cuboid, wherein the width of each cuboid is the same, and the length and height are not completely the same.

[0027] Optionally, determining the skeleton reference line corresponding to each of the two opposite main surfaces of the target shell, and adaptively generating the internal shell skeleton of the target shell according to the two skeleton reference lines includes:

[0028] determining the two opposite main surfaces of the target shell, which are a pair of opposite surfaces constituting the thickness direction of the target shell;

[0029] identifying the thickness change position of each main surface as an inflection point of the extension path of the corresponding skeleton reference line;

[0030] segmenting the target shell into a plurality of independent blocks with the inflection points as the demarcation points;

[0031] determining a sub-skeleton reference line for each of the two opposite sub-main surfaces of each independent block, wherein the sub-skeleton reference line is a straight line arranged along the extension direction of the corresponding sub-main surface, and the sub-skeleton reference line is located within the contour boundary of the independent block;

[0032] respectively, each of the sub-skeleton reference lines in each of the independent blocks is extended along a second direction, wherein the second direction is a vertical direction of a corresponding sub-main surface, and an end point of an extension path of each of the sub-skeleton reference lines is a shallowest position of a vertical side main surface of the target shell;

[0033] According to the extended boundaries in each of the independent blocks, a corresponding cuboid is generated respectively, and the cuboids are spliced at the inflection points to obtain an internal shell skeleton of the target shell.

[0034] Optionally, after obtaining the shell skeleton discrete model corresponding to the internal shell skeleton and the contour profile discrete model corresponding to each contour profile feature, the method further comprises:

[0035] Identifying specific regions to be corrected in the shell skeleton discrete model and each of the contour profile discrete models, the specific regions including a circular surface region, a quasi-rectangular surface region, a transition region from narrow to wide, and a transition region from wide to narrow;

[0036] Correcting an original mesh flow of the circular surface region into an O-shaped mesh flow to adapt to the contour shape of the circular surface;

[0037] Adjusting the distribution of mesh nodes of the quasi-rectangular surface region to keep the number of mesh nodes on four boundaries of the quasi-rectangular surface consistent;

[0038] Performing transition processing on the mesh of the transition region from narrow to wide and the transition region from wide to narrow in a combination of quadrilateral mesh and triangular mesh to eliminate mesh deformities at the wide-narrow transition;

[0039] After completing the mesh flow correction of each of the specific regions, outputting the corrected shell skeleton discrete model and the corrected contour profile discrete model.

[0040] Optionally, combining the shell skeleton discrete model and each of the contour profile discrete models, and outputting the combined discrete model as a standard finite element mesh comprises:

[0041] Unifying the coordinate system of the shell skeleton discrete model and the coordinate system of each of the contour profile discrete models to a global coordinate system of the target shell;

[0042] Based on the position correlation between the internal shell skeleton and each of the contour profile features, taking the shell skeleton discrete model as a basic framework, mapping each of the contour profile discrete models to a corresponding spatial region;

[0043] Performing coordinate alignment on overlapping mesh nodes of adjacent discrete models, and performing fusion processing on adjacent mesh contact surfaces to form an integrated combined discrete model;

[0044] After the uniform physical property parameters are configured for the combined discrete model and the finite element analysis constraint identification is marked, the combined discrete model is converted into a standard finite element grid.

[0045] In a second aspect, the application provides a finite element grid generation device for a target shell, the device comprising:

[0046] An acquisition module is configured to acquire an external contour shape of the target shell and a plurality of external contour features.

[0047] A generation module is configured to determine a respective skeleton reference line for each of two opposite main surfaces of the target shell, and to generate an internal shell skeleton of the target shell based on the two skeleton reference lines, wherein the two opposite main surfaces include a first main surface and a second main surface, the distance between the first main surface and the second main surface refers to the maximum thickness of the target shell in the vertical direction of the first main surface, the internal shell skeleton is a regular structure formed by splicing at least one cuboid, and the internal shell skeleton maximally fits the external contour shape of the target shell.

[0048] A matching module is configured to match a respective finite element feature model for the internal shell skeleton and each external contour feature from a preset database.

[0049] A segmentation module is configured to perform finite element segmentation using the respective finite element feature model to obtain a shell skeleton discrete model corresponding to the internal shell skeleton and an external contour discrete model corresponding to each external contour feature.

[0050] An output module is configured to combine the shell skeleton discrete model and each external contour discrete model, and to output the combined discrete model as a standard finite element grid.

[0051] In a third aspect, the application provides an electronic device, comprising: at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor connected to the at least one bus; and at least one memory connected to the at least one bus.

[0052] In a fourth aspect, the application further provides a computer storage medium storing computer executable instructions for executing the finite element grid generation method for a target shell according to any one of the above aspects.

[0053] Compared with the prior art, the above technical solution provided by the embodiments of the present application has the following advantages: firstly, the shape contour form and the plurality of shape contour features of the target shell are acquired, and the overall structural layout and the local detail information of the shell are completely captured; then, the respective skeleton reference lines of the two opposite main surfaces of the target shell are determined, and the internal shell skeleton formed by splicing the cuboids is adaptively generated according to the two skeleton reference lines, which can maximize the adaptation of the shape contour form of the target shell, and can convert the originally irregular complex shell form into a regular skeleton structure while completely retaining the irregular shape contour features of the shell outside; then, for the internal shell skeleton and each shape contour feature, the finite element feature model adapted thereto is used to perform finite element segmentation, to generate the corresponding shell skeleton discrete model and shape contour discrete model, and the generation accuracy of each discrete model is improved by avoiding the insufficient adaptation of a single segmentation logic to different structures through the targeted segmentation mode; finally, all the discrete models are combined and output as a standard finite element mesh. By disassembling the target shell into an internal regular skeleton and external contour features and using the adapted finite element feature model to complete the segmentation, the local mesh distortion problem caused by structural heterogeneity during overall segmentation can be effectively avoided, and the accuracy of the shell finite element mesh generation is finally improved. BRIEF DESCRIPTION OF DRAWINGS

[0054] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate one embodiment consistent with the present application and, together with the description, serve to explain the principles of the application.

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained from these drawings without any creative labor.

[0056] One or more embodiments are exemplarily illustrated by the pictures in the corresponding drawings, which do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, unless otherwise specified. The drawings in the drawings do not constitute a proportional limitation.

[0057] Figure 1 The finite element mesh generation system principle diagram of the target shell provided by the embodiments of the present application;

[0058] Figure 2 The finite element mesh generation method flowchart of the target shell provided by the embodiments of the present application;

[0059] Figure 3 The plane structure schematic diagram of the complex shell provided by the embodiments of the present application;

[0060] Figure 4 A schematic diagram of the planar structure of the target housing provided in the embodiments of this application;

[0061] Figure 5 A flowchart illustrating a finite element mesh generation method for a target shell, as provided in an embodiment of this application;

[0062] Figure 6 A schematic diagram of a finite element mesh generation device for a target shell provided in an embodiment of this application;

[0063] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0065] The following disclosure provides numerous different embodiments or examples for implementing various structures of this application. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of this application. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed.

[0066] Optionally, in the embodiments of this application, the above-described finite element mesh generation method for the target shell can be applied to, for example... Figure 1 The hardware environment shown consists of terminal 101 and server 103. Figure 1 As shown, server 103 is connected to terminal 101 via a mesh network and can be used to provide services to the terminal or clients installed on the terminal. Database 105 can be set up on the server or independently of the server to provide data storage services for server 103. The mesh network includes, but is not limited to, wide area network, metropolitan area network or local area network. Terminal 101 includes, but is not limited to, PC, mobile phone, tablet computer, etc.

[0067] The following will describe in detail a finite element mesh generation method for a target shell provided in this application embodiment, taking a server as an example, with specific implementation details. Figure 2 As shown, the specific steps are as follows:

[0068] Step 201: Obtain the shape contour form and various shape contour features of the target shell;

[0069] Step 202: Determine the respective skeleton reference lines of the two opposite main surfaces of the target shell, and generate an internal shell skeleton of the target shell according to the two skeleton reference lines, wherein the two opposite main surfaces include a first main surface and a second main surface, the distance between the first main surface and the second main surface refers to the maximum thickness of the target shell in the vertical direction of the first main surface, the internal shell skeleton is a regular structure formed by splicing at least one cuboid, and the internal shell skeleton maximizes the adaptation of the external contour form of the target shell;

[0070] Step 203: Match the finite element feature model corresponding to the internal shell skeleton and each shape contour feature from a preset database, respectively;

[0071] Step 204: Perform finite element segmentation using the corresponding finite element feature model to obtain a shell skeleton discrete model corresponding to the internal shell skeleton and a shape contour discrete model corresponding to each shape contour feature;

[0072] Step 205: Combine the shell skeleton discrete model and each shape contour discrete model, and output the combined discrete model as a standard finite element mesh.

[0073] In step 201, in the pre-processing stage of the finite element analysis of the casting part, the technician imports the three-dimensional design model of the casting part into the terminal, and the three-dimensional design model automatically extracts the shape contour form of the target shell of the casting part through the geometric topology analysis function, including the overall extension range of the shell, the surface transition trend, and the wall thickness distribution of different regions. At the same time, the feature recognition function is used to capture various shape contour features on the target shell, such as reinforcement intersection features, spring disc features, and sub-frame mounting point features, etc., and the specific geometric parameters of each type of shape contour feature are recorded, such as the thickness of the reinforcement intersection, the diameter of the spring disc, the hole diameter of the 3D hole, etc.

[0074] After completing data collection, the terminal packages the 3D model file of the shape contour form and the parameter table of the shape contour feature into structured data and transmits it to the background server.

[0075] In step 202, after the server receives the data uploaded by the terminal, it first analyzes the shape and contour of the target shell through a geometric structure analysis algorithm, screens and determines two opposite main surfaces. In the actual processing logic, any set of opposite surfaces in the target shell has the basic condition of serving as the opposite main surface. The two opposite main surfaces include a first main surface and a second main surface. The first main surface can be any surface of the target shell, can be a surface with the widest extension range, or can be a surface that can reflect the structure trend of the shell. When measuring along the vertical direction of the first main surface, the corresponding surface that forms the maximum distance with the first main surface is the second main surface. The distance between the second main surface and the first main surface is the maximum thickness of the target shell in the vertical direction.

[0076] The main surface has a trend reference line that reflects its extension trend. The trend reference line of the first main surface can be parallel to the trend reference line of the second main surface, or can be slightly inclined within 5°. This slight inclination angle does not affect the distance between the second main surface and the first main surface in the corresponding vertical direction, which is the maximum thickness of the target shell.

[0077] When selecting the first main surface, the surface with a wider extension area that can reflect the structure trend of the shell can be selected as the first main surface according to the coverage priority principle. For example, for a target shell with irregular structure, among the surfaces in the same direction, if one side is a long strip-shaped surface with a larger extension area and the other side is a small area short surface with local protrusions, the long strip-shaped surface and the surface on the opposite side are defined as the opposite main surfaces. This selection can make the main surface more fully cover the structure area of the target shell, provide a reference for the generation of the skeleton reference line that fits the overall shape of the shell, and effectively avoid the situation that the skeleton reference line cannot accurately map the overall contour shape of the target shell due to the small coverage range of the main surface.

[0078] After determining each opposite main surface, the server generates a skeleton reference line corresponding to each main surface based on the extension direction of the main surface. The skeleton reference line is a straight line that is consistent with the overall trend of the corresponding main surface. For example, if the main surface extends along the front-back horizontal direction of the target shell, the skeleton reference line is arranged along the front-back direction, and the skeleton reference line and the main surface maintain a certain distance.

[0079] The server uniformly extends in the internal space of the shell based on the skeleton reference lines of the two opposite main surfaces. During the extension process, the boundary adaptation algorithm is used to fit the internal wall contour boundary of the shell to form an internal shell skeleton that adapts to the internal space form of the shell. The internal shell skeleton is a splicing of multiple cuboids.

[0080] Specifically, different forms of the shell correspond to different internal shell skeleton structures. For example, if the target shell is a cuboid, the internal shell skeleton is a maximum cuboid structure that fits the inner wall of the shell and has the same length, width, and height ratio as the shell as a whole. This cuboid structure can minimize the deviation between the internal shell skeleton and the target shell in actual form, reducing errors in subsequent block dispersion. If the target shell is a circle, the internal shell skeleton is a maximum cuboid that can be accommodated inside the shell. The four corners of the cuboid are fitted to the tangent positions of the circular inner wall, although there is a certain gap with the circular inner wall, the regular cuboid structure can greatly simplify the calculation logic of subsequent block dispersion, reducing the complexity of grid generation. If the shell is a combination of a rectangle and a small hexagon, as shown in Figure 3 Figure 3 is a planar structure diagram of a complex shell. As can be seen, the internal shell skeleton is divided into two parts. The large cuboid in the rectangular area is adapted to the size of the area, and the small cuboid in the small hexagonal area is adapted to the outline of the area. The two cuboids are seamlessly spliced through coordinate calibration at the junction to ensure that the skeleton of each area can match the structural characteristics of the corresponding area.

[0081] In this application, the cuboid is a regular geometric structure with a mature grid division algorithm and controllable calculation error, which can effectively avoid problems such as grid distortion and substandard unit quality that may occur when irregular structures are dispersed.

[0082] In step 203, the server side pre-builds and stores a finite element feature database, which contains two types of model resources: one is a basic finite element feature model adapted to the cuboid, and the other is a standard finite element feature model covering multiple types of outline features. All finite element models stored in the database have been verified through multiple rounds of actual simulation, and their grid quality and element accuracy meet the precision threshold requirements of scenarios such as casting strength analysis and vibration analysis.

[0083] In the model matching stage, the server first obtains the core parameters of each cuboid in the internal shell skeleton, including geometric dimensions (length, width, height), spatial coordinates, and associated positions with other cuboids. Then, based on the extracted core parameters, the server calls the parameter similarity algorithm to select the corresponding basic finite element feature model from the database. For each type of outline feature, the server extracts the geometric parameters and structural attributes of the outline feature one by one, and determines the corresponding standard finite element feature model through geometric parameter and attribute matching logic. The geometric parameters can be the overlap width of a flange, the hole diameter of a 3D hole, etc., and the structural attributes can be the connection type of a fastener, etc.

[0084] By directly matching the corresponding finite element feature model in the database, on the one hand, the modeling period of the finite element pre-processing can be greatly shortened, and on the other hand, the standardized model can avoid the parameter deviation that may occur in manual modeling.​

[0085] In step 204, for the regular internal shell skeleton, the server first calls the matched basic finite element feature model, and then calls the special segmentation algorithm adapted to the model, such as the hexahedral element sweep algorithm. This algorithm is suitable for regular structures such as cuboids and can generate hexahedral grid elements uniformly along the extension direction of the skeleton, ensuring the adaptability of the grid to the internal shell skeleton structure, and finally generating a shell skeleton discrete model completely corresponding to the internal shell skeleton.

[0086] For each type of contour feature, the server calls the corresponding standard finite element feature model, selects a dedicated segmentation algorithm according to the structural complexity and analysis requirements of the feature, and generates a contour discrete model corresponding to each contour feature. For example, the reinforcement intersection feature requires high structural stress concentration, so a first-order tetrahedral element segmentation algorithm is used to ensure grid density; the flange feature is used to ensure the accuracy of the connection strength analysis, and a three-layer shell element segmentation algorithm is used; the logos do not affect the structural performance, and a simplified segmentation method of face deletion and hole filling is used.

[0087] In the above process of determining the discrete model, the server supports multi-thread parallel computing, that is, the shell skeleton discrete model of the internal shell skeleton and the contour discrete model of each type of contour feature can be generated simultaneously without waiting in sequence. Through reasonable allocation of computing resources, the generation time of the discrete model is shortened.

[0088] In step 205, after obtaining each discrete model, the server maps each contour discrete model to the corresponding spatial position of the shell skeleton discrete model, such as mapping the discrete model of the spring disc feature to the top pre-set area of the skeleton, and mapping the discrete model of the 3D hole feature to the corresponding hole position of the side wall of the skeleton. Through spatial coordinate calibration, all discrete models are accurately fitted to finally combine into a complete overall finite element model.

[0089] After completing the model combination, the server calls the format conversion module to convert the overall finite element model into the standard format in the field of finite element simulation, including the.inp format or the.cdb format, etc., to ensure that the overall finite element model can be directly recognized by mainstream simulation software. At the same time, the server will automatically generate a grid quality report containing key indicators such as element pass rate, grid distortion, and element length-width ratio, which can help subsequent users quickly judge whether the model quality meets the analysis requirements.

[0090] In the present application, firstly, the shape contour form and various shape contour features of the target shell are acquired, and the overall structural layout and local detail information of the shell are completely captured; then, the skeleton reference lines corresponding to the two opposite main surfaces of the target shell are determined, and the internal shell skeleton formed by splicing cuboids is adaptively generated according to the two skeleton reference lines. The internal shell skeleton can not only maximize adapt to the shape contour form of the target shell, but also convert the originally irregular complex shell form into a regular skeleton structure while completely retaining the irregular shape contour features of the shell outside. Then, for the internal shell skeleton and each shape contour feature, the corresponding shell skeleton discrete model and shape contour discrete model are generated by performing finite element segmentation on the adaptively matched finite element feature model, and the generation accuracy of each discrete model is improved by avoiding the insufficient adaptation of a single segmentation logic to different structures through a targeted segmentation mode. Finally, all discrete models are combined and output as standard finite element grids. By disassembling the target shell into an internal regular skeleton and external contour features and performing segmentation on the adaptively matched finite element feature model, the present application can effectively avoid the local grid distortion problem caused by structural heterogeneity during overall segmentation, and finally improve the accuracy of shell finite element grid generation.

[0091] As an optional implementation, in step 202, determining the skeleton reference lines corresponding to the two opposite main surfaces of the target shell and adaptively generating the internal shell skeleton of the target shell according to the two skeleton reference lines includes two embodiments, which are described below.

[0092] In the first embodiment,

[0093] Step S11: determining the two opposite main surfaces of the target shell, the two opposite main surfaces being a pair of opposite surfaces constituting the thickness direction of the target shell;

[0094] Step S12: determining a skeleton reference line for each main surface in the target shell, the skeleton reference line being a straight line arranged along the extension direction of the corresponding main surface, and the skeleton reference line being located within the contour boundary of the target shell;

[0095] Step S13: extending along a first direction and a second direction in the target shell with each skeleton reference line as a reference, wherein the first direction is the extension direction of the corresponding main surface, the turning point of the extension path in the first direction coincides with the thickness change position of the corresponding main surface, and the second direction is perpendicular to the first direction, and the end point of the extension path in the second direction is the shallowest position of the vertical side main surface;

[0096] Step S14: generating a plurality of cuboids according to the extended boundary, and the plurality of cuboids are spliced at the turning points to obtain the internal shell skeleton.

[0097] In step S11, the server first analyzes the shape contour of the target shell to screen out two opposite main surfaces. The vertical distance between the two opposite main surfaces is the wall thickness of the target shell, and can completely cover the main extension range of the shell in this direction, rather than a small local surface. The selection of the opposite main surfaces is flexible, and can be the upper and lower surfaces, left and right surfaces, or front and back surfaces of the shell, which is determined by the structural layout of the target shell.

[0098] Figure 4 The figure is a schematic diagram of the planar structure of the target shell. The outer frame part is the shape contour of the target shell. In the embodiment of the present application, the upper and lower surfaces are taken as the two opposite main surfaces.

[0099] In step S12, for each main surface, the server sets a skeleton reference line along the main extension direction of the main surface in the target shell. The skeleton reference line is in a straight line form and completely located within the contour boundary of the target shell. Figure 4 In the embodiment, the main surfaces are the upper and lower surfaces in the horizontal direction, and the skeleton reference lines are two horizontal straight lines respectively adhering to the upper and lower surfaces.

[0100] The skeleton reference line serves as the reference axis of the extension of the internal shell skeleton. The straight line form can provide a stable and unified extension direction, avoiding the uncertainty of the extension path caused by the curved form. The skeleton reference line reflects the trend of the target shell in the extension direction, such as the length direction or the width direction, so that the extension of the subsequent internal shell skeleton always adheres to the overall structure of the shell. For example, if the main surface extends in the horizontal direction, the skeleton reference line is set in the horizontal direction, serving as the scale of the extension of the internal skeleton in this direction, and ensuring that the extension direction of the skeleton reference line is consistent with the main structure trend of the target shell.

[0101] In step S13, the server respectively takes the two skeleton reference lines as the reference to extend each skeleton reference line in the target shell along the first direction and the second direction, to construct the boundary that adapts to the internal space of the shell.

[0102] The first direction is the extension direction of the main surface, such as the horizontal direction. In this direction, the inflection point of the extension path of the skeleton reference line must coincide with the position of the thickness change of the corresponding main surface. Since the shell has thickness changes due to functional requirements, such as reinforcing ribs, grooves, protrusions and other designs, the setting of the inflection point can make the extension boundary of the skeleton reference line accurately adhere to these shape changes, ensuring the adaptability of the boundary of the internal shell skeleton to the shape contour of the target shell. For example, when the surface of the shell has reinforcing ribs, the extension path will form an inflection point at the edge of the reinforcing ribs, so that the skeleton boundary wraps around the structure of the reinforcing ribs.

[0103] From Figure 4As can be seen from the drawings, the skeleton reference line does not remain a straight line during its extension in the horizontal direction, but adjusts its path synchronously with the thickness changes of the main surface. When the main surface has a position with increased or decreased wall thickness due to design requirements, the skeleton reference line naturally forms a turning point at this position, and the spatial position of the turning point completely coincides with the boundary of the thickness change of the main surface, without deviation. The purpose of this precise alignment of the turning point with the position of the thickness change is to enable the extension path of the skeleton reference line to completely adapt to the actual shape of the corresponding main surface. Where the main surface is thickened or thinned, the skeleton reference line adjusts its direction accordingly, neither exceeding the boundary of the shell nor lagging behind the shape change of the shell, and always keeping consistent with the contour trend of the main surface. In addition, the skeleton reference line is the boundary reference for constructing the internal shell skeleton. When the skeleton reference line closely follows the shape of the main surface, the boundary of the internal shell skeleton formed based on the extension of the skeleton reference line can naturally best fit the external contour shape of the target shell.

[0104] From Figure 4 As can be seen from the drawings, a continuous horizontal line segment is formed between every two adjacent turning points. The server extends the turning points on the skeleton reference line as the dividing line of the cuboid, and these turning points themselves are the boundaries of the shape change of the main surface of the target shell, such as the transition between the convex and flat areas of the main surface. The actual length of this line segment can be directly mapped as the length of a single cuboid. In short, the turning points divide the complex shell into multiple regular length segments, each corresponding to a cuboid. This segmentation method not only preserves the change characteristics of the shell contour, but also provides clear size references for subsequent formation of a regular internal shell skeleton through cuboid assembly.

[0105] The second direction is perpendicular to the first direction, for example, perpendicular to the front-to-back direction of the paper. In this direction, the end point of the extension path of the skeleton reference line is the shallowest position of the vertical side main surface. The shallowest position refers to the minimum size of the target shell in this direction. This can strictly constrain the extension range of the skeleton reference line, avoiding exceeding the actual contour of the target shell. If the extension end point exceeds the shallowest position, the skeleton reference line will extend outside the target shell, resulting in the generated internal shell skeleton exceeding the range of the target shell. If the shallowest position is not reached, a large amount of voids will be formed between the generated internal shell skeleton and the target shell, affecting the maximum adaptability of the internal shell skeleton to the shell contour. When the skeleton reference line accurately extends to the shallowest position, the extension boundary formed thereby can closely fit the inner wall contour of the target shell in the second direction, neither breaking through the external boundary of the target shell nor missing the internal space, and finally constructing a boundary that is highly adaptable to the internal space of the target shell.

[0106] The extension distance of the skeleton reference line in the second direction is the width of the cuboid generated subsequently. Since all skeleton reference lines extend to the shallowest position, the extension distance of the skeleton reference line in the second direction at each position remains consistent, ensuring that the widths of all cuboids are uniform, avoiding gaps or overlaps during subsequent splicing due to differences in width, and laying the foundation for the regular splicing of the internal shell skeleton.

[0107] In step S14, the size of each cuboid is set to match the structural features of the target shell. Specifically, the length of the cuboid is determined by the distance between two adjacent inflection points in the first direction, ensuring that each cuboid completely covers a consistent structural region of the target shell. The height of the cuboid is determined by the vertical distance between the two skeleton reference lines. Since the shape of the target shell's main surface at different inflection points differs, such as some areas having a convex main surface and some areas having a flat main surface, the vertical distance between the two skeleton reference lines at different inflection points is different, and thus the height of different cuboids is adjusted accordingly. The width of the cuboid corresponds to the second direction, and all position points of the skeleton reference line extend in the second direction according to a unified standard, i.e., to the shallowest position of the vertical side main surface of the target shell. Therefore, the widths of all cuboids remain consistent, avoiding gaps during subsequent splicing. Through this size logic, multiple cuboids with regular shapes but different heights are gradually formed.

[0108] These cuboids are seamlessly spliced at the inflection points. Specifically, one side of each cuboid near an inflection point is completely aligned with one side of the adjacent cuboid near the same inflection point, without gaps or overlaps, forming a complete internal shell skeleton. The final constructed internal shell skeleton retains the regularity of the cuboid structure, which greatly simplifies subsequent finite element segmentation operations and avoids problems such as mesh distortion and poor cell quality that often occur with irregular structures. On the other hand, by combining cuboids of different heights, the irregular shape of the target shell is accurately adapted, such as the convex part of the main surface corresponding to a cuboid with a greater height and the flat part of the main surface corresponding to a cuboid with a smaller height, achieving the adaptation of the regular internal shell skeleton to the actual contour shape of the target shell.

[0109] In the present application, the target shell contour is first analyzed and two opposite main surfaces are screened out, and then a straight skeleton reference line is set for each main surface along the main extension direction, which not only provides a stable and unified extension direction in the form of a straight line, but also makes the skeleton reference line fit the overall structure trend of the shell, laying a foundation for the reference axis of the internal shell skeleton; then the inflection point is accurately aligned with the thickness change position of the main surface when the skeleton reference line extends along the first direction, and extends to the shallowest position of the vertical side main surface along the second direction, which not only ensures that the extension boundary closely fits the irregular shape of the target shell and does not exceed the actual contour, but also clearly defines the length and width size basis of the subsequent cuboid, realizing the accurate adaptation of the internal space of the shell; finally, the inflection point is taken as the cuboid boundary line, the length, width and height are set, and the seamless splicing of the cuboid is realized, which not only retains the regular structure of the cuboid to simplify the subsequent finite element segmentation and avoid mesh quality problems, but also accurately adapts to the irregular contour of the shell through the combination of cuboids of different heights.

[0110] In step S12, determining a skeleton reference line for each main surface respectively includes: obtaining a minimum mesh wall thickness and a wall thickness coefficient in finite element analysis, wherein the wall thickness coefficient is used to indicate the number of layers of the minimum mesh wall thickness contained between the skeleton reference line and the corresponding main surface; taking the minimum distance between the two opposite main surfaces as the local shell wall thickness; determining the target distance between the skeleton reference line and the corresponding main surface according to the minimum mesh wall thickness, the wall thickness coefficient and the local shell wall thickness; drawing a straight line as the skeleton reference line along the extension direction of the corresponding main surface at the internal position of the target shell maintaining the target distance with the main surface.

[0111] In the present application, two key parameters in the finite element analysis scenario need to be determined first: the minimum mesh wall thickness and the wall thickness coefficient, which jointly determine the spacing logic between the reference line and the main surface.

[0112] The minimum mesh wall thickness t refers to the minimum size of the mesh element in the wall thickness direction that can be divided under the current finite element analysis requirements and the constraints of the shell material properties to ensure calculation accuracy and efficiency.

[0113] The wall thickness coefficient k refers to the number of layers of the minimum mesh wall thickness that needs to be contained between the skeleton reference line and the corresponding main surface to ensure that the mesh density in the wall thickness direction meets the analysis requirements, and is a positive integer such as 1, 2, 3, etc. Its core role is to avoid calculation errors caused by too few layers of mesh in the wall thickness direction. For example, when analyzing the bending deformation of the shell, at least 2 layers of mesh are needed to capture the stress gradient of the cross section, so the wall thickness coefficient k is 2; if higher accuracy is required, for example, to analyze local plastic deformation, k can be 3 to ensure that the mesh around the reference line can fully reflect the stress distribution law, and k is a variable parameter.

[0114] The local shell wall thickness r refers to the minimum vertical distance between the two opposite main surfaces of the target shell.

[0115] The target distance d between the skeleton reference line and the corresponding main surface needs to meet two core conditions: one is that the distance can completely accommodate the minimum grid wall thickness of k layers, and the other is that the remaining space between the two opposite skeleton reference lines is sufficient to accommodate the basic grid.

[0116] The following formula needs to be met: r-2d≥t, d=kxt.

[0117] d=kxt ensures that there is enough space between the skeleton reference line and the corresponding main surface to accommodate k layers of minimum grid units, avoiding the problem that the main surface and the skeleton reference line coincide to cause no grid analysis of the features on the contour. Since the target shell has two opposite main surfaces corresponding to two opposite skeleton reference lines, an intermediate area will be formed between the two skeleton reference lines, which must be able to accommodate at least 1 layer of minimum grid wall thickness t, i.e. r-2d≥t.

[0118] Therefore, the formula is to determine the basic value of d according to the number of single-side grid layers first, and then verify whether this value can be used by the overall space constraint; if the constraint is not met, k needs to be adjusted until the two conditions are met simultaneously.

[0119] After determining the target distance d, the position of the skeleton reference line is obtained by offsetting the distance d from the corresponding main surface to the interior of the shell, and then drawing a straight line at this position, which is the skeleton reference line.

[0120] In a second embodiment.

[0121] Step S21: Determine the two opposite main surfaces of the target shell, which are a pair of opposite surfaces constituting the thickness direction of the target shell;

[0122] Step S22: Identify the thickness change position of each main surface as the inflection point of the extension path of the corresponding skeleton reference line;

[0123] Step S23: Divide the target shell into multiple independent blocks with the inflection points as the dividing points;

[0124] Step S24: Determine a sub-skeleton reference line for each of the two opposite sub-main surfaces of each independent block, the sub-skeleton reference line being a straight line arranged along the extension direction of the corresponding sub-main surface, and the sub-skeleton reference line being located within the contour boundary of the independent block;

[0125] Step S25: Extend the sub-skeleton reference line in each independent block along the second direction, wherein the second direction is the vertical direction of the corresponding sub-main surface, and the end point of the extension path of each sub-skeleton reference line is the shallowest position of the vertical side main surface of the target shell;

[0126] Step S26: According to the boundary extended in each independent block, the corresponding cuboid is generated respectively, and the cuboids are spliced at the inflection points to obtain the internal shell skeleton of the target shell.

[0127] In step S21, the server first analyzes the contour shape of the target shell, and selects two opposite main surfaces therefrom. The vertical distance between the two opposite main surfaces is the wall thickness of the target shell. The two opposite main surfaces are parallel to each other in the wall thickness dimension, and can completely cover the main extension range of the shell in this direction, rather than a small area surface. The selection of the opposite main surfaces is flexible, and can be the upper and lower surfaces, the left and right surfaces, or the front and back surfaces of the shell, which is determined by the structural layout of the target shell.

[0128] In step S22, the server identifies the thickness change positions on each opposite main surface, and directly sets these positions as the inflection points of the extension path of the subsequent skeleton reference line. The thickness change positions are not randomly distributed, but are a direct manifestation of the functional design of the shell, such as the raised edge of the reinforcing rib, which is manifested at the transition from the flat to the thickened main surface, and the recessed boundary of the mounting groove, which is manifested at the transition from the thickened to the thinned main surface, which are structural boundaries where the main surface shape and thickness change.

[0129] In step S23, the server uses the inflection points as the segmentation reference line, and accurately cuts along the three-dimensional contour of the target shell to divide the target shell into multiple independent blocks. Each independent block corresponds to a shell region with uniform internal structure and stable morphological characteristics, such as a region containing a single reinforcing rib, which becomes an independent block, an adjacent region without reinforcing ribs, which becomes another independent block, and a local raised region with mounting holes, which forms a new independent block.

[0130] In step S24, for each independent block after segmentation, a relative sub-main surface needs to be set inside it, i.e. the two surfaces of each independent block in the core extension direction, which are the relative sub-main surfaces of the independent block. Then, the server sets a straight sub-skeleton reference line for each sub-main surface, and the sub-skeleton reference line must be completely located within the contour boundary of the independent block and cannot exceed the block range. Figure 5 The figure shows that the target shell is divided into five independent blocks by the dashed line, and each independent block has a corresponding sub-skeleton reference line generated inside it.

[0131] The setting process of the sub-skeleton reference line is similar to the setting process of the skeleton reference line in the first embodiment described above, which will not be repeated here.

[0132] In step S25, the server controls the two sub-skeleton reference lines in each independent block to extend in the second direction within the block, and the endpoints of all the extension paths of the sub-reference lines are uniformly set as the shallowest position of the vertical side main surface of the target shell. The vertical side main surface here refers to the end surface of the target shell in the second direction, such as the front side and the rear side of the shell when the second direction is the front-rear direction, and the shallowest position is the minimum dimension of the target shell in the second direction, which can be accurately positioned through the size scanning of the target shell to ensure that the extension path does not exceed the actual contour of the target shell. The extension distance of the sub-reference line in the second direction is the width of the cuboid generated subsequently. Since the endpoints of the sub-reference lines of all the blocks are the same shallowest position, the widths of all the cuboids remain the same, which eliminates the size obstacle for subsequent block splicing.

[0133] Optionally, the extension of each sub-skeleton reference line in the second direction can also be determined based on the minimum dimension of the independent block in the second direction, so that the extension of each sub-skeleton reference line in the second direction is not completely the same, the width of the cuboid formed is not completely the same, and the internal shell skeleton formed can better fit the contour of the target shell.

[0134] In step S26, for each independent block, the server generates a corresponding cuboid according to the boundary of the sub-reference line extended within the block: the length of the cuboid is determined by the distance between the inflection points of the independent block in the first direction, to ensure that the structure range of the independent block can be completely covered; the height is determined by the vertical distance between the two sub-skeleton reference lines within the independent block, since the shapes of different independent blocks are different, the distance between the sub-skeleton reference lines is different, and the height of the cuboid is adjusted accordingly to fit the local shape of the block; and the width is the extension distance in the second direction. If the extension distance is determined according to the shallowest position of the vertical side main surface of the target shell, the widths of all the cuboids remain the same, and if the extension distance is determined according to the shallowest position of the vertical side main surface of each independent block, the widths of all the cuboids are not completely the same. Through this size logic, a regular internal shell sub-skeleton is generated for each independent block.

[0135] Subsequently, the inflection points of these cuboids are set as the splicing positions to achieve seamless splicing, and the internal shell skeleton finally constructed retains the regularity of the cuboid structure, greatly simplifies the subsequent finite element segmentation operation, and avoids problems such as mesh distortion and poor unit quality that are prone to occur in irregular structures. On the other hand, through the logic of generating cuboids by block and then splicing, the irregular shape of the target shell is accurately fitted, such as the high cuboid corresponding to the reinforcing rib block and the low cuboid corresponding to the main block, to achieve a high degree of fit between the regular structure and the actual contour of the target shell.

[0136] Compared with the logic of extending the reference line first and then dividing the cuboid in the first embodiment, the first block design of this embodiment can better adapt to complex shells. When the target shell integrates multiple functional structures, such as containing multiple sets of reinforcing ribs, or containing mounting slots of different specifications, direct overall processing may not be able to adapt to the local area, and after being disassembled into independent blocks, each independent block can be designed independently according to its own shape. The internal shell skeleton greatly improves the adaptation accuracy. For example, on the side of the automobile gearbox shell, the side area with the heat dissipation rib and the top area with the bolt hole are divided into different blocks, and the internal shell skeleton of each block can respectively match the shape requirements of the heat dissipation structure and the connection structure, avoiding the loss of details caused by uniform processing.

[0137] In the embodiments of the present application, the contour features include but are not limited to: reinforcing rib intersection features, spring disc features, subframe mounting point features, shock absorber mounting point features, logos, fasteners, 3D holes, 2D holes, round corners, right angles, and flanges.

[0138] Among them, the casting part is generally designed to be relatively thick and has different shapes at the intersection of the reinforcing ribs. The shape of the reinforcing rib intersection feature has multiple types, including: cylindrical features located on a plane or on an arc with a large angle; cylindrical features located at the intersection of features; semi-cylindrical features; X-shaped features.

[0139] The spring disc feature is the support seat of the spiral spring or air spring. In the automobile suspension system, the spring is a key component that bears the weight of the vehicle body and buffers the impact. It needs a flat and solid platform to place and fix. Different types of suspensions have different positions of the spring disc on the vehicle body. For MacPherson suspension, the installation position is located at the top of the front shock tower of the vehicle body. It is a reinforced metal ring connected to the shock absorber assembly and the spring through bolts; for multi-link suspension mounting points, it is usually a specially reinforced area located below the vehicle body floor.

[0140] The subframe mounting point feature is the position of the subframe. The subframe is an independent structural member, usually made of stamped and welded steel plate or aluminum casting. It integrates key components of the chassis (such as engine, transmission, control arm, stabilizer bar, and steering machine) on itself, and then installs them on the vehicle body through the subframe mounting point. The subframe mounting point will bear the weight of large components such as the engine and transmission, as well as the impact and vibration from the road.

[0141] Suspension mounting point feature, is a specific location for fixing and connecting the suspension assembly, usually composed of bolt holes, brackets, bases and other structures; Suspension mounting point transmits all impact forces in vehicle driving, such as bumps, corners, brakes, these impact forces will be transmitted to the vehicle body through the tires, suspension, suspension mounting point is the key structure to ensure the handling and safety of the vehicle, the mounting point needs to withstand high frequency vibration, if the design is poor, it is easy to appear metal fatigue and cracking problem.

[0142] Logos, a unique graphic, symbol or text design, is a trademark or logo of a company, organization or product, Logos will not affect the stiffness or strength of the structure, when normal meshing, it is necessary to remove it to ensure the correctness of the mesh flow.

[0143] Fastener, a kind of fastener commonly used to connect one or more parts together, commonly used are bolts, rivets, screws, SPR (Self-Piercing Rivet, self-piercing rivet), FDS (Flow Drill Screw, flow drill screw) and the like.

[0144] 3D hole, a hole that appears when meshing solid elements, this hole can be a mounting hole, a weight reduction hole or a wire harness hole.

[0145] 2D hole, a hole that appears when meshing shell elements, this hole can be a mounting hole, a weight reduction hole or a wire harness hole.

[0146] Round corner, refers to the intersection of two planes or curved surfaces as an arc-shaped surface, round corners can eliminate sharp corners, from an engineering point of view, round corners can disperse stress and avoid stress concentration at right angles, thereby enhancing the durability and preventing cracking of the product.

[0147] Right angle, refers to the intersection of two planes or curved surfaces as a plane.

[0148] Flange, refers to the overlapping part of two parts.

[0149] Each type of contour feature corresponds to a finite element feature model, i.e. a discretizer module, the discretizer module is composed of seven parts, first discretizer, second discretizer, third discretizer, logos eliminator (logo eliminator), bolt connection generator, washer generator (washer generator) and flange edge processor, the functions of each part in the discretizer module are as follows.

[0150] First discretizer: discretize the feature into hexahedral elements.

[0151] Second discretizer: discretize the feature into first-order tetrahedral elements.

[0152] Third discretizer: Discretize features into first order tetrahedron elements and ensure at least 2 rows of mesh in thickness direction.

[0153] Logos eliminator: Close the face by hole fill command after Logos is deleted.

[0154] Bolt generator: Generate a kind of element to replace the bolt.

[0155] Washer generator: Generate a washer with the same size as the contact area of the bolt for the bolt joint.

[0156] Flange edge processor: Process the flange edge into 3 rows or more of mesh to ensure the accuracy of the finite element result.

[0157] In which, the process of dividing mesh for each profile feature is as follows.

[0158] Cylindrical features with the cylinder lying on a plane or on a large angle arc surface are discretized by the first discretizer.

[0159] Cylindrical features at the intersection of features, semi-cylindrical features and X-shaped features are discretized by the second discretizer.

[0160] Spring disc features, auxiliary frame mounting point features, and shock absorber mounting point features are discretized by the third discretizer.

[0161] Logos, call the Logos eliminator to delete the Logos.

[0162] Fastener, call the bolt generator to generate a bolt joint.

[0163] 3D hole, call the washer generator to generate a circle with the same size as the contact surface of the bolt.

[0164] 2D hole, call the washer generator to generate a circle with the same size as the contact surface of the bolt.

[0165] Rounded corner, automatically determine the number of rows of mesh to be divided on the rounded corner according to the size of the rounded corner. If the size of the rounded corner is too small, the rounded corner will be deleted.

[0166] Right angle, automatically determine the number of rows of mesh to be divided on the right angle according to the size of the right angle. If the size of the right angle is too small, the right angle will be deleted.

[0167] Flange, call the flange edge processor to generate 3 rows of shell element mesh on the flange edge.

[0168] In the present application, for repetitive or similar shape profile features, it is not necessary to repeat the calculation and debugging work of mesh division for each shape profile feature. By directly calling the mature mesh division method stored in the database, the discrete processing of similar features can be quickly completed, greatly reducing the redundant calculation links, shortening the overall cycle of finite element preprocessing, and realizing the leap-forward improvement of mesh generation efficiency. The mesh division schemes for various features stored in the database are high-quality schemes verified by a large number of engineering practices, covering key technical points such as element type selection, mesh density control, and transition area processing. Based on these mature schemes to carry out mesh division, it can ensure that whether for different structural models or by different operators to carry out work, high-standard meshes with stable quality can be output, effectively avoiding mesh quality problems caused by human operation differences or improper scheme design.

[0169] In the present application, the feature recognition module identifies the typical features of the casting part, such as fillets, bolts, and reinforcement rib intersections. Based on this recognition result, the server queries the pre-stored optimal mesh division scheme in the data storage and update module, and simultaneously calls the mesh division module containing seven discretizers and processors to perform mesh division. Then, the mesh flow correction module corrects the division results to solve the mesh flow rationality problem. Next, the export module outputs the final mesh model. At the same time, if new features or new optimal division schemes are found in the export results, these new contents can be stored into the database through the two updating methods of the data storage and update module. When the server subsequently identifies the same or similar features again, it can directly call the optimized scheme in the updated database to realize iterative optimization, thereby forming a complete adaptive mesh division closed loop.

[0170] Among them, the updating method of the data storage and update module includes two kinds. One is to enter the model browser and add new feature data to the database. When adding, the corresponding feature type and the corresponding mesh division need to be selected. The other is to classify similar features, and then select the features to be added to the database and add them.

[0171] As an optional implementation, after obtaining the shell skeleton discrete model corresponding to the internal shell skeleton and the shape profile discrete model corresponding to each shape profile feature, the method further comprises:

[0172] Step S31: identifying specific areas to be corrected in the shell skeleton discrete model and each shape profile discrete model, the specific areas including circular face areas, rectangular face areas, transition areas from narrow to wide, and transition areas from wide to narrow;

[0173] Step S32: correcting the original mesh flow of the circular face area to an O-shaped mesh flow to adapt to the profile shape of the circular face;

[0174] Step S33: Adjust the grid node distribution of the quasi-rectangular face area to keep the number of grid nodes on the four boundaries of the quasi-rectangular face consistent.

[0175] Step S34: Perform transition processing on the grid of the narrow-to-wide transition area and the wide-to-narrow transition area by combining quadrilateral grids and triangular grids to eliminate grid distortion at the wide-to-narrow transition.

[0176] Step S35: After completing the grid flow correction of each specific area, output the corrected shell skeleton discrete model and the corrected discrete model of each contour profile.

[0177] In step S31, the server accurately locates the specific areas that need to be optimized in the shell skeleton discrete model and the discrete model of each contour profile by combining geometric feature analysis and grid quality pre-detection. These specific areas are prone to quality defects due to the particularity of their geometric shapes, and if not corrected, they will directly affect the accuracy of subsequent finite element calculations. Specifically, first, through the geometric topology analysis function of the finite element pre-processing software, four typical areas are selected from the discrete model.

[0178] Circular face areas are commonly found in the circular bearing planes of spring discs, the end faces of 3D and 2D holes, and the circular connecting faces of flanges. The default grid flow of this type of area is prone to form small-sized units in the center due to the uniform curvature of the ring-shaped contour.

[0179] Quasi-rectangular face areas are mostly rectangular bases of sub-frame mounting points and rectangular end faces of reinforcing ribs, etc. The length ratio of their sides is close to that of a rectangle, but the default grid may have uneven node distribution due to the length difference of the boundaries.

[0180] Narrow-to-wide and wide-to-narrow transition areas are commonly found in the connecting parts of reinforcing ribs from thick sections to thin sections, and the transition sections of shock absorber mounting points from wide brackets to narrow bolt holes. The default grid of this type of area is prone to form elongated or twisted units due to rapid size changes.

[0181] The server performs grid quality scanning on the initially identified areas to verify whether there are defects through indicators such as unit distortion rate and node connectivity. For example, if the length-to-width ratio of the center unit of a circular face exceeds a certain threshold, the difference between the number of nodes on adjacent sides of a quasi-rectangular face exceeds a certain value, or the unit distortion rate of a transition area exceeds a certain proportion, it is determined that the area needs to be corrected, ensuring that the identification result is neither missed nor misjudged.

[0182] In step S32, the server reconstructs the original grid stream into an O-shaped grid stream to better adapt to the contour shape of the circular face for the grid defect of the circular face region. The original grid stream such as a radiation type or a C type is prone to converge into small triangular units at the center of the circular face, resulting in grid distortion and affecting the calculation accuracy, while the O-shaped grid stream is nested and diffused in a ring shape with the center of the circular face as the reference, which can effectively avoid such problems.

[0183] In the correction process, first, the center coordinates, radius size of the circular face, and the target grid size of the region are obtained through geometric measurement. The target grid size needs to match the surrounding model grid size to prevent new calculation errors due to density mutation. Then, a series of concentric circles are generated with the center as the origin as the grid reference line, and the circle spacing is set according to the target grid size. At the same time, radial auxiliary lines are generated through the center, and the number of radial auxiliary lines matches the number of ring lines to ensure that the divided units are close to regular quadrilaterals. Finally, the quadrilateral units are divided along the intersection of the ring lines and the radial auxiliary lines, and the unit quality is verified to ensure that the aspect ratio of each unit is controlled within a fixed threshold and there is no distortion or overlap. For example, after the correction of the circular bearing surface of the spring disc, the grid is uniformly distributed from the center to the edge, completely fitting the circular contour, avoiding the small unit defect of the original grid at the center.

[0184] In step S33, the server adjusts the grid node distribution of the quasi-long rectangular face region to keep the number of grid nodes on the four boundaries of the quasi-long rectangular face consistent, solving the unit distortion problem caused by the default grid due to the mismatch of the number of nodes. If the number of nodes on the four boundaries of the quasi-long rectangular face is different, the grid units of adjacent boundaries will be forced to splice, which is prone to form triangular gaps or elongated quadrilaterals, which will reduce the accuracy of finite element calculation.

[0185] First, the actual lengths of the four sides of the quasi-long rectangular face are measured, and the theoretical number of nodes of each side is calculated based on the target grid size of the overall model. The theoretical number of nodes is calculated by dividing the length of each side by the target grid size and adding 1, ensuring that the first and last nodes of each side can accurately connect with the nodes of the adjacent boundary. Then, the number of nodes on the four sides is unified based on the least common multiple or the maximum adaptive value of the theoretical number of nodes. For example, if the theoretical number of nodes of the long side is 11 and the short side is 6, the number of nodes of the short side can be adjusted to 11 or the number of nodes of the long side can be adjusted to 6, depending on the surrounding grid density and calculation accuracy requirements. After the number of nodes is unified, the grid is re-divided to ensure that the unit shape is close to a regular quadrilateral, the aspect ratio is controlled within 1.5, and the grid density of the entire quasi-long rectangular face is uniform. For example, the rectangular base of the sub-frame mounting point, after adjustment, the number of nodes on the four sides is consistent, and the grid is arranged neatly, which can more realistically simulate the bearing stress of the base.

[0186] In step S34, the server performs mesh transition processing on the transition region from wide to narrow and from narrow to wide by combining quadrilateral meshes and triangular meshes to eliminate mesh distortion at the wide-narrow transition. Due to the rapid change in size, the use of only quadrilateral meshes in such a transition region can cause the wide section mesh to be forcibly contracted or the narrow section mesh to be forcibly stretched, resulting in an elongated element with a length-width ratio exceeding 5, which can cause the calculation result to be distorted or even not to converge.

[0187] In the processing, the range of the transition region is first defined, and the start point and end point of the wide-narrow transition are determined. The length of the transition region is usually set to 2-3 times the size difference between the wide section and the narrow section to ensure a smooth transition. Then, quadrilateral meshes are laid in the non-transition core areas of the wide section and the narrow section. The wide section is divided according to the target mesh size, for example, a 20mm wide section is divided into 2 columns of quadrilateral elements, each column being 10mm wide. The narrow section is divided according to the appropriate size, for example, an 8mm narrow section is divided into 1 column of quadrilateral elements. Then, 1-2 rows of triangular meshes are used to connect the quadrilateral meshes of the wide section and the narrow section in the transition core area to fill the difference in the number of columns between the wide section and the narrow section, achieving a smooth transition. For example, the wide-narrow transition section of the shock absorber mounting point bracket, through triangular meshes, 4 columns of quadrilateral elements of the wide section are smoothly merged into 2 columns of quadrilateral elements of the narrow section, completely eliminating the original elongated elements. Finally, the quality of the elements in the transition region is checked to ensure that the length-width ratio of all elements does not exceed 3, the twist degree does not exceed 10°, and there are no overlapping or missing elements.

[0188] In step S35, the service performs a comprehensive mesh quality final inspection on the corrected shell skeleton discrete model and each contour discrete model. The core indicators include that the element distortion rate, the length-width ratio, and the shell element twist degree all do not exceed their respective threshold values, and that there are no isolated nodes in the node connectivity, ensuring that there are no missed quality defects. Then, according to the format requirements of the finite element analysis software, the corrected model is exported in standardized formats such as.inp,.cdb, or.hwf, while retaining the element type, material property association, and the relative position relationship of each discrete model, ensuring the compatibility of the model between different software. Finally, a mesh quality report is generated to mark the key data comparison before and after the correction of each specific area, providing quality traceability for subsequent analysts, and finally outputting the corrected shell skeleton discrete model and each contour discrete model.

[0189] In this application, the corrected mesh flow is more suitable for the geometric contour. The O-shaped mesh of the circular surface can accurately reflect the annular stress distribution of the spring disc and the hole end surface. The uniform node distribution of the rectangular-like surface avoids false stress concentration in the mounting point of the sub-frame, flange and other load-bearing areas, and the stress transfer simulation is closer to the actual situation. The smooth mesh transition in the wide-narrow transition area ensures the accurate calculation of the mechanical properties of the reinforcing ribs, shock absorber brackets and other force transmission paths, providing reliable data support for fatigue strength checking and stiffness analysis, and greatly reducing the calculation errors caused by mesh defects.

[0190] As an optional implementation, in step 205, the shell skeleton discrete model and each contour profile discrete model are combined, and the combined discrete model is output as a standard finite element mesh, including:

[0191] Step S41: Coordinate systems of the shell skeleton discrete model and each contour profile discrete model are unified to a global coordinate system of the target shell;

[0192] Step S42: Based on the position association relationship between the internal shell skeleton and each contour profile feature, each contour profile discrete model is mapped to the corresponding spatial region based on the shell skeleton discrete model as the basic framework;

[0193] Step S43: Coordinate alignment is performed on the overlapping grid nodes of adjacent discrete models, and fusion processing is performed on the adjacent grid contact surfaces to form an integrated combined discrete model;

[0194] Step S44: The combined discrete model is configured with uniform physical property parameters, and after marking the finite element analysis constraint identifier, the combined discrete model is converted into a standard finite element mesh.

[0195] In step S41, in the early discrete model generation stage, the internal shell skeleton and different contour profile features may be modeled based on their respective local coordinate systems. The origin positions and axis system directions of these local coordinate systems may differ. If direct integration is performed, it will inevitably lead to position misplacement of the discrete model, and the overall discrete model that conforms to the actual structure of the target shell cannot be formed. Therefore, the server performs coordinate conversion on the shell skeleton discrete model and each contour profile discrete model through a coordinate transformation algorithm, and converts the local coordinate parameters of all discrete models into global coordinate parameters.

[0196] In step S42, based on the unified global coordinate system, the server accurately maps each contour profile discrete model to the spatial region corresponding to the shell skeleton discrete model according to the position association relationship between the internal shell skeleton and each contour profile feature, realizing the spatial position matching of the skeleton and the contour. This position association relationship is derived from the comprehensive analysis of the target shell structure in the early stage, which clearly defines the relative positions of each contour profile feature in the overall structure of the shell, and is the core basis for ensuring the accuracy of the mapping.

[0197] During the operation, firstly, the pre-stored position correlation data is called to determine the target space position of each contour discrete model in the global coordinate system based on the shell skeleton discrete model as the basic framework. During the mapping process, the relative offset, angle and other parameters in the position correlation relationship need to be strictly followed to ensure that the contour discrete model can be accurately mapped to the specified area of the internal shell skeleton. After the mapping is completed, position compliance inspection needs to be carried out, on the one hand, to confirm that each contour discrete model does not exceed the global contour range of the target shell, and on the other hand, to check whether there is non-design spatial overlap between different contour discrete models. If problems are found, the mapping parameters are adjusted in time until all contour discrete models are in the correct spatial position.

[0198] In step S43, after the completion of coordinate unification and space mapping, the connection parts of adjacent discrete models may have problems such as misalignment of grid nodes, gaps in contact surfaces, etc. These problems will cause the stress transmission to be interrupted during finite element calculation, seriously affecting the accuracy of the analysis results. The server first aligns the coordinates of overlapping grid nodes. Through the grid node search algorithm, it identifies the overlapping nodes with similar spatial positions between adjacent discrete models. Taking the node coordinates of the shell skeleton discrete model as the benchmark, the coordinates of the corresponding overlapping nodes of the contour discrete model are adjusted to make the coordinates of the two completely consistent, ensuring the connection continuity at the node level. Then the adjacent grid contact surfaces are fused. For the contact surfaces of adjacent models, the contact surface grids of the two discrete models are merged into a single continuous grid surface, and the duplicate grid elements are deleted to eliminate the gap between the contact surfaces. After the fusion operation is completed, the grid quality of the fusion area needs to be checked to ensure that the shape of the fused grid elements meets the requirements of finite element analysis and avoids introducing new grid deformation problems due to fusion operation.

[0199] In step S44, the server configures uniform physical property parameters and marks finite element analysis constraint identifiers for the combined discrete model, finally converts the integrated combined discrete model into a standard finite element grid that can be directly used for finite element analysis, and completes the key conversion from the structure model to the analysis model. This step needs to ensure that the model not only has a continuous structure shape, but also can accurately reflect the mechanical properties and analysis requirements of the target shell.

[0200] The physical property parameters include the elastic modulus, Poisson's ratio, density and the like of the material, and ensure that the mechanical response of the model is consistent with the actual shell. If the target shell has a multi-material region, the physical parameters of the corresponding materials are configured according to the region, and the transition of the properties at the boundary of different material regions is ensured to be clear. The marking of the finite element analysis constraint mark is marked according to the specific analysis requirements, and the constraint position and constraint type are clearly marked on the discrete model, and the load application region is provided for the clear guidance of the constraint and load setting in the subsequent finite element analysis. Finally, the standard finite element grid conversion is performed, the combined discrete model is converted into the corresponding standard format according to the format requirements of the target finite element analysis software, and the key information such as the unit type, node number, physical property association and constraint mark needs to be completely retained in the conversion process. After the conversion is completed, the integrity and accuracy of the standard finite element grid are checked to ensure that the output grid can be directly used for subsequent finite element analysis calculation.

[0201] Based on the same technical concept, the application provides a finite element grid generation device for a target shell, as shown in Figure 6 The device comprises:

[0202] The acquisition module 601 is configured to acquire the contour shape of the target shell and a plurality of contour characteristics.

[0203] The generation module 602 is configured to determine two corresponding skeleton reference lines of the two opposite main surfaces of the target shell, and to generate an internal shell skeleton of the target shell according to the two skeleton reference lines, wherein the two opposite main surfaces comprise a first main surface and a second main surface, the distance between the first main surface and the second main surface refers to the maximum thickness of the target shell in the vertical direction of the first main surface, the internal shell skeleton is a regular structure formed by splicing at least one cuboid, and the internal shell skeleton maximizes the adaptation to the external contour shape of the target shell.

[0204] The matching module 603 is configured to match the internal shell skeleton and the finite element characteristic model corresponding to each contour characteristic from the preset database respectively.

[0205] The segmentation module 604 is configured to perform finite element segmentation by using the corresponding finite element characteristic model to obtain a shell skeleton discrete model corresponding to the internal shell skeleton and a contour discrete model corresponding to each contour characteristic.

[0206] The output module 605 is configured to combine the shell skeleton discrete model and each contour discrete model, and output the combined discrete model as a standard finite element grid.

[0207] Optionally, the generation module 602 is configured to:

[0208] determine the two opposite main surfaces of the target shell, and the two opposite main surfaces are a pair of opposite surfaces in the thickness direction of the target shell.​

[0209] In the target shell, a skeleton reference line is determined for each main surface, the skeleton reference line being a straight line arranged along the extension direction of the corresponding main surface, and the skeleton reference line being located within the contour boundary of the target shell;

[0210] Each skeleton reference line is taken as a reference to extend in the target shell along a first direction and a second direction, wherein the first direction is the extension direction of the corresponding main surface, the turning point of the extension path in the first direction coincides with the position of the thickness change of the corresponding main surface, and the second direction is perpendicular to the first direction, and the end point of the extension path in the second direction is the shallowest position of the vertical side main surface;

[0211] A plurality of cuboids are generated according to the extended boundary, the plurality of cuboids are spliced at the turning points, and an internal shell skeleton is obtained.

[0212] Optionally, the generating module 602 is specifically configured to:

[0213] Obtain the minimum grid wall thickness and the wall thickness coefficient in the finite element analysis, wherein the wall thickness coefficient is used to indicate the number of grid layers of the minimum grid wall thickness accommodated between the skeleton reference line and the corresponding main surface;

[0214] The minimum distance between the two opposite main surfaces is taken as the local shell wall thickness;

[0215] According to the minimum grid wall thickness, the wall thickness coefficient and the local shell wall thickness, the target distance between the skeleton reference line and the corresponding main surface is determined;

[0216] A straight line is drawn as a skeleton reference line along the extension direction of the corresponding main surface at a position inside the target shell that maintains the target distance with the main surface.

[0217] Optionally, the generating module 602 is specifically configured to:

[0218] The distance between the two adjacent turning points in the first direction is mapped as the length of a cuboid;

[0219] The extension distance in the second direction is mapped as the width of the cuboid;

[0220] The vertical distance between the two skeleton reference lines is mapped as the height of the cuboid;

[0221] According to the length, width and height of each cuboid, a cuboid is constituted, wherein the width of each cuboid is the same, and the length and the height are not completely the same.

[0222] Optionally, the generating module 602 is configured to:

[0223] Two opposite main surfaces of the target shell are determined, and the two opposite main surfaces are a pair of opposite surfaces in the thickness direction of the target shell.

[0224] identify the position of the thickness variation of each main surface as the inflection point of the extension path of the corresponding skeleton reference line;

[0225] segment the target shell into multiple independent blocks with the inflection point as the demarcation point;

[0226] determine a sub-skeleton reference line for each of the two opposite sub-main surfaces of each independent block, the sub-skeleton reference line being a straight line arranged along the extension direction of the corresponding sub-main surface, and the sub-skeleton reference line being located within the contour boundary of the independent block;

[0227] extend the sub-skeleton reference line in each independent block along a second direction, wherein the second direction is the vertical direction of the corresponding sub-main surface, and the end point of the extension path of each sub-skeleton reference line is the shallowest position of the vertical side main surface of the target shell;

[0228] generate a corresponding cuboid according to the extended boundary in each independent block, and obtain the internal shell skeleton of the target shell by corresponding splicing of the cuboids at the inflection points.

[0229] Optionally, the device is further used for:

[0230] identify the specific area to be corrected in the shell skeleton discrete model and each contour profile discrete model, the specific area including a circular surface area, a quasi-rectangular surface area, a narrow-to-wide transition area, and a wide-to-narrow transition area;

[0231] correct the original grid flow of the circular surface area to an O-shaped grid flow to adapt to the contour shape of the circular surface;

[0232] adjust the distribution of grid nodes of the quasi-rectangular surface area to keep the number of grid nodes on the four boundaries of the quasi-rectangular surface consistent;

[0233] perform transition processing on the grid of the narrow-to-wide transition area and the wide-to-narrow transition area in a combination of quadrilateral grid and triangular grid to eliminate the grid deformity at the wide-to-narrow transition;

[0234] after completing the grid flow correction of each specific area, output the corrected shell skeleton discrete model and the corrected contour profile discrete model.

[0235] Optionally, the output module 605 is specifically configured to:

[0236] unify the coordinate system of the shell skeleton discrete model and the coordinate system of each contour profile discrete model to a global coordinate system of the target shell;

[0237] Based on the position correlation relationship between the internal shell skeleton and each contour feature, each contour discrete model is mapped to a corresponding space region based on the shell skeleton discrete model as a basic framework.

[0238] Coordinate alignment is performed on the overlapping grid nodes of adjacent discrete models, and fusion processing is performed on the adjacent grid contact surfaces to form an integrated combined discrete model.

[0239] Uniform physical property parameters are configured for the combined discrete model, and after marking the finite element analysis constraint identifier, the combined discrete model is converted into a standard finite element grid.

[0240] As shown in Figure 7 The embodiment of the present application provides an electronic device, which comprises a processor 701, a communication interface 702, a memory 703 and a communication bus 704, wherein the processor 701, the communication interface 702 and the memory 703 complete mutual communication through the communication bus 704.

[0241] The memory 703 is used for storing a computer program.

[0242] In an embodiment of the present application, the processor 701 is used for executing the program stored in the memory 703, so as to realize the finite element grid generation method of the target shell provided in any one of the preceding method embodiments.

[0243] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps of the finite element grid generation method of the target shell provided in any one of the preceding method embodiments.

[0244] The device embodiments described above are only schematic, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0245] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software plus a general hardware platform, and of course, it can also be realized by hardware. Based on such understanding, the above technical solutions or the part that contributes to the related art can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a grid device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0246] It is to be understood that the terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The terms "comprises", "comprising", "includes", "including" and "has" are inclusive and therefore specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring their performance in the particular order

[0247] The above description is that of current embodiments of the application. Various modifications and changes can be made thereto without departing from the spirit and scope of the application as set forth. The scope of the application is not to be limited to the specific illustrative embodiments therein but only by the claims.

Claims

1. A method for generating a finite element mesh for a target shell, characterized in that, The method includes: Obtain the outer contour shape and various outer contour features of the target shell; The skeleton reference lines corresponding to the two opposing main surfaces of the target shell are determined, and the internal shell skeleton of the target shell is adaptively generated according to the two skeleton reference lines. The two opposing main surfaces include a first main surface and a second main surface. The distance between the first main surface and the second main surface refers to the maximum thickness of the target shell in the vertical direction of the first main surface. The internal shell skeleton is a regular structure formed by splicing at least one cuboid. The internal shell skeleton maximizes the adaptation to the external contour shape of the target shell. Match the finite element feature models corresponding to the internal shell skeleton and each external contour feature from the preset database; Finite element segmentation is performed using the corresponding finite element feature model to obtain the discrete model of the shell skeleton corresponding to the internal shell skeleton and the discrete model of the external contour corresponding to each external contour feature. The discrete model of the shell skeleton and each of the discrete models of the outer contour are combined, and the combined discrete model is output as a standard finite element mesh.

2. The method according to claim 1, characterized in that, Determining the skeleton reference lines corresponding to the two opposing main surfaces of the target shell, and adaptively generating the internal shell skeleton of the target shell based on the two skeleton reference lines includes: Two opposing main surfaces of the target housing are determined, wherein the two opposing main surfaces are a pair of opposing surfaces constituting the wall thickness direction of the target housing; Inside the target housing, a skeleton reference line is defined for each of the main surfaces. The skeleton reference line is a straight line set along the extension direction of the corresponding main surface and is located within the contour boundary of the target housing. Taking each of the skeleton reference lines as a reference, the extension is carried out inside the target shell along a first direction and a second direction. The first direction is the extension direction corresponding to the main surface, and the inflection point of the extension path in the first direction coincides with the thickness change position of the corresponding main surface. The second direction is perpendicular to the first direction, and the end point of the extension path in the second direction is the shallowest position of the vertical side main surface. Multiple cuboids are generated based on the extended boundary, and the multiple cuboids are spliced ​​together at the inflection point to obtain the internal shell skeleton.

3. The method according to claim 2, characterized in that, Determining a skeleton reference line for each of the main surfaces includes: Obtain the minimum mesh wall thickness and wall thickness coefficient in the finite element analysis, wherein the wall thickness coefficient is used to indicate the number of mesh layers with the minimum mesh wall thickness accommodated between the skeleton reference line and the corresponding main surface; The minimum distance between the two opposing main surfaces is taken as the local shell wall thickness; The target spacing between the skeleton reference line and the corresponding main surface is determined based on the minimum mesh wall thickness, the wall thickness coefficient, and the local shell wall thickness. At a location inside the target shell that maintains the target distance from the main surface, a straight line is drawn along the extension direction of the corresponding main surface as the skeleton reference line.

4. The method according to claim 2, characterized in that, Multiple cuboids are generated based on the extended boundary, including: Map the distance between two adjacent inflection points along the first direction to the length of a cuboid; Map the extension distance along the second direction to the width of the cuboid; The vertical distance between the two skeleton reference lines is mapped to the height of the cuboid. A cuboid is constructed by combining the length, width, and height of each cuboid, where each cuboid has the same width, but its length, height, and height are not exactly the same.

5. The method according to claim 1, characterized in that, Determining the skeleton reference lines corresponding to the two opposing main surfaces of the target shell, and adaptively generating the internal shell skeleton of the target shell based on the two skeleton reference lines includes: Two opposing main surfaces of the target housing are determined, wherein the two opposing main surfaces are a pair of opposing surfaces constituting the wall thickness direction of the target housing; Identify the thickness variation locations of each of the main surfaces and use them as inflection points of the extension paths of the corresponding skeleton reference lines; The target shell is divided into multiple independent blocks using the inflection point as the dividing point; For each independent block, a sub-skeleton reference line is determined for each of the two opposing sub-main surfaces. The sub-skeleton reference line is a straight line set along the extension direction of the corresponding sub-main surface and is located within the contour boundary of the independent block. The sub-skeleton reference lines within each of the independent blocks are extended along the second direction, wherein the second direction is the direction perpendicular to the corresponding sub-main surface, and the endpoint of the extension path of each sub-skeleton reference line is the shallowest position of the vertical side main surface of the target shell. Based on the extended boundary of each independent block, a corresponding cuboid is generated, and the cuboids are spliced ​​together at the inflection point to obtain the internal shell skeleton of the target shell.

6. The method according to claim 1, characterized in that, After obtaining the discrete model of the shell skeleton corresponding to the internal shell skeleton and the discrete model of the external contour corresponding to each external contour feature, the method further includes: Identify specific regions to be corrected in the discrete models of the shell skeleton and the discrete models of each of the external contours. These specific regions include circular surface regions, rectangular surface regions, transition regions that change from narrow to wide, and transition regions that change from wide to narrow. The original mesh flow of the circular surface region is modified into an O-type mesh flow to adapt to the contour shape of the circular surface; Adjust the distribution of grid nodes in the rectangular-face region to ensure that the number of grid nodes on the four boundaries of the rectangular-face remains consistent. A combination of quadrilateral and triangular meshes is used to perform transition processing on the meshes in the transition regions from narrow to wide and from wide to narrow, in order to eliminate mesh distortion at the transition points. After completing the mesh flow correction for each specific region, the corrected shell skeleton discrete model and the corrected external contour discrete models are output.

7. The method according to claim 1, characterized in that, Combining the discrete model of the shell skeleton and each of the discrete models of the outer contour, and outputting the combined discrete model as a standard finite element mesh includes: The coordinate system of the discrete model of the shell skeleton is aligned with the coordinate system of each discrete model of the outer contour to the global coordinate system of the target shell; Based on the positional relationship between the internal shell skeleton and each of the external contour features, and using the discrete model of the shell skeleton as the basic framework, each of the external contour discrete models is mapped to the corresponding spatial region. Coordinate alignment is performed on overlapping mesh nodes of adjacent discrete models, and fusion processing is performed on the contact surfaces of adjacent meshes to form an integrated combined discrete model; After configuring unified physical property parameters for the combined discrete model and marking the finite element analysis constraint identifiers, the combined discrete model is converted into a standard finite element mesh.

8. A finite element mesh generation device for a target shell, characterized in that, The device includes: The acquisition module is used to acquire the external contour shape and various external contour features of the target shell; The generation module is used to determine the skeleton reference lines corresponding to the two opposing main surfaces of the target shell, and adaptively generate the internal shell skeleton of the target shell according to the two skeleton reference lines. The two opposing main surfaces include a first main surface and a second main surface. The distance between the first main surface and the second main surface refers to the maximum thickness of the target shell in the vertical direction of the first main surface. The internal shell skeleton is a regular structure formed by splicing at least one cuboid. The internal shell skeleton maximizes the adaptation to the external contour shape of the target shell. The matching module is used to match the internal shell skeleton and the finite element feature model corresponding to each external contour feature from the preset database respectively; The segmentation module is used to perform finite element segmentation using the corresponding finite element feature model to obtain the discrete model of the shell skeleton corresponding to the internal shell skeleton and the discrete model of the external contour corresponding to each external contour feature. The output module is used to combine the discrete model of the shell skeleton and each of the discrete models of the outer contour, and output the combined discrete model as a standard finite element mesh.

9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images