Industrial design system and method

Abstract

The invention relates to the technical field of industrial parameter design, in particular to an industrial design system and method. The method comprises the following steps: obtaining a to-be-laid industrial structural member; performing to-be-laid layer curved surface model reconstruction on the to-be-laid layer industrial structural member, and performing surface element discretization processing to generate to-be-laid layer discrete geometric surface element data; according to the discrete geometric surface metadata of the to-be-laid layer, carrying out laying layer direction field identification, and generating laying layer direction field data; according to the layering direction field data, carrying out layering process analogue simulation, solving stress, strain and displacement distribution on each surface element, and obtaining surface element mechanical state data; performing process defect analysis according to the surface element mechanical state data to obtain layering process defect data; and performing intelligent layering process parameter combination based on the layering process defect data so as to realize industrial structural member shell layering design. According to the invention, the intelligent design of the layering process of the complex curved surface industrial structural member is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial parameter design, and particularly to an industrial design system and method. Background Art

[0002] Composite industrial structural parts have characteristics such as high specific strength, high specific stiffness, good fatigue resistance, and design flexibility; especially for industrial structural parts with complex curved surface shapes, composites are widely used for shell manufacturing. The prepreg layup process of composites involves laying fibers (such as carbon fibers, glass fibers, etc.) pre-impregnated with resin layer by layer on a mold or the shell of a structural part according to design requirements, and then through hot pressing and curing to finally form a composite product with a specific shape and performance. The quality of the layup process directly determines the mechanical properties, dimensional accuracy, and appearance quality of the final industrial product. However, the traditional layup process design method for industrial structures mainly relies on the experience of engineers and repeated tests, which is not only time-consuming and laborious, but also difficult to ensure the layup quality. Especially when dealing with complex curved surfaces, since the prepreg will deform such as bending and torsion during the laying process, various process defects are likely to occur. Especially when dealing with areas with high curvature changes, when the prepreg is laid on a hyperbolic surface area, the material needs to adapt to the curvature changes in two directions simultaneously, and wrinkles are likely to occur; while in the transition area, the sudden change in the fiber direction will cause fiber breakage or overlap. Summary of the Invention

[0003] Based on this, the present invention provides an industrial design system and method to solve at least one of the above technical problems.

[0004] To achieve the above object, an industrial design method includes the following steps:

[0005] Step S1: Obtain the industrial structural part to be laid up; reconstruct the surface model to be laid up of the industrial structural part to be laid up to generate a three-dimensional surface model; perform surface element discretization processing on the three-dimensional surface model to generate the data of the discrete geometric surface elements to be laid up.

[0006] Step S2: Analyze the surface element features within the neighborhood based on the data of the discrete geometric surface elements to be laid up to obtain the local neighborhood surface element feature data; perform geodesic tracing based on the local neighborhood surface element feature data, and then perform layup direction field recognition to generate the layup direction field data.

[0007] Step S3: Perform simulation of the layup process based on the layup direction field data, solve the stress, strain, and displacement distributions on each surface element to obtain the surface element mechanical state data; perform process defect analysis based on the surface element mechanical state data to obtain the layup process defect data.

[0008] Step S4: Based on the data of laying process defects, perform intelligent combination of laying process parameters to achieve the laying design of the shell of industrial structural parts.

[0009] The present invention reconstructs the three-dimensional surface model of the complex industrial structural part to be laid, and provides a reliable data basis for the subsequent laying design and simulation through the accurate surface model. Discretize the surface elements of the reconstructed three-dimensional surface model and decompose it into a large number of tiny discrete geometric surface elements. Based on the discrete geometric surface element data, perform the analysis of the surface element characteristics in the neighborhood. By calculating geometric characteristics such as curvature and normal variation in the local neighborhood, the local morphology of the surface can be grasped more accurately, providing a basis for the reasonable planning of the laying direction. Then, use the local neighborhood surface element characteristic data to perform geodesic tracing to obtain the laying direction field that conforms to the geometric characteristics of the surface. Geodesic tracing can ensure that the fibers are laid along the "shortest path" of the surface, reducing the bending and stretching of the fibers, thereby reducing the risk of defects such as wrinkles and fiber fractures. According to the geometric characteristics of the surface, automatically generate the optimal laying direction, avoiding the deviations and errors in manual design, and improving the accuracy and efficiency of the laying design. According to the generated laying direction field data, perform the simulation of the laying process. Through simulation, the stress, strain, and displacement distributions on each surface element can be predicted, and the mechanical problems occurring in the laying process can be comprehensively evaluated. Use the surface element mechanical state data to perform process defect analysis, and potential laying defects such as wrinkles, fiber overlap, and pores can be identified in advance. Through simulation and process defect analysis, problems can be discovered and improved before actual laying, avoiding material waste and quality problems caused by process defects, and reducing production costs. The innovation of this method lies in automatically adjusting the combination of laying process parameters, such as laying angle, prepreg tension, curing temperature, and pressure, based on the laying process defect data through intelligent algorithms, so as to minimize or eliminate laying defects to achieve the optimization of the laying design of the shell of industrial structural parts. The realization of the intelligent combination of laying process parameters improves the efficiency and quality of the laying design. Therefore, an industrial design method of the present invention reconstructs the three-dimensional surface model to accurately obtain the shell geometric information of the industrial structural part to be laid, uses geodesic tracing to generate an optimized laying path according to the geometric characteristics of the surface, and reduces fiber deformation; constructs a finite element model to simulate the laying process and identify potential process defects; based on the defect analysis results, perform intelligent process parameter optimization (including laying direction, pressure, speed, temperature, etc.) to realize the automatic adjustment and optimization of the laying process, which can significantly improve the efficiency and quality of the laying design of complex curved surface composite material structural parts, and finally realize the manufacturing of high-quality composite material structural part shells.

[0010] Preferably, the reconstruction of the to-be-laid surface model of the industrial structural part to be laid in step S1 includes:

[0011] Obtain the CAD model of the industrial structural part;

[0012] Use a three - dimensional scanning device to scan the geometric information on the outer surface of the industrial structural part to be laid, and obtain the original surface point cloud data of the structural part.

[0013] Perform multi - view point cloud registration on the original surface point cloud data of the structural part to obtain the optimized surface point cloud data of the structural part.

[0014] According to the optimized surface point cloud data of the structural part and the CAD model of the industrial structural part, reconstruct the surface model to be laid, and generate a three - dimensional surface model.

[0015] The present invention obtains the CAD model of the industrial structural part as a reference benchmark, uses a three - dimensional scanning device to scan the geometric information on the outer surface of the industrial structural part to be laid, and obtains the original surface point cloud data of the structural part. Since there are occlusions or other problems in single - view scanning, multi - view point cloud registration is performed on the original surface point cloud data of the structural part. By aligning and fusing the point cloud data from multiple views, the optimized surface point cloud data of the structural part is obtained, thus making up for the deficiencies of single - view scanning and improving the integrity and accuracy of the data. Combine the optimized surface point cloud data of the structural part with the CAD model of the industrial structural part to reconstruct the surface model to be laid. This combination method makes full use of the prior knowledge of the CAD model, can effectively eliminate the noise and errors in the point cloud data, smooth and repair the point cloud data, and obtain a more accurate three - dimensional surface model. The generated three - dimensional surface model not only has high precision but also has a good topological structure.

[0016] Preferably, the surface element discretization process according to the three - dimensional surface model in step S1 includes:

[0017] Perform differential geometry calculation of the laying surface on the three - dimensional surface model to generate the laying surface curvature distribution data;

[0018] According to the laying surface curvature distribution data, divide the surface area type to generate the surface area type data;

[0019] Set the initial surface element size for the three - dimensional surface model through the surface area type data, and then perform paving mesh division to generate the initial paving mesh division data;

[0020] Perform the minimum laying radius constraint and the fiber direction control accuracy constraint according to the initial paving mesh division data to obtain the constrained paving mesh division data;

[0021] Extract the discrete geometric surface element data to be laid according to the constrained paving mesh division data.

[0022] The present invention realizes the refined control of the discretization of the surface elements of a three-dimensional surface model by introducing the differential geometry calculation of the layup surface and the division of the surface region types, effectively solving the problems of poor mesh generation quality and inability to adapt to complex surface geometric features in traditional methods. According to the layup surface curvature distribution data, the surface is divided into different types of regions, such as flat regions, single-curvature regions, double-curvature regions, etc. This division can better adapt to the geometric features of different regions and provide guidance for the subsequent setting of the surface element size. For different types of regions, different initial surface element sizes are set, which can ensure that smaller surface element sizes are used in regions with large curvature changes to improve the calculation accuracy, while larger surface element sizes are used in regions with small curvature changes to reduce the calculation amount. According to the set initial surface element size, the layup grid is generated, and the minimum layup radius constraint and the fiber direction control accuracy constraint are introduced to optimize the initial layup grid. The minimum layup radius constraint can ensure that the size of the grid is not less than the preset minimum value, avoiding the appearance of overly small surface elements, thus ensuring the feasibility of the layup; the fiber direction control accuracy constraint can ensure that the deviation between the edge of the grid and the fiber direction does not exceed the preset threshold, thereby improving the control accuracy of the layup direction.

[0023] Preferably, the extraction of the discrete geometric surface element data of the layer to be laid according to the constrained layup grid generation data includes:

[0024] Performing surface element discretization processing on the three-dimensional surface model through the constrained layup grid generation data to obtain the layup discrete surface element data;

[0025] Reading the vertex coordinates of the surface elements according to the layup discrete surface element data;

[0026] Calculating the geometric center coordinates of the surface elements according to the vertex coordinates of the surface elements;

[0027] Calculating the normal vector of the surface elements according to the layup discrete surface element data;

[0028] Calculating the area of the surface elements according to the layup discrete surface element data;

[0029] Integrating the geometric center coordinates of the surface elements, the normal vector of the surface elements, and the area of the surface elements to generate the discrete geometric surface element data of the layer to be laid.

[0030] The present invention performs element discretization on a three-dimensional surface model by constraining the paving grid division data. The constrained paving grid division can ensure the quality of the grid, avoid the appearance of overly distorted or too small elements, transform the complex surface into a series of discrete elements, read the vertex coordinates of the elements from the paving discrete element data, and calculate the geometric center coordinates of the elements accordingly. The geometric center coordinates of the elements represent the position information of the elements and are important reference points for local neighborhood analysis and direction field calculation. Calculate the normal vector of the element according to the paving discrete element data. The normal vector of the element is an important parameter describing the direction of the element, which reflects the local orientation of the surface at this point. Calculate the area of the element according to the paving discrete element data. The area of the element is an important parameter for mechanical analysis and energy calculation. Integrate the geometric characteristics of the elements such as the geometric center coordinates of the elements, the normal vector of the elements, and the area of the elements to construct a complete and structured data set, which is convenient for the subsequent module to call and process.

[0031] Preferably, step S2 includes the following steps:

[0032] Step S21: Construct a KD-tree spatial index based on the geometric center coordinates of the elements in the discrete geometric element data of the layer to be paved;

[0033] Step S22: Based on the KD-tree spatial index, with each geometric center coordinate of the element as the center, search for the elements in its neighborhood to obtain an element neighborhood set;

[0034] Step S23: Extract the geometric center coordinates and normal vectors of all elements in the target element neighborhood according to the element neighborhood set to obtain local neighborhood element feature data;

[0035] Step S24: Perform local surface fitting according to the local neighborhood element feature data, and then calculate the average curvature of each neighborhood element to obtain neighborhood curvature distribution data;

[0036] Step S25: Solve the neighborhood surface Gaussian eigenvalue and neighborhood principal curvature direction vector of the Gaussian map according to the neighborhood curvature distribution data;

[0037] Step S26: Use the area of the element in the discrete geometric element data of the layer to be paved as the weight, and perform weighted calculation of the principal curvature direction according to the neighborhood surface Gaussian eigenvalue and neighborhood principal curvature direction vector to generate weighted principal curvature direction data;

[0038] Step S27: Perform geodesic tracing on the three-dimensional surface model through the weighted principal curvature direction data, and then perform layer direction field identification to generate layer direction field data.

[0039] By constructing a KD - tree spatial index, the present invention efficiently searches for neighboring patches, and combines local surface fitting and weighted principal curvature direction calculation to achieve accurate identification of the ply direction field of complex surfaces, overcoming the problem of direction deviation that easily occurs in traditional methods when dealing with regions of high curvature change. Organizing discrete patches using the KD - tree spatial index greatly improves the efficiency of neighborhood search. Compared with brute - force search, the KD - tree can quickly locate neighboring patches near a specified patch, reducing the computational complexity. By extracting the geometric center coordinates and normal vectors of neighboring patches, local neighborhood patch feature data is constructed. Using this feature data for local surface fitting can more accurately estimate the local curvature information of the surface, overcoming the errors generated by directly calculating curvature using discrete patches. By solving the eigenvalues and eigenvector of the Gauss map, the principal curvature direction information of the surface is obtained. And using the patch area as a weight, weighted calculation of the eigenvalues and eigenvector is performed, considering the influence of different patches on the ply direction, guiding the fibers to be laid along the principal curvature direction of the surface, minimizing the bending and stretching of the fibers, and reducing the risk of wrinkles and defects. Compared with traditional methods, this solution can more accurately identify the principal curvature direction of the surface and generate a more reasonable ply direction field, improving the ply quality and efficiency.

[0040] Preferably, step S27 includes the following steps:

[0041] Step S271: Select the maximum principal curvature according to the weighted principal curvature direction data, and use the corresponding direction as the initial geodesic direction;

[0042] Step S272: Set the geodesic step size to 3 mm, and calculate the coordinates of the next target point reached after advancing one such geodesic step in this direction based on the initial geodesic direction through the three - dimensional surface model, obtaining the coordinates of the next target point;

[0043] Step S273: Search for the patch containing this target point through the three - dimensional surface model based on the coordinates of the next target point, and project the target point onto this patch along the patch normal vector direction to calculate the projection distance; when the projection distance is greater than 1 mm, it is determined that the projection fails and the tracking of this geodesic is stopped, otherwise, an initial geodesic segment with a length of 5 mm is generated on the three - dimensional surface model, and this target point is used as the new target point;

[0044] Step S274: Repeat step S272 and step S273 until the geodesic segment reaches the surface boundary or the angle between the geodesic direction and the initial direction exceeds 90°, thereby obtaining the initial ply direction line;

[0045] Step S275: Calculate the angle and distance between adjacent ply direction lines according to the initial ply direction line, obtaining the adjacent ply direction line relationship data;

[0046] Step S276: According to the adjacent ply orientation line relationship data, for adjacent ply orientation lines with an included angle less than the angle threshold and a distance less than the distance threshold, merge them to generate merged ply orientation line data. Here, the angle threshold is set to 15°, the distance threshold is 2 mm, and the merging direction of the new ply orientation line is the average direction of the two initial ply orientation lines;

[0047] Step S277: Using the geometric center coordinates of the element as interpolation nodes, perform ply orientation field interpolation on the merged ply orientation line data to generate ply orientation field data.

[0048] Through precise geodesic tracing and intelligent ply orientation line merging and interpolation processing, the present invention realizes the refined construction of the ply orientation field for complex surfaces, effectively solving the problems of sparse, discontinuous ply orientation lines and easy occurrence of conflicts in traditional methods. By selecting the maximum principal curvature direction as the initial geodesic direction, it ensures that the geodesic extends along the direction with the fastest change of the surface, thus more effectively covering the entire surface. Setting a reasonable geodesic step size not only ensures the accuracy of the tracing but also avoids excessive computational amount. By judging whether the tracing fails through the projection distance, the ineffective tracing is stopped in time, improving the computational efficiency. By calculating the included angle and distance between adjacent ply orientation lines, the distribution of ply orientation lines is analyzed. For adjacent ply orientation lines with an included angle and distance less than the threshold, merge them to reduce the number of ply orientation lines, avoid process conflicts caused by overly dense ply orientation lines, and improve the integrity of the ply orientation field. The average direction is adopted for the merged new ply orientation line to ensure the smooth transition of the ply orientation field. Using the geometric center coordinates of the element as interpolation nodes, perform ply orientation field interpolation on the merged ply orientation line data to achieve the coverage of the ply orientation field for the entire surface. The interpolation processing can fill the blank areas between ply orientation lines, optimize the distribution of ply orientation lines, and improve the quality of the ply orientation field.

[0049] Preferably, step S3 includes the following steps:

[0050] Step S31: Obtain the prepreg property parameters and the preliminary laying process parameters of the prepreg;

[0051] Step S32: Construct a local finite element model based on the element according to the ply orientation field data and the prepreg property parameters;

[0052] Step S33: Analyze the element loading sequence of the local finite element model through the preliminary laying process parameters of the prepreg to generate element ply laying load sequence data;

[0053] Step S34: Calculate the element laying load according to the laying process parameters to generate discrete element load data;

[0054] Step S35: Based on the ply laying load sequence data of the panel elements and the discrete panel element load data, perform a ply laying process simulation on the local finite element model, solve for the stress, strain, and displacement distributions on each panel element, and obtain the panel element mechanical state data, where the panel element mechanical state data includes panel element stress field data, panel element displacement field data, and panel element shear strain data;

[0055] Step S36: Perform a process defect analysis based on the panel element mechanical state data to obtain ply laying process defect data, where the ply laying process defect data includes wrinkling risk panel element data, gap risk area data, and fiber misalignment risk panel element data.

[0056] The present invention realizes a refined simulation of the ply laying process by constructing a local finite element model based on panel elements and combining the ply laying order analysis of panel elements and accurate load calculation, so as to more accurately predict the stress, strain, and displacement distributions on the panel elements. Combining the property parameters of the prepreg and the preliminary laying process parameters, a local finite element model based on panel elements is constructed. This local model can effectively reduce the calculation scale, improve the simulation efficiency, and ensure the accuracy of the simulation at the same time. Through the ply laying order analysis of panel elements, the ply laying order of each panel element is determined. The determination of this order can more realistically reflect the actual ply laying process and avoid simulation errors caused by unreasonable loading orders. According to the laying process parameters, the ply laying load of the panel elements is calculated to obtain the load information on each panel element. Accurate load calculation can more accurately simulate the actual ply laying mechanical environment and improve the reliability of the simulation results. Based on the ply laying load sequence data of panel elements and the discrete panel element load data, a ply laying process simulation is performed on the local finite element model. By solving the stress, strain, and displacement distributions on each panel element, the mechanical behavior of the panel element during the ply laying process can be comprehensively reflected. By analyzing the stress concentration area, displacement abnormal area, and shear strain excessive area, the generation positions and severity of process defects such as wrinkles, voids, and delaminations can be predicted.

[0057] Preferably, the process defect analysis based on the panel element mechanical state data includes:

[0058] Calculate the deviation angle between the principal stress direction and the ply laying direction of each panel element according to the panel element stress field data to obtain stress-ply deviation angle data;

[0059] When the stress-ply deviation angle data is greater than 15° and the compressive stress value of this panel element is greater than 0.5 MPa, it is determined that this panel element has a wrinkling risk, and the wrinkling risk panel element data is obtained;

[0060] Calculate the relative displacement between adjacent panel elements according to the panel element displacement field data. When the relative displacement is greater than 0.8 mm, it is determined that there is a gap risk in this area, and the gap risk area data is obtained;

[0061] Calculate the shear strain of each element according to the element shear strain data. When the shear strain is greater than the preset strain threshold, it is determined that there is a risk of fiber misalignment in the element, and the fiber misalignment risk element data is obtained.

[0062] The present invention can effectively identify the area at risk of wrinkling by calculating the deviation angle between the principal stress direction and the ply direction of each element. When the deviation angle is large and the element is subjected to a large compressive stress, it indicates that the fibers in this area are prone to buckling and deformation, thus generating wrinkles, and can more accurately predict the occurrence position of wrinkles. By calculating the relative displacement between adjacent elements, the area at risk of gap can be effectively identified. When the relative displacement is large, it indicates that there is obvious debonding or separation between adjacent elements, which is likely to form a gap, and can more accurately predict the occurrence position of the gap. By calculating the shear strain of each element, the area at risk of fiber misalignment can be effectively identified. When the shear strain exceeds the preset strain threshold, it indicates that the fibers in this area are prone to slipping and misalignment, thus affecting the mechanical properties of the ply structure, and can more accurately predict the occurrence position of fiber misalignment.

[0063] Preferably, step S4 includes the following steps:

[0064] Step S41: Adjust the ply direction according to the wrinkling risk element data to obtain the optimized ply direction data;

[0065] Step S42: Adjust the ply pressure and the pressure holding time according to the gap risk area data to obtain the optimized ply pressure-time parameters;

[0066] Step S43: Adjust the laying speed and the heating temperature according to the fiber misalignment risk element data to obtain the optimized laying speed-temperature parameters;

[0067] Step S44: Combine the optimized ply direction data, the optimized ply pressure-time parameters, and the optimized laying speed-temperature parameters for the intelligent ply process parameters to realize the ply design of the industrial structural part shell.

[0068] The present invention adjusts the laying direction according to the wrinkle risk element data. By optimizing the laying direction of the fibers, the bending and compression of the fibers during the laying process are reduced, thereby reducing the possibility of wrinkles. This targeted adjustment of the laying direction can more effectively control the deformation of the fibers and improve the stability of the laid structure. The laying pressure and holding time are adjusted according to the gap risk area data. Increasing the laying pressure can improve the adhesion between the prepreg and the mold surface and reduce the generation of gaps; extending the holding time can enable the prepreg to infiltrate the fibers more fully and improve the interlayer bonding strength, thereby reducing the risk of gaps. This strategy of synergistically adjusting the laying pressure and holding time can more effectively control the interlayer bonding quality. The laying speed and heating temperature are adjusted according to the fiber misalignment risk element data. Reducing the laying speed can reduce the shear force on the fibers during the laying process and reduce the possibility of fiber misalignment; adjusting the heating temperature can change the viscosity of the prepreg, thereby affecting the fluidity and arrangement of the fibers. By optimizing the laying speed and heating temperature, the arrangement and distribution of the fibers can be more effectively controlled.

[0069] Preferably, the present invention further provides an industrial design system that executes the industrial design method as described above. The industrial design system includes:

[0070] An industrial structure modeling module for obtaining the industrial structure part to be laid; reconstructing the surface model to be laid of the industrial structure part to generate a three-dimensional surface model; performing element discretization processing on the three-dimensional surface model to generate the discrete geometric element data to be laid.

[0071] A laying direction field generation module for analyzing the element features in the neighborhood according to the discrete geometric element data to be laid to obtain the local neighborhood element feature data; performing geodesic tracking according to the local neighborhood element feature data, and then performing laying direction field identification to generate the laying direction field data.

[0072] A defect analysis module for performing simulation of the laying process according to the laying direction field data, solving the stress, strain and displacement distributions on each element to obtain the element mechanical state data; performing process defect analysis according to the element mechanical state data to obtain the laying process defect data.

[0073] An intelligent parameter optimization module for performing intelligent combination of laying process parameters based on the laying process defect data to achieve the laying design of the industrial structure part shell. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 is a schematic flow chart of the steps of an industrial design method of the present invention;

[0075] Figure 2 is Figure 1 a detailed implementation step flow chart of step S3 in

[0076] Figure 3 For Figure 1 a schematic flow chart of the detailed implementation steps of step S4 in

[0077] The realization, functional features and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. Specific embodiments

[0078] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0079] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0080] It should be understood that although the terms "first", "second", etc. may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be called the second unit, and similarly the second unit may be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed associated items.

[0081] To achieve the above object, please refer to Figures 1 to 3 , the present invention provides an industrial design method, including the following steps:

[0082] Step S1: Obtain the industrial structural member to be laid; reconstruct the surface model to be laid of the industrial structural member to generate a three-dimensional surface model; perform surface element discretization processing on the three-dimensional surface model to generate the data of the discrete geometric surface elements to be laid;

[0083] Step S2: Analyze the surface element features in the neighborhood according to the data of the discrete geometric surface elements to be laid to obtain the local neighborhood surface element feature data; perform geodesic tracking according to the local neighborhood surface element feature data, and then perform layup direction field identification to generate the layup direction field data;

[0084] Step S3: Perform a simulation of the laying process based on the laying direction field data, solve for the stress, strain, and displacement distributions on each surface element, and obtain the mechanical state data of the surface element; perform a process defect analysis based on the mechanical state data of the surface element to obtain the laying process defect data;

[0085] Step S4: Perform an intelligent combination of laying process parameters based on the laying process defect data to achieve the laying design of the industrial structural part housing.

[0086] In the embodiment of the present invention, the industrial design method includes the following steps:

[0087] Step S1: Obtain the industrial structural part to be laid; reconstruct the surface model to be laid of the industrial structural part to generate a three-dimensional surface model; perform a discretization process on the three-dimensional surface model to generate the discrete geometric surface element data to be laid;

[0088] In the embodiment of the present invention, for example, obtain the engine hood part to be laid from the database of the industrial design system, including but not limited to the engine hood part. This part is a small industrial structural part housing with a complex surface. The system reads the STEP format CAD model file of the engine hood through the built-in CAD model interface. Parse the STEP file, extract the geometric information therein, including surfaces, boundaries, and topological relationships, and convert it into the internal data structure of the system, such as NURBS surface representation. Then, use a high-precision structured light three-dimensional scanner to scan the actual engine hood part to obtain its surface point cloud data. The scanner projects a coded structured light pattern, captures the reflected pattern through a binocular camera, and calculates the three-dimensional coordinates corresponding to each pixel using the principle of triangulation. Change the relative positions of the scanner and the part multiple times to obtain multi-viewpoint clouds. Use the iterative closest point (ICP) algorithm to register the multi-viewpoint clouds to the unified coordinate system. Based on the registered point cloud data and the CAD model, perform surface reconstruction. Use non-uniform rational B-spline (NURBS) surfaces to fit the point cloud, and at the same time refer to the boundary and shape constraints of the CAD model to generate an accurate three-dimensional surface model. Finally, use the constrained Delaunay triangulation algorithm to discretize the three-dimensional surface model. Define the constraint edges according to the laying process requirements (such as fiber direction), and generate a series of triangular surface elements with these constraint edges as the limit. Each surface element consists of three vertices and three edges. The system stores information such as the vertex index, coordinates, normal vector, and area of each surface element to generate the discrete geometric surface element data to be laid.

[0089] Step S2: Perform an analysis of the surface element features in the neighborhood based on the discrete geometric surface element data to be laid to obtain the local neighborhood surface element feature data; perform geodesic tracing based on the local neighborhood surface element feature data, and then perform a laying direction field identification to generate the laying direction field data;

[0090] In the embodiment of the present invention, the geometric center coordinates of all triangular elements are extracted according to the discrete geometric element data of the layer to be laid. A KD-tree spatial index is constructed using these coordinates. For each element, using its geometric center as the query point, k nearest neighbor elements (e.g., k = 20) are searched using the KD-tree to form an element neighborhood set. The geometric center coordinates and normal vectors of all elements within the neighborhood set are extracted to constitute local neighborhood element feature data. Using these feature data, local quadratic surface fitting is performed, and the quadratic surface equation is fitted using the least squares method. The principal curvatures k1 and k2 of the quadratic surface at the geometric center point of each element are calculated, and the mean curvature H = (k1 + k2) / 2. The eigenvector corresponding to the principal curvature with a larger absolute value is selected as the initial geodesic direction. Starting from the starting element, the geodesic is traced on the surface along the initial geodesic direction. The fourth-order Runge-Kutta method is used to calculate the coordinates of the next target point, and the step size is set to 3 mm. The target point is projected onto the nearest element, and if the projection distance is greater than 1 mm, the tracing stops. The starting point and the target point are connected to generate a geodesic segment. This process is repeated until the geodesic reaches the boundary or the direction changes by more than 90 degrees. Multiple geodesics are traced, and the laying direction field is identified according to the geodesic directions to generate laying direction field data.

[0091] Step S3: Perform a simulation of the laying process according to the laying direction field data, solve the stress, strain, and displacement distributions on each element to obtain element mechanical state data; perform process defect analysis based on the element mechanical state data to obtain laying process defect data;

[0092] In the embodiments of the present invention, the prepreg property parameters (such as E1, E2, in-plane shear modulus G12, Poisson's ratio V12, thickness, and density of carbon fiber / epoxy resin) and preliminary laying process parameters (such as laying pressure, temperature, speed, and radius of the laying roller) are obtained from the material database. According to the laying direction field data, the prepreg material properties are transformed into the local coordinate system of each element. Taking each triangular element as a finite element unit, a local finite element model is constructed. The stiffness matrix of each element is calculated and assembled into a global stiffness matrix. According to the laying path, the element loading sequence is determined. The incremental loading method is used to simulate the layer-by-layer laying of the prepreg. For each activated element, its pressure load and the contact force load caused by the laying roller are calculated and added to the global load vector. This process is repeated until all elements are activated. The stress, strain, and displacement data of each element are stored as the mechanical state data of the element. Based on these data, process defect analysis is carried out: calculate the deviation angle between the principal stress direction and the laying direction. If the deviation is greater than 15° and there is a compressive stress greater than 0.5 MPa, it is determined as the risk of wrinkling; calculate the relative displacement between adjacent elements. If it is greater than 0.8 mm, it is determined as the risk of gap; calculate the shear strain of the element. If it is greater than the preset threshold (such as 5%), it is determined as the risk of fiber misalignment. All defective elements and their defect types are stored as the laying process defect data.

[0093] Step S4: Based on the laying process defect data, perform intelligent combination of laying process parameters to achieve the laying design of the industrial structural part shell.

[0094] In the embodiments of the present invention, for the elements at risk of wrinkling, the laying direction is adjusted to be closer to the principal stress direction. Calculate the weighted average of the principal stress direction and the original laying direction as the new laying direction. For the gap risk area, increase the laying pressure (such as increasing by 0.1 MPa) and extend the pressure holding time (such as increasing by 5 seconds). For the elements at risk of fiber misalignment, reduce the laying speed (such as reducing by 2 mm / s) and adjust the heating temperature as needed (if heating is required). Integrate the optimized laying direction, laying pressure-time parameters, and laying speed-temperature parameters. Create a data structure, indexed by the element ID, to store the optimized process parameters of each element. For example, for each element, store its optimized laying direction, optimized laying pressure (or initial pressure), optimized pressure holding time (or initial time), optimized laying speed (or initial speed), and optimized heating temperature (or "NA"). Further, these element-level parameters can be aggregated to the laying path level to calculate the average parameters of each path. Finally, a process parameter combination including element-level and laying path-level parameters is output. This combination can be directly used to control the numerically controlled laying machine to achieve the automated laying of the engine hood shell, improve the laying quality, and reduce defects.

[0095] Preferably, the reconstruction of the surface model to be laid for the industrial structural member in step S1 includes:

[0096] Obtain the CAD model of the industrial structural member;

[0097] Use a three-dimensional scanning device to scan the geometric information of the outer shell surface of the industrial structural member to be laid, and obtain the original structural member surface point cloud data;

[0098] Perform multi-viewpoint cloud registration on the original structural member surface point cloud data to obtain the optimized structural member surface point cloud data;

[0099] Reconstruct the surface model to be laid according to the optimized structural member surface point cloud data and the CAD model of the industrial structural member, and generate a three-dimensional surface model.

[0100] In the embodiments of the present invention, a CAD model of the target small industrial structural part housing is directly extracted from the database of the industrial design system. The housing is an engine hood part with complex curved surfaces, and the CAD model is stored in the STEP format. The system reads the geometric information in the STEP file, including the curved surfaces, boundaries, topological relationships, etc. of the housing, through a built-in STEP format parsing module. The parsing module converts the geometric primitives (such as NURBS surfaces, B-Rep solids) in the STEP file into the internal data structure of the system, such as triangular meshes or parametric surface representations. The actual engine hood part (the housing of the industrial structural part to be laminated) is scanned using a high-precision structured light 3D scanner. The scanner projects a structured light pattern with a specific code onto the surface of the engine hood and captures the reflected light pattern through two or more cameras. Using the principle of triangulation, the control unit of the scanner calculates the three-dimensional spatial coordinates corresponding to each camera pixel point. The specific operation is as follows: Place the engine hood part within the working range of the scanner, and adjust the parameters of the scanner, including the exposure time, the density of the projected pattern, etc., to ensure high-quality point cloud data is obtained. During the scanning process, the relative position and angle between the scanner and the engine hood part need to be changed multiple times to cover the entire housing surface and avoid scanning dead angles. After each scan is completed, the control unit of the scanner generates a set of raw point cloud data, and each set of data contains a large number of three-dimensional coordinate points (x, y, z) and reflection intensity information. Manually or automatically select at least three feature points that can be clearly identified in the point clouds from different viewpoints as the initial corresponding points. Then, calculate a rigid body transformation matrix (including a rotation matrix and a translation vector) to minimize the distance between these initial corresponding points. Using the calculated transformation matrix, transform one viewpoint point cloud into the coordinate system of another viewpoint point cloud. Next, based on the transformed point cloud, search for more accurate corresponding point pairs in the overlapping area of the two point clouds. For each point, search for the point with the closest distance in its nearest neighborhood as the corresponding point. Using all the found corresponding point pairs, recalculate the transformation matrix. Repeat the process of searching for corresponding points and calculating the transformation matrix until the average distance between the corresponding points is less than a preset threshold or the maximum number of iterations is reached. The finally obtained transformation matrix is used to unify all viewpoint point clouds into the same coordinate system to generate a complete and optimized point cloud data. The point cloud is segmented using the normal vector information of the point cloud, and the point cloud is divided into different surface patches. The region growing algorithm is used for segmentation: Starting from a seed point, calculate the normal vectors of the points in its neighborhood. If the angle between the normal vector of the neighborhood point and the normal vector of the seed point is less than a set threshold, then classify the neighborhood point into the same surface patch; repeat this process until all points are divided into different surface patches. Then, parametric fitting is performed on each surface patch. Non-uniform rational B-spline (NURBS) surfaces are used for fitting.By adjusting the control points, knot vectors, and weights of the NURBS surface, the NURBS surface is made to approximate the point cloud data on the surface patch. During the fitting process, the CAD model is used as a reference to constrain the boundaries and shape of the NURBS surface to make it consistent with the CAD model. Finally, all the fitted NURBS surface patches are stitched together to form a complete three-dimensional surface model.

[0101] Preferably, the surface element discretization process according to the three-dimensional surface model in step S1 includes:

[0102] Perform differential geometry calculations on the layup surface of the three-dimensional surface model to generate layup surface curvature distribution data;

[0103] Divide the surface region types according to the layup surface curvature distribution data to generate surface region type data;

[0104] Set the initial surface element size for the three-dimensional surface model through the surface region type data, and then perform paving mesh division to generate initial paving mesh division data;

[0105] Perform minimum layup radius constraint and fiber direction control accuracy constraint according to the initial paving mesh division data to obtain constrained paving mesh division data;

[0106] Extract the discrete geometric surface element data to be laid according to the constrained paving mesh division data.

[0107] In the embodiments of the present invention, a set of constraint edges are defined on a three-dimensional surface model according to the requirements of the laying process (such as fiber direction, laying tape width). These constraint edges can be along a specific fiber direction or along the boundary of the laying tape. Then, with these constraint edges as limiting conditions, the three-dimensional surface model is triangulated. The triangulation algorithm generates a series of triangular surface elements that completely cover the entire surface model, and all the edges of the triangles coincide or are parallel to the constraint edges. Each triangular surface element is a plane composed of three vertices and three edges. The system stores the vertex indices, vertex coordinates, and adjacent surface element information of each triangular surface element as the laid discrete surface element data. Traverse the laid discrete surface element data generated in the previous step. For each triangular surface element, according to its vertex indices, read the three-dimensional coordinates (x, y, z) of the three vertices from the vertex coordinate array. For example, for the i-th triangular surface element with vertex indices (v1, v2, v3), read the three coordinate values at indices v1, v2, and v3 in the vertex coordinate array respectively to obtain the three vertex coordinates of this triangular surface element. Use the three vertex coordinates of the triangular surface element to calculate the normal vector of the triangular surface element respectively, calculate the cross product of the vectors, and the obtained cross product vector is the normal vector (nx, ny, nz) of this triangular surface element, and calculate the area of this triangular surface element. Integrate the geometric center coordinates (xc, yc, zc), unit normal vector (nx, ny, nz), and area Area of each triangular surface element calculated in the previous steps into a structure or data object. This structure also includes the vertex indices and adjacent surface element information of this surface element. Traverse all triangular surface elements to generate the laid discrete geometric surface element data of all surface elements and store these data in a list.

[0108] Preferably, the extraction of the laid discrete geometric surface element data according to the constraint laying grid division data includes:

[0109] Perform surface element discretization processing on the three-dimensional surface model through the constraint laying grid division data to obtain the laid discrete surface element data;

[0110] Read the surface element vertex coordinates according to the laid discrete surface element data;

[0111] Calculate the geometric center coordinates of the surface element according to the surface element vertex coordinates;

[0112] Calculate the surface element normal vector according to the laid discrete surface element data;

[0113] Calculate the surface element area according to the laid discrete surface element data;

[0114] Integrate the geometric center coordinates of the surface element, the surface element normal vector, and the surface element area for the geometric feature integration of the laid surface elements to generate the laid discrete geometric surface element data.

[0115] In the embodiment of the present invention, the constrained Delaunay triangulation algorithm is used to mesh the three-dimensional surface model of the engine hood component reconstructed in the previous step. First, a set of constrained edges are defined on the three-dimensional surface model according to the layup process requirements (such as fiber direction, width of the layup tape). These constrained edges can be along a specific fiber direction or along the boundary of the layup tape. Then, with these constrained edges as the limiting conditions, the three-dimensional surface model is triangulated. The triangulation algorithm generates a series of triangular elements, which completely cover the entire surface model, and all the edges of the triangles coincide with or are parallel to the constrained edges. Each triangular element is a plane, composed of three vertices and three edges. The system stores the vertex indices, vertex coordinates, and adjacent element information of each triangular element as the discrete ply element data. Traverse the discrete ply element data generated in the previous step. For each triangular element, according to its vertex indices, read the three-dimensional coordinates (x, y, z) of the three vertices from the vertex coordinate array. For example, for the i-th triangular element with vertex indices (v1, v2, v3), read the three coordinate values at indices v1, v2, and v3 in the vertex coordinate array respectively to obtain the three vertex coordinates of this triangular element: (x1, y1, z1), (x2, y2, z2), (x3, y3). Using the three vertex coordinates (x1, y1, z1), (x2, y2, z2), (x3, y3) of the triangular element, calculate the geometric center coordinates (xc, yc, zc) of this triangular element. The arithmetic mean method is used to calculate the geometric center: xc = (x1 + x2 + x3) / 3; yc = (y1 + y2 + y3) / 3; zc = (z1 + z2 + z3) / 3. This calculation formula directly performs a weighted average on the coordinates of the three vertices to obtain the geometric center of the triangular element. Using the three vertex coordinates (x1, y1, z1), (x2, y2, z2), (x3, y3) of the triangular element, calculate the normal vector (nx, ny, nz) of this triangular element. Calculate two vectors: vector V1 = (x2 - x1, y2 - y1, z2 - z1); vector V2 = (x3 - x1, y3 - y1, z3 - z1). Then, calculate the cross product of these two vectors: V1×V2 = ((y2 - y1)(z3 - z1) - (z2 - z1)(y3 - y1), (z2 - z1)(x3 - x1) - (x2 - x1)(z3 - z1), (x2 - x1)(y3 - y1) - (y2 - y1)(x3 - x1)). The resulting cross product vector is the normal vector (nx, ny, nz) of this triangular element. Calculate the cross product of these two vectors according to vector V1 and vector V2, and then calculate the magnitude of the cross product vector. The area of the triangular element is half of the magnitude of the cross product vector, and the element area is obtained.Create a data structure to store the geometric feature information of each patch element, such as a structure or class that contains member variables like patch element index, geometric center coordinates, normal vector, and area. Traverse all patch elements and store the information such as geometric center coordinates, normal vector, and area of each patch element into the corresponding data structure to obtain the discrete geometric patch element data of the layer to be laid.

[0116] Preferably, step S2 includes the following steps:

[0117] Step S21: Construct a KD - tree spatial index based on the geometric center coordinates of the patch elements in the discrete geometric patch element data of the layer to be laid;

[0118] Step S22: Based on the KD - tree spatial index, with the geometric center coordinates of each patch element as the center, search for the patch elements within its neighborhood to obtain the patch element neighborhood set;

[0119] Step S23: Extract the geometric center coordinates and normal vectors of all patch elements within the target patch element neighborhood according to the patch element neighborhood set to obtain the local neighborhood patch element feature data;

[0120] Step S24: Perform local surface fitting based on the local neighborhood patch element feature data, and then calculate the average curvature of each neighborhood patch element to obtain the neighborhood curvature distribution data;

[0121] Step S25: Solve the neighborhood surface Gaussian eigenvalues and neighborhood principal curvature direction vectors of the Gaussian map according to the neighborhood curvature distribution data;

[0122] Step S26: Take the patch element area in the discrete geometric patch element data of the layer to be laid as the weight, and perform weighted calculation of the principal curvature direction according to the neighborhood surface Gaussian eigenvalues and neighborhood principal curvature direction vectors to generate weighted principal curvature direction data;

[0123] Step S27: Perform geodesic tracing on the three - dimensional surface model through the weighted principal curvature direction data, and then perform lay - up direction field recognition to generate lay - up direction field data.

[0124] In the embodiments of the present invention, the geometric center coordinates (xc, yc, zc) of all triangular elements are extracted. These geometric center coordinates are used as point cloud data to construct a KD tree (K-Dimensional Tree) spatial index. A KD tree is a binary tree structure used to efficiently organize and search points in a multi-dimensional space. The construction process is as follows: First, select a dimension (such as the x-axis), and calculate the median of all geometric center coordinates in this dimension; use this median as the splitting point to divide all geometric center coordinates into two groups, one group with x coordinates less than the median and the other group with x coordinates greater than or equal to the median. Then, recursively split these two groups of data. Calculate the median in the next dimension (such as the y-axis) and group them until the number of geometric center coordinates contained in each leaf node is less than a preset threshold (such as 10). Each node of the KD tree stores a splitting dimension, a splitting value, and pointers to its left and right subtrees. The constructed KD tree can be used to quickly find the nearest neighbor or k nearest neighbors of a certain point. Traverse all triangular elements, use the geometric center coordinates (xc, yc, zc) of each element as the query point, and perform k-nearest neighbor search using the KD tree constructed in the previous step. Set a neighborhood radius r (for example, 2 times the average curvature radius of the hood component), or set a number k of neighborhood elements (for example, 20). Start from the root node of the KD tree and recursively visit the subtrees. For each node, calculate the distance from the query point to the corresponding splitting hyperplane of the node. If this distance is less than the current nearest distance (the initial value is infinity), then visit the two subtrees of the node. During the process of visiting the subtrees, continuously update the current nearest distance and the k nearest neighbor points. When visiting a leaf node, calculate the distances from the query point to all geometric center coordinates in the leaf node and update the k nearest neighbor points. After the search is completed, the k nearest neighbor elements of each element are obtained. Traverse all triangular elements. For each target element, according to the element indices in the element neighborhood set obtained in the previous step, extract the geometric center coordinates (xc, yc, zc) and unit normal vectors (nx, ny, nz) of these neighborhood elements from the discrete geometric element data of the layer to be laid. For example, for the i-th target element, its neighborhood set is {n1, n2,..., nk}, then extract the geometric center coordinates of these neighborhood elements respectively: (xc1, yc1, zc1), (xc2, yc2, zc2),...,(xck, yck, zck); and the normal vectors: (nx1, ny1, nz1), (nx2, ny2, nz2),...,(nxk, nyk, nzk). Combine these extracted geometric center coordinates and normal vectors, together with the geometric center coordinates and normal vectors of the target element itself, into a local neighborhood element feature data set. For each target element, perform local quadratic surface fitting using the local neighborhood element feature data obtained in the previous step.A quadratic surface equation is fitted using the least squares method: z = ax^2 + by^2 + cxy + dx + ey + f. Substitute the geometric center coordinates (x, y, z) of the neighborhood surface element into this equation to construct an overdetermined system of equations. By solving the least squares solution of this overdetermined system of equations, the coefficients a, b, c, d, e, f of the quadratic surface are obtained. Using these coefficients, calculate the principal curvatures k1 and k2 of the quadratic surface at the geometric center point of each neighborhood surface element. The mean curvature H is defined as the average of the two principal curvatures: H = (k1 + k2) / 2. Calculate the mean curvature H of each neighborhood surface element and store these mean curvature values as a neighborhood curvature distribution data. Based on the neighborhood curvature distribution data obtained in the previous step, construct the Weingarten matrix (also known as the shape operator matrix) for each surface element. The Weingarten matrix is a 2x2 matrix that describes the local linear transformation of the Gaussian map of the surface at that point. The elements of the Weingarten matrix can be calculated from the coefficients of the quadratic surface fitting. For each surface element, calculate the eigenvalues λ1 and λ2 of its Weingarten matrix, and the corresponding eigenvectors v1 and v2. The eigenvalues λ1 and λ2 correspond to the two principal curvatures k1 and k2 at that surface element (sorted from largest to smallest in absolute value), and the eigenvectors v1 and v2 correspond to the two principal curvature directions. These eigenvalues and eigenvectors describe the local bending characteristics of the surface at that point. For each triangular surface element, extract its area Area from the discrete geometric surface element data of the layer to be laid. Using the eigenvalues λ1, λ2 and eigenvectors v1, v2 calculated in the previous step, perform a weighted calculation of the principal curvature directions. Using the area Area as the weight, perform a weighted average of the two principal curvature directions. The calculation formula is as follows: w_d = (Area × |λ1| × v1 + Area × |λ2| × v2) / (Area × (|λ1| + |λ2|)). If |λ1| + |λ2| is close to zero (indicating that this surface element is close to a plane), then set the weighted principal curvature direction to a predefined default direction (for example, along the main axis direction of the CAD model). This formula performs a weighted average of the two principal curvature directions according to the area of the surface element and the magnitudes of the principal curvatures, obtaining a direction that can better reflect the requirements of the laying process., Perform geodesic tracing on the three-dimensional surface model of the engine hood. Starting from a starting point, along the weighted principal curvature direction of that point, trace a geodesic of a fixed length on the surface. A geodesic is the shortest path between two points on the surface. Use a numerical integration method (such as the Runge-Kutta method) to calculate the geodesic. During the tracing process, continuously update the weighted principal curvature direction of the current point and move forward a small step along this direction. Repeat this process until the geodesic reaches the preset length or reaches the surface boundary. Perform geodesic tracing on multiple starting points on the three-dimensional surface model to obtain a set of geodesics. Based on the directions of these geodesics, identify the laying direction field.Take the weighted principal curvature direction of each surface element as the ply direction of that surface element. Finally, a ply direction field data is generated, which assigns a ply direction to each triangular surface element.

[0125] Preferably, step S27 includes the following steps:

[0126] Step S271: Select the maximum principal curvature according to the weighted principal curvature direction data, and take the corresponding direction as the initial geodesic direction;

[0127] Step S272: Set the geodesic step size to 3 mm, and calculate the coordinates of the next target point reached after advancing one such geodesic step length in this direction based on the initial geodesic direction through the three-dimensional surface model, to obtain the coordinates of the next target point;

[0128] Step S273: Search for the surface element containing this target point through the three-dimensional surface model based on the coordinates of the next target point, and project the target point onto this surface element along the surface element normal vector direction to calculate the projection distance; when the projection distance is greater than 1 mm, it is determined that the projection fails and the tracking of this geodesic is stopped, otherwise an initial geodesic segment with a length of 5 mm is generated on the three-dimensional surface model, and this target point is used as the new target point;

[0129] Step S274: Repeat step S272 and step S273 until the geodesic segment reaches the surface boundary or the angle between the geodesic direction and the initial direction exceeds 90°, so as to obtain the initial ply direction line;

[0130] Step S275: Calculate the angle and distance between adjacent ply direction lines according to the initial ply direction line to obtain the adjacent ply direction line relationship data;

[0131] Step S276: According to the adjacent ply direction line relationship data, perform a merging process on adjacent ply direction lines with an angle less than the angle threshold and a distance less than the distance threshold to generate merged ply direction line data, where the angle threshold is set to 15°, the distance threshold is 2 mm, and the merging direction of the new ply direction line is the average direction of the two initial ply direction lines;

[0132] Step S277: Use the geometric center coordinates of the surface element as interpolation nodes, and perform ply direction field interpolation processing on the merged ply direction line data to generate ply direction field data.

[0133] In the embodiments of the present invention, the weighted principal curvature direction data is traversed. For each triangular element, its weighted principal curvature direction is obtained by weighted averaging of two principal curvature directions. Recall the previous calculation, each element has two principal curvatures λ1 and λ2, and the corresponding eigenvectors (i.e., principal directions) v1 and v2. Compare the absolute values |λ1| and |λ2| of the two principal curvatures, and select the eigenvector corresponding to the principal curvature with the larger absolute value as the initial geodesic direction. For example, if |λ1| > |λ2|, then select v1 as the initial geodesic direction; otherwise, select v2 as the initial geodesic direction. This selection is based on the principle that the direction with a larger curvature indicates that the surface bends more severely in this direction and is more suitable as the starting direction of the laminate. Store the selected initial geodesic direction as the initial geodesic direction data for each element. Select a starting element, and starting from the geometric center of this element, advance 3 mm (geodesic step length) along the initial geodesic direction determined in the previous step on the three-dimensional surface model of the engine hood. Use a numerical integration method, such as the fourth-order Runge-Kutta method, to calculate the coordinates of the next target point. The specific operation is as follows: Substitute the coordinates of the current point (starting point) and the initial geodesic direction into the Runge-Kutta formula. This formula estimates the position of the next target point by calculating the slopes (i.e., directional derivatives) of several points in the current point and its neighborhood. It should be noted that the "advance" here means advancing along the geodesic of the surface, not in a straight line. Therefore, it is necessary to continuously project the advancing direction onto the surface to ensure that the calculated points are always on the surface. Store the three-dimensional coordinates (x_next, y_next, z_next) of the calculated next target point as the coordinates of the next target point. Using the coordinates (x_next, y_next, z_next) of the next target point calculated in the previous step, find the triangular element containing this point in the three-dimensional surface model. A point-triangle inclusion relationship test algorithm, such as the barycentric coordinate method, can be used. Traverse all triangular elements and calculate the barycentric coordinates of the target point relative to each triangular element. If the barycentric coordinates of the target point relative to a certain triangular element are all greater than or equal to 0 and less than or equal to 1, then the target point is inside this triangular element. After finding the element containing the target point, calculate the projection distance from the target point to this element. Project the target point onto the plane where the element is located along the normal vector direction of this element. Calculate the Euclidean distance between the target point and the projection point. If this distance is greater than 1 mm, it means that the target point deviates too far from the surface and the geodesic tracking is inaccurate, and stop the tracking of this geodesic. If the projection distance is less than or equal to 1 mm, then the projection is considered successful. On the three-dimensional surface model, connect the starting point and the target point to generate a geodesic segment with a length of 3 mm (geodesic step length, not the straight-line distance) (because the cumulative geodesic distance in step 272 is already 3 mm).To generate a geodesic with a length of 5 mm, step S272 needs to be repeated once, and then move forward 3 mm. At this time, the cumulative length is 6 mm, which is greater than 5 mm. A geodesic segment is generated by connecting the starting point and the latest target point. The latest target point is used as the new starting point, and steps S272 and S273 are continuously repeated. Move forward along the new geodesic direction and perform projection and judgment. Keep iterating until the geodesic segment reaches the boundary of the engine hood surface model, or the angle between the current geodesic direction and the initial geodesic direction exceeds 90 degrees (indicating that the geodesic has a large bend and is no longer suitable as the ply direction). Connect the traced geodesic segments to form an initial ply direction line. Select several sampling points on it (for example, select a sampling point every 5 mm). For each sampling point, find the sampling point on another initial ply direction line that is the closest within its neighborhood (for example, within a range with a radius of 10 mm centered on the sampling point). Calculate the distance between these two sampling points. At the same time, calculate the angle between the tangent vectors of the two ply direction lines at these two sampling points (which can be calculated using the dot product of vectors). Average the distance and angle data of all sampling points to obtain the average distance and average angle between these two initial ply direction lines. Repeat this calculation for all pairs of adjacent initial ply direction lines to obtain the average distance and average angle between all adjacent ply direction lines. Set the angle threshold to 15 degrees and the distance threshold to 2 mm. Traverse all pairs of adjacent ply direction lines. If the average angle between two ply direction lines is less than 15 degrees and the average distance is less than 2 mm, then it is considered that these two ply direction lines are too close and have similar directions, and they need to be merged. The merging operation is as follows: Select several corresponding points on the two initial ply direction lines respectively (for example, select the closest points as the corresponding points). For each pair of corresponding points, calculate their geometric center and use this center point as a point on the new ply direction line. At the same time, calculate the average vector of the tangent vectors of the two initial ply direction lines at this corresponding point and use this average vector as the direction of the new ply direction line at this point. Connect the center points of all corresponding points to form a new ply direction line. Repeat this merging operation until there are no pairs of ply direction lines that meet the merging conditions. Store all the merged ply direction lines as merged ply direction line data. Use the geometric center coordinates (xc, yc, zc) of all triangular elements as interpolation nodes. For each element, find the several closest merged ply direction lines (for example, 3 lines) within its neighborhood (for example, within a range with a radius of 15 mm centered on the geometric center of the element). Calculate the ply direction of this element according to the sampling points and directions of these ply direction lines in the neighborhood. The inverse distance weighted interpolation method can be used: Calculate a weight according to the distance from each sampling point to the geometric center of the element; the closer the distance, the greater the weight. Then, perform a weighted average on the directions of all sampling points to obtain the ply direction of this element.Repeat this interpolation operation for all triangular elements to obtain the ply orientation of each element. Store the ply orientations of all elements as a ply orientation field data.

[0134] As an example of the present invention, refer to Figure 2 shown in Figure 1 the detailed implementation step flow diagram of step S3 in

[0135] Step S31: Obtain the prepreg property parameters and the preliminary laying process parameters of the prepreg;

[0136] In the embodiment of the present invention, the prepreg property parameters are obtained from the material database of the industrial design system. For example, if the prepreg used is a carbon fiber / epoxy prepreg, the property parameters to be obtained include: the thickness of a single-layer prepreg (e.g., 0.125 mm), the tensile modulus E1 in the fiber direction (e.g., 150 GPa), the tensile modulus E2 perpendicular to the fiber direction (e.g., 10 GPa), the in-plane shear modulus G12 (e.g., 5 GPa), the Poisson's ratio ν12 (e.g., 0.3), and the density ρ of the prepreg (e.g., 1600 kg / m 3 ³). At the same time, the preliminary laying process parameters of the prepreg are obtained from the process parameter database. These parameters include: the laying pressure P (e.g., 0.5 MPa), the laying temperature T (e.g., room temperature 25 °C), the laying speed V (e.g., 10 mm / s), the radius R of the laying roller (e.g., 50 mm), the contact angle θ between the laying roller and the laying surface (e.g., 10 degrees), etc.

[0137] Step S32: Construct a local finite element model based on the element according to the ply orientation field data and the prepreg property parameters;

[0138] In the embodiment of the present invention, each triangular element is used as a finite element unit. For each element, according to its ply orientation, the material property parameters (E1, E2, G12, ν12) of the prepreg are transformed into the local coordinate system of the element. The specific operation is as follows: Construct a material property matrix D, which has a larger stiffness (E1) in the fiber direction and a smaller stiffness (E2) perpendicular to the fiber direction. Then, according to the angle between the ply orientation and the local coordinate system of the element, a rotation transformation is performed on the material property matrix D to obtain the material property matrix D' of the element in the local coordinate system. The geometric shape of each element is defined by the coordinates of its three vertices. Each vertex has three degrees of freedom (displacements in the x, y, and z directions). Therefore, each triangular element has 9 degrees of freedom. According to the finite element theory, the stiffness matrix K of each element can be calculated. The stiffness matrix K is a 9x9 matrix that describes the deformation characteristics of the element under the action of external forces. Assemble the stiffness matrices K of all elements into a global stiffness matrix.

[0139] Step S33: Analyze the element loading sequence of the local finite element model according to the preliminary laying process parameters of the prepreg to generate element layer laying loading sequence data;

[0140] In the embodiment of the present invention, a loading sequence based on the laying path is adopted: according to the laying path (geodesic) generated in the previous step, arrange the elements on the laying path in the laying order. For each laying path, starting from the starting point, add the elements to the loading sequence one by one along the path. For multiple laying paths, they can be sorted according to the starting point position or path length of the path to determine the loading order between the paths. For example, it can start from the top of the engine hood and lay layer by layer along multiple laying paths. In the loading sequence, each element appears only once. The total length of the loading sequence is equal to the total number of elements. Store the generated element loading sequence as element layer laying loading sequence data.

[0141] Step S34: Calculate the element laying load according to the laying process parameters to generate discrete element load data;

[0142] In the embodiment of the present invention, for the laying pressure, distribute it evenly on the area of the element. Calculate the area Area of the element (which has been calculated in the previous step). The pressure load F_pressure received by the element = P×Area, where P is the laying pressure, and the direction is perpendicular to the element plane and points to the inside of the element. For the contact force caused by the laying roller, it needs to be calculated according to parameters such as the radius R, contact angle θ, and laying speed V of the laying roller. Since the contact between the laying roller and the curved surface is a complex process, a simplified model can be used for calculation. For example, the contact force can be decomposed into a normal force perpendicular to the curved surface and a tangential force along the laying direction. The magnitude of the normal force is related to the laying pressure and the contact area, and the magnitude of the tangential force is related to the normal force, the friction coefficient, and the laying speed. Combine the calculated pressure load and contact force load (including normal force and tangential force) of each element into a load vector F. Store the load vectors of all elements as discrete element load data.

[0143] Step S35: Perform a laying process simulation on the local finite element model based on the element layer laying loading sequence data and the discrete element load data, solve the stress, strain, and displacement distributions on each element to obtain element mechanical state data, where the element mechanical state data includes element stress field data, element displacement field data, and element shear strain data;

[0144] In the embodiment of the present invention, a simulation of the layup process is performed on the constructed local finite element model. The incremental loading method is adopted to simulate the process of laying prepreg layer by layer on the surface of the engine hood. According to the surface element layup loading sequence, each surface element is activated in turn. For each activated surface element, its load vector F is added to the global load vector F_global. Then, the global equilibrium equation is solved: K_global×U = F_global. Where K_global is the global stiffness matrix, U is the global displacement vector, and F_global is the global load vector. Since K_global is a large sparse matrix, a sparse matrix solver (such as Cholesky decomposition or conjugate gradient method) can be used to solve this equation to obtain the global displacement vector U. Based on the global displacement vector U, the displacement, strain, and stress of each surface element can be calculated. The surface element displacement is the displacement of the three vertices of the surface element. The surface element strain can be calculated from the surface element displacement and the geometry of the surface element. The surface element stress can be calculated from the surface element strain and the material property matrix D' of the surface element. After each loading increment, the global stiffness matrix K_global and the global load vector F_global are updated, and the equilibrium equation is solved again. This process is repeated until all surface elements are activated. The displacement, strain, and stress data of each surface element at each loading increment are stored as surface element mechanical state data, including surface element stress field data (stress components), surface element displacement field data (displacement components), and surface element shear strain data (shear strain components).

[0145] Step S36: Perform process defect analysis based on the surface element mechanical state data to obtain layup process defect data, where the layup process defect data includes wrinkling risk surface element data, gap risk area data, and fiber misalignment risk surface element data.

[0146] In the embodiment of the present invention, the shear strain value of each surface element is obtained. A shear strain threshold is set. This threshold needs to be adjusted according to the performance of the prepreg and the layup process. The selection of the threshold can refer to the shear strength of the material and the empirical data during the layup process. The shear strain data is traversed. For each surface element, if its shear strain is greater than the set threshold, it is determined that the surface element has a fiber misalignment risk. The information of the surface element with a fiber misalignment risk is stored in the fiber misalignment risk surface element data. The fiber misalignment risk surface element data can include information such as the index, coordinates, and shear strain value of the surface element. For example, if the shear strain of a certain surface element is 0.05 and the set threshold is 0.04, it is determined that the surface element has a fiber misalignment risk, and its information is added to the fiber misalignment risk surface element data. The fiber misalignment risk surface element data can help process engineers identify areas prone to fiber misalignment and take corresponding measures for prevention. The preventive measures can include adjusting the layup path, reducing the layup speed, or optimizing the fiber arrangement of the prepreg.

[0147] Preferably, the process defect analysis based on the mechanical state data of the surface element includes:

[0148] Calculating the deviation angle between the principal stress direction and the ply direction of each surface element according to the surface element stress field data to obtain stress-ply deviation angle data;

[0149] When the stress-ply deviation angle data is greater than 15° and the compressive stress value of the surface element is greater than 0.5 MPa, it is determined that there is a risk of wrinkling in the surface element to obtain wrinkling risk surface element data;

[0150] Calculating the relative displacement between adjacent surface elements according to the surface element displacement field data. When the relative displacement is greater than 0.8 mm, it is determined that there is a risk of gap in the area to obtain gap risk area data;

[0151] Calculating the shear strain of each surface element according to the surface element shear strain data. When the shear strain is greater than the preset strain threshold, it is determined that there is a risk of fiber misalignment in the surface element to obtain fiber misalignment risk surface element data.

[0152] In the embodiments of the present invention, for each triangular element, its stress field data includes the stress tensor of the element in the local coordinate system. This stress tensor is a 3x3 symmetric matrix, including three normal stress components (σx, σy, σz) and three shear stress components (τxy, τyz, τzx). Since the element is two-dimensional, the stress component σz perpendicular to the plane of the element and the shear stress components τyz, τzx can be ignored, and only the in-plane stress state, namely σx, σy, and τxy, is considered. Based on these three stress components, the principal stresses σ1 and σ2 of the element, as well as the principal stress direction θp, can be calculated. The principal stress direction θp is the angle relative to the x-axis of the local coordinate system of the element. At the same time, the ply direction θf of the element (also the angle relative to the x-axis of the local coordinate system of the element) is obtained from the ply direction field data. Calculate the deviation angle between the principal stress direction θp and the ply direction θf: Δθ = |θp - θf|. If Δθ is greater than 180 degrees, then take its supplementary angle (360 degrees - Δθ). Store the deviation angle Δθ of all elements as stress-ply deviation angle data. For each triangular element, check whether its deviation angle Δθ is greater than 15 degrees. At the same time, check the principal stresses σ1 and σ2 of the element to determine whether there is a compressive stress (i.e., the principal stress is negative). If there is a compressive stress and its absolute value is greater than 0.5 MPa, then it is considered that the element meets the wrinkle risk criterion. Store the indices and relevant information (such as deviation angle, compressive stress value) of all elements that meet the wrinkle risk criterion as wrinkle risk element data. These data indicate that at these elements, due to the large deviation between the stress direction and the ply direction and the existence of a large compressive stress, the prepreg will wrinkle. Traverse the element displacement field data. For each triangular element, its displacement field data includes the displacement vectors of the three vertices of the element. For each pair of adjacent elements (i.e., two elements sharing an edge), calculate the relative displacement between them. The specific operation is as follows: Select two vertices on the shared edge of the adjacent elements. Calculate the difference between the displacement vectors of these two vertices after lamination. This displacement difference vector represents the relative displacement between these two vertices. Calculate the magnitude of this displacement difference vector, that is, the magnitude of the relative displacement. If this relative displacement is greater than 0.8 mm, then it is considered that there is a gap risk between these two adjacent elements. Store the indices and relevant information (such as the magnitude of the relative displacement) of all adjacent element pairs with a gap risk as gap risk region data. Traverse the element shear strain data. For each triangular element, its shear strain data includes the shear strain γxy of the element in the local coordinate system. According to the material properties of the prepreg, set a shear strain threshold γ_threshold (for example, for carbon fiber / epoxy prepreg, it can be set to 0.05, that is, 5% shear strain). Check whether the absolute value of the shear strain γxy of each element is greater than this threshold γ_threshold. If it is greater than the threshold, then it is considered that the element has a fiber misalignment risk.Store the indices and related information (such as shear strain values) of all the surface elements at risk of fiber misalignment as fiber misalignment risk surface element data.

[0153] As an example of the present invention, refer to Figure 3 shown in Figure 1 the detailed implementation step flow diagram of step S4 in

[0154] Step S41: Adjust the laying direction according to the wrinkle risk surface element data to obtain optimized laying direction data;

[0155] In the embodiment of the present invention, for each surface element at risk of wrinkle, review the previously calculated stress-laying deviation angle data. This data indicates that at these surface elements, the principal stress direction deviates greatly from the laying direction. To reduce the deviation, the laying direction needs to be adjusted. A stress-direction-based adjustment method is adopted: adjust the laying direction of this surface element to be closer to its principal stress direction. The specific operation is as follows: Calculate the weighted average of the principal stress direction θp and the original laying direction θf as the new laying direction θn. The weight can be determined according to the size of the deviation angle Δθ. For example, the weight w = 1 - Δθ / 90° (if Δθ is greater than 90 degrees, then w = 0). The new laying direction θn = w×θf + (1 - w)×θp. Adjust the laying direction for all surface elements at risk of wrinkle. Update the adjusted laying direction into the laying direction field data to obtain optimized laying direction data.

[0156] Step S42: Adjust the laying pressure and the pressure holding time according to the gap risk area data to obtain optimized laying pressure-time parameters;

[0157] In the embodiment of the present invention, the laying pressure and the pressure holding time are adjusted. For each pair of adjacent surface elements at risk of gap, analyze the magnitude and direction of their relative displacement. If the relative displacement is large, it means that the contact between the prepregs during the laying process is not tight enough. To improve the contact, the laying pressure can be appropriately increased. For example, the initial laying pressure P can be increased by an increment ΔP (such as 0.1 MPa). At the same time, the pressure holding time can be extended to allow the prepregs to have more time to deform and flow to fill the gap. For example, the initial pressure holding time T can be increased by an increment ΔT (such as 5 seconds). For different gap risk areas, different pressure and time increments can be set according to the magnitude and direction of their relative displacement. Store the adjusted laying pressure and pressure holding time parameters as optimized laying pressure-time parameters.

[0158] Step S43: Adjust the laying speed and the heating temperature according to the fiber misalignment risk surface element data to obtain optimized laying speed-temperature parameters;

[0159] In the embodiments of the present invention, for each surface element at risk of fiber misalignment, the magnitude of its shear strain is analyzed. If the shear strain is too large, it indicates that excessive shear deformation has occurred in the prepreg during the laying process. To reduce the shear deformation, the laying speed can be decreased. For example, the initial laying speed V can be reduced by an increment ΔV (e.g., 2 mm / s). At the same time, if the laying process of the prepreg requires heating, the heating temperature can be appropriately adjusted. For example, if the initial heating temperature is low, the heating temperature can be appropriately increased to reduce the viscosity of the prepreg, making it easier to flow, thereby reducing the shear strain. Conversely, if the initial heating temperature is high, the heating temperature can be appropriately decreased to increase the viscosity of the prepreg, making it less likely to undergo shear deformation. For different surface elements at risk of fiber misalignment, different speed and temperature adjustment amounts can be set according to the magnitude of their shear strain. The adjusted laying speed and heating temperature parameters are stored as optimized laying speed-temperature parameters.

[0160] Step S44: Combine the optimized ply orientation data, optimized ply pressure-time parameters, and optimized laying speed-temperature parameters for intelligent ply laying process parameter combination to achieve the ply laying design of the industrial structural part housing.

[0161] In the embodiments of the present invention, a parameter database is established, which contains the optimized ply orientation data, optimized ply pressure-time parameters, and optimized laying speed-temperature parameters. This database needs to be able to index parameters according to different regions of the housing model. For example, the housing model can be divided into multiple regions, and each region corresponds to a set of optimized process parameters. Then, an intelligent ply laying parameter combination algorithm is developed. This algorithm can automatically select appropriate process parameters according to the geometric shape, material properties, and ply laying requirements of the housing model. The algorithm needs to consider the following factors: smoothness of parameter transition between adjacent regions, minimization of material usage, minimization of laying time, and process feasibility, etc. The output of the parameter combination algorithm is an ordered list containing the process parameters of all regions, and this list can be used as the control instruction for the ply laying equipment. For example, the algorithm outputs the following instructions: Region 1: Ply orientation 30 degrees, ply pressure 0.12 MPa, pressure holding time 12 seconds, laying speed 0.15 m / s; Region 2: Ply orientation 45 degrees, ply pressure 0.1 MPa, pressure holding time 10 seconds, laying speed 0.2 m / s; and so on. Next, the output result of the parameter combination algorithm is converted into control codes that can be recognized by the ply laying equipment. The control codes need to contain all necessary instructions, such as starting point coordinates, ending point coordinates, ply orientation, ply pressure, pressure holding time, and laying speed, etc. Finally, the control codes are uploaded to the control system of the ply laying equipment to start the automatic ply laying process. The ply laying equipment will accurately control parameters such as the movement trajectory of the ply laying head, ply pressure, and heating temperature according to the instructions of the control codes, thereby achieving the ply laying design of the engine hood.

[0162] Preferably, the present invention further provides an industrial design system that executes the industrial design method as described above. The industrial design system includes:

[0163] An industrial structure modeling module, configured to obtain an industrial structure part to be laminated; reconstruct a surface model to be laminated for the industrial structure part to be laminated to generate a three-dimensional surface model; perform surface element discretization processing on the three-dimensional surface model to generate data of discrete geometric surface elements to be laminated;

[0164] A lamination direction field generation module, configured to perform neighborhood surface element feature analysis on the data of discrete geometric surface elements to be laminated to obtain local neighborhood surface element feature data; perform geodesic tracing based on the local neighborhood surface element feature data, and then perform lamination direction field identification to generate lamination direction field data;

[0165] A defect analysis module, configured to perform lamination process simulation based on the lamination direction field data, solve the stress, strain, and displacement distributions on each surface element to obtain surface element mechanical state data; perform process defect analysis based on the surface element mechanical state data to obtain lamination process defect data;

[0166] An intelligent parameter optimization module, configured to perform intelligent lamination process parameter combination based on the lamination process defect data to achieve the lamination design of the industrial structure part housing.

[0167] The present application lies in that through precise three-dimensional surface model reconstruction and surface element discretization processing, the shell geometric information of the industrial structure part to be laminated can be accurately obtained. By using the geodesic tracing technology, an optimized lamination path is generated according to the geometric characteristics of the surface, reducing fiber deformation and avoiding the deviations and errors that occur in traditional manual design, thereby improving the accuracy and efficiency of the lamination design. At the same time, based on the defect analysis results, intelligent process parameter optimization is performed to achieve automatic adjustment and optimization of the lamination process, improving the efficiency and quality of the lamination design. By optimizing process parameters such as the lamination direction, lamination pressure, laying speed, and heating temperature, it is ensured that the prepreg can better adapt to the geometric characteristics of complex surfaces during the laying process, reducing deformations such as fiber bending, torsion, and misalignment, thereby improving the mechanical properties and appearance quality of the final industrial structure part. Especially when dealing with areas with high curvature changes, the generation of defects such as wrinkles, fiber breakage, or overlap can be effectively avoided. The present application is particularly suitable for the lamination design of industrial structure parts with complex surface shapes. In the fields of aerospace, automotive manufacturing, high-end medical devices, etc., the application of complex surface structural parts is becoming more and more widespread. The present application can effectively solve the problems of traditional lamination processes when dealing with complex surfaces, meet the manufacturing requirements of high-precision and high-quality industrial structure parts in these fields, and has broad application prospects.

[0168] Therefore, in all respects, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Thus, all changes that fall within the meaning and scope of the equivalent elements of the application documents are intended to be encompassed within the present invention.

[0169] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features invented herein.

Claims

1. An industrial design method, characterized in that: The following steps are involved: Step S1: obtaining an industrial structural part to be laid; reconstructing a surface model of the industrial structural part to be laid to generate a three-dimensional surface model; performing surface element discretization processing according to the three-dimensional surface model to generate discrete geometric surface element data to be laid; Step S2: performing a surface element feature analysis in the neighborhood according to the discrete geometric surface element data of the layer to be laid, and obtaining local neighborhood surface element feature data; According to the local neighborhood surface element feature data, geodesic line tracking is performed, and then the ply direction field is identified to generate ply direction field data; Step S3: simulating the ply laying process according to the ply laying direction field data, solving the stress, strain and displacement distribution on each surface element, and obtaining the surface element mechanical state data; Perform process defect analysis based on the surface element mechanical state data to obtain the ply process defect data; Step S4: Based on the layup process defect data, intelligent layup process parameter combination is performed to realize the shell layup design of industrial structural parts.

2. The industrial design method according to claim 1, characterized in that: The step S1 of reconstructing the surface model of the industrial structural part to be laid includes: Obtain CAD models of industrial structural parts; Use 3D scanning equipment to scan the geometric information of the shell surface of the industrial structural parts to be laid, and obtain the surface point cloud data of the original structural parts; Perform multi-view point cloud registration on the original structural component surface point cloud data to obtain optimized structural component surface point cloud data; The surface model of the layer to be laid is reconstructed based on the optimized surface point cloud data of the structural parts and the CAD model of the industrial structural parts to generate a three-dimensional surface model.

3. The industrial design method according to claim 1, characterized in that: The process of performing the surface element discretization process according to the three-dimensional surface model in step S1 includes: Perform differential geometry calculation on the ply surface of the three-dimensional surface model to generate ply surface curvature distribution data; Classify the surface area types according to the curvature distribution data of the pavement surface, and generate surface area type data; The initial surface element size of the three-dimensional surface model is set by using the surface area type data, and then the surface mesh is divided to generate initial surface mesh division data; According to the initial pavement mesh division data, the minimum pavement radius constraint and the fiber direction control accuracy constraint are performed to obtain the constrained pavement mesh division data; The discrete geometric surface data of the layer to be paved are extracted according to the constrained paving meshing data.

4. The industrial design method according to claim 3, characterized in that: The step of extracting discrete geometric facet data of the layer to be paved according to the constrained paving meshing data comprises: The three-dimensional surface model is processed into surface elements by constraining the pavement meshing data to obtain the pavement discrete surface element data; Read the coordinates of the facet vertices according to the discrete facet data of the pavement; Calculate the coordinates of the geometric center of the face element according to the coordinates of the face element vertices; Calculate the surface element normal vector according to the discrete surface element data of the pavement; Calculate the area of ​​the facet according to the discrete facet data of the pavement; The geometric center coordinates of the facet, the facet normal vector and the facet area are integrated with the facet geometric features of the layer to be laid to generate discrete geometric facet data of the layer to be laid.

5. The industrial design method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: constructing a KD tree spatial index according to the geometric center coordinates of the facets in the discrete geometric facet data to be laid; Step S22: Based on the KD tree spatial index, taking the geometric center coordinates of each facet as the center, searching for facets in its neighborhood to obtain a facet neighborhood set; Step S23: extracting the geometric center coordinates and normal vectors of all facets in the target facet neighborhood according to the facet neighborhood set, and obtaining local neighborhood facet feature data; Step S24: performing local surface fitting according to the local neighborhood face element feature data, and then calculating the average curvature of each neighborhood face element to obtain neighborhood curvature distribution data; Step S25: solving the neighborhood surface Gaussian eigenvalues ​​and neighborhood principal curvature direction vectors of the Gaussian mapping according to the neighborhood curvature distribution data; Step S26: using the area of ​​the facet in the discrete geometric facet data of the to-be-laid layer as a weight, and performing a weighted calculation of the principal curvature direction according to the Gaussian eigenvalue of the neighborhood surface and the principal curvature direction vector of the neighborhood, to generate weighted principal curvature direction data; Step S27: geodesic tracing is performed on the three-dimensional surface model by using the weighted principal curvature direction data, and then ply direction field recognition is performed to generate ply direction field data.

6. The industrial design method according to claim 5, characterized in that: Step S27 includes the following steps: Step S271: selecting the maximum principal curvature according to the weighted principal curvature direction data, and using the corresponding direction as the initial geodesic direction; Step S272: Setting the geodesic step length to 3 mm, and calculating the coordinates of the next target point reached after advancing one geodesic step length in the direction through the three-dimensional surface model based on the initial geodesic direction, to obtain the coordinates of the next target point; Step S273: based on the coordinates of the next target point, search for the face element containing the target point through the three-dimensional surface model, and project the target point onto the face element along the face element normal vector direction to calculate the projection distance; when the projection distance is greater than 1 mm, it is determined that the projection fails and the tracking of the geodesic line is stopped, otherwise an initial geodesic line segment with a length of 5 mm is generated on the three-dimensional surface model, and the target point is used as the new target point; Step S274: repeating steps S272 and S273 until the geodesic line segment reaches the surface boundary or the angle between the geodesic line direction and the initial direction exceeds 90°, thereby obtaining an initial ply direction line; Step S275: calculating the angle and distance between adjacent ply direction lines according to the initial ply direction lines, and obtaining the adjacent ply direction line relationship data; Step S276: according to the adjacent ply direction line relationship data, adjacent ply direction lines whose angle is less than the angle threshold and whose distance is less than the distance threshold are merged to generate merged ply direction line data, wherein the angle threshold is set to 15°, the distance threshold is set to 2 mm, and the merged direction of the new ply direction line is the average direction of the two initial ply direction lines; Step S277: using the geometric center coordinates of the face element as interpolation nodes, performing ply direction field interpolation processing on the merged ply direction line data to generate ply direction field data.

7. The industrial design method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: obtaining prepreg property parameters and preliminary prepreg placement process parameters; Step S32: constructing a local finite element model based on surface elements according to the ply direction field data and prepreg property parameters; Step S33: performing a panel loading sequence analysis on the local finite element model according to the preliminary placement process parameters of the prepreg, and generating panel ply loading sequence data; Step S34: performing calculation of the surface element placement load according to the placement process parameters to generate discrete surface element load data; Step S35: simulating the ply laying process of the local finite element model based on the panel element ply loading sequence data and the discrete panel element load data, solving the stress, strain and displacement distribution on each panel element, and obtaining the panel element mechanical state data, wherein the panel element mechanical state data includes the panel element stress field data, the panel element displacement field data and the panel element shear strain data; Step S36: Perform process defect analysis based on the surface element mechanical state data to obtain ply process defect data, wherein the ply process defect data includes wrinkle risk surface element data, gap risk area data, and fiber dislocation risk surface element data.

8. The industrial design method according to claim 7, characterized in that: The process defect analysis according to the surface element mechanical state data includes: Calculate the deviation angle between the principal stress direction of each surface element and the ply direction according to the surface element stress field data to obtain stress-ply deviation angle data; When the stress-layment deviation angle data is greater than 15° and the compressive stress value of the surface element is greater than 0.5MPa, it is determined that the surface element has a wrinkle risk, and the wrinkle risk surface element data is obtained; The relative displacement between adjacent facets is calculated based on the facet displacement field data. When the relative displacement is greater than 0.8 mm, it is determined that there is a gap risk in the area, and the gap risk area data is obtained. The shear strain of each facet is calculated according to the facet shear strain data. When the shear strain is greater than a preset strain threshold, it is determined that the facet has a fiber dislocation risk, and fiber dislocation risk facet data is obtained.

9. The industrial design method according to claim 7, characterized in that: Step S4 includes the following steps: Step S41: adjusting the ply direction according to the wrinkle risk facet metadata to obtain optimized ply direction data; Step S42: adjusting the ply pressure and holding time according to the gap risk area data to obtain optimized ply pressure-time parameters; Step S43: adjusting the laying speed and heating temperature according to the fiber dislocation risk facet metadata to obtain optimized laying speed-temperature parameters; Step S44: combining the optimized laying direction data, the optimized laying pressure-time parameters and the optimized laying speed-temperature parameters into intelligent laying process parameter combinations to achieve the shell laying design of industrial structural parts.

10. An industrial design system, characterized in that: For executing the industrial design method according to claim 1, the industrial design system comprises: The industrial structure modeling module is used to obtain the industrial structure parts to be laid; reconstruct the surface model of the industrial structure parts to be laid to generate a three-dimensional surface model; discretize the surface elements according to the three-dimensional surface model to generate discrete geometric surface element data to be laid; The ply direction field generation module is used to perform surface element feature analysis in the neighborhood according to the discrete geometric surface element data of the ply to be laid, and obtain local neighborhood surface element feature data; perform geodesic line tracking according to the local neighborhood surface element feature data, and then perform ply direction field recognition to generate ply direction field data; The defect analysis module is used to simulate the layup process according to the layup direction field data, solve the stress, strain and displacement distribution on each surface element, and obtain the surface element mechanical state data; perform process defect analysis according to the surface element mechanical state data to obtain the layup process defect data; The intelligent parameter optimization module is used to combine intelligent layup process parameters based on layup process defect data to achieve the layup design of industrial structural parts shell.

Images