A method for overall optimization of a multi-directional layup composite curved component layup process

Abstract

A multi-directional layer composite curved surface component laying process overall optimization method, characterized in that, suitable for laying process optimization of complex curved surface composite material component, comprising the following steps: 1) according to the given fiber reference direction information of multi-directional layer design, the frame field composed of fiber reference direction is constructed on the target curved surface; 2) according to the laying process target and constraint, the frame field is optimized, and the optimized frame field is obtained; 3) the geometric characteristics of the target curved surface are combined to analyze the layability, a plurality of segmentation lines are formed on the curved surface according to the optimized frame field, the adaptability region division of the target curved surface is carried out, and the curved surface partition laying process result is obtained, which is used for guiding subsequent automatic laying track generation or manual laying layout design. The application changes the optimization object from single layer vector field to frame field composed of multi-directional layer, guarantees the interlayer included angle, and meets the laying quality.

Description

Technical Field

[0001] This invention belongs to the field of composite material manufacturing technology, and particularly relates to a method for optimizing the layup process of a complex curved surface component made of carbon fiber reinforced composite material, specifically a method for overall optimization of the layup process of a multi-directional layup composite material curved surface. Background Technology

[0002] Carbon fiber reinforced composites possess advantages such as high specific strength and high specific stiffness, and are widely used in the manufacturing of high-performance structures in aerospace and other fields. To leverage their anisotropic load-bearing advantages, composite components typically require multi-directional layup design, and the predetermined layup direction is achieved through automated fiber placement processes. However, the curvature variations and geometric boundaries of complex surfaces can limit fiber path generation, easily leading to defects such as fiber orientation deviation, wrinkles, and bridging, thus affecting the forming quality of automatically placed components. Existing layup optimization methods mostly focus on a single layup direction, such as the fixed angle method, geodesic method, reference path offset method, and single vector field optimization method. These methods primarily focus on the geometric layability or local directional control of a single layup, lacking collaborative constraints and overall optimization of the angular relationships between multi-directional layups. This would allow for satisfying the layability of curved surfaces while also ensuring the consistency of angular constraints between multiple layups, thereby improving layup quality and structural performance.

[0003] In summary, this invention provides an overall optimization method for the laying process of multi-directional ply composite curved surfaces, which shifts the optimization object from the vector field of a single ply to the frame field composed of multi-directional plies, ensuring the interlayer angle while satisfying the laying quality. Summary of the Invention

[0004] The purpose of this invention is to address the problem that existing layup optimization methods, which are mostly designed for single-layup directions, cannot adapt to complex curved surfaces and multi-layer layup requirements. The invention proposes an overall optimization method for the layup process of multi-directional layup composite curved surface components.

[0005] The technical solution of this invention is:

[0006] A method for overall optimization of the layup process of multi-directional ply composite curved surface components, characterized in that it is applicable to the optimization of the layup process of complex curved surface composite components, and includes the following steps:

[0007] Step 1) Based on the fiber reference direction information given in the multi-directional layup design, construct a frame field composed of fiber reference directions on the target surface;

[0008] Step 2) Optimize the marker rack area according to the laying process objectives and constraints to obtain the optimized marker rack area;

[0009] Step 3) Analyze the pavingability of the target surface based on its geometric characteristics. Based on the optimized frame field, form several dividing lines on the surface to divide the target surface into adaptive regions and obtain the surface zoning paving process results, which are used to guide the subsequent automatic paving trajectory generation or manual paving layout design.

[0010] Furthermore:

[0011] In step 1), the fiber reference direction information given by the layup design includes a combination of 0°, 90° and ±45° reference directions, or a combination of any given angles; a frame field is constructed on the target surface according to the reference direction information, that is, any point on the surface contains all of the above-mentioned direction information.

[0012] In step 2), the direction variables of each fiber reference direction in the target surface tangent plane in the standard frame field are used as optimization variables to establish a target optimization model containing constraint terms and target terms. The constraint terms are used to limit the acceptable range of deviation of the layup quality and layup performance, including but not limited to one or more of the following: direction deviation constraint, adjacent direction continuity constraint, surface layability constraint, and equipment layup capability constraint. The target terms are used to further optimize the layup effect and performance retention effect on the basis of satisfying the constraint terms, including at least one or more of the following: minimizing direction deviation, optimizing direction field smoothness, and minimizing the risk of layup defects. By adjusting the weights or thresholds of each constraint term and target term, the optimized standard frame field that meets the layup process requirements is obtained.

[0013] The adaptive region division in step 3) is to divide the target surface according to the optimized multi-directional layup direction results and the layup spreadability of the target surface, dividing the target surface into several sub-regions suitable for laying trajectory generation, surface flattening or piece generation respectively; the layability is used to characterize the risk of local stretching, compression, wrinkling or bridging defects in the target surface during the laying or flattening process.

[0014] Details are as follows:

[0015] A method for overall optimization of the layup process of multi-directional ply composite curved surface components is proposed. For complex non-developable curved surface components, the method represents the frame field formed by the multi-directional ply as a globally optimizable representation vector field. Based on layup constraints, multi-objective optimization is performed on the represented vector field. Based on the optimization results, the surface is adaptively partitioned and post-processed, thereby improving the consistency between the optimized ply direction results and the designed interlayer angle constraints. Specifically, the method includes the following steps:

[0016] S1 constructs a frame field composed of fiber reference directions based on the fiber reference direction information given in the multi-directional layup design. Specifically, after generating reference lines based on the bearing direction, a frame field composed of corresponding directional combination information is generated based on the reference lines.

[0017] S2 optimizes the benchmark field based on the laying process objectives and constraints to obtain the optimized benchmark field. The surface and theoretical direction are conformally mapped to a two-dimensional plane, and the representation vector field is obtained through a rotationally symmetric representation model. The surface model is input into a neural network solver, and the optimization results with corresponding optimization weights are obtained by setting the loss function and parameters. The corresponding benchmark field is generated based on the optimization results, and the benchmark field is implicitly represented using a potential energy field.

[0018] S3 analyzes the layability of the target surface based on its geometric characteristics. According to the optimized frame field, several dividing lines are formed on the surface to adaptively divide the target surface into regions, obtaining the surface laying process results to guide subsequent automatic laying trajectory generation or manual laying layout design. The equipotential lines of the potential energy field are extracted as dividing boundaries, and the surface is divided into regions using a bisection method to obtain partitioning results that meet the set layability threshold. During the partitioning process, when the partition scale is less than or equal to the set minimum partition threshold, further subdivision stops. For automatic laying, post-processing involves path generation, specifically generating equal-width lines for different partitioning results and performing generalization optimization, using the optimized result as the laying trajectory for that region. For manual laying, post-processing involves surface flattening, specifically flattening different partitioning results, and inputting the layout into a cutting machine for cutting.

[0019] The beneficial effects of this invention are:

[0020] This invention shifts the optimization target from a vector field of a single ply to a frame field composed of multi-directional plies, ensuring both the interlayer angle and the ply quality. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the process of the present invention.

[0022] Figure 2 This is a schematic diagram illustrating the principle of implicit representation of layup direction based on potential energy field in this invention.

[0023] Figure 3 This is a schematic diagram of the simulated curved surface of the aircraft inlet to be laid and its theoretical layup direction in an embodiment of the present invention.

[0024] Figure 4 This is a schematic diagram of the surface conformal mapping and multi-directional plywood field rotational symmetry characterization process in an embodiment of the present invention.

[0025] Figure 5 This is a schematic diagram of the convergence process of multi-objective optimization based on implicit neural representation network in an embodiment of the present invention.

[0026] Figure 6This is a schematic diagram of the results of adaptive zoning of the surface based on the optimized potential energy field in an embodiment of the present invention.

[0027] Figure 7 This is a schematic diagram illustrating the layability distribution under different cutting lines and the post-layout process of automatic and manual laying in an embodiment of the present invention. Detailed Implementation

[0028] To better understand the above-described objects, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted, in particular, that the described embodiments are merely some embodiments of the present invention, and not all embodiments. Without conflict, the embodiments of the present invention and the features obtained from those embodiments can be combined with each other. Furthermore, the present invention can be implemented in other ways than those described herein; therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0029] like Figure 1 As shown.

[0030] A method for overall optimization of the layup process of multi-directional ply composite curved surface components includes the following steps:

[0031] S1 constructs a frame field composed of fiber reference directions based on the fiber reference direction information given by the multi-directional layup design.

[0032] S101. After generating a reference line based on the bearing direction, generate a frame field based on the reference line, which is composed of information from the corresponding directions.

[0033] Specifically, reference lines are obtained based on the working environment and stress points of the components. For discrete surface models, the theoretical reference vector can be obtained based on the tangent vector of the nearest reference line, and this vector is used as the 0° reference to obtain the corresponding other reference direction information. The combination of these information forms the corresponding theoretical frame field.

[0034] S2 optimizes the label rack field according to the laying process objectives and constraints to obtain an optimized label rack field.

[0035] S201. Conformally map the surface and theoretical direction to a two-dimensional plane, and obtain the representation vector field through a rotationally symmetric representation model.

[0036] Specifically, if a certain vector group is rotated in each rotation If the vectors remain overlapping afterward, they are said to have N-rotational symmetry. This is because the vector group that makes up this set of vectors... ,satisfy Therefore, this single vector It can be used as a representation vector. For orthogonal ply directions, a 4-rotational-symmetric representation is used, and its corresponding unique, non-singular representation vector can be mathematically described as: For multi-directional layups with multiple base directions, an N-rotationally symmetric representation can be constructed based on the characteristics of the base directions. Its unique, non-singular representation vector can be mathematically described as follows: .

[0037] S202. Input the surface model into the neural network solver, and obtain the optimization results of the corresponding optimization weights by setting the loss function and parameters.

[0038] Specifically, the neural network solver can employ an implicit neural representation network, whose inputs are the corresponding surface geometric features and the representation vector to be optimized. The output is the optimized representation vector at that position. The given multi-objective optimization involves two or more optimization objectives, constrained by a constructed loss function. The total loss is controlled by weight ratios, and its mathematical expression is: Its loss term includes geometric metrics. and performance metrics Geometric metrics are used to maintain fiber consistency, while performance metrics are used to maintain alignment with theoretical directions.

[0039] S203. Generate the corresponding frame field based on the optimization results, and use the potential energy field to implicitly represent the frame field.

[0040] Specifically, such as Figure 2 As shown, let For one direction of the optimized frame field, gradient potential is now used. To make a substitution, therefore by minimizing and The inner product between them makes the direction of the equipotential lines parallel to the direction of the inner product between them. Maintain consistency. Its energy function can be expressed as: .

[0041] S3 analyzes the pavingability of the target surface by combining its geometric characteristics. Based on the optimized template field, several dividing lines are formed on the surface to divide the target surface into adaptive regions, thereby obtaining the surface paving process results, which are used to guide the subsequent automatic paving trajectory generation or manual paving layout design.

[0042] Specifically, layability characterizes the degree of risk of a target surface experiencing localized stretching, compression, or wrinkling during flattening or laying. This layability can be measured by Gaussian curvature, mapped area distortion, side length distortion, or a weighted combination thereof. Regions with a large absolute value of Gaussian curvature are more prone to deformation during flattening or laying; regions with Gaussian curvature close to zero are closer to developable surfaces and have higher layability.

[0043] S301. Extract the equipotential lines of the potential energy field as the dividing boundary, divide the surface into partitions using the bisection method, and obtain the partitioning results that satisfy the set developable threshold.

[0044] S302. During the process of obtaining partitioning results, when the partitioning scale is less than or equal to the set minimum partitioning threshold, stop further subdivision.

[0045] Specifically, for the potential energy field obtained in step S203, in the potential energy field Candidate equipotential values ​​are selected from the range of values. and with The corresponding equipotential lines serve as candidate segmentation boundaries; the candidate equipotential values ​​are adjusted using a binary search. The process involves dividing the region into two sub-regions that meet a preset expandability threshold. If both sub-regions meet the threshold, the division proceeds directly; otherwise, the best region from one side is selected as the division result. If the current region still does not meet the expandability threshold after reaching the preset minimum partition size, further subdivision is stopped, and the current region is output as a restricted playable region, or marked as requiring manual intervention or special playability processing. The purpose of this step is to limit the number of region divisions from becoming too large.

[0046] S303. For automatic tiling type, post-processing is path generation, which specifically includes generating equal width lines for different partition results and performing generalization optimization on them, and using the optimization result as the tiling trajectory for that area.

[0047] S304. For manual laying type, the post-processing is surface flattening, which specifically includes flattening the surface of different partition results, and then inputting the layout into the cutting machine for cutting.

[0048] To illustrate the effectiveness of the method provided by this invention, the following detailed description of the above technical solution is provided through a specific embodiment:

[0049] As shown in Figures 3 to 7.

[0050] In this embodiment, the object to be laid is a simulated curved surface of an aircraft air intake, and its layup direction combination is [0° / ±45° / 90°]. The reference frame field based on the S1 reference line and its corresponding direction combination information is as follows: Figure 3 As shown. For S201, conformal mapping is performed on the surface model and theoretical ply directions, mapping the three-dimensional surface and its theoretical directions to a two-dimensional parameter plane, while maintaining the relative positional relationship between the theoretical directions and the surface mesh during the mapping process. Furthermore, a 4-rotational symmetric representation model is used to uniformly represent the mapped multi-directional ply frame field, thereby obtaining a unique and unambiguous representation vector field. The representation process is as follows: Figure 4 As shown.

[0051] For S202, after obtaining the representation vector field, multi-objective optimization is performed on the representation vector field. Specifically, for S202, the surface model and the representation vector field are input into an implicit neural representation network, and the optimization results under corresponding weight conditions are obtained by setting a loss function and related optimization parameters. During the optimization process, a loss function containing geometric and performance metrics is constructed to jointly constrain the continuity of the layup direction, the preservation of the theoretical direction, and the consistency of the interlayer angle. By adjusting the weights of different loss terms, an optimization result that satisfies the layup constraints is obtained. This optimization process is as follows: Figure 5 As shown. Further, for S203, a corresponding multi-directional ply template field is generated based on the optimization results, and a potential energy field is constructed to implicitly represent the template field, so that the equipotential lines of the potential energy field are consistent with the optimized ply direction.

[0052] For S3, the layability of the target surface is analyzed based on its geometric characteristics. Several dividing lines are formed on the surface according to the optimized frame field, and the target surface is adaptively divided into regions to obtain the surface laying process results. These results guide subsequent automatic laying trajectory generation or manual laying layout design. Specifically, for S301, equipotential lines of the potential energy field are used as candidate dividing boundaries, and the target surface is divided into regions through iterative segmentation. In this embodiment, the layability threshold is set to 1%. When the sub-regions after segmentation meet the preset layability threshold, the partitioning result is retained. Further, for S302, when the region scale reaches the preset minimum partitioning scale, further subdivision is stopped to avoid an excessive number of partitions affecting the implementation of subsequent laying processes. The partitioning results under different layup directions are as follows: Figure 6 As shown. By Figure 6 As can be seen, the degree of local surface distortion decreases with the reduction of the iterative segmentation region, indicating a decrease in the risk of surface deformation during play and improved playability. For S303 and S304, after obtaining the surface partitioning results, post-processing is performed based on the play type and surface partitioning results, such as... Figure 7 As shown. Specifically, for S303, for the automatic layup type, the post-processing is path generation. First, the geodesic distance distribution of each zone is calculated, then the contour lines are extracted based on the geodesic distance distribution, and the contour lines are optimized for smoothness. The optimized contour lines are used as the automatic layup trajectory. For S304, for the manual layup type, the post-processing is surface flattening. The contour lines of each zone are flattened, and the flattened results are used for layout and cutting to obtain the cut pieces for manual layup. Thus, this embodiment realizes a complete process optimization flow from multi-directional ply matrix characterization, multi-objective optimization, surface adaptive zoning to layup post-processing. This flow can improve the feasibility of layup in complex curved surface areas while taking into account the consistency of interlayer angles between multi-directional plies, thereby improving the overall rationality of the composite material curved surface component layup process.

[0053] The parts not covered in this invention are the same as or can be implemented using existing technologies.

Claims

1. A method for overall optimization of the layup process of multi-directional ply composite curved surface components, characterized in that, The optimization of the layup process for complex curved surface composite material components includes the following steps: 1) Constructing a frame field composed of fiber reference directions on the target curved surface based on the fiber reference direction information given by the multi-directional layup design; 2) Optimizing the frame field according to the layup process objectives and constraints to obtain the optimized frame field; 3) Analyzing the layability of the target curved surface in combination with its geometric characteristics, forming several dividing lines on the curved surface based on the optimized frame field, and dividing the target curved surface into adaptive regions to obtain the curved surface partitioned layup process results, which are used to guide the subsequent automatic layup trajectory generation or manual layup layout design.

2. The method for overall optimization of the layup process of multi-directional ply composite curved surface components according to claim 1, characterized in that, In step 1, the fiber reference direction information given by the layup design includes a combination of 0°, 90° and ±45° reference directions, or a combination of any given angles; a frame field is constructed on the target surface according to the reference direction information, that is, any point on the surface contains all of the above-mentioned direction information.

3. The method for overall optimization of the layup process of multi-directional ply composite curved surface components according to claim 1, characterized in that, In step 2, the direction variables of each fiber reference direction in the target surface tangent plane in the standard frame field are used as optimization variables to establish a target optimization model containing constraint terms and target terms. The constraint terms are used to limit the acceptable range of deviations in layup quality and ply performance, including one or more of the following: direction deviation constraints, adjacent direction continuity constraints, surface layability constraints, and equipment layup capability constraints. The target terms are used to further optimize the layup effect and performance retention effect based on satisfying the constraint terms, including at least one or more of the following: minimizing direction deviation, optimizing direction field smoothness, and minimizing the risk of layup defects. By adjusting the weights or thresholds of each constraint term and target term, an optimized standard frame field that meets the layup process requirements is obtained.

4. The method for overall optimization of the layup process of multi-directional ply composite curved surface components according to claim 1, characterized in that, The adaptive region division in step 3 is to divide the target surface according to the optimized multi-directional layup direction results and the layup spreadability of the target surface, dividing the target surface into several sub-regions suitable for laying trajectory generation, surface flattening or piece generation respectively; the layability is used to characterize the risk of local stretching, compression, wrinkling or bridging defects in the target surface during the laying or flattening process.

Images