A method and system for optimizing the design of a carbon fiber fuselage structure layup

Abstract

The application discloses a kind of carbon fiber fuselage structure layer optimization design method and system, it is related to material structure intelligent design technical field, including, import the three-dimensional CAD model of fuselage structure, collect physical field input parameter, integration generates digital engineering package, based on digital engineering package, using thermal coupling solver to carry out temperature field calculation, obtain the temperature distribution of fuselage, the temperature distribution of fuselage is mapped to thermal stress field by thermoelastic constitutive equation, and with mechanical load is vector superposition, generate comprehensive stress nephogram, based on comprehensive stress nephogram, start asymmetric layer design in high stress area, and the mechanical balance of layer angle combination is checked by balance optimization algorithm, obtain layer optimization scheme.The application realizes the accurate evaluation of thermal deformation sensitivity by using thermal coupling solver and finite element perturbation analysis method, can effectively predict and control thermal deformation, so as to improve the dimensional stability and surface quality of final product.

Description

Technical Field

[0001] This invention relates to the field of intelligent material structure design technology, and in particular to a method and system for optimizing the layup design of carbon fiber fuselage structures. Background Technology

[0002] Carbon fiber composites, due to their excellent specific strength, specific stiffness, and fatigue resistance, have become one of the core materials in modern airframe structural design. In recent years, with the increasing demands for lightweight and high strength in the aerospace industry, carbon fiber layup optimization design methods have received widespread attention and research. Traditional layup design is usually based on empirical rules or simple finite element analysis, manually adjusting the layup angle and thickness to meet mechanical performance requirements.

[0003] With the development of computer-aided engineering technology, optimization algorithms have been gradually introduced into ply design. However, some shortcomings still exist. On the one hand, existing technologies also have limitations in controlling thermal deformation, failing to fully consider the influence of ply angle combinations on thermal deformation sensitivity, thus making it difficult to achieve high-precision deformation control. On the other hand, current ply optimization schemes ignore potential process defects that may occur during actual manufacturing, such as excessively large ply angle jumps or non-compliant curvature radii, which not only reduce the quality of the finished product but may also lead to a significant increase in manufacturing costs. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a carbon fiber fuselage structure layup optimization design method to solve the problems of inaccurate control of thermal deformation and process defects in the manufacturing process.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a carbon fiber fuselage structure layup optimization design method, which includes importing a three-dimensional CAD model of the fuselage structure, collecting physical field input parameters, and integrating them to generate a digital engineering package.

[0008] Based on the digital engineering package, a thermo-coupling solver is used to calculate the temperature field and obtain the temperature distribution of the fuselage. The temperature distribution of the fuselage is mapped to a thermal stress field through the thermoelastic constitutive equation and then vector-superimposed with the mechanical load to generate a comprehensive stress cloud map.

[0009] Based on the comprehensive stress cloud map, asymmetric ply design is initiated in the high stress area, and the mechanical balance of the ply angle combination is verified by the balance optimization algorithm to obtain the ply optimization scheme.

[0010] The layer optimization scheme is introduced into the thermal coupling solver, secondary temperature field calculation is performed, and the thermal deformation amount is detected, based on the thermal deformation amount, the weight coefficient of the layer angle combination is adjusted, and a secondary layer optimization scheme is output;

[0011] Based on the secondary layer optimization scheme, the layer feasibility is verified through the automatic fiber laying machine process database, and after the verification is completed, a manufacturing ready package is output using the angle gradual transition strategy;

[0012] The digital twin engine loads the manufacturing ready package, simulates the actual layering process, and outputs the final layering process parameter set.

[0013] As a preferred scheme of the carbon fiber fuselage structure layer optimization design method of the application, wherein: the physical field input parameters include mechanical load, temperature field parameter, material anisotropy parameter, fuselage vibration parameter and manufacturing process parameter.

[0014] As a preferred scheme of the carbon fiber fuselage structure layer optimization design method of the application, wherein: the comprehensive stress nephogram has the following specific steps,

[0015] Load the material anisotropy parameters in the thermal coupling solver, and obtain the temperature field distribution of the fuselage surface through the transient heat conduction algorithm;

[0016] Extract the thermal expansion tensor in the material anisotropy parameter, and map the temperature field distribution to the thermal stress field based on the thermal expansion tensor and the thermoelastic constitutive equation;

[0017] The thermal stress field and the mechanical load are superimposed by the weight tensor superposition algorithm to generate a three-dimensional comprehensive stress tensor field, and the maximum principal stress algorithm is used to obtain the comprehensive stress nephogram based on the three-dimensional comprehensive stress tensor field.

[0018] As a preferred scheme of the carbon fiber fuselage structure layer optimization design method of the application, wherein: the layer optimization scheme has the following specific steps,

[0019] Based on the comprehensive stress nephogram, the region coordinates of the high stress area are extracted by the DBSCAN clustering algorithm, and are classified into dominant areas of tension, compression and shear, and an asymmetric layering strategy is designed for each dominant area by the layer driving method;

[0020] After the design of the asymmetric layering of the high stress area is completed, the NSGA-II multi-objective balance optimization algorithm is used to dynamically adjust the layer angle and thickness, and the ABD matrix calculation is used to verify the mechanical balance, and finally the layer optimization scheme is output.

[0021] As a preferred scheme of the carbon fiber fuselage structure layer optimization design method of the application, wherein:

[0022] The specific steps of the ply secondary optimization scheme are as follows,

[0023] The ply optimization scheme is imported into the thermal coupling solver through the parameterized script interface, the secondary temperature field calculation is performed through the dynamic heat conduction algorithm, the updated three-dimensional temperature gradient field is generated, and the thermal deformation amount is detected based on the updated three-dimensional temperature gradient field through the displacement extreme value extraction algorithm;

[0024] Based on the thermal deformation amount, the partial derivative matrix of the ply angle is obtained through the finite element perturbation analysis method, and the deformation sensitivity gradient analysis is performed to generate the deformation sensitivity weight distribution map;

[0025] Based on the generated deformation sensitivity weight distribution map, the ply angle parameters are adjusted through the gradient descent adjustment algorithm, and the ply secondary optimization scheme is output through the multi-objective optimization framework.

[0026] As a preferred scheme of the carbon fiber fuselage structure ply optimization design method, the specific steps of the manufacturing ready package are as follows,

[0027] The ply secondary optimization scheme is executed by the automatic fiber placement machine process database through multi-rule joint verification:

[0028] The rule violation area is removed through the dynamic curvature radius algorithm, the adjacent ply angle jump gradient is detected in real time by using the sliding window method, and the process defect probability is predicted by calling the random forest classification model;

[0029] After completing the multi-rule joint verification, based on the angle gradual transition strategy, the NC code of the fiber placement machine is generated through the transition layer insertion algorithm and the B-spline curve path fitting method, and the manufacturing ready package is generated by inputting the digital twin engine.

[0030] As a preferred scheme of the carbon fiber fuselage structure ply optimization design method, the specific steps of the output final ply process parameter set are as follows,

[0031] The fiber placement process is laid through the kinematics modeling and real-time physical simulation engine in the digital twin engine, and the final ply process parameter set is output by iteratively adjusting the laying speed, heating temperature and pressure parameters through the multi-objective optimization algorithm.

[0032] Secondly, the present application provides a carbon fiber fuselage structure ply optimization design system, comprising a parameter integration module, a thermal stress coupling module, a primary optimization module, a secondary optimization module, a verification module, and a process simulation module,

[0033] The parameter integration module is used to import the three-dimensional CAD model of the fuselage structure, collect physical field input parameters, and integrate to generate a digital engineering package;

[0034] The thermal stress coupling module is used to calculate the temperature field based on the digital engineering package and the thermal coupling solver to obtain the temperature distribution of the fuselage. The temperature distribution of the fuselage is mapped to a thermal stress field through the thermoelastic constitutive equation and vector superimposed with the mechanical load to generate a comprehensive stress cloud map.

[0035] The initial optimization module is used to initiate asymmetric ply design in high-stress areas based on the comprehensive stress cloud map, and to verify the mechanical balance of the ply angle combination through the balance optimization algorithm to obtain the ply optimization scheme.

[0036] The secondary optimization module is used to import the ply optimization scheme into the thermo-mechanical coupling solver, perform secondary temperature field calculations, detect thermal deformation, adjust the weight coefficients of the ply angle combination based on the thermal deformation, and output the ply secondary optimization scheme.

[0037] The verification module is used to verify the feasibility of the layup based on the secondary optimization scheme of the layup and through the automatic fiber placement machine process database. After the verification is completed, a manufacturing ready package is output using an angle gradual transition strategy.

[0038] The process simulation module is used to load the manufacturing-ready package into the digital twin engine, simulate the actual layup process, and output the final layup process parameter set.

[0039] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the carbon fiber fuselage structure layup optimization design method as described in the first aspect of the present invention.

[0040] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the carbon fiber fuselage structure layup optimization design method as described in the first aspect of the present invention.

[0041] The beneficial effects of this invention are as follows: Firstly, by utilizing a thermo-coupling solver and finite element perturbation analysis, accurate assessment of thermal deformation sensitivity is achieved. Secondly, by dynamically adjusting the weighting coefficients of the layup angle combination and employing a gradient descent adjustment algorithm, thermal deformation can be effectively predicted and controlled, thereby improving the dimensional stability and surface quality of the final product. Thirdly, addressing the issue of process defects during manufacturing, this invention combines an automatic fiber placement machine process database with a multi-rule joint verification mechanism to ensure the manufacturing feasibility of the layup scheme. Simultaneously, it uses a random forest classification model to predict the probability of potential process defects and optimizes the layup design based on a gradient angle transition strategy, significantly reducing the risk of process defects during manufacturing and improving finished product quality and production efficiency. Attached Figure Description

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0043] Fig. 1 Flow chart of carbon fiber fuselage structure layer optimization design method.

[0044] Fig. 2 Schematic diagram of carbon fiber fuselage structure layer optimization design system.

[0045] Fig. 3 Flow chart of obtaining process of comprehensive stress diagram.

[0046] Fig. 4 Flow chart of output process of final layer process parameter set. DETAILED DESCRIPTION

[0047] In order to make the above objectives, features and advantages of the present application more apparent and comprehensible, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0048] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can be practiced in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the scope of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0049] Secondly, the "one embodiment" or "embodiment" referred to herein can include specific features, structures or characteristics contained in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0050] Reference Figs. 1-4 For one embodiment of the present application, the embodiment provides a carbon fiber fuselage structure layer optimization design method, comprising the following steps:

[0051] S1, import the three-dimensional CAD model of the fuselage structure, collect the physical field input parameters, and integrate to generate a digital engineering package.

[0052] Specifically, the following steps are included,

[0053] The CATIA software of computer aided design is used to convert the design drawing of the fuselage structure into a digital three-dimensional CAD model to ensure that all details are accurately captured and expressed. Next, in order to fully understand the response of the fuselage in the actual working environment, a series of physical field input parameters need to be collected, including mechanical load, temperature field parameter, material anisotropy parameter, fuselage vibration and manufacturing process parameter;

[0054] The data collection process relies on a variety of existing technical means. In specific operation, mechanical testing equipment such as strain gauges and force sensors are installed at the stressed parts of the fuselage. By measuring the deformation and stress response of the structure under external load, the size and distribution of mechanical load are recorded to collect mechanical load data.

[0055] Temperature sensors are arranged in the area of the fuselage structure, and at the same time, thermal imaging technology is used to scan the surface of the fuselage by infrared thermal imager to obtain the temperature distribution under different working conditions to collect temperature field parameters.

[0056] Based on laboratory testing means, by applying different mechanical loads to the material sample and recording its response, combined with finite element analysis technology, the anisotropy data of the material inside is derived to collect material anisotropy parameters.

[0057] Acceleration sensors and spectrum analyzers are arranged at different positions of the fuselage to monitor vibration signals in real time, and signal processing technology is used to extract key frequency components and their corresponding amplitude information to collect fuselage vibration parameters. Finally, through the historical process database and on-site monitoring equipment, the process parameters of the automatic fiber placement machine in the actual manufacturing process are collected and recorded, such as laying speed, heating temperature and pressure, etc. to form a complete set of manufacturing process parameters. The collected physical field input parameters provide a solid data foundation for the subsequent generation of digital engineering package.

[0058] The collected physical field input parameters are accurately associated with the three-dimensional CAD model geometry coordinates through data mapping technology: the mechanical load data, temperature field parameters, material anisotropy parameters, fuselage vibration parameters, environmental humidity field parameters and manufacturing process parameters are converted into node force distribution through a finite element pre-processor, a continuous temperature gradient field is generated by fusing discrete temperature field parameters through an interpolation algorithm, material anisotropy parameters are bound layer by layer through a composite material attribute distribution tool (such as ANSYS Composite PrepPos), fuselage vibration parameters are calibrated through finite element analysis, manufacturing constraint conditions of manufacturing process parameters are extracted through a Java process rule analysis tool, and mechanical load, temperature field parameters, material anisotropy parameters, fuselage vibration parameters, environmental humidity field parameters and manufacturing process parameters are structured and packaged in layers through HDF5 format, the geometric matching degree is verified through curvature continuity detection algorithm, and the missing data is completed through Kriging interpolation algorithm, then various data files are integrated through standardized compression format, and finally efficient data calling is realized through layered index interface, and a digital engineering package is generated.

[0059] S2, based on the digital engineering package, using a thermal coupling solver to calculate the temperature field, obtaining the temperature distribution of the fuselage, mapping the temperature distribution of the fuselage into the thermal stress field through the thermoelastic constitutive equation, and superimposing the mechanical load to generate a comprehensive stress cloud chart.

[0060] Specifically includes the following steps,

[0061] After the thermal coupling solver loads the digital engineering package, the material anisotropy parameters need to be analyzed first, and the thermal conductivity tensor is read through the attribute extraction interface to calculate the temperature field distribution. The thermal conductivity tensor reflects the difference in thermal conductivity of carbon fiber material in different directions. According to the layer angle and the direction of the material principal axis, the thermal conductivity tensor is accurately assigned. After the thermal conductivity tensor is assigned, the transient heat conduction algorithm is used to calculate the temperature field. In the specific operation, the transient heat conduction calculation needs to consider the temperature change in the time dimension. The finite volume method is usually used to discretize the spatial domain of the thermal conductivity tensor to better handle complex geometric shapes and material discontinuity problems, and the implicit Euler format algorithm is used to time advance the thermal conductivity tensor. The unconditional stability of the implicit Euler format algorithm can avoid the strict restriction of time step in the explicit method.

[0062] After the spatial domain discretization and time marching are completed, transient temperature field data containing temperature gradient distribution of the fuselage surface and internal structure are generated, and the pressure distribution and boundary layer characteristics of the aircraft surface are calculated by aerodynamic thermodynamic simulation software to obtain aerodynamic heat flow calculation results. The transient temperature field data and the aerodynamic heat flow calculation results are integrated by a boundary condition fusion processing method, and the high temperature area is refined by using a local grid refinement technique. Combined with real-time updating and iteration of the anisotropic parameters of the material, the temperature field distribution of the fuselage surface is finally obtained.

[0063] After obtaining the temperature field distribution, the temperature field distribution result needs to be mapped to a thermal stress field. In the specific operation, first, the thermal expansion tensor is extracted from the anisotropic parameters of the material by a material attribute parser. The thermal expansion tensor is a key parameter for calculating thermal strain. The thermal expansion tensor has directionality and needs to be assigned according to the material principal axis direction of each ply. Based on the assigned thermal expansion tensor, the thermal elastic constitutive equation is used, and the incremental method is used to gradually apply temperature load to the carbon fiber material to obtain the temperature change. The specific mathematical formula is as follows:

[0064]

[0065] Where ΔT represents the temperature change, T represents the temperature, T1 represents the sum of the principal diagonal element values of the tensor, ε represents the thermal strain tensor, k represents the correction factor, α represents the thermal expansion tensor, T1(ε) represents the sum of the principal diagonal element values of the thermal strain tensor, k(T) represents the correction factor of the temperature, and T1(α) represents the sum of the principal diagonal element values of the thermal expansion tensor.

[0066] It should be noted that the specific value of k(T) depends on the properties of the material, the working temperature range and the experimental analysis results, and is generally between 0 and 2.

[0067] The thermal strain field is calculated in each temperature increment step based on the obtained temperature change. In the calculation process, the thermal elastic balance equation is established by finite element discretization based on the virtual work principle, and the high-efficiency conjugate gradient method is used to map the temperature field distribution to the thermal strain field. Special attention needs to be paid to the strain compatibility conditions at the ply interface to ensure the deformation continuity and physical reasonableness between different angle plies.

[0068] After obtaining the thermal stress field, the grid matching algorithm is used to check the grid consistency of the thermal stress field and the mechanical load, to ensure that the stress fields of the thermal stress field and the mechanical load have the same cell topological structure. For the case of inconsistent grid, the stress field interpolation technology based on shape function is used to realize the grid consistency. After ensuring the grid consistency, the weight tensor superposition algorithm is used for vector superposition of the thermal stress field and the mechanical load. In the specific operation, the material stiffness tensor is extracted from the material anisotropy parameters by the material property parser, the thermal stress component is obtained by the thermoelasticity solver, and the mechanical stress component is obtained by the statics finite element analysis. The material stiffness tensor is normalized to establish a weight coefficient matrix reflecting the anisotropy characteristics of the material.

[0069] Based on the weight coefficient matrix, the tensor eigenvalue decomposition algorithm is used to decompose the thermal stress component and the mechanical stress component into tensor, and the normal stress and shear stress components are extracted respectively. The normal stress component and the shear stress component are weighted and synthesized according to the weight, wherein the shear stress weight of the high temperature area is reduced (considering the material softening effect), and the normal stress component of the high stiffness direction obtains higher weight. The continuity of the stress component is ensured by using the Gauss integral point algorithm in the process of weighted synthesis. After the weighted synthesis is completed, a three-dimensional comprehensive stress tensor field containing six independent components is generated.

[0070] Based on the three-dimensional comprehensive stress tensor field, a comprehensive stress contour map is generated. In the specific operation, based on the three-dimensional comprehensive stress tensor field, the finite element unit is generated by the finite element preprocessor, the three-dimensional stress tensor is generated by the thermal force coupling solver, and then the maximum principal stress algorithm is used to traverse all Gauss integral points of the finite element unit. The eigenvalue of the three-dimensional stress tensor at each Gauss integral point is solved by the Jacobi iteration method to extract the maximum value in the three-dimensional stress tensor, and the maximum value is taken as the characteristic stress value of the Gauss integral point.

[0071] After completing the eigenvalue solution of all Gauss integral points, each characteristic stress value is mapped to the unit node by the bilinear interpolation technology to generate a node stress field, and the node stress field is smoothly transitioned by using the Lagrange shape function. The node stress field is rendered by using the adaptive color scale generation technology, and the color distribution range is automatically adjusted according to the stress gradient. The local color scale encryption processing is used for the high stress concentration area (such as the opening edge, the connection joint, etc.), and the stress mutation boundary is identified by the edge detection algorithm. Combined with the contour surface extraction technology, clear stress contour lines are generated.

[0072] Based on the stress contour line, a parallel computing accelerated graphics rendering pipeline is generated by the CUDA parallel computing framework. According to the comprehensive stress contour map finally generated by the parallel computing accelerated graphics rendering pipeline, the comprehensive stress contour map contains a dynamic color scale ruler, a maximum stress value label and a key area zoom view, supports the display of any cross section of the three-dimensional model and the playback of stress distribution animation.

[0073] S3, based on the comprehensive stress nephogram, starting the asymmetric layer design in the high stress area, and checking the mechanical balance of the layer angle combination through the balance optimization algorithm, obtaining the layer optimization scheme.

[0074] Specifically includes the following steps,

[0075] When carrying out asymmetric layer design based on the comprehensive stress nephogram, first, the high stress area is accurately identified and reasonably partitioned through the DBSCAN clustering algorithm, and the regional coordinate set of the high stress area is obtained. In the specific operation, the DBSCAN clustering algorithm sets appropriate neighborhood search radius and minimum sample number threshold through the parameter optimization tool (such as the multi-objective optimization component in ANSYS Workbench), and based on the set neighborhood search radius, the density accessibility is analyzed to automatically scan the entire stress nephogram, and the area with stress value exceeding the minimum sample number threshold is identified as the high stress area. The regional coordinate set of the high stress area is obtained through the space index technology, and the regional coordinate set of the high stress area is divided into several independent clusters based on the Euclidean distance measurement method;

[0076] In the clustering process, the principal stress direction in each independent cluster is analyzed by the principal stress direction algorithm, the first principal stress direction is obtained based on the principal stress direction through the eigenvalue decomposition algorithm, the angle between the first principal stress direction and the material axial direction is calculated, and the independent clusters are further subdivided into tensile dominant area, compression dominant area and shear dominant area by using K-means clustering technology. The judgment standard of tensile dominant area is that the first principal stress direction is consistent with the material axial direction, the compression dominant area is that the first principal stress direction is perpendicular to the material axial direction, and the shear dominant area is that the maximum shear stress is significantly higher than the normal stress;

[0077] After the division of each dominant area, the asymmetric layer strategy is adopted for different types of dominant areas by using layer driving technology through rule engine: zero-degree layer is mainly arranged in tensile dominant area to fully utilize the high strength characteristics of fiber direction, forty-five-degree layer proportion is increased in compression dominant area to improve the in-plane shear stiffness and buckling resistance, and positive and negative forty-five-degree layers are alternately arranged in shear dominant area to optimize the shear performance. The initial layer scheme is automatically generated by the parameterized modeling tool (such as the layer parameterization tool in CATIA Composites) according to the layer strategy through the layer driving technology, and reasonable layer transition zone is set between adjacent high stress areas to ensure the smooth transition of mechanical properties through the gradual change of angle and thickness adjustment, avoiding the occurrence of stiffness mutation;

[0078] After the design of the asymmetric layup in the high stress area is completed, the NSGA-II multi-objective balance optimization algorithm is used to iteratively improve the initial layup scheme. In the specific operation, in the initialization process of the NSGA-II multi-objective balance optimization algorithm, a plurality of layup schemes are randomly generated by a Latin hypercube sampling method to form an initial population, wherein each layup scheme in the initial population includes a complete sequence of layup angles and a thickness distribution;

[0079] Each layup scheme is evaluated based on the initial population through a multi-objective evaluation technique. First, the ABD matrix is calculated through the classical laminate theory, and the balance degree of tensile stiffness and compressive stiffness is obtained through stiffness matrix decomposition based on the ABD matrix. At the same time, the proportional relationship between in-plane stiffness and bending stiffness is obtained through the ratio of stiffness components, and the numerical value of each coupling effect is obtained through the coupling stiffness term analysis. At the same time, the weight index of each layup scheme is calculated through the layup parameter integral, and the process feasibility is evaluated through the process constraint checking technique. The process feasibility includes whether the layup angle change rate exceeds the maximum allowable value of the automatic laying machine, whether the minimum curvature radius meets the material laying requirements, etc.

[0080] The multi-objective fitness function is constructed by a weighted linear combination method based on the balance degree of tensile stiffness and compressive stiffness, the proportional relationship between in-plane stiffness and bending stiffness, the numerical value of each coupling effect, the weight index of the layup scheme, and the process feasibility. Based on the multi-objective fitness function, the non-dominated sorting mechanism of the NSGA-II algorithm verifies the mechanical balance.

[0081] At the same time, in the evolution process of each generation of the NSGA-II multi-objective balance optimization, the NSGA-II multi-objective balance optimization algorithm selects parent individuals through a tournament selection mechanism, and generates a child population through a simulated binary crossover and a polynomial mutation operation. The binary crossover intelligently recombines the layup angle sequence of the parent individual, and the polynomial mutation applies adaptive disturbance to the layup thickness parameter to dynamically adjust the layup angle and thickness.

[0082] Next, all parent individuals, child populations, and initial populations are divided into multiple frontiers according to the superiority and inferiority degree through the Pareto dominance relationship, and an improved crowding degree formula is defined based on the layup angle continuity constraint and the thickness gradient limit. The improved crowding degree formula ensures that the parent individuals, child populations, and initial populations on each frontier can be uniformly distributed under the premise of meeting the layup process requirements through adaptive neighborhood search technology.

[0083] Based on the multiple frontiers, the NSGA-II multi-objective balance optimization algorithm performs multi-generation evolution through an elite reservation strategy, dynamically adjusts the angle and thickness combination in each generation (such as increasing the 0° layup proportion for schemes with insufficient stiffness, and optimizing the ±45° layup thickness distribution for overweight schemes), and outputs a set of Pareto optimal solutions.

[0084] The Pareto optimal solution achieves the best balance between weight, stiffness and process feasibility, and the final decision is based on the optimal solution to generate a layer optimization scheme through a multi-criteria decision method. The layer optimization scheme not only meets the mechanical performance requirements, but also considers the manufacturing process limitations, thereby providing a reliable basis for the detailed design of the carbon fiber fuselage structure.

[0085] S4, importing the layer optimization scheme into the thermal coupling solver, performing secondary temperature field calculation, and detecting the thermal deformation amount, based on the thermal deformation amount, adjusting the weight coefficient of the layer angle combination, and outputting the secondary layer optimization scheme.

[0086] Specifically includes the following steps,

[0087] In the process of importing the layer optimization scheme into the thermal coupling solver, the parameterized script interface first obtains the layer angle sequence and thickness distribution parameters through the CATIA CAA secondary development interface, and performs format conversion through the ANSYS ACT custom converter. The format conversion process needs to accurately match the topological relationship between the material axial direction and the grid, so as to ensure that the angle sequence and thickness distribution parameters of each layer unit can be accurately mapped to the corresponding grid area;

[0088] After the format conversion is completed, the thermal coupling solver starts the dynamic heat conduction calculation. In the specific operation, the nonlinear transient thermal analysis algorithm adopts a time step control strategy, which automatically increases the time step in the area where the temperature changes sharply to improve the calculation accuracy of the temperature field evolution. At the same time, in the layer transition area, the anisotropic heat conduction algorithm automatically adjusts the heat flow calculation direction to reflect the change of the heat conduction performance caused by the gradual change of the layer angle. During the temperature field evolution calculation, the thermal coupling solver monitors the temperature field convergence in real time. When it is detected that the local temperature field convergence gradient exceeds the convergence threshold, the grid refinement program is automatically triggered, and new calculation nodes are inserted in the area where the temperature changes sharply;

[0089] Based on the new calculation nodes, the updated three-dimensional temperature gradient field is generated through the cubic spline interpolation technology. The convergence threshold is defined based on the material thermal expansion coefficient and the layer angle distribution characteristics through the thermal-mechanical coupling sensitivity analysis;

[0090] Based on the updated three-dimensional temperature gradient field, the node displacement and temperature change are obtained through the thermal coupling solver. The displacement extreme value extraction algorithm establishes the mapping relationship between the node displacement and the temperature change through the least square regression analysis. Based on the mapping relationship between the node displacement and the temperature change, the displacement extreme value points are identified through the Delaunay triangulation area scanning technology, and the complete history of the displacement of the displacement extreme value points with time is recorded. The thermal deformation amount is obtained through the numerical differentiation algorithm.

[0091] Based on the thermal deformation, the finite element perturbation analysis method obtains the partial derivative matrix of the layer angle, and in the specific operation, the central difference method is used to perturb each layer angle parameter slightly, and after each perturbation, the thermal force coupling solver is used to calculate the thermal force coupling to obtain new displacement field data. By comparing the displacement field data before and after the perturbation, the complete layer angle partial derivative matrix is constructed by the matrix assembly algorithm. Each element in the layer angle partial derivative matrix quantifies the control effect of the specific layer angle adjustment on the local thermal deformation.

[0092] After generating the layer angle partial derivative matrix, the deformation sensitivity gradient analysis is carried out by the gradient projection method, and the deformation sensitivity distribution data is generated. Based on the generated deformation sensitivity distribution data, the layer angle partial derivative matrix is converted into a visual deformation sensitivity weight distribution graph by the Matplotlib visualization library. The deformation sensitivity weight distribution graph is displayed in the form of a heat map, and the gradient color scale from blue to red is used to represent the sensitivity of different regions. The red area represents the area with high sensitivity in the layer angle adjustment process, and the blue area represents the area with low sensitivity in the layer angle adjustment process. At the same time, the visualization algorithm will automatically mark the position of the sensitivity mutation. The position of the sensitivity mutation usually corresponds to the layer transition zone or the area with discontinuous material properties.

[0093] Based on the deformation sensitivity weight distribution graph, the layer parameter optimization is carried out by the gradient descent adjustment algorithm. In the specific operation, when the gradient descent adjustment algorithm is initialized, the high sensitivity areas in the deformation sensitivity weight distribution graph are first identified by the region clustering analysis algorithm, and high adjustment weights are assigned to these high sensitivity areas. At the same time, in the iteration process of the gradient descent adjustment algorithm, the sensitivity gradient direction in the high sensitivity area is obtained by the finite difference analysis method, and the gradient descent adjustment algorithm adjusts the layer angle parameters along the sensitivity gradient direction through the step control strategy. At the same time, the displacement change of the key nodes (such as the displacement extreme point, the temperature gradient mutation point and the layer transition zone) is monitored. When it is detected that the adjustment of a certain layer angle leads to displacement improvement, the gradient descent adjustment algorithm will increase the adjustment weight of the layer angle, and vice versa.

[0094] During the adjustment of the layer angle parameters, the gradient descent adjustment algorithm enforces the layer process constraints through constraint processing technology to ensure that the layer angle change rate and the thickness gradient are always within the manufacturable range. At the same time, a multi-objective optimization framework is constructed based on the Pareto frontier theory. The multi-objective optimization framework plays a coordinating role in the entire layer angle parameter adjustment process. Through the dynamic priority algorithm, three independent optimization target queues are maintained: the stiffness performance queue, the thermal deformation control queue and the process feasibility queue.

[0095] After each adjustment of the layer parameters, the multi-objective optimization framework evaluates the impact of the layer parameter adjustment on the three optimization target queues through the multi-criteria decision analysis method. Only when the layer parameter adjustment is beneficial to at least two optimization target queues and does not harm the third optimization target queue, the layer parameter adjustment is retained. The multi-objective optimization framework also sets a target balance monitoring mechanism through the performance deviation monitoring algorithm. When the optimization of a certain target queue is obviously lagging behind, the optimization priority of the target queue is automatically increased. Based on the retained layer parameters, the parameterized modeling engine outputs a secondary optimization scheme of the layer.

[0096] S5, based on the secondary optimization scheme of the layer, the feasibility of the layer is verified through the automatic fiber placement machine process database. After verification, the manufacturing ready package is output using the angle gradual transition strategy.

[0097] Specifically includes the following steps,

[0098] In the process of realizing the manufacturing ready package, the secondary optimization scheme of the layer is first imported into the automatic fiber placement machine process database through the parameterized data interface for multi-rule joint verification. In the specific operation, the entire secondary optimization scheme of the layer is scanned through the dynamic curvature radius algorithm, and the area where the curvature exceeds the minimum allowable value of the automatic fiber placement machine is identified through the rate extreme detection algorithm, and is marked as a violation area through the hash marking algorithm. The minimum allowable value is defined by the minimum allowable curvature radius parameter in the technical specification of the automatic fiber placement machine. At the same time, the sliding window method moves along the layer path with a fixed width, the adjacent layer angle change rate is calculated in real time through the difference angle calculator, and the adjacent layer angle jump gradient is obtained through the gradient extraction algorithm;

[0099] When the adjacent layer angle jump gradient exceeds the gradient threshold value, a warning is immediately issued. The gradient threshold value is defined based on the maximum allowable angle change rate of adjacent layers specified in the composite material process specification. At the same time, by using the historical physical field input parameters, a random forest classification model is built through the Scikit-learn machine learning framework, and is trained through the feature weighting training method. The trained random forest classification model comprehensively evaluates multiple factors such as fiber orientation, layer sequence and thickness change through a multi-dimensional feature fusion evaluator, and outputs the process defect probability value of each area;

[0100] After the multi-rule joint verification is completed, the angle gradual transition strategy starts to be implemented. In the specific operation, the transition layer insertion algorithm first analyzes the area to be optimized to determine the insertion position and quantity of the transition layer. The transition layer insertion algorithm prioritizes the high-risk areas with higher process defect probability values and increases the transition layer in the high-risk areas through the adaptive layer number allocation algorithm. The increased transition layer smoothly changes the angle through linear interpolation, and the B-spline curve path fitting technology is used to convert the discrete layer path points into a continuous and smooth curve through the node parameterization method, ensuring that the automatic fiber placement head can smoothly perform the placement action.

[0101] In the path fitting process of the B-spline curve path fitting technology, material tension compensation and laying speed optimization are considered to avoid fiber wrinkles or stretching at turns. Through the processing of the transition layer insertion algorithm and the B-spline curve path fitting technology, the optimized layer path data is generated. Based on the optimized layer path data, the NC code generator is used to convert it into fiber placement NC code. Multiple verifications are implemented during the NC code generation process, including instruction syntax checking, motion range verification, and process parameter compliance confirmation.

[0102] After receiving the generated fiber placement NC code, the digital twin engine constructs a virtual manufacturing environment through three-dimensional scene reconstruction technology, completely reproducing the actual production workshop equipment configuration and process conditions. The digital twin engine performs full-process simulation to simulate the actual motion trajectory and laying process of the automatic fiber placement machine, and detects possible interference or collision through the collision detection algorithm to generate a simulation verification report. Finally, the optimized layer path data, NC code file, and simulation verification report are integrated through manufacturing data integration and packaging technology to generate a manufacturing-ready package. The entire manufacturing-ready package generation process is fully automated, from the input of the layer optimization scheme to the output of the ready package without human intervention, ensuring the accuracy and efficiency of process conversion.

[0103] S6, load the manufacturing-ready package in the digital twin engine, simulate the actual layering process, and output the final layering process parameter set to optimize the layer design.

[0104] Specifically includes the following steps,

[0105] In the digital twin engine, the layering process is optimized. In the specific operation, the dedicated data parser is used to start loading and analyzing the manufacturing-ready package, and the structured data conversion algorithm is used to convert it into an internal format that can be processed by the digital twin engine.

[0106] After the format conversion is completed, a composite material constitutive model is built through a multi-body dynamics simulation framework using kinematics modeling technology. In the process of building the composite material constitutive model, all degrees of freedom of the motion of the mechanical structure of the fiber placement machine are completely reproduced through a parameterized joint modeling method, including the six-axis motion capability of the fiber placement head and the dynamic characteristics of the fiber feeding mechanism. After the composite material constitutive model is built, the real-time physics simulation engine is initialized through a physics engine protocol, and the composite material constitutive model and the temperature field data are loaded through a distributed data loader, thereby providing a physical basis for subsequent simulation;

[0107] In the simulation stage of the laying process, the NC code is converted into action instructions through a G code interpreter, and the material constitutive model is driven to perform the laying action according to the action instructions through kinematics modeling technology. At the same time of performing the laying action, the real-time physics simulation engine is applied to perform calculation at a fixed time simulation step: the stress and strain state vector of the carbon fiber in the laying process is calculated through the finite element method, the heating temperature field distribution value is calculated through the heat conduction equation, the interaction force value between the fiber placement head and the mold surface is calculated based on the contact mechanics algorithm, and the laying process state report is generated through multi-physical field coupling analysis based on the stress and strain state vector, the temperature field distribution size and the interaction force value between the fiber placement head and the mold surface.

[0108] Based on the generated laying process state report, the multi-objective optimization algorithm dynamically adjusts the process parameters. In specific operation, the multi-objective optimization algorithm constructs a weighted objective function through a target weight distribution algorithm, and based on the weighted objective function, three parallel optimization targets are maintained through a dynamic priority algorithm: laying quality index, production efficiency and energy consumption control, and the laying parameters are adjusted through the gradient descent method. Under the premise of ensuring the fiber wetting quality, the motion efficiency is maximized, and after each laying parameter adjustment, the digital twin engine quickly performs local re-simulation, and evaluates the laying parameter adjustment effect based on the three parallel optimization targets through multi-criteria decision analysis.

[0109] Based on the laying parameter adjustment effect, the heating temperature control is realized through the PID algorithm, and the simulation verification is completed through the physical field convergence verification method. The execution result of the simulation verification is integrated through the parameter normalization technology, and the final laying process parameter set is output.

[0110] The embodiment also provides a carbon fiber fuselage structure laying optimization design system, which comprises a parameter integration module, a thermal stress coupling module, a primary optimization module, a secondary optimization module, a verification module and a process simulation module.

[0111] The parameter integration module is used to import a three-dimensional CAD model of the fuselage structure, collect physical field input parameters, and integrate to generate a digital engineering package.

[0112] a thermal stress coupling module, configured to perform temperature field calculation based on a digital engineering package using a thermal-mechanical coupling solver, to obtain a temperature distribution of the fuselage, to map the temperature distribution of the fuselage into a thermal stress field through a thermo-elastic constitutive equation, and to perform vector superposition with mechanical loads to generate a comprehensive stress cloud map;

[0113] a primary optimization module, configured to start asymmetric layering design in a high stress area based on the comprehensive stress cloud map, and to check mechanical balance of a layering angle combination through a balance optimization algorithm to obtain a layering optimization scheme;

[0114] a secondary optimization module, configured to import the layering optimization scheme into the thermal-mechanical coupling solver, to perform secondary temperature field calculation, and to detect a thermal deformation amount, to adjust a weight coefficient of the layering angle combination based on the thermal deformation amount, and to output a secondary layering optimization scheme;

[0115] a verification module, configured to check layering feasibility through an automatic fiber placement machine process database based on the secondary layering optimization scheme, and to output a manufacturing-ready package using an angle gradual transition strategy after completion of the checking;

[0116] a process simulation module, configured to load the manufacturing-ready package in a digital twin engine, to simulate an actual layering process, and to output a final layering process parameter set.

[0117] The embodiment also provides a computer device suitable for the carbon fiber fuselage structure layering optimization design method, which comprises a memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions to implement the carbon fiber fuselage structure layering optimization design method proposed in the above embodiment.

[0118] The computer device can be a terminal, which comprises a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with external terminals. The wireless communication can be achieved through WIFI, operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0119] The embodiment also provides a storage medium on which a computer program is stored, the program being executed by a processor to implement the method for optimizing the structure of a carbon fiber fuselage, as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage device, a flash memory, a magnetic disk, or an optical disk.

[0120] To sum up, the present application realizes accurate evaluation of thermal deformation sensitivity through a thermal coupling solver and a finite element perturbation analysis method, effectively predicts and controls thermal deformation by dynamically adjusting the weight coefficients of the layer angle combination and using a gradient descent adjustment algorithm, thereby improving the dimensional stability and surface quality of the final product. Secondly, in view of the process defect problem in the manufacturing process, the present application combines an automatic fiber placement machine process database and a multi-rule joint verification mechanism to ensure the manufacturing feasibility of the layer scheme, simultaneously calls a random forest classification model to predict the probability of potential process defects, and optimizes the layer design based on an angle gradual transition strategy, thereby greatly reducing the risk of process defects in the manufacturing process and improving the product quality and production efficiency.

[0121] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application, although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A method of carbon fiber fuselage structure layup optimization design, characterized in that: Comprising, Import the three-dimensional CAD model of the fuselage structure, collect the physical field input parameters, and integrate to generate a digital engineering package; Based on the digital engineering package, use the thermal coupling solver to calculate the temperature field, obtain the temperature distribution of the fuselage, map the temperature distribution of the fuselage to the thermal stress field through the thermoelastic constitutive equation, and perform vector superposition with the mechanical load to generate a comprehensive stress cloud map; Based on the comprehensive stress cloud map, start the asymmetric layup design in the high stress area, and check the mechanical balance of the layup angle combination through the balance optimization algorithm to obtain the layup optimization scheme, the specific steps are as follows, Based on the comprehensive stress cloud map, the region coordinates of the high stress area are extracted by the DBSCAN clustering algorithm, and are classified into dominant areas of tension, compression and shear. The asymmetric layup strategy is designed for each dominant area through the layup driving method; Use the NSGA-II multi-objective balance optimization algorithm to dynamically adjust the layup angle and thickness, and verify the mechanical balance through the ABD matrix calculation, and finally output the layup optimization scheme; Import the layup optimization scheme into the thermal coupling solver, perform secondary temperature field calculation, and detect the thermal deformation, adjust the weight coefficient of the layup angle combination based on the thermal deformation, and output the secondary layup optimization scheme; Based on the secondary layup optimization scheme, check the layup feasibility through the automatic fiber placement machine process database, and output the manufacturing ready package after the checking is completed using the angle gradual transition strategy; Load the manufacturing ready package into the digital twin engine, simulate the actual layup process, and output the final layup process parameter set.

2. The method of claim 1, wherein: The physical field input parameters include mechanical load, temperature field parameters, material anisotropy parameters, fuselage vibration parameters, and manufacturing process parameters.

3. The method of claim 1, wherein: The comprehensive stress cloud map has the following specific steps, Load the material anisotropy parameters in the thermal coupling solver, and obtain the temperature field distribution on the surface of the fuselage through the transient heat conduction algorithm; Extract the thermal expansion tensor in the material anisotropy parameters, and map the temperature field distribution to the thermal stress field based on the thermal expansion tensor using the thermoelastic constitutive equation; Perform vector superposition on the thermal stress field and the mechanical load through the weight tensor superposition algorithm to generate a three-dimensional comprehensive stress tensor field, and obtain the comprehensive stress cloud map based on the three-dimensional comprehensive stress tensor field through the maximum principal stress algorithm.

4. The method of claim 1, wherein: The secondary layup optimization scheme has the following specific steps, Import the layup optimization scheme into the thermal coupling solver through the parameterized script interface, perform secondary temperature field calculation through the dynamic heat conduction algorithm, generate an updated three-dimensional temperature gradient field, and detect the thermal deformation based on the updated three-dimensional temperature gradient field through the displacement extreme extraction algorithm; Based on the thermal deformation, obtain the partial derivative matrix of the layup angle through the finite element perturbation analysis method, and perform deformation sensitivity gradient analysis to generate a deformation sensitivity weight distribution map; Based on the generated deformation sensitivity weight distribution map, adjust the layup angle parameters through the gradient descent adjustment algorithm, and output the secondary layup optimization scheme through the multi-objective optimization framework.

5. The method of claim 4, wherein: The manufacturing ready package has the following specific steps, Use the automatic fiber placement machine process database to perform multi-rule joint verification on the secondary layup optimization scheme: The dynamic curvature radius algorithm is used to remove the illegal area, the sliding window method is used to detect the gradient of the angle jump of adjacent layers in real time, and the random forest classification model is called to predict the process defect probability; After completing the multi-rule joint verification, based on the angle gradual transition strategy, the transition layer insertion algorithm and the B-spline curve path fitting method are used to generate the NC code of the fiber placement machine, which is input into the digital twin engine to generate the manufacturing ready package.

6. The method of claim 5, wherein: The specific steps of outputting the final layer process parameter set are as follows, In the digital twin engine, the fiber placement process is modeled and simulated in real time through the kinematics modeling and real-time physics simulation engine, and the final layer process parameter set is output by iteratively adjusting the placement speed, heating temperature and pressure parameters through the multi-objective optimization algorithm.

7. A carbon fiber fuselage structure layup optimization design system based on the carbon fiber fuselage structure layup optimization design method of any one of claims 1-6, characterized in that: It includes a parameter integration module, a thermal stress coupling module, a primary optimization module, a secondary optimization module, a verification module, and a process simulation module, The parameter integration module is used to import the three-dimensional CAD model of the fuselage structure, collect physical field input parameters, and integrate to generate a digital engineering package; The thermal stress coupling module is used to calculate the temperature field based on the digital engineering package using a thermal-mechanical coupling solver, obtain the temperature distribution of the fuselage, map the temperature distribution of the fuselage to a thermal stress field through a thermoelastic constitutive equation, and perform vector superposition with mechanical loads to generate a comprehensive stress cloud map; The primary optimization module is used to start asymmetric layer design in high stress areas based on the comprehensive stress cloud map, and check the mechanical balance of the layer angle combination through a balance optimization algorithm to obtain a layer optimization scheme, the specific steps are as follows, Based on the comprehensive stress cloud map, the region coordinates of the high stress area are extracted through the DBSCAN clustering algorithm, and are classified into dominant areas of tension, compression and shear. An asymmetric layer strategy is designed for each dominant area through a layer driving method; NSGA-II multi-objective balance optimization algorithm is used to dynamically adjust the layer angle and thickness, and the mechanical balance is verified through ABD matrix calculation, and finally the layer optimization scheme is output; The secondary optimization module is used to import the layer optimization scheme into the thermal-mechanical coupling solver, perform secondary temperature field calculation, and detect thermal deformation, adjust the weight coefficient of the layer angle combination based on the thermal deformation, and output the secondary optimization scheme of the layer; The verification module is used to verify the layer feasibility based on the secondary optimization scheme of the layer through the automatic fiber placement machine process database, and output the manufacturing ready package using the angle gradual transition strategy after verification is completed; The process simulation module is used to load the manufacturing ready package into the digital twin engine to simulate the actual layering process and output the final layer process parameter set.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that: The processor executes the computer program to realize the steps of the carbon fiber fuselage structure layer optimization design method of any one of claims 1-6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the carbon fiber fuselage structure layer optimization design method of any one of claims 1-6.

Images