Integrated design method for woven composite variable-curvature reinforced thin-wall structure

Abstract

The invention discloses an integrated design method for a woven composite variable-curvature reinforced thin-wall structure, and the method comprises the steps: achieving the parametric modeling of a design model of the woven composite variable-curvature reinforced thin-wall structure through a Python language; the method comprises the following steps of: modeling a woven composite material variable curvature reinforced thin-wall structure design model for each sample point sampled in a design space of design variables, performing instability mechanical response solution by utilizing finite element software explicit dynamics, and then selecting a design variable optimal solution based on a deep neural network model in combination with a multi-objective optimization algorithm; and according to the optimal solution of the design variable, the variable-curvature reinforced thin-wall structure of the woven composite material is constructed. According to the method, integrated structural design is conducted on the woven composite material and the variable-curvature reinforced thin-wall structure, the design space of the woven composite material variable-curvature reinforced thin-wall structure is rich, the bearing efficiency of the woven composite material variable-curvature reinforced thin-wall structure is improved, and the further lightweight design requirement of the related engineering field is met.

Description

Technical Field

[0001] The invention belongs to the field of finite element simulation and structural optimization design, and specifically relates to an integrated design method for a braided composite material variable curvature reinforced thin-wall structure. Background Art

[0002] Composite materials have the advantages of light weight, high modulus, high strength, designability, high temperature resistance, excellent thermal stability, fatigue resistance, corrosion resistance, and good processability. They are very suitable for fields with high requirements for both load-bearing and lightweight. Advanced composite materials have shown excellent application prospects, especially in the aviation field. Composite materials have developed from the initial non-load-bearing components to the secondary and main load-bearing components, which can achieve a significant effect of reducing the weight by 20%-30%. Nowadays, the composite materials used in large aircraft can even account for about 30% of the total materials used. In the future, the use of composite materials in helicopters and small aircraft will reach about 70%-80%.

[0003] At the same time, as a common engineering structure, the reinforced thin shell structure is widely used in the aerospace field. This structure improves its load-bearing capacity and stability by adding reinforcements in the thin shell structure, such as longitudinal or transverse reinforcing ribs and frames. The reinforcement can not only effectively improve the strength and stiffness of the shell structure under in-plane and out-of-plane loads, but also help control the local and overall buckling of the shell, thereby improving the stability of the structure; at the same time, the reasonable arrangement of reinforcement can improve the structural performance without significantly increasing the weight. The reinforced shell structure can form various complex curved shapes to meet specific functional and aesthetic requirements, and can improve the fatigue resistance and corrosion resistance of the shell.

[0004] Therefore, the combination of the two, the reinforced thin-walled structure made of woven composite materials has the advantages of light weight, excellent fatigue resistance, and easy processing. However, for the design of woven composite reinforced thin-walled structures, linear reinforcement is widely used and the rib height is consistent. The design space is limited and there is a lack of more design solutions. In addition, for such load-bearing components, the problem of buckling instability during the load-bearing process is also extremely critical, and stability design is also indispensable in the design and construction links. Traditional linear buckling analysis can only obtain the critical buckling load. Although the nonlinear buckling analysis method based on explicit dynamics can obtain the mechanical response of the entire process from nonlinear post-buckling behavior to collapse, this method has a high computational cost and low analysis efficiency, and it is difficult to promote it to engineering applications. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and to provide an integrated design method for a braided composite material variable curvature reinforced thin-wall structure.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The present invention provides a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, comprising the following steps:

[0008] Step S1: using Python language to develop a design model for a braided composite material variable curvature reinforced thin-walled structure, wherein the design variables are mechanical parameters, geometric parameters and braiding parameters of the braided yarn, and the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, rib curve parameters, and rib height, wherein the rib curve parameters include the rib curve form, rib curve endpoints, and rib curve control points;

[0009] Step S2: Sampling is performed in the design space of the design variables using a sampling algorithm, and a design model of a braided composite material variable curvature reinforced thin-walled structure is built for each sample point. Then, the finite element software Abaqus explicit dynamics is used to solve the instability mechanical response of each braided composite material variable curvature reinforced thin-walled structure design model, and key indicators describing the structural bearing capacity are extracted. The key indicators include the critical buckling load P cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co ; Obtain a data set consisting of design variables and key indicators;

[0010] Step S3: construct a deep neural network model, divide the data set into a training set and a validation set, train and validate the deep neural network model, and obtain a trained and validated deep neural network model;

[0011] Step S4: Combine the multi-objective optimization algorithm with the deep neural network model to construct a multi-objective optimization problem with key indicators as optimization objectives and design variables as optimization variables, solve the optimal solution of the design variables that meets the weight upper limit requirements and has the best key indicators, and construct a variable curvature reinforced thin-walled structure of a woven composite material with the optimal solution of the design variables.

[0012] Preferably, step S1 is specifically as follows:

[0013] Step S11: Open the notepad++ editor and import the Abaqus, Part, Material, Section, Assembly, Step, Interaction, Load, Mesh, Optimization, Job, Sketch, and Visualization modules;

[0014] Step S12: setting the curvature radius R of the thin-walled shell surface, establishing a thin-walled shell surface without a stiffener, unfolding the thin-walled shell surface into a two-dimensional plane through coordinate transformation, and then constructing two endpoints and two control points of a cubic spline curve of each stiffener in the two-dimensional plane to determine the geometric shape of the stiffener; wherein the two endpoints of the cubic spline curve are located on any two sides in the two-dimensional plane; the two control points CP1 and CP2 of the cubic spline curve are set to coordinate information in a rectangular plane coordinate system using a random sampling algorithm;

[0015] Step S13: transforming the thin-walled shell surface without reinforcing ribs in the two-dimensional plane and the two endpoints and two control points of the cubic spline curve of each reinforcing rib into a cylindrical coordinate system, and generating a spatial spline curve on the thin-walled shell surface in the cylindrical coordinate system;

[0016] Step S14: define the height of each reinforcement rib and complete the establishment of the parametric geometric model;

[0017] Step S15: Based on the mechanical parameters, geometric parameters and weaving parameters of the weaving yarn, the TexGen software is called by Python command to establish a mesoscale model of the plain weave composite material; wherein the mechanical parameters are the longitudinal elastic modulus E1, the transverse elastic modulus E2, the Poisson's ratio v 12 , longitudinal shear modulus G1, transverse shear modulus G2; the geometric parameters are the width a of a single yarn and the thickness b of a single yarn; the weaving parameters are the weft spacing S weft and warp spacing S warp ;

[0018] Step S16: importing the established mesoscopic scale model into the finite element software Abaqus to apply periodic boundary conditions, and calculating the ABD matrix corresponding to the mechanical parameters, geometric parameters and weaving parameters of different weaving yarns through the ABD matrix calculator plug-in of the finite element software Abaqus;

[0019] Step S17: Use the shell element of the finite element software Abaqus to establish a finite element model of a variable curvature reinforced thin-walled structure of a woven composite material. The material parameters in the finite element model are defined by the obtained ABD matrix. Macro recording is performed during the process of establishing the finite element model, and the macro-recorded Python code is imported into the Notepad++ editor. The parametric modeling of the design model of the variable curvature reinforced thin-walled structure of the woven composite material is completed.

[0020] More preferably, the constitutive equation of the ABD matrix is ​​as follows:

[0021]

[0022] Among them, ε x represents the positive strain of the braided yarn in the x direction, εy It represents the positive strain of the braided yarn in the y direction. The x and y directions are two mutually perpendicular directions in the two-dimensional plane of the thin-walled shell surface. The positive strain represents the relative measure of the stretching or compression of the braided yarn along the direction of the force under the action of the tensile or compressive force. The positive value represents the stretching and the negative value represents the compression. xy represents the shear strain on the two-dimensional plane; κ x represents the curvature of the braided yarn in the x direction; κ y represents the curvature of the braided yarn in the y direction; κ xy represents the shear curvature; N x is the resultant force in the x direction, N y is the resultant force in the y direction, N xy represents the shear force on the two-dimensional plane; M x ,M y ,M xy Respectively represent N x ,N y ,N xy The sub-matrices A and D represent the in-plane tensile stiffness matrix and the out-of-plane bending stiffness matrix, respectively, and the sub-matrix B represents the coupled stiffness matrix.

[0023] in,

[0024] Preferably, the sampling algorithm in step S2 adopts Latin hypercube sampling. During the Latin hypercube sampling process, the design interval of each parameter in the design variable is evenly divided into sub-intervals, points are randomly selected in the sub-intervals, and there is only one sample point in each sub-interval.

[0025] Preferably, the process of using the finite element software Abaqus explicit dynamics in step S2 to solve the instability mechanical response of the design model of the braided composite material variable curvature reinforced thin-walled structure is specifically as follows:

[0026] ① Set a reference point on the upper and lower surfaces of the thin-walled shell, couple the reference point and the upper and lower surfaces of the thin-walled shell through coupling instructions, and apply compression displacement boundary conditions to the reference point;

[0027] ② The finite element software Abaqus is used to display dynamics to simulate the compression process of the variable curvature reinforced thin-walled structure design model of the braided composite material, and the compression displacement and reaction force are extracted. The calculation process ensures that the total energy is 0 and the kinetic energy is less than 10% of the strain energy;

[0028] ③ Extract the critical buckling load P of the braided composite material variable curvature reinforced thin-walled structure design model during compression from the reaction force-displacement curve of the braided composite material variable curvature reinforced thin-walled structure design model during compression. cr , pre-buckling slope Kpre , post-buckling slope K post and critical collapse load P co .

[0029] More preferably, the critical buckling load P cr It refers to the minimum load value at which the design model of the braided composite material variable curvature reinforced thin-walled structure begins to buckle; the pre-buckling slope K pre It refers to the proportional coefficient of the linear relationship between load and displacement before the buckling of the design model of the braided composite material variable curvature reinforced thin-walled structure occurs; the post-buckling slope K post It refers to the proportional coefficient of the nonlinear relationship between load and displacement after the buckling of the design model of the braided composite material variable curvature reinforced thin-walled structure occurs; the critical collapse load P co It refers to the load value at which the variable curvature reinforced thin-walled structure design model of the braided composite material completely loses its bearing capacity and collapses when the load is further increased after buckling.

[0030] Preferably, the input of the deep neural network model in step S3 is the design variable, and the output is the key indicator.

[0031] Preferably, step S4 is specifically as follows:

[0032] Step S41: Construct to maximize the critical buckling load P cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co To optimize the target, the weight M is within the upper weight limit M * The following are constraints, and a multi-objective optimization problem with mechanical parameters, geometric parameters and weaving parameters of the braided yarn, as well as the curvature radius of the thin-walled shell surface of the thin-walled structure with variable curvature, rib curve parameters, and rib height as optimization variables;

[0033] Step S42: using the NSGA-II multi-objective optimization algorithm to solve the multi-objective optimization problem, specifically, using the NSGA-II multi-objective optimization algorithm to call the deep neural network model to iteratively optimize the mechanical parameters, geometric parameters and weaving parameters of the braided yarn, as well as the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, the rib curve parameters, and the rib height, and solving the Pareto solution set of the optimal parameter combination in the entire design space of the design variables, and selecting the optimal solution of the design variables in the Pareto solution set according to the emphasis on each optimization objective;

[0034] Step S43: constructing a braided composite material variable curvature reinforced thin-walled structure with the optimal solution of the design variables.

[0035] More preferably, the mathematical expression of the multi-objective optimization problem is:

[0036] Design variables: R,x1,x2,y1,y2,x3,x3,y4,y4,h1,h2,

[0037] E1,E2,v 12 ,G1,G2,S weft ,S warp ,a,b

[0038] Optimization goal: P cr ,K pre ,K post ,P co

[0039] Constraints: M≤M *

[0040] Among them, there are two stiffeners designed, x1, y1 and x2, y2 are the coordinates of the cubic spline curve control points CP1, CP2 of one of the stiffeners, x3, y3 and x4, y4 are the coordinates of the cubic spline curve control points CP1, CP2 of the other stiffener, and h1, h2 are the heights of the two stiffeners.

[0041] The present invention has the following beneficial effects:

[0042] The present invention provides an integrated design method for a woven composite material variable curvature reinforced thin-walled structure, which performs integrated structural design on the woven composite material and the variable curvature reinforced thin-walled structure, and studies the influence of the mechanical parameters of the composite material, the geometric parameters and weaving parameters of the woven yarn, and the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, the rib curve parameters, and the rib height on the complex instability mechanical response of the woven composite material variable curvature reinforced thin-walled structure. A finite element model for predicting the complex instability behavior of the woven composite material variable curvature reinforced thin-walled structure is established based on data-driven, and the influencing factors of the complex instability behavior of the woven composite material variable curvature reinforced thin-walled structure are comprehensively and detailedly explored, so as to achieve rapid prediction of nonlinear mechanical response. On this basis, the present invention performs reverse design of a braided composite material variable curvature reinforced thin-walled structure driven by optimal anti-instability performance of the structure, and uses the mechanical parameters of the composite material, the geometric parameters and braiding parameters of the braided yarn, and the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, the rib curve parameters, and the rib height as design variables to greatly enrich the design space of the braided composite material variable curvature reinforced thin-walled structure, and accurately and efficiently designs the optimal solution of the design variables that meet the weight upper limit requirements and have the best anti-instability performance based on the deep neural network model combined with the multi-objective optimization algorithm, and constructs a braided composite material variable curvature reinforced thin-walled structure according to the optimal solution of the design variables. The present invention is conducive to improving the load-bearing efficiency of the braided composite material variable curvature reinforced thin-walled structure and meeting the further lightweight design requirements in the relevant engineering field. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flow chart of the method of the present invention;

[0044] Figure 2 It is a schematic diagram of a braided composite material variable curvature reinforced thin-walled structure in a specific embodiment of the present invention;

[0045] Figure 3 It is a schematic diagram of the structure of the braided yarn in a specific embodiment of the present invention;

[0046] Figure 4 A flow chart of a design model for a braided composite material variable curvature reinforced thin-wall structure developed in the present invention;

[0047] Figure 5 It is a reaction force-displacement curve diagram in a specific implementation manner of the present invention. DETAILED DESCRIPTION

[0048] The present invention is further described below in conjunction with the accompanying drawings.

[0049] like Figure 1As shown, an integrated design method for a braided composite material variable curvature reinforced thin-walled material structure is provided. The Python language is used to implement the parametric modeling of the braided composite material variable curvature reinforced thin-walled structure design model. The braided composite material variable curvature reinforced thin-walled structure design model is modeled for each sample point sampled in the design space of the design variable. The finite element software Abaqus explicit dynamics is used to solve the instability mechanical response. Subsequently, the optimal solution of the design variable is selected based on the deep neural network model (DNN) combined with the multi-objective optimization algorithm. A braided composite material variable curvature reinforced thin-walled structure is constructed according to the optimal solution of the design variable. The method has the following steps:

[0050] Step S1: Develop a design model of a variable curvature reinforced thin-walled structure of a woven composite material using Python language to avoid the tedious process of repeatedly using the Abaqus interface to build the model. The design variables include the mechanical parameters, geometric parameters and weaving parameters of the woven yarn, and the curvature radius, rib curve parameters and rib height of the thin-walled shell surface of the variable curvature reinforced thin-walled structure. The rib curve parameters include the rib curve form, rib curve endpoints and rib curve control points. Each design variable can be parameterized.

[0051] like Figure 4 As shown, step S1 is specifically as follows:

[0052] Step S11: Open the notepad++ editor (Python editor), and import the Abaqus, Part, Material, Section, Assembly, Step, Interaction, Load, Mesh, Optimization, Job, Sketch, and Visualization modules required in the programming process;

[0053] Step S12: setting the radius of curvature R of the thin-walled shell surface, establishing a thin-walled shell surface without ribs, unfolding the thin-walled shell surface into a two-dimensional plane through coordinate transformation, and then constructing two endpoints and two control points of a cubic spline curve (rib curve form) of each rib in the two-dimensional plane to determine the geometric shape of the rib; wherein the two endpoints of the cubic spline curve are located on any two sides in the two-dimensional plane; the two control points CP1 and CP2 of the cubic spline curve are set to coordinate information in a rectangular plane coordinate system using a random sampling algorithm, that is, the control points CP1 and CP2 can be located at any position in the two-dimensional plane unfolded from the thin-walled shell surface;

[0054] Step S13: transforming the thin-walled shell surface without reinforcing ribs in the two-dimensional plane and the two endpoints and two control points of the cubic spline curve of each reinforcing rib into a cylindrical coordinate system, and generating a spatial spline curve on the thin-walled shell surface in the cylindrical coordinate system;

[0055] Step S14: define the height of each reinforcement rib and complete the establishment of the parametric geometric model;

[0056] Step S15: Based on the mechanical parameters, geometric parameters and weaving parameters of the weaving yarn, the TexGen software is called by Python command to establish a mesoscopic model of the plain weave composite material; wherein the mechanical parameters are five independent mechanical engineering constants, namely, the longitudinal elastic modulus E1, the transverse elastic modulus E2, the Poisson's ratio v 12 , longitudinal shear modulus G1, transverse shear modulus G2; E1 is the elastic modulus of a single yarn in the longitudinal direction (the long axis direction of the elliptical cross section), which describes the stiffness of the material when it is stretched or compressed in the longitudinal direction; E2 is the elastic modulus of a single yarn in the transverse direction (i.e. the short axis direction of the elliptical cross section), which describes the stiffness of the material when it is stretched or compressed in the transverse direction; v 12 It is the ratio of the transverse strain to the longitudinal strain of the woven composite material of the yarn when it is stretched, which describes the shrinkage characteristics of the material when it is stretched transversely; G1 is the shear modulus of a single yarn in the longitudinal direction, which describes the stiffness of the material when it is subjected to longitudinal shear force; G2 is the shear modulus of a single yarn in the transverse direction, which describes the stiffness of the material when it is subjected to transverse shear force; the geometric parameters are the width s of a single yarn (i.e. the length of the major axis of the elliptical yarn) and the thickness b of a single yarn (i.e. the length of the minor axis of the elliptical yarn); the weaving parameter is the weft spacing S weft (i.e. the distance between adjacent weft yarns in the fabric) and the warp spacing S warp (i.e. the distance between adjacent warp yarns in a fabric), e.g. Figure 3 As shown;

[0057] Step S16: Import the established mesoscale model into the finite element software Abaqus to apply periodic boundary conditions to ensure the homogenization of the yarn and matrix properties in the microstructure (i.e., the system is periodically repeated in space). This can reduce the boundary effect caused by the finite size and simulate the properties of the infinite system in a smaller calculation area to improve the calculation efficiency. The braided composite material of the braided yarn is composed of a periodically distributed mesostructure. The microstructural characteristics of the braided composite material are used to predict the macroscopic behavior of the material. The ABD matrix corresponding to the mechanical parameters, geometric parameters and braiding parameters of different braided yarns is calculated by the ABD matrix calculator plug-in of the finite element software Abaqus, which characterizes the basic concept of the effective stiffness characteristics of the braided composite material. The strain ε x ,ε y ,γ xy and curvature κ x ,κ y ,κ xy With the combined force N x ,N y ,N xyand bending moment M x ,M y ,M xy Its accurate calculation is crucial to analyze the complex performance of woven composite structures. The constitutive equation of the ABD matrix is ​​as follows:

[0058]

[0059] Among them, ε x represents the positive strain of the braided yarn in the x direction, ε y It represents the positive strain of the braided yarn in the y direction. The x and y directions are two mutually perpendicular directions in the two-dimensional plane of the thin-walled shell surface. The positive strain represents the relative measure of the stretching or compression of the braided yarn along the direction of the force under the action of the tensile or compressive force. The positive value represents the stretching and the negative value represents the compression. xy Represents the shear strain on a two-dimensional plane, which describes the displacement difference of the braided yarn in two perpendicular directions and reflects the degree of deformation of the braided yarn under the action of shear force. x represents the curvature of the braided yarn in the x direction, which describes the degree of bending of the braided yarn after being subjected to force in the x direction; κ y represents the curvature of the braided yarn in the y direction, which describes the degree of bending of the braided yarn after being subjected to force in the y direction; κ xy N stands for shear curvature, which describes the degree of bending of the braided yarn in a two-dimensional plane when subjected to shear force. x is the resultant force in the x direction, N y is the resultant force in the y direction, N xy M represents the shear force on a two-dimensional plane. x ,M y ,M xy Respectively represent N x ,N y ,N xy The sub-matrices A and D represent the in-plane tensile stiffness matrix and the out-of-plane bending stiffness matrix respectively, and the sub-matrix B represents the coupling stiffness matrix, which describes the mutual coupling between bending and stretching. All three sub-matrices are 3×3 symmetric matrices.

[0060] in,

[0061] Step S17: Use the shell element of the finite element software Abaqus to establish a finite element model of a braided composite material variable curvature reinforced thin-walled structure. The material parameters in the finite element model are defined by the obtained ABD matrix. Macro recording is performed during the establishment of the finite element model, and the macro-recorded Python code is imported into the Notepad++ editor. The parametric modeling of the design model of the braided composite material variable curvature reinforced thin-walled structure is completed. Figure 2 shown.

[0062] Step S2: Use the sampling algorithm to sample in the design space of the design variables, and build a design model of the braided composite material variable curvature reinforced thin-walled structure for each sample point. Then, use the finite element software Abaqus explicit dynamics to solve the instability mechanical response of each braided composite material variable curvature reinforced thin-walled structure design model, and extract the key indicators that describe the structural bearing capacity. The key indicators include the critical buckling load P cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co , and obtain a data set consisting of design variables and key indicators.

[0063] The sampling algorithm adopts Latin hypercube sampling. In the Latin hypercube sampling process, the design interval of each parameter in the design variable is evenly divided into sub-intervals, and points are randomly selected in the sub-interval to ensure that each sample point is randomly distributed in the sub-interval, and there is only one sample point in each sub-interval, that is, each level of each parameter is only studied once, and more sample points and more combinations can be studied for each parameter.

[0064] Among them, the process of using the finite element software Abaqus explicit dynamics to solve the instability mechanical response of the design model of the braided composite material variable curvature reinforced thin-walled structure is as follows:

[0065] ① Set a reference point on the upper and lower surfaces of the thin-walled shell respectively, couple the reference point and the upper and lower surfaces of the thin-walled shell through coupling instructions, and apply compression displacement boundary conditions to the reference point; the reference point can be used to control the entire surface; the reference point is a virtual point;

[0066] ② The finite element software Abaqus is used to display dynamics to simulate the compression process of the variable curvature reinforced thin-walled structure design model of the braided composite material, and the compression displacement and reaction force (equal to the applied load and opposite in direction) are extracted. The calculation process ensures that the total energy is 0 and the kinetic energy is less than 10% of the strain energy to ensure the realization of the quasi-static compression process;

[0067] ③ From the reaction force-displacement curve of the compression process of the design model of the braided composite material variable curvature reinforced thin-walled structure (such as Figure 5 As shown, Figure 5 The critical buckling load P of the braided composite material variable curvature reinforced thin-walled structure design model under compression is extracted from the load-reaction force and displacement-compression displacement. cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co.

[0068] The critical buckling load P cr It refers to the minimum load value at which the design model of the braided composite variable curvature reinforced thin-walled structure begins to buckle. When the applied load reaches this value, the structure will change from its initial stable state to an unstable state and begin to bend or twist. The pre-buckling slope K pre It refers to the proportional coefficient of the linear relationship between load and displacement before the buckling of the design model of the braided composite variable curvature reinforced thin-walled structure occurs, which represents the stiffness response of the structure to the load before buckling; the post-buckling slope K post It refers to the proportional coefficient of the nonlinear relationship between load and displacement after the buckling of the design model of the braided composite variable curvature reinforced thin-walled structure. After buckling, the stiffness of the structure will decrease, so the slope after buckling is usually smaller than the slope before buckling; the critical collapse load P co It refers to the load value at which the design model of the variable curvature reinforced thin-walled structure of the braided composite material further increases the load after buckling, resulting in a complete loss of bearing capacity and collapse. This load value is usually higher than the critical buckling load and is related to factors such as the nonlinear behavior of the material and the geometric defects of the structure.

[0069] Step S3: construct a deep neural network model, divide the data set into a training set and a validation set, train and validate the deep neural network model, and obtain a trained and validated deep neural network model. The input of the deep neural network model is the design variable, and the output is the key indicator.

[0070] Step S4: Combine the multi-objective optimization algorithm with the deep neural network model to construct a multi-objective optimization problem with key indicators as optimization targets and design variables as optimization variables, solve the optimal solution of the design variables that meets the weight upper limit requirements and describes the key indicators of the structural bearing capacity, and construct a braided composite material variable curvature reinforced thin-walled structure with the optimal solution of the design variables. Among them, the combination of the multi-objective optimization algorithm and the deep neural network model significantly improves the optimization efficiency. The details are as follows:

[0071] Step S41: Construct to maximize the critical buckling load P cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co To optimize the target, the weight M is within the upper weight limit M * The following are constraints, and a multi-objective optimization problem with mechanical parameters, geometric parameters and weaving parameters of the braided yarn, as well as the curvature radius of the thin-walled shell surface of the thin-walled structure with variable curvature, rib curve parameters, and rib height as optimization variables;

[0072] Step S42: Solve the multi-objective optimization problem using the NSGA-II multi-objective optimization algorithm. Specifically, the NSGA-II multi-objective optimization algorithm is used to call the deep neural network model to iteratively optimize the mechanical parameters, geometric parameters and weaving parameters of the braided yarn, as well as the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, the rib curve parameters, and the rib height, and the Pareto solution set of the optimal parameter combination is solved in the entire design space of the design variables, and the optimal solution of the design variables is selected from the Pareto solution set according to the emphasis on each optimization goal.

[0073] Step S43: constructing a braided composite material variable curvature reinforced thin-walled structure with the optimal solution of the design variables.

[0074] Among them, the mathematical expression of the multi-objective optimization problem is:

[0075] Design variables: R,x1,x2,y1,y2,x3,x3,y4,y4,h1,h2,

[0076] E1,e2,v 12 ,G1,G2,S weft ,S warp ,a,b

[0077] Optimization goal: P cr ,K pre ,K post ,P co

[0078] Constraints: M≤M *

[0079] Among them, there are two stiffeners designed, x1, y1 and x2, y2 are the coordinates of the cubic spline curve control points CP1, CP2 of one of the stiffeners, x3, y3 and x4, y4 are the coordinates of the cubic spline curve control points CP1, CP2 of the other stiffener, and h1, h2 are the heights of the two stiffeners.

Claims

1. A method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: The steps are as follows: Step S1: using Python language to develop a design model for a braided composite material variable curvature reinforced thin-walled structure, wherein the design variables are mechanical parameters, geometric parameters and braiding parameters of the braided yarn, and the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, rib curve parameters, and rib height, wherein the rib curve parameters include the rib curve form, rib curve endpoints, and rib curve control points; Step S2: Sampling is performed in the design space of the design variables using a sampling algorithm, and a design model of a braided composite material variable curvature reinforced thin-walled structure is built for each sample point. Then, the finite element software Abaqus explicit dynamics is used to solve the instability mechanical response of each braided composite material variable curvature reinforced thin-walled structure design model, and key indicators describing the structural bearing capacity are extracted. The key indicators include the critical buckling load P cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co ; Obtain a data set consisting of design variables and key indicators; Step S3: construct a deep neural network model, divide the data set into a training set and a validation set, train and validate the deep neural network model, and obtain a trained and validated deep neural network model; Step S4: Combine the multi-objective optimization algorithm with the deep neural network model to construct a multi-objective optimization problem with key indicators as optimization objectives and design variables as optimization variables, solve the optimal solution of the design variables that meets the weight upper limit requirements and has the best key indicators, and construct a variable curvature reinforced thin-walled structure of a woven composite material with the optimal solution of the design variables.

2. According to claim 1, a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: Step S1 is specifically as follows: Step S11: Open the notepad++ editor and import the Abaqus, Part, Material, Section, Assembly, Step, Interaction, Load, Mesh, Optimization, Job, Sketch, and Visualization modules; Step S12: setting the curvature radius R of the thin-walled shell surface, establishing a thin-walled shell surface without a stiffener, unfolding the thin-walled shell surface into a two-dimensional plane through coordinate transformation, and then constructing two endpoints and two control points of a cubic spline curve of each stiffener in the two-dimensional plane to determine the geometric shape of the stiffener; wherein the two endpoints of the cubic spline curve are located on any two sides in the two-dimensional plane; the two control points CP1 and CP2 of the cubic spline curve are set to coordinate information in a rectangular plane coordinate system using a random sampling algorithm; Step S13: transforming the thin-walled shell surface without reinforcing ribs in the two-dimensional plane and the two endpoints and two control points of the cubic spline curve of each reinforcing rib into a cylindrical coordinate system, and generating a spatial spline curve on the thin-walled shell surface in the cylindrical coordinate system; Step S14: define the height of each reinforcement rib and complete the establishment of the parametric geometric model; Step S15: Based on the mechanical parameters, geometric parameters and weaving parameters of the weaving yarn, the TexGen software is called by Python command to establish a mesoscale model of the plain weave composite material; wherein the mechanical parameters are the longitudinal elastic modulus E1, the transverse elastic modulus E2, the Poisson's ratio v 12 , longitudinal shear modulus G1, transverse shear modulus G2; the geometric parameters are the width a of a single yarn and the thickness b of a single yarn; the weaving parameters are the weft spacing S weft and warp spacing S warp ; Step S16: importing the established mesoscopic scale model into the finite element software Abaqus to apply periodic boundary conditions, and calculating the ABD matrix corresponding to the mechanical parameters, geometric parameters and weaving parameters of different weaving yarns through the ABD matrix calculator plug-in of the finite element software Abaqus; Step S17: Use the shell element of the finite element software Abaqus to establish a finite element model of a variable curvature reinforced thin-walled structure of a woven composite material. The material parameters in the finite element model are defined by the obtained ABD matrix. Macro recording is performed during the process of establishing the finite element model, and the macro-recorded Python code is imported into the Notepad++ editor. The parametric modeling of the design model of the variable curvature reinforced thin-walled structure of the woven composite material is completed.

3. According to claim 2, a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: The constitutive equation of the ABD matrix is ​​as follows: Among them, ε x represents the positive strain of the braided yarn in the x direction, ε y It represents the positive strain of the braided yarn in the y direction. The x and y directions are two mutually perpendicular directions in the two-dimensional plane of the thin-walled shell surface. The positive strain represents the relative measure of the stretching or compression of the braided yarn along the direction of the force under the action of the tensile or compressive force. The positive value represents the stretching and the negative value represents the compression. xy represents the shear strain on the two-dimensional plane; κ x represents the curvature of the braided yarn in the x direction; κ y represents the curvature of the braided yarn in the y direction; κ xy represents the shear curvature; N x is the resultant force in the x direction, N y is the resultant force in the y direction, N xy represents the shear force on the two-dimensional plane; M x ,M y ,M xy Respectively represent N x ,N y ,N xy The bending moment caused by the bending moment; sub-matrices A and D represent the in-plane tensile stiffness matrix and the out-of-plane bending stiffness matrix respectively, and sub-matrix B represents the coupling stiffness matrix; in, 4. According to claim 1, a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: The sampling algorithm described in step S2 adopts Latin hypercube sampling. During the Latin hypercube sampling process, the design interval of each parameter in the design variable is evenly divided into subintervals, points are randomly selected in the subintervals, and there is only one sample point in each subinterval.

5. According to claim 1, a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: The specific process of solving the instability mechanical response of the braided composite material variable curvature reinforced thin-walled structure design model using the finite element software Abaqus explicit dynamics in step S2 is as follows: ① Set a reference point on the upper and lower surfaces of the thin-walled shell, couple the reference point and the upper and lower surfaces of the thin-walled shell through coupling instructions, and apply compression displacement boundary conditions to the reference point; ② The finite element software Abaqus is used to display dynamics to simulate the compression process of the variable curvature reinforced thin-walled structure design model of the braided composite material, and the compression displacement and reaction force are extracted. The calculation process ensures that the total energy is 0 and the kinetic energy is less than 10% of the strain energy; ③ Extract the critical buckling load P of the braided composite material variable curvature reinforced thin-walled structure design model during compression from the reaction force-displacement curve of the braided composite material variable curvature reinforced thin-walled structure design model during compression. cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co .

6. The method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure according to claim 5, characterized in that: The critical buckling load P cr It refers to the minimum load value at which the design model of the braided composite material variable curvature reinforced thin-walled structure begins to buckle; the pre-buckling slope K pre It refers to the proportional coefficient of the linear relationship between load and displacement before the buckling of the design model of the braided composite material variable curvature reinforced thin-walled structure occurs; the post-buckling slope K post It refers to the proportional coefficient of the nonlinear relationship between load and displacement after the buckling of the design model of the braided composite material variable curvature reinforced thin-walled structure occurs; the critical collapse load P co It refers to the load value at which the variable curvature reinforced thin-walled structure design model of the braided composite material completely loses its bearing capacity and collapses when the load is further increased after buckling.

7. According to claim 1, a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: The input of the deep neural network model in step S3 is the design variable, and the output is the key indicator.

8. According to claim 1, a method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure, characterized in that: Step S4 is specifically as follows: Step S41: Construct to maximize the critical buckling load P cr , pre-buckling slope K pre , post-buckling slope K post and critical collapse load P co To optimize the target, the weight M is within the upper limit M * The following are constraints, and a multi-objective optimization problem with mechanical parameters, geometric parameters and weaving parameters of the braided yarn, as well as the curvature radius of the thin-walled shell surface of the thin-walled structure with variable curvature, rib curve parameters, and rib height as optimization variables; Step S42: using the NSGA-II multi-objective optimization algorithm to solve the multi-objective optimization problem, specifically, using the NSGA-II multi-objective optimization algorithm to call the deep neural network model to iteratively optimize the mechanical parameters, geometric parameters and weaving parameters of the braided yarn, as well as the curvature radius of the thin-walled shell surface of the variable curvature reinforced thin-walled structure, the rib curve parameters, and the rib height, and solving the Pareto solution set of the optimal parameter combination in the entire design space of the design variables, and selecting the optimal solution of the design variables in the Pareto solution set according to the emphasis on each optimization objective; Step S43: constructing a braided composite material variable curvature reinforced thin-walled structure with the optimal solution of the design variables.

9. The method for integrated design of a braided composite material variable curvature reinforced thin-walled material structure according to claim 8, characterized in that: The mathematical expression of the multi-objective optimization problem is: Design variables: R,x1,x2,y1,y2,x3,x3,y4,y4,h1,h2, E1,E2,v 12 ,G1,G2,S wef t,S warp ,a,b Optimization goal: P cr ,K pre ,K post ,P co Constraints: M≤M * Among them, there are two stiffeners designed, x1, y1 and x2, y2 are the coordinates of the cubic spline curve control points CP1, CP2 of one of the stiffeners, x3, y3 and x4, y4 are the coordinates of the cubic spline curve control points CP1, CP2 of the other stiffener, and h1, h2 are the heights of the two stiffeners.

Images