Prediction method for curing deformation of composite laminated plate component

Abstract

The invention relates to a method for predicting curing deformation of a composite laminated plate component, which comprises the following steps of: firstly, acquiring data of a single-layer composite material by using an experiment to obtain characteristic strain data of the single-layer composite material in a curing process, and then, calculating characteristic strain in a single-layer plane plate based on a total-amount theory to obtain the characteristic strain of the single-layer plane plate. And finally, by utilizing a characteristic strain transfer principle, mapping the inherent strain of the single-layer composite material into the composite material component, predicting a curing deformation result of the composite material laminated plate component, and finally establishing a bridge between the microstructure change of the material and the macromechanical response in the curing process based on the characteristic strain. The curing deformation of the composite laminated plate component is predicted. According to the method, complex manufacturing mechanism exploration is avoided, the curing deformation quantity of the composite laminated plate component is directly predicted through the total theory and the like, the flow-thermosetting three-field coupling evolution process is simplified into a structure deformation model, the prediction process is simplified, and the operability and practicability of prediction are improved.

Description

Technical Field

[0001] The invention relates to a method for predicting the curing deformation of a composite material laminate component, and belongs to the technical field of resin-based composite material curing. Background Art

[0002] Composite laminate components are plate-like structural materials formed by bonding two or more layers of fiber-reinforced composite materials together through a specific process. The layup of composite laminate components usually includes two structures: unidirectional tape and two-dimensional fabric. Each single layer can have a different fiber layup direction. By rationally designing the parameters of each layer, the laminate can meet specific mechanical performance requirements in different directions.

[0003] Composite laminate components are widely used in aviation, aerospace, and civil industries due to their high strength-to-weight ratio, corrosion resistance, and fatigue resistance. However, residual stress during the curing process may lead to curing deformation, making the structure unable to be assembled or even scrapped, which restricts manufacturing quality and application effects. Therefore, it is necessary to simulate the curing process of composite laminate components (unidirectional tape or fabric) and predict the curing deformation to optimize the curing process.

[0004] Curing deformation introduced during the manufacturing process is influenced by a variety of factors, including curing temperature, mold effects, material properties, and curing pressure. Traditional methods for predicting curing deformation often rely on complex process modeling, requiring the coupling of multiple physical fields involved in the curing process, including temperature, chemical (resin curing reaction), and stress. These methods are limited by the modeling difficulty, time consumption, and high cost, as well as the difficulty in capturing dynamic material properties. This results in low efficiency and inability to quickly and accurately predict curing deformation.

[0005] The applicant searched and found that the Chinese patent with publication number CN115470669A discloses a method for predicting the curing deformation of composite material structures. This method generates sample sets in batches based on finite element models, and by constructing a convolutional neural network, establishes a mapping relationship between the laying angle of the composite material and the curing deformation cloud map of a given component structure, thereby realizing the rapid prediction of the component curing deformation cloud map. However, this method requires the generation of a large number of training sample sets, which not only consumes a lot of computing resources and time costs, but also has high requirements on the performance of hardware equipment; on the other hand, the construction, training and parameter tuning process of the convolutional neural network is relatively complicated, involving a lot of professional algorithm knowledge and programming operations. Engineering and technical personnel need to have a deep theoretical foundation and practical experience to master it proficiently. The above problems make the promotion and application of this method among the engineering and technical personnel group face great obstacles, and it is difficult to serve actual engineering production quickly and efficiently. Summary of the Invention

[0006] The present invention aims at composite material laminate components and proposes a method for predicting the curing deformation of composite material laminate components, which improves simulation efficiency while ensuring prediction accuracy. It can effectively solve the defects of traditional prediction methods such as long time consumption, high cost, and difficulty in measuring material parameters.

[0007] In order to solve the above technical problems, the present invention proposes a technical solution: a method for predicting the curing deformation of a composite laminate component, comprising the following steps: Step 1: Select multiple composite single-layer planar panels (unidirectional tapes or two-dimensional woven fabrics), each with a different fiber layup orientation; perform a curing test on each single-layer planar panel to obtain deformation results of multiple nodes on each single-layer planar panel before and after curing; Step 2: For each single-layer planar plate from Step 1, divide the grid into units based on the nodes. Based on the full-quantity theory, calculate the curing deformation characteristic strain of each grid unit of the single-layer planar plate based on the deformation results of all nodes before and after curing. Then, select the average characteristic strain of all units on the single-layer planar plate as the representative characteristic strain of the single-layer planar plate.

[0008] Step 3: Divide the composite laminate to be predicted into grid units, wherein the composite component has at least two plies, and the overall deformation of the composite laminate component is regarded as the superposition of the characteristic strains of each ply, wherein the characteristic strain of each grid unit of each ply is the representative characteristic strain of a single-layer planar plate with the same fiber ply direction as that of the ply; Step 4: using the finite element method to solve the thermal-mechanical static equilibrium of the composite laminate component to obtain the curing deformation of the composite laminate component.

[0009] The present invention simulates the curing process of composite laminate components through three stages: "simulation experiment + full-quantity theory + characteristic strain transfer". First, the data of the single-layer plane plate is collected by experiment (which can be a simulation experiment or an actual experiment), and the characteristic strain data of the single-layer plane plate during the curing process is systematically obtained. Then, based on theoretical formulas such as the full-quantity theory, the characteristic strain in the single-layer plane plate is calculated. Finally, the characteristic strain transfer principle is used to map the inherent strain of the single-layer plane plate to the composite laminate component, and the curing deformation result of the composite laminate component is predicted. Finally, a bridge is established between the microstructure change of the material and the macroscopic mechanical response during the curing process based on the characteristic strain, so as to realize the prediction of the curing deformation of the composite laminate component. The establishment of this inverse model avoids the complex exploration of manufacturing mechanisms, breaks through the limitations of traditional methods in predicting the curing deformation of composite laminate structures, can quickly predict the curing deformation of complex structures, reduce the number of trial and error tests, and provide theoretical support and technical means for quality control and performance optimization in the manufacturing process of composite components.

[0010] The present invention avoids the complex exploration of manufacturing mechanisms and directly predicts the solidification deformation of composite laminate components through full-quantity theory, simplifies the flow-heat-solid three-field coupled evolution process into a structural deformation model, simplifies the prediction process, and improves the operability and practicality of the prediction.

[0011] The present invention establishes a mapping relationship between the laying angle and the curing deformation value based on the idea of the characteristic strain method to obtain the prediction result. There is no need to obtain some difficult-to-obtain material performance data, nor does it require manual experience to accurately set the parameters. The curing deformation cloud map of the continuous fiber reinforced composite laminate component can be obtained within seconds. The calculation speed is fast and has high accuracy. It overcomes the low calculation efficiency of the existing finite element analysis method, provides a new perspective and method for the prediction of the strain of the composite component laminate after curing, and also promotes the further development and wide application of composite material technology. It is of great significance to improving the quality and reliability of composite products, reducing production costs, and promoting scientific and technological progress in related industries. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Schematic diagram of a U-shaped complex component in an embodiment of the present invention.

[0013] Figure 2 Schematic diagram of the curing deformation prediction results of the method according to an embodiment of the present invention.

[0014] Figure 3 Schematic diagram of the solidification deformation prediction results of the traditional prediction method. DETAILED DESCRIPTION

[0015] This embodiment shows a method for predicting the curing deformation of a composite material laminate component to improve the simulation efficiency of the curing deformation of a composite material laminate component. Figure 1 As shown, this embodiment takes a U-shaped fairing as an example to explain in detail the method for predicting the curing deformation of a composite laminate component. The U-shaped fairing is composed of 15 plies (all of which are two-dimensional woven plies), forming a U-shaped complex laminate component with 15 plies. The ply information is shown in Table 1:

[0016] It can be seen that the 15 plies of the U-shaped fairing have only two fiber laying directions, namely 0 degrees and 45 degrees.

[0017] The method for predicting curing deformation of a composite laminate member of this embodiment includes the following steps: Step 1: Select multiple single-layer composite planar panels (which can be unidirectional tapes or two-dimensional woven fabrics), each with a different fiber layup orientation; perform a curing simulation test on each single-layer planar panel to obtain the deformation results of multiple nodes on each single-layer planar panel before and after curing.

[0018] As previously mentioned, the U-shaped fairing in this embodiment is constructed from 15 two-dimensional woven plies, each containing only two fiber layup orientations. Therefore, this embodiment requires only two single-layer composite panels, with fiber layup orientations of 0 and 45 degrees, respectively. Curing tests on these single-layer panels can be conducted in the laboratory by first preparing the panels and, after curing, measuring the deformation at pre-determined nodes to collect deformation data. Alternatively, curing simulations can be conducted using existing finite element simulation models to obtain deformation results at different nodes.

[0019] This embodiment preferably uses a curing simulation test. During implementation, ABAQUS software can be used to perform a curing simulation of a single-layer flat plate. The simulation process is divided into two parts: thermochemical simulation and thermodynamic simulation. First, the mold and single-layer flat plate are imported into ABAQUS software, meshed, and mesh properties are set. Basic properties such as material and layup angle (i.e., fiber layup direction) are set. Initial conditions and boundary conditions are set to simulate the thermochemical temperature field of the composite material component during curing deformation. Then, a composite material curing constitutive model is established, and the temperature field results obtained from the thermochemical simulation are read in as the temperature field loading. Initial conditions and boundary conditions related to curing and demolding are set to obtain the curing deformation results of the single-layer flat plate. The deformation results of each node and unit are stored and accessed in the form of a txt file.

[0020] Step 2: For each single-layer planar panel from Step 1, calculate the curing deformation characteristic strain of each unit in the single-layer planar panel based on the deformation results of all nodes before and after curing, using the full-quantity theory. This is a prior art technique and will not be described in detail here. The average of the characteristic strains of all units in the single-layer planar panel is selected as the representative characteristic strain of the single-layer planar panel, thereby correlating the characteristic strain with the corresponding material and fiber layup direction.

[0021] Step 3: Divide the composite laminate component to be predicted (made of the same material as the single-layer plane plate) into grid units, wherein the composite laminate component has at least two plies, and the overall deformation of the composite laminate component is regarded as the superposition of the characteristic strains of each ply, wherein the characteristic strain of each unit of each ply is the representative characteristic strain of the single-layer plane plate with the same fiber ply direction as that of the ply.

[0022] This embodiment utilizes the principle of characteristic strain transfer to map the characteristic strain of a single-layer planar plate to a complex composite laminate component. The normal and fiber directions of specific elements in the composite laminate model are determined. By analyzing the fiber layup orientation of the complex component, the representative characteristic strain of the single-layer planar plate is read for the same layup angle. Each element in the composite component is assigned a corresponding characteristic strain (element strain), while maintaining the normal direction of each element consistent with the composite laminate model. This results in the global characteristic strain (element strain) of the final complex component.

[0023] Since composite materials have consistent thermal, chemical, and mechanical behaviors during the curing process (such as resin curing shrinkage and fiber constraint effects), the overall deformation of composite laminate components can be regarded as the superposition result of the characteristic strains of each ply. This embodiment utilizes the characteristic strain transfer principle and distributes the characteristic strains of simple laminates to the various plies of a complex structure through transfer rules, and maps the characteristic strains of simple laminates to complex composite components. This avoids the complex exploration of manufacturing mechanisms and directly predicts the curing deformation of composite materials through full-quantity theory, simplifying the flow-heat-solid three-field coupled evolution process into a structural deformation model, thereby simplifying the prediction process and improving the operability and practicality of the prediction.

[0024] Step 4: using the finite element method to solve the thermal-mechanical static equilibrium of the composite laminate component to obtain the curing deformation of the composite laminate component.

[0025] For example, this embodiment is based on the composite laminate component model that has been given characteristic strain in step three, establishes a shell unit composite model in ABAQUS, introduces the characteristic strain tensor as a pseudo thermal expansion coefficient tensor, and calculates the curing deformation of the composite laminate component by applying a unit temperature field to perform thermal-mechanical static equilibrium analysis. This is a prior art and relevant literature can be referred to and will not be repeated here.

[0026] The applicant conducted comparative tests using a traditional prediction method and the method of this embodiment. The traditional prediction method uses the existing flow field-thermochemistry-thermodynamics sequential coupling method to simulate the curing molding process, and performs autoclave flow field analysis, composite material thermal-chemical analysis and thermo-mechanical analysis in sequence to finally obtain the stress, strain and deformation results of the composite material component. Obviously, although the results of the traditional prediction method are relatively accurate, the analysis process is relatively complex and inefficient, and it takes a long time, and it is impossible to quickly obtain simulation results.

[0027] Figure 2 The prediction results (displacement deformation values) of the method of this embodiment are shown. Figure 3This is the prediction result (displacement deformation value) obtained by the traditional prediction method. It can be seen that the overall deformation trend of the method in this embodiment is consistent with that of the traditional prediction method, that is, the overall deformation trend is a trend of both sides converging toward the middle, the deformation values at the outermost ends on both sides are the largest, and the error of the maximum deformation value is within 10%. Furthermore, since the representative characteristic strains of the single-layer planar plate obtained in steps 1 and 2 of the method of this embodiment can be reused, that is, the same set of data can be used for multiple predictions, a database of curing deformation parameters of single-layer planar plates covering different materials and different layup angles can be established in the early stages: the database can store characteristic strain parameters for various materials (e.g., carbon fiber / epoxy, glass fiber / polyester) and layups (e.g., unidirectional, symmetric, asymmetric), allowing direct call-up of similar cases during design, avoiding repeated experiments or simulations and significantly shortening the R&D cycle; the diverse data accumulated in the database can provide statistical support for characteristic strain transfer rules, for example, through interpolation or machine learning algorithms, predicting characteristic strains of unverified material combinations or complex layups, thus overcoming the limitations of traditional empirical models; the database provides high-quality training sets for supervised learning (e.g., neural networks and random forests), enabling the construction of nonlinear mapping models between characteristic strains and material / process parameters, replacing some analytical calculations and further improving efficiency; the database supports the dynamic addition of new experimental or simulation data to continuously improve model accuracy, forming a closed loop of "data acquisition → model update → prediction optimization."

[0028] According to the applicant's experimental results, after removing the time for calculating the representative characteristic strain of a single-layer plane plate, this embodiment saves about 75% of the calculation time compared to traditional prediction methods, greatly improving the prediction speed of solidification deformation.

[0029] It should be noted that in this embodiment, when performing a curing simulation test on a single-layer flat panel in step one, the curing process parameters such as the heating rate, holding temperature, cooling rate, and pressure are kept consistent with the process parameters in steps three and four, and the curing time required for both must be able to ensure that the composite material is completely cross-linked.

[0030] This embodiment can also be improved as follows: in step 1, when calculating the deformation results of multiple nodes on the single-layer plane plate before and after solidification, only the deformation results of the nodes on the middle surface of the single-layer plane plate before and after solidification are calculated; Accordingly, in step 2, the grid units are divided according to the nodes on the middle surface of the single-layer plane plate, and the curing deformation characteristic strain of each unit on the middle surface of the single-layer plane plate is calculated based on the deformation results of the nodes on the middle surface of the single-layer plane plate before and after curing. The representative characteristic strain of the single-layer plane plate is the average value of the curing deformation characteristic strain of each unit on the middle surface of the single-layer plane plate.

[0031] The midplane of the single-layer planar plate is a plane located directly in the middle of the plate's thickness, dividing it into two equal parts, upper and lower, and geometrically located at the center of the plate. Through the aforementioned improvements, this embodiment only calculates the deformation results of the nodes on the midplane of the single-layer planar plate. This simplifies the three-dimensional problem into a two-dimensional one, reducing the computational effort and complexity, improving computational efficiency, and simultaneously capturing the key characteristics of the plate's deformation, as the deformation of the midplane can, to a certain extent, reflect the overall bending and deformation trends of the plate.

Claims

1. A method for predicting curing deformation of a composite laminate component, comprising the following steps: Step 1: Select multiple composite single-layer planar panels, each single-layer planar panel having a different fiber layup direction; perform a curing test on each single-layer planar panel to obtain deformation results of multiple nodes on each single-layer planar panel before and after curing; Step 2: For each single-layer plane plate in step 1, divide the grid cells according to the nodes. Based on the full quantity theory, calculate the curing deformation characteristic strain of each grid cell of the single-layer plane plate according to the deformation results of all nodes before and after curing, and select the average value of the characteristic strain of all cells on the single-layer plane plate as the representative characteristic strain of the single-layer plane plate. Step 3: Divide the composite laminate component to be predicted into grid units, wherein the composite laminate component has at least two plies, and regard the overall deformation of the composite laminate component as the superposition of the characteristic strains of each ply, wherein the characteristic strain of each grid unit of each ply is the representative characteristic strain of a single-layer planar plate with the same fiber ply direction as that of the ply; Step 4: using the finite element method to solve the thermal-mechanical static equilibrium of the composite laminate structure to obtain the curing deformation of the composite structure.

2. The method for predicting curing deformation of composite laminated plate components according to claim 1, characterized in that: In step 1, the curing test for each single-layer flat panel is a curing simulation test.

3. The method for predicting curing deformation of composite laminated plate components according to claim 1, characterized in that: In step 1, when calculating the deformation results of multiple nodes on the single-layer plane plate before and after solidification, only the deformation results of the nodes on the middle surface of the single-layer plane plate before and after solidification are calculated; Accordingly, in step 2, the grid units are divided according to the nodes on the middle surface of the single-layer plane plate, and the curing deformation characteristic strain of each unit on the middle surface of the single-layer plane plate is calculated based on the deformation results of the nodes on the middle surface of the single-layer plane plate before and after curing. The representative characteristic strain of the single-layer plane plate is the average value of the curing deformation characteristic strain of each unit on the middle surface of the single-layer plane plate.

Images