Method for obtaining ablation morphology of c / c composite material based on physical information neural network

Abstract

The application discloses a method for obtaining C / C composite ablation morphology based on physical information neural network, and relates to the field of aerospace, comprising the following steps: taking the coordinate data of a C / C composite three-dimensional space model under a Cartesian coordinate system, an ablation time, a dimensionless parameter Sh and a dimensionless parameter A as inputs of a pre-trained multilayer perceptron network, outputting oxygen molar concentration and solid volume fraction through the pre-trained multilayer perceptron network; solving a gas phase diffusion equation and a solid phase evolution equation through the oxygen molar concentration and the solid volume fraction output by the multilayer perceptron network to obtain the C / C composite ablation morphology. The method can efficiently and accurately solve the gas-solid coupling equation, overcome the limitations of traditional numerical methods, accurately obtain the ablation morphology of the C / C composite under the gas-solid coupling effect at a low calculation cost, and provide strong theoretical support and technical support for the application of the C / C composite in the field of aerospace and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of aerospace, in particular to a method for obtaining ablation morphology of C / C composite material based on physical information neural network. BACKGROUND

[0002] C / C composite material refers to a pure carbon multi-phase structure composed of carbon fiber or its fabric as the reinforcing phase and pyrolytic carbon or resin carbon, pitch carbon as the matrix by chemical vapor infiltration. Due to its low density, excellent ablation resistance, good high-temperature mechanical properties and other outstanding advantages, it has been widely used in aerospace fields such as thermal protection system of space vehicles, throat liner of rocket engines, etc. In the actual service process, C / C composite material will inevitably face extreme complex gas-solid coupling environment such as high temperature and high speed airflow, and thermal chemical ablation phenomenon will inevitably occur. It is crucial to accurately predict the ablation morphology of C / C composite material under the action of gas-solid coupling. Traditional experimental test methods, such as oxygen-ethyne ablation test, plasma ablation test, etc., can directly obtain the ablation data and surface morphology of the material, but have problems such as high cost, long cycle, complex working condition control, etc., and it is difficult to fully reveal the internal physical mechanism in the ablation process. Traditional numerical methods based on finite element method, finite difference method, etc. have low computational efficiency, difficulty in meshing, limited ability to handle complex boundary conditions, etc. when dealing with complex geometry, highly nonlinear and multi-scale gas-solid coupling problems.

[0003] In recent years, deep learning technology has developed rapidly. As a new numerical calculation method, physical information neural network (PINN) can solve differential equations without mesh by embedding physical prior knowledge such as physical equations and boundary conditions into neural network. However, there is no mature and systematic method or system based on PINN to solve gas-solid coupling equations to obtain the ablation morphology of C / C composite material. Therefore, it is of great significance to develop an efficient and accurate method or system based on PINN to deeply understand the ablation mechanism of C / C composite material, optimize material design and improve the reliability of related engineering applications. SUMMARY

[0004] In view of the above shortcomings in the prior art, the method for obtaining ablation morphology of C / C composite material based on physical information neural network provided by the present application can efficiently and accurately solve gas-solid coupling equations, overcome the limitations of traditional numerical methods, and accurately obtain the ablation morphology of C / C composite material under the action of gas-solid coupling with low computational cost, providing strong theoretical support and technical guarantee for the application of C / C composite material in aerospace and other fields.

[0005] In order to achieve the above application purposes, the technical scheme adopted by the present application is as follows:

[0006] A method for obtaining C / C composite ablation morphology based on physical information neural network is provided, comprising the following steps:

[0007] Constructing a gas-solid coupling equation of C / C composite in a micro-ablation model;

[0008] Obtaining dimensionless parameter Sh and dimensionless parameter A of micro-ablation behavior of C / C composite;

[0009] Constructing a pre-trained multi-layer perception network; constructing a three-dimensional space model of C / C composite;

[0010] Taking the coordinate data of the three-dimensional space model of C / C composite in the Cartesian coordinate system, the ablation time, the dimensionless parameter Sh and the dimensionless parameter A as inputs of the pre-trained multi-layer perception network, and outputting the oxygen molar concentration and the solid volume fraction through the pre-trained multi-layer perception network;

[0011] Based on the gas-solid coupling equation, obtaining a gas phase diffusion equation and a solid phase evolution equation of C / C composite in the micro-ablation model; setting initial conditions and boundary conditions of the gas phase diffusion equation and the solid phase evolution equation;

[0012] Solving the gas phase diffusion equation and the solid phase evolution equation through the oxygen molar concentration and the solid volume fraction output by the multi-layer perception network to obtain the ablation morphology of C / C composite.

[0013] Further, the expression of the gas-solid coupling equation is:

[0014]

[0015]

[0016] wherein represents a solid phase; represents time; represents an interface recession velocity, , represents a solid molar volume of the phase, represents a molar rate of oxidation at the interface, , is an oxidation reaction rate, represents a local normal vector at the interface, , represents a gradient operator, represents a gradient of the solid phase, represents a module of represents an oxygen molar concentration; is a diffusion coefficient.

[0017] Further, the calculation expression of the dimensionless parameter Sh is:

[0018]

[0019] wherein is the fiber radius of the C / C composite material; is the oxidation reaction rate of the matrix of the C / C composite material; is the diffusion coefficient.

[0020] Further, the calculation expression of the dimensionless parameter A is:

[0021]

[0022] wherein is the oxidation reaction rate of the matrix of the C / C composite material; is the molar volume of the matrix phase; is the oxidation reaction rate of the carbon fiber in the C / C composite material; is the molar volume of the fiber phase.

[0023] Further, the construction method of the pre-trained multi-layer perception network comprises the following steps:

[0024] The number of hidden layers and the number of neurons in each hidden layer of the multi-layer perception network are set, and an activation function is used for nonlinear transformation in the hidden layer; wherein the number of hidden layers and the number of neurons are determined by the open source framework Optuna;

[0025] A gas phase diffusion loss function, a solid phase evolution loss function and an initial condition loss function are constructed, and then a total loss function of the multi-layer perception network is obtained;

[0026] A C / C composite material three-dimensional space model with known ablation morphology and its initial condition are taken as a calculation domain, and points are taken in the calculation domain; wherein the point taking mode is uniform sampling, random sampling or adaptive sampling; the taken points contain oxygen molar concentration and solid volume fraction;

[0027] The coordinates, ablation time, dimensionless parameter Sh and dimensionless parameter A of the taken points are input into the multi-layer perception network as training samples, and the oxygen molar concentration and solid volume fraction predicted by the multi-layer perception network are obtained;

[0028] The oxygen molar concentration and solid volume fraction contained in the taken points are taken as real labels, combined with the oxygen molar concentration and solid volume fraction predicted by the multi-layer perception network, and the total loss value is calculated by the total loss function of the multi-layer perception network;

[0029] The parameters of the multi-layer perception network are adjusted to minimize the total loss value until a preset training end condition is reached, and a pre-trained multi-layer perception network is obtained.

[0030] Further, the expression of the total loss function of the multi-layer perception network is:

[0031]

[0032] Among them is the total loss function; , and is the weight coefficient; is the gas phase diffusion loss function, ; is the solid phase evolution loss function, ; is the initial condition loss function, ; represents the molar concentration of oxygen; represents time; represents the gradient operator; is the solid volume fraction, when represents the gas phase unit, when represents the solid phase unit; is the diffusion coefficient; is the molar concentration of the reaction gas; is the square of the L2 norm; is the speed of solid retreat; is the outer edge concentration of the boundary layer.

[0033] Further, the specific method for constructing the three-dimensional space model of C / C composite material is:

[0034] A three-dimensional space model of C / C composite material is established, the coordinate data of the three-dimensional space model of C / C composite material in the Cartesian coordinate system is obtained, and the cross-sectional explanation of the three-dimensional space model of C / C composite material is given.

[0035] Further, the gas phase diffusion equation is:

[0036]

[0037] Among them, in the fluid domain and the solid domain , at the interface ; is the solid volume fraction; C is the molar concentration of the reaction gas; is the heterogeneous reaction rate; represents the inverse of the gradient module length.

[0038] Further, the solid phase evolution equation is:

[0039]

[0040] wherein is the solid volume fraction; is the solid recession velocity; in the fluid domain and the solid domain , at the interface ; is the solid volume fraction; the molar concentration of the reaction gas; is the heterogeneous reaction rate; represents the reciprocal of the gradient modulus.

[0041] Further, the specific method for obtaining the ablation morphology of the C / C composite material by solving the gas phase diffusion equation and the solid phase evolution equation through the oxygen molar concentration and the solid volume fraction output by the multilayer perception network is:

[0042] Through the oxygen molar concentration and the solid volume fraction output by the multilayer perception network, the evolution of the oxygen molar concentration and the solid volume fraction distribution of the material at different times is tracked through the gas phase diffusion equation and the solid phase evolution equation, and then the development process of the ablation morphology of the three-dimensional C / C composite material with time is obtained.

[0043] The beneficial effects of the present application are: the present method can efficiently and accurately solve the gas-solid coupling equation, overcome the limitations of traditional numerical methods, accurately obtain the ablation morphology of the C / C composite material under the action of gas-solid coupling with low calculation cost, and provide strong theoretical support and technical support for the application of C / C composite materials in the field of aerospace and the like. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 is a flowchart of the present method;

[0045] Figure 2 is a schematic diagram of the calculation domain of the three-dimensional model in the embodiment;

[0046] Figure 3 is a y-z cross-sectional view of the three-dimensional model;

[0047] Figure 4 is a schematic diagram in which the data points of each region when solving the coupled Stokes-Darcy equation are represented by different colors and symbols;

[0048] Figure 5 is a schematic diagram of the point selection principle of the PINN in the calculation domain;

[0049] Figure 6 is a schematic diagram of the configuration points of the three-dimensional model and the cross-sectional configuration points; wherein (a) is the three-dimensional configuration point, and (b) is the cross-sectional configuration point;

[0050] Figure 7 Variation curves of equation loss, boundary loss and total loss for a multi-layer perception network trained for 120 times;

[0051] Figure 8 Evolution of the gas-solid interface at different times;

[0052] Figure 9 Ablation morphology of a single fiber at different Da numbers. DETAILED DESCRIPTION

[0053] The specific embodiments of the present application are described below to facilitate the understanding of the present application for those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments, and for those skilled in the art, it is obvious that various changes are within the spirit and scope of the present application defined and determined by the appended claims, and all the inventions utilizing the concept of the present application are within the scope of protection.

[0054] As shown in Figure 1 The method for obtaining the ablation morphology of C / C composite material based on the physical information neural network comprises the following steps:

[0055] S1, constructing a gas-solid coupling equation of C / C composite material in a micro-ablation model;

[0056] S2, obtaining dimensionless parameter Sh and dimensionless parameter A of micro-ablation behavior of C / C composite material;

[0057] S3, constructing a pre-trained multi-layer perception network; constructing a three-dimensional space model of C / C composite material;

[0058] S4, taking the coordinate data of the three-dimensional space model of C / C composite material in the Cartesian coordinate system, the ablation time, the dimensionless parameter Sh and the dimensionless parameter A as the input of the pre-trained multi-layer perception network, and outputting the oxygen molar concentration and the solid volume fraction through the pre-trained multi-layer perception network;

[0059] S5, based on the gas-solid coupling equation, obtaining a gas phase diffusion equation and a solid phase evolution equation of C / C composite material in the micro-ablation model; setting initial conditions and boundary conditions of the gas phase diffusion equation and the solid phase evolution equation;

[0060] S6, solving the gas phase diffusion equation and the solid phase evolution equation through the oxygen molar concentration and the solid volume fraction output by the multi-layer perception network to obtain the ablation morphology of C / C composite material.

[0061] The material of the carbonized composite at the ablation surface mainly contains carbon elements (carbon fibers and carbon matrix), and the thermal-chemical ablation phenomenon on the surface mainly includes oxidation, nitridation and sublimation of carbon. Generally, the carbonized composite mainly occurs thermal-chemical ablation phenomenon caused by oxidation, so we only study the oxidation reaction of the carbonized composite. There is a thermal-chemical ablation phenomenon near the ablation surface of the C / C composite, and a gas flow boundary layer appears around it. In the boundary layer, complex heat and mass transfer processes occur, which affect the ablation behavior. Previous studies have shown that the oxygen transport in the boundary layer at the microscale is controlled by the pure diffusion mechanism, and oxygen will enter the interior of the boundary layer from the outer edge of the channel in the form of pure diffusion and react with the carbon elements on the material surface, which will also cause the degradation of the gas-solid interface.

[0062] In this embodiment, the molar concentration distribution of oxygen in the boundary layer and the carbon reaction rate are the key parameters to characterize the recession of the ablation surface of the C / C composite. Gas diffusion determines how oxygen is transported to the surface of the composite and distributed inside the material, while geometric changes mainly involve the shape and size changes of the composite due to the consumption of carbon during oxidation, which not only affects the mechanical properties of the material, but also affects the gas diffusion path and the characteristics of the boundary layer. Therefore, the oxidation process of the carbon fiber reinforced composite can be described based on gas diffusion and geometric changes. Specifically, the expression of the gas-solid coupling equation is:

[0063]

[0064]

[0065] wherein represents the solid phase; represents time; represents the interface recession velocity, , represents the solid molar volume of the phase, represents the molar rate of oxidation at the interface, , is the oxidation reaction rate, represents the local normal vector at the interface, , represents the gradient operator, represents the gradient with respect to the solid phase, represents the modulus of ; represents the molar concentration of oxygen; is the diffusion coefficient.

[0066] In this embodiment, the dimensionless parameter Sh is calculated as follows:

[0067]

[0068] wherein is the fiber radius of the C / C composite material; is the oxidation reaction rate of the matrix of the C / C composite material; is the diffusion coefficient.

[0069] In the present embodiment, the calculation expression of the dimensionless parameter A is:

[0070]

[0071] wherein is the oxidation reaction rate of the matrix of the C / C composite material; is the molar volume of the matrix phase; is the oxidation reaction rate of the carbon fiber of the C / C composite material; is the molar volume of the fiber phase.

[0072] The construction method of the pre-trained multi-layer perception network in step S3 includes the following steps:

[0073] S3-1, set the number of hidden layers of the multi-layer perception network and the number of neurons in each hidden layer, and use an activation function for nonlinear transformation in the hidden layer; wherein the number of hidden layers and the number of neurons are determined by the open source framework Optuna; common activation functions include ReLU function, Tanh function, Sigmoid function, etc. The ReLU function is selected in the present embodiment, which has the advantages of simple calculation and fast convergence speed, and can effectively introduce nonlinear factors, so that the multi-layer perception network can learn more complex function relationships; therefore, selecting the ReLU function can enhance the multi-layer perception network's ability to capture various nonlinear physical phenomena in the gas-solid coupling process;

[0074] S3-2, construct a gas phase diffusion loss function, a solid phase evolution loss function and an initial condition loss function, and then obtain a total loss function of the multi-layer perception network;

[0075] S3-3, take points in the calculation domain by taking the three-dimensional space model of the C / C composite material with known ablation morphology and its initial conditions as the calculation domain; wherein the taking points are uniform sampling, random sampling or adaptive sampling; the taken points include oxygen molar concentration and solid volume fraction;

[0076] S3-4, input the coordinates of the taken points, the ablation time, the dimensionless parameter Sh and the dimensionless parameter A into the multi-layer perception network as training samples, and obtain the oxygen molar concentration and the solid volume fraction predicted by the multi-layer perception network;

[0077] S3-5, taking the molar concentration of oxygen and the solid volume fraction contained by the taken point as the true label, combining the molar concentration of oxygen and the solid volume fraction obtained by the multi-layer perception network, and calculating a total loss value through a total loss function of the multi-layer perception network;

[0078] S3-6, adjusting parameters of the multi-layer perception network until a preset training end condition is reached, so as to obtain a pre-trained multi-layer perception network.

[0079] In the embodiment, x, y and z coordinates in the Cartesian coordinate system are taken as part of the input layer, and the coordinate values can locate each point in the calculation region inside and around the C / C composite material, thereby providing a basis for subsequent calculation of physical quantities at the point. By introducing a time variable, the multi-layer perception network can capture the evolution law of the morphology over time in the ablation process.

[0080] The expression of the total loss function of the multi-layer perception network is:

[0081]

[0082] wherein is the total loss function; , and are weight coefficients; is a gas phase diffusion loss function, ; is a solid phase evolution loss function, ; is an initial condition loss function, ; represents the molar concentration of oxygen; represents time; represents a gradient operator; is the solid volume fraction, and when represents a gas phase unit, and when represents a solid phase unit; is a diffusion coefficient; is the molar concentration of the reaction gas; is the square of the L2 norm; is the speed of solid retreat; is the boundary layer outer edge concentration.

[0083] As shown in Figure 2 and Figure 3 , the specific method for constructing the three-dimensional space model of the C / C composite material in step S3 is:

[0084] A three-dimensional space model of C / C composite material is established, and the coordinate data of the three-dimensional space model of C / C composite material in the Cartesian coordinate system is obtained, and the cross section of the three-dimensional space model of C / C composite material is explained.

[0085] In the micro-ablation model, the gas-solid interface will ablate and retreat under the action of oxidation, and the gas phase diffusion and interface retreat are coupled with each other. The retreat rate of the interface is related to the gas concentration at the interface, and the interface retreat will change the gas concentration distribution; the solid phase will be converted into the gas phase during the oxidation process, and the geometry of the solid phase region will undergo complex deformation such as fusion and disappearance. Based on this, the gas phase diffusion equation in this embodiment is:

[0086]

[0087] Wherein, in the fluid domain and the solid domain , at the interface ; is the solid volume fraction; C is the molar concentration of the reaction gas; is the heterogeneous reaction rate; represents the reciprocal of the gradient module length.

[0088] The solid phase evolution equation is:

[0089]

[0090] Wherein is the solid volume fraction; is the solid retreat speed; in the fluid domain and the solid domain , at the interface ; is the solid volume fraction; the molar concentration of the reaction gas; is the heterogeneous reaction rate; represents the reciprocal of the gradient module length.

[0091] In computational fluid dynamics, the gas / liquid two-phase flow problem and the gas / solid problem in this paper have similarities, so the simulation method of gas / liquid two-phase flow can be used to solve the gas phase diffusion equation and the solid phase evolution equation. The fluid volume fraction method is commonly used to solve the gas / liquid two-phase flow problem, and similarly, the solid volume fraction method is introduced in this method, and the evolution of the fluid domain and the solid domain has a similar form of control equation , wherein is the solid phase volume contained in the unit volume V. Specifically, the oxygen molar concentration and the solid volume fraction output by the multi-layer perception network are used to solve the gas phase diffusion equation and the solid phase evolution equation, and the specific method for obtaining the ablation morphology of the C / C composite material is:

[0092] The oxygen molar concentration and the solid volume fraction output by the multi-layer perception network are used to track the evolution of the oxygen molar concentration and the solid volume fraction distribution of the material at different times through the gas phase diffusion equation and the solid phase evolution equation, and then the ablation morphology of the three-dimensional C / C composite material is obtained.

[0093] In an embodiment of the present application, when solving the partial differential equation using the physical information neural network (pre-trained multi-layer perception network), the selection of points in the calculation domain, the boundary and the initial condition is very important. The selection of points in this study only involves the calculation domain and the initial condition. First, the selection of points in the calculation domain of the three-dimensional space model of C / C composite material is used to describe the change of physical field in the whole region. Commonly used calculation domain points include uniform sampling, random sampling and adaptive sampling.

[0094] From the calculation domain diagram, it can be seen that the gas is uniformly distributed on the surface of the solid phase region, so uniform sampling is adopted, that is, uniform sampling is performed in the calculation domain according to a certain grid spacing, which can ensure that the model has relatively uniform learning and fitting in each region of the calculation domain, avoiding overfitting in some regions and insufficient fitting in other regions. For two-dimensional solution of coupled Stokes-Darcy equation, the data points in each region are represented by different colors and symbols, as shown in Figure 4 .

[0095] The gas phase diffusion equation and the solid phase evolution equation describe the change of the physical system in the whole calculation domain. By selecting points in the calculation domain, the PINN can obtain the physical information at these points, such as the change of the gas-solid interface movement. The model uses this information to learn the physical process represented by the partial differential equation, thereby establishing a mapping relationship between the input (spatial and temporal coordinates, etc.) and the output (physical quantity) to fit the physical phenomenon in the whole calculation domain. Then, for the problem of the ablation morphology of C / C composite material evolving with time, the distribution of physical quantities at the initial time is given as the starting condition of the problem. For the initial condition, sampling can be performed on the time slice at the initial time, and the sampling point method is similar to the calculation domain point. For example, for one-dimensional heat transfer solving problem, the point selection principle diagram of PINN in the calculation domain is shown in Figure 5 .

[0096] The initial condition taking points provide specific state information of the physical system at the initial time for the PINN. By giving accurate initial conditions, the PINN can start modeling and predicting the dynamic process of the physical system from the correct starting point, ensuring that the simulation results conform to the development law of the actual physical phenomenon. During the learning process, the network needs to adjust the network parameters so that the model can accurately reproduce the physical quantity values at these taking points at the initial time. This helps guide the model to train in the correct direction, avoids deviations in the training results that do not conform to the actual physical process, and thus improves the accuracy and reliability of the model. Therefore, the three-dimensional configuration points and cross-section configuration points of the three-dimensional model in the embodiment are as shown in Figure 6 .

[0097] In the specific implementation process, the pre-trained multi-layer perception network is applied to new test points or regions to calculate the gas-solid physical quantities and ablation morphology parameters of the C / C composite material under the action of gas-solid coupling. The calculation results are compared with the results of traditional numerical methods (such as the calculation results of the finite volume coupling method) to evaluate the accuracy and reliability of the solution. The results are visually analyzed, such as drawing the ablation rate distribution cloud map of the C / C composite material surface and the three-dimensional visualization image of the ablation morphology, so as to intuitively understand the ablation process and morphology evolution of the C / C composite material under the action of gas-solid coupling.

[0098] In the embodiment, as shown in Figure 7 , Figure 8 and Figure 9 , the model is trained to obtain the loss function, the gas-solid interface change at different times, and the ablation morphology of a single fiber under different Da numbers.

[0099] In summary, the present application can efficiently and accurately solve the gas-solid coupling equation, overcome the limitations of traditional numerical methods, and accurately obtain the ablation morphology of the C / C composite material under the action of gas-solid coupling with low computational cost, thereby providing strong theoretical support and technical guarantee for the application of C / C composite materials in the field of aerospace and other fields.

Claims

1. A method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network, characterized in that, Includes the following steps: Construct gas-solid coupling equations for C / C composite materials in a microscopic ablation model; The dimensionless parameters Sh and A of the microscopic ablation behavior of C / C composite materials were obtained. Construct a pre-trained multilayer perceptron network; construct a three-dimensional spatial model of C / C composite materials; The coordinate data of the three-dimensional spatial model of C / C composite material in the Cartesian coordinate system, ablation time, dimensionless parameter Sh and dimensionless parameter A are used as inputs to a pre-trained multilayer perceptron network. The pre-trained multilayer perceptron network outputs oxygen molar concentration and solid volume fraction. Based on the gas-solid coupling equation, the gas phase diffusion equation and solid phase evolution equation of C / C composite material in the microscopic ablation model are obtained; the initial conditions and boundary conditions of the gas phase diffusion equation and solid phase evolution equation are set. The gas phase diffusion equation and solid phase evolution equation were solved by using the oxygen molar concentration and solid volume fraction output by the multilayer perceptron network to obtain the ablation morphology of the C / C composite material. The expression for calculating the dimensionless parameter A is: in The matrix oxidation reaction rate in the C / C composite material; The molar volume of the matrix phase; The oxidation reaction rate of carbon fibers in C / C composite materials; This represents the molar volume of the fibrous phase. The expression for calculating the dimensionless parameter Sh is: in The fiber radius of the C / C composite material; The matrix oxidation reaction rate of the C / C composite material; The diffusion coefficient is denoted as . The method for constructing a pre-trained multilayer perceptron network includes the following steps: Set the number of hidden layers and the number of neurons in each hidden layer of the multilayer perceptron network, and use activation functions to perform nonlinear transformations in the hidden layers; The number of hidden layers and neurons were determined using the open-source framework Optuna. The gas-phase diffusion loss function, solid-phase evolution loss function, and initial condition loss function are constructed, and then the total loss function of the multilayer perceptron network is obtained. The three-dimensional spatial model of the C / C composite material with known ablation morphology and its initial conditions are used as the computational domain, and points are selected within the computational domain. The point selection method is uniform sampling, random sampling, or adaptive sampling. The selected points include oxygen molar concentration and solid volume fraction. The coordinates of the selected points, ablation time, dimensionless parameter Sh, and dimensionless parameter A are used as training samples and input into the multilayer perceptron network to obtain the oxygen molar concentration and solid volume fraction predicted by the multilayer perceptron network. The oxygen molar concentration and solid volume fraction of the selected points are used as the true labels. Combined with the oxygen molar concentration and solid volume fraction predicted by the multilayer perceptron network, the total loss value is calculated through the total loss function of the multilayer perceptron network. The parameters of the multilayer perceptron network are adjusted with the goal of minimizing the total loss value until the preset training termination condition is met, thus obtaining the pre-trained multilayer perceptron network.

2. The method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network according to claim 1, characterized in that, The expression for the gas-solid coupling equation is: in Indicates the solid phase; Indicates time; Indicates the speed at which the screen scrolls backward. , Represents the solid molar volume of a phase. This indicates the molar rate of oxidation at the interface. , The oxidation reaction rate, This represents the local normal vector at the interface. , Represents the gradient operator. This indicates the gradient with respect to the solid phase. express The model; Indicates the molar concentration of oxygen; is the diffusion coefficient.

3. The method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network according to claim 1, characterized in that, The expression for the total loss function of a multilayer perceptron network is: in This is the total loss function; , and These are the weighting coefficients; Let be the gas phase diffusion loss function. ; Let be the solid-state evolution loss function. ; The initial conditional loss function, ; Indicates the molar concentration of oxygen; Indicates time; Represents the ladder operator; For solid volume fraction, when When represents the gas phase unit, when Time represents a solid phase unit; The diffusion coefficient is denoted as . This represents the molar concentration of the reacting gas; The square of the L2 norm; The velocity of the solid as it recedes; This represents the concentration at the outer edge of the boundary layer.

4. The method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network according to claim 1, characterized in that, The specific method for constructing a three-dimensional spatial model of C / C composite materials is as follows: A three-dimensional spatial model of C / C composite material with a three-dimensional matrix encapsulating fiber monofilaments is established. The coordinate data of the three-dimensional spatial model of C / C composite material in the Cartesian coordinate system is obtained, and the cross-sectional description of the three-dimensional spatial model of C / C composite material is given.

5. The method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network according to claim 2, characterized in that, The gas phase diffusion equation is: Among them, in the fluid domain and the solid domain At the interface ; is the solid volume fraction; C is the molar concentration of the reactant gas; This represents a heterogeneous reaction rate. It represents the reciprocal of the gradient magnitude.

6. The method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network according to claim 2, characterized in that, The solid phase evolution equation is: in It represents the volume fraction of solids. The velocity of the solid's retreat; in the fluid and solid domains At the interface ; Solid volume fraction; molar concentration of the reactant gas; This represents a heterogeneous reaction rate. It represents the reciprocal of the gradient magnitude.

7. The method for obtaining the ablation morphology of C / C composite materials based on a physical information neural network according to claim 1, characterized in that, The specific method for obtaining the ablation morphology of C / C composite materials by solving the gas phase diffusion equation and the solid phase evolution equation using the oxygen molar concentration and solid volume fraction output by the multilayer perceptron network is as follows: By using the oxygen molar concentration and solid volume fraction output by the multilayer perceptron network, the evolution of the oxygen molar concentration and solid volume fraction distribution of the material at different times is tracked through the gas phase diffusion equation and the solid phase evolution equation, thereby obtaining the development process of the ablation morphology of the three-dimensional C / C composite material over time.

Images