Method, device, equipment and medium for predicting creep performance of aluminum composite material

Abstract

This invention discloses a method, apparatus, equipment, and medium for predicting the creep properties of aluminum composite materials, belonging to the field of performance prediction. The method comprises: acquiring experimental data of the aluminum composite material; determining matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters based on the experimental data, wherein the interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix; constructing a finite element model using the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters; and simulating and solving the finite element model based on the working conditions corresponding to the creep experiment to obtain the creep performance prediction parameters. This invention can construct a simulation model of the material and make predictions based on the interface cracking parameters determined by the contact state between the fiber and the matrix, which can closely reflect the relationship between the fiber and the matrix at high temperatures, thereby reducing prediction deviations and improving prediction accuracy.

Description

Technical Field

[0001] This invention relates to the technical field of performance prediction, and in particular to a method, apparatus, equipment and medium for predicting the creep performance of aluminum composite materials. Background Technology

[0002] Creep is the phenomenon where the strain of a material gradually increases over time under constant temperature and load. For aluminum-based composite materials, high-temperature creep is a significant factor affecting their long-term service reliability. Creep performance refers to the ability of a material to slowly deform over time under constant stress, reflecting its stability and durability under long-term load.

[0003] Therefore, creep performance prediction is necessary for different materials before they are put into application. One commonly used method for creep performance prediction is the finite element method (FEM). This method establishes a representative volume element model (RVE) of the composite material and simulates the matrix creep, fiber reinforcement constraint, and interface load transfer process in finite element software to obtain the predicted parameters of creep performance. This method allows for intuitive analysis of microscopic responses such as matrix creep strain distribution, fiber axial stress, interface slip, and damage evolution.

[0004] However, the above method has the following technical problems: When predicting creep performance, traditional finite element models need to set the connection between the fiber and the matrix as completely bonded. However, at high temperatures, the fiber and matrix may experience tangential slippage, interface cracking, interface stiffness degradation, and interface damage evolution. This prediction method leads to the model overestimating the fiber reinforcement effect, resulting in an underestimation of the creep rate, thus causing the prediction results to deviate from reality and resulting in low prediction accuracy. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for predicting the creep properties of aluminum composite materials, which can solve one or more technical problems existing in the prior art.

[0006] A first aspect of this invention provides a method for predicting the creep properties of aluminum composite materials, the method comprising: Obtain experimental data of aluminum composite materials, wherein the experimental data is data collected during creep experiments on aluminum composite materials within a preset temperature range; Based on the experimental data, the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters were determined respectively. Among them, the interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix. A finite element model is constructed using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters. The finite element model is simulated and solved based on the working conditions corresponding to the creep experiment to obtain the creep performance prediction parameters.

[0007] In conjunction with the first aspect, in one implementation, the operation of determining the matrix creep constitutive parameters includes: Creep test curves are extracted from the experimental data, and the stable creep stage of the aluminum composite material is determined based on the slope of the creep test curves. The stable characteristic parameters corresponding to the stable creep stage are extracted from the creep test curve, and the matrix constitutive model is constructed using the stable characteristic parameters; The parameters are solved based on the matrix constitutive model, and the solved parameters are verified based on the simulation curves of the creep experiment curves to obtain the matrix creep constitutive parameters.

[0008] In conjunction with the first aspect, in one implementation, the operation of determining the fiber creep constitutive parameters includes: The elastic modulus and Poisson's ratio are extracted from the experimental data, and the Lamé constant is calculated based on the elastic modulus and Poisson's ratio, respectively. The fiber elastic stiffness tensor is determined using the Lamé constant, and the elastic constitutive relation of the alumina fiber is determined based on the fiber elastic stiffness tensor to obtain the fiber creep constitutive parameters.

[0009] In conjunction with the first aspect, in one implementation, the operation of determining the geometric model parameters includes: Fiber characteristic parameters are extracted from the experimental data, and fiber characteristic parameters are used to construct fiber-penetrating representative volume units; The engineering equivalent matrix model and the polycrystalline matrix model are constructed using the fiber-penetrating representative volume unit. The engineering equivalent matrix model is a model in which the aluminum alloy matrix is ​​a uniform and continuous medium. The polycrystalline matrix model is a model in which grain seed points are generated in the aluminum alloy matrix region after deducting the fiber volume, and the polycrystalline matrix model is formed by three-dimensional partitioning. The model parameters of the engineering equivalent matrix model and the polycrystalline matrix model are extracted respectively to obtain the geometric model parameters.

[0010] In conjunction with the first aspect, in one implementation, the construction of a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters includes: Based on the Boolean topology operation algorithm, a volume element model of the fiber and matrix is ​​constructed according to the geometric model parameters; The interface damage constitutive parameters, the matrix creep constitutive parameters, and the fiber creep constitutive parameters are bound to the volume element model to obtain the finite element model.

[0011] In conjunction with the first aspect, in one implementation, the step of simulating and solving the finite element model based on the operating conditions corresponding to the creep experiment to obtain creep performance prediction parameters includes: The periodic boundary conditions of the finite element model are determined by determining the working conditions corresponding to the creep test, wherein the periodic boundary conditions include: the displacement distance of corresponding nodes on the relative boundary; The finite element model is simulated using the periodic boundary conditions to record the variation parameters of the finite element model under the periodic boundary conditions; The creep performance prediction parameters are calculated using the aforementioned changing parameters, which include: creep strain-time curve, steady-state creep rate, matrix equivalent creep strain contour map, fiber axial stress contour map, interface slip, interface damage variable, grain boundary damage variable, and creep performance variation law.

[0012] In conjunction with the first aspect, in one implementation, after the step of simulating and solving the finite element model based on the working conditions corresponding to the creep experiment to obtain the creep performance prediction parameters, the method further includes: The creep strain-time curve of the predicted creep performance parameters and the curve deviation parameter of the creep test curve of the experimental data are determined. The correction process is performed based on the curve deviation parameters, wherein the correction process includes correction of the aluminum alloy matrix creep parameters, correction of the fiber-matrix interface parameters, and correction of the fiber characteristic parameters.

[0013] A second aspect of the present invention provides a creep property prediction device for aluminum composite materials, the device comprising: The acquisition module is used to acquire experimental data of aluminum composite materials, wherein the experimental data is data collected during creep experiments on aluminum composite materials based on a preset temperature range; The determination module is used to determine the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters based on the experimental data, wherein the interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix. The construction module is used to construct a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters; The prediction module is used to simulate and solve the finite element model based on the working conditions corresponding to the creep experiment to obtain creep performance prediction parameters.

[0014] Compared to existing technologies, the creep performance prediction method, apparatus, equipment, and medium for aluminum composite materials provided in this invention offer the following advantages: This invention can acquire experimental data of aluminum composite materials; determine matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters based on the experimental data, wherein the interface damage constitutive parameters are interface cracking parameters determined based on the fiber-matrix contact state; construct a finite element model using the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters; and simulate and solve the finite element model based on the working conditions corresponding to the creep experiment to obtain creep performance prediction parameters. This invention can determine the interface cracking parameters determined by the fiber-matrix contact state, construct a simulation model of the material based on the interface cracking parameters, and perform predictions. This approach closely reflects the relationship between the fiber and the matrix at high temperatures, thereby avoiding overestimation of the fiber reinforcement effect, reducing prediction deviations, and improving prediction accuracy. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a method for predicting the creep properties of aluminum composite materials according to an embodiment of the present invention. Figure 2 This is a flowchart of the operation for determining matrix creep constitutive parameters according to an embodiment of the present invention; Figure 3 This is a flowchart of the operation of constructing a finite element model according to an embodiment of the present invention; Figure 4 This is a schematic diagram of a creep performance prediction device for aluminum composite materials provided in an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] To address the aforementioned issues, the following specific embodiments will provide a detailed description and explanation of the creep performance prediction method, apparatus, equipment, and medium for aluminum composite materials provided in this application.

[0018] To address the technical problems of biased and low-accuracy predictions in existing technologies, referencing Figure 1 The diagram shows a flowchart of a method for predicting the creep properties of aluminum composite materials according to an embodiment of the present invention.

[0019] As an example, the method for predicting the creep properties of the aluminum composite material may include: S11. Obtain experimental data of aluminum composite material, wherein the experimental data is data collected during creep experiments on aluminum composite material based on a preset temperature range.

[0020] In one embodiment, the aluminum composite material to be predicted can be determined first, which may be a material of an alumina fiber reinforced aluminum matrix composite system.

[0021] The aluminum alloy matrix can be Al-Si, Al-Cu, Al-Mg, Al-Si-Cu-Mg, or heat-resistant aluminum alloy. The alumina fibers can be chopped alumina fibers, continuous alumina fibers, or alumina fiber preforms. The composite material can be obtained through extrusion casting, liquid infiltration, powder metallurgy, hot pressing sintering, or other metal matrix composite material preparation processes.

[0022] The alumina fiber in this invention is a ceramic fiber with Al2O3 as its main component. Aluminum-based composite materials are composite materials formed by adding ceramic particles, whiskers, short fibers, or continuous fibers as reinforcing phases, with aluminum or aluminum alloys as the matrix.

[0023] The service conditions of the material are then determined. For creep prediction of alumina fiber-reinforced aluminum matrix composites within a preset temperature range, the preferred temperature analysis range is 250℃ to 450℃. If the material's service scenario involves lower thermal exposure, the range can be extended to 150℃ to 250℃; if extreme condition analysis is required, the range can be extended to close to 500℃. However, in this case, it is not advisable to use only the simple creep constitutive model; the risks of matrix softening, reinforcing phase dissolution, interfacial reactions, and material instability should also be considered.

[0024] The applied stress level should be determined based on the high-temperature yield strength of the composite material at the corresponding temperature and the actual service load. For short fiber reinforced aluminum matrix composites, the uniaxial creep test stress is preferably 20 MPa to 100 MPa. For continuous fiber reinforced systems or low-temperature high-strength systems, the stress can be appropriately extended to 120 MPa, but it should be ensured that the specimen enters the stable creep stage within the set time, rather than experiencing significant plastic instability or fracture in a short period of time.

[0025] The prediction objectives of this invention include: creep strain-time curve, steady-state creep rate, creep life, fiber axial stress distribution, matrix equivalent creep strain distribution, fiber-matrix interface slip, interface damage variables, and the influence of different microscopic parameters on the creep properties of composite materials.

[0026] Next, high-temperature creep experiments can be carried out on aluminum alloy matrix materials and alumina fiber reinforced aluminum matrix composite materials respectively based on preset temperature ranges to obtain experimental data and determine model input parameters.

[0027] Specifically, for the aluminum alloy matrix, uniaxial constant stress creep experiments were conducted at multiple temperatures and stress levels to obtain the creep strain-time curves of the matrix material. Characteristic parameters of the initial creep stage, steady-state creep stage, and accelerated creep stage were extracted from the experimental curves, especially the steady-state creep rate.

[0028] For alumina fiber reinforced aluminum matrix composites, creep experiments were conducted under the same temperature and stress conditions to obtain the overall creep curve of the composite material, which was used for subsequent model verification and interface parameter inversion.

[0029] Simultaneously, the microstructure of the composite material was characterized. Fiber length distribution, diameter distribution, orientation distribution, volume fraction, porosity, and fiber agglomeration were obtained using scanning electron microscopy, metallographic microscopy, X-ray micro-CT, or image analysis methods, and the dimensional relationship between fiber length and aluminum alloy matrix grain size was clarified. If necessary, electron backscatter diffraction was used to obtain information on the aluminum alloy matrix grain size and orientation.

[0030] This invention models alumina fibers as a linear elastic reinforcing phase and determines the fiber length, diameter, aspect ratio, volume fraction, and orientation distribution based on the statistical results of microstructure, thus avoiding the problem of inconsistency between the regular fiber arrangement model and the actual material structure.

[0031] Through the above experiments and characterization, the original data and initial input parameters used for subsequent constitutive calibration, geometric modeling and model verification were obtained, and the experimental data were obtained.

[0032] Specifically, experimental data may include the following: (1) Elastic modulus, Poisson's ratio, coefficient of thermal expansion, high-temperature yield strength of the base aluminum alloy, and measured creep strain-time curves and steady-state creep rates at various temperatures and stress levels; it should be noted that the creep constant required by the Norton equation is... Stress index Creep activation energy and hardening index Instead of being a direct measurement in this step, it is treated as a state variable to be determined, and will be determined by mathematical fitting and inverse solution based on the measured curves obtained in this step in subsequent steps. (2) Elastic modulus of alumina fiber Poisson's ratio Coefficient of thermal expansion and fiber length ,diameter Aspect ratio, volume fraction and statistical parameters such as orientation distribution; (3) Measurable parameters or inversion initial values ​​such as the equivalent elastic modulus, equivalent thickness, interface strength, interface fracture energy and interface friction coefficient of the interface layer; interface traction force, separation displacement and damage variables are state variables calculated from the interface constitutive relation during the finite element solution process; (4) Macroscopic creep curves and steady-state creep rates of composite materials under corresponding temperature and stress conditions are used for subsequent model verification and interface parameter inversion.

[0033] Steady-state creep rate: This is the strain growth rate in the approximately linear stage of the creep curve, and it is commonly used to evaluate the creep resistance of materials. Generally, the lower the steady-state creep rate, the stronger the material's creep resistance.

[0034] The aforementioned material constants, structural statistical parameters, and interface parameters are preferably obtained from creep experiments. If measured data for the corresponding material are unavailable, initial values ​​can be set by referring to literature on similar materials or material handbooks, and then corrected by inversion based on experimental curves in subsequent steps.

[0035] S12. Determine the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters based on the experimental data, wherein the interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix.

[0036] Since the experimental data includes multiple different data and parameters, four parameters can be determined based on the experimental data, including matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters.

[0037] Wherein, the matrix creep constitutive parameters are the constitutive parameters of matrix creep in the steady-state creep stage of aluminum composite materials; The fiber creep constitutive parameters are the constitutive parameters of the fibers in the aluminum composite material during the creep process. The interface damage constitutive parameters are determined based on the contact state between the fibers and the matrix of the aluminum composite material, and are obtained by cracking parameters at the fiber-matrix contact interface. The geometric model parameters are the relevant parameters of the fiber and matrix geometry in the aluminum composite material.

[0038] In an optional embodiment, determining the matrix creep constitutive parameters may include the following sub-steps: S21. Extract the creep test curve from the experimental data, and determine the stable creep stage of the aluminum composite material based on the slope of the creep test curve.

[0039] S22. Extract the stable characteristic parameters corresponding to the stable creep stage from the creep experiment curve, and use the stable characteristic parameters to construct a matrix constitutive model.

[0040] S23. Solve the parameters based on the matrix constitutive model, and verify the solved parameters based on the simulation curve of the creep experiment curve to obtain the matrix creep constitutive parameters.

[0041] Reference Figure 2 The diagram illustrates an operation flowchart for determining matrix creep constitutive parameters according to an embodiment of the present invention.

[0042] The present invention preferably employs a step-by-step calibration method. First, at the same temperature, the stress exponent is fitted according to the steady-state creep rate under different stress levels. Secondly, under the same stress level, the creep activation energy is fitted based on the steady-state creep rate at different temperatures. Finally, the least squares method was used to correct all creep curves as a whole to obtain the final matrix creep constitutive parameters.

[0043] Specifically, Figure 2 The material information, creep test scheme, matrix creep curve, and steady-state creep rate are provided in step S11; subsequently, feature parameters are extracted and fitted sequentially. and ,Sure and hardening parameters The data is written into finite element analysis software (such as Abaqus) and compared with the experimental curves of the matrix. If the error exceeds the limit, the data is returned to the parameter fitting stage for correction, and finally the matrix creep constitutive parameters are output.

[0044] Abaqus is a finite element analysis software used for structural mechanics, thermodynamics, contact, damage, creep, and multiphysics coupling analysis. This invention uses Abaqus to establish a mesoscopic finite element model of alumina fiber-reinforced aluminum matrix composites and to perform creep solutions.

[0045] In one embodiment, the mechanism for determining the creep stage is based on the slope of the macroscopic experimental curve, i.e., the creep rate. Divided by dynamic evolution characteristics over time: when creep rate When the creep rate gradually decreases over time, the material is considered to be in the initial creep stage; when the creep rate... When the creep rate remains approximately constant over a continuous time interval and its rate of change is below a preset threshold, the material is considered to have entered the steady-state creep stage; when the creep rate... When the creep continues to increase over time, accompanied by the expansion of microscopic damage within the material, the material is determined to be in the accelerated creep stage.

[0046] The specific judgment logic and technical means for selecting the appropriate creep equation for different creep stages and applied load conditions are as follows: (1) When the simulation prediction target only focuses on the steady-state creep stage of the material under constant load, the Norton type creep equation is selected for calculation; (2) When the simulation prediction target needs to describe the complete deformation behavior of the initial creep stage and the steady-state creep stage at the same time, and the load remains constant, the time-hardening Norton-Bailey creep equation is selected for calculation. Norton-Bailey creep model: is a creep model that introduces time or strain hardening terms on the basis of Norton model. It can be used to describe the initial creep stage and unsteady creep behavior. Norton creep model: is a commonly used power-law creep model used to describe the relationship between steady-state creep rate and stress. It is usually expressed as εcr=Aσn. (3) When the set service conditions involve variable stress loading or multi-stage step loading process, select strain hardening creep equation to ensure the continuity of historical strain accumulation; (4) When the simulation analysis needs to cover the accelerated creep stage until material failure, damage variables or tissue softening variables are further coupled on the basis of the above corresponding equations, and the solution is performed jointly by the user-defined material subroutine of the finite element software (such as CREEP or UMAT).

[0047] Among them, the UMAT subroutine: UMAT is an abbreviation for User Material, which can be used to define complex material constitutive relations in finite element analysis software.

[0048] CREEP subroutine: CREEP is a user subroutine in finite element analysis software used to define user-defined creep constitutive relations. The CREEP subroutine allows for the creation of temperature-dependent, damage-dependent, or piecewise creep models that are difficult to express using built-in material models.

[0049] This invention establishes a temperature-dependent creep constitutive model for aluminum alloy matrices, which can employ Norton-type, Norton-Bailey-type, time-hardening, strain-hardening, or hyperbolic sinusoidal creep equations, and can be implemented using the built-in Creep model, CREEP subroutine, or UMAT subroutine in finite element analysis software.

[0050] The creep behavior of the aluminum alloy matrix under high-temperature stress is described using a temperature-dependent power-law creep model, preferably a Norton-type or Norton-Bailey-type creep equation. The purpose of constructing this creep equation is to convert the matrix creep experimental curves obtained in step S11 into local material constitutive relations that can be called by the finite element analysis software. This allows the solver to calculate the creep strain rate and increment at each matrix volume integral point based on the current equivalent stress, temperature, creep time, or cumulative creep strain. This constitutive relation can be assigned to the aluminum matrix region in subsequent steps, participating in stress updates and iterative solutions, and forming the macroscopic creep response of the composite material through volume averaging.

[0051] The creep stage is determined by the change in the slope of the experimental curve over time: when the creep rate... When the creep gradually decreases over time, it is considered to be in the initial creep stage; when When the temperature is approximately constant over a continuous time interval and the rate of change is below a preset threshold, it is determined to be in the steady-state creep stage; when... When the damage continues to increase and is accompanied by damage propagation, it is determined to be the accelerated creep stage. Norton-type equations are used when only the steady-state stage is predicted and the load is constant; time-hardening Norton-Bailey equations are used when the initial stage needs to be described simultaneously; strain-hardening equations are used when there is variable stress or graded loading; when the accelerated stage needs to be described, damage variables and tissue softening variables are coupled to the above equations, or the CREEP / UMAT user subroutine is used.

[0052] The steady-state creep stage can take the following form: ; In the formula, This is the equivalent creep strain rate. The creep constant is For equivalent stress, Stress index For creep activation energy, The gas constant is This refers to absolute temperature.

[0053] When it is necessary to describe the initial creep stage, the time-hardening form can be used: ; In the formula, Creep time, This is the time hardening index.

[0054] When a material undergoes alternating stress loading or multi-stage loading, strain hardening is preferred. ; In the formula, To accumulate creep strain, The strain hardening index is denoted as .

[0055] In finite element analysis software, if a standard creep model is used, time hardening, strain hardening, or hyperbolic-sine creep laws can be set through the Creep material model; if it is necessary to introduce temperature dependence, damage variables, tissue softening variables, or interface coupling variables, this can be achieved through the CREEP or UMAT user subroutines.

[0056] The purpose of constructing the creep equation is to transform the macroscopic high-temperature creep experimental data obtained in step S11 into local material constitutive relations that can be directly identified and called by the finite element solver.

[0057] In terms of the logical connections between steps, this step constitutes a crucial data loop for the transformation from macroscopic experimental data to microscopic finite element calculations. Specifically: using the high-temperature creep experimental curve output from step S11 as input, constitutive parameters (creep constant A, stress exponent n, creep activation energy Q, etc.) are obtained through mathematical analysis and nonlinear fitting, thereby establishing the creep equation; simultaneously, the creep equation output in this step is directly called in subsequent modeling and solving, that is, in subsequent steps, this creep equation is forcibly bound as a material property to the constructed three-dimensional geometric domain of the aluminum alloy matrix; and in the finite element incremental iterative solving stage, the solver will call the creep equation in real time at each integration point of the matrix mesh, and calculate the local creep strain increment and stress update based on the equivalent stress and time state of the current incremental step, ultimately forming the overall macroscopic creep response of the composite material.

[0058] In one embodiment, determining the fiber creep constitutive parameters may include the following sub-steps: S31. Extract the elastic modulus and Poisson's ratio from the experimental data, and calculate the Lamé constant based on the elastic modulus and Poisson's ratio.

[0059] S32. The fiber elastic stiffness tensor is determined using the Lamé constant, and the elastic constitutive relation of the alumina fiber is determined based on the fiber elastic stiffness tensor to obtain the fiber creep constitutive parameters.

[0060] In aluminum matrix composites, alumina fibers primarily serve to transfer loads and constrain deformation. Since the creep deformation of alumina fibers within the temperature range of 250℃ to 450℃ is typically much smaller than that of the aluminum alloy matrix, a linear elastic model is preferably used to describe the alumina fibers in this invention.

[0061] The elastic constitutive relation of alumina fiber is: ; In the formula, For fiber stress tensor, For fiber strain tensor, This is the fiber elastic stiffness tensor.

[0062] It should be noted that the fiber stress tensor in the above formula... and fiber strain tensor It is the state parameter of the finite element model during the incremental iterative solution process.

[0063] Specifically, the construction process of the mechanical model of alumina fiber-reinforced phase is as follows: extract the macroscopic elastic modulus of alumina fibers from experimental data. Compared to Poisson Identical material constants; input based on experimental data and Calculate Lamé constant and This leads to the construction of the fiber elastic stiffness tensor. In the subsequent finite element solution process, the fiber strain tensor The fiber stress tensor will be calculated by the solver based on the fiber node displacement field. The solver then calls the stiffness tensor established in this step. The mapping was calculated in real time.

[0064] For isotropic alumina fibers, the elastic modulus can be used. Compared to Poisson Description. Step S2 only requires providing... , Equal material constants, no need to obtain the fiber stress tensor in advance. Fiber strain tensor and elastic stiffness tensor .according to and Calculate Lamé constant and And construct In the finite element method solution process, Calculated from the displacement field of the fiber nodes, The values ​​are obtained in real time by the solver. The preferred elastic modulus of alumina fiber is 250 GPa to 380 GPa, and the preferred Poisson's ratio is 0.20 to 0.27. The specific values ​​are determined based on the fiber manufacturer's data, material handbook, or independent tensile tests.

[0065] ; ; The alumina fiber-reinforced phase mechanical model is used to characterize the high stiffness load-bearing capacity of the fibers and their constraint on the creep deformation of the matrix. Its material constants are derived from experimental data, and its geometric dimensions and spatial orientation are determined by geometric model parameters. In subsequent step 13, the fibers can be assigned solid domains, and in subsequent step 14, they participate in the equilibrium equations and load transfer calculations together with the matrix and interface models. In step S14, the microscopic responses such as fiber axial stress are output. Therefore, this model is an essential component for a complete RVE solution. (RVE: an abbreviation for Representative Volume Element. It is a local microscopic model that can represent the overall statistical characteristics of a material. In this invention, RVE is used to characterize the spatial distribution, periodic extension relationship, and, when necessary, the grain structure of the aluminum alloy matrix of alumina fibers in the aluminum matrix.) It should be noted that the fiber stress tensor in the above formula... and fiber strain tensor These are the state parameters in the incremental iterative solution process of the finite element model. The purpose of constructing the mechanical model of the alumina fiber-reinforced phase is to quantitatively characterize the high stiffness axial load-bearing capacity of alumina fibers at the microscale, as well as their strong motion constraint effect on the high-temperature creep deformation of the aluminum alloy matrix, thereby providing a mechanical constitutive description of the reinforcing phase for the representative volume element (RVE) of the composite material.

[0066] This step is an indispensable part of solving the mechanical equilibrium of multiphase materials. The specific logical relationship and subsequent calling mechanism are as follows: First, this step uses the macroscopic elastic modulus of alumina fiber obtained in step S11. Compared to Poisson Using this as the basic input data, a complete fiber linear elastic stiffness tensor is constructed internally. .

[0067] Secondly, the mechanical model output in this step has a strong coupling calling relationship with several subsequent key steps: (1) In step S13, the fiber mechanical model, as the core constitutive property, is forcibly bound and assigned to the step of determining the geometric model parameters to construct the fiber three-dimensional geometric entity domain through the automated script, thus completing the spatial intersection of geometric data flow and material data flow; (2) In step S14, when the finite element solver assembles the global stiffness matrix of RVE and solves the overall equilibrium equation, it must call the elastic stiffness tensor constructed in this step at the integration point of each mesh element belonging to the fiber domain, so as to calculate the instantaneous stress update inside the fiber and share the complex load transmitted from the matrix in collaboration with the interface model. (3) In step S14, the integral point state variables calculated based on the mechanical model directly determine and output key microscopic response results such as the fiber axial stress cloud diagram. If the mechanical model in this step is missing, the solver in subsequent steps will cause the global finite element matrix to become unstable or unable to solve because it cannot obtain the stiffness contribution of the reinforcing phase.

[0068] For chopped or discontinuous alumina fibers, the fibers can be simplified to cylinders, elliptical cylinders, or cylinders with rounded ends. The average fiber length should not be limited to tens to hundreds of micrometers; when the alumina fibers in the actual material are in the millimeter range, the average fiber length is preferably set to 0.5 mm to 5 mm based on actual measurements, or extended to a wider range depending on the specific material. The average fiber diameter, length distribution, and aspect ratio distribution should be determined statistically by microscopic images, fracture analysis, or X-ray micro-CT. For millimeter-sized fibers, the fiber length is usually larger than the grain size of the aluminum alloy matrix; the model should show the fiber passing through one or more grains, rather than the grains surrounding or enclosing a single fiber.

[0069] For continuous alumina fibers or fiber preforms, the fibers can be modeled as continuous cylinders running through the RVE, periodically extended fiber segments, or periodic array structures constructed according to the actual weaving method. The modeling of continuous fibers or millimeter-long fibers should be determined according to the actual arrangement, and can be unidirectional arrangement, two-dimensional woven arrangement, three-dimensional preform arrangement, or measured orientation distribution.

[0070] In one embodiment, constitutive parameters of interfacial damage can be determined based on experimental data. The fiber-matrix interface is a key region affecting the creep properties of alumina fiber-reinforced aluminum matrix composites. To avoid prediction bias caused by the assumption of perfect bonding, this invention introduces an interface model between the fiber and the matrix.

[0071] The interface model can take one of the following three forms: The first type is the fully bonded model. This model is suitable for preliminary analysis where the interface bonding is good, the interface slippage is weak, and the prediction target is mainly the overall steady-state creep rate.

[0072] The second type is the surface-based cohesive contact model. This model is suitable for simplified engineering analysis and can describe interface normal cracking, tangential slip, and damage degradation, with high modeling efficiency.

[0073] The third type is the zero-thickness Cohesive element model. This model is suitable for fine analysis and can insert COH3D6 or COH3D8 elements between the fiber and the matrix to describe the interfacial traction-separation behavior, interfacial damage initiation, and damage evolution.

[0074] The Cohesive element is a finite element used to simulate interface cracking, slippage, and damage evolution. In this invention, the Cohesive element is used to describe the interfacial behavior between alumina fibers and an aluminum alloy matrix.

[0075] This invention introduces a fully bonded model, a surface-substrate cohesive contact model, or a zero-thickness cohesive element model between the fiber and the matrix, which can describe interface normal cracking, tangential slip, damage initiation, and damage evolution.

[0076] In a preferred embodiment, the interface is described using a bilinear traction-separation model. When the interface is undamaged, its normal and tangential traction forces satisfy: ; ; ; In the formula, For the interface normal traction force, and These are the traction forces in two mutually orthogonal tangential directions at the interface; , , , and These are the interface normal separation displacement and two tangential slip displacements, respectively.

[0077] The maximum nominal stress criterion can be used to initiate interface damage: ; In the formula, , , These are the critical intensities in the interface normal and the two tangential directions, respectively.

[0078] The evolution of interface damage is described using the energy criterion: ; In the formula, For interface damage variables, Release energy to the current interface. The energy released at the onset of damage, This represents the interfacial fracture energy.

[0079] When it is necessary to consider the slow interface slippage at high temperatures, an equivalent interface creep slip relationship can be further introduced: ; In the formula, The tangential creep slip rate of the interface. Let be the interface slip constant. The interfacial stress index. This is the activation energy for interface sliding.

[0080] It should be noted that the interfacial creep slip constant and interfacial fracture energy should not be directly specified as fixed values, but should be obtained through single fiber push-out experiments, microdroplet debonding experiments, interfacial shear experiments, or inversion of the overall creep curve of composite materials.

[0081] According to the following Figure 2 After constructing the Norton-type or Norton-Bailey-type creep equation, the creep constant A, stress exponent n, and creep activation energy Q will be determined.

[0082] The input to the fiber-matrix interface model is the normal stiffness. Tangential stiffness Interfacial critical strength and Interfacial fracture energy And slip parameters, etc. and Based on the equivalent elastic modulus of the interface layer shear modulus and equivalent thickness The critical strength and fracture energy can be estimated by single-fiber push-out, microdroplet debonding, or interfacial shear experiments, or inverted from the creep curve of the composite material. Interfacial normal traction force. and tangential traction force It refers to the relative displacement of the nodes on the fiber side and the matrix side during the solution process. , and damage variables Calculated.

[0083] ; ; It should be noted that the interface normal traction force in the above formula and tangential traction force , It can be the state parameter of the finite element model during the incremental iterative solution process.

[0084] The above parameters can be combined to form the constitutive parameters for interface damage.

[0085] Specifically, the construction process and data logic of the fiber-matrix interface model are as follows: Experimental data provides the equivalent elastic modulus of the interface layer. shear modulus and equivalent thickness Initial material constants are constant; this step can directly estimate the initial normal stiffness of the interface based on the initial constants input from experimental data. and tangential stiffness , In the subsequent finite element solution process of step S14, the solver will calculate the real-time relative displacement (i.e., separation displacement) between the fiber-side and matrix-side nodes. , , and interface damage variables The bilinear traction-separation model established in this step is invoked to dynamically calculate and update the interface traction force under the current incremental step.

[0086] The purpose of constructing the fiber-matrix interface model is to quantitatively describe the load transfer, tangential slip, normal cracking, and interface stiffness degradation mechanisms between alumina fibers and aluminum alloy matrices under high-temperature service conditions, thereby overcoming the prediction bias caused by the "complete bonding assumption" in traditional finite element models that overestimates the fiber reinforcement effect.

[0087] In terms of logical connections and data flow between steps, this step constitutes a crucial link in realizing multiphase nonlinear collaborative calculation of composite materials and is not unused in subsequent steps. The specific logical relationship and subsequent calling mechanism are as follows: First, this step uses the equivalent elastic modulus, thickness, and strength of the interface layer from experimental data as initial input data to construct the initial normal / tangential stiffness and damage criteria of the interface. Secondly, the interface model output in this step can be frequently called in subsequent steps: (1) In step S13, the interface model, as the core contact constitutive property, is forcibly bound and assigned to the common contact surface of the fiber and matrix with determined geometric model parameters through an automated script (i.e., binding is achieved through surface-based Cohesive contact or zero-thickness Cohesive elements), completing the convergence of interface material data flow and spatial topological information; (2) In step S14, when the finite element solver performs incremental iterative solution, it must call the interface model established in this step in real time at each contact node of the interface, calculate the instantaneous traction force and damage variables based on the relative displacement of the nodes, thereby realizing the dynamic redistribution of load between the fiber and the matrix; (3) In step S14, based on the state variables calculated by the interface model, the microscopic results such as interface slip and damage area are directly output; and in step S14, the interface parameters set in this step are inverted and corrected according to the macroscopic experimental curve of the composite material. If the interface model in this step is missing, the subsequent finite element solution will be unable to simulate the real high-temperature damage evolution behavior of the interface.

[0088] The interface model is used to describe load transfer, tangential slip, normal cracking, and stiffness degradation between the fiber and the matrix. In step S13, the fiber-matrix contact surface is provided by surface-based Cohesive contacts or zero-thickness Cohesive elements. In step S14, the traction force and damage variables are updated incrementally, and the interface slip and damage region are extracted. Steps S15-16 then correct the interface parameters based on the composite material experimental curves.

[0089] In one embodiment, the operation of determining the geometric model parameters may include the following sub-steps: S41. Extract fiber characteristic parameters from the experimental data, and use the fiber characteristic parameters to construct a fiber-penetrating representative volume unit.

[0090] S42. Using the fiber-penetrating representative volume unit, construct an engineering equivalent matrix model and a polycrystalline matrix model respectively. The engineering equivalent matrix model is a model in which the aluminum alloy matrix is ​​a uniform and continuous medium. The polycrystalline matrix model is a model in which grain seed points are generated in the aluminum alloy matrix region after deducting the fiber volume, and a polycrystalline matrix model is formed by three-dimensional partitioning.

[0091] S43. Extract the model parameters of the engineering equivalent matrix model and the polycrystalline matrix model respectively to obtain the geometric model parameters.

[0092] In one embodiment, representative volumetric units can be established based on the actual microstructure of the composite material. The Reinforcing Vessel (RVE) can be set as a cube or cuboid, and its dimensions should be determined through statistical convergence rather than fixed to a certain empirical value. For materials with fiber lengths in the millimeter range and matrix grain sizes ranging from 10 μm to several hundred μm, fiber-scale RVEs and grain-scale RVEs should be distinguished; fibers act as independent reinforcing phases, passing through multiple grains or RVE boundaries, while grains belong only to the aluminum alloy matrix region.

[0093] It should be noted that this step can use a spatial topology generation algorithm to transform the discrete experimental data obtained in step S11 into a computer-recognizable three-dimensional geometric model. Its core data mapping and automatic construction mechanism is as follows: (1) Boundary initialization and data transfer: The algorithm first extracts the target fiber volume fraction from the experimental data. and the average fiber length, combined with the preset volume fraction tolerance. Initialize and define the global three-dimensional spatial domain of the representative volume element. .

[0094] (2) Feature data sampling and single fiber generation: For the first fiber that needs to be generated... For the root fiber, the algorithm directly calls the macroscopic statistical distribution function obtained in step S11, and uses the Monte Carlo sampling method to randomly extract the specific length of the fiber. ,diameter Polar angle and azimuth Its center point in the spatial domain is generated by a random number generator. Based on these sampling parameters, the algorithm calculates the unit vector of the fiber direction. And generate the corresponding spatial cylindrical fragments.

[0095] (3) Interference test and periodic extension: The algorithm calculates in real time the minimum distance between the centerline of the newly generated candidate fiber and the centerline of the existing fiber. Based on the diameter data determined in step S11, it is determined whether geometric interference has occurred. If overlap occurs, it is discarded and resampling is performed; if there is no overlap but the fiber penetrates the boundary region... Then, cutting and periodic mirroring are performed according to the relative plane coordinates.

[0096] (4) Logical termination determination: The algorithm continuously executes the above data sampling and geometry generation process in a loop, and accumulates the volume of generated fibers in real time until the total volume fraction of fibers in the current domain reaches the target volume fraction input in step S11. The error is less than the tolerance. When the orientation distribution statistical deviation meets the threshold, the construction is automatically terminated and the complete fiber-penetrating RVE geometric flow data is output.

[0097] Based on the fiber length distribution, diameter distribution, orientation distribution, target volume fraction, and porosity data from experiments, a fiber-penetrating RVE was constructed. First, the RVE region was defined. and volume fraction tolerance Then, for the i-th fiber, the length is extracted from the measured statistical distribution. ,diameter Polar angle and azimuth And randomly generate the center point. Fiber direction unit vector The center line is represented as , A radius is generated with this centerline as the axis. A cylinder or a cylinder with rounded ends, and with Find the fiber segments within the RVE by finding their intersections; periodically replicate the portions that cross the boundary according to their relative surface coordinates.

[0098] ; ; For each candidate fiber generated, the minimum distance between its centerline and the centerlines of existing fibers is calculated. If the non-overlapping condition is not met, discard and regenerate. Repeat the above process until the fiber volume fraction error is less than [value missing]. The statistical deviation between the orientation distribution and the measured distribution is less than a preset threshold, and the fiber cross-sections at the relative boundaries meet the periodic matching requirements. This is achieved by changing... , , Multiple RVEs are generated using random seeds, and the final size is determined based on the steady-state creep rate convergence criterion described later.

[0099] Fiber-penetrating RVE is used to transform the statistical structure information of experimental data into a computable three-dimensional mesoscopic geometric domain, ensuring that millimeter-scale fibers can penetrate multiple matrix grains or RVE boundaries, rather than being simplified to short rods within a single grain. An aluminum matrix is ​​built within this geometric domain. Step S13 combines the fibers, matrix, and interface and meshes them. Step S14 applies periodic boundary conditions, solves on the RVE, and obtains macroscopic creep properties through volume averaging.

[0100] The purpose of constructing the fiber-penetrating representative volume element (RVE) is to transform the discrete, macroscopic microstructural statistical information of experimental data into a computer-computable three-dimensional mesoscopic geometric domain with clear spatial coordinates, and to establish a rigorous periodic topological extension relationship. This enables the realistic characterization of the motion enhancement constraint features of millimeter-scale or high aspect ratio discontinuous fibers passing through multiple matrix grains or RVE boundaries, eliminating the load transfer mechanism deviation caused by artificially cutting fibers or forcibly isolating fibers inside a single grain in traditional finite element models.

[0101] In terms of logical connections and data flow between steps, this step serves as the geometric foundation for the entire mesoscopic simulation finite element solution, forming the indispensable underlying data basis for all subsequent physics calculations and material assignments. It is not unused in subsequent steps. The specific logical relationships and subsequent calling mechanisms are as follows: (1) This step uses the statistical functionals of fiber length distribution, diameter distribution, orientation distribution and volume fraction of experimental data as basic inputs, and instantiates them into spatial geometric entities through a spatial topology generation algorithm; (2) The construction of the aluminum matrix model directly depends on the RVE global spatial domain output in this step. and fiber geometry By subtracting fiber and pore volumes through Boolean cutting operations, the aluminum alloy matrix domain is precisely derived. ; (3) In step S13, the fiber geometry and the matrix geometry generated in this step are recombined in the CAE kernel through spatial topology intersection, and the mesh is discretized guided by the RVE size convergence criterion set in this step, and a complete finite element mesh model containing the solid domain and the contact surface is automatically created. (4) In step S14, the solver applies periodic boundary conditions to the assembled global stiffness matrix based on the RVE relative boundary node coordinates defined in this step, and performs multiphase nonlinear time-dependent stress incremental iterative solution entirely within the geometric mesh space derived from this step. (5) In step S14, when performing the calculation of the macroscopic equivalent properties of the composite material, the total RVE volume defined in this step must be called. The volume average is used as the denominator for the integral. If the RVE geometric domain constructed in this step is missing, the subsequent matrix generation, mesh discretization, boundary application, and volume average integration will completely lose their spatial dependence.

[0102] The RVE size is determined based on the average grain size, fiber diameter, fiber spacing, target volume fraction, and computing power: the grain-scale RVE should contain a sufficient number of matrix grains and be able to accommodate representative fiber-penetrating segments; when the fiber length is greater than the RVE side length, the fibers that cross the RVE boundary are periodically extended or treated with equivalent long fibers.

[0103] The RVE size satisfies the following statistical convergence condition: ; In the formula, For the first Steady-state creep rate corresponding to each RVE dimension The convergence threshold is preferably set to 5%.

[0104] The spatial distribution of fibers can be generated using random sequential adsorption algorithms, Monte Carlo algorithms, repulsive potential function iterative algorithms, or measured geometry based on CT reconstruction. Each fiber is determined by its centerline position, diameter, orientation angle, and intersection with the RVE boundary. When the fiber length is greater than or close to the RVE size, only the penetrating segment of the fiber can be retained in the RVE, without requiring the entire fiber to be contained within the RVE. The fiber centerline parameters can be expressed as: ; Fiber orientation is represented by polar angle and azimuth angle. If the measured results show that the fibers are approximately three-dimensionally random oriented, they can be uniformly distributed within the range of 0 to 2π, and... The material is uniformly distributed within the range of -1 to 1. If the material exhibits orientation segregation caused by extrusion, infiltration, or preformation, the orientation distribution function obtained from microscopic images or CT statistics should be used.

[0105] The fiber volume fraction is determined based on the actual material preparation results. For randomly chopped or millimeter-scale discontinuous fiber-reinforced aluminum matrix composites, the fiber volume fraction is preferably 5% to 20%. When using fiber preforms or continuous fiber reinforcement systems, the volume fraction can be extended to 25% to 35%, but the effects of fiber agglomeration, infiltration defects, and increased porosity on material properties need to be considered.

[0106] The following conditions should be met during fiber formation: no geometric overlap between fibers; the fiber-RVE boundary meets the periodic extension requirement; the total fiber integral reaches the target value; the fiber orientation distribution is consistent with the microscopic statistical results; pores or defect regions can be introduced depending on the actual material conditions. In the polycrystalline model, grain boundaries should be located inside the aluminum alloy matrix and terminate at the fiber-matrix interface; grains should not be constructed into ring structures surrounding a single fiber.

[0107] In one embodiment, the present invention can construct a two-level aluminum matrix model using fiber-penetrating representative volume units according to the prediction accuracy requirements, thereby obtaining an engineering equivalent matrix model and a polycrystalline matrix model.

[0108] The first level is the engineering equivalent matrix model. This model treats the aluminum alloy matrix as a homogeneous continuous medium and uses a creep constitutive model to describe its high-temperature deformation behavior. This model has high computational efficiency and is suitable for material parameter screening, fiber volume fraction optimization, and preliminary engineering design.

[0109] The second level is the polycrystalline Voronoi matrix model. Based on the actual grain size distribution, this model generates grain seed points within the aluminum alloy matrix region after deducting the fiber volume, and forms a polycrystalline matrix through three-dimensional Voronoi partitioning. Alumina fibers act as independent reinforcing phases, penetrating one or more matrix grains; grain boundary cohesive units can be inserted between adjacent grains to describe grain boundary slip and grain boundary damage, but these units are only arranged between the aluminum matrix grains.

[0110] The Voronoi model is a polyhedral structure model based on spatial seed point partitioning, commonly used to simulate the grain structure in polycrystalline materials. In this invention, a three-dimensional Voronoi partitioning can be used to establish the grain structure of the aluminum alloy matrix; alumina fibers, as independent reinforcing phases, penetrate the matrix grains, and Voronoi grain boundaries represent only the grain boundaries within the aluminum alloy matrix.

[0111] This invention employs a two-tiered modeling approach: an engineering equivalent matrix model and a polycrystalline Voronoi matrix model. In the polycrystalline model, Voronoi grains only delineate the aluminum alloy matrix region, while alumina fibers, acting as independent reinforcing phases, penetrate multiple grains. Grain boundaries terminate at the fiber-matrix interface, balancing computational efficiency with refined failure analysis capabilities.

[0112] Optionally, the number of grains should not be directly fixed at 64, 100, or other values, but should be determined based on the matrix volume after deducting fibers and pores, and the actual average grain size. The number of grains can be estimated using the following formula: ; In the formula, For the number of grains, The volume of the aluminum alloy matrix after deducting fibers and pores, The average grain volume is denoted as .

[0113] It should be noted that when calculating the number of grains in a polycrystalline matrix... In this process, the simultaneous subtraction of pore volume and fiber volume is achieved through the collaborative data flow transmission of the finite element preprocessing script and the geometric volume measurement algorithm, thereby ensuring that the number of grain seed points generated accurately matches the actual microstructure density of the matrix. The specific simultaneous subtraction and construction mechanism is as follows: (1) Theoretical subtraction of data flow dimension: The preprocessing script first receives the statistical material characteristic parameters of the experimental data and directly retrieves the target fiber volume fraction. With target porosity At the numerical calculation level, through formulas The theoretical volume of the net matrix is ​​calculated in advance. Furthermore, this is combined with the measured average grain volume. Solve for the total number of grains to be generated. .

[0114] (2) Digital calibration of geometric topology: In order to eliminate the small volume deviations caused by random fibers and pores in spatial arrangement and boundary truncation, the preprocessing script calls the geometric measurement API query interface of the finite element kernel in the geometric reconstruction stage to obtain the complete fiber entity domain in real time. precise digital volume and porous solid domain precise digital volume Subsequently, the total geometric volume of RVE was used. Subtracting the sum of the numerical volumes of the two mentioned above, the volume calculated in the first step... Dynamic boundary calibration is performed to obtain the absolutely accurate actual geometric volume of the matrix.

[0115] (3) Spatial domain constraint and seed point placement: The final number of grains determined after dynamic calibration. This will serve as a loop control variable, guiding the topology generation algorithm to only perform Boolean cutting operations on the effective geometric domain of the matrix after subtracting fiber and porous entities. Internally, deployment A number of polycrystalline seed points are used, and three-dimensional Voronoi subdivision is performed based on the nearest neighbor criterion. This ensures that the spatial density of the grains is not artificially diluted by the presence of fibers and pores, achieving complete synchronous subtraction of pore and fiber volumes at both the quantitative and spatial geometric levels.

[0116] It should be noted that the import mapping of the EBSD grain orientation data and the grain-fiber Boolean cutting operation follow a strict temporal logic of "geometric topology freezing takes precedence over material property assignment" to eliminate orientation assignment failures caused by topological reconstruction and ID rearrangement of geometric elements (cells) during Boolean operations in the finite element preprocessing kernel. EBSD is an abbreviation for Electron Backscatter Diffraction. It is a microscopic analysis technique used to measure grain orientation, grain boundary distribution, and texture information.

[0117] The specific execution steps and the logical order of data flow are as follows: (1) First stage: Polycrystalline basic topological partitioning. The preprocessing script first divides the polycrystalline basic topology into complete representative volumetric unit spatial domains. Within, a global polycrystalline Voronoi geometric entity is generated based on the average grain size of the experimental data, completing the initial polycrystalline grain boundary topological space partitioning. At this point, each grain is a continuous and complete convex polyhedron.

[0118] (2) Second stage: Fiber cross Boolean cut. Subsequently, the script introduces random alumina fiber geometry and uses the fiber geometry as the tool to perform fully automatic Boolean cut operations on the polycrystalline geometric entities generated in the first stage. In this process, the fiber geometric lines smoothly pass through multiple grain geometries and cut them, thereby accurately deriving the net geometric domain of the aluminum alloy matrix after being divided by the fiber at the geometric level. At this point, all geometric boundaries and topological units of the entire RVE model are completely frozen, and the final geometric features (Cells) of each segmented grain are assigned a fixed index ID.

[0119] (3) Third stage: EBSD orientation data mapping assignment. After the geometric topology is completely determined, the script activates the mapping of material property flow. The script traverses each independent matrix grain geometric unit generated after Boolean cutting, and synchronously reads the discrete EBSD grain orientation data file in the experimental data; using the spatial centroid matching algorithm or index association algorithm, the three-dimensional Euler angles in the EBSD data are mapped. The orientation matrix is ​​converted into a local crystal orientation matrix, and then precisely assigned to the corresponding grain geometry by calling the underlying orientation assignment API of the finite element analysis software.

[0120] By employing the aforementioned "cut first, then assign" topological reorganization logic, it is ensured that even after undergoing drastic geometric mesh changes, each independent matrix grain can still accurately and stably carry its measured anisotropic crystal orientation information, providing a unique orientation boundary flow for solving the crystal plasticity or anisotropic creep equilibrium equations in subsequent steps.

[0121] When the RVE size is small or the grain size is large, a fine model with a small number of grains can be used; when the RVE size is large, the fiber length is in the millimeter range or the grain size is small, the number of grains should be increased appropriately, or an engineering equivalent matrix model combined with a local polycrystalline model should be used to avoid excessive computation.

[0122] In one embodiment, an aluminum matrix model can be constructed based on experimental data such as grain size distribution, porosity, and fiber volume fraction, combined with matrix creep constitutive modeling. For the engineering equivalent model, the RVE parent region is first established. Then subtract the fiber entity from it. and pore region The aluminum matrix region was obtained. And assign elastic, plastic and creep parameters .

[0123] For the polycrystalline Voronoi model, based on the effective volume of the matrix and average grain volume Calculate the number of grains ,exist Internally generated grain seed points satisfying minimum spacing constraints and according to the nearest neighbor criterion Divided into grain domains If EBSD data is available, then each grain is assigned a crystal orientation based on the measured orientation distribution. Specifically, the effective matrix volume, number of grains, and grain domains can be calculated as follows: ; .

[0124] It should be noted that this step uses a finite element preprocessing script to automatically convert the macroscopic statistical parameters of the experimental data into a geometric topological model of the aluminum matrix that includes the real physical boundaries. The specific data flow call and automatic construction mechanism for the two-level matrix model is as follows: (1) Data mapping and construction of engineering equivalent matrix model: First, extract fiber volume fraction from experimental data. and porosity Parameters; pre-built in the computer kernel with a size of global parent space domain Subsequently, as an independent fibrous entity and the pre-defined porous entity This is a tool for Boolean operations, performing a complete Boolean subtraction operation on the parent space domain to accurately output the net aluminum matrix continuous geometric domain after deducting reinforcing phases and defects. .

[0125] (2) Data mapping and construction of polycrystalline Voronoi matrix model: The average grain size data of aluminum alloy matrix was obtained from experimental data and converted into average grain volume. Next, the script calls the net matrix geometric domain volume obtained from the above Boolean operation. The total number of seed crystal points to be deployed is calculated based on the formula. Subsequently, the spatial coordinate generator in the net matrix domain Internal distribution seed points Furthermore, the minimum safe spacing between seed points is strictly set based on the lower limit of grain size determined by experimental data to avoid generating non-physically abnormally small grains. Finally, based on the nearest neighbor geometry rule, a polyhedral grain network covering the entire domain is automatically constructed by a topology generation algorithm. After obtaining the two-level model, all the model data can be obtained, and the geometric model parameters can be derived.

[0126] The purpose of constructing the aluminum matrix model is to provide a continuous medium space for the composite material to bear the main time-dependent creep deformation, and to provide a unique topological matrix for subsequent fiber embedding, interface generation and grain boundary division.

[0127] The geometric model parameters can be spatially combined and discretized into finite element meshes within the software kernel using the matrix model, fiber, and interface models generated in subsequent step S13; alternatively, in the incremental solution of subsequent step S14, the finite element solver must traverse every mesh integration point within this model (aluminum matrix) and call the creep constitutive equations to update the matrix stress and creep strain; ultimately, the equivalent creep strain and grain boundary damage within this model domain can be extracted. Without this model, continuous load transfer of the entire composite material cannot be achieved.

[0128] The aluminum matrix model is used to bear the main time-dependent creep deformation and provides a continuous geometric domain for fiber embedding, interface generation and grain boundary separation.

[0129] This invention uses a random sequential adsorption algorithm, Monte Carlo algorithm, repulsive potential function iterative algorithm, or CT reconstruction method to generate a fiber-penetrating RVE model. For millimeter-scale alumina fibers, the fiber is allowed to pass through the RVE boundary and be periodically extended, thereby avoiding the incorrect simplification of the fiber into a micrometer-scale short rod located only inside a single grain.

[0130] S13. Construct a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters.

[0131] In one embodiment, after obtaining the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters, a three-dimensional simulation model of the aluminum composite material can be constructed using the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters to obtain a finite element model.

[0132] By simulating aluminum composite materials, the creep properties of aluminum composite materials can be measured in a finite element model to obtain the predicted parameters of the creep properties of aluminum composite materials.

[0133] In one embodiment, constructing the finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters may include the following sub-steps: S131. Construct a volumetric element model of the fiber and matrix based on the geometric model parameters using the Boolean topology operation algorithm.

[0134] S132. Bind the interface damage constitutive parameters, the matrix creep constitutive parameters, and the fiber creep constitutive parameters to the volume element model to obtain the finite element model.

[0135] Reference Figure 3 The diagram illustrates a flowchart of the operation for constructing a finite element model according to an embodiment of the present invention.

[0136] In one embodiment, a geometric combination can be performed based on geometric model parameters, and then the geometric combination can be meshed according to interface damage constitutive parameters. Finally, material values ​​can be assigned according to matrix creep constitutive parameters and fiber creep constitutive parameters to construct a three-dimensional simulation model of aluminum composite material, thereby obtaining a finite element model.

[0137] It should be noted that the assembly of the multiphase material geometric model in this step does not rely on manual interaction and assembly through a graphical user interface (GUI). Instead, it utilizes a pre-written Python automated script based on the finite element software kernel architecture to call the underlying geometric modeling API (Application Programming Interface) to perform fully automated spatial topology assembly and mesh discretization. The automated assembly techniques and execution logic for fiber geometry, matrix grain geometry, and interface regions are as follows: (1) Geometric array reading and spatial instantiation: Read the fiber three-dimensional coordinate parameter array from the geometric model parameters And the topological information of the matrix grain seed points; through code instructions, the above abstract data stream is automatically instantiated into a set of mutually independent matrix entities (Part-Matrix) and fiber entities (Part-Fiber) located in the same global coordinate system.

[0138] (2) Automated Boolean topology operations: The script calls the kernel's Boolean operation API interface (BooleanOperations), using the fiber entity set as the tool body, to force the Boolean subtraction and intersection operations (Boolean Cut / Intersect) to be performed on the matrix entities. Through the computer's underlying algorithm, it automatically addresses and eliminates spatial overlapping volumes, thereby deriving a complex matrix net space domain that penetrates the polycrystalline boundary, and completely freezing the topology index IDs of all geometric units.

[0139] (3) Common Boundary Retrieval and Automatic Interface Generation: After the geometric topology reconstruction is completed, the script uses the topology iteration algorithm and the surface normal vector matching algorithm to automatically retrieve and extract the shared boundary geometric surfaces (Shared Boundaries) that coincide between the fiber and the matrix in the whole domain. Then, the script automatically constructs surface base Cohesive contact pairs along the common boundary, or automatically inserts a layer of zero-thickness Cohesive geometric units through the node offset algorithm, thereby technically completing the automatic encapsulation and generation of the interface region.

[0140] By employing the aforementioned computer script-driven geometric combination techniques, the discreteness of manual operations is eliminated, ensuring the automated and high-fidelity construction of complex discontinuous fiber-penetrating polycrystalline matrix models.

[0141] Next, the fiber geometry, matrix grain geometry, and interface region can be combined to form a complete volume element (RVE) model.

[0142] The preferred modeling sequence is as follows: first, generate an aluminum alloy polycrystalline matrix based on the grain size distribution, then insert millimeter-sized fibers or fiber fragments according to the measured orientation and perform Boolean cutting on the matrix grains so that the fibers pass through multiple grains.

[0143] For simplified engineering models, embedded fiber models or Boolean cut models can be used. For models that require analysis of interface damage, Boolean cut is preferred to form a real fiber-matrix interface, and Cohesive units are inserted at this interface; grain boundary units and fiber interface units should be distinguished from each other, and grain boundaries should not penetrate into the alumina fiber domain.

[0144] The matrix region can use C3D8R, C3D10, or C3D4 elements. For complex geometries, quadratic tetrahedral elements (C3D10) are preferred to improve the accuracy of stress calculations near the curved fiber surface. The fiber region can use C3D8R or C3D10 elements. The interface region can use COH3D6 or COH3D8 elements.

[0145] C3D8R is a three-dimensional eight-node linear hexahedral reduced integral element in finite element analysis software, commonly used in finite element analysis of solid structures.

[0146] C3D10 is a three-dimensional ten-node quadratic tetrahedral element in finite element analysis software, suitable for meshing complex geometric structures.

[0147] COH3D6 is a three-dimensional six-node cohesive element in finite element analysis software, which can be used to simulate three-dimensional interface damage.

[0148] COH3D8 is a three-dimensional eight-node cohesive element in finite element analysis software, which can be used to simulate traction-separation and damage evolution at three-dimensional interfaces.

[0149] The mesh size should be determined through mesh convergence analysis, rather than being fixed to a single value. Preferably, at least 3 to 5 layers of elements should be used along the fiber diameter direction; for millimeter-scale fibers, the number of elements along the fiber length direction should be determined based on the interface slip gradient, fiber end position, and RVE boundary periodic extension requirements, and should not be described by only 8 to 15 elements for the entire fiber. Local mesh refinement should be performed at fiber ends, interface regions, grain boundary-interface junction regions, and densely fiber regions.

[0150] The mesh convergence criterion is: after further refining the mesh, the change in steady-state creep rate, maximum interface damage variable, or average creep strain in the loading direction is less than 3% to 5%.

[0151] It should be noted that the volume element (RVE) model is mathematically and geometrically represented as a set of non-overlapping spaces of a multiphase medium. Its specific geometric combination formulas, physical parameter definitions, and material mapping relationships are as follows: (1) Global spatial topology combination formula: ; ; In the formula, A global three-dimensional geometric space domain representing a representative volume element; The net solid domain of the aluminum alloy matrix represents the geometric model parameters; It represents all individual fiber segments The total solid domain of the alumina fibers; This represents a solid domain at the fiber-matrix interface with physically equivalent thickness (when a surface-matrix contact model is used, this domain degenerates into a common boundary geometry). ); This represents the pore defect domain inside the composite material; This indicates the search for the internal space of the corresponding geometric set. The above formula, from a mathematical topological perspective, strictly defines the RVE as a continuous space completely filled by four phases: matrix, fiber, interface, and pores, without any volumetric penetration between them.

[0152] (2) Component volume fraction constraint formula: ; ; In the formula, This indicates the global volume fraction of alumina fibers in the RVE; Represents the total geometric volume of the RVE model ( ); This represents the absolute summation of the volumes of all independent individual fibers. Indicates the porosity of the material; Represents the total physical volume of all pores; Indicates the first not truncated by the RVE boundary The geometric volume of a cylindrical alumina fiber, in which and These represent the measured diameter and length of the individual fiber, respectively. This set of formulas establishes a precise numerical mapping boundary from the independent microscopic dimensions of a single fiber to the macroscopic compositional content of the composite material.

[0153] (3) Spatial domain-constitutive material mapping function: The interface damage constitutive parameters, the matrix creep constitutive parameters, and the fiber creep constitutive parameters can be bound to the volume element model to obtain a finite element model.

[0154] At the level of finite element integral solution, any spatial coordinate point within the RVE The material response is parameterized and forcibly assigned using a region-specific constitutive mapping function: When spatial coordinates When the matrix creep constitutive parameters are used, the matrix elastoplastic-creep constitutive matrix is ​​automatically called to update the stress. When spatial coordinates At that time, the fiber elastic stiffness tensor of the fiber creep constitutive parameters is automatically called to calculate the stiffness contribution; When spatial coordinates or When the interface damage constitutive parameters are applied, the interface traction-separation rule and damage evolution constitutive calculation are automatically invoked to calculate the relative displacement.

[0155] Finite element models can be automatically generated using Python scripts in finite element analysis software / CAE, rather than relying on manual step-by-step operations. The script reads the fiber parameter array. After determining the RVE size and grain seed point, the interfaces for geometry creation, Boolean operation, partitioning, set creation, section assignment, and mesh generation are called sequentially: First, the RVE parent entity is created; then, each fiber entity is generated according to the centerline and diameter, and periodic mirrors are generated for fibers that cross the boundary; subsequently, the parent entity is Boolean-cut using the fiber entities and pore entities as tools, retaining the cut aluminum matrix entity and independent fiber entities; fiber surface sets and matrix surface sets are automatically established based on the common boundary of the two entities, and surface-based Cohesive contacts are defined, or zero-thickness Cohesive unit layers are generated along the common boundary; finally, sets of matrix domains, fiber domains, interface domains, grain boundary domains, and pore domains are established, and interface damage constitutive parameters, matrix creep constitutive parameters, and fiber creep constitutive parameters are assigned respectively.

[0156] A complete finite element model can be represented as ,in For aluminum matrix domain, For fiber domain, For an interface domain with equivalent thickness, It is a porous domain; each entity domain does not overlap with the others.

[0157] When using the surface-based cohesive model, Degenerate into interface Fiber volume fraction and porosity are respectively calculated according to... and Calculation; cylindrical fiber without being truncated by the boundary The material response within the finite element model employs a partitioned constitutive model: Time-invoking matrix creep constitutive model, Time-invoking fiber elastic constitutive, or The interface is invoked to traction-separation and damage constitutive model.

[0158] ; ; In one embodiment, the data flow convergence and binding method for interface damage constitutive parameters, matrix creep constitutive parameters, and fiber creep constitutive parameters is as follows: In the underlying architecture of finite element analysis, the fiber geometry model and the material constitutive model belong to two independent data streams. To achieve the physical mapping and intersection of the two, this step uses a Python automated script to execute the following binding procedure: (1) Geometric flow reception and set creation: The script automatically traverses the RVE three-dimensional topology structure after Boolean cutting, and uses the bit field retrieval algorithm to automatically filter and extract the geometric entities (Instances) belonging to alumina fibers in the geometric model parameters, and combine them to define a globally unique geometric unit set (Element Set, set as Set-Fiber), thereby receiving and locking the geometric data flow.

[0159] (2) Material flow reception and section creation: The script calls the linear elastic stiffness matrix and material constants to create a solid material constitutive model inside the software. Then, a homogeneous solid section property (Homogeneous Solid Section, set as Section-Fiber) is established, and the above solid material constitutive model is designated as the only data source for the section, thereby receiving and locking the material data flow.

[0160] (3) Data flow mapping and intersection (Section Assignment): The script calls the underlying API interface function (SectionAssignment) of the finite element analysis software to forcibly assign the section-fiber carrying mechanical constitutive properties to the set-fiber carrying geometric spatial information.

[0161] At this point, the geometric spatial information of alumina fiber and the material constitutive matrix have been strictly mathematically bound at the element integration point level. The two independent data streams have achieved a logical closed loop, providing complete input conditions for the assembly and solution of the overall stiffness matrix in the subsequent step S14.

[0162] S14. Based on the working conditions corresponding to the creep experiment, the finite element model is simulated and solved to obtain the creep performance prediction parameters.

[0163] To improve the accuracy of predictions and better reflect actual conditions, the operating conditions corresponding to the creep experiment can be determined. Then, periodic boundary conditions, temperature fields, and loads are applied to the finite element model based on the operating conditions. Simulation is performed under these conditions, and various data are measured to obtain the creep performance prediction parameters.

[0164] As an example, the step of simulating and solving the finite element model based on the working conditions corresponding to the creep experiment to obtain creep performance prediction parameters may include the following sub-steps: S141. Determine the working conditions corresponding to the creep experiment and determine the periodic boundary conditions of the finite element model, wherein the periodic boundary conditions include: the displacement distance of corresponding nodes on the relative boundary.

[0165] S142. Simulate the finite element model using the periodic boundary conditions to record the changing parameters of the finite element model under the periodic boundary conditions.

[0166] S143. Calculate creep performance prediction parameters using the aforementioned changing parameters, wherein the creep performance prediction parameters include: creep strain-time curve, steady-state creep rate, matrix equivalent creep strain cloud map, fiber axial stress cloud map, interface slip, interface damage variable, grain boundary damage variable, and creep performance variation law.

[0167] To enable the finite element model to represent the local microstructure in an infinitely large composite material, periodic boundary conditions should be applied.

[0168] Given a finite element model with dimensions of For the corresponding nodes on the relative boundary, their displacements satisfy: ; In the formula, and These represent the displacements of the nodes corresponding to the relative boundaries. For macroscopic average strain tensor. and These are the coordinates of the corresponding node.

[0169] In finite element analysis software, periodic boundary conditions can be automatically applied through *Equation constraints, reference point coupling, or Python scripts. To avoid rigid body displacement, a reference node or the necessary degrees of freedom of a reference point should be constrained. Periodic boundary conditions are a commonly used boundary constraint method in RVE models, which can make corresponding nodes on opposite boundaries satisfy periodic displacement relationships, thereby simulating infinitely periodically repeating local structures inside the material.

[0170] The temperature field can be set as a uniform isothermal field, or a non-uniform temperature field obtained from the thermal analysis of the actual component can be imported. If residual stress after cooling is considered, a thermo-mechanical coupling analysis should be performed first to calculate the thermal residual stress caused by the difference in thermal expansion coefficients between the matrix and the fiber, and then creep analysis should be performed.

[0171] The load can be constant tensile stress, constant displacement, graded stress, or actual service load spectrum. For uniaxial creep simulation, constant tensile stress loading is preferred, and macroscopic stress control is achieved through reference point coupling or uniformly distributed loads on the boundary surface.

[0172] After applying periodic boundary conditions, a creep analysis step can be set in the finite element analysis software and the solution can be obtained. Specifically, a creep analysis step can be established in the finite element analysis software.

[0173] This invention links the microscopic creep response (RVE) with the macroscopic creep response through periodic boundary conditions, and obtains the macroscopic creep strain and steady-state creep rate of the composite material through volume averaging.

[0174] Preferably, a static equilibrium analysis step can be set first to make the model reach stress equilibrium under the initial load; then a Visco analysis step or a Static General creep analysis step can be set to solve for long-term creep.

[0175] The total creep time should be determined based on experimental or service conditions. In the examples, experimental times of 100h to 500h can be used; for life extrapolation analysis, it can be extended to more than 1000h, but the rationality of the extrapolation needs to be verified by experimental data.

[0176] The time increment is automatically controlled. The initial time increment is preferably 1 / 1000 to 1 / 200 of the total duration, and the maximum time increment is preferably no more than 1 / 20 to 1 / 10 of the total duration. If the model contains interface damage or contact nonlinearity, the maximum time increment should be appropriately reduced to ensure that the damage evolution process can be accurately captured.

[0177] For models containing cohesive damage, viscous regularization can be used to improve convergence if necessary, but the proportion of viscous dissipation energy to total energy should be controlled at a low level, preferably not exceeding 5%.

[0178] In a specific operational method, the creep analysis step and solution process of finite element analysis software specifically includes the following steps: (1) In the Initial analysis step, establish the periodic node equations, reference point coupling relationship and minimum constraint to prevent rigid body displacement, and define the temperature field, material initial state and necessary initial values ​​of thermal residual stress.

[0179] (2) Establish Static, General preload, and smoothly load the external stress or displacement from zero to the target value so that RVE satisfies static equilibrium at the start of creep; the displacement, stress and interface traction force output in this step are used as the initial state of subsequent creep steps.

[0180] (3) Establish the Visco creep analysis step and set the total creep time. Initial time increment Minimum time increment Maximum time increment The maximum increment and creep integral error tolerance (CETOL) are set, and automatic time increment is enabled. For cases involving interface damage, further limits are imposed based on the damage evolution rate. .

[0181] (4) During the k-th time increment, the creep strain increment is calculated by calling the creep law selected by the matrix creep constitutive parameters at the matrix volume integral point. The fiber integration point calls the elastic constitutive parameters of the fiber creep constitutive model; the interface nodes update the traction force and damage variables according to the relative displacement increment. Taking the time-hardening type as an example, the incremental relationship can be written as: ; (5) Update the matrix stress using the total strain decomposition relationship and establish the finite element equilibrium residuals. The updated matrix stress and the overall equilibrium equations are as follows: ; ; (6) Solve using the Newton-Raphson method If the force residual, displacement increment, and energy error are all less than the convergence tolerance, the current increment is accepted and the process proceeds to the next increment; if convergence fails, the time increment is automatically reduced and the iteration restarts. Below If convergence is still not achieved, check the interface stiffness, damage softening slope, mesh quality, and boundary constraints. Viscous regularization can also be used to improve stability.

[0182] (7) Output matrix creep strain, fiber stress, relative interface displacement, interface damage variables, reaction force, and volume average quantity at preset time intervals. In the above formulas, , , and Constitutive parameters derived from matrix creep Let be the matrix elastic stiffness tensor. For matrix equivalent stress, The gas constant is... Absolute temperature The current creep time, The strain-displacement matrix, For interface shape functions, For interface traction force, Here is the tangent stiffness matrix.

[0183] Through the above simulation, the parameters mentioned above can be recorded, and the varying parameters of the finite element model can be obtained. These varying parameters can then be used to calculate creep performance prediction parameters.

[0184] This invention employs a method combining experimental calibration with finite element simulation using finite element analysis software. This not only directly fits the overall creep curve of the composite material, but also establishes microscopic models of the aluminum alloy matrix, alumina fibers, and fiber-matrix interface, thereby improving the physical interpretation capability of creep performance prediction.

[0185] Specifically, volume averaging can be performed to obtain the macroscopic stress, macroscopic strain, and creep strain of the composite material, thereby obtaining creep performance prediction parameters.

[0186] In one operating mode, the macroscopic average stress is: ; The macroscopic average strain is: ; In the formula, This represents the total volume of the RVE.

[0187] The creep strain in the loading direction can be obtained by subtracting the instantaneous elastic strain from the total strain: ; Steady-state creep rate can be calculated during the approximately linear phase of the creep curve: ; Post-processing results include: creep strain-time curves, steady-state creep rate, matrix equivalent creep strain contour map, fiber axial stress contour map, interface slip, interface damage variables, grain boundary damage variables, and the variation law of creep performance under different microstructural parameters.

[0188] To verify whether the predicted parameters meet the requirements and whether there are any deviations, in one embodiment, after the step of simulating and solving the finite element model based on the working conditions corresponding to the creep experiment to obtain the creep performance prediction parameters, the method may further include the following steps: S15. Determine the creep strain-time curve of the creep performance prediction parameters and the curve deviation parameter of the creep test curve of the experimental data.

[0189] S16. Correction processing is performed based on the curve deviation parameters, wherein the correction processing includes correction of the aluminum alloy matrix creep parameters, correction of the fiber-matrix interface parameters, and correction of the fiber characteristic parameters.

[0190] In one embodiment, the simulated creep strain-time curve can be compared with the experimental curve. If there is a significant deviation between the simulation results and the experimental results, the following corrections are made in that order.

[0191] First, the creep parameters of the aluminum alloy matrix were corrected to make the simulation curve of the unreinforced matrix material consistent with the experimental curve of the matrix.

[0192] Secondly, the fiber-matrix interface stiffness, interface strength, interface fracture energy, and interface slip parameters were modified to make the initial creep stage and steady-state creep stage of the composite material consistent with the experimental results.

[0193] The fiber volume fraction, fiber length distribution, fiber orientation distribution, and porosity were further adjusted to make the model closer to the real microstructure.

[0194] Finally, RVE size convergence and mesh convergence checks are performed.

[0195] Model validation metrics include: steady-state creep rate error, creep strain error at a specified time point, consistency of creep curve trend, and rationality of damage location. Preferably, the steady-state creep rate error is controlled within 15%, and the creep strain error at a specified time point is controlled within 20%. If the experimental data exhibits significant dispersion, the error standards can be appropriately relaxed based on experimental repeatability.

[0196] The validated model can be used to conduct parametric predictions, including the influence of fiber volume fraction, fiber aspect ratio, fiber orientation distribution, interfacial strength, interfacial fracture energy, temperature, stress level and porosity on the creep properties of composite materials.

[0197] This invention improves the reliability of the model through experimental curve verification and parameter inversion correction. The verified model can be used for parameterized prediction of temperature, stress, fiber volume fraction, interfacial strength and porosity.

[0198] To further illustrate the implementation process of the present invention, three practical application embodiments are described.

[0199] Example 1 is a prediction of the creep performance of millimeter-scale alumina fiber reinforced ZL109 aluminum alloy composite material under constant temperature and load conditions. It is used to illustrate how the present invention completes the experimental parameter calibration, fiber-penetrating RVE modeling, interface model establishment, application of periodic boundary conditions, creep solution and experimental verification.

[0200] In Example 1 of a specific application, the creep performance of millimeter-scale alumina fiber reinforced ZL109 aluminum alloy composite material can be predicted.

[0201] This embodiment takes millimeter-scale alumina fiber reinforced ZL109 aluminum alloy matrix composite material as the research object and predicts its creep properties under different tensile stress conditions at 350℃.

[0202] 1. Materials and experimental conditions.

[0203] The matrix material is ZL109 aluminum alloy, and the reinforcing phase is alumina fiber. Based on microscopic images, fracture surface observation, or X-ray micro-CT, the average length of the alumina fibers is in the millimeter range, for example, 0.5 mm to 3 mm; the average diameter can be taken as 6 μm to 15 μm based on actual measurements; the average aspect ratio is typically greater than 100; and the fiber volume fraction is 15%. The aluminum alloy matrix grain size ranges from 10 μm to several hundred μm, therefore a single fiber can penetrate multiple matrix grains. The composite material is prepared using an extrusion casting process.

[0204] The creep test temperature was set at 350℃, and the applied tensile stresses were 40MPa, 60MPa, and 80MPa, respectively. The total test time was 200 hours. During the experiment, the creep strain-time curves of the matrix aluminum alloy and composite material under different stress levels were recorded.

[0205] 2. Calibration of matrix creep parameters.

[0206] The matrix creep parameter calibration process is as follows: Figure 2 As shown, a high-temperature creep test was first conducted on the unreinforced ZL109 aluminum alloy. Based on the steady-state creep rate at different stress levels at 350℃, the Norton power-law equation was used for fitting.

[0207] In this embodiment, the matrix creep model adopted is: ; The temperature is fixed at 350℃, and the temperature term is incorporated into a constant. In China. Through the study of and By performing linear fitting, the stress index of the matrix material is obtained. Then, the experimental curves were used for correction. This ensures that the simulated curves of the matrix match the experimental curves of the matrix.

[0208] If it is necessary to predict creep performance in the range of 300℃ to 450℃ simultaneously, an Arrhenius-type creep equation with a temperature term is used, and the creep activation energy is fitted through multi-temperature experiments. .

[0209] 3. Establishment of RVE geometric model.

[0210] Grain-scale RVEs are established based on fiber volume fraction, fiber spacing, and matrix grain size. For millimeter-scale fibers, the RVE does not need to completely contain the entire fiber; instead, fiber segments penetrating the RVE boundary are periodically extended. The RVE size is determined by the statistical convergence of steady-state creep rate, average strain, and interfacial damage variables.

[0211] Alumina fiber segments were generated in the RVE using a Python script. Each fiber was determined by its centerline position, diameter, orientation angle, and intersection with the RVE boundary. Fiber lengths were determined using millimeter-level measured distributions; fiber diameters followed a measured statistical distribution. Fiber orientation could be determined using a three-dimensional random distribution or an orientation distribution obtained from CT or microscopic cross-section statistics.

[0212] When generating fibers, the centerline position and orientation of candidate fibers are first randomly generated; then it is determined whether the fiber overlaps with an existing fiber; if they do not overlap, the fiber is retained; if they overlap, the fiber is regenerated. This process is repeated until the total fiber integral reaches 15%, ensuring that the fiber can pass through multiple grains rather than being surrounded by a single grain.

[0213] For fibers crossing the RVE boundary, a periodic extension method is used to make corresponding parts appear on the relative boundary, ensuring the periodicity of the RVE structure and the continuity of millimeter-level fibers.

[0214] 4. Material properties and interface model.

[0215] The aluminum alloy matrix adopts an elastoplastic-creep constitutive model, in which the elastic stage is described by the elastic modulus and Poisson's ratio, and the creep stage is described by the Norton-type creep equation.

[0216] The alumina fiber was modeled using a linear elasticity. Its elastic modulus was taken from experimental or handbook values ​​in the range of 250 GPa to 380 GPa, and its Poisson's ratio was taken from 0.22 to 0.25.

[0217] Fiber-matrix interfaces can be described using fully bonded, surface-matrix cohesive contacts, or zero-thickness cohesive elements. The random fiber RVE and interface modeling process is as follows: Figure 3 As shown. The initial stiffness of the interface is determined by the interface layer thickness and the equivalent elastic modulus, while the interface strength is obtained from single fiber push-out experiments, microdroplet debonding experiments, or literature inversion. The tangential creep slip of the interface is described by the following equation: ; in, and The results were obtained through inversion of the creep curve of the composite material. The interface damage was controlled by an energy criterion; when the interface release energy reached the critical fracture energy, the interface stiffness gradually degraded.

[0218] 5. Grid division.

[0219] The matrix is ​​meshed using C3D10 tetrahedral elements, while the fibers are meshed using C3D8R or C3D10 elements. For the polycrystalline matrix model, aluminum matrix grains are first generated, and optional Cohesive elements are placed at the grain boundaries. Then, penetrating alumina fibers are inserted to form the fiber-matrix interface. COH3D6 or COH3D8 elements are inserted at the fiber-matrix interface, and local mesh refinement is applied to the fiber ends, fiber-grain boundary junctions, and dense fiber regions.

[0220] Mesh convergence is determined by comparing the steady-state creep rate under different mesh densities. When the change in steady-state creep rate is less than 3% after further mesh refinement, the mesh is considered to meet the accuracy requirements.

[0221] 6. Boundary conditions and load settings.

[0222] Periodic boundary conditions are applied to the RVE. The displacement relationships are established between corresponding nodes on the opposite surfaces in the X, Y, and Z directions using the *Equation constraint function of the finite element analysis software.

[0223] Macroscopic tensile loads were applied in the X direction, corresponding to three stress levels: 40 MPa, 60 MPa, and 80 MPa. The model temperature was uniformly set to 350℃. First, a static equilibrium analysis was performed to ensure that the RVE reached stress equilibrium at the initial moment of loading; then, a creep analysis was conducted for 200 hours.

[0224] 7. Solving and post-processing.

[0225] After submitting the solution to the finite element analysis software, the average strain of the RVE in the X direction is extracted. The creep strain is obtained by subtracting the instantaneous elastic strain from the total strain, and the creep strain-time curve is plotted.

[0226] Simultaneously, the maximum creep strain of the matrix, the axial stress of the fibers, the tangential slip at the interface, and the interface damage variables were extracted. The analysis results show that in the initial stage of loading, the fibers bear part of the load, and the creep of the matrix is ​​significantly constrained; as time increases, slippage occurs at the interface of some fiber ends, and the local creep strain of the matrix increases; when the interface damage further develops, the overall creep rate shows an upward trend.

[0227] 8. Result verification.

[0228] The simulated creep curves at 40MPa, 60MPa, and 80MPa were compared with the experimental curves. If the simulated curve is too low in the initial creep stage, it indicates that the interface constraint may be set too strong, and the interface tangential stiffness should be reduced or the interface creep slip rate should be increased. If the steady-state creep rate is too high, it indicates that there may be deviations in the matrix creep parameters or fiber volume fraction, and the matrix creep parameters and the actual fiber content should be checked first.

[0229] After parameter correction, the model can predict the creep deformation trend of composite materials under different stress levels quite well. This embodiment shows that the method of the present invention can be used to predict the creep properties of short fiber reinforced aluminum matrix composites.

[0230] Example 2 is a prediction of the effect of different fiber volume fractions on the creep properties of composite materials. It is used to illustrate how the present invention changes the volume fraction of alumina fibers based on the verification model and analyzes the effect of fiber content on steady-state creep rate, interface damage and creep resistance of materials.

[0231] In Example 2 of a specific application, the influence of different fiber volume fractions on the creep properties of composite materials can be predicted.

[0232] This embodiment illustrates the application of the method of the present invention in the optimization design of material composition.

[0233] 1. Predict the target.

[0234] Keeping the matrix material, millimeter-scale fiber size distribution, interface parameters, and service temperature constant, we only changed the volume fraction of alumina fibers to predict the effect of fiber volume fraction on the steady-state creep rate and interface damage evolution of the composite material.

[0235] The fiber volume fractions were set to 5%, 10%, 15%, 20%, and 25%, respectively. The temperature was set to 350℃, the applied stress was set to 60MPa, and the creep time was set to 300h.

[0236] 2. Model establishment.

[0237] For each fiber volume fraction, a fiber-through RVE was established. To avoid bias caused by randomness, three RVE models with different random seeds were generated for each volume fraction, and the results were averaged.

[0238] The RVE size can remain consistent while satisfying statistical convergence, or it can be adjusted according to the fiber volume fraction and fiber spacing. As the fiber volume fraction increases, the number of fibers penetrating the RVE or the equivalent fiber spacing changes accordingly. In the low volume fraction model, the fiber spacing is larger, and interface damage is mainly concentrated at a few fiber ends or in the fiber-grain boundary junction region; in the high volume fraction model, the fiber spacing decreases, and matrix creep is more strongly constrained, but local fiber aggregation leads to enhanced stress concentration.

[0239] 3. Analysis process.

[0240] The same matrix creep model and interface model as in Example 1 were used. The same periodic boundary conditions, temperature field, and applied stress were applied to all models, and creep analysis was performed on each model.

[0241] After the calculation is completed, the creep strain-time curve of each model is extracted, and the average steady-state creep rate in the range of 100h to 250h is calculated.

[0242] 4. Results Analysis.

[0243] The prediction results show that as the fiber volume fraction increases from 5% to 20%, the steady-state creep rate of the composite material decreases significantly. This is because more alumina fibers participate in the load-bearing process, resulting in stronger constraint on the creep deformation of the matrix.

[0244] When the fiber volume fraction was further increased to 25%, the decline in steady-state creep rate slowed down, and some models even showed premature occurrence of local interface damage. This indicates that a higher fiber volume fraction is not necessarily better. When the fiber content is too high, the fiber spacing is too small, local stress concentration is enhanced, and pores and fiber agglomeration are more likely to occur during the preparation process, thereby weakening the overall performance of the material.

[0245] Therefore, this embodiment can be used to determine a reasonable range of fiber volume fraction. For the material system, under operating conditions of 350°C and 60 MPa, the preferred range of fiber volume fraction is 15% to 20%.

[0246] Example 3 is a fine creep prediction considering polycrystalline matrix and grain boundary slip. It is used to illustrate how to use the Voronoi model of fiber penetration polycrystalline matrix for further fine analysis when the grain structure, grain boundary slip and intergranular damage of aluminum alloy matrix have a significant impact on creep behavior.

[0247] In Example 3 of the specific application, fine creep prediction of polycrystalline matrix and grain boundary slip is considered.

[0248] This embodiment illustrates the application of the method of the present invention in high-precision microscopic prediction.

[0249] 1. Applicable scenarios.

[0250] When the aluminum alloy matrix has a large grain size and significant differences in grain orientation, or when grain boundary slip, grain boundary voids, and intergranular damage are observed in experiments, a homogeneous matrix model alone may not be able to accurately describe the creep damage evolution. For alumina fibers in the millimeter range and matrix grain sizes ranging from 10 μm to hundreds of μm, a single fiber typically penetrates multiple grains. In such cases, a Voronoi model of fiber penetration through a polycrystalline matrix can be used for detailed analysis.

[0251] 2. Establishment of the polycrystalline model.

[0252] The required number of grains is calculated based on the RVE size, fiber volume fraction, and measured grain size. Grain seed points are generated in the aluminum alloy matrix region after deducting the fiber volume, and a polycrystalline matrix structure is formed using three-dimensional Voronoi sectioning. Subsequently, millimeter-scale alumina fibers or their penetrating segments are inserted according to the measured fiber orientation, and Boolean cutting is performed on each grain to allow the fiber to pass through multiple grains. Each grain is assigned a random crystal orientation; if EBSD experimental data is available, the grain is assigned a value based on the actual orientation distribution function.

[0253] The grain interior is modeled using a matrix creep constitutive model, with zero-thickness cohesive elements inserted at grain boundaries. These grain boundary cohesive elements describe grain boundary slip and damage and are only placed between aluminum alloy matrix grains. When a grain boundary encounters the surface of alumina fibers, it should terminate at the fiber-matrix interface and not enter the fiber solid domain. The grain boundary tangential slip rate can be expressed as: ; In the formula, This represents the grain boundary creep slip rate. Where is the grain boundary slip constant. This is the tangential traction force at the grain boundary. The grain boundary stress index, This is the activation energy for grain boundary slip.

[0254] 3. Comparison with the engineering equivalent model.

[0255] Under the same temperature and stress conditions, engineering equivalent matrix RVE and polycrystalline Voronoi matrix RVE were established respectively. The creep curves and damage distributions obtained from the two models were compared.

[0256] The results show that the engineering equivalent matrix model can predict the overall steady-state creep rate well, but it is difficult to reflect local damage at grain boundaries. The fiber-penetrating polycrystalline Voronoi model can reveal strain concentration at grain boundaries, grain boundary slip, local damage at fiber-grain boundary junctions, and damage propagation processes, making it suitable for failure mechanism analysis and high-precision lifetime prediction.

[0257] 4. Engineering significance.

[0258] This embodiment demonstrates that the present invention does not simply stack complex models, but rather selects different modeling levels based on the prediction objective. When only material parameter screening is required, an engineering equivalent matrix model can be used; when analyzing grain boundary damage or local failure mechanisms, a polycrystalline Voronoi fine model is employed. This layered strategy balances computational efficiency and prediction accuracy.

[0259] In this embodiment, the present invention provides a method for predicting the creep properties of aluminum composite materials. Its advantages are as follows: the present invention can acquire experimental data of aluminum composite materials; determine matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters based on the experimental data, wherein the interface damage constitutive parameters are interface cracking parameters determined based on the fiber-matrix contact state; construct a finite element model using the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters; and simulate and solve the finite element model based on the working conditions corresponding to the creep experiment to obtain creep property prediction parameters. The present invention can determine the interface cracking parameters determined by the fiber-matrix contact state, construct a simulation model of the material based on the interface cracking parameters, and make predictions. This can closely reflect the relationship between the fiber and the matrix at high temperatures, thereby avoiding overestimation of the fiber reinforcement effect, reducing prediction deviation, and improving prediction accuracy.

[0260] This invention also provides a device for predicting the creep properties of aluminum composite materials, see [link to relevant documentation]. Figure 4 The diagram shows a schematic diagram of a creep performance prediction device for aluminum composite materials provided in an embodiment of the present invention.

[0261] As an example, the creep performance prediction device for the aluminum composite material may include: The acquisition module 201 is used to acquire experimental data of aluminum composite materials, wherein the experimental data is data collected during creep experiments on aluminum composite materials based on a preset temperature range; The determination module 202 is used to determine the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters respectively based on the experimental data. The interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix. Construction module 203 is used to construct a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters; The prediction module 204 is used to perform simulation and solution of the finite element model based on the working conditions corresponding to the creep experiment to obtain creep performance prediction parameters.

[0262] Optionally, the operation of determining the matrix creep constitutive parameters includes: Creep test curves are extracted from the experimental data, and the stable creep stage of the aluminum composite material is determined based on the slope of the creep test curves. The stable characteristic parameters corresponding to the stable creep stage are extracted from the creep test curve, and the matrix constitutive model is constructed using the stable characteristic parameters; The parameters are solved based on the matrix constitutive model, and the solved parameters are verified based on the simulation curves of the creep experiment curves to obtain the matrix creep constitutive parameters.

[0263] Optionally, the operation of determining the fiber creep constitutive parameters includes: The elastic modulus and Poisson's ratio are extracted from the experimental data, and the Lamé constant is calculated based on the elastic modulus and Poisson's ratio, respectively. The fiber elastic stiffness tensor is determined using the Lamé constant, and the elastic constitutive relation of the alumina fiber is determined based on the fiber elastic stiffness tensor to obtain the fiber creep constitutive parameters.

[0264] Optionally, the operation of determining the geometric model parameters includes: Fiber characteristic parameters are extracted from the experimental data, and fiber characteristic parameters are used to construct fiber-penetrating representative volume units; The engineering equivalent matrix model and the polycrystalline matrix model are constructed using the fiber-penetrating representative volume unit. The engineering equivalent matrix model is a model in which the aluminum alloy matrix is ​​a uniform and continuous medium. The polycrystalline matrix model is a model in which grain seed points are generated in the aluminum alloy matrix region after deducting the fiber volume, and the polycrystalline matrix model is formed by three-dimensional partitioning. The model parameters of the engineering equivalent matrix model and the polycrystalline matrix model are extracted respectively to obtain the geometric model parameters.

[0265] Optionally, the step of constructing a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters includes: Based on the Boolean topology operation algorithm, a volume element model of the fiber and matrix is ​​constructed according to the geometric model parameters; The interface damage constitutive parameters, the matrix creep constitutive parameters, and the fiber creep constitutive parameters are bound to the volume element model to obtain the finite element model.

[0266] Optionally, the step of simulating and solving the finite element model based on the operating conditions corresponding to the creep experiment to obtain creep performance prediction parameters includes: The periodic boundary conditions of the finite element model are determined by determining the working conditions corresponding to the creep test, wherein the periodic boundary conditions include: the displacement distance of corresponding nodes on the relative boundary; The finite element model is simulated using the periodic boundary conditions to record the variation parameters of the finite element model under the periodic boundary conditions; The creep performance prediction parameters are calculated using the aforementioned changing parameters, which include: creep strain-time curve, steady-state creep rate, matrix equivalent creep strain contour map, fiber axial stress contour map, interface slip, interface damage variable, grain boundary damage variable, and creep performance variation law.

[0267] Optionally, the device further includes: The deviation module is used to determine the curve deviation parameters between the creep strain-time curve of the creep performance prediction parameters and the creep test curve of the experimental data after the step of simulating and solving the finite element model based on the working conditions corresponding to the creep test to obtain the creep performance prediction parameters. The correction module is used to perform correction processing based on the curve deviation parameters, wherein the correction processing includes correction of the aluminum alloy matrix creep parameters, correction of the fiber-matrix interface parameters, and correction of the fiber characteristic parameters.

[0268] Those skilled in the art will understand that, for ease of description and brevity, the specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0269] Furthermore, this application also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the creep performance prediction method for aluminum composite materials as described in the above embodiments.

[0270] Furthermore, embodiments of this application also provide a computer-readable storage medium storing a computer-executable program for causing a computer to execute the creep performance prediction method for aluminum composite materials as described in the above embodiments.

[0271] In the description of the embodiments of the present invention, it should be noted that the terms "above," "below," etc., indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. When an element such as a layer, region, or substrate is referred to as being "above" or "on top of" another element, it may be directly on the other element, or there may be an intermediate element. Conversely, when an element is referred to as being "directly on" or "above" another element, there is no intermediate element. It should also be understood that when an element is referred to as being "below" or "under" another element, it may be directly below or under the other element, or there may be an intermediate element. Conversely, when an element is referred to as being "directly below" or "under" another element, there is no intermediate element. Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0272] Those skilled in the art will understand that embodiments of this application may also include computer program products. Therefore, this application may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0273] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), devices, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0274] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0275] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0276] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting the creep properties of aluminum composite materials, characterized in that, The method includes: Obtain experimental data of aluminum composite materials, wherein the experimental data is data collected during creep experiments on aluminum composite materials within a preset temperature range; Based on the experimental data, the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters were determined respectively. Among them, the interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix. A finite element model is constructed using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters. The finite element model is simulated and solved based on the working conditions corresponding to the creep experiment to obtain the creep performance prediction parameters.

2. The method for predicting the creep properties of aluminum composite materials according to claim 1, characterized in that, The operation of determining the matrix creep constitutive parameters includes: Creep test curves are extracted from the experimental data, and the stable creep stage of the aluminum composite material is determined based on the slope of the creep test curves. The stable characteristic parameters corresponding to the stable creep stage are extracted from the creep test curve, and the matrix constitutive model is constructed using the stable characteristic parameters; The parameters are solved based on the matrix constitutive model, and the solved parameters are verified based on the simulation curves of the creep experiment curves to obtain the matrix creep constitutive parameters.

3. The method for predicting the creep properties of aluminum composite materials according to claim 1, characterized in that, The operation of determining the fiber creep constitutive parameters includes: The elastic modulus and Poisson's ratio are extracted from the experimental data, and the Lamé constant is calculated based on the elastic modulus and Poisson's ratio, respectively. The fiber elastic stiffness tensor is determined using the Lamé constant, and the elastic constitutive relation of the alumina fiber is determined based on the fiber elastic stiffness tensor to obtain the fiber creep constitutive parameters.

4. The method for predicting the creep properties of aluminum composite materials according to claim 1, characterized in that, The operation of determining the geometric model parameters includes: Fiber characteristic parameters are extracted from the experimental data, and fiber characteristic parameters are used to construct fiber-penetrating representative volume units; The engineering equivalent matrix model and the polycrystalline matrix model are constructed using the fiber-penetrating representative volume unit, respectively. The engineering equivalent matrix model is a model in which the aluminum alloy matrix is ​​a uniform and continuous medium. The polycrystalline matrix model is a model in which grain seed points are generated in the aluminum alloy matrix region after deducting the fiber volume, and the polycrystalline matrix model is formed by three-dimensional partitioning. The model parameters of the engineering equivalent matrix model and the polycrystalline matrix model are extracted respectively to obtain the geometric model parameters.

5. The method for predicting the creep properties of aluminum composite materials according to claim 1, characterized in that, The construction of a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters includes: Based on the Boolean topology operation algorithm, a volume element model of the fiber and matrix is ​​constructed according to the geometric model parameters; The interface damage constitutive parameters, the matrix creep constitutive parameters, and the fiber creep constitutive parameters are bound to the volume element model to obtain the finite element model.

6. The method for predicting the creep properties of aluminum composite materials according to claim 1, characterized in that, The finite element model is simulated and solved based on the working conditions corresponding to the creep experiment to obtain creep performance prediction parameters, including: The periodic boundary conditions of the finite element model are determined by determining the working conditions corresponding to the creep test, wherein the periodic boundary conditions include: the displacement distance of corresponding nodes on the relative boundary; The finite element model is simulated using the periodic boundary conditions to record the variation parameters of the finite element model under the periodic boundary conditions; The creep performance prediction parameters are calculated using the aforementioned changing parameters, which include: creep strain-time curve, steady-state creep rate, matrix equivalent creep strain contour map, fiber axial stress contour map, interface slip, interface damage variable, grain boundary damage variable, and creep performance variation law.

7. The method for predicting the creep properties of aluminum composite materials according to any one of claims 1-6, characterized in that, After the step of simulating and solving the finite element model based on the working conditions corresponding to the creep experiment to obtain the creep performance prediction parameters, the method further includes: The creep strain-time curve of the predicted creep performance parameters and the curve deviation parameter of the creep test curve of the experimental data are determined. The correction process is performed based on the curve deviation parameters, wherein the correction process includes correction of the aluminum alloy matrix creep parameters, correction of the fiber-matrix interface parameters, and correction of the fiber characteristic parameters.

8. A device for predicting the creep properties of aluminum composite materials, characterized in that, The device includes: The acquisition module is used to acquire experimental data of aluminum composite materials, wherein the experimental data is data collected during creep experiments on aluminum composite materials based on a preset temperature range; The determination module is used to determine the matrix creep constitutive parameters, fiber creep constitutive parameters, interface damage constitutive parameters, and geometric model parameters based on the experimental data, wherein the interface damage constitutive parameters are interface cracking parameters determined based on the contact state between the fiber and the matrix. The construction module is used to construct a finite element model using the matrix creep constitutive parameters, the fiber creep constitutive parameters, the interface damage constitutive parameters, and the geometric model parameters; The prediction module is used to simulate and solve the finite element model based on the working conditions corresponding to the creep experiment to obtain creep performance prediction parameters.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, when the processor executes the program, it implements the creep performance prediction method for aluminum composite materials as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the creep property prediction method for aluminum composite materials as described in any one of claims 1-7.

Images