Fatigue failure test method suitable for planar three-way fabric reinforced flexible composite material

Abstract

The invention discloses a fatigue failure test method suitable for a planar three-way fabric reinforced flexible composite material, and relates to the technical field of high-performance fibers, products and composite materials. The method comprises the following steps: compounding a flexible matrix with a planar three-way fabric through a compression molding process, and performing compression molding to obtain a composite material; preparing a sample of the flexible composite material reinforced by the planar three-way fabric; a microstructure model of the composite material is constructed, and a three-dimensional geometric solid model is established based on a yarn section form and an interweaving path; carrying out quasi-static out-of-plane load test on the sample, and determining mechanical response and failure form; fatigue damage characteristics of the sample are tested through an alternating cyclic load test, and fatigue life data and a performance degradation rule are determined; and analyzing a damage evolution process based on finite element simulation, and establishing an out-of-plane deformation fatigue failure criterion in combination with a simulation analysis result and a test result.

Description

Technical Field

[0001] The present application relates to the technical field of high-performance fibers and products and composite materials, and more specifically to the field of structural design and performance testing of flexible composite materials. Specifically, it relates to a fatigue failure testing method suitable for planar three-dimensional fabric-reinforced flexible composite materials. Background Art

[0002] Flexible composite materials, thanks to their lightweight, flexible, and multifunctional integration, are increasingly being used in cutting-edge fields such as lightweight aerospace structures, flexible electronics, and smart wearable devices. Planar triaxial fabric-reinforced flexible composites, a key structural form in this field, possess unique mechanical response properties due to their three-dimensional interwoven yarn structure. However, this also significantly influences their fatigue failure process due to microstructural parameters.

[0003] Existing technologies still face technical bottlenecks in evaluating the fatigue performance of these materials. Traditional testing methods primarily focus on macroscopic mechanical properties, providing only macroscopic data such as fatigue life and fracture strength, but struggle to reveal microscopic mechanisms such as stress concentration effects in yarn interlacing areas.

[0004] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0005] The embodiments of the present application provide a fatigue failure testing method applicable to planar three-dimensional fabric-reinforced flexible composite materials to solve the above-mentioned technical problems.

[0006] The present application provides a fatigue failure test method applicable to planar triaxial fabric reinforced flexible composite materials, comprising:

[0007] The flexible matrix was compounded with the planar three-dimensional fabric by a compression molding process to prepare a sample of the planar three-dimensional fabric reinforced flexible composite material.

[0008] Construct a microstructure model of the composite material and establish a three-dimensional geometric solid model based on the yarn cross-sectional morphology and interweaving path;

[0009] Performing a quasi-static out-of-plane load test on the sample to determine the mechanical response and failure mode;

[0010] Testing the fatigue damage characteristics of the sample by an alternating cyclic load test to determine fatigue life data and performance degradation law;

[0011] The damage evolution process is analyzed based on finite element simulation, and the out-of-plane deformation fatigue failure criterion is established by combining the simulation analysis results with the test results.

[0012] Furthermore, the parameters of the compression molding process include:

[0013] The flexible substrate is a butyl rubber compound, and the flat three-way fabric is woven with 1260D nylon 66 filaments with a fabric density of 40 strands / 10cm;

[0014] The hot pressing temperature is controlled at 120°C to 180°C, the pressure is 5 to 15 MPa, and the hot pressing time is 10 to 30 minutes;

[0015] By adjusting the yarn linear density and adopting an interwoven structure with an angle of 60°, samples of planar three-dimensional fabrics with different tissue parameters were prepared.

[0016] Furthermore, the construction of the microstructure model includes:

[0017] Using a super-eye digital electron microscope to collect cross-sectional images of the composite material, the yarn's non-circular cross-sectional morphology, buckling state, and spatial path in the flexible matrix were analyzed;

[0018] Establish a three-dimensional geometric solid model to characterize the yarn interweaving angle, cross-sectional shape and matrix filling state;

[0019] Periodic symmetric boundary conditions are set in the microstructure model to simulate the cooperative deformation and stress transfer characteristics of the yarn and the matrix under out-of-plane loads.

[0020] Furthermore, the quasi-static out-of-plane load test includes:

[0021] A hemispherical bursting device was used to apply a normal load at a loading rate of 10 to 50 mm / min;

[0022] Real-time monitoring of the circumferential deformation uniformity of the specimen and capturing of local stress concentration areas through digital image correlation strategies;

[0023] The damaged surface characteristics were observed by fracture peeling experiments, and microscopic analysis was performed using scanning electron microscopy to distinguish three damage modes: yarn breakage, matrix debonding, and interface delamination.

[0024] Furthermore, the implementation of the alternating cyclic load test includes:

[0025] The edge of the specimen is fixed with an arc-shaped fixture, and the inner surface of the arc-shaped fixture is covered with a non-slip rubber layer;

[0026] The fatigue loading stress levels were set to 50%, 60%, and 70% of the static bursting strength of the specimen, the stress ratio was set to 0.1, and the loading frequency was set to 5 Hz.

[0027] At least five groups of parallel specimens were tested at each fatigue loading stress level to determine the load-displacement curve, stiffness degradation rate, and deflection increment in order to draw the fatigue life curve.

[0028] Furthermore, the finite element simulation analysis includes:

[0029] Based on the microstructure model, the Hashin three-dimensional space criterion is used to define the initial damage of the yarn, and the maximum stress criterion is used to define the matrix damage.

[0030] The damage accumulation evolution under cyclic loading is realized through the user subroutine UMAT, simulating the yarn interface debonding, matrix crack propagation and stress redistribution process;

[0031] The stress and number curves and strain field distribution were output by the ABAQUS high cycle fatigue module. After comparison with the experimental data, the model mesh density, interface bonding strength and damage evolution rate parameters were modified.

[0032] Furthermore, the establishment of the out-of-plane deformation fatigue failure criterion includes:

[0033] Based on the critical stress concentration factor, a fatigue life prediction formula is defined under different fabric structural parameters, which is related to yarn density, matrix elastic modulus and interface peeling energy.

[0034] Infrared thermal imaging was used to detect the temperature changes in the damaged area of ​​the specimen and verify the correlation between the stress concentration location and the damage extension path;

[0035] The fatigue failure modes of composite materials reinforced with planar tri-directional fabrics and warp and weft isotropic woven fabrics were compared, and the stress concentration resistance advantage of planar tri-directional fabrics was determined by the deflection increment threshold and stiffness degradation rate.

[0036] Furthermore, the simulation of damage accumulation evolution includes:

[0037] Based on the curvature radius of the yarn interweaving path and the local stress gradient distribution, the mesh density of the finite element model was adjusted in real time, increasing the mesh density in high stress concentration areas to 3 times the baseline value and reducing it to 0.5 times in low gradient areas. Dynamic re-meshing was achieved by embedding a custom script in the ABAQUS high-cycle fatigue module.

[0038] An interface debonding evolution equation based on chaotic dynamics is constructed to dynamically correlate the yarn interface debonding rate with the local strain energy density and matrix creep resistance. Parameters are calibrated experimentally to predict the sudden damage propagation pattern under non-uniform loads.

[0039] The fractal feature data of the yarn fracture surface in the microscopic SEM image is converted into the interface bonding strength correction coefficient of the macroscopic model to realize the coupled analysis of microscopic morphology and macroscopic mechanical behavior.

[0040] Furthermore, the finite element simulation analysis further includes:

[0041] The yarn interweaving nodes of a planar tri-directional fabric are abstracted into a graph structure. The node attributes include yarn tension, curvature, and damage status, and the edge attributes include interweaving angle and contact pressure. The node damage status is updated through the graph convolution operator, and a yarn breakage probability distribution map is output.

[0042] Based on the damage evolution database of different matrix materials, a meta-learning framework is used to fit damage parameters from matrix experimental data;

[0043] The equilibrium equation constraints based on fracture mechanics are introduced into the meta-learning framework, and the constraint strength is dynamically adjusted through adaptive weights to adapt to the Poisson's ratio characteristics of different matrix materials.

[0044] Furthermore, the method further comprises:

[0045] The fractal features of SEM images extracted by the pre-trained visual model are embedded into node attributes, and the microscopic features related to the current damage stage are screened through the attention mechanism;

[0046] Combined with a reinforcement learning algorithm, the action parameters associated with the physical mechanism of yarn breakage are defined, including the interface peeling threshold increment, crack growth rate coefficient, and mesh refinement priority. The action selection is optimized using a double-delayed deep deterministic policy gradient algorithm.

[0047] A knowledge base of matrix material damage is constructed, and the evolution data of butadiene rubber and silicone rubber are used as prior knowledge and transferred to the agent model training of new matrix materials.

[0048] Based on the embodiments provided in this application, a controllable composite of a flexible matrix and a planar three-dimensional fabric is achieved through a compression molding process, ensuring that the sample can truly reflect the actual structural characteristics of the material, providing a reliable carrier for subsequent microscopic modeling and macroscopic testing. Based on the three-dimensional geometric solid model constructed by the yarn cross-sectional morphology and interweaving path, the fabric microscopic structural parameters (such as yarn interweaving angle, cross-sectional shape) are incorporated into the fatigue analysis system for the first time, so that the stress concentration effect, interface damage and other microscopic evolution processes can be quantitatively analyzed. The combination of quasi-static out-of-plane load testing and alternating cyclic load testing realizes the full-cycle data collection of the material's "static failure characteristics-dynamic fatigue performance". The former provides a benchmark reference for the starting point of fatigue damage by determining the mechanical response and failure morphology; the latter establishes a damage evolution curve under dynamic load by obtaining fatigue life data and performance degradation laws. The two complement each other to form a three-dimensional representation of the entire material fatigue process, solving the one-sided problem of the existing technology that relies solely on macroscopic fatigue life data.

[0049] Based on the damage evolution analysis of finite element simulation, the microstructure model is deeply coupled with the macro test data, and through cross-validation of simulation results and test results, an out-of-plane deformation fatigue failure criterion with cross-scale correlation is established. This criterion can not only predict the failure mode of materials under complex load conditions, but also reversely optimize the microstructure design of fabrics (such as adjusting the yarn interweaving density), forming a positive R&D cycle of "testing-simulation-design", filling the gap in the lack of micro-theoretical support for fatigue failure criteria in existing technologies. In response to the stringent requirements for material reliability in scenarios such as aerospace lightweight structures and flexible electronic devices, this method integrates process preparation, structural modeling, multi-load testing and simulation analysis to construct a systematic fatigue assessment solution. Compared with traditional methods, it can more accurately reveal the fatigue weaknesses of materials in actual applications (such as early damage at the yarn interweaving point), provide a scientific basis for optimizing material design and extending service life, and significantly reduce application risks in high-reliability scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The drawings described herein are used to provide a further understanding of the embodiments of the present invention and constitute a part of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. In the drawings:

[0051] Figure 1 This is a flow chart of an optional fatigue failure test method applicable to planar tri-directional fabric reinforced flexible composite materials according to an embodiment of the present application;

[0052] Figure 2 This is a technical roadmap for an optional fatigue failure testing method applicable to planar tri-directional fabric reinforced flexible composite materials according to an embodiment of the present application;

[0053] Figure 3 Schematic diagram of an optional method for preparing a rubber composite material by hot pressing molding according to an embodiment of the present application.

[0054] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0056] Fabric-reinforced flexible composites are made of textiles reinforced with a surface coating of polymer materials such as plastic or rubber. They exhibit large deformation, high load-bearing capacity, and excellent fatigue resistance, and are widely used in rubber diaphragms, buildings, ships, airships, and inflatable space structures. During their application, fabric-reinforced flexible composites, such as rubber diaphragms, are often subjected to out-of-plane loads or out-of-plane cyclic loads, which can damage the composite. Continued use can damage the material structure and compromise safety.

[0057] In the existing technology, due to the complex failure mechanism and the limitations of existing test methods, the research on the damage evolution mechanism and mechanical property simulation of textile-reinforced flexible composite materials under out-of-plane loads is incomplete. It is necessary to establish research on the damage evolution and failure mechanism under out-of-plane loads to improve the structural effect and reliability of textile-reinforced flexible composite materials.

[0058] Planar tri-directional fabrics are made of three groups of yarns that are angled to each other (usually at 60° to each other). Compared with woven fabrics with perpendicular warp and weft interweaving, the load-bearing capacity of this fabric is more uniform in all directions. Regardless of whether the force direction is along the yarn direction, there is no obvious weak direction for resisting shear deformation and shear failure. Even thinner fabrics have good dimensional stability. When impacted by normal external force, the fabric is more uniformly stressed and deformed in the circumferential direction. It is not easy for local stress concentration in planar two-directional fabrics to cause yarn breakage. It is suitable for flexible composite material reinforcements in fields with high requirements for isotropic mechanical properties, such as rubber diaphragm base fabrics.

[0059] The interwoven structure of three yarn groups in planar triaxial fabrics presents challenges in studying the mechanical properties and analyzing the damage evolution of planar triaxial fabric-reinforced flexible composites. Therefore, revealing the structural effects, damage evolution, and failure mechanisms of planar triaxial fabric-reinforced flexible composites under out-of-plane loading has important theoretical and engineering implications for broadening the application of planar triaxial fabrics as reinforcements in flexible composites.

[0060] Planar triaxial fabrics are a new type of two-dimensional textile structure with quasi-isotropic mechanical properties, excellent tear and burst resistance, and overcome the anisotropic mechanical properties of traditional woven fabrics. They hold great promise for use as reinforcements in flexible composites. However, research on planar triaxial fabrics has primarily focused on planar triaxial fabric-reinforced resin-based composites, with fewer reports on planar triaxial fabric-reinforced flexible composites. The unique structural characteristics of planar triaxial fabrics and the large deformability of flexible composites determine the unique and diverse failure mechanisms under load. Therefore, analyzing the failure mechanisms of planar triaxial fabric-reinforced flexible composites is of great significance.

[0061] Currently, theoretical research on textile-reinforced flexible composites lags far behind their practical application. There is an urgent need to analyze their failure mechanisms using out-of-plane deformation structural analysis methods under out-of-plane loads to improve the structural effectiveness and reliability of textile-reinforced flexible composites. A reasonable and accurate microscopic model describing planar triaxial textile composites is an important prerequisite for accurately predicting their mechanical properties and failure mechanisms.

[0062] The microscopic geometric structure model of a planar triaxial fabric composite primarily includes the yarn cross-section and the spatial structure of the yarn within the fabric. Therefore, the geometric form and positional relationships of the yarns within the planar triaxial fabric must be determined, including the yarn cross-sectional shape, the yarn buckling state, and the interweaving state of the yarns within the fabric. A microscopic geometric solid model of a fabric-reinforced flexible composite was established, and the mechanical characteristics of the composite were numerically simulated using finite element methods. The effects of the fabric structure and yarn cross-sectional shape on the mechanical properties of the planar triaxial fabric composite were analyzed, and the damage mechanism was explored. This approach is of great significance for the application of planar triaxial fabrics as composite reinforcements.

[0063] Planar three-dimensional fabric reinforced flexible composite materials have a complex microstructure. Therefore, how to establish a mechanical analysis model that can accurately reflect the actual microstructural characteristics of planar three-dimensional fabric reinforced flexible composite materials and is suitable for numerical solution is a key technical problem solved by this application.

[0064] During numerical simulation, how to establish the damage criterion of the material and the constitutive model of fatigue damage evolution, reasonably characterize the fatigue failure characteristics of planar tri-directional fabric-reinforced flexible composite materials under out-of-plane loads, and accurately explain their failure mechanism is another key technical problem solved by this application.

[0065] In summary, how to establish a mechanical analysis model and material damage criterion that can accurately reflect the actual microstructural characteristics of planar tri-directional fabric reinforced flexible composites and is suitable for solution, and then effectively explain the progressive failure mechanism of planar tri-directional fabric reinforced flexible composites under out-of-plane loads is a challenging topic.

[0066] To achieve this goal, the applicant, based on the structural design, mechanical property analysis, and engineering application of planar three-dimensional fabric reinforced flexible composite materials, took the basic structure and double plain weave of planar three-dimensional fabric as the research objects, and prepared planar three-dimensional fabric by adjusting the fabric structural parameters. First, the influence of the planar three-dimensional fabric structure and structural parameters on the mechanical properties of the planar three-dimensional fabric reinforced flexible composite materials was analyzed, and a finite element analysis model was established. Secondly, a fatigue experimental device was used to study the effects of different structural parameters on the failure characteristics and damage mechanism of the planar three-dimensional fabric reinforced flexible composite materials under out-of-plane loads. Finally, a numerical simulation method was used to analyze the damage evolution process of the planar three-dimensional fabric reinforced flexible composite materials under out-of-plane loads, revealing the fatigue failure mechanism of the planar three-dimensional fabric reinforced flexible composite materials. This project belongs to the frontier field of multidisciplinary intersection of textile materials, material mechanics, and computer programming.

[0067] The research results will provide theoretical support for the structural design of planar three-dimensional fabric reinforced flexible composite materials under out-of-plane loads, and further provide ideas for the establishment of a theoretical system for parameter optimization and failure evaluation of planar three-dimensional fabric reinforced flexible composite materials.

[0068] The applicant used theoretical and numerical simulation methods to systematically study the tensile, tearing and bursting properties of planar three-dimensional fabric-based reinforced rubber composites. The applicant is familiar with the theoretical models related to the mechanics of planar three-dimensional fabric-reinforced flexible composites, the theoretical basis of finite element analysis and simulation. Based on this, the applicant proposed the problem of "research on the damage evolution and failure mechanism of planar three-dimensional fabric-reinforced flexible composites under out-of-plane loads", which needs to be studied and explored. Based on reading a large number of related research hotspots and progress, a detailed theoretical and numerical simulation research plan was formulated, which has real feasibility and is expected to establish theoretical and numerical simulation methods for the damage evolution process of planar three-dimensional fabric-reinforced flexible composites. At the same time, the members of the applicant's research team have been engaged in textile fiber materials and structural design, testing and mechanical analysis for a long time, which can ensure that the key technical difficulties of this project can be overcome.

[0069] In summary, the technical route and research plan proposed in this project are scientific and reasonable, the research method is feasible, and it has a good foundation for preliminary experiments and theoretical research work, and can achieve the project and its research objectives.

[0070] Alternatively, as Figure 1 As shown, the planar three-dimensional fabric reinforced flexible composite material of the present application relates to the technical field of high-performance fibers and products and composite materials. The present application provides a fatigue failure test method applicable to planar three-dimensional fabric reinforced flexible composite materials, including:

[0071] S101, compounding a flexible matrix with a planar triaxial fabric through a compression molding process to prepare a specimen of a planar triaxial fabric-reinforced flexible composite material; S102, constructing a microstructure model of the composite material, and establishing a three-dimensional geometric solid model based on the cross-sectional morphology and interweaving paths of the yarns;

[0072] S103, performing a quasi-static out-of-plane load test on the specimen to determine the mechanical response and failure mode;

[0073] S104, testing the fatigue damage characteristics of the specimens through alternating cyclic loading tests to determine fatigue life data and performance degradation patterns;

[0074] S105, based on finite element simulation analysis of the damage evolution process, combining the simulation analysis results with the test results to establish the out-of-plane deformation fatigue failure criterion.

[0075] The technical roadmap adopted by this application is as follows Figure 2 shown.

[0076] Based on the embodiments provided in this application, a controllable composite of a flexible matrix and a planar three-dimensional fabric is achieved through a compression molding process, ensuring that the sample can truly reflect the actual structural characteristics of the material, providing a reliable carrier for subsequent microscopic modeling and macroscopic testing. Based on the three-dimensional geometric solid model constructed by the yarn cross-sectional morphology and interweaving path, the fabric microscopic structural parameters (such as yarn interweaving angle, cross-sectional shape) are incorporated into the fatigue analysis system for the first time, so that the stress concentration effect, interface damage and other microscopic evolution processes can be quantitatively analyzed. The combination of quasi-static out-of-plane load testing and alternating cyclic load testing realizes the full-cycle data collection of the material's "static failure characteristics-dynamic fatigue performance". The former provides a benchmark reference for the starting point of fatigue damage by determining the mechanical response and failure morphology; the latter establishes a damage evolution curve under dynamic load by obtaining fatigue life data and performance degradation laws. The two complement each other to form a three-dimensional representation of the entire material fatigue process, solving the one-sided problem of the existing technology that relies solely on macroscopic fatigue life data.

[0077] Based on the damage evolution analysis of finite element simulation, the microstructure model is deeply coupled with the macro test data, and through cross-validation of simulation results and test results, an out-of-plane deformation fatigue failure criterion with cross-scale correlation is established. This criterion can not only predict the failure mode of materials under complex load conditions, but also reversely optimize the microstructure design of fabrics (such as adjusting the yarn interweaving density), forming a positive R&D cycle of "testing-simulation-design", filling the gap in the lack of micro-theoretical support for fatigue failure criteria in existing technologies. In response to the stringent requirements for material reliability in scenarios such as aerospace lightweight structures and flexible electronic devices, this method integrates process preparation, structural modeling, multi-load testing and simulation analysis to construct a systematic fatigue assessment solution. Compared with traditional methods, it can more accurately reveal the fatigue weaknesses of materials in actual applications (such as early damage at the yarn interweaving point), provide a scientific basis for optimizing material design and extending service life, and significantly reduce application risks in high-reliability scenarios.

[0078] Currently, there is no published research on the fatigue mechanical behavior and damage of planar tri-directional fabrics in flexible composite materials under out-of-plane loads. This application uses a fatigue testing machine to study the effects of the structure and structural parameters of planar tri-directional fabrics on the fatigue damage performance of their reinforced composite materials under out-of-plane loads, providing a direct basis for achieving material-structure performance matching in design.

[0079] Finite element analysis is used to study the mechanical response of planar tri-directional fabric reinforced flexible composites under out-of-plane loads at the microscopic structural scale. The stress concentration effect, damage initiation, expansion and failure mechanism of the material under out-of-plane loads are analyzed, and an out-of-plane deformation fatigue failure criterion applicable to planar tri-directional fabric / rubber composites is established.

[0080] Furthermore, the parameters of the compression molding process include:

[0081] The flexible matrix is ​​a butadiene rubber compound, the flat three-way fabric is woven with 1260D nylon 66 filaments, and the fabric density is 40 strands / 10cm; the warp and weft homogeneous woven fabric is woven with an SGA598 fully automatic rapier weaving machine, and the fabric specification is 60 strands / 10cm.

[0082] The hot pressing temperature is controlled at 120°C to 180°C, the pressure is 5 to 15 MPa, and the hot pressing time is 10 to 30 minutes;

[0083] By adjusting the yarn linear density and adopting an interwoven structure with an angle of 60°, samples of planar three-dimensional fabrics with different tissue parameters were prepared.

[0084] In this embodiment, if Figure 3 FIG. 1 is a schematic diagram of a method for preparing a rubber composite material by hot pressing molding.

[0085] Based on the examples provided in this application, by limiting the molding process parameters (such as using a composite of a butadiene rubber compound and a 1260D nylon 66 filament fabric), combined with the controllable adjustment of hot pressing temperature, pressure, and time, it is ensured that the interface stress between the flexible substrate and the planar three-dimensional fabric is evenly transferred, thus avoiding the problem of discrete sample performance caused by parameter fluctuations in traditional processes. By adjusting the yarn linear density and the interwoven structure with a 60° angle between them, the influence of fabric structure parameters (such as density and angle) on the fatigue performance of the material can be systematically studied, providing a process basis for the customized design of materials in multiple scenarios (such as lightweight aerospace structures and flexible electronic devices), significantly improving the controllability of sample preparation and the flexibility of structural design.

[0086] Furthermore, the construction of the mesostructure model includes:

[0087] Using a super-eye digital electron microscope to collect cross-sectional images of the composite material, the non-circular cross-sectional morphology, buckling state, and spatial path of the yarn in the flexible matrix were analyzed;

[0088] SolidWorks 3D mechanical design software was used to build a 3D geometric solid model to characterize the yarn interlacing angle, cross-sectional shape, and matrix filling state.

[0089] Periodic symmetric boundary conditions are set in the microstructure model to simulate the cooperative deformation and stress transfer characteristics of the yarn and matrix under out-of-plane loads.

[0090] Based on the examples provided in this application, cross-sectional images of composite materials were captured using a super-eye digital electron microscope, overcoming the resolution limitations of traditional optical observations for non-circular yarn cross-sections and buckling states, ensuring that the microstructural model faithfully reproduces the spatial path of the yarn within the matrix. By setting periodic symmetric boundary conditions to simulate cooperative deformation under out-of-plane loads, microscopic parameters such as yarn interweaving angles and cross-sectional shapes are linked to macroscopic mechanical responses. This overcomes the oversimplification of the complex interweaving characteristics of fabrics by existing modeling methods, enabling finite element simulations to accurately capture the stress transfer characteristics of the yarn-matrix interface, providing a reliable model foundation for the microscopic mechanistic analysis of fatigue damage.

[0091] Furthermore, the quasi-static out-of-plane load test includes:

[0092] A hemispherical bursting device was used to apply a normal load at a loading rate of 10 to 50 mm / min;

[0093] Real-time monitoring of the circumferential deformation uniformity of the specimen and capturing of local stress concentration areas through digital image correlation strategies;

[0094] The damaged surface characteristics were observed by fracture peeling experiments, and microscopic analysis was performed using scanning electron microscopy to distinguish three damage modes: yarn breakage, matrix debonding, and interface delamination.

[0095] Based on the examples provided in this application, a hemispherical bursting device was used to apply a normal load, simulating the out-of-plane stress scenarios experienced by flexible composite materials in practical applications (e.g., compressive deformation of flexible sensors). Dynamic capture of the material failure process under quasi-static loads was achieved by controlling the loading rate at 10-50 mm / min. Combining digital image correlation strategies with scanning electron microscopy analysis, macroscopic deformation uniformity data and microscopic damage patterns (yarn breakage, matrix debonding, and interface delamination) were simultaneously acquired. This approach addresses the disconnect between macroscopic and microscopic analysis in traditional testing and provides an experimental basis for multi-dimensional determination of fatigue failure onset.

[0096] Furthermore, the implementation of the alternating cyclic load test includes:

[0097] The edge of the specimen is fixed with an arc-shaped fixture, and the inner surface of the arc-shaped fixture is covered with a non-slip rubber layer to ensure uniform load transmission and prevent the specimen from slipping;

[0098] The fatigue loading stress levels were set to 50%, 60%, and 70% of the static bursting strength of the specimen, the stress ratio was set to 0.1, and the loading frequency was set to 5 Hz.

[0099] At least five groups of parallel specimens were tested at each fatigue loading stress level to determine the load-displacement curve, stiffness degradation rate, and deflection increment in order to draw the fatigue life curve.

[0100] Based on the embodiments provided in this application, the design of the arc-shaped fixture combined with the anti-slip rubber layer effectively avoids the test error caused by the edge slip of the specimen under alternating cyclic loads, thereby improving data reliability. By setting multiple stress levels of 50%-70% of the static bursting strength, a stress ratio of 0.1, and a loading frequency of 5Hz, typical load conditions of the material in actual service are covered, and combined with statistical analysis of at least 5 groups of parallel specimens, a fatigue life curve with statistical significance can be established to avoid the one-sidedness of a single load test. This standardized test protocol provides a repeatable engineering test paradigm for the fatigue performance evaluation of materials in complex vibration environments (such as flexible connection components of aviation equipment).

[0101] Furthermore, the finite element simulation analysis includes:

[0102] Based on the microstructure model, the Hashin three-dimensional space criterion is used to define the initial damage of the yarn, and the maximum stress criterion is used to define the matrix damage.

[0103] The damage accumulation evolution under cyclic loading is realized through the user subroutine UMAT, simulating the yarn interface debonding, matrix crack propagation and stress redistribution process;

[0104] The stress and number curves and strain field distribution were output by the ABAQUS high cycle fatigue module. After comparison with the experimental data, the model mesh density, interface bonding strength and damage evolution rate parameters were modified.

[0105] Based on the examples provided in this application, initial damage is defined based on the Hashin three-dimensional spatial criterion and the maximum stress criterion, and a dynamic simulation of damage accumulation under cyclic loading is implemented using the UMAT subroutine. This accurately characterizes the temporal relationship between yarn interface debonding and matrix crack propagation. By comparing and correcting experimental data with the ABAQUS high-cycle fatigue module, dynamic adjustment of parameters such as mesh density and interfacial bond strength resolves the disconnect between damage evolution patterns in existing simulations and actual operating conditions. This enables the finite element model to truly reflect the stress redistribution process of the material under multi-cyclic loading, providing a reliable numerical analysis tool for cross-scale validation of fatigue failure criteria.

[0106] Furthermore, the establishment of the out-of-plane deformation fatigue failure criterion includes:

[0107] Based on the critical stress concentration factor, a fatigue life prediction formula is defined under different fabric structural parameters, which is related to yarn density, matrix elastic modulus and interface peeling energy.

[0108] Infrared thermal imaging was used to detect the temperature changes in the damaged area of ​​the specimen and verify the correlation between the stress concentration location and the damage extension path;

[0109] The fatigue failure modes of composite materials reinforced with planar tri-directional fabrics and warp and weft isotropic woven fabrics were compared, and the stress concentration resistance advantage of planar tri-directional fabrics was determined by the deflection increment threshold and stiffness degradation rate.

[0110] Based on the examples provided in this application, a fatigue life prediction formula is constructed by introducing a critical stress concentration factor, quantitatively correlating parameters such as yarn density and matrix elastic modulus, thus overcoming the limitations of traditional criteria that rely on empirical parameters. Infrared thermal imaging technology is used to verify the correlation between stress concentration locations and damage extension, enabling non-contact monitoring of damage evolution under thermal-mechanical multi-physics fields. Combined with comparative analysis of planar triaxial fabrics and warp-and-weft isotropic woven fabrics, the advantages of triaxial interwoven structures in suppressing stress concentration are intuitively revealed, providing theoretical support for material selection in engineering design (such as high-reliability flexible structural components).

[0111] Based on the examples provided in this application, the mesh density is dynamically adjusted according to the yarn curvature radius and stress gradient, and a custom script is used to achieve mesh densification in high-stress areas and sparseness in low-gradient areas, thereby improving simulation efficiency while ensuring computational accuracy. An interface debonding evolution equation based on chaotic dynamics is constructed, dynamically correlating strain energy density with matrix creep resistance. This allows prediction of sudden damage expansion under non-uniform loads, resolving the problem of delayed response to complex stress paths in traditional simulations. By combining the fractal characteristics of SEM images to correct interfacial bonding strength, a cross-scale coupled analysis of microscopic damage morphology and macroscopic mechanical behavior is achieved, improving the model's adaptability to actual materials.

[0112] In a specific embodiment of the present application, the present application adopts a method that combines experimental characterization, theoretical analysis and numerical simulation to conduct a systematic study on the microstructure, out-of-plane deformation fatigue behavior and damage failure mechanism of a planar triaxial fabric reinforced flexible composite material, specifically including:

[0113] (1) Preparation of planar three-dimensional fabric reinforced flexible composite materials:

[0114] Raw materials: 1260D nylon 66 filament; Nylon rubber compound or other flexible matrix

[0115] Weaving of reinforced skeleton: a homemade flat three-way loom prototype is used to weave flat three-way fabric with a fabric specification of 40 threads / 10cm. The warp and weft homogeneous woven fabric is woven using the SGA598 fully automatic rapier prototype machine with a fabric specification of 60 threads / 10cm.

[0116] Hot pressing process: temperature, pressure, and hot pressing time.

[0117] (2) Determination of the microstructure of planar three-dimensional fabric reinforced flexible composite materials:

[0118] The cross-sectional morphology and path of the yarn were analyzed using the Super Eye digital magnified electron microscope B008 combined with the cross-sectional view of the planar three-dimensional fabric reinforced flexible composite material, and the solid model of the planar three-dimensional fabric reinforced flexible composite material was established using the 3D modeling software SolidWorks.

[0119] (3) Determination of damage behavior of planar triaxial fabric reinforced flexible composites under out-of-plane loads:

[0120] With reference to the bursting standard of fabric, an electronic universal material testing machine was used to conduct quasi-static out-of-plane deformation mechanical experiments on planar three-dimensional fabric reinforced flexible composite materials, and the failure strength, modulus, load-displacement curve and final failure morphology were obtained. The fatigue properties of planar three-dimensional fabric reinforced flexible composite materials were tested using a fatigue testing machine (the corresponding chuck was designed by ourselves). The fatigue loading stress levels were formulated according to the results of the quasi-static out-of-plane deformation test, and were 50%, 60% and 70% of the static bursting strength respectively; the stress ratio at each level was 0.1; the load loading frequency was 5 Hz, and mechanical indicators such as load-displacement curve, stress-number curve, displacement-number curve, strength and modulus were obtained; the failure mode of planar three-dimensional fabric (warp and weft isotropic plain fabric) reinforced composite materials was analyzed by means of macroscopic observation, fracture analysis, peeling test, SEM and infrared detection; the influence of planar three-dimensional fabric structural parameters on the structural response and failure mechanism of the fatigue behavior of its reinforced composite materials was discussed.

[0121] (4) Determination of damage mechanism of planar triaxial fabric reinforced flexible composite materials under out-of-plane loads:

[0122] Based on the established micro-element solid model, the initial damage criteria for the planar triaxial fabric yarn and the flexible matrix were the Hashin three-dimensional space criterion and the maximum stress criterion, respectively. A damage evolution model was established, and a material user subroutine (UMAT) was developed. The quasi-static out-of-plane deformation stiffness and strength of the micro-element model were obtained using ABAQUS, compared with experimental results, and the model was modified. Based on the continuum damage mechanics approach, the initial damage criteria, damage evolution properties, and failure criteria for the component materials were defined, considering factors such as stress level, load, static failure strength, and damage mode under cyclic loading. The material user subroutine (UMAT) was developed. Using the high-cycle fatigue analysis module in ABAQUS, the mechanical behavior and damage process of planar triaxial fabric-reinforced flexible composites under out-of-plane loading were numerically simulated. The stress distribution, load-displacement curves, stress-order curves, strain-order curves, strength, and modulus were obtained. The results were compared with experimental results and the model was modified. The stress concentration effects, damage initiation, propagation, and failure mechanisms of the planar triaxial fabric under out-of-plane loading were analyzed, and a fatigue failure criterion applicable to planar triaxial fabric-reinforced flexible composites was established.

[0123] Furthermore, the simulation of damage accumulation evolution includes:

[0124] Based on the curvature radius of the yarn interweaving path and the local stress gradient distribution, the mesh density of the finite element model was adjusted in real time, increasing the mesh density in high stress concentration areas to 3 times the baseline value and reducing it to 0.5 times in low gradient areas. Dynamic re-meshing was achieved by embedding a custom script in the ABAQUS high-cycle fatigue module.

[0125] In this example, high stress concentration areas include yarn interlacing points (e.g., 60° intersections) in a planar triaxial fabric, significant yarn bends (with minimal curvature), and direct load application areas (e.g., the contact center of the bursting device). These areas are characterized by localized stresses significantly exceeding the average stress, representing the starting points of fatigue damage (e.g., yarn breakage and interfacial debonding), requiring a denser mesh for focused simulation.

[0126] Low-gradient regions include the base between straight yarn segments, areas with large curvature (gentle bends), or areas far from the load. These regions are characterized by uniform stress distribution with minimal variation, low damage risk, and the ability to reduce mesh density for improved computational efficiency.

[0127] An interface debonding evolution equation based on chaotic dynamics is constructed to dynamically correlate the yarn interface debonding rate with the local strain energy density and matrix creep resistance. Parameters are calibrated experimentally to predict the sudden damage propagation pattern under non-uniform loads.

[0128] The fractal feature data of the yarn fracture surface in the microscopic SEM image is converted into the interface bonding strength correction coefficient of the macroscopic model to realize the coupled analysis of microscopic morphology and macroscopic mechanical behavior.

[0129] In an optional embodiment, the dynamic grid density is adjusted based on the following formula:

[0130]

[0131] Where x, y, and z are three-dimensional spatial coordinates (unit: mm) used to locate any point in the composite microstructure, where the z axis is defined as the thickness direction (out-of-plane direction) of the planar triaxial fabric; ρ(x, y, z) is the mesh density at the spatial point (x, y, z), defined as the number of finite element units per unit volume (unit: number of units / mm). 3 ), reflecting the calculation accuracy of the area, high stress or high curvature area is automatically encrypted; ρ0 is the reference grid density, the global unified initial grid density (unit: number of elements / mm 3 ), the value is determined by the performance of the computing device and is usually set to 100 units / mm 3, used for low stress and low curvature background areas; λ is the curvature sensitivity factor, ranging from 0.5 to 2, and is dynamically adjusted by the yarn interweaving curvature radius. For example, when the yarn interweaving angle is 60°, the curvature radius is small, and λ is taken as 1.5 to enhance the mesh densification effect; κ(x, y, z) is the inverse of the curvature radius of the yarn at the point (x, y, z) (curvature), which reflects the degree of yarn bending and is obtained through electron microscope image analysis. The greater the curvature (such as at the interweaving node), the greater the value; k0 is the standard curvature, defined as the average curvature of the yarn in a planar triaxial fabric. It is obtained by taking the average value of the curvature of all yarn interweaving points and is used for normalization processing; η is the stress gradient suppression factor (unit: mm / MPa), ranging from 0.1 to 0.5, which is determined by the viscosity of the matrix material. For example, the viscosity of butadiene rubber compound is relatively high, and η is taken as 0.3 to suppress excessive densification in low stress gradient areas; is the stress gradient at the point (x, y, z) (unit: MPa / mm), obtained through preliminary finite element calculation, which indicates the rate of change of stress in space. The larger the gradient, the more significant the stress concentration.

[0132] Furthermore, the finite element simulation analysis further includes:

[0133] The yarn interweaving nodes of a planar tri-directional fabric are abstracted into a graph structure. The node attributes include yarn tension, curvature, and damage status, and the edge attributes include interweaving angle and contact pressure. The node damage status is updated through the graph convolution operator, and a yarn breakage probability distribution map is output.

[0134] Based on the damage evolution database of different matrix materials, a meta-learning framework is used to fit damage parameters from matrix experimental data;

[0135] The equilibrium equation constraints based on fracture mechanics are introduced into the meta-learning framework, and the constraint strength is dynamically adjusted through adaptive weights to adapt to the Poisson's ratio characteristics of different matrix materials.

[0136] In an optional embodiment, the graph convolution damage state is updated based on the following formula:

[0137]

[0138] Where v and u are nodes in the graph structure. Each node represents a yarn interlacing point in a planar three-dimensional fabric, and the node number corresponds to the coordinates of the yarn intersection in the three-dimensional model. N(v) is the set of adjacent nodes of node v, that is, the nodes that are directly interlaced with node v through yarns. For example, in a 60° interlaced structure, each node is usually connected to three adjacent nodes. is the damage state vector of node v in the l-th layer of graph convolution, which includes: yarn tension: the normalized dynamic tension value (corresponding to the alternating load stress level); curvature: the normalized curvature value of the yarn at this node; damage degree: a dimensionless value from 0 to 1, where 0 indicates no damage and 1 indicates complete breakage. is the damage state vector of the adjacent node u in layer l; c uv W is the contact pressure normalization coefficient between nodes v and u, defined as the ratio of the actual contact pressure to the maximum contact pressure, ranging from 0.1 to 1. The greater the pressure, the smaller the cvu (for example, when the contact pressure is 80% of the maximum pressure, cvu = 0.8). (l) is the neighborhood weight matrix of the l-th layer graph convolution, with a dimension of d×d (d is the state vector dimension, such as 50 dimensions), which is fitted from experimental data through the meta-learning framework and is used to aggregate the damage information of neighboring nodes. is the self-loop weight matrix of the l-th layer graph convolution, with the same dimension as W (l) Consistent, used to preserve the impact of the node’s own damage state. B (l) is a fracture mechanics constraint matrix with dimensions d × 1. It introduces equilibrium equation constraints (e.g., force and moment balance), with adaptive weights ranging from 0.01 to 0.3. When the matrix Poisson's ratio increases (e.g., for silicone rubber), the constraint weights automatically increase to enhance physical consistency. θ(v,u) is the yarn interweaving angle (in degrees) between nodes v and u, directly corresponding to the interweaving structure parameter of 60° in Weight 2. Input is converted to radians (e.g., 60° = π / 3 radians) for matrix operations.

[0139] Based on the embodiments provided in this application, yarn interweaving nodes are abstracted into a graph structure, and the node damage status is updated through a graph convolution operator. This innovatively introduces graph neural networks into composite material fatigue analysis, effectively addressing the nonlinear stress transfer problem in yarn interweaving networks. By fitting the damage parameters of different matrix materials using a meta-learning framework and dynamically adjusting the model weights by introducing fracture mechanics constraints, this method addresses the lack of adaptability of traditional finite element methods to novel matrix materials. The method is particularly suitable for the rapid modeling of flexible matrices such as butadiene rubber and silicone rubber, significantly improving the generalization capability of fatigue simulation in multi-material systems.

[0140] Furthermore, the method further comprises:

[0141] The fractal features of SEM images extracted by the pre-trained visual model are embedded into node attributes, and the microscopic features related to the current damage stage are screened through the attention mechanism;

[0142] Combined with a reinforcement learning algorithm, the action parameters associated with the physical mechanism of yarn breakage are defined, including the interface peeling threshold increment, crack growth rate coefficient, and mesh refinement priority. The action selection is optimized using a double-delayed deep deterministic policy gradient algorithm.

[0143] A knowledge base of matrix material damage is constructed, and the evolution data of butadiene rubber and silicone rubber are used as prior knowledge and transferred to the agent model training of new matrix materials.

[0144] In an optional embodiment, the attention-reinforcement learning action is optimized based on the following formula:

[0145]

[0146] Among them, a t is the action vector at time t, which contains three independent parameters: Δτ: the interface peeling threshold increment (unit: MPa), which is used to adjust the interface damage initiation threshold of the Hashin criterion, with a value range of ±5 MPa, and is dynamically adjusted according to the current damage stage (for example, when damage intensifies in the late fatigue stage, Δτ is -3 MPa to trigger damage in advance); k: the crack growth rate coefficient, which is used to modify the growth rate of the chaotic dynamics equation, with a value of 0.1 to 1.5, and a larger value is taken in high stress concentration areas (such as the bursting load center); p: the mesh encryption priority (dimensionless), a normalized value between 0 and 1, with 1 indicating the highest priority, used for dynamic meshing with weight 8, and associated with the curvature radius and stress gradient (for example, when the curvature radius is less than 0.5 mm, p = 0.9). t is the state vector at time t, consisting of: Ft, a fractal feature vector extracted from SEM images by the pre-trained visual model, such as crack roughness and fracture surface fractal dimension (obtained through SEM image analysis); St, the current fatigue loading stress level (in MPa), corresponding to 50%, 60%, and 70% of the set static bursting strength; and Dt, a damage stage indicator, with 0 indicating the initial state and 1 indicating complete failure, determined by the stiffness degradation rate (e.g., Dt = 0.6 for stiffness degradation > 30%).

[0147] Q(s t ,a;ω) is the Q value function, which is used to evaluate the t The long-term cumulative reward for executing action a is output as a reward score (usually in the range [-1, 1]). A larger value indicates a better action. ω is a trainable parameter of the reinforcement learning model (such as the weight of a neural network), which is optimized through algorithmic iteration and has no physical unit. i is the attention weight of the i-th fractal feature, which is obtained through visual model training and is used to dynamically screen key damage features (for example, the weight of crack roughness can be 0.6); f i (s t ) is the normalization function output of the i-th fractal feature, which converts the SEM image features (such as fractal dimension) into a value of [0,1] (such as the fractal dimension of the fracture surface is 0.8 after normalization).

[0148] Based on the embodiments provided in this application, a pre-trained visual model is used to extract the fractal features of SEM images, and the key damage features are screened through the attention mechanism, thereby realizing the intelligent analysis of microscopic image data and reducing the error of manual interpretation. The reinforcement learning algorithm is combined to optimize the damage simulation parameters (such as interface peeling threshold and crack propagation rate), and the physical interpretability of action selection is improved through the double-delayed deep deterministic policy gradient algorithm, providing an adaptive optimization framework for fatigue simulation of complex fabric structures. The introduction of the matrix material damage knowledge base for transfer learning can quickly transfer the prior knowledge in the historical data to the training of new material models, shorten the R&D cycle, and adapt to the engineering needs of rapid iteration of flexible composite materials.

[0149] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A fatigue failure test method for planar three-dimensional fabric reinforced flexible composite materials, characterized in that: include: The flexible matrix was compounded with the planar three-dimensional fabric by a compression molding process to prepare a sample of the planar three-dimensional fabric reinforced flexible composite material. Construct a microstructure model of the composite material and establish a three-dimensional geometric solid model based on the yarn cross-sectional morphology and interweaving path; Performing a quasi-static out-of-plane load test on the sample to determine the mechanical response and failure mode; Testing the fatigue damage characteristics of the sample by an alternating cyclic load test to determine fatigue life data and performance degradation law; The damage evolution process is analyzed based on finite element simulation, and the out-of-plane deformation fatigue failure criterion is established by combining the simulation analysis results with the test results.

2. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 1, characterized in that: The parameters of the compression molding process include: The flexible substrate is a butyl rubber compound, and the flat three-way fabric is woven with 1260D nylon 66 filaments with a fabric density of 40 strands / 10cm; The hot pressing temperature is controlled at 120°C to 180°C, the pressure is 5 to 15 MPa, and the hot pressing time is 10 to 30 minutes; By adjusting the yarn linear density and adopting an interwoven structure with an angle of 60°, samples of planar three-dimensional fabrics with different tissue parameters were prepared.

3. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 1, characterized in that: The construction of the microstructure model includes: Using a super-eye digital electron microscope to collect cross-sectional images of the composite material, the yarn's non-circular cross-sectional morphology, buckling state, and spatial path in the flexible matrix were analyzed; Establish a three-dimensional geometric solid model to characterize the yarn interweaving angle, cross-sectional shape and matrix filling state; Periodic symmetric boundary conditions are set in the microstructure model to simulate the cooperative deformation and stress transfer characteristics of the yarn and the matrix under out-of-plane loads.

4. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 1, characterized in that: The quasi-static out-of-plane load test includes: A hemispherical bursting device was used to apply a normal load at a loading rate of 10 to 50 mm / min; Real-time monitoring of the circumferential deformation uniformity of the specimen and capturing of local stress concentration areas through digital image correlation strategies; The damaged surface characteristics were observed by fracture peeling experiments, and microscopic analysis was performed using scanning electron microscopy to distinguish three damage modes: yarn breakage, matrix debonding, and interface delamination.

5. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 1, characterized in that: The implementation of the alternating cyclic load test includes: The edge of the specimen is fixed with an arc-shaped fixture, and the inner surface of the arc-shaped fixture is covered with a non-slip rubber layer; The fatigue loading stress levels were set to 50%, 60%, and 70% of the static bursting strength of the specimen, the stress ratio was set to 0.1, and the loading frequency was set to 5 Hz. At least five groups of parallel specimens were tested at each fatigue loading stress level to determine the load-displacement curve, stiffness degradation rate, and deflection increment in order to draw the fatigue life curve.

6. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 1, characterized in that: The finite element simulation analysis includes: Based on the microstructure model, the Hashin three-dimensional space criterion is used to define the initial damage of the yarn, and the maximum stress criterion is used to define the matrix damage. The damage accumulation evolution under cyclic loading is realized through the user subroutine UMAT, simulating the yarn interface debonding, matrix crack propagation and stress redistribution process; The stress and number curves and strain field distribution were output by the ABAQUS high cycle fatigue module. After comparison with the experimental data, the model mesh density, interface bonding strength and damage evolution rate parameters were modified.

7. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 1, characterized in that: The establishment of the out-of-plane deformation fatigue failure criterion includes: Based on the critical stress concentration factor, a fatigue life prediction formula is defined under different fabric structural parameters, which is related to yarn density, matrix elastic modulus and interface peeling energy. Infrared thermal imaging was used to detect the temperature changes in the damaged area of ​​the specimen and verify the correlation between the stress concentration location and the damage extension path; The fatigue failure modes of composite materials reinforced with planar tri-directional fabrics and warp and weft isotropic woven fabrics were compared, and the stress concentration resistance advantage of planar tri-directional fabrics was determined by the deflection increment threshold and stiffness degradation rate.

8. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 6, characterized in that: Simulation of damage accumulation evolution, including: Based on the curvature radius of the yarn interweaving path and the local stress gradient distribution, the mesh density of the finite element model was adjusted in real time, increasing the mesh density in high stress concentration areas to 3 times the baseline value and reducing it to 0.5 times in low gradient areas. Dynamic re-meshing was achieved by embedding a custom script in the ABAQUS high-cycle fatigue module. An interface debonding evolution equation based on chaotic dynamics is constructed to dynamically correlate the yarn interface debonding rate with the local strain energy density and matrix creep resistance. Parameters are calibrated experimentally to predict the sudden damage propagation pattern under non-uniform loads. The fractal feature data of the yarn fracture surface in the microscopic SEM image is converted into the interface bonding strength correction coefficient of the macroscopic model to realize the coupled analysis of microscopic morphology and macroscopic mechanical behavior.

9. The fatigue failure testing method for planar three-dimensional fabric reinforced flexible composite materials according to claim 6, characterized in that: The finite element simulation analysis further includes: The yarn interweaving nodes of a planar tri-directional fabric are abstracted into a graph structure. The node attributes include yarn tension, curvature, and damage status, and the edge attributes include interweaving angle and contact pressure. The node damage status is updated through the graph convolution operator, and a yarn breakage probability distribution map is output. Based on the damage evolution database of different matrix materials, a meta-learning framework is used to fit damage parameters from matrix experimental data; The equilibrium equation constraints based on fracture mechanics are introduced into the meta-learning framework, and the constraint strength is dynamically adjusted through adaptive weights to adapt to the Poisson's ratio characteristics of different matrix materials.

10. The fatigue failure testing method for planar tri-directional fabric reinforced flexible composite materials according to claim 9, characterized in that: The method further comprises: The fractal features of SEM images extracted by the pre-trained visual model are embedded into node attributes, and the microscopic features related to the current damage stage are screened through the attention mechanism; Combined with a reinforcement learning algorithm, the action parameters associated with the physical mechanism of yarn breakage are defined, including the interface peeling threshold increment, crack growth rate coefficient, and mesh refinement priority. The action selection is optimized using a double-delayed deep deterministic policy gradient algorithm. A knowledge base of matrix material damage is constructed, and the evolution data of butadiene rubber and silicone rubber are used as prior knowledge and transferred to the agent model training of new matrix materials.

Images