织物橡胶复合结构兼具高柔韧性、可设计性和优异的综合特性，被广泛应用于航天航空等诸多工程领域，其力学性状直接决定大型客机的安全性与可靠性。航空织物橡胶复合结构的复杂材料构型、苛刻巡航工况与极高性能要求，导致其研制与验证周期长、成本高，织物橡胶复合结构力学性能预测及数字孪生模型设计是亟待解决的、面向世界科技前沿、面向国家重大需求的关键科学问题。本论文以织物橡胶复合结构为主题，对其非线性各向异性力学行为、数字孪生预测与在位评价模型和全尺寸服役力学性能分析等方面进行了系列研究。主要内容如下：首先，建立了非正交网状纤维织物橡胶复合材料的各向异性超弹性本构模型，理论推导了此框架的应力-应变关系；发展了非正交复杂织物橡胶复合材料宏细观建模策略；准确预测不同材料取向下织物橡胶复合材料及其复合密封结构的力学响应；探究了网状纤维织物橡胶复合材料各向异性超-粘-伪弹性行为，基于非线性连续介质力学、损伤力学与伪弹性理论框架，拓展了类橡胶复合材料各向异性大变形理论；基于此理论框架，推导了显式应力解并发展复杂本构参数的高效识别方法，揭示了织物纤维取向和纤维-基体相互作用项对大变形回滞行为的影响。其次，发展了织物橡胶复合结构数字孪生预测模型及计算框架；基于深度学习构造三种深度神经网络架构，开展模型性能评估并择优构建预测模型，建立了基于数字图像相关技术的实时力学交互实验平台，证明了直接从特征传感数据实时高保真预测复杂织物橡胶复合结构三维位移场与应力场的可行性与可扩展性。再次，基于数字孪生预测模型、高斯过程回归和动态贝叶斯网络，建立了复杂织物橡胶复合结构数字孪生在位评价模型及其数据交互框架；研究了复杂织物橡胶复合结构在位运行阶段数据交互与大型客机巡航环境下力学性能预测问题。最后，建立了登机门复合密封结构全尺寸数值模型，通过复合密封结构服役力学性能分析与实际装机状态试验验证了该模型的有效性，揭示了整体压缩力学响应分布规律，给出了登机门复合密封结构全尺寸服役力学性能评估的经验公式。研究成果将为织物橡胶复合材料及其复杂织物橡胶复合结构的力学性能表征预测、在位评价及服役力学性能评估提供理论依据、建模方法和实验基础，对于大型客机舱门复合密封结构的国产化与智能化发展具有重要的科学意义。
Fabric rubber composite structures are widely used in many engineering fields such as aerospace because of their high flexibility, designability and excellent comprehensive properties. Their mechanical properties directly determine the safety and reliability of large passenger aircraft. Aviation fabric rubber composite structures have complex material configurations, harsh service environments, and extremely high performance requirements, resulting in long development and verification cycles and high costs. The mechanical performance prediction and digital twin design of fabric rubber composite structures are key scientific issues that need to be solved, and target the global sci-tech frontiers and strive to fulfill the significant needs of the country. This dissertation focuses on the fabric rubber composite structures and conducts a series of studies on their nonlinear anisotropic mechanical behaviors, digital twin prediction and in-situ evaluation models, and full-size service mechanical performance analysis. The main contents are as follows:First, an anisotropic hyperelastic constitutive model of non-orthogonal reticulated fiber fabric composites is established, and the stress-strain relationship based on this framework is theoretically derived. A macro-mesoscopic modeling strategy for non-orthogonal complex fabric rubber composite structures is developed. The changes in mechanical performance of fabric rubber composites and composite sealing structures with different material orientations are accurately predicted. Based on the theoretical framework of nonlinear continuum mechanics, damage mechanics and pseudoelasticity, an anisotropic hyper-visco-pseudo-elastic constitutive model of reticulated fiber fabric rubber composites is established, expanding the theory of anisotropic composites under large deformation. Based on this theoretical framework, an explicit stress solution is derived and an efficient identification method for complex constitutive parameters is developed, revealing the influence of fabric fiber orientation and fiber-matrix interaction terms on the hysteresis behavior under large deformation.Second, a digital twin prediction model and its calculation framework of the complex fabric rubber composite structure are developed. Based on deep learning, three deep neural network architectures are constructed, and the digital twin prediction model is selected based on the performance evaluation functions. A real-time mechanical interactive experimental platform based on digital image correlation technology is built, and a series of experiments are designed to prove the feasibility and scalability of directly realizing real-time high-fidelity prediction of the three-dimensional displacement field and stress field for complex fabric rubber composite structures from characteristic sensing data.Third, based on the digital twin prediction model, Gaussian process regression and dynamic Bayesian network, a digital twin in-situ evaluation model for complex fabric-rubber composite structures and its data interaction framework are established. The data interaction process of complex fabric rubber composite structure during the operation stage and the prediction of mechanical properties of large passenger aircraft under cruising environment are studied.Finally, a full-size numerical simulation model of the composite sealing structure for a large passenger aircraft cabin door is established, and the effectiveness of the model is verified through the service mechanics analysis of the composite sealing structure and the actual installation state experiments. The distribution law of the overall compression mechanical response are revealed, and an empirical formula for the full-size mechanical performance evaluation of the cabin door composite sealing structure is given.These research results will provide theoretical support, modeling methods and experimental foundation for the characterization, prediction, in-situ evaluation, and service mechanical assessment of the mechanical performance for fabric rubber composites and their complex fabric rubber composite structures, holding significant scientific significance for the localization process and intelligent development of large aircraft cabin composite sealing structures.
