Mesoscopic damage analysis of braided composites using mesh-superposition techniques for static failure and impact loading

编织是一种经济、快速的连续型纺织预制体生产方法。通过编织角度和编织模式的变化，可以获得最适合应用负载的结构。此外，编织复合材料具有优异的损伤容限，损伤传播受纱线结构的影响。为了准确分析这些损伤过程，必须考虑结构的几何复杂性以及可能发生的各种纤维或纱线、基体和纱线-基体界面失效机制。到目前为止，这是不可能的，大多数损伤模型使用的是近似损伤的均质方法。特别是，动态问题的冲击建模是具有挑战性的，因为详细的体系结构建模和必要的计算能力迄今为止使这种类型的模拟不切实际。为了克服这些限制，本提案将研究网格叠加技术来离散地模拟纱线和基体相，并通过运动耦合将这两个成分结合起来。这种方法大大减少了实际解决方案所需的模型大小和自由度。该方法的一个缺点是纱线和基体体积的叠加导致刚度的重复。与传统模型相比，必须考虑到这一点，以获得等效的应力分布。对于损伤建模和损伤传播，这在整个数值模拟运行时是必要的。此外，如果不采用网格叠加技术来模拟静态破坏或冲击载荷的损伤或分层，则无法使用先前的建模技术。本研究计划探讨编织复合材料在静态破坏和冲击载荷下的细观损伤模型预测的新方法。此外，现有的基于有限元的建模方法和经过验证的复合材料损伤模型将得到应用，新的研究重点是将这些模型与新的网格叠加技术相结合。因此，发生的破坏机制不再是均匀的，而是可以单独和整体地捕获，这导致比以前的建模方法更高的预测质量和更好的材料理解。在本研究中，将推导出一种载荷工况优化的编织结构，并对从静态非对称剪切和裸眼拉伸到冲击和冲击后压缩的不同载荷工况进行验证。该方法不仅局限于编织结构，还将适用于其他编织纺织复合材料，并提高当前在静态破坏和冲击载荷下对几何复杂部件进行建模的能力。

Braiding is an economical and fast manufacturing method to produce continuous textile preforms. Via variations in the braiding angle and braiding pattern it is possible to obtain architectures that can be optimally suited for applied loadings. In addition, braided composites have excellent damage tolerance in which damage propagation is influenced by the yarn architecture. For an accurate analysis of these damage processes it is essential to account for geometric complexity of the architecture and the various fibre or yarn, matrix and yarn-matrix interface failure mechanisms that can occur. To date this has not been possible, and most damage models use a homogenised approach that approximates damage. In particular, impact modelling of dynamic problems is challenging since detailed modelling of the architecture and the necessary computing capacities has so far made this type of simulation impractical.To overcome these limitations, this proposal will investigate mesh superposition techniques to discretely model the yarn and matrix phases, with kinematic coupling to combine the two constituents. This approach greatly reduces the model sizes and the degrees of freedom necessary for a practical solution. One drawback of the method is superposition of yarn and matrix volumes causing duplication of stiffness. This must be accounted for to obtain an equivalent stress distribution compared to a conventional model. For damage modelling and damage propagation this is necessary throughout the runtime of the numerical simulation. Furthermore, previous modelling techniques cannot be used without adaption for the mesh superposition technique to model damage or delamination for static failure or impact loading.This research proposal investigates new methods to predict damage modelling at the mesoscopic scale of braided composites for both static failure and impact loading. Furthermore, existing finite element-based modelling methods and well validated composites damage models will be applied, with new research focussed on combining these models with the new mesh superposition technique. As a result, the occurring damage mechanisms are no longer homogenised, but can be captured separately and holistically, which leads to a higher forecast quality and better material understanding than with previous modelling approaches. In the proposed study, a load case optimized braided architecture will be derived, which is validated for different load cases from static non-symmetrical shear and open-hole tension to impact and compression-after-impact tests. This method is not only limited to braided structures and will be applicable to other woven textile composites and advance current capabilities to model geometrically complex components under static failure and impact loading.