Multi-scale mechanical modelling of braided composites including process induced defects

点击翻译按钮获取中文摘要

In this project proposal a methodology for multi-scale FEM simulation of braided composites, with defects, will be developed and used to obtain homogenised composites damage models for industrial macro-scale analysis. For materials characterisation a detailed manufacture and test program will be undertaken to characterise a wide range of braid mechanical properties. Defects in braids, in particular fibre damage and architecture irregularities always occur and a variety of braids with different machine settings will be produced to create such defects, which will then be quantified using scanning and computer tomography. Therefore ITA has recently developed an online laser scanning sensor. Furthermore ITA has a wide range of process knowledge as far as the correlation between machine settings and the appearance of braiding defects is concerned. After the infusion, coupon preparation and mechanical testing for stiffness, damage and failure will be undertaken. Digital image correlation techniques will be used to identify and measure variations in surface damage, with cyclic loading to measure damage and plasticity evolution. New models will then be incorporated in the process simulation of the braided preforms to produce numerical models of braids that contain defects. These will be compared to test measurements for validation. IFB have recently de-veloped a novel approach to extract 3D representative volume elements (RVE) from process simulation results which have shown potential for stiffness and damage/failure modelling. These capabilities will be extended here to include the process induced defects and will be supported by FE micromechanical analysis for failure prediction of impregnated yarns. Finally, the work will be used to calibrate homogenised composites damage models in com-mercial FE codes and validated against a test program of coupons and a demonstrator struc-ture. The main new results of this research will be new knowledge of defects generated dur-ing processing of braids. Which will be quantified, and used in prediction of real (imperfect) braid mechanical properties. The work will make use of and link, analysis techniques across all scales from micro- to meso- and finally macro-analysis. Finally, the research will provide important new knowledge of damage/failure of braids under complex loading and make a critical evaluation of current homogenised composites damage models used in commercial FE codes, with suggestions for future improvements.