Progressive Failure Modeling of Multi-Layered Textile Composites
Technical Report,15 Jul 2013,30 Apr 2018
University of Michigan - Ann Arbor Ann Arbor United States
Pagination or Media Count:
Major Goals The major goals of the work is on physics based prediction of the progressive damage and failure response of hybrid 3D woven textile composites H3DWTC using a novel two-scale computational mechanics framework. Here the term hybrid refers to different constituent fibers, including carbon, glass and kevlar that are infused with SC-15 matrix and integrally woven into a single preform. The H3DWTCs are made through a 3D textile weaving process. Three different versions of hybridized architectures will be examined at Unit cell level to determine the progression of damage and failure under loading. A micro-CT analysis of multi-layered textile composites will be conducted to study the effect of microstructure imperfections on predicting the progressive damage and failure response. The micro-CT analysis delivers a list of inputs like identifying the size of a Representative Unit Cell RUC, the fiber tows cross sectional details and the porosity analysis. These subscale microstructure inputs are the building blocks for 3D CAD modeling and FE meshing of Unit Cell at a global scale. A multiscale investigation to study the progressive damage and failure at different length scales will be carried out. In the computational model, the macroscale finite element analysis FEA is carried out at the RUC level, while a novel micromechanics analysis will be implemented simultaneously at the subscale level using material properties of the constituents fiber and matrix as input. Accomplishments The subscale micromechanics analysis was developed based on the N-layers concentric cylinder model NCYL to compute the local fields in the fiber and matrix cylinders. The influence of matrix micro-damage at the subscale that causes the progressive degradation of fiber tow stiffness at the macroscale was studied by a secant moduli approach, to capture the observed pre-peak nonlinear response.
- Numerical Mathematics