A systematic approach to transforming composite 3D images into meso-scale computational models

High performance polymer matrix composites (PMC) have a high specific stiffness and can be used to easily manufacture highly complex components. Many types of defects can occur during molding. Flaws and damage degrade the resulting mechanical properties of the composites material. It is difficult to assess the actual stiffness, strength and fatigue limit of flawed and damaged structures. Among these the fatigue limit is the most difficult to predict. Through a combination of modern imaging techniques and finite element analysis of in-situ fiber bundles, it is now becoming possible to estimate fatigue limits for polymer matrix composites structures with flaws or damage. Composite materials can be imaged with 3D X-ray Computed Tomography (CT) in a sufficient detail to view 3D fiber bundle matrix interfaces. These images can then be directly imported into physical models to be used in finite element analysis. The process of converting these images into computer models for analysis is currently extremely time consuming, difficult and subjective. The method presented here has been developed to bridge this gap.