Author:
Bricker Stephen,Simmons J.P.,Przybyla Craig,Hardie Russell
Abstract
ABSTRACTContinuous fiber reinforced composites (RFC) are hierarchal and complex at multiple scales. In this work, tools are developed to automate the 3D characterization and quantification of the overall microstructure. Structure quantification enables accurate representation for material simulation and property prediction for the integrated computation materials engineering (ICME) of RFC based components. Relationships are developed to describe the key attributes of the microstructure at multiple scales including the individual fibers, tows, weave, porosity, and secondary matrix phases, which are treated as 'gestalts' of the structure. Here gestalt refers to the essence of shape or complete form of key features of the microstructure such as those of the tow architecture of the textile. Visualization tools are developed based on an artificial color scheme that allow the visual recognition of whole tows instead of just the collection of simple lines and curves representative of the fibers, which provides means whereby the gestalt of the microstructure can be visualized at the tow scale. These tools are demonstrated using a 3D dataset of the SiNC/SiC S200 ceramic matrix composite material (CMC) obtained via automated serial sectioning. Methods are then demonstrated to generate microstructure models representative of the characterized material for finite element analyses (FEA).
Publisher
Springer Science and Business Media LLC
Reference13 articles.
1. On the Velocity Gradient Tensor and Fluid Feature Extraction;Haimes;AIAA paper,1999
2. Maximum Likelihood from Incomplete Data via the EM Algorithm;Demptster;Journal of the Royal Statistical Society, B,1977
3. Statistical characterization of SiC/SiC ceramic matrix composites at the filament scale with Bayesian segmentation, hough transform feature extraction, and pair correlation statistics;Przybyla;International SAMPE Technical Conference,2013
4. The EM/MPM algorithm for segmentation of textured images: analysis and further experimental results
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献