Abstract
Usage of discontinuous glass fibers in injection and compression molded
resin components is rapidly increasing to improve their mechanical
properties. Since added fiber contributes to more strength along the fiber
direction compared to transverse direction, the mechanical properties of
such components strongly depend on the fiber orientation. Therefore it is
important to estimate the fiber orientation distribution in such materials. In
this paper we are presenting recently developed method to estimate fiber
orientation from 3D image of fibers generated using micro CT
scan. Typically the 3D images from CT scan are very large and it becomes
difficult to separate each fiber and extract end point information. In this
paper a novel method to address this challenge is presented. The micro-CT
images were broken into finite volume, reducing data size, and then each
fiber was reduced to its own center line, further reducing the data size,
using Mimics® Innovation Suite (Materialise NV). These 3D centerlines
were then used to quantify the 2nd order orientation tensor. The results
from the proposed method are compared with the measurements using
well established industry standard approach called method of ellipses for
validation. The key challenges in estimating the fiber orientation are
identified and future improvements are proposed.
