Automated segmentation of computed tomography images of fiber-reinforced composites by deep learning
- PDF / 3,135,246 Bytes
- 17 Pages / 595.276 x 790.866 pts Page_size
- 95 Downloads / 187 Views
Automated segmentation of computed tomography images of fiber-reinforced composites by deep learning Aly Badran1,* , David Marshall1, Zacharie Legault2, Ruslana Makovetsky3, Benjamin Provencher3, Nicolas Piche´3, and Mike Marsh3 1
Department of Aerospace Engineering and Sciences, University of Colorado, Boulder, 3775 Discovery Dr. Boulder, Boulder, CO 80303, USA 2 Department of Computer and Software Engineering, Polytechnique Montreal, Montreal, Canada 3 Object Research Systems, Montreal, Canada
Received: 17 April 2020
ABSTRACT
Accepted: 21 August 2020
A deep learning procedure has been examined for automatic segmentation of 3D tomography images from fiber-reinforced ceramic composites consisting of fibers and matrix of the same material (SiC), and thus identical image intensities. The analysis uses a neural network to distinguish phases from shape and edge information rather than intensity differences. It was used successfully to segment phases in a unidirectional composite that also had a coating with similar image intensity. It was also used to segment matrix cracks generated during in situ tensile loading of the composite and thereby demonstrate the influence of nonuniform fiber distribution on the nature of matrix cracking. By avoiding the need for manual segmentation of thousands of image slices, the procedure overcomes a major impediment to the extraction of quantitative information from such images. The analysis was performed using recently developed software that provides a general framework for executing both training and inference.
Ó
Springer Science+Business
Media, LLC, part of Springer Nature 2020
Handling Editor: Avinash Dongare.
Address correspondence to E-mail: [email protected]
https://doi.org/10.1007/s10853-020-05148-7
J Mater Sci
GRAPHIC ABSTRACT
Introduction X-ray micro-CT imaging has become a valuable tool for analysis of structural materials, both for visualizing complex 3D microstructures and for imaging internal defects and damage introduced during manufacture or service [1–10]. In the case of fiber-reinforced composites (polymer or ceramic matrices), useful mechanical properties are achieved by arranging high strength fibers in sometimes complex 3D arrangements and surrounding them with a matrix [11]. The performance and reliability of these composites are dependent on the internal fiber architecture and the nature of
damage caused by external loads. During the past 10 years, synchrotron micro-CT imaging has been used to measure internal fiber architectures, including statistical deviations from the ideal (or intended) architecture, and provide the input needed to build numerical models for life prediction in both ceramic and polymer matrix composites [1, 4–7]. It has also been used with in situ mechanical loading and/or heating to detect and image the initiation and development of internal damage [2–6, 12], a critical step in guiding and validating life prediction and processing models. In most studies, the identification and representation of image features of intere
Data Loading...
