Phase segmentation of uncured prepreg X-Ray CT micrographs

Pedro Galvez Hernandez; Karolina Gaska; James Kratz

doi:10.1016/j.compositesa.2021.106527

Phase segmentation of uncured prepreg X-Ray CT micrographs

Pedro Galvez Hernandez, Karolina Gaska, James Kratz^*

^*Corresponding author for this work

Research output: Contribution to journal › Article (Academic Journal) › peer-review

15 Downloads (Pure)

Abstract

This paper explores methods to investigate voids and dry fibre areas in pre-impregnated aligned carbon fibre reinforced epoxy lay-ups. A deep learning segmentation approach was compared to conventional thresholding techniques to characterise the interlaminar voids (entrapped air) and dry areas (unsaturated fibre bed) phases obtained by micro-CT scanning of samples from uncured laminates. The performance of both approaches was quantitatively assessed in three regions of interest having different levels of porosity, ranging from a low 1% to a high 25%. Deep learning consistently outperformed thresholding in the segmentation of both interlaminar voidage and dry areas. Furthermore, deep learning improved the ability to detect small voids and was able to accurately segment voids in volumes with less than 2% voidage, whereas thresholding techniques fail in this task. Finally, the application of deep learning to the segmentation of dry areas in micro-CT scans provided sharper results than thresholding, without needing filtering.

Original language	English
Article number	106527
Number of pages	14
Journal	Composites Part A: Applied Science and Manufacturing
Volume	149
Early online date	17 Jun 2021
DOIs	https://doi.org/10.1016/j.compositesa.2021.106527
Publication status	Published - 1 Oct 2021

Bibliographical note

Funding Information:
The authors would like to acknowledge the Engineering and Physical Sciences Research Council (EPSRC) for their support of this research through Investigation of Fine-Scale Flows in Composites Processing [ EP/S016996/1 ]. A PhD studentship for P. Galvez-Hernandez was supported through the Rolls-Royce Composites University Technology Centre at the University of Bristol.

Funding Information:
The authors would like to acknowledge the Engineering and Physical Sciences Research Council (EPSRC) for their support of this research through Investigation of Fine-Scale Flows in Composites Processing [EP/S016996/1]. A PhD studentship for P. Galvez-Hernandez was supported through the Rolls-Royce Composites University Technology Centre at the University of Bristol.

Publisher Copyright:
© 2021 The Authors

Keywords

A. Prepreg
B. Porosity
D. CT analysis
Machine learning

Access to Document

10.1016/j.compositesa.2021.106527

Full-text PDF (accepted author manuscript)
This is the accepted author manuscript (AAM). The final published version (version of record) is available online via Elsevier at https://doi.org/10.1016/j.compositesa.2021.106527. Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 2.06 MBLicence: CC BY-NC-ND

Handle.net

Persistent link

Investigation of fine-scale flows in composites processing
Rendall, T. C. S.
1/04/19 → 31/03/23
Project: Research

Cite this

@article{c8741a91c27e42e6b8a63dba05ca0552,

title = "Phase segmentation of uncured prepreg X-Ray CT micrographs",

abstract = "This paper explores methods to investigate voids and dry fibre areas in pre-impregnated aligned carbon fibre reinforced epoxy lay-ups. A deep learning segmentation approach was compared to conventional thresholding techniques to characterise the interlaminar voids (entrapped air) and dry areas (unsaturated fibre bed) phases obtained by micro-CT scanning of samples from uncured laminates. The performance of both approaches was quantitatively assessed in three regions of interest having different levels of porosity, ranging from a low 1% to a high 25%. Deep learning consistently outperformed thresholding in the segmentation of both interlaminar voidage and dry areas. Furthermore, deep learning improved the ability to detect small voids and was able to accurately segment voids in volumes with less than 2% voidage, whereas thresholding techniques fail in this task. Finally, the application of deep learning to the segmentation of dry areas in micro-CT scans provided sharper results than thresholding, without needing filtering. ",

keywords = "A. Prepreg, B. Porosity, D. CT analysis, Machine learning",

author = "{Galvez Hernandez}, Pedro and Karolina Gaska and James Kratz",

note = "Funding Information: The authors would like to acknowledge the Engineering and Physical Sciences Research Council (EPSRC) for their support of this research through Investigation of Fine-Scale Flows in Composites Processing [ EP/S016996/1 ]. A PhD studentship for P. Galvez-Hernandez was supported through the Rolls-Royce Composites University Technology Centre at the University of Bristol. Funding Information: The authors would like to acknowledge the Engineering and Physical Sciences Research Council (EPSRC) for their support of this research through Investigation of Fine-Scale Flows in Composites Processing [EP/S016996/1]. A PhD studentship for P. Galvez-Hernandez was supported through the Rolls-Royce Composites University Technology Centre at the University of Bristol. Publisher Copyright: {\textcopyright} 2021 The Authors",

year = "2021",

month = oct,

day = "1",

doi = "10.1016/j.compositesa.2021.106527",

language = "English",

volume = "149",

journal = "Composites Part A: Applied Science and Manufacturing",

issn = "1359-835X",

publisher = "Elsevier Science",

}

TY - JOUR

T1 - Phase segmentation of uncured prepreg X-Ray CT micrographs

AU - Galvez Hernandez, Pedro

AU - Gaska, Karolina

AU - Kratz, James

N1 - Funding Information: The authors would like to acknowledge the Engineering and Physical Sciences Research Council (EPSRC) for their support of this research through Investigation of Fine-Scale Flows in Composites Processing [ EP/S016996/1 ]. A PhD studentship for P. Galvez-Hernandez was supported through the Rolls-Royce Composites University Technology Centre at the University of Bristol. Funding Information: The authors would like to acknowledge the Engineering and Physical Sciences Research Council (EPSRC) for their support of this research through Investigation of Fine-Scale Flows in Composites Processing [EP/S016996/1]. A PhD studentship for P. Galvez-Hernandez was supported through the Rolls-Royce Composites University Technology Centre at the University of Bristol. Publisher Copyright: © 2021 The Authors

PY - 2021/10/1

Y1 - 2021/10/1

N2 - This paper explores methods to investigate voids and dry fibre areas in pre-impregnated aligned carbon fibre reinforced epoxy lay-ups. A deep learning segmentation approach was compared to conventional thresholding techniques to characterise the interlaminar voids (entrapped air) and dry areas (unsaturated fibre bed) phases obtained by micro-CT scanning of samples from uncured laminates. The performance of both approaches was quantitatively assessed in three regions of interest having different levels of porosity, ranging from a low 1% to a high 25%. Deep learning consistently outperformed thresholding in the segmentation of both interlaminar voidage and dry areas. Furthermore, deep learning improved the ability to detect small voids and was able to accurately segment voids in volumes with less than 2% voidage, whereas thresholding techniques fail in this task. Finally, the application of deep learning to the segmentation of dry areas in micro-CT scans provided sharper results than thresholding, without needing filtering.

AB - This paper explores methods to investigate voids and dry fibre areas in pre-impregnated aligned carbon fibre reinforced epoxy lay-ups. A deep learning segmentation approach was compared to conventional thresholding techniques to characterise the interlaminar voids (entrapped air) and dry areas (unsaturated fibre bed) phases obtained by micro-CT scanning of samples from uncured laminates. The performance of both approaches was quantitatively assessed in three regions of interest having different levels of porosity, ranging from a low 1% to a high 25%. Deep learning consistently outperformed thresholding in the segmentation of both interlaminar voidage and dry areas. Furthermore, deep learning improved the ability to detect small voids and was able to accurately segment voids in volumes with less than 2% voidage, whereas thresholding techniques fail in this task. Finally, the application of deep learning to the segmentation of dry areas in micro-CT scans provided sharper results than thresholding, without needing filtering.

KW - A. Prepreg

KW - B. Porosity

KW - D. CT analysis

KW - Machine learning

U2 - 10.1016/j.compositesa.2021.106527

DO - 10.1016/j.compositesa.2021.106527

M3 - Article (Academic Journal)

VL - 149

JO - Composites Part A: Applied Science and Manufacturing

JF - Composites Part A: Applied Science and Manufacturing

SN - 1359-835X

M1 - 106527

ER -

Phase segmentation of uncured prepreg X-Ray CT micrographs

Abstract

Bibliographical note

Keywords

Access to Document

Handle.net

Fingerprint

Projects

Investigation of fine-scale flows in composites processing

Cite this