@inproceedings{9a844a620dc343c289db7f8d9e6f3da0,
title = "Quantitative evaluation of induced porosity reference samples implemented by drill hole in CFRP using XCT and various segmentation parameters",
abstract = "Determination of porosity in carbon fibre reinforced polymers (CFRP) using X-ray computed tomography (XCT) is an important reference method for quality control in aerospace industry. The main objective of this work is to create reference samples of carbon fibre reinforced polymer (CFRP) and evaluate porosity quantitatively using X-ray computed tomography (XCT). Artificial porosity was induced in three nonporous CFRP laminates by multiple cylindrical holes drilled with a diameter of 200 and 300 μm. These plates were arranged in different permutations to obtain various combinations of a pre-determined void content. The combinations of CFRP plates were scanned using XCT for porosity determination. The diameters of the drilled cylinders implanted were measured with voxel size of 5 μm3 XCT scans with an accuracy of 295 ± 2.75 μm and 196 ± 2.75 μm to obtain reference porosity limits. The reference samples results show a void content ranging from 0.96 up to 4.81vol. %. Voxel variation scans (VVS) were performed for a reference sample with 4.52 ± 0.07 vol. % of porosity from (11 μm)3 to (120 μm)3 voxel sizes to obtain comparative results of various global thresholds as well as visualize acceptable air-material boundary segmentation. Voxel size (VS) of (11μm)3 was identified as an optimum resolution for these purposes and sample size. Five global threshold methods were investigated for segmentation and results were compared. A novel segmentation method was introduced from the adaptation of ISO50 threshold. The adapted method from ISO50 threshold led to results within the expected porosity range. It provided accurate porosity results for macro voids with smallest individual void volume above 30×106 μm3 in CFRP. We show that the selection of a global threshold at an optimum resolution is very important in determination of porosity quantitatively in CFRP material system. The repeatability of the experiment was performed by scanning a reference sample with 4.4 ± 0.07 vol. % of porosity at optimum resolution of (11 μm)3 voxel size to consider variations in sample preparation, measurements and evaluation uncertainties. The standard deviation in porosity volume evaluated with ISO50 adapted thresholds were within ± 0.03 vol. % showing reasonable reproducibility.",
keywords = "CFRP, Reference porosity, Segmentation parameters, X-ray computed tomography",
author = "Guruprasad Rao and Bernhard Plank and Christoph Heinzl and Johann Kastner",
note = "Publisher Copyright: {\textcopyright} 2016, European Conference on Composite Materials, ECCM. All rights reserved.; 17th European Conference on Composite Materials, ECCM 2016 ; Conference date: 26-06-2016 Through 30-06-2016",
year = "2016",
language = "English",
isbn = "978-3-00-053387-7",
series = "ECCM 2016 - Proceeding of the 17th European Conference on Composite Materials",
publisher = "European Conference on Composite Materials, ECCM",
pages = "1--11",
booktitle = "ECCM 2016 - Proceeding of the 17th European Conference on Composite Materials",
}