CFRP porosity characterization using µ-Computed Tomography with optimized test parameters supported by XCT-simulation

Industrial X-ray Computed Tomography (XCT) techniques are used in a wide variety of different applica-tions. In recent years the inspection of Carbon Fiber Reinforced Polymers (CFRP) has also been added to these applications. Besides defect detection, an increasing interest is observed in characterizing the inner structure and in providing in-depth information for the design and manufacturing of new components. This paper presents the results of volume porosity characterization and parameter optimization obtained using different µ-XCT systems as well as CT-simulation. The scans were performed on the EADS Innova-tion Works’ µ-XCT system, which was especially built for CFRP characterization. It is an open system, which allows different kinds of acquisition, reconstruction, and data evaluation. In addition, scans were carried out at the University of Applied Sciences,Upper Austria, which operates a GE phoenix|x-ray Nanotom 180 device. The XCT scans were carried out on two specimens with the focus on quantitative characterization of porosity. In addition, the choice of optimized measurement and evaluation parameters is of particular importance, especially with varying dimensions and pore shape. Several studies of porosity determination [1, 2] have shown that the evaluated degree of porosity and its uncertainty are dependent on CT measurement parameters and evaluation. Due to the CFRP manufacturing process it is difficult to obtain samples with predefined degrees of porosity, as well as homogeneous pore distribution, which are needed for porosity studies of this kind. Thus, the XCT-simulation of synthetically-generated volume porosity datasets is a promising method for test parameter optimization. The correlation of XCT-simulation and XCT-measurement is mandatory in order to choose the best possible aquisition parameters and thus improve the precision of quantitative non-destructive testing of CFRP materials using µ-XCT.