S. Ganti, M. Velez, B. Geier, B. Hayes, B. Turner, E. Jenkins
January 28, 2017
Porosity is a typical defect in additively manufactured (AM) parts. Such defects limit the properties and performance of AM parts, and therefore need to be characterized accurately. Current methods for characterization of defects and microstructure rely on classical stereological methods that extrapolate information from two dimensional images. The automation of serial sectioning provides an opportunity to precisely and accurately quantify porosity in three dimensions in materials. In this work, we analyzed the porosity of an additively manufactured Ti 6Al 2Sn 4Zr 2Mo sample using Robo-Met.3D®, an automated serial sectioning system. Image processing for three dimensional reconstruction of the serial-sectioned two dimensional images was performed using open source image analysis software (Fiji/ImageJ, Dream.3D, Paraview). The results from this 3D serial sectioning analysis were then compared to classical 2D stereological methods (Saltykov stereological theory). We found that for this dataset, the classical 2D methods underestimated the porosity size and distributions of the larger pores; a critical attribute to fatigue behavior of the AM part. The results suggest that acquiring experimental data with equipment such as Robo-Met.3D® to measure the number and size of particles such as pores in a volume irrespective of knowing their shape is a better choice.