Image data analysis of high resolution μCT data for the characterization of pore orientation and pore space interconnectivity in freeze cast ceramics

Abstract Unidirectional and non-unidirectional freeze cast SiOC monoliths were characterized by means of digital image processing. For this purpose a novel method of pore segmentation in complex dendritic pore morphologies was developed. For the first time, the proposed method made it possible to perform an extensive geometrical and structural pore space characterization and analysis on single pore level. The orientation of the primary dendritic channels was obtained from high resolution μCT images. Based on the mean directional vector distribution, the pore space was segmented. As a result, 95% (unidirectional) and 88% (non-unidirectional) of the original binary pore volume could be allocated to individual pores. The pore interconnectivity could be derived from this segmented pore space. It has been shown that the degree of interconnectivity between pores of the same orientation is higher than between pores of different orientation. This could be the reason for peculiar wicking behavior in non-unidirectional samples.