OBTAINING PORE-SIZE DISTRIBUTION BY IMAGE ANALYSIS OF STACKS WITH LARGE IMAGE SPACING WITHOUT RESIZING

Resumo

The characterization of porous media is a crucial process to study the flow in porous
structures and the pore-size distribution (PSD) is a fundamental parameter to perform this
characterization. The PSD describes the porous volume fraction of each different pore-size. The PSD
is usually estimated by laboratorial tests, which have long durations and may be sample destructive.
Yang and collaborators developed an alternative capable to estimate the PSD of porous media by
image analysis, which has a fast result and preserves the sample. The method requires the digital
reconstruction of the porous medium by creating a stack of images that should be spaced by a length
equal to the pixel size used on these images. To overcome this problem, Yang suggests to resize the
image stack to match the image spacing to the pixel size, which reduces the image stack resolutions
and increase the errors. To avoid the resizing procedure and its consequences, this paper tested three
different approaches: the first one consisted in using the method proposed by Yang and collaborators
without resizing the images; the second approach consisted in approximating the pores to spheres and
analytically calculate their volumes; and the last approach was to approximate the pores to prolate
spheroids and analytically calculate their volumes. The results showed that the first and second
approaches were more efficient than the resizing alternative and that the third approach could be used
to improve the results from the other two approaches.