2016-Sustainable Industrial Processing Summit
SIPS 2016 Volume 1: D'Abreu Intl. Symp. / Iron and Steel Making
| Editors: | Kongoli F, Noldin JH, Takano C, Lins F, Gomez Marroquin MC, Contrucci M |
| Publisher: | Flogen Star OUTREACH |
| Publication Year: | 2016 |
| Pages: | 320 pages |
| ISBN: | 978-1-987820-37-9 |
| ISSN: | 2291-1227 (Metals and Materials Processing in a Clean Environment Series) |
CD shopping page
Pellet Porosimetry by X-Ray MicroCT - An Environment Friendly Approach
Karen Augusto1; Sidnei Paciornik1; Otavio Gomes2; Marcos Mauricio1;1DEQM PUC-RIO, Rio de Janeiro, Brazil; 2CETEM/MCTI, Rio de Janeiro, Brazil;
Type of Paper: Regular
Id Paper: 295
Topic: 2
Abstract:
One of the main steps in the characterization of iron ore pellets is the quantification of porosity. Traditionally it is measured by mercury intrusion porosimetry (porHg) - that only reveals open pores - and/or optical microscopy (OM) - that only provides 2D information. Both processes are destructive and require expensive and/or toxic consumables, with direct impact on the environment after disposal.
X-ray microtomography (microCT) is a promising alternative. It is non-destructive, requires no specimen preparation, can analyze full pellets and provides 3D information of total, open and closed porosity. The main drawbacks of the technique are its limited resolution (approx. 2 to 4 µm when scanning a full 12 mm diameter pellet) and the acquisition and processing time (approximately several hours per pellet).
This paper describes the development of the 3D methodology to measure the porosity of iron ore samples. It involves optimization of image acquisition parameters, a sequence of automatic 3D image processing and analysis steps, 3D visualization, and comparison with porHg and OM.
Employing a Carl Zeiss Versa 510 x-ray microscope, pellets were scanned with different resolutions to evaluate the impact on the measured pore distribution. Noise filtering, edge enhancement and careful manual thresholding were used to segment pores. Initial tests with 8 µm resolution, which allows visualization of a full pellet, showed that the main porosity peak revealed by porHg (approx. at 10 µm) is not detected. The second set of experiments exploiting the geometric flexibility of source-sample-detector setup reached 4 µm resolution, albeit with the loss of a small peripheral layer of the pellets. In this case, the main porosity peak is detected. The numerical results, however, are still very different from porHg.
To compare with OM, after microCT the pellet was cut in a plane orthogonal to the microCT rotation axis, then mounted and polished. Computer controlled OM was used to create a mosaic image covering the full cross section with 0.53 µm resolution. The acquired image was used in an automatic correlation procedure to detect the closest microCT layer image. Due to specimen preparation errors, the OM image is not perfectly parallel to this closest MicroCT layer. Thus, the OM image was automatically registered to the microCT image. Finally, the porosity values were compared. As expected, the porosity measured by OM was much larger than the value obtained from microCT.
Each of the techniques has its own limitations, with microCT providing true 3D information in a non-destructive sequence that requires no specimen preparation. Further quantitative comparison is expected to provide a road map to correlate the values obtained from the different techniques. In principle, this can allow microCT to become the main technique for measuring porosity in iron ore pellets.
Practical application of this methodology will also require a substantial reduction in microCT scanning and processing time. New methods such as DART – Discrete Algebraic Reconstruction Tomography – are currently under development, in order to reduce the required number of image projections while preserving the pore segmentation quality.