Characterization and comparison of pore distribution in weathered volcanic rocks by different techniques - Volcanic Rock Mechanics - page 78

characteristics can be obtained from a set of
contiguous two dimensional X-ray computerized
tomography images (cT) at different resolutions
(5 to 650 µm). X-ray cT reconstructs internal
images based on the distribution of the X-ray lin-
ear absorption coefficient deduced from the pro-
jection of X-rays through a sample (ohtani et al.,
2000). X-ray cT images have been obtained by
means of a Ge D-600 medical cT hybrid scanner
and a BiR actis 130/150 Micro cT/DR system.
The advantages of medical scanner include high
image acquisition velocity and configuration ver-
satility. on the other side, a Micro cT/DR system
allows the acquisition of images at higher resolu-
tion (40-60 µm) for small samples (<2 cm). image
processing, data extraction and data analysis were
elaborated by a 3D visualization software (avizo
6) ( Figure 3 ) . The results of image processing are:
a geometrical, morphological and topological
description of the features inside the investigated
volume, the identification and isolation of elements
of interest (e.g. porosity, minerals or particles and
voids distribution). in any case, the quality of final
results depends on a series of noise reduction, fil-
tering, thresholding, and particle separation steps
(Gualda and Rivers, 2006; ketcham, 2005).

rapid and precise method to obtain pore area, vol-
ume, shape, frequency, and spatial distribution.
Results from X-ray tomography and pycnom-
eter tests reveal that total porosity increase pro-
gressively with weathering grade with some minor
changes for the fifth alteration grade.
Differences in the results from Micro-cT and
Medical cT can result from the different resolu-
tion and from the consequent averaging. in par-
ticular, results reported in Figure 5 are obtained
starting from an evaluated density values from
which total porosity is computed by knowing the
specific weight of the solid phase.
This could also depend on the required manual
thresholding process and image sharpening, which

4 ResUlTs

The adopted classification scheme, based on min-
eralogical and structural characteristics, allowed a
preliminary classification of the collected samples
in base of their grade of alteration from fresh to
slightly, moderately and highly altered lava (sl1,
sl2, sl3, sl4, sl5, respectively).
effective porosity was obtained from bulk-
specific weight measurements, water immersion and
hg-porosimetry ( Table 1 and Figure 4 ) . according
to porosimetry results, a direct relationship between
the effective porosity and alteration/weathering
degree of samples exists. on the other hand, effective
porosity from bulk-specific weight measurements
seems to have no clear relationship with weathering
grade. The reason could be the percentage and sizes
of interconnected pores and fractures contained in
each samples. it means that porous system connec-
tivity do not increase progressively with weather-
ing degrees. on the contrary, a strong connectivity
increase is found between second and third weath-
ering grade. The decrease in the computed porosity
by imbibition could also be the result of a decrease
in size of the pores with the increased alteration and
then with a consequent difficulty in their saturation
under low air vacuum conditions.
Total porosity was obtained from thin sections,
X-ray tomographic image analysis and pycnometer
test method ( Figure 5 ) . image analysis represents a

Figure 4. effective porosity obtained from bulk specific
weight measurements, water immersion and mercury
porosimeter.

Figure 5. Total porosity obtained from pycnometer test,
thin section analysis and X-ray tomography by Medical
scanner and Micro cT. Porosity from Medical scanner
images has been obtained from bulk density estimates
and specific weight (assumed constant value: 2.78).