Granulometry
First, we compute pore sizes using granulometry, a purely geometrical algorithm. PoroDict Granulometry uses the following assertion when defining a pore diameter:
A voxel is part of a pore with a diameter equal or larger than D, if it is included in a sphere of diameter D, which is completely included in the pore space.
So, for the example above with a voxel length of 0.5 µm, if a sphere of diameter 1 µm can be found that contains a certain voxel, but no sphere of diameter 1.5 µm can be found that contains the voxel, this voxel is assigned a pore diameter of 1 µm. This means, that each voxel is linked to the diameter of the largest sphere that fits inside the pore and contains this voxel.
Granulometry seems to find three large pore bodies (green), but these belong to pores with a diameter larger or equal than 2 µm.
In the cumulative volume fraction histogram, observe that nearly 80% of the pore space is associated to a diameter larger than 2 µm. The small closed pore at the top has a diameter smaller than 2 µm
When looking at the pores with a diameter larger or equal than 4.5 µm, only one of the large pore bodies is detected. This causes the increase in the volume fraction at the far right of the graph. The two other large pore bodies add up to the peak, at just below 4 µm.
|
Note! Granulometry assigns small spheres even into pore space that human observers consider as pore throats. But these small spheres fit, of course, also in the pore bodies. Thus, Granulometry is useful to detect the larger pore bodies. Even closed pores can be identified with this method. |
In the image below, the pore space with a diameter larger than 1 µm has green color. A part of the closed pore is colored white, which means these voxels have a diameter smaller than 1 µm.
