Closed alexeytopolnitskiy closed 3 years ago
You need to do network extraction - there are a few examples https://github.com/PMEAL/porespy/blob/dev/examples/networks/snow_basic.ipynb https://github.com/PMEAL/porespy/blob/dev/examples/networks/snow_advanced.ipynb
You can then also do simulations using our OpenPNM software https://github.com/PMEAL/OpenPNM
As @TomTranter says, the definition of a pore only applies if you do a pore network extraction to identify pores and throats. But pore networks have some assumptions built in. The purest estimation of the size distribution is the local thickness filter, binned as desired. It's a statistical value after all.
@jgostick @TomTranter Thank you for your replies! Now it is a bit understandable for me. If I got it right there are two ways of obtaining desired pore size distribution:
op.network.GenericNetwork()
, extract geometry with op.geometry.GenericGeometry()
and then, using geo["geometry"]
take values of pore diameters;ps.filters.local_thickness()
and analyze histogram of voxel values.First approach gives me close to something that I need, but second returns really strange distribution. Am I getting it right or you meant something else?
Sorry for the late reply. If you want to do network extraction to get the 'stick and ball' representation, follow these examples
If you want to get the statistical info about pores sizes, use local thickness as described here
Hello! I have a 3D model of a porous media. How can I estimate number of pores in this cube using PoresPy? And also, is it possible to obtain pore size distribution where each pore and its diameter are included (not like
porespy.metrics.pore_size_distribution
does, because it depends on number of bins). I mean for all possible pores and its diameter (smth like GeoDict does). Thanks!