A re‐examination of throats

Abstract

[1] We critically re-examine the concept of a throat in a porous medium as a geometric quantity defined independently of an entry meniscus in a drainage process. To maintain the standard notion of a throat as a locally minimum-area cross section in the pore network, we demonstrate with examples that throats must intersect each other. Using flow simulation, we show that these intersecting throats correspond to capillary pressure controlled entry points during drainage. We have designed a throat-finding algorithm that explicitly locates intersecting throats, using a planar approximation for robustness and speed. The capability of the new algorithm was compared against an existing algorithm in the construction of pore-throat networks from X-ray computed tomography images of consolidated sandstones (7.5-22% porosity) and of an unconsolidated sand pack (32.5% porosity). We show that the probability of throat intersection increases significantly with porosity above 20%; in the sand pack, over 1/4 of all throats intersect with another.

Versatile and efficient pore network extraction method using marker-based watershed segmentation

“…The present work also outlined how to convert this segmented image into a pore network by extracting the pertinent information such as connectivity, pore sizes, and throat sizes. This process of calculating network properties from the watershed image was by no means the final word on the matter [47,48,63], and many improvements could be made such as finding surface area [64] and perimeter more rigorously. The networks extracted using the SNOW algorithm were compared to several known materials (see Table II).…”

“…GETNET.py in the Supplemental Material [37] is PYTHON code that performs these steps and outputs the data in a format suitable for importing into OPENPNM [46]. It should be stressed that the following interpretation of pore and throat size information represents only the most general approach, and much more sophisticated analyses could be brought to bear [47,48].…”

Machine learning framework for analysis of transport through complex networks in porous, granular media: A focus on permeability

“…To construct the pore network, nodes are assigned to pores, connected by an edge if the corresponding pores share a throat. What constitutes a 'pore' and 'throat' remains ambiguous [83]. We opt to use the modified Delaunay tessellation approach by Al-Raoush and Willson [64], with several adaptations [84].…”

Throat Finding Algorithms based on Throat Types

“…For cubic samples (512*512*512 voxel 3 ), the total CPU time consumed for the throat finding algorithms and probability density function of throat area, pore volume, and coordination number is 4 to 10 hours. Our new algorithms include a crossed-throat algorithm [4]. The Pdf of throat area, pore volume, and coordination number is shown in Figure 13.…”

“…For non-crossed throats, their outer perimeter voxels have to exist on the boundary grain voxels. The 3DMA-Rock software package [2] has three major throat-finding algorithms; (1) the wedgebased algorithm [3],(2) the Dijkstra-based shortest length algorithm [1], and (3) the planar dilation algorithm [4]. The wedge-based algorithm yields acceptable measures in only low porosity samples.…”