(188c) Modeling Interfacial Behavior in Complex Porous Materials | AIChE

(188c) Modeling Interfacial Behavior in Complex Porous Materials

Authors 

Thompson, K. E. - Presenter, Louisiana State University


The behavior of fluid-fluid interfaces in porous materials is of interest in many applications, including composite-materials manufacturing, oil and gas recovery, separations processes, coatings applications, environmental transport, and more. Modeling interfacial behavior in these systems is challenging because of the complicated, interconnected pore structure found in most real materials, which in turn leads to highly complex interfacial dynamics.

In the past, fundamental (microscale) simulation techniques have involved one of two approaches: 1. rigorous fluid-mechanics modeling of interfaces in single, idealized pores; 2. pore-scale modeling of interfacial flows in network models or similar structures. The first approach, although rigorous, is not readily applied to most real materials, especially over reasonably large characteristic lengths. The second approach is ideal for larger, heterogeneous structures. However, the interfacial geometries and fluid mechanics are necessarily oversimplified (e.g., equations based on triangular capillaries of constant cross section).

In this work, we present an intermediate technique that bridges the two approaches described above. A series of test interfaces are used to probe for the location (and interface stability) of wetting-phase fluids. This information is then translated into polynomial expressions that describe the capillary pressure versus phase saturation functionality within each individual pore in a material. This information can be used in two ways: for modeling fluid displacement or for predicting fluid configurations given a capillary pressure distribution. Results are presented for a computer-generated sphere packing, which is a common prototype for disordered particulate materials. These tests are performed on digital images of the sphere packing, which helps to emphasize how the algorithm will be generalized for use with tomography images of real materials.