InterPore2026

Network-based modeling of fluid flow through membranes

21 May 2026, 09:05

15m

Oral Presentation (MS07) Mathematical and numerical methods for multi-scale multi-physics, nonlinear coupled processes MS07

Speaker

Lou Kondic (NJIT)

Description

We model a porous medium as a random pore network and focus on how the medium’s internal structure affects its flow and adsorptive behavior (see the figure for an example of considered membrane structure). A particular emphasis is on modeling suspension flow, where particles adsorb onto the pore walls. We start by formulating the governing equations of fluid flow on a general network. Then, we model adsorption by applying an advection equation with a sink term in each pore and examine how network parameters influence flow and transport; see [1-3] for some of our recent work.

The presentation will focus on linking the medium's topology (pore network) to its flow properties. The challenging aspect of understanding and quantifying the evolving pore network structure is addressed by using topological methods that provide simplified network descriptions, both of the networks’ initial properties and their time evolution. For this purpose, we use tools based on persistent homology. These tools enable us to connect structure, transport, and adsorption as key steps toward designing membranes with desired properties. Most of the material presented is in [4].

The final section of the presentation will focus on new results related to evolving networks and the tools used to measure network development. Specifically, we demonstrate that the measures quantifying topological changes can clearly differentiate between different filtration regimes and help improve understanding of the factors influencing filtration performance.

Acknowledgements: This work was supported by the NSF Grants DMS-2201627 and DMR-2410985, as well as by the NJIT Artificial Intelligence Grace Hopper Institute grants.