Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2609.39041

The Ganglion Network Model: Evolving Trapped Phases in Porous Media

Yashar Mehmani
View original source ↗Read source document ↗Export metadata ↓

Abstract

Partially miscible ganglia trapped within porous media, spanning one or multiple pores and evolving through diffusive mass transfer, are common in subsurface (e.g., CO$_2$ and H$_2$ storage) and manufacturing (e.g., fuel cells) applications. We present the ganglion network model (GNM), a reduced-order method for simulating how a population of such ganglia evolves inside an arbitrary porous microstructure. GNM operates on a tree graph, the ganglion network, extracted from the pore-scale image of a given sample. Each point on the graph encodes a possible ganglion configuration in the void space, without any loss of geometric or topological complexity. The evolution of a population is modeled by representing each ganglion as a particle on the graph and tracking it according to a set of rules formulated herein. The rules capture capillary events such as pore invasion, retraction, snap-off, fragmentation, and merger. Unlike pore-network models, another graph-based modeling tool at the pore scale, GNM solves no system of equations and its cost scales with ganglion count, not domain size. We validate GNM against an image-based pore-network model in 2.5D and 3D domains with populations undergoing ripening, dissolution, and growth. We find good agreement in ganglion statistics, aggregate properties, and spatial configuration. We further argue that the ganglion network is the statistical space needed for extending kinetic theories of Ostwald ripening from single- to multi-pore ganglia, and provide an outline for how to do this. GNM opens the door to modeling other dynamics of trapped phases in porous media.

Explore related subjects

physics.comp-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Yashar Mehmani. 2026-09-30. The Ganglion Network Model: Evolving Trapped Phases in Porous Media. https://arxiv.org/abs/2609.39041

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

T-matrix scattering formalism for electron-beam spectroscopy

Advanced computational tools that describe the interaction of electrons with structured nanophotonic devices are crucial for theoretical predictions, specific design tasks, and the interpretation of experimental results. These tools open the door to a systematic exploration of light-matter interactions in complex photonic environments and support the development of next-generation nanophotonic platforms. Here, we report on the implementation of electron-beam spectroscopy in a T-matrix-based scattering formulation. Such a framework is quite versatile in predicting the electromagnetic response of complex photonic materials composed of periodically or aperiodically arranged individual scatterers. By extending this formalism to describe interactions with fast electrons, we provide a fast and accurate numerical tool for simulating cathodoluminescence (CL) and electron energy-loss spectroscopy (EELS) measurements. The desired functionalities are implemented in treams_ebeam, an extension of the existing software suite treams for electromagnetic scattering computations, available at https://github.com/tfp-photonics/treams_ebeam. We demonstrate the implementation details on a carefully selected set of problems, including single scatterers of various shapes and materials, a periodic chain of elliptical nanodisks, and a finite cluster of nanospheres arranged in a two-dimensional (2D) lattice. By uniting fast-electron physics with advanced scattering theory, our framework unlocks new possibilities for designing, understanding, and engineering next-generation nanoscale light-matter interactions.

physics.comp-ph↗
arXiv2026

Machine-learning-driven kinetic discovery of carbon interstitial color centers in diamond

Diamond hosts optically active point defects central to quantum technologies, yet the carbon self-interstitials introduced during growth and irradiation compete with them and form new defects whose configurational landscape is poorly charted, as subtle energy differences govern the competing minima and pathways. Here we build an interstitial-focused dataset by active learning and benchmark three machine-learning interatomic potentials -- GAP, NEP and the equivariant MACE -- against density functional theory for energies, forces and migration barriers. MACE reproduces the reference energetics and relative stabilities, whereas the others can misorder the ground states. Annealing molecular dynamics with the validated potentials uncovers a series of previously unreported carbon interstitial clusters, from di- to octa-interstitials -- several introducing in-gap states of interest as colour centres -- and shows that their metastability is governed by kinetically accessible pathways rather than energetic ordering. These results chart the interstitial defect landscape and accelerate defect discovery for quantum technologies.

physics.comp-ph↗
arXiv2026

Atomic Design Transformer: Scaffold-Conditioned 3D Molecule Generation with xTB-Verified Reinforcement Learning

We present an autoregressive 3D-molecule generator with SE(3)-invariant tokenization, the Atomic Design Transformer (ADT). ADT places atoms one at a time, autoregressively. SE(3) invariance is achieved by tokenization: each new atom's position is encoded in the local coordinate frame of a previously placed atom. The backbone is a plain causal transformer. The token stream fully specifies a 3D structure together with its chemical-bond graph G, without any bond-order assignment. The model emits heavy-atom skeletons; hydrogens are added by separate learned models before the xTB relaxation. To score generated molecules we introduce the xTB topology-preservation rate (XTP): the fraction of molecules for which an xTB GFN2 relaxation preserves G specified by the token stream. For XTP-accepted molecules we also report the relaxation energy and the root mean square of the atomic displacement (RMSD). We evaluate two ADT models. The first is ADT pretrained on the GEOM-Drugs $\le\!30$-heavy-atom dataset; we benchmark scaffold-conditioned 3D generation across seven drug-like scaffolds from the model. It reaches an XTP of ${\sim}55\%$ and a valid-molecule yield $N^{\mathrm{gen}}/N$ of ${\sim}53\%$, where $N^{\mathrm{gen}}/N$ is the fraction of samples that are distinct, topology-preserving, and RDKit-readable. The second model continues from the first by reinforcement learning against the verifiable xTB reward (RLVR), using no external molecules. RLVR raises XTP to ${\sim}98\%$ and $N^{\mathrm{gen}}/N$ to ${\sim}95\%$, while approximately preserving the GEOM-Drugs size and composition distributions. Finally, we present an Inverse-Kinematics Transformer that recovers XTP for large molecules, where discretization error accumulates. ADT thus enables direct 3D generation.

physics.comp-ph↗