Estimating permeability of 3D micro-CT images by physics-informed CNNs based on DNS
Gärttner S, Alpak FO, Meier A, Ray N, Frank F (2023)
Publication Language: English
Publication Status: Submitted
Publication Type: Journal article, Original article
Future Publication Type: Journal article
Publication year: 2023
Journal
URI: https://link.springer.com/article/10.1007/s10596-022-10184-0
DOI: 10.1007/s10596-022-10184-0
Open Access Link: https://link.springer.com/article/10.1007/s10596-022-10184-0
Abstract
In recent years, convolutional neural networks (CNNs) have experienced an increasing interest in their ability to perform a fast approximation of effective hydrodynamic parameters in porous media research and applications. This paper presents a novel methodology for permeability prediction from micro-CT scans of geological rock samples. The training data set for CNNs dedicated to permeability prediction consists of permeability labels that are typically generated by classical lattice Boltzmann methods (LBM) that simulate the flow through the pore space of the segmented image data. We instead perform direct numerical simulation (DNS) by solving the stationary Stokes equation in an efficient and distributed-parallel manner. As such, we circumvent the convergence issues of LBM that frequently are observed on complex pore geometries, and therefore, improve the generality and accuracy of our training data set. Using the DNS-computed permeabilities, a physics-informed CNN (PhyCNN) is trained by additionally providing a tailored characteristic quantity of the pore space. More precisely, by exploiting the connection to flow problems on a graph representation of the pore space, additional information about confined structures is provided to the network in terms of the maximum flow value, which is the key innovative component of our workflow. The robustness of this approach is reflected by very high prediction accuracy, which is observed for a variety of sandstone samples from archetypal rock formations.
Authors with CRIS profile
Stephan Gärttner
Andreas Meier
Graduiertenkolleg 2339/2 IntComSin: Grenzflächen, komplexe Strukturen und singuläre Grenzwerte in der Kontinuumsmechanik - Analysis und Numerik
Nadja Ray
Lehrstuhl für Angewandte Mathematik (Modellierung und Numerik)
Florian Frank
Professur für Wissenschaftliches Rechnen
Involved external institutions
Netherlands (NL)How to cite
APA:
Gärttner, S., Alpak, F.O., Meier, A., Ray, N., & Frank, F. (2023). Estimating permeability of 3D micro-CT images by physics-informed CNNs based on DNS. Computational Geosciences. https://doi.org/10.1007/s10596-022-10184-0
MLA:
Gärttner, Stephan, et al. "Estimating permeability of 3D micro-CT images by physics-informed CNNs based on DNS." Computational Geosciences (2023).
BibTeX: Download