Physics-informed neural networks for heterogeneous poroelastic media

This study presents a novel physics-informed neural network (PINN) framework for modeling poroelasticity in heterogeneous media with material interfaces. The approach introduces a composite neural network (CoNN) where separate neural networks predict displacement and pressure variables for each mate...

Full description

Saved in:
Bibliographic Details
Published in:International journal of computational methods in engineering science and mechanics Vol. 26; no. 2; pp. 187 - 207
Main Authors: Roy, Sumanta, Annavarapu, Chandrasekhar, Roy, Pratanu, Valiveti, Dakshina M.
Format: Journal Article
Language:English
Published: United States Taylor & Francis 04.03.2025
Subjects:
coupled problems
deep neural network
ENGINEERING
physics-constrained machine learning
Physics-informed neural networks
poromechanics
scientific machine learning
ISSN:1550-2287, 1550-2295
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study presents a novel physics-informed neural network (PINN) framework for modeling poroelasticity in heterogeneous media with material interfaces. The approach introduces a composite neural network (CoNN) where separate neural networks predict displacement and pressure variables for each material. While sharing identical activation functions, these networks are independently trained for all other parameters. To address challenges posed by heterogeneous material interfaces, the CoNN is integrated with the Interface-PINNs (I-PINNs) framework (Sarma et al., Comput. Methods Appl. Mech. Eng. 429: 117135, 2024), allowing different activation functions across material interfaces. This ensures accurate approximation of discontinuous solution fields and gradients. Performance and accuracy of this combined architecture were evaluated against the conventional PINNs approach, a single neural network (SNN) architecture, and the eXtended PINNs (XPINNs) framework through two one-dimensional benchmark examples with discontinuous material properties. The results show that the proposed CoNN with I-PINNs architecture achieves an RMSE that is two orders of magnitude better than the conventional PINNs approach and is at least 40 times faster than the SNN framework. Compared to XPINNs, the proposed method achieves an RMSE at least one order of magnitude better and is 40% faster.
Bibliography:AC52-07NA27344
USDOE National Nuclear Security Administration (NNSA)
LLNL--JRNL-860569
ISSN:1550-2287
1550-2295
DOI:10.1080/15502287.2024.2440420