Physics-informed neural networks for heterogeneous poroelastic media

Roy, Sumanta; Annavarapu, Chandrasekhar; Roy, Pratanu; Valiveti, Dakshina M.

doi:10.1080/15502287.2024.2440420

Physics-informed neural networks for heterogeneous poroelastic media

Journal Article · Thu Dec 26 00:00:00 EST 2024 · International Journal for Computational Methods in Engineering Science and Mechanics

DOI:https://doi.org/10.1080/15502287.2024.2440420· OSTI ID:2556941

Roy, Sumanta ^[1]; Annavarapu, Chandrasekhar ^[1]; Roy, Pratanu ^[2]; Valiveti, Dakshina M. ^[3]

Indian Inst. of Technology (IIT), Madras (India)
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
ExxonMobil Technology and Engineering, Spring, TX (United States)

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. Further, 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.

View Accepted Manuscript (DOE)

Research Organization:: Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)

Sponsoring Organization:: USDOE National Nuclear Security Administration (NNSA)

Grant/Contract Number:: AC52-07NA27344

OSTI ID:: 2556941

Report Number(s):: LLNL--JRNL-860569; 1091510

Journal Information:: International Journal for Computational Methods in Engineering Science and Mechanics, Journal Name: International Journal for Computational Methods in Engineering Science and Mechanics Journal Issue: 2 Vol. 26; ISSN 1550-2287

Publisher:: Taylor & FrancisCopyright Statement

Country of Publication:: United States

Language:: English

References (44)

Modeling of tensile strength of rocks materials based on support vector machines approaches Ceryan, Nurcihan; Okkan, Umut; Samui, Pijush International Journal for Numerical and Analytical Methods in Geomechanics, Vol. 37, Issue 16 https://doi.org/10.1002/nag.2154	journal	November 2012
Google Colaboratory Bisong, Ekaba Building Machine Learning and Deep Learning Models on Google Cloud Platform https://doi.org/10.1007/978-1-4842-4470-8_7	book	January 2019
Neural networks in computational mechanics Yagawa, G.; Okuda, H. Archives of Computational Methods in Engineering, Vol. 3, Issue 4 https://doi.org/10.1007/BF02818935	journal	December 1996
Physics-informed neural networks (PINNs) for fluid mechanics: a review Cai, Shengze; Mao, Zhiping; Wang, Zhicheng Acta Mechanica Sinica, Vol. 37, Issue 12 https://doi.org/10.1007/s10409-021-01148-1	journal	December 2021
Support vector machine for structural reliability analysis Li, Hong-shuang; Lü, Zhen-zhou; Yue, Zhu-feng Applied Mathematics and Mechanics, Vol. 27, Issue 10 https://doi.org/10.1007/s10483-006-1001-z	journal	October 2006
Numerical simulation and modeling of a poroelastic media for detection and discrimination of geo-fluids using finite difference method Adnan Ahmad, Qazi; Irfan Ehsan, Muhammad; Khan, Nasir Alexandria Engineering Journal, Vol. 61, Issue 5 https://doi.org/10.1016/j.aej.2021.08.064	journal	May 2022
An immersed interface method for a 1D poroelasticity problem with discontinuous coefficients Bean, Maranda; Yi, Son-Young Journal of Computational and Applied Mathematics, Vol. 272 https://doi.org/10.1016/j.cam.2014.05.009	journal	December 2014
Generalized Multiscale Finite Element Method for thermoporoelasticity problems in heterogeneous and fractured media Ammosov, Dmitry; Vasilyeva, Maria; Chung, Eric T. Journal of Computational and Applied Mathematics, Vol. 407 https://doi.org/10.1016/j.cam.2021.113995	journal	June 2022
The neural particle method – An updated Lagrangian physics informed neural network for computational fluid dynamics Wessels, Henning; Weißenfels, Christian; Wriggers, Peter Computer Methods in Applied Mechanics and Engineering, Vol. 368 https://doi.org/10.1016/j.cma.2020.113127	journal	August 2020
A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics Haghighat, Ehsan; Raissi, Maziar; Moure, Adrian Computer Methods in Applied Mechanics and Engineering, Vol. 379 https://doi.org/10.1016/j.cma.2021.113741	journal	June 2021
CAN-PINN: A fast physics-informed neural network based on coupled-automatic–numerical differentiation method Chiu, Pao-Hsiung; Wong, Jian Cheng; Ooi, Chinchun Computer Methods in Applied Mechanics and Engineering, Vol. 395 https://doi.org/10.1016/j.cma.2022.114909	journal	May 2022
Physics-informed neural network simulation of multiphase poroelasticity using stress-split sequential training Haghighat, Ehsan; Amini, Danial; Juanes, Ruben Computer Methods in Applied Mechanics and Engineering, Vol. 397 https://doi.org/10.1016/j.cma.2022.115141	journal	July 2022
Respecting causality for training physics-informed neural networks Wang, Sifan; Sankaran, Shyam; Perdikaris, Paris Computer Methods in Applied Mechanics and Engineering, Vol. 421 https://doi.org/10.1016/j.cma.2024.116813	journal	March 2024
Interface PINNs (I-PINNs): A physics-informed neural networks framework for interface problems Sarma, Antareep Kumar; Roy, Sumanta; Annavarapu, Chandrasekhar Computer Methods in Applied Mechanics and Engineering, Vol. 429 https://doi.org/10.1016/j.cma.2024.117135	journal	September 2024
Exact enforcement of temporal continuity in sequential physics-informed neural networks Roy, Pratanu; Castonguay, Stephen T. Computer Methods in Applied Mechanics and Engineering, Vol. 430 https://doi.org/10.1016/j.cma.2024.117197	journal	October 2024
RPINNs: Rectified-physics informed neural networks for solving stationary partial differential equations Peng, Pai; Pan, Jiangong; Xu, Hui Computers & Fluids, Vol. 245 https://doi.org/10.1016/j.compfluid.2022.105583	journal	September 2022
How to achieve optimal and sustainable use of the subsurface for Aquifer Thermal Energy Storage Bloemendal, Martin; Olsthoorn, Theo; Boons, Frank Energy Policy, Vol. 66 https://doi.org/10.1016/j.enpol.2013.11.034	journal	March 2014
Physics-informed neural networks for heat transfer prediction in two-phase flows Jalili, Darioush; Jang, Seohee; Jadidi, Mohammad International Journal of Heat and Mass Transfer, Vol. 221 https://doi.org/10.1016/j.ijheatmasstransfer.2023.125089	journal	April 2024
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations Raissi, M.; Perdikaris, P.; Karniadakis, G. E. Journal of Computational Physics, Vol. 378 https://doi.org/10.1016/j.jcp.2018.10.045	journal	February 2019
Multiscale finite volume method for finite-volume-based simulation of poroelasticity Sokolova, Irina; Bastisya, Muhammad Gusti; Hajibeygi, Hadi Journal of Computational Physics, Vol. 379 https://doi.org/10.1016/j.jcp.2018.11.039	journal	February 2019
A combined Lattice-Boltzmann-finite element approach to modeling unsaturated poroelastic behavior of heterogeneous media Adia, J. -L.; Yvonnet, J.; He, Q. -C. Journal of Computational Physics, Vol. 437 https://doi.org/10.1016/j.jcp.2021.110334	journal	July 2021
On Physics-Informed Neural Networks training for coupled hydro-poromechanical problems Millevoi, Caterina; Spiezia, Nicolò; Ferronato, Massimiliano Journal of Computational Physics, Vol. 516 https://doi.org/10.1016/j.jcp.2024.113299	journal	November 2024
Numerical methods for hydraulic fracture propagation: A review of recent trends Lecampion, Brice; Bunger, Andrew; Zhang, Xi Journal of Natural Gas Science and Engineering, Vol. 49 https://doi.org/10.1016/j.jngse.2017.10.012	journal	January 2018
Multi-domain physics-informed neural network for solving forward and inverse problems of steady-state heat conduction in multilayer media Zhang, Benrong; Wu, Guozheng; Gu, Yan Physics of Fluids, Vol. 34, Issue 11 https://doi.org/10.1063/5.0116038	journal	November 2022
Fuzzy Control for Nonlinear Systems via Neural-Network-Based Approach Hsiao, Feng-Hsiag; Chiang, Wei-Ling; Chen, Cheng-Wu International Journal for Computational Methods in Engineering Science and Mechanics, Vol. 6, Issue 3 https://doi.org/10.1080/15502280590923612	journal	July 2005
Artificial Neural Network (ANN) Model for Prediction of Mixing Behavior of Granular Flows Mujumdar, Ajit; Robi, P. S.; Malik, Moinuddin International Journal for Computational Methods in Engineering Science and Mechanics, Vol. 8, Issue 3 https://doi.org/10.1080/15502280701252495	journal	April 2007
A Mechanism Contributing to Subsidence Above Gas Reservoirs and its Application to a Case Study Menin, Andrea; Salomoni, Valentina A.; Santagiuliana, Raffaella International Journal for Computational Methods in Engineering Science and Mechanics, Vol. 9, Issue 5 https://doi.org/10.1080/15502280802225234	journal	July 2008
Performance Evaluation of 14 Neural Network Architectures Used for Predicting Heat Transfer Characteristics of Engine Oils Al-Ajmi, R. M.; Abou-Ziyan, H. Z.; Mahmoud, M. A. International Journal for Computational Methods in Engineering Science and Mechanics, Vol. 13, Issue 1 https://doi.org/10.1080/15502287.2011.636789	journal	January 2012
Machine learning and molecular dynamics based models to predict the temperature dependent elastic properties of silver nanowires Joshi, S. K.; Singh, Sanjeev K.; Dubey, Santosh International Journal for Computational Methods in Engineering Science and Mechanics, Vol. 24, Issue 5 https://doi.org/10.1080/15502287.2023.2186974	journal	March 2023
Neural-network approximation of piecewise continuous functions: application to friction compensation Selmic, R. R.; Lewis, F. L. IEEE Transactions on Neural Networks, Vol. 13, Issue 3 https://doi.org/10.1109/TNN.2002.1000141	journal	May 2002
Physics-Informed Neural Networks for Heat Transfer Problems Cai, Shengze; Wang, Zhicheng; Wang, Sifan Journal of Heat Transfer, Vol. 143, Issue 6 https://doi.org/10.1115/1.4050542	journal	April 2021
Analyses of internal structures and defects in materials using physics-informed neural networks Zhang, Enrui; Dao, Ming; Karniadakis, George Em Science Advances, Vol. 8, Issue 7 https://doi.org/10.1126/sciadv.abk0644	journal	February 2022
Physics-informed neural networks for inverse problems in nano-optics and metamaterials Chen, Yuyao; Lu, Lu; Karniadakis, George Em Optics Express, Vol. 28, Issue 8 https://doi.org/10.1364/OE.384875	journal	January 2020
A Coupled Reservoir and Geomechanical Simulation System Settari, Antonin; Mourits, F. M. SPE Journal, Vol. 3, Issue 03 https://doi.org/10.2118/50939-PA	journal	September 1998
Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations Karniadakis, Ameya D. Jagtap & George Em Communications in Computational Physics, Vol. 28, Issue 5 https://doi.org/10.4208/cicp.OA-2020-0164	journal	June 2020
Adam: A Method for Stochastic Optimization Kingma, Diederik P.; Ba, Jimmy arXiv https://doi.org/10.48550/arXiv.1412.6980	preprint	January 2014
Physics-informed deep learning for flow and deformation in poroelastic media Bekele, Yared W. arXiv https://doi.org/10.48550/arXiv.2010.15426	preprint	January 2020
A Novel Adaptive Causal Sampling Method for Physics-Informed Neural Networks Guo, Jia; Wang, Haifeng; Hou, Chenping arXiv https://doi.org/10.48550/arXiv.2210.12914	preprint	January 2022
Enforcing Dirichlet boundary conditions in physics-informed neural networks and variational physics-informed neural networks Berrone, S.; Canuto, C.; Pintore, M. arXiv https://doi.org/10.48550/arXiv.2210.14795	text	January 2022
Physical informed neural networks with soft and hard boundary constraints for solving advection-diffusion equations using Fourier expansions Li, Xi'an; Deng, Jiaxin; Wu, Jinran arXiv https://doi.org/10.48550/arXiv.2306.12749	preprint	January 2023
An Expert's Guide to Training Physics-informed Neural Networks Wang, Sifan; Sankaran, Shyam; Wang, Hanwen arXiv https://doi.org/10.48550/arXiv.2308.08468	preprint	January 2023
Physics-informed neural networks with curriculum training for poroelastic flow and deformation processes Bekele, Yared W. arXiv https://doi.org/10.48550/arXiv.2404.13909	preprint	January 2024
Adaptive Interface-PINNs (AdaI-PINNs): An Efficient Physics-informed Neural Networks Framework for Interface Problems Roy, Sumanta; Annavarapu, Chandrasekhar; Roy, Pratanu arXiv https://doi.org/10.48550/arXiv.2406.04626	preprint	January 2024
Variational Interface Physics Informed Neural Networks (VI-PINNs) for Heterogeneous Subsurface Systems Sarma, A. K.; Annavarapu, C.; Roy, P. 57th U.S. Rock Mechanics/Geomechanics Symposium https://doi.org/10.56952/ARMA-2023-0760	conference	June 2023

Similar Records

Adaptive Interface-PINNs (AdaI-PINNs) for inverse problems: Determining material properties for heterogeneous systems

Journal Article · Wed May 28 20:00:00 EDT 2025 · Finite Elements in Analysis and Design · OSTI ID:2571677

Interface PINNs (I-PINNs): A physics-informed neural networks framework for interface problems

Journal Article · Wed Jun 19 20:00:00 EDT 2024 · Computer Methods in Applied Mechanics and Engineering · OSTI ID:2425908

Adaptive Interface-PINNs (AdaI-PINNs): An Efficient Physics-Informed Neural Networks Framework for Interface Problems

Journal Article · Fri Feb 28 19:00:00 EST 2025 · Communications in Computational Physics · OSTI ID:2571702

Related Subjects

42 ENGINEERING
Energy - Geothermal
Energy
Geosciences
Energy - Petroleum
Physics-informed neural networks
coupled problems
deep neural network
physics-constrained machine learning
poromechanics
scientific machine learning

Physics-informed neural networks for heterogeneous poroelastic media

Citation Formats

References (44)

Similar Records

Related Subjects