Deep Learning Based Hybrid Modelling of Poromechanics and Fluid Dynamics

Dear Colleagues,

Background:

The interplay between poromechanics and fluid dynamics is crucial for understanding and predicting the behavior of fluid-saturated porous materials. These phenomena are critical in various fields, including geomechanics, petroleum engineering, and biomechanics. Traditional modeling approaches, while effective, often fall short in capturing the complex, multiscale interactions inherent in these systems, leading to challenges in efficiency and accuracy.

Aim and scope:

This Special Issue aims to explore the integration of deep learning techniques with traditional modelling in hybrid frameworks to advance the understanding of poromechanics and fluid dynamics. By leveraging the strengths of data-driven methods alongside established physical models, the contributions seek to enhance predictive capabilities and provide new insights into the behavior of coupled fluid–structure systems.

History:

Hybrid modeling approaches, which combine empirical data with theoretical models, have a long history in the study of poromechanics and fluid dynamics. However, the advent of deep learning has opened new avenues for more sophisticated and accurate models. Such approaches like neural operators and physics-informed neural networks (PINNs) are advanced techniques used for solving partial differential equations (PDEs). The main idea of these approaches is to combine deep knowledge of physical laws (described by PDEs) and the learning capabilities of neural networks. Though they differ fundamentally in their concepts, applications, and underlying methodologies, these methods of hybrid modeling have their unique strengths and should be chosen based on the specific requirements and constraints of the PDE problem at hand. This Issue is focusing on how deep learning can be harnessed to overcome existing limitations and push the boundaries of current methodologies.

What kind of papers we are soliciting:

We are soliciting papers that present innovative applications of deep learning in hybrid modeling frameworks. This includes, but is not limited to, novel algorithmic approaches, advancements in computational efficiency, and the integration of deep learning with traditional numerical methods. Contributions that provide new theoretical insights, develop practical applications, or showcase interdisciplinary combinations are especially encouraged.

We invite original research articles, reviews, and case studies including, but not limited, to the following specific topics:

• Deep learning models for (coupled) poromechanical and fluid dynamics simulations.
• Hybrid modeling frameworks integrating deep learning and physical laws.
• Applications of AI in geosciences, petroleum engineering, and related fields.
• Data-driven approaches for multiscale and multiphysics problems in poromechanics and fluid dynamics.
• Theoretical advancements in hybrid modeling using deep learning, especially in reservoir simulations.

Dr. Leonid B. Sheremetov
Dr. Caterina Millevoi
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Computation is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1800 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• coupled hydro-poromechanics
• hybrid modeling
• deep learning
• data-driven scientific computing
• fluid flow modeling
• neural operator
• partial differential equations
• physics-informed neural networks
• real-life applications
• reservoir simulation