MétaCan
Menu
Back to cohort
Record W4389584862 · doi:10.17118/11143/21035

Combining reduced-order modeling and PINNs to model fluid-structureinteractions

2023· article· en· W4389584862 on OpenAlex
Omar Tazi Labzour, Hamid Reza Karbasian, Sébastien Houde, Frédérick P. Gosselin

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

Venuenot available
Typearticle
Languageen
FieldPhysics and Astronomy
TopicModel Reduction and Neural Networks
Canadian institutionsUniversité LavalPolytechnique Montréal
Fundersnot available
KeywordsComputer scienceOrder (exchange)

Abstract

fetched live from OpenAlex

Abstract: Hydraulic turbines are subjected to Fluid-Structure Interactions (FSI) during operation, potentially leading to severe damage and shortening their life expectancy. This type of phenomenon that requires combining fluid and structure dynamics can be simulated using high-fidelity approaches, such as Computational Fluid Dynamics (CFD) coupled with a Finite Element (FE) analysis. The coupled CFD-FE tool for FSI problems usually is computationally expensive as it requires coupling governing equations of the flow field and structure. Moreover, this high-fidelity tool needs to generate dynamic mesh at each iteration. Therefore, it is essential to develop a cheaper model with acceptable accuracy to avoid the computational cost of such high-fidelity tools for FSI problems. Reduced-Order Models (ROMs) are approaches that have attracted much attention for engineering problems. In this study, a ROM is developed using Physics-Informed Neural Networks (PINNs) to model the flow around a bluff body. PINNs are new architectures in deep learning that approximate physical fields by training a neural network constrained by governing equations over data points distributed in the domain. We leverage Proper Orthogonal Decomposition (POD) on data from a precomputed CFD solution to decompose the temporal and spatial terms containing significant coherent structures in the flow fields. Furthermore, the spatial modes from the fluid flow are combined with the kinematic modes of the structural mechanics obtained for a moving solid in the domain. Therefore, the governing equation for the FSI problem can be represented as a set of Ordinary Differential Equations (ODE). PINNs are used to solve these ODEs for physical quantities, such as velocity, pressure, and solid displacement. The neural network is trained by minimizing the residuals of the governing equations, which are the incompressible Navier-Stokes equations. This work considers laminar flow around a 2D cylinder, a canonical example in fluid mechanics. The results show that the proposed approach can predict physical quantities with significantly lower computational costs in time with acceptable accuracy. This ROM approach could offer the possibility to leverage existing CFD data to create at reduced computational and implementation cost a way to predict the FSI response of a structure, say to vortex-induced vibrations with better accuracy than a one-way coupling simulation.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.356
Threshold uncertainty score0.424

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0000.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.033
GPT teacher head0.297
Teacher spread0.264 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Quick stats

Citations0
Published2023
Admission routes1
Has abstractyes

Explore more

Same topicModel Reduction and Neural NetworksFrench-language works237,207