Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder in a fluid flow using SINDy

Joshua A. Foster; Jan Rik Decuyper; Tim De Troyer; Mark Runacres

Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder in a fluid flow using SINDy

Joshua A. Foster, Jan Rik Decuyper, Tim De Troyer, Mark Runacres

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › Research

263 Downloads (Pure)

Abstract

The Sparse Identification of Nonlinear Dynamics (SINDy) toolbox can be used to estimate a nonlinear model of dynamical systems. SINDy is a dictionary method that applies sparse regression to a library of candidate functions. The effectiveness of SINDy has been demonstrated in a variety of fields. One such application is the modelling of the wake of a submerged cylinder in a flow. This is regarded as a canonical system for fluid-structure interactions. The SINDy method was found to be successful in modelling the flow around a stationary cylinder. In this work, the technique is applied to the wake of a submerged cylinder undergoing an imposed periodic oscillation. The experiment therefore strongly relates to the case of vortex-induced vibrations (VIV). VIV is challenging to model given that it exhibits much richer nonlinear dynamics than the stationary case. The study is carried out on the vorticity field in the wake of the cylinder. This work demonstrates that SINDy is capable of capturing the observed nonlinear dynamics.

Original language	English
Title of host publication	Conference on Noise and Vibration Engineering ISMA 2022
Publisher	ISMA 2022
Pages	tba
Number of pages	11
Edition	41
Publication status	Accepted/In press - 12 Sep 2022

Keywords

Sparse Identification of Nonlinear Dynamics
Fluid dynamics
Data-driven modelling

Access to Document

Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder using SINDyAccepted author manuscript, 1.51 MBLicence: CC BY

1 Finished

SRP60: SRP-Groeifinanciering: A system identification framework for multi-fidelity modelling
De Troyer, T., Runacres, M., Blondeau, J., Bram, S., Bellemans, A. & Contino, F.
1/03/19 → 29/02/24
Project: Fundamental

1 Conference paper
1 Poster

Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder: Using Sparse Identification of Nonlinear Dynamics (SINDy)
Foster, J. A., Decuyper, J. R., De Troyer, T. & Runacres, M., 2022.
Research output: Unpublished contribution to conference › Poster

Open Access
File
Modelling Vortex-Induced Loads Using Recurrent Neural Networks
Foster, J. A., Decuyper, J. R., Runacres, M. & De Troyer, T., 25 Oct 2021, Modeling, Estimation and Control Conference MECC 2021. Wang, J., Fathy, H., Wang, Q. & Ren, B. (eds.). 20 ed. Austin, Texas, USA: IFAC - PapersOnLine, Vol. 54. p. 32-37 6 p. (IFAC Proceedings Volumes; vol. 54, no. 20).
Research output: Chapter in Book/Report/Conference proceeding › Conference paper › Research

Open Access
File
3 Citations (Scopus)

91 Downloads (Pure)

Cite this

@inproceedings{d371812f511b44308d0ffe5644ee9ecd,

title = "Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder in a fluid flow using SINDy",

abstract = "The Sparse Identification of Nonlinear Dynamics (SINDy) toolbox can be used to estimate a nonlinear model of dynamical systems. SINDy is a dictionary method that applies sparse regression to a library of candidate functions. The effectiveness of SINDy has been demonstrated in a variety of fields. One such application is the modelling of the wake of a submerged cylinder in a flow. This is regarded as a canonical system for fluid-structure interactions. The SINDy method was found to be successful in modelling the flow around a stationary cylinder. In this work, the technique is applied to the wake of a submerged cylinder undergoing an imposed periodic oscillation. The experiment therefore strongly relates to the case of vortex-induced vibrations (VIV). VIV is challenging to model given that it exhibits much richer nonlinear dynamics than the stationary case. The study is carried out on the vorticity field in the wake of the cylinder. This work demonstrates that SINDy is capable of capturing the observed nonlinear dynamics.",

keywords = "Sparse Identification of Nonlinear Dynamics, Fluid dynamics, Data-driven modelling",

author = "Foster, {Joshua A.} and Decuyper, {Jan Rik} and {De Troyer}, Tim and Mark Runacres",

year = "2022",

month = sep,

day = "12",

language = "English",

pages = "tba",

booktitle = "Conference on Noise and Vibration Engineering ISMA 2022",

publisher = "ISMA 2022",

edition = "41",

}

Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder in a fluid flow using SINDy. / Foster, Joshua A.; Decuyper, Jan Rik ; De Troyer, Tim ; Runacres, Mark.

Conference on Noise and Vibration Engineering ISMA 2022. 41. ed. ISMA 2022, 2022. p. tba.

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › Research

TY - GEN

T1 - Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder in a fluid flow using SINDy

AU - Foster, Joshua A.

AU - Decuyper, Jan Rik

AU - De Troyer, Tim

AU - Runacres, Mark

PY - 2022/9/12

Y1 - 2022/9/12

N2 - The Sparse Identification of Nonlinear Dynamics (SINDy) toolbox can be used to estimate a nonlinear model of dynamical systems. SINDy is a dictionary method that applies sparse regression to a library of candidate functions. The effectiveness of SINDy has been demonstrated in a variety of fields. One such application is the modelling of the wake of a submerged cylinder in a flow. This is regarded as a canonical system for fluid-structure interactions. The SINDy method was found to be successful in modelling the flow around a stationary cylinder. In this work, the technique is applied to the wake of a submerged cylinder undergoing an imposed periodic oscillation. The experiment therefore strongly relates to the case of vortex-induced vibrations (VIV). VIV is challenging to model given that it exhibits much richer nonlinear dynamics than the stationary case. The study is carried out on the vorticity field in the wake of the cylinder. This work demonstrates that SINDy is capable of capturing the observed nonlinear dynamics.

AB - The Sparse Identification of Nonlinear Dynamics (SINDy) toolbox can be used to estimate a nonlinear model of dynamical systems. SINDy is a dictionary method that applies sparse regression to a library of candidate functions. The effectiveness of SINDy has been demonstrated in a variety of fields. One such application is the modelling of the wake of a submerged cylinder in a flow. This is regarded as a canonical system for fluid-structure interactions. The SINDy method was found to be successful in modelling the flow around a stationary cylinder. In this work, the technique is applied to the wake of a submerged cylinder undergoing an imposed periodic oscillation. The experiment therefore strongly relates to the case of vortex-induced vibrations (VIV). VIV is challenging to model given that it exhibits much richer nonlinear dynamics than the stationary case. The study is carried out on the vorticity field in the wake of the cylinder. This work demonstrates that SINDy is capable of capturing the observed nonlinear dynamics.

KW - Sparse Identification of Nonlinear Dynamics

KW - Fluid dynamics

KW - Data-driven modelling

M3 - Conference paper

SP - tba

BT - Conference on Noise and Vibration Engineering ISMA 2022

PB - ISMA 2022

ER -

Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder in a fluid flow using SINDy

Abstract

Keywords

Access to Document

Fingerprint

Projects

SRP60: SRP-Groeifinanciering: A system identification framework for multi-fidelity modelling

Research output

Estimating a sparse nonlinear dynamical model of the flow around an oscillating cylinder: Using Sparse Identification of Nonlinear Dynamics (SINDy)

Modelling Vortex-Induced Loads Using Recurrent Neural Networks

Cite this