Super-resolution reconstruction of turbulent flows with machine learning

Kai Fukami, Koji Fukagata, Kunihiko Taira

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

We use machine learning to perform super-resolution analysis of grossly under-resolved turbulent flow field data to reconstruct the high-resolution flow field. Two machine learning models are developed, namely, the convolutional neural network (CNN) and the hybrid downsampled skip-connection/multi-scale (DSC/MS) models. These machine learning models are applied to a two-dimensional cylinder wake as a preliminary test and show remarkable ability to reconstruct laminar flow from low-resolution flow field data. We further assess the performance of these models for two-dimensional homogeneous turbulence. The CNN and DSC/MS models are found to reconstruct turbulent flows from extremely coarse flow field images with remarkable accuracy. For the turbulent flow problem, the machine-leaning-based super-resolution analysis can greatly enhance the spatial resolution with as little as 50 training snapshot data, holding great potential to reveal subgrid-scale physics of complex turbulent flows. With the growing availability of flow field data from high-fidelity simulations and experiments, the present approach motivates the development of effective super-resolution models for a variety of fluid flows.

Original languageEnglish
Pages (from-to)106-120
Number of pages15
JournalJournal of Fluid Mechanics
Volume870
DOIs
Publication statusPublished - 2019 Jul 10

Fingerprint

machine learning
turbulent flow
Turbulent flow
Learning systems
flow distribution
Flow fields
multiscale models
homogeneous turbulence
Neural networks
laminar flow
wakes
fluid flow
availability
Laminar flow
education
spatial resolution
Flow of fluids
Turbulence
Physics
physics

Keywords

  • Computational methods
  • Homogeneous turbulence
  • Wakes

ASJC Scopus subject areas

  • Condensed Matter Physics
  • Mechanics of Materials
  • Mechanical Engineering

Cite this

Super-resolution reconstruction of turbulent flows with machine learning. / Fukami, Kai; Fukagata, Koji; Taira, Kunihiko.

In: Journal of Fluid Mechanics, Vol. 870, 10.07.2019, p. 106-120.

Research output: Contribution to journalArticle

@article{5278a4a75be64d74a3a1e541babe70cc,
title = "Super-resolution reconstruction of turbulent flows with machine learning",
abstract = "We use machine learning to perform super-resolution analysis of grossly under-resolved turbulent flow field data to reconstruct the high-resolution flow field. Two machine learning models are developed, namely, the convolutional neural network (CNN) and the hybrid downsampled skip-connection/multi-scale (DSC/MS) models. These machine learning models are applied to a two-dimensional cylinder wake as a preliminary test and show remarkable ability to reconstruct laminar flow from low-resolution flow field data. We further assess the performance of these models for two-dimensional homogeneous turbulence. The CNN and DSC/MS models are found to reconstruct turbulent flows from extremely coarse flow field images with remarkable accuracy. For the turbulent flow problem, the machine-leaning-based super-resolution analysis can greatly enhance the spatial resolution with as little as 50 training snapshot data, holding great potential to reveal subgrid-scale physics of complex turbulent flows. With the growing availability of flow field data from high-fidelity simulations and experiments, the present approach motivates the development of effective super-resolution models for a variety of fluid flows.",
keywords = "Computational methods, Homogeneous turbulence, Wakes",
author = "Kai Fukami and Koji Fukagata and Kunihiko Taira",
year = "2019",
month = "7",
day = "10",
doi = "10.1017/jfm.2019.238",
language = "English",
volume = "870",
pages = "106--120",
journal = "Journal of Fluid Mechanics",
issn = "0022-1120",
publisher = "Cambridge University Press",

}

TY - JOUR

T1 - Super-resolution reconstruction of turbulent flows with machine learning

AU - Fukami, Kai

AU - Fukagata, Koji

AU - Taira, Kunihiko

PY - 2019/7/10

Y1 - 2019/7/10

N2 - We use machine learning to perform super-resolution analysis of grossly under-resolved turbulent flow field data to reconstruct the high-resolution flow field. Two machine learning models are developed, namely, the convolutional neural network (CNN) and the hybrid downsampled skip-connection/multi-scale (DSC/MS) models. These machine learning models are applied to a two-dimensional cylinder wake as a preliminary test and show remarkable ability to reconstruct laminar flow from low-resolution flow field data. We further assess the performance of these models for two-dimensional homogeneous turbulence. The CNN and DSC/MS models are found to reconstruct turbulent flows from extremely coarse flow field images with remarkable accuracy. For the turbulent flow problem, the machine-leaning-based super-resolution analysis can greatly enhance the spatial resolution with as little as 50 training snapshot data, holding great potential to reveal subgrid-scale physics of complex turbulent flows. With the growing availability of flow field data from high-fidelity simulations and experiments, the present approach motivates the development of effective super-resolution models for a variety of fluid flows.

AB - We use machine learning to perform super-resolution analysis of grossly under-resolved turbulent flow field data to reconstruct the high-resolution flow field. Two machine learning models are developed, namely, the convolutional neural network (CNN) and the hybrid downsampled skip-connection/multi-scale (DSC/MS) models. These machine learning models are applied to a two-dimensional cylinder wake as a preliminary test and show remarkable ability to reconstruct laminar flow from low-resolution flow field data. We further assess the performance of these models for two-dimensional homogeneous turbulence. The CNN and DSC/MS models are found to reconstruct turbulent flows from extremely coarse flow field images with remarkable accuracy. For the turbulent flow problem, the machine-leaning-based super-resolution analysis can greatly enhance the spatial resolution with as little as 50 training snapshot data, holding great potential to reveal subgrid-scale physics of complex turbulent flows. With the growing availability of flow field data from high-fidelity simulations and experiments, the present approach motivates the development of effective super-resolution models for a variety of fluid flows.

KW - Computational methods

KW - Homogeneous turbulence

KW - Wakes

UR - http://www.scopus.com/inward/record.url?scp=85065398025&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065398025&partnerID=8YFLogxK

U2 - 10.1017/jfm.2019.238

DO - 10.1017/jfm.2019.238

M3 - Article

AN - SCOPUS:85065398025

VL - 870

SP - 106

EP - 120

JO - Journal of Fluid Mechanics

JF - Journal of Fluid Mechanics

SN - 0022-1120

ER -