Predicting Coherent Turbulent Structures via Deep Learning
Turbulent flow is widespread in many applications, such as airplane wings or turbine blades. Such flow is highly chaotic and impossible to predict far into the future. Some regions exhibit a coherent physical behavior in turbulent flow, satisfying specific properties; these regions are denoted as co...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-04-01
|
| Series: | Frontiers in Physics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fphy.2022.888832/full |
| _version_ | 1818018953048358912 |
|---|---|
| author | D. Schmekel F. Alcántara-Ávila S. Hoyas R. Vinuesa |
| author_facet | D. Schmekel F. Alcántara-Ávila S. Hoyas R. Vinuesa |
| author_sort | D. Schmekel |
| collection | DOAJ |
| description | Turbulent flow is widespread in many applications, such as airplane wings or turbine blades. Such flow is highly chaotic and impossible to predict far into the future. Some regions exhibit a coherent physical behavior in turbulent flow, satisfying specific properties; these regions are denoted as coherent structures. This work considers structures connected with the Reynolds stresses, which are essential quantities for modeling and understanding turbulent flows. Deep-learning techniques have recently had promising results for modeling turbulence, and here we investigate their capabilities for modeling coherent structures. We use data from a direct numerical simulation (DNS) of a turbulent channel flow to train a convolutional neural network (CNN) and predict the number and volume of the coherent structures in the channel over time. Overall, the performance of the CNN model is very good, with a satisfactory agreement between the predicted geometrical properties of the structures and those of the reference DNS data. |
| first_indexed | 2024-04-14T07:46:32Z |
| format | Article |
| id | doaj.art-d176aacf23fb403781aaff8e0a807ac3 |
| institution | Directory Open Access Journal |
| issn | 2296-424X |
| language | English |
| last_indexed | 2024-04-14T07:46:32Z |
| publishDate | 2022-04-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Physics |
| spelling | doaj.art-d176aacf23fb403781aaff8e0a807ac32022-12-22T02:05:20ZengFrontiers Media S.A.Frontiers in Physics2296-424X2022-04-011010.3389/fphy.2022.888832888832Predicting Coherent Turbulent Structures via Deep LearningD. Schmekel0F. Alcántara-Ávila1S. Hoyas2R. Vinuesa3FLOW, Engineering Mechanics, KTH Royal Institute of Technology, Stockholm, SwedenFLOW, Engineering Mechanics, KTH Royal Institute of Technology, Stockholm, SwedenInstituto de Matemática Pura y Aplicada, Universitat Politècnica de València, València, SpainFLOW, Engineering Mechanics, KTH Royal Institute of Technology, Stockholm, SwedenTurbulent flow is widespread in many applications, such as airplane wings or turbine blades. Such flow is highly chaotic and impossible to predict far into the future. Some regions exhibit a coherent physical behavior in turbulent flow, satisfying specific properties; these regions are denoted as coherent structures. This work considers structures connected with the Reynolds stresses, which are essential quantities for modeling and understanding turbulent flows. Deep-learning techniques have recently had promising results for modeling turbulence, and here we investigate their capabilities for modeling coherent structures. We use data from a direct numerical simulation (DNS) of a turbulent channel flow to train a convolutional neural network (CNN) and predict the number and volume of the coherent structures in the channel over time. Overall, the performance of the CNN model is very good, with a satisfactory agreement between the predicted geometrical properties of the structures and those of the reference DNS data.https://www.frontiersin.org/articles/10.3389/fphy.2022.888832/fullturbulencecoherent turbulent structuresmachine learningconvolutional neural networksdeep learning |
| spellingShingle | D. Schmekel F. Alcántara-Ávila S. Hoyas R. Vinuesa Predicting Coherent Turbulent Structures via Deep Learning Frontiers in Physics turbulence coherent turbulent structures machine learning convolutional neural networks deep learning |
| title | Predicting Coherent Turbulent Structures via Deep Learning |
| title_full | Predicting Coherent Turbulent Structures via Deep Learning |
| title_fullStr | Predicting Coherent Turbulent Structures via Deep Learning |
| title_full_unstemmed | Predicting Coherent Turbulent Structures via Deep Learning |
| title_short | Predicting Coherent Turbulent Structures via Deep Learning |
| title_sort | predicting coherent turbulent structures via deep learning |
| topic | turbulence coherent turbulent structures machine learning convolutional neural networks deep learning |
| url | https://www.frontiersin.org/articles/10.3389/fphy.2022.888832/full |
| work_keys_str_mv | AT dschmekel predictingcoherentturbulentstructuresviadeeplearning AT falcantaraavila predictingcoherentturbulentstructuresviadeeplearning AT shoyas predictingcoherentturbulentstructuresviadeeplearning AT rvinuesa predictingcoherentturbulentstructuresviadeeplearning |