Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions
Multi-scale system remains a classical scientific problem in fluid dynamics, biology, etc. In the present study, a scheme of multi-scale Physics-informed neural networks (msPINNs) is proposed to solve the boundary layer flow at high Reynolds numbers without any data. The flow is divided into several...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-03-01
|
| Series: | Theoretical and Applied Mechanics Letters |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2095034924000072 |
| _version_ | 1797345935958736896 |
|---|---|
| author | Jianlin Huang Rundi Qiu Jingzhu Wang Yiwei Wang |
| author_facet | Jianlin Huang Rundi Qiu Jingzhu Wang Yiwei Wang |
| author_sort | Jianlin Huang |
| collection | DOAJ |
| description | Multi-scale system remains a classical scientific problem in fluid dynamics, biology, etc. In the present study, a scheme of multi-scale Physics-informed neural networks (msPINNs) is proposed to solve the boundary layer flow at high Reynolds numbers without any data. The flow is divided into several regions with different scales based on Prandtl’s boundary theory. Different regions are solved with governing equations in different scales. The method of matched asymptotic expansions is used to make the flow field continuously. A flow on a semi infinite flat plate at a high Reynolds number is considered a multi-scale problem because the boundary layer scale is much smaller than the outer flow scale. The results are compared with the reference numerical solutions, which show that the msPINNs can solve the multi-scale problem of the boundary layer in high Reynolds number flows. This scheme can be developed for more multi-scale problems in the future. |
| first_indexed | 2024-03-08T11:26:01Z |
| format | Article |
| id | doaj.art-a7cc0f744e6a46bca8c7ae3c9b667674 |
| institution | Directory Open Access Journal |
| issn | 2095-0349 |
| language | English |
| last_indexed | 2024-03-08T11:26:01Z |
| publishDate | 2024-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Theoretical and Applied Mechanics Letters |
| spelling | doaj.art-a7cc0f744e6a46bca8c7ae3c9b6676742024-01-26T05:33:11ZengElsevierTheoretical and Applied Mechanics Letters2095-03492024-03-01142100496Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansionsJianlin Huang0Rundi Qiu1Jingzhu Wang2Yiwei Wang3Key Laboratory for Mechanics in Fluid Solid Coupling Systems, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, PR China; School of Future Technology, University of Chinese Academy of Sciences, Beijing 100049, PR ChinaKey Laboratory for Mechanics in Fluid Solid Coupling Systems, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, PR China; School of Future Technology, University of Chinese Academy of Sciences, Beijing 100049, PR ChinaKey Laboratory for Mechanics in Fluid Solid Coupling Systems, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, PR China; School of Engineering Science, University of Chinese Academy of Sciences, Beijing 100049, PR China; Guangdong Aerospace Research Academy, Guangzhou 511458, PR ChinaCorresponding author at: Key Laboratory for Mechanics in Fluid Solid Coupling Systems, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, PR China.; Key Laboratory for Mechanics in Fluid Solid Coupling Systems, Institute of Mechanics, Chinese Academy of Sciences, Beijing 100190, PR China; School of Future Technology, University of Chinese Academy of Sciences, Beijing 100049, PR China; School of Engineering Science, University of Chinese Academy of Sciences, Beijing 100049, PR ChinaMulti-scale system remains a classical scientific problem in fluid dynamics, biology, etc. In the present study, a scheme of multi-scale Physics-informed neural networks (msPINNs) is proposed to solve the boundary layer flow at high Reynolds numbers without any data. The flow is divided into several regions with different scales based on Prandtl’s boundary theory. Different regions are solved with governing equations in different scales. The method of matched asymptotic expansions is used to make the flow field continuously. A flow on a semi infinite flat plate at a high Reynolds number is considered a multi-scale problem because the boundary layer scale is much smaller than the outer flow scale. The results are compared with the reference numerical solutions, which show that the msPINNs can solve the multi-scale problem of the boundary layer in high Reynolds number flows. This scheme can be developed for more multi-scale problems in the future.http://www.sciencedirect.com/science/article/pii/S2095034924000072Physics-informed neural networks (PINNs)Multi-scaleFluid dynamicsBoundary layer |
| spellingShingle | Jianlin Huang Rundi Qiu Jingzhu Wang Yiwei Wang Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions Theoretical and Applied Mechanics Letters Physics-informed neural networks (PINNs) Multi-scale Fluid dynamics Boundary layer |
| title | Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions |
| title_full | Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions |
| title_fullStr | Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions |
| title_full_unstemmed | Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions |
| title_short | Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions |
| title_sort | multi scale physics informed neural networks for solving high reynolds number boundary layer flows based on matched asymptotic expansions |
| topic | Physics-informed neural networks (PINNs) Multi-scale Fluid dynamics Boundary layer |
| url | http://www.sciencedirect.com/science/article/pii/S2095034924000072 |
| work_keys_str_mv | AT jianlinhuang multiscalephysicsinformedneuralnetworksforsolvinghighreynoldsnumberboundarylayerflowsbasedonmatchedasymptoticexpansions AT rundiqiu multiscalephysicsinformedneuralnetworksforsolvinghighreynoldsnumberboundarylayerflowsbasedonmatchedasymptoticexpansions AT jingzhuwang multiscalephysicsinformedneuralnetworksforsolvinghighreynoldsnumberboundarylayerflowsbasedonmatchedasymptoticexpansions AT yiweiwang multiscalephysicsinformedneuralnetworksforsolvinghighreynoldsnumberboundarylayerflowsbasedonmatchedasymptoticexpansions |