RECONSTRUCTION OF THE FLOW FIELD AROUND A CYLINDER IN A TWO-DIMENSIONAL SHEAR FLOW BASED ON PINN

Yuan Guangyao; Wang Junsong; Zhao Xuanlie; Geng Jing

doi:10.6052/0459-1879-24-417

Volume 57 Issue 2

Feb. 2025

Turn off MathJax

Article Contents

Abstract

References

Chinese Journal of Theoretical and Applied Mechanics > 2025 > 57(2): 436-452. > DOI: 10.6052/0459-1879-24-417 CSTR: 32045.14.0459-1879-24-417

Yuan Guangyao, Wang Junsong, Zhao Xuanlie, Geng Jing. Reconstruction of flow field around two-dimensional shear flow cylinder based on PINN. Chinese Journal of Theoretical and Applied Mechanics, 2025, 57(2): 436-452. DOI: 10.6052/0459-1879-24-417

Citation:

PDF (9887 KB)

RECONSTRUCTION OF THE FLOW FIELD AROUND A CYLINDER IN A TWO-DIMENSIONAL SHEAR FLOW BASED ON PINN

Yuan Guangyao^*,
Wang Junsong^*,
Zhao Xuanlie^{*, †, ,},
Geng Jing^{*, †}

*.
College of Shipbuilding Engineering, Harbin Engineering University, Harbin 150000, China
†.
Sanya Nanhai Innovation and Development Base of Harbin Engineering University, Sanya 572024, Hainan, China

Received Date: August 27, 2024
Accepted Date: January 02, 2025
Available Online: January 02, 2025
Published Date: January 08, 2025

Graphical Abstract

Abstract

Abstract

With the advancement of machine learning techniques, data-driven flow field prediction has become a significant research focus in fluid mechanics. Compared with traditional neural networks, which often require large amounts of training data and suffer from poor interpretability, physics-informed neural networks (PINN) embed physical constraints directly into the loss function. This enables accurate flow field predictions with only minimal training data. This study investigates the reconstruction of shear flow around a circular cylinder under low Reynolds number conditions using a PINN-based approach. Numerical simulation data generated by the open-source computational fluid dynamics (CFD) software OpenFOAM were used to construct the training dataset. A PINN model incorporating shear flow boundary conditions for circular cylinder flow was developed, and velocity and pressure information from sampling points within the flow field were employed to train the model. The goal was to predict the velocity and pressure distributions across various regions of the flow field. To evaluate the performance of the PINN model, its predictions were compared with results from numerical simulations, demonstrating the model's ability to effectively reconstruct complex flow fields. Furthermore, this study investigated the impact of neural network parameters, such as the number of layers, the number of nodes per layer, and the spatial location of monitoring points, on the prediction accuracy. The mechanisms underlying the optimization of these parameters were analyzed in detail. The results indicate that the PINN model can accurately reconstruct the velocity and pressure distributions of shear flow around a circular cylinder, achieving a prediction accuracy that closely aligns with numerical simulation results. This validates the practical applicability and accuracy of PINN for reconstructing flow fields in low Reynolds number shear flows. Moreover, comparative analyses revealed that the location of monitoring points has a significant influence on the prediction results. Proper selection of monitoring points can enhance the prediction accuracy by 1 to 2 orders of magnitude.
- physical information neural network,
- Navier-Stokes equation,
- shear flow field reconstruction,
- flow around a cylinder

FullText(HTML)

References (32)

References

[1]	张伟伟, 王旭, 寇家庆. 面向流体力学的多范式融合研究展望. 力学进展, 2023, 53(2): 433-467 (Zhang Weiwei, Wang Xu, Kou Jiaqing. Prospects of multi-paradigm fusion methods for fluid mechanics research. Advances in Mechanics, 2023, 53(2): 433-467 (in Chinese) doi: 10.6052/1000-0992-22-050 Zhang Weiwei, Wang Xu, Kou Jiaqing. Prospects of multi-paradigm fusion methods for fluid mechanics research. Advances in Mechanics, 2023, 53(2): 433-467 (in Chinese) doi: 10.6052/1000-0992-22-050
[2]	Xu SR, Zhao JZ, Wu HK, et al. A review of solution stabilization techniques for RANS CFD solvers. Aerospace, 2023, 10(3): 230-230 doi: 10.3390/aerospace10030230
[3]	Subhrangshu P, Saud MA. Review of smooth particle hydrodynamics and its applications for environmental flows. Journal of the Institution of Engineers (India): Series A, 2022, 103(3): 921-941 doi: 10.1007/s40030-022-00650-4
[4]	El-Emam AM, Zhou L, Shi W, et al. Theories and applications of CFD-DEM coupling approach for granular flow: A review. Archives of Computational Methods in Engineering, 2021, 28(7): 1-42
[5]	张伟伟, 寇家庆, 刘溢浪. 智能赋能流体力学展望. 航空学报, 2021, 42(4): 26-71 (Zhang Weiwei, Kou Jiaqing, Liu Yilang. Prospect of artificial intelligence empowered fluid mechanics. Acta Aeronautica et Astronautica Sinica. 2021, 42(4): 26-71 (in Chinese) Zhang Weiwei, Kou Jiaqing, Liu Yilang. Prospect of artificial intelligence empowered fluid mechanics. Acta Aeronautica et Astronautica Sinica. 2021, 42(4): 26-71 (in Chinese)
[6]	查文舒, 李道伦, 沈路航等. 基于神经网络的偏微分方程求解方法研究综述. 力学学报, 2022, 54(3): 543-556 (Zha Wenshu, Li Daolun, Shen Luhang. Review of neural network-based methods for solving partial differential equations. Chinese Journal of Theoretical and Applied Mechanics, 2022, 54(3): 543-556 (in Chinese) doi: 10.6052/0459-1879-21-617 Zha Wenshu, Li Daolun, Shen Luhang. Review of neural network-based methods for solving partial differential equations. Chinese Journal of Theoretical and Applied Mechanics, 2022, 54(3): 543-556 (in Chinese) doi: 10.6052/0459-1879-21-617
[7]	Jin XW, Peng C, Li WC, et al. Prediction model of velocity field around circular cylinder over various Reynolds numbers by fusion convolutional neural networks based on pressure on the cylinder. Physics of Fluids, 2018, 30(4): 047105 doi: 10.1063/1.5024595
[8]	Xu H, Chang HB, Zhang DX. DL-PDE: Deep-learning based data-driven discovery of partial differential equations from discrete and noisy data. Communications in Computational Physics, 2021, 29(3): 698-728 doi: 10.4208/cicp.OA-2020-0142
[9]	Raissi M, Perdikaris P, Karniadakis G. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 2018, 378: 686-707
[10]	刘霞, 冯文晖, 连峰等. 基于物理信息神经网络的气动数据融合方法. 空气动力学学报, 2023, 41(8): 87-96 (Liu Xia, Feng Wenhui, Lian Feng, et al. Aerodynamic data fusion method based on physics-informed neural network. Acta Aerodynamica Sinica, 2023, 41(8): 87-96 (in Chinese) doi: 10.7638/kqdlxxb-2023.0060 Liu Xia, Feng Wenhui, Lian Feng, et al. Aerodynamic data fusion method based on physics-informed neural network. Acta Aerodynamica Sinica, 2023, 41(8): 87-96 (in Chinese) doi: 10.7638/kqdlxxb-2023.0060
[11]	侯先瑞, 周星宇, 黄骁骋. 基于物理信息神经网络的船舶螺旋桨尾流场重构. 上海交通大学学报, 2023, 101: 1-20 (Hou Xianrui, Zhou Xingyu, Huang Xiaopin. Reconstruction of ship propeller wake field based on physics-informed neural networks. Journal of Shanghai Jiaotong University, 2023, 101: 1-20 (in Chinese) Hou Xianrui, Zhou Xingyu, Huang Xiaopin. Reconstruction of ship propeller wake field based on physics-informed neural networks. Journal of Shanghai Jiaotong University, 2023, 101: 1-20 (in Chinese)
[12]	赵暾, 周宇, 程艳青等. 基于内嵌物理机理神经网络的热传导方程的正问题及逆问题求解. 空气动力学学报, 2021, 39(5): 19-26 (Zhao Tun, Zhou Yu, Cheng Yanqing, et al. Solving forward and inverse problems of the heat conduction equation using physics-informed neural networks. Acta Aerodynamica Sinica, 2021, 39(5): 19-26 (in Chinese) doi: 10.7638/kqdlxxb-2020.0176 Zhao Tun, Zhou Yu, Cheng Yanqing, et al. Solving forward and inverse problems of the heat conduction equation using physics-informed neural networks. Acta Aerodynamica Sinica, 2021, 39(5): 19-26 (in Chinese) doi: 10.7638/kqdlxxb-2020.0176
[13]	Ang WHE, Wang G, Ng FB. Physics-informed neural networks for low reynolds number flows over cylinder. Energies, 2023, 16(12): 4558
[14]	Sun L, Gao H, Pan S, et al. Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data. Computer Methods in Applied Mechanics and Engineering, 2019, 361(C): 112732
[15]	Maziar R, Alireza Y, Em GK. Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations. Science, 2020, 367(6481): 1026-1030 doi: 10.1126/science.aaw4741
[16]	Jagtap DA, Kawaguchi K, Karniadakis EG. Adaptive activation functions accelerate convergence in deep and physics-informed neural networks. Journal of Computational Physics, 2020, 404: 109136
[17]	Chen C, Zhang GT. Deep learning method based on physics informed neural network with resnet block for solving fluid flow problems. Water, 2021, 13(4): 423 doi: 10.3390/w13040423
[18]	Chen C, Hao M, Li YZ, et al. Deep learning based on PINN for solving 2 DOF vortex induced vibration of cylinder. Ocean Engineering, 2021, 240: 109932
[19]	宋家豪, 曹文博, 张伟伟. FD-PINN: 频域物理信息神经网络. 力学学报, 2023, 55(5): 1195-1205 (Song Jiahao, Cao Wenbo, Zhang Weiwei. FD-PINN: Frequency domain physics-informed neural network. Chinese Journal of Theoretical and Applied Mechanics, 2023, 55(5): 1195-1205 (in Chinese) doi: 10.6052/0459-1879-23-169 Song Jiahao, Cao Wenbo, Zhang Weiwei. FD-PINN: Frequency domain physics-informed neural network. Chinese Journal of Theoretical and Applied Mechanics, 2023, 55(5): 1195-1205 (in Chinese) doi: 10.6052/0459-1879-23-169
[20]	杨茵. 剪切流动掺混特性分析及其流动控制研究. [博士论文]. 西安: 西北工业大学, 2019 (Yang yin. Mixing characteristic analysis and flow control of shear flow. [PhD Thesis]. Xi’an: Northwestern Polytechnical University, 2019 (in Chinese) Yang yin. Mixing characteristic analysis and flow control of shear flow. [PhD Thesis]. Xi’an: Northwestern Polytechnical University, 2019 (in Chinese)
[21]	李野, 陈松灿. 基于物理信息的神经网络: 最新进展与展望. 计算机科学, 2022, 49(4): 254-262 (Li Ye, Chen Songcan. Physics-informed neural networks: recent advances and prospects. Computer Science, 2022, 49(4): 254-262 (in Chinese) Li Ye, Chen Songcan. Physics-informed neural networks: recent advances and prospects. Computer Science, 2022, 49(4): 254-262 (in Chinese)
[22]	郭远, 傅卓佳, 闵建等. 课程-迁移学习物理信息神经网络用于长时间非线性波传播模拟. 力学学报, 2024, 56(3): 763-773 (Guo Yuan, Fu Zhuojia, Min Jian, et al. Curriculum-transfer-learning based physics-informed neural networks for long-time simulation of nonlinear wave propagation. Chinese Journal of Theoretical and Applied Mechanics, 2024, 56(3): 763-773 (in Chinese) doi: 10.6052/0459-1879-23-457 Guo Yuan, Fu Zhuojia, Min Jian, et al. Curriculum-transfer-learning based physics-informed neural networks for long-time simulation of nonlinear wave propagation. Chinese Journal of Theoretical and Applied Mechanics, 2024, 56(3): 763-773 (in Chinese) doi: 10.6052/0459-1879-23-457
[23]	陈旷奇, 欧阳晗青, 朱志成等. 基于稀疏数据的流场结构重构方法研究. 北京理工大学学报, 2023, 43(9): 950-961 (Chen Kuangqi, Ouyang Hanqing, Zhu Zhicheng. Research on flow field structure prediction method driven by physical information. Transactions of Beijing Institute of Technology, 2023, 43(9): 950-961 (in Chinese) Chen Kuangqi, Ouyang Hanqing, Zhu Zhicheng. Research on flow field structure prediction method driven by physical information. Transactions of Beijing Institute of Technology, 2023, 43(9): 950-961 (in Chinese)
[24]	Omar S, Mirjam F. On the use of fourier features-physics informed neural networks (FF-PINN) for forward and inverse fluid mechanics problems. Proceedings of the Institution of Mechanical Engineers, Part M: Journal of Engineering for the Maritime Environment, 2023, 237(4): 846-866 doi: 10.1177/14750902231166424
[25]	Michael PDAJ, Shandian Z, et al. A unified scalable framework for causal sweeping strategies for physics-informed neural networks (PINNs) and their temporal decompositions. Journal of Computational Physics, 2023, 493: 112464
[26]	Xue Y, Wang YK, Liang JY. A self-adaptive gradient descent search algorithm for fully-connected neural networks. Neurocomputing, 2022, 478: 70-80
[27]	Sirignano J, Spiliopoulos K. DGM: A deep learning algorithm for solving partial differential equations. Journal of Computational Physics, 2018, 375: 1339-1364
[28]	Emanuele A, Alessandro T, Enrico S. Quantification of uncertainty in a defect-based physics-informed neural network for fatigue evaluation and insights on influencing factors. Engineering Fracture Mechanics, 2023, 292: 109595
[29]	He YC, Wang ZC, Xiang H, et al. An artificial viscosity augmented physics-informed neural network for incompressible flow. Applied Mathematics and Mechanics, 2023, 44(7): 1101-1110 doi: 10.1007/s10483-023-2993-9
[30]	Laurin ES, Alf G, Jens L, et al. Physics informed neural networks: A case study for gas transport problems. Journal of Computational Physics, 2023, 481: 112041
[31]	徐健, 朱海龙, 朱江乐等. 基于物理信息神经网络的Burgers-Fisher方程求解方法. 浙江大学学报(工学版), 2023, 57(11): 2160-2169 (Xu Jian, Zhu Hailong, Zhu Jiangle, et al. Solution approach of Burgers-Fisher equation based on physics-informed neural networks. Journal of Zhejiang University (Engineering Science), 2023, 57(11): 2160-2169 (in Chinese) doi: 10.3785/j.issn.1008-973X.2023.11.003 Xu Jian, Zhu Hailong, Zhu Jiangle, et al. Solution approach of Burgers-Fisher equation based on physics-informed neural networks. Journal of Zhejiang University (Engineering Science), 2023, 57(11): 2160-2169 (in Chinese) doi: 10.3785/j.issn.1008-973X.2023.11.003
[32]	刘强. OpenFOAM在绕流问题中的应用研究. [硕士论文]. 哈尔滨: 哈尔滨工程大学, 2012 (Liu Qiang. Application Research of OpenFOAM in the study of flow around obstacle. [PhD Thesis]. Harbin: Harbin Engineering University, 2012 (in Chinese) Liu Qiang. Application Research of OpenFOAM in the study of flow around obstacle. [PhD Thesis]. Harbin: Harbin Engineering University, 2012 (in Chinese)

[1]	Shen Feng, Chen Jie, Cheng Siyuan, Pang Yan, Liu Zhaomiao. EFFECTS OF GAS-LIQUID INTERFACE POSITION ON THE TWO-PHASE FLOW FIELDS AND DRAG REDUCTION IN MICRO-GROOVES[J]. Chinese Journal of Theoretical and Applied Mechanics, 2024, 56(9): 2511-2521. DOI: 10.6052/0459-1879-24-105
[2]	Guo Peiyang, Zhang Yi, Zhang Mengzhuo, Hu Haibao. EXPERIMENTAL STUDY ON DRAG REDUCTION BY AIR INJECTION ON HYDROPHILIC AND ALTERNATED SUPERHYDROPHOBIC SURFACES[J]. Chinese Journal of Theoretical and Applied Mechanics, 2024, 56(1): 94-100. DOI: 10.6052/0459-1879-23-303
[3]	Bai Jianxia, Zhao Kaifang, Cheng Xiaoqi, Jiang Nan. EXPERIMENTAL INVESTIGATION ON ACTIVE CONTROL TURBULENT BOUNDARY LAYER DRAY REDUCTION BY SYNCHRONOUS AND ASYNCHRONOUS VIBRATION OF DUAL VIBRATORS[J]. Chinese Journal of Theoretical and Applied Mechanics, 2023, 55(1): 52-61. DOI: 10.6052/0459-1879-22-248
[4]	Shi Pengfei, Du Wei, Hu Haibao, Feng Jiaxing, Xie Luo. EXPERIMENTAL STUDY ON DRAG REDUCTION CHARACTERISTICS OF DIUTAN GUM SOLUTION BY SLIT INJECTION[J]. Chinese Journal of Theoretical and Applied Mechanics, 2022, 54(5): 1257-1263. DOI: 10.6052/0459-1879-21-567
[5]	Feng Jiaxing, Hu Haibao, Lu Bingju, Qin Liping, Zhang Mengzhuo, Du Peng, Huang Xiao. EXPERIMENTAL STUDY ON DRAG REDUCTION CHARACTERISTICS OF SUPERHYDROPHOBIC GROOVE SURFACES WITH VENTILATION[J]. Chinese Journal of Theoretical and Applied Mechanics, 2020, 52(1): 24-30. DOI: 10.6052/0459-1879-19-279
[6]	Shuaijie Wang, Xiaotong Cui, Jianxia Bai, Zhanqi Tang, Nan Jiang. THE EFFECT OF PERIODIC PERTURBATION ON MULTI SCALES IN A TURBULENT BOUNDARY LAYER FLOW UNDER DRAG REDUCTION[J]. Chinese Journal of Theoretical and Applied Mechanics, 2019, 51(3): 767-774. DOI: 10.6052/0459-1879-18-409
[7]	Cai Shupeng, Wang Zhineng, Duan Chuanwei, Li Dan. DRAG CHARACTERISTICS OF A DRAG-REDUCING SURFACTANT SOLUTION FLOWING OVER A SUDDEN-EXPANSION PIPE[J]. Chinese Journal of Theoretical and Applied Mechanics, 2018, 50(2): 274-283. DOI: 10.6052/0459-1879-17-328
[8]	Li Shan, Yang Shaoqiong, Jiang Nan. TRPIV MEASUREMENT OF DRAG-REDUCTION IN THE TURBULENT BOUNDARY LAYER OVER RIBLETS PLATE[J]. Chinese Journal of Theoretical and Applied Mechanics, 2013, 45(2): 183-192. DOI: 10.6052/0459-1879-12-262
[9]	Shidong Luo, Chunxiao Xu, Guixiang Cui. Direct numerical simulation of turbulent pipe flow controlled by MHD for drag reduction[J]. Chinese Journal of Theoretical and Applied Mechanics, 2007, 23(3): 311-319. DOI: 10.6052/0459-1879-2007-3-2006-296
[10]	THE PHENOMENON OF DRAG REDUCTION IN FLOWS OF CLAY SUSPENSIONS[J]. Chinese Journal of Theoretical and Applied Mechanics, 1996, 28(5): 522-531. DOI: 10.6052/0459-1879-1996-5-1995-365

Cited By

Get Citation

PDF

XML

Article Metrics

Article views (269) PDF downloads (108)

RECONSTRUCTION OF THE FLOW FIELD AROUND A CYLINDER IN A TWO-DIMENSIONAL SHEAR FLOW BASED ON PINN

Abstract

References

Related Articles

Catalog

Article Metrics

Related

Download Center

Author Center

RECONSTRUCTION OF THE FLOW FIELD AROUND A CYLINDER IN A TWO-DIMENSIONAL SHEAR FLOW BASED ON PINN

Abstract

References

Related Articles

Catalog

Article Metrics

Related

Download Center

Author Center

Export File

Citation

Format

Content