Solution Method of Flow Field in the Narrow Rectangular Channel Based on Physics-informed Neural Network

Zhang Xiaoying; Yuan Dewen; Bi Jingliang; Huang Yanping

doi:10.13832/j.jnpe.2024.080040

Volume 46 Issue 4

Aug. 2025

Turn off MathJax

Article Contents

Article Navigation > Nuclear Power Engineering > 2025 > 46(4): 266-272

Zhang Xiaoying, Yuan Dewen, Bi Jingliang, Huang Yanping. Solution Method of Flow Field in the Narrow Rectangular Channel Based on Physics-informed Neural Network[J]. Nuclear Power Engineering, 2025, 46(4): 266-272. doi: 10.13832/j.jnpe.2024.080040

Citation:

Zhang Xiaoying, Yuan Dewen, Bi Jingliang, Huang Yanping. Solution Method of Flow Field in the Narrow Rectangular Channel Based on Physics-informed Neural Network[J]. Nuclear Power Engineering, 2025, 46(4): 266-272. doi: 10.13832/j.jnpe.2024.080040

Citation:

PDF( 0 KB)

Solution Method of Flow Field in the Narrow Rectangular Channel Based on Physics-informed Neural Network

doi: 10.13832/j.jnpe.2024.080040

State Key Laboratory of Advanced Nuclear Energy Technology, Nuclear Power Institute of China, Chengdu, 610213, China

Received Date: 2024-08-15
Rev Recd Date: 2025-04-30
Publish Date: 2025-08-15

Abstract

Abstract

To explore the application potential of Physics-informed Neural Network (PINN) in the field of thermal and hydraulic calculation, multiple working conditions for both laminar and turbulent flow in a narrow rectangular channel were calculated in this study. Computational Fluid Dynamics (CFD) was utilized to obtain label data, and the continuity equation and N-S equations were embedded into the neural network model for prediction. The results show that for the incompressible flow in the narrow rectangular channel, the PINN model can accurately restore the flow field characteristics for laminar flow conditions. For turbulent conditions, the weight of the loss term of the model can be adjusted to achieve good consistency between the predicted solution and the CFD numerical solution. Therefore, the PINN model can be applied to the flow field calculation of narrow rectangular channels, and can further accumulate experience for the rapid analysis of flow fields in more scenarios.
- Physics-informed neural network (PINN),
- Narrow rectangular channel,
- Navier-Stokes equations(N-S equations)

FullText(HTML)

References(15)

References

[1]	HORNIK K, STINCHCOMBE M, WHITE H. Universal approximation of an unknown mapping and its derivatives using multilayer feedforward networks[J]. Neural Networks, 1990, 3(5): 551-560. doi: 10.1016/0893-6080(90)90005-6
[2]	PSICHOGIOS D C, UNGAR L H. A hybrid neural network‐first principles approach to process modeling[J]. AIChE Journal, 1992, 38(10): 1499-1511. doi: 10.1002/aic.690381003
[3]	LAGARIS I E, LIKAS A, FOTIADIS D I. Artificial neural networks for solving ordinary and partial differential equations[J]. IEEE Transactions on Neural Networks, 1998, 9(5): 987-1000. doi: 10.1109/72.712178
[4]	RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378: 686-707. doi: 10.1016/j.jcp.2018.10.045
[5]	CAI S Z, MAO Z P, WANG Z C, et al. Physics-informed neural networks (PINNs) for fluid mechanics: a review[J]. Acta Mechanica Sinica, 2021, 37(12): 1727-1738. doi: 10.1007/s10409-021-01148-1
[6]	RAO C P, SUN H, LIU Y. Physics-informed deep learning for incompressible laminar flows[J]. Theoretical and Applied Mechanics Letters, 2020, 10(3): 207-212. doi: 10.1016/j.taml.2020.01.039
[7]	JIN X W, CAI S Z, LI H, et al. NSFnets (Navier-Stokes flow nets): physics-informed neural networks for the incompressible Navier-Stokes equations[J]. Journal of Computational Physics, 2021, 426: 109951. doi: 10.1016/j.jcp.2020.109951
[8]	CAO W B, SONG J H, ZHANG W W. A solver for subsonic flow around airfoils based on physics-informed neural networks and mesh transformation[J]. Physics of Fluids, 2024, 36(2): 027134. doi: 10.1063/5.0188665
[9]	CAI S Z, WANG Z C, WANG S F, et al. Physics-informed neural networks for heat transfer problems[J]. Journal of Heat and Mass Transfer, 2021, 143(6): 060801.
[10]	HENNIGH O, NARASIMHAN S, NABIAN M A, et al. NVIDIA SimNet^TM: an ai-accelerated multi-physics simulation framework[M]//PASZYNSKI M, KRANZLMÜLLER D, KRZHIZHANOVSKAYA V V, et al. Computational Science–ICCS 2021. Krakow: Springer International Publishing, 2021: 447-461.
[11]	陆至彬,瞿景辉,刘桦,等. 基于物理信息神经网络的传热过程物理场代理模型的构建[J]. 化工学报,2021, 72(3): 1496-1503.
[12]	张程宾,陈永平,施明恒,等. 表面粗糙度的分形特征及其对微通道内层流流动的影响[J]. 物理学报,2009, 58(10): 7050-7056. doi: 10.3321/j.issn:1000-3290.2009.10.062
[13]	幸奠川,阎昌琪,曹夏昕,等. 高宽比对矩形窄通道内单相水流动特性的影响机理[J]. 原子能科学技术,2013, 47(1): 43-47. doi: 10.7538/yzk.2013.47.01.0043
[14]	尧少波,何伟峰,陈丽华,等. 融合物理的神经网络方法在流场重建中的应用[J]. 空气动力学学报,2022, 40(5): 30-38. doi: 10.7638/kqdlxxb-2021.0080
[15]	王浩蔓. 基于物理信息神经网络的大地电磁正演研究[D]. 长春: 吉林大学,2022.