基于GPU的二维梯形空腔流的格子Boltzmann模拟与分析

陈百慧; 施保昌; 汪垒; 柴振华

doi:10.7498/aps.72.20230430

摘要

采用格子Boltzmann方法模拟上下壁面驱动的二维梯形空腔流, 并使用GPU-CUDA程序进行加速计算. 主要采用本征正交分解方法, 分析了4种壁面驱动条件的流场模态, 并探究了雷诺数和驱动速度方向对流场形态的影响. 结果表明: 1)当上壁面单驱动(T1a)时, 若雷诺数为1000—8000, 流场处于稳态流动; 雷诺数为8500时, 流场处于周期性非稳态流动; 雷诺数大于10000时, 流场处于非周期非稳态流动. 2)当下壁面单驱动(T1b)时, 若雷诺数在1000—8000之间, 流场处于稳态流动; 雷诺数增大至11500时, 流场处于周期性非稳态流动; 雷诺数大于12500时, 流场进入非周期非稳态流动. 3)当上下壁面同方向同速度双驱动(T2a)时, 若雷诺数在1000—10000区间, 流场均为稳态流动; 雷诺数为12500—15000时, 流场处于周期性非稳态流动; 当雷诺数大于20000时, 流场为非周期非稳态流动. 4)当上下壁面反方向同速度双驱动(T2b)时, 若雷诺数在1000—5000之间, 流场处于稳态流动; 雷诺数为6000时, 流场处于周期性非稳态流动; 雷诺数大于8000时, 流场为非周期非稳态流动.

关键词:

Abstract

In this study, we utilize the lattice Boltzmann method to investigate the flow behavior in a two-dimensional trapezoidal cavity, which is driven by both sides on the upper wall and lower wall. Our calculations are accelerated through GPU-CUDA software. We conduct an analysis of the flow field mode by using proper orthogonal decomposition. The effects of various parameters, such as Reynolds number (Re) and driving direction, on the flow characteristics are examined through numerical simulations. The results are shown below. 1) For the upper wall drive (T1a), the flow field remains stable, when the Re value varies from 1000 to 8000. However, when Re = 8500, the flow field becomes periodic but unstable. The velocity phase diagram at the monitoring point is a smooth circle, and the energy values of the first two modes dominate the energy of the whole field. Once Re exceeds 10000, the velocity phase diagram turns irregular and the flow field becomes aperiodic and unsteady. 2) For the lower wall drive (T1b), the flow is stable when Re value is in a range of 1000－8000, and it becomes periodic and unsteady when Re = 11500. The energy values of the first three modes appear relatively large. When Re is greater than 12500, the flow field becomes aperiodic and unsteady. At this time, the phase diagram exhibits a smooth circle, with the energy values of the first two modes almost entirely dominating the entire energy. 3) For the case of upper wall and lower wall moving in the same direction at the same speed (T2a), the flow field remains stable when Re changes from 1000 to 10000. When Re varies from 12500 to 15000, the flow becomes periodic and unstable. The velocity phase diagram is still a smooth circle, with the first two modes still occupying a large portion of the energy. Once Re exceeds 20000, the energy proportions of the first three modes significantly decrease, and the flow becomes aperiodic and unsteady. 4) For the case in which the upper wall and lower wall are driven in opposite directions at the same velocity (T2b), the flow field remains stable when Re changes from 1000 to 5000. When Re = 6000, the energy of the first mode accounts for 86%, and the flow field becomes periodic but unstable. When Re exceeds 8000, the energy proportions of the first three modes decrease significantly, and the flow field becomes aperiodic and unsteady.

Keywords:

作者及机构信息

1.
华中科技大学数学与统计学院, 武汉　430074

2.
华中科技大学, 工程建模与科学计算湖北重点实验室, 武汉　430074

3.
华中科技大学数学与应用学科交叉创新研究院, 武汉　430074

4.
中国地质大学(武汉)数学与物理学院, 武汉　430074

5.
中国地质大学(武汉)数学科学中心, 武汉　430074

通信作者: 施保昌, shibc@hust.edu.cn

基金项目: 国家自然科学基金(批准号: 12072127, 51836003)资助的课题

Authors and contacts

1.
School of Mathematics and Statistics, Huazhong University of Science and Technology, Wuhan 430074, China

2.
Hubei Key Laboratory of Engineering Modeling and Scientific Computing, Huazhong University of Science and Technology, Wuhan 430074, China

3.
Institute of Interdisciplinary Research for Mathematics and Applied Science, Huazhong University of Science and Technology, Wuhan 430074, China

4.
School of Mathematics and Physics, China University of Geosciences, Wuhan 430074, China

5.
Center for Mathematical Science, China University of Geosciences, Wuhan 430074, China

Corresponding author: Shi Bao-Chang, shibc@hust.edu.cn

Funds: Project supported by the National Natural Science Foundation of China (Grant Nos. 12072127, 51836003)

文章全文 : translate this paragraph

参考文献

[1]	Peng Y F, Shiau Y H, Hwang R R 2003 Comput. Fluids 32 337 Google Scholar
[2]	Isaev S A, Baranov P A, Sudakov A G, et al. 2008 Thermophys. Aeromech. 15 463 Google Scholar
[3]	Balachandar S, Eaton J K 2010 Annu. Rev. Fluid Mech. 42 111 Google Scholar
[4]	Kuhlmann H C, Romanò F 2019 The Lid-Driven Cavity (Austria: Springer International Publishing) p233
[5]	Nonino C, Del Giudice S 1988 Int. J. Numer. Methods Eng. 25 313 Google Scholar
[6]	Alleborn N, Raszillier H, Durst F 1999 Int. J. Heat Mass Tran. 42 833 Google Scholar
[7]	Talebi F, Mahmoudi A H, Shahi M 2010 Int. Commun. Heat Mass 37 79 Google Scholar
[8]	Chowdhury M, Kumar B V R 2023 J. Nonnewton. Fluid Mech. 312 104975 Google Scholar
[9]	Ribbens C J, Watson L T, Wang C Y 1994 J. Comput. Phys. 112 173 Google Scholar
[10]	McQuain W D, Ribbens C J, Wang C Y, et al. 1994 Comput. Fluids 23 613 Google Scholar
[11]	Ismael M A, Chamkha A J 2015 Numer. Heat Transfer, Part A 68 312 Google Scholar
[12]	Mebarek-Oudina F, Laouira H, Hussein A K, et al. 2022 Mathematics 10 929 Google Scholar
[13]	Mondal P, Mahapatra T R 2021 Int. J. Mech. Sci. 208 106665 Google Scholar
[14]	Kareem A K, Mohammed H A, Hussein A K, et al. 2016 Int. Commun. Heat Mass 77 195 Google Scholar
[15]	Rashad A M, Sivasankaran S, Mansour M A, et al. 2017 Numer. Heat Transfer, Part A 71 1223 Google Scholar
[16]	Darr J H, Vanka S P 1991 Phys. Fluids 3 385 Google Scholar
[17]	Paramane S B, Sharma A 2008 Numer. Heat Transfer, Part B 54 84 Google Scholar
[18]	张恒, 任峰, 胡海豹 2021 物理学报 70 184703 Google Scholar Zhang H, Ren F, Hu H B 2021 Acta Phys. Sin. 70 184703 Google Scholar
[19]	Yang F, Shi X, Guo X, et al. 2012 Energy Procedia 16 639 Google Scholar
[20]	Nagendra K, Lakshmisha K N 2009 Int. J. Numer. Methods Heat Fluid Flow 19 790 Google Scholar
[21]	Perumal D A, Dass A K 2011 Comput. Math. Appl. 61 3711 Google Scholar
[22]	Nemati H, Farhadi M, Sedighi K, et al. 2010 Int. Commun. Heat Mass 37 1528 Google Scholar
[23]	Guo X, Zhong C, Zhuo C, et al. 2014 Theor. Comput. Fluid Dyn. 28 215 Google Scholar
[24]	Patil D V, Lakshmisha K N, Rogg B 2006 Comput. Fluids 35 1116 Google Scholar
[25]	Hou S, Zou Q, Chen S, et al. 1995 J. Comput. Phys. 118 329 Google Scholar
[26]	Chai Z H, Shi B C, Zheng L 2006 Chin. Phys. 15 1855 Google Scholar
[27]	Yuana K A, Budiana E P 2019 IOP Conf. Ser.: Mater. Sci. Eng. 546 052088 Google Scholar
[28]	Perumal D A, Dass A K 2008 WIT Trans. Eng. Sci. 59 45 Google Scholar
[29]	Bo A N, Mellibovsky F, Bergada J M, et al. 2020 Appl. Math. Modell. 82 469 Google Scholar
[30]	Shahid H, Yaqoob I, Khan W A, et al. 2021 Int. Commun. Heat Mass 129 105658 Google Scholar
[31]	Perumal D A 2018 Therm. Sci. Eng. Prog. 6 48 Google Scholar
[32]	Guo Y, Bennacer R, Shen S, et al. 2010 Int. J. Numer. Methods Heat Fluid Flow 20 130 Google Scholar
[33]	Shahid H, Yaqoob I, Khan W A, et al. 2021 Int. Commun. Heat Mass 128 105552 Google Scholar
[34]	An B, Bergada J M, Mellibovsky F 2019 J. Fluid Mech. 875 476 Google Scholar
[35]	An B, Guo S, Bergadà J M 2023 Appl. Sci. 13 888 Google Scholar
[36]	Sidik N A C, Munir F A 2012 Arab. J. Sci. Eng. 37 1723 Google Scholar
[37]	Zhang T, Shi B, Chai Z 2010 Compu. Fluids 39 1977 Google Scholar
[38]	Qian Y H, D’Humières D, Lallemand P 1992 EPL 17 479 Google Scholar
[39]	Ziegler D P 1993 J. Stat. Phys. 71 1171 Google Scholar
[40]	Guo Z L, Zheng C G, Shi B C 2002 Chin. Phys. 11 366 Google Scholar
[41]	Guo Z L, Zheng C G, Shi B C 2002 Phys. Fluids 14 2007 Google Scholar

施引文献

图 1 二维等腰梯形空腔示意图

Fig. 1. Coordinate of two-dimensional isosceles trapezoidal cavity

下载: 全尺寸图片幻灯片

图 2 曲边界上各节点示意图

Fig. 2. Geometry condition of curved wall boundary

下载: 全尺寸图片幻灯片

图 3 对比Re = 100和500时垂直和水平中线上的速度　(a), (c) 沿垂直中线的速度u和纵坐标y的关系图; (b), (d) 沿水平中线的速度v和横坐标x的关系图

Fig. 3. Comparison of the velocity profiles along the vertical and horizontal lines through cavity center at Re = 100, 500: (a) and (c) The y-component of velocity u profiles along the vertical line; (b) and (d) x-component of velocity v profiles along the horizontal line.

下载: 全尺寸图片幻灯片

图 4 对比网格数为$ 256\times128 $, $ 512\times256 $, $ 1024\times512 $, Re = 500, 10000时垂直和水平中线上的速度. $ U_1=0.1 $, $ U_2=0 $, Re = 500时(a)沿垂直中线的速度u和纵坐标y的关系图, (b)沿水平中线的速度v和横坐标x的关系图. $ U_1=0.1 $, $ U_2=0.1 $, Re = 10000时(c) 沿垂直中线的速度u和纵坐标y的关系图, (d) 沿水平中线的速度v和横坐标x的关系图

Fig. 4. Comparison of the velocity profiles along the vertical and horizontal lines at Re = 500, 10000 for $ 256\times128 $, $ 512\times256 $, $ 1024\times512 $. (a) The y-component of velocity u profiles along the vertical line, (b) x-component of velocity v profiles along the horizontal line at $ U_1=0.1 $, $ U_2=0 $ and Re = 500. (c) The y-component of velocity u profiles along the vertical line, (d) x-component of velocity v profiles along the horizontal line at $ U_1=0.1 $, $ U_2=0.1 $ and Re = 10000.

下载: 全尺寸图片幻灯片

图 5 4种不同壁面驱动速度下的梯形空腔示意图　(a) Case T1a, 上壁面向右单驱动; (b) Case T1b, 下壁面向右单驱动; (c) Case T2a, 上壁面和下壁面同时向右双驱动; (d) Case T2b, 上壁面向右驱动, 下壁面向左双驱动

Fig. 5. Trapezoidal cavity under four different velocity boundaries: (a) Case T1a, only the upper wall-driven; (b) Case T1b, only the lower wall-driven; (c) Case T2a, the upper and lower wall are driven at the same speed and in the same direction; (d) Case T2b, the upper and lower wall are driven at the same speed and in the opposite direction

下载: 全尺寸图片幻灯片

图 6 Case T1a条件下, Re为1000—8000时的流场形态　(a) Re = 1000; (b) Re = 3200; (c) Re = 5000; (d) Re = 6000; (e) Re = 8000

Fig. 6. Stream function contours of Reynolds numbers from 1000 to 8000 for Case T1a: (a) Re = 1000; (b) Re = 3200; (c) Re = 5000; (d) Re = 6000; (e) Re = 8000

下载: 全尺寸图片幻灯片

图 7 Case T1a条件下, Re为1000—8000时, (a) $ x=1 $的中线上, 纵坐标y和水平速度u的关系, (b) $ y=0.5 $ 的中线上, 横坐标x和竖直速度v的关系

Fig. 7. (a) Variation of the velocity u magnitude along $ x=1 $, (b) variation of the velocity v magnitude along $ y=0.5 $ for Re from 1000 to 8000 for Case T1a.

下载: 全尺寸图片幻灯片

图 8 Case T1a条件下, Re = 8500时呈现周期性非稳态流动　(a) 时间步$ t=3000000 $时的流场图; (b) 监控点处速度相图; (c) 速度监控点处速度随时间的变化

Fig. 8. Stream function contours are periodic and unsteady at Re = 8500 for Case T1a: (a) Stream function contours at time steps $ t=3000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 9 Case T1a条件下, Re = 8500时的各阶模态图　(a)速度平均场模态图; (b) 前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 9. Each order of mode for Case T1a when Re = 8500: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 10 Case T1a条件下, Re = 10000时呈现非周期非稳态流动　(a)时间步$ t=5000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 10. Stream function contours are non-periodic and unsteady at Re = 10000 for Case T1a: (a) Stream function contours at time steps $ t=5000000 $; (b) velocity phase diagrams at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 11 Case T1a条件下, Re = 12500时呈现非周期非稳态流动　(a)时间步$ t=5000000 $时的流场图; (b)监控点处速度相图; (c)监控点处速度随时间的变化

Fig. 11. Stream function contours are non-periodic and unsteady at Re = 12500 for Case T1a: (a) Stream function contours at time steps $ t=5000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 12 Case T1b条件下, Re为1000—8000时的流场形态　(a) Re = 1000; (b) Re = 3200; (c) Re = 5000; (d) Re = 6000; (e) Re = 8000

Fig. 12. Stream function contours of Reynolds numbers from 1000 to 8000 for Case T1b: (a) Re = 1000; (b) Re = 3200; (c) Re = 5000; (d) Re = 6000; (e) Re = 8000

下载: 全尺寸图片幻灯片

图 13 Case T1b条件下, Re为1000—8000时, (a) $ x=1 $的中线上, 纵坐标y 和水平速度u的关系; (b) $ y=0.5 $的中线上, 横坐标x和竖直速度v的关系

Fig. 13. (a) Variation of the velocity u magnitude along $ x=1 $, (b) variation of the velocity v magnitude along $ y=0.5 $ for Re from 1000 to 8000 for Case T1b.

下载: 全尺寸图片幻灯片

图 14 Case T1b条件下, Re = 11500时呈现周期性非稳态流动　(a)时间步$ t=5000000 $时的流场图; (b)监控点处速度相图; (c)监控点处速度随时间变化图

Fig. 14. (a) Stream function contours are periodic and unsteady at Re = 11500 for Case T1b: (a) Stream function contours at time steps $ t=5000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 15 Case T1b条件下, Re = 11500时的各阶模态图　(a)速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 15. Each order of mode for Case T1b when Re = 11500: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 16 Case T1b条件下, Re = 12500时呈非周期非稳态流动　(a)时间步$ t=6000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 16. Stream function contours are non-periodic and unsteady at Re = 12500 for Case T1b: (a) Stream function contours at time steps $ t=6000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 17 Case T1b条件下, Re = 15000时呈非周期非稳态流动　(a)时间步$ t=6000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 17. Stream function contours are non-periodic and unsteady at Re = 15000 for Case T1b: (a) Stream function contours at time steps $ t=6000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 18 Case T2a条件下, Re为1000—10000时的流场形态　(a) Re = 1000; (b) Re = 3200; (c) Re = 5000; (d) Re = 6000; (e) Re = 8000; (f) Re = 10000

Fig. 18. Stream function contours of Reynolds numbers from 1000 to 10000 for Case T2a: (a) Re = 1000; (b) Re = 3200; (c) Re = 5000; (d) Re = 6000; (e) Re = 8000; (f) Re = 10000

下载: 全尺寸图片幻灯片

图 19 Case T2a条件下, Re为1000—10000时, (a) $ x=1 $的中线上, 纵坐标y 和水平速度u的关系; (b) $ y=0.5 $的中线上, 横坐标x和竖直速度v的关系

Fig. 19. (a) Variation of the velocity u magnitude along $ x=1 $, (b) variation of the velocity v magnitude along $ y=0.5 $ at Re = 1000–10000 for Case T2a.

下载: 全尺寸图片幻灯片

图 20 Case T2a条件下, Re = 12500时呈现周期性非稳态流动　(a)时间步$ t=6000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 20. Stream function contours are periodic and unsteady at Re = 12500 for Case T2a: (a) Stream function contours at time steps $ t=6000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 21 Case T2a条件下, Re = 12500时的各阶模态图　(a)速度平均场模态图; (b)前n 阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 21. Each order of mode for Case T2a when Re = 12500: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 22 Case T2a条件下, Re = 15000时呈现周期性非稳态流动　(a)时间步$ t=6000000 $时的流场图; (b) 监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 22. Stream function contours are periodic and unsteady at Re = 15000 for Case T2a: (a) Stream function contours at time steps $ t=6000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 23 Case T2a条件下, Re = 15000时的各阶模态图　(a)速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 23. Each order of mode for Case T2a when Re = 15000: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 24 Case T2a条件下, Re = 20000时呈现非周期非稳态流动　(a)时间步$ t=6000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 24. Stream function contours are non-periodic and unsteady at Re = 20000 for Case T2a: (a) Stream function contours at time steps $ t=6000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 25 Case T2a条件下, Re = 20000时的各阶模态图　(a)速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 25. Each order of mode for Case T2a when Re = 20000: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity.

下载: 全尺寸图片幻灯片

图 26 Case T2a条件下, Re = 25000时呈非周期非稳态流动　(a)时间步$ t=8000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 26. Stream function contours are non-periodic and unsteady at Re = 25000 for Case T2a: (a) Stream function contours at time steps $ t=8000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 27 Case T2a条件下, Re = 25000时的各阶模态图　(a)速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 27. Each order of mode for Case T2a when Re = 25000: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 28 Case T2b条件下, Re为1000—5000时的流场形态　(a) Re = 1000; (b) Re = 3200; (c) Re = 5000

Fig. 28. Stream function contours of Reynolds numbers from 1000 to 5000 for Case T2b: (a) Re = 1000; (b) Re = 3200; (c) Re = 5000

下载: 全尺寸图片幻灯片

图 29 Case T2b条件下, Re为1000—5000时, (a) $ x=1 $的中线上, 纵坐标y 和水平速度u的关系; (b) $ y=0.5 $的中线上, 横坐标x和竖直速度v的关系

Fig. 29. (a) Variation of the velocity u magnitude along $ x=1 $, (b) variation of the velocity v magnitude along $ y=0.5 $ at Re = 1000–5000 for Case T2b.

下载: 全尺寸图片幻灯片

图 30 Case T2b条件下, Re = 6000时呈现周期性非稳态流动　(a)时间步$ t=3000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 30. Stream function contours are periodic and unsteady at Re = 6000 for Case T2b: (a) Stream function contours at time steps $ t=3000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 31 Case T2b条件下, Re = 6000时的各阶模态图　(a)速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 31. Each order of mode for Case T2b when Re = 6000: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 32 Case T2b条件下, Re = 8000时呈现非周期非稳态流动　(a)时间步$ t=5000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间变化图

Fig. 32. Stream function contours are non-periodic and unsteady at Re = 8000 for Case T2b: (a) Stream function contours at time steps $ t=5000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 33 Case T2b条件下, Re = 8000时的各阶模态图　(a)速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 33. Each order of mode for Case T2b when Re = 8000: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 34 Case T2b条件下, Re = 10000时呈现非周期性非稳态流动　(a)时间步$ t=10000000 $时的流场图; (b)监控点处速度相图; (c)速度监控点处速度随时间的变化

Fig. 34. Stream function contours are non-periodic and unsteady at Re = 10000 for Case T2b: (a) Stream function contours at time steps $ t=10000000 $; (b) velocity phase diagram at the velocity monitoring point; (c) velocity over time at the velocity monitoring point

下载: 全尺寸图片幻灯片

图 35 Case T2b条件下, Re = 10000时的各阶模态图　(a) 速度平均场模态图; (b)前n阶模态的能量占比; (c)各阶模态的能量占比; (d)—(f)一阶、二阶、三阶速度模态图

Fig. 35. Each order of mode for Case T2b when Re = 10000: (a) The mean field modal diagrams of velocity; (b) energy share of the first n order of mode for velocity; (c) energy share of each order of mode for velocity; (d)–(f) the first-order, second-order and third-order mode of velocity

下载: 全尺寸图片幻灯片

图 36 4种驱动情况下, Re对流场形态的影响图. 黑色圆圈表示流场处于稳态流动状态, 红色三角形表示流场处于周期性非稳态流动状态, 蓝色方框表示流场处于非周期性非稳态流动状态

Fig. 36. Effect of Reynolds number on the flow pattern for four different cases. The black circle indicates that the flow field is steady, the red triangle shows that the flow field is periodic but still unsteady, and the blue box indicates that the flow field is non-periodic and unsteady

下载: 全尺寸图片幻灯片

表 1 CPU, OpenAcc, CUDA的计算时间和加速比

Table 1. Compute time and acceleration ratio of CPU, OpenAcc, CUDA

计算方式	CPU	OpenAcc	CUDA	加速比(CPU/OpenAcc)	加速比(CPU/CUDA)
计算时间/s	6459.28	408.75	48.95	15.80	131.96

下载: 导出CSV

表 2 对比Re = 100, 1000, 3200时流场中的涡心位置

Table 2. Comparison of the vortex center position in the cavity at Re = 100, 1000, 3200

Re		$x_1$	$y_1$	$x_2$		$y_2$
Re		$x_1$	$y_1$	$x_{\rm 2 l}$	$x_{\rm 2 r}$	$y_{\rm 2 l}$	$y_{\rm 2 r}$
100	Zhang et al.^[37]	0.5721	0.4212	—	—	—	—
	Present	0.5722	0.4210	—	—	—	—
	Error/%	0.03	0.05	—	—	—	—
1000	Zhang et al.^[37]	0.5473	0.3558	0.3423	0.0180	0.6351	0.0450
	Present	0.5479	0.3561	0.3428	0.0179	0.6370	0.0451
	Error/%	0.11	0.09	0.15	0.42	0.30	0.12
3200	Zhang et al.^[37]	0.714	0.4392	0.3378	0.3491	0.4504	0.0788
	Present	0.7225	0.4448	0.3427	0.3432	0.4539	0.0809
	Error/%	1.18	1.28	1.46	1.70	0.77	2.66

下载: 导出CSV

表 3 Case T1a条件下, Re = 1000, 3200, 5000, 6000, 8000时流场中一级涡、二级涡和三级涡的涡心位置

Table 3. The first, second and third primary eddies characteristics at Re = 1000, 3200, 5000, 6000, 8000 for Case T1a.

Re	$x_1$	$y_1$	$x_2$		$y_2$		$x_3$		$y_3$
Re	$x_1$	$y_1$	$x_{\rm 2 l}$	$x_{\rm 2 r}$	$y_{\rm 2 l}$	$y_{\rm 2 r}$	$x_{\rm 3 l}$	$x_{\rm 3 r}$	$y_{\rm 3 l}$	$y_{\rm 3 r}$
1000	286.89	148.44	153.65	358.66	9.93	11.39	—	—	—	—
3200	359.03	180.97	163.84	158.09	142.90	15.17	—	—	—	—
5000	192.02	134.52	241.44	384.21	241.30	192.66	155.94	347.52	13.77	25.36
6000	200.42	132.50	261.28	393.16	242.12	196.81	155.06	353.98	12.93	35.50
8000	210.51	130.69	290.17	405.75	243.39	202.60	154.62	358.04	12.48	37.51

下载: 导出CSV

表 4 Case T1b条件下, Re = 1000, 3200, 5000, 6000, 8000时流场中一级涡, 二级涡和三级涡的涡心位置

Table 4. The first, second and third primary eddies characteristics at Re = 1000, 3200, 5000, 6000, 8000 for Case T1b.

Re	$x_1$	$y_1$	$x_2$		$y_2$		$x_3$		$y_3$
Re	$x_1$	$y_1$	$x_{\rm 2 l}$	$x_{\rm 2 r}$	$y_{\rm 2 l}$	$y_{\rm 2 r}$	$x_{\rm 3 l}$	$x_{\rm 3 r}$	$y_{\rm 3 l}$	$y_{\rm 3 r}$
1000	282.15	113.99	102.26	426.01	194.18	207.84	17.85	492.65	237.54	247.61
3200	280.78	118.95	110.43	412.34	158.93	218.72	25.99	487.95	229.41	245.42
5000	280.79	118.95	110.43	412.34	158.95	218.72	25.97	487.81	229.78	245.57
6000	280.79	118.95	110.43	412.34	158.94	218.73	26.10	487.84	229.99	245.57
8000	283.22	121.00	121.00	408.63	117.03	222.46	61.80	469.33	210.11	233.00

下载: 导出CSV

表 5 Case T2a条件下, Re = 1000, 3200, 5000, 6000, 8000, 10000时流场中的涡心位置

Table 5. The first, second and third primary eddies characteristics at Re = 1000, 3200, 5000, 6000, 8000, 10000 for Case T2a.

Re	$x_1$	$y_1$	$x_2$		$y_2$		$x_3$			$y_3$
			$x_{\rm 2 l}$	$x_{\rm 2 r}$	$y_{\rm 2 l}$	$y_{\rm 2 r}$	$x_{\rm 3 l}$	$x_{\rm 3 r}$		$y_{\rm 3 l}$	$y_{\rm 3 r}$
			$x_{\rm 2 l}$	$x_{\rm 2 r}$	$y_{\rm 2 l}$	$y_{\rm 2 r}$	$x_{\rm 3 l}$	$x_{\rm 3 r_1}$	$x_{\rm 3 r_2}$	$y_{\rm 3 l}$	$y_{\rm 3 r_1}$	$y_{\rm 3 r_2}$
1000	265.65	83.21	166.69	390.56	209.14	202.40	—	412.69	422.93	—	101.61	117.56
3200	276.64	90.43	139.51	404.39	190.46	205.49	34.89	410.36	421.46	216.82	115.62	134.97
5000	282.10	92.07	141.76	408.64	186.10	206.44	40.16	410.52	421.23	222.49	120.92	138.91
6000	284.21	92.69	142.88	410.13	184.42	206.76	41.25	410.77	421.29	223.56	122.65	140.12
8000	287.07	93.68	144.54	412.15	182.08	207.14	42.18	411.69	422.02	224.22	124.32	141.38
10000	288.94	94.48	145.67	413.49	180.45	207.33	43.32	412.89	423.35	223.99	124.48	141.80

下载: 导出CSV

表 6 Case T2b条件下, Re = 1000, 3200, 5000时流场中的涡心位置

Table 6. The first, second and third primary eddies characteristics at Re = 1000, 3200, 5000 for Case T2b.

Re	$x_1$	$y_1$	$x_2$	$y_2$	$x_3$		$y_3$
Re	$x_1$	$y_1$	$x_2$	$y_2$	$x_{3 {\rm l}}$	$x_{3 {\rm r} }$	$y_{3 {\rm l} }$	$y_{3 {\rm r}}$
1000	259.99	132.44	59.83	183.94	—	—	—	—
3200	260.97	131.12	66.16	206.26	84.50	369.55	245.42	14.67
5000	261.42	130.74	65.46	208.22	88.80	365.54	246.86	14.39

下载: 导出CSV

[1]	Peng Y F, Shiau Y H, Hwang R R 2003 Comput. Fluids 32 337 Google Scholar
[2]	Isaev S A, Baranov P A, Sudakov A G, et al. 2008 Thermophys. Aeromech. 15 463 Google Scholar
[3]	Balachandar S, Eaton J K 2010 Annu. Rev. Fluid Mech. 42 111 Google Scholar
[4]	Kuhlmann H C, Romanò F 2019 The Lid-Driven Cavity (Austria: Springer International Publishing) p233
[5]	Nonino C, Del Giudice S 1988 Int. J. Numer. Methods Eng. 25 313 Google Scholar
[6]	Alleborn N, Raszillier H, Durst F 1999 Int. J. Heat Mass Tran. 42 833 Google Scholar
[7]	Talebi F, Mahmoudi A H, Shahi M 2010 Int. Commun. Heat Mass 37 79 Google Scholar
[8]	Chowdhury M, Kumar B V R 2023 J. Nonnewton. Fluid Mech. 312 104975 Google Scholar
[9]	Ribbens C J, Watson L T, Wang C Y 1994 J. Comput. Phys. 112 173 Google Scholar
[10]	McQuain W D, Ribbens C J, Wang C Y, et al. 1994 Comput. Fluids 23 613 Google Scholar
[11]	Ismael M A, Chamkha A J 2015 Numer. Heat Transfer, Part A 68 312 Google Scholar
[12]	Mebarek-Oudina F, Laouira H, Hussein A K, et al. 2022 Mathematics 10 929 Google Scholar
[13]	Mondal P, Mahapatra T R 2021 Int. J. Mech. Sci. 208 106665 Google Scholar
[14]	Kareem A K, Mohammed H A, Hussein A K, et al. 2016 Int. Commun. Heat Mass 77 195 Google Scholar
[15]	Rashad A M, Sivasankaran S, Mansour M A, et al. 2017 Numer. Heat Transfer, Part A 71 1223 Google Scholar
[16]	Darr J H, Vanka S P 1991 Phys. Fluids 3 385 Google Scholar
[17]	Paramane S B, Sharma A 2008 Numer. Heat Transfer, Part B 54 84 Google Scholar
[18]	张恒, 任峰, 胡海豹 2021 物理学报 70 184703 Google Scholar Zhang H, Ren F, Hu H B 2021 Acta Phys. Sin. 70 184703 Google Scholar
[19]	Yang F, Shi X, Guo X, et al. 2012 Energy Procedia 16 639 Google Scholar
[20]	Nagendra K, Lakshmisha K N 2009 Int. J. Numer. Methods Heat Fluid Flow 19 790 Google Scholar
[21]	Perumal D A, Dass A K 2011 Comput. Math. Appl. 61 3711 Google Scholar
[22]	Nemati H, Farhadi M, Sedighi K, et al. 2010 Int. Commun. Heat Mass 37 1528 Google Scholar
[23]	Guo X, Zhong C, Zhuo C, et al. 2014 Theor. Comput. Fluid Dyn. 28 215 Google Scholar
[24]	Patil D V, Lakshmisha K N, Rogg B 2006 Comput. Fluids 35 1116 Google Scholar
[25]	Hou S, Zou Q, Chen S, et al. 1995 J. Comput. Phys. 118 329 Google Scholar
[26]	Chai Z H, Shi B C, Zheng L 2006 Chin. Phys. 15 1855 Google Scholar
[27]	Yuana K A, Budiana E P 2019 IOP Conf. Ser.: Mater. Sci. Eng. 546 052088 Google Scholar
[28]	Perumal D A, Dass A K 2008 WIT Trans. Eng. Sci. 59 45 Google Scholar
[29]	Bo A N, Mellibovsky F, Bergada J M, et al. 2020 Appl. Math. Modell. 82 469 Google Scholar
[30]	Shahid H, Yaqoob I, Khan W A, et al. 2021 Int. Commun. Heat Mass 129 105658 Google Scholar
[31]	Perumal D A 2018 Therm. Sci. Eng. Prog. 6 48 Google Scholar
[32]	Guo Y, Bennacer R, Shen S, et al. 2010 Int. J. Numer. Methods Heat Fluid Flow 20 130 Google Scholar
[33]	Shahid H, Yaqoob I, Khan W A, et al. 2021 Int. Commun. Heat Mass 128 105552 Google Scholar
[34]	An B, Bergada J M, Mellibovsky F 2019 J. Fluid Mech. 875 476 Google Scholar
[35]	An B, Guo S, Bergadà J M 2023 Appl. Sci. 13 888 Google Scholar
[36]	Sidik N A C, Munir F A 2012 Arab. J. Sci. Eng. 37 1723 Google Scholar
[37]	Zhang T, Shi B, Chai Z 2010 Compu. Fluids 39 1977 Google Scholar
[38]	Qian Y H, D’Humières D, Lallemand P 1992 EPL 17 479 Google Scholar
[39]	Ziegler D P 1993 J. Stat. Phys. 71 1171 Google Scholar
[40]	Guo Z L, Zheng C G, Shi B C 2002 Chin. Phys. 11 366 Google Scholar
[41]	Guo Z L, Zheng C G, Shi B C 2002 Phys. Fluids 14 2007 Google Scholar

[1]	赖瑶瑶, 陈鑫梦, 柴振华, 施保昌. 基于格子Boltzmann方法的钉扎螺旋波反馈控制. 物理学报, 2024, 73(4): 040502. doi: 10.7498/aps.73.20231549
[2]	张恒, 任峰, 胡海豹. 基于格子Boltzmann方法的幂律流体二维顶盖驱动流转捩研究. 物理学报, 2021, 70(18): 184703. doi: 10.7498/aps.70.20210451
[3]	臧晨强, 娄钦. 复杂微通道内非混相驱替过程的格子Boltzmann方法. 物理学报, 2017, 66(13): 134701. doi: 10.7498/aps.66.134701
[4]	王佐, 张家忠, 王恒. 非正交多松弛系数轴对称热格子Boltzmann方法. 物理学报, 2017, 66(4): 044701. doi: 10.7498/aps.66.044701
[5]	周光雨, 陈力, 张鸿雁, 崔海航. 基于格子Boltzmann方法的自驱动Janus颗粒扩散泳力. 物理学报, 2017, 66(8): 084703. doi: 10.7498/aps.66.084703
[6]	张娅, 潘光, 黄桥高. 疏水表面减阻的格子Boltzmann方法数值模拟. 物理学报, 2015, 64(18): 184702. doi: 10.7498/aps.64.184702
[7]	黄桥高, 潘光, 宋保维. 疏水表面滑移流动及减阻特性的格子Boltzmann方法模拟. 物理学报, 2014, 63(5): 054701. doi: 10.7498/aps.63.054701
[8]	任晟, 张家忠, 张亚苗, 卫丁. 零质量射流激励下诱发液体相变及其格子Boltzmann方法模拟. 物理学报, 2014, 63(2): 024702. doi: 10.7498/aps.63.024702
[9]	解文军, 滕鹏飞. 声悬浮过程的格子Boltzmann方法研究. 物理学报, 2014, 63(16): 164301. doi: 10.7498/aps.63.164301
[10]	史冬岩, 王志凯, 张阿漫. 任意复杂流-固边界的格子Boltzmann处理方法. 物理学报, 2014, 63(7): 074703. doi: 10.7498/aps.63.074703
[11]	刘邱祖, 寇子明, 贾月梅, 吴娟, 韩振南, 张倩倩. 改性疏水固壁润湿性反转现象的格子Boltzmann方法模拟. 物理学报, 2014, 63(10): 104701. doi: 10.7498/aps.63.104701
[12]	郭亚丽, 徐鹤函, 沈胜强, 魏兰. 利用格子Boltzmann方法模拟矩形腔内纳米流体Raleigh-Benard对流. 物理学报, 2013, 62(14): 144704. doi: 10.7498/aps.62.144704
[13]	曾建邦, 李隆键, 蒋方明. 气泡成核过程的格子Boltzmann方法模拟. 物理学报, 2013, 62(17): 176401. doi: 10.7498/aps.62.176401
[14]	刘邱祖, 寇子明, 韩振南, 高贵军. 基于格子Boltzmann方法的液滴沿固壁铺展动态过程模拟. 物理学报, 2013, 62(23): 234701. doi: 10.7498/aps.62.234701
[15]	曾建邦, 李隆键, 廖全, 蒋方明. 池沸腾中气泡生长过程的格子Boltzmann方法模拟. 物理学报, 2011, 60(6): 066401. doi: 10.7498/aps.60.066401
[16]	曾建邦, 李隆键, 廖全, 陈清华, 崔文智, 潘良明. 格子Boltzmann方法在相变过程中的应用. 物理学报, 2010, 59(1): 178-185. doi: 10.7498/aps.59.178
[17]	石自媛, 胡国辉, 周哲玮. 润湿性梯度驱动液滴运动的格子Boltzmann模拟. 物理学报, 2010, 59(4): 2595-2600. doi: 10.7498/aps.59.2595
[18]	卢玉华, 詹杰民. 三维方腔温盐双扩散的格子Boltzmann方法数值模拟. 物理学报, 2006, 55(9): 4774-4782. doi: 10.7498/aps.55.4774
[19]	李华兵, 黄乒花, 刘慕仁, 孔令江. 用格子Boltzmann方法模拟MKDV方程. 物理学报, 2001, 50(5): 837-840. doi: 10.7498/aps.50.837
[20]	吕晓阳, 李华兵. 用格子Boltzmann方法模拟高雷诺数下的热空腔黏性流. 物理学报, 2001, 50(3): 422-427. doi: 10.7498/aps.50.422

计量

文章访问数: 2766
PDF下载量: 59
被引次数: 0

姓名
邮箱
手机号码
标题
留言内容
验证码

搜索

留言板