一种基于块雅可比迭代的高阶FR格式隐式方法

于要杰; 刘锋; 高超; 冯毅

doi:10.6052/0459-1879-20-404

一种基于块雅可比迭代的高阶FR格式隐式方法

doi: 10.6052/0459-1879-20-404

于要杰^*,**,
刘锋^†,
高超^*,
冯毅^**,2), ,

^*西北工业大学航空学院, 西安 710072
^†加州大学尔湾分校机械与航天工程系, 美国加利福尼亚州 92697-3975
^**成都流体动力创新中心, 成都 610071

基金项目: 1)国家重点研发计划资助项目(2018YFB1501102)

详细信息

作者简介:
2)冯毅, 副研究员, 主要研究方向: 飞行器设计, 计算流体力学. E-mail: fengyi0218@163.com

通讯作者:
冯毅

中图分类号: O355
计量
- 文章访问数: 646
- HTML全文浏览量: 77
- PDF下载量: 272
- 被引次数: 0
出版历程
- 收稿日期: 2020-11-30
- 刊出日期: 2021-06-01

AN IMPLICIT BLOCK JACOBI APPROACH FOR HIGH-ORDER FLUX RECONSTRUCTION METHOD

Yu Yaojie^*,**,
Liu Feng^†,
Gao Chao^*,
Feng Yi^{**,2)
, ,}

^*Department of Fluid Mechanics, Northwestern Polytechnical University, Xi'an 710072, China
^†Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA 92697-3975, USA
^**Chengdu Fluid Dynamics Innovation Center, Chengdu 610071, China

摘要

摘要: 最近, 基于非结构网格的高阶通量重构格式(flux reconstruction, FR)因其构造简单且通用性强而受到越来越多人的关注. 但将FR格式应用于大规模复杂流动的模拟时仍面临计算开销大、求解时间长等问题. 因此, 亟需发展与之相适应的高效隐式求解方法和并行计算技术. 本文提出了一种基于块Jacobi迭代的高阶FR格式求解定常二维欧拉方程的单GPU隐式时间推进方法. 由于直接求解FR格式空间和隐式时间离散后的全局线性方程组效率低下并且内存占用很大. 而通过块雅可比迭代的方式, 能够改变全局线性方程组左端矩阵的特征, 克服影响求解并行性的相邻单元依赖问题, 使得只需要存储和计算对角块矩阵. 最终将求解全局线性方程组转化为求解一系列局部单元线性方程组, 进而又可利用LU分解法在GPU上并行求解这些小型局部线性方程组. 通过二维无黏Bump流动和NACA0012无黏绕流两个数值实验表明, 该隐式方法计算收敛所用的迭代步数和计算时间均远小于使用多重网格加速的显式Runge-Kutta格式, 且在计算效率方面至少有一个量级的提升.
- 高精度格式 /
- 通量重构格式 /
- 隐式时间推进方法 /
- 块雅可比迭代 /
- GPU计算
Abstract: Recently, the flux reconstruction (FR) method has attracted more and more attentions for its simplicity and generality. However, it is still computationally expensive and time consuming when simulating the complex flow problems by FR method. There is a huge demand for developing appropriate efficient implicit solvers and parallel computing techniques for FR. This paper proposes an implicit high-order flux reconstruction solver on GPU platform based on the block Jacobi iteration method. As it is inefficient to solve the large global linear system resulting from spatial and implicit temporal discretization of FR directly. A block Jacobi approach is used to change the characteristics of the lift-hand matrix of the global linear system and this avoids the dependence of neighboring elements. Therefore, only the diagonal blocks of global matrix need to be stored and calculated. Then, the problem of solving the huge global linear system is transformed into solving a series of local linear equations simultaneously. Finally, these small local linear equations would be solved by the LU decomposition method in parallel on GPU platforms. Two typical cases, including subsonic flows over a bump and a NACA0012 airfoil, were simulated and compared with the multi-grid explicit Runge-Kutta scheme. The numerical results demonstrated that the present implicit method can reduce the iterations significantly. Meanwhile, the implicit solver has shown at least 10x speedup over the multi-grid Runge-Kutta scheme in all cases.
- high-order methods /
- flux reconstruction method /
- implicit time-marching methods /
- block Jacobi iteration /
- GPU computing

HTML全文

参考文献(39)

[1]	Slotnick J, Khodadoust A, Alonso J, et al. CFD Vision 2030 Study: A path to revolutionary computational aerosciences. NASA/CR-2014-218178, Washington: NASA, 2014
[2]	阎超, 屈峰, 赵雅甜等. 航空航天CFD物理模型和计算方法的述评与挑战. 空气动力学学报, 2020, 38(5): 829-857 (Yan Chao, Qu Feng, Zhao Yatian, et al. Review of development and challenges for physical modeling and numerical scheme of CFD in aeronautics and astronautics. Chinese Journal of Theoretical and Applied Mechanics, 2020, 38(5): 829-857 (in Chinese))
[3]	Jiang GS, Shu CW. Efficient implementation of weighted ENO schemes. Journal of Computational Physics, 1996, 126(1): 202-228
[4]	Deng XG, Zhang HX. Developing high-order weighted compact nonlinear schemes. Journal of Computational Physics, 2000, 165(1): 22-44
[5]	Tsoutsanis P, Titarev A, Drikakis D. WENO schemes on arbitrary mixed-element unstructured meshes in three space dimensions. Journal of Computational Physics, 2011, 230(4): 1585-1601
[6]	Wang Q, Ren YX, Pan J, et al. Compact high order finite volume method on unstructured grids III: Variational reconstruction. Journal of Computational Physics, 2017, 337: 1-26
[7]	Cockburn B, Lin SY, Shu CW. TVB Runge-Kutta local projection discontinuous Galerkin finite element method for conservation laws III: One-dimensional systems. Journal of Computational Physics, 1989, 84(1): 90-113
[8]	Cockburn B, Shu CW. Runge-Kutta discontinuous Galerkin methods for convection-dominated problems. Journal of Scientific Computing, 2001, 16(3): 173-261
[9]	Wang ZJ. Spectral (finite) volume method for conservation laws on unstructured grids: Basic formulation. Journal of Computational Physics, 2002, 178(1): 210-251
[10]	Liu Y, Vinokur M, Wang ZJ. Spectral difference method for unstructured grids I: Basic formulation. Journal of Computational Physics, 2006, 216(2): 780-801
[11]	Huynh HT. A flux reconstruction approach to high-order schemes including discontinuous Galerkin methods//18th AIAA Computational Fluid Dynamics Conference, 2007
[12]	Huynh HT. A reconstruction approach to high-order schemnes including discontinuous Galerkin for diffusion//47th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition, 2009
[13]	Wang ZJ, Gao H. A unifying lifting collocation penalty formulation including the discontinuous Galerkin, spectral volume/difference methods for conservation laws on mixed grids. Journal of Computational Physics, 2009, 228(21): 8161-8186
[14]	Huynh HT, Wang ZJ, Vincent P. High-order methods for computational fluid dynamics: A brief review of compact differential formulations on unstructured grids. Computers & Fluids, 2014, 98: 209-220
[15]	Vincent P, Castonguay P, Jameson A. Insights from von Neumann analysis of high-order flux reconstruction schemes. Journal of Computational Physics, 2011, 230(22): 8134-8154
[16]	Lopez-Morales M, Bull J, Crabill J, et al. Verification and validation of HiFiLES: A high-order LES unstructured solver on multi-GPU platforms//32nd AIAA Applied Aerodynamics Conference, Atlanta, Georgia, USA, 2014: 16-20
[17]	Witherden F, Farrington A, Vincent P. PyFR: An open source framework for solving advection-diffusion type problems on streaming architectures using the flux reconstruction approach. Computer Physics Communications, 2014, 185(11): 3028-3040
[18]	Huynh HT. Flux reconstruction/correction procedure via reconstruction, a unified approach to high-order accuracy//37th Advanced VKI CFD Lecture Series: Recent Developments in Higher Order Methods and Industrial Application in Aeronautics, 2013
[19]	Allaneau Y, Jameson A. Connections between the filtered discontinuous Galerkin method and the flux reconstruction approach to high order discretizations. Computer Methods in Applied Mechanics and Engineering, 2011, 200(49-52): 3628-3636
[20]	De Grazia D, Mengaldo G, Moxey D, et al. Connections between the discontinuous Galerkin method and high-order flux reconstruction schemes. International Journal for Numerical Methods in Fluids, 2014, 75(12): 860-877
[21]	Gottlieb S, Shu CW. Total variation diminishing Runge-Kutta schemes. Mathematics of Computation, 1998, 67(221): 73-85
[22]	刘伟, 张来平, 赫新等. 基于Newton/Gauss-Seidel 迭代的高阶精度DGM隐式计算方法研究. 力学学报, 2012, 44(4): 792-796 (Liu Wei, Zhang Laiping, He Xin, et al. An implicit algorithm for high-order discontinuous Galerkin method based on Newton/Gauss-Seidel iterations. Chinese Journal of Theoretical and Applied Mechanics, 2017, 49(3): 627-637 (in Chinese))
[23]	Yu ML, Wang Lai. A high-order flux reconstruction/correction procedure via reconstruction formulation for unsteady incompressible flow on unstructured moving grids. Computers & Fluids, 2016, 139: 161-173
[24]	Vandenhoeck R, Lani A. Implicit high-order flux reconstruction solver for high-speed compressible flows. Computer Physics Communications, 2019, 242: 1-24
[25]	Yan ZG, Pan Y, Castiglioni G, et al. Nektar++: Design and implementation of an implicit, spectral/hp element, compressible flow solver using a Jacobian-free Newton Krylov approach. Computers and Mathematics with Applications, 2021, 81: 351-372
[26]	Watkins J, Romero J, Jameson A. Multi-GPU, implicit time stepping for high-order methods on unstructured grids//46th AIAA Fluid Dynamics Conference, 2016
[27]	Vincent P, Witherden F, Vermeire B, et al. Towards green aviation with python at petascale//SC16: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, Salt Lake City, UT, USA, IEEE, 2016: 1-11
[28]	徐传福, 车永刚, 李大力等. 天河超级计算机上超大规模高精度计算流体力学并行计算研究进展. 计算机工程与科学, 2020, 42(10): 1815-1826 (Xu Chuanfu, Che Yonggang, Li Dali, et al. Research progresses of large-scale parallel computing for high-order CFD on the Tianhe supercomputer. Computer Engineering & Science, 2020, 42(10): 1815-1826 (in Chinese))
[29]	Zimmerman B, Wang ZJ. The efficient implementation of correction procedure via reconstruction with graphics processing unit computing. Computers & Fluids, 2014, 101: 263-272
[30]	Romero J, Asthana K, Jameson A. A simplified formulation of the flux reconstruction method. Journal of Scientific Computing, 2015, 67(1): 351-374
[31]	Vermeire B, Witherden F, Vincent P. On the utility of GPU accelerated high-order methods for unsteady flow simulations: A comparison with industry-standard tools. Journal of Computational Physics, 2017, 334: 497-521
[32]	Jourdan E, Wang ZJ. Efficient implementation of the FR/CPR method on GPU clusters for industrial large eddy simulation// AIAA Aviation Forum, 2020
[33]	Roe PL. Approximate Riemann solvers, parameter vectors, and difference schemes. Journal of Computational Physics, 1981, 43(2): 357-372
[34]	Rusanov V. Calculation of interaction of non-steady shock waves with obstacles. Journal of Computational Math and Physics USSR, 1961, 1: 261-279
[35]	Sun Y, Wang ZJ, Liu Y. Efficient implicit non-linear LU-SGS approach for compressible flow computation using high-order spectral difference method. Communications in Computational Physics, 2013, 5(2): 760-778
[36]	Xia Y, Luo H, Frisbey M, et. al. A set of parallel, implicit methods for a reconstructed discontinuous galerkin method for compressible flows on 3d hybrid grids. Computers & Fluids, 2014, 98: 134-151
[37]	Pazner W, Persson P. Approximate tensor-product preconditioners for very high order discontinuous Galerkin methods. Journal of Computational Physics, 2018, 354: 344-369
[38]	Phipps E, Gay D. Sacado automatic differentiation package. http://trilinos.sandia.gov/packages/sacado/
[39]	https://how5.cenaero.be/