摘要
高精度算法和亚网格尺度(SGS)模型是大涡模拟的关键,对声源流场和气动声场的准确预测有着重要的影响。本文采用前人具有部分DNS数据的平板射流为考核算例,使用具有4阶空间精度和3阶时间精度的高精度算法,系统地比较Smagorinsky模型(SM)、选择多尺度模型(SMSM)、以及前人采用动态Smagorinsky模型(DSM)的计算结果。对射流速度、剪切层、湍流脉动以及涡结构的多尺度特性的结果分析表明,Smagorinsky模型的黏性耗散过大,而选择多尺度模型则能揭示转捩过程中的黏性变化,其结果比DSM更接近DNS的数据。同时,SMSM是局部模型,在大规模分布式内存的并行计算方面易于编程,具有良好的应用前景。
High resolution algorithms and SGS models are two important factors for the accuracy of large eddy simulation, they are also decisive to the accurate prediction of radiated aeroacoustic sound.In this paper, a plane jet with some DNS data is employed as a test case for validations of algorithm and SGS models. The Favre filtered Navier-Stokes equations are discretized by fourth-order spatial and third order Runge-Kutta temporal schemes, respectively. Comparisons of subgrid-scale（SGS）models are carried out between the standard Smagorinsky model（SM） and the selective mixed-scale model（SMSM）. Meanwhile, the results calculated by the dynamic Smagorinsky model（DSM） in literature are also included for a reference. The decay of jet centerline velocity, the growth of shear layer, the turbulent fluctuation and the multi-scale vortical structures are analyzed. The Smagorinsky model is shown to be too dissipative, while the selective mixed-scale model displays reasonable changes of viscosity during the flow transition. In the meantime, the results predicted using the SMSM are closer to the DNS data, as compared with those calculated by the DSM. Besides, SMSM is a localized SGS model, which is generally preferred in coding for large-scale parallel computation.
作者
刘琪麟
何诚
赖焕新
LIU Qi-Lin;HE Cheng;LAI Huan-Xin(School of MechanicM and Power Engineering, East China University of Science and Technology, Shanghai 200237, Chin)
出处
《工程热物理学报》
EI
CAS
CSCD
北大核心
2018年第6期1272-1278,共7页
Journal of Engineering Thermophysics
基金
国家自然科学基金项目(No.51576067)
关键词
射流
大涡模拟
局部模型
选择混合尺度模型
jet flow
large eddy simulation
localized SGS model
selective mixed-scale model
