scholarly journals Subgrid-scale characterization and asymptotic behavior of multidimensional upwind schemes for the vorticity transport equations

2021 ◽  
Vol 6 (2) ◽  
Author(s):  
Daniel Foti ◽  
Karthik Duraisamy
1973 ◽  
Vol 95 (3) ◽  
pp. 429-438 ◽  
Author(s):  
J. W. Deardorff

A three-dimensional numerical model of turbulence in an atmospheric boundary layer has been revised to utilize subgrid transport equations for the subgrid Reynolds stresses and fluxes rather than subgrid eddy coefficients. It was applied to a daytime boundary layer over heated ground in a region of horizontal area 8km square and 2km deep, utilizing 40×40×40 grid points. The constraints involved in selecting four important subgrid closure constants are discussed in some detail, along with maintenance of realizability on the subgrid scale. The results indicate that the subgrid transport equations produce subgrid Reynolds stresses and fluxes which realistically simulate the transfer of larger scale variance to subgrid scales, provided truncation errors due to advective terms are not too large. They also show the superiority of this method over the use of (nonstability dependent) nonlinear eddy coefficients in maintaining the sharpness of the inversion base which lies above the mixed layer.


Author(s):  
Y. Morinishi ◽  
E. Nakamura

The objective of this study is to specify the wall asymptotic behavior of the subgrid scale stress for the large eddy simulation (LES) of wall bounded turbulent flows. The grid scale (GS) and subgrid scale (SGS) turbulent statistics are extracted from the numerical data of direct numerical simulation (DNS) of the turbulent channel flow at the Reynolds number of 300, based on the friction velocity and channel half width, in order to achieve this objective. The basic equations of the DNS are the Navier-Stokes and continuity equations for incompressible flows. The GS and SGS statistics are computed corresponding to various LES filters, that is, two-dimensional (plane) and three-dimensional filters with different filter widths. The Gaussian filter is used in the periodic directions. The filtering function in the wall-normal direction is top-hat for the three-dimensional filter. Then, the components of the SGS stress are compared with the analytical results by the Taylor expansion. The results estimated by the Taylor expansion and the DNS data indicate different wall asymptotic behaviors of the SGS stress for the plane and three-dimensional filters. In particular, the wall asymptotic behavior of the Reynolds stress, which has been assumed in the SGS modeling, is not true of the SGS stress for the three-dimensional filter.


Sign in / Sign up

Export Citation Format

Share Document