CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Wall function treatment (http://www.cfd-online.com/Forums/cfx/18671-wall-function-treatment.html)

VEEBS October 9, 2001 02:16

Wall function treatment
 
Hi All,

What is the use of fixed_wall_distance_model = t, function in tascflow. It is given it will work for y+ less than 100. Does that mean even if we have y+ less than 20 ( lower limit for standard log law wall function), calculation will be accurate ? Thanks for your time Veera

Robin Steed October 9, 2001 23:19

Re: Wall function treatment
 
Hi Veera,

With k-epsilon and a standard wall treatment, y+ values must be maintained between 20 and 100. If y+ is too large, the boundary layer will not be sufficiently resolved; if y+ is too small, the wall function approximation will be inaccurate. The latter becomes problematic if one wishes to perform grid refinement studies, because as more nodes are added and y+ drops too low, the results will get worse.

The fixed_wall_distance_model flag selects the scalable wall function model in CFX-TASCflow (also available in CFX-5). This formulation will not be adversely affected by small y+ values, but still requires that y+ be below 100. Thus the scalable wall function allows grids to be refined without violating any wall function assumptions.

The theory documentation provides detailed information about these and other turbulence models available.

Regards, Robin


All times are GMT -4. The time now is 18:54.