What does "direct" turbulence input type mean?
Hi all,
In BC - surfaces - choose an inflow surface - simple boundary condition, there is a 'turbulence input type'.
1) What does the choice of 'direct' mean?
2) Does 'direct' mean direct numerical simulation? (Which means that there is no turbulence)
3) If I choose 'Intensity and length scale', what is the proper dimension for the length scale? ( I am modelling a 2D 4m*4m square in a large rectangular fluid domain, interested in the drag force)
Any help is highly appreciated.
Thanks all.
Fangyuan
Answers
-
When performing RANS or DES analyses, AcuSolve solves additional partial differential equations to model the impact of turbulence. So, at inlet boundaries, you need to assign suitable boundary conditions to the variables transported by the turbulence model equations (turbulent kinetic energy, eddy frequency, eddy viscosity, etc.). When you select 'direct' as the input method, that means you will manually enter the values for each of these variables at the boundary. However, since it is often uncommon to know what these values should actually be, the options of 'Intensity and Length Scale', etc. are available to compute an estimate of the turbulence variables based on other characteristics of the geometry/flow.
If you are modeling an internal flow, then the length scale corresponds to the largest scale eddy that can be present at the inlet. For the case of a duct you can use the height of the duct as a first approximation.
0 -
Altair Forum User said:
When performing RANS or DES analyses, AcuSolve solves additional partial differential equations to model the impact of turbulence. So, at inlet boundaries, you need to assign suitable boundary conditions to the variables transported by the turbulence model equations (turbulent kinetic energy, eddy frequency, eddy viscosity, etc.). When you select 'direct' as the input method, that means you will manually enter the values for each of these variables at the boundary. However, since it is often uncommon to know what these values should actually be, the options of 'Intensity and Length Scale', etc. are available to compute an estimate of the turbulence variables based on other characteristics of the geometry/flow.
If you are modeling an internal flow, then the length scale corresponds to the largest scale eddy that can be present at the inlet. For the case of a duct you can use the height of the duct as a first approximation.
Thank you cfdguru for your quick reply.
1)When I chose 'direct' for the turbulence input, I didn't see any place pops up asking for the variables (eg. turbulent kinetic energy, eddy frequency), only eddy viscosity is available. Is the setting somewhere else in the GUI? Or setting eddy viscosity is enough?
2) If to choose 'direct', how can I find out what turbulence intensity it is using?
3) My case is an external flow over a square obstacle. Shall I use the dimension of the square side for turbulence length scale?
Thank you very much.
Regards,
Fangyuan
0 -
1. With 'Direct' the input depends on the turbulence model being used.
If SA model is used, then eddy viscosity is visible
if K-Omega is used, then Kinetic energy and eddy freq are visible
2. I guess the intensity can be calculated based upon a calculator like on
http://www.cfd-online.com/Tools/turbulence.php
3- Not sure. It seems there are no exact recommendations in literature for this.
0