CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FloEFD, FloWorks & FloTHERM (
-   -   Boundaries in external flow (

crzbur August 10, 2009 09:22

Boundaries in external flow

I'm new in this community and first of all I would like to thank the administrator and team for this wonderful site dedicated to CFD.

I have to model an external flow on a plate at a certain speed. The plate simulates a flat-plate solar panel, and i want to set one of the walls as a no-slip boundary, so that the flow velocity be 0 there. The thing is that i just find 'default', 'symmetry' and 'periodicity' as the three possible options. How can i fix that problem?


j.gerber August 12, 2009 05:32


The options "default", "symmetry" and "periodicity" are settings at the minimum and maximum of the three coordinate axes of the Computational Domain.

The default velocity boundary condition at solid walls is the no-slip condition.
For a slip condition you would have to specify an ideal wall boundary condition.

I hope this helps you.

crzbur August 13, 2009 05:10


Well, the thing is that when i do a cut plot of the velocities at the tunnel floor, it appears that the velocity is similar to the inlet velocity specified at the inlet lid.

i kind of get confused with that. do you know if i'm doing something wrong or if i miss something?

j.gerber August 13, 2009 06:28

Which analysis type have you specified under General Settings, External or Internal?

crzbur August 13, 2009 06:54

internal... that's why it kind of gets me confused. it should behave like the flow though a pipe... but don't know why it does not. jesus!

j.gerber August 13, 2009 09:42

Have you got heat transfer in solids enabled?
And did you apply simmetry at any of the computational walls?

crzbur August 13, 2009 10:18

I didn't activate the option of heat transfer. If so, which material should i pick up from the list? The material of the wind tunnel could be interpreted as an insulator?

And yeah, i applied symmetry to the side walls of the tunnel (computational domain). The rest are under the default condition.

Thanks. It seems you know what's gotta be done!

Im trying with heat transfer activated right now to see if anything changes...

j.gerber August 13, 2009 10:51

You probably need to adjust your mesh at the tunnel walls in order detect the variation in velocity.

Compare the size of the cells at the tunnel walls with the boundary layer thickness. This could give you an indication of how much refinement is needed.

crzbur August 13, 2009 11:14

I did a cut plot of velocities at the bottom boundary of the CD and checked the "display boundary layer" button and got it. I could see the velocity there was 0, as I had set the default condition there. _Thank you very much for your help, j.gerber.

But my concern now is this one: In an external flow analysis, how can I set a velocity profile (given by a power law expression)?? So if i set different cut plots of the velocity at different heights I could see increasing speed with an increasing height measured from the bottom.

j.gerber August 18, 2009 11:37

My apologies for my previous statement. FloEFD uses modified wall functions to characterize flow at the walls, thus a finer mesh in these regions are not needed to detect the velocity variation.

Are you saying that you need to set up a velocity profile at the wall irrespective of what the calculated profile should be?

Say if you make a horizontal cut plot close to the floor, just use the "Display Boundary Layer" option again and apply the plot. It seems to not show the boundary layer variation in velocity in preview mode.

crzbur August 19, 2009 15:06

yeah. I think FloWorks has a default uniform velocity profile, as opposed to a parabolic-kinda profile like the one in pipes flow. I am running now an external analysis, and was wondering if i could simulate the atmospheric boundary layer, where the velocity is 0 at the earth's surface but grows exponentially according to the power law (and consequently, the "ceiling" on the computational domain must be set up so that the velocity is not 0 there).

And about the display boundary layer, i could see how the velocity was 0 or almost 0 in the computational domain bottom boundary but if i applied another cut plot just 0.1 meters above, the velocity was almost fully developed, at the inlet's velocity.

j.gerber, you spanish?

All times are GMT -4. The time now is 18:16.