CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   directMappedPatch for LES (http://www.cfd-online.com/Forums/openfoam-solving/64182-directmappedpatch-les.html)

syoon May 1, 2009 11:05

directMappedPatch for LES
 
I’m simulating a backward facing step using LES (oodles).
In order to generate inlet turbulence, I’m using directMappedPatch.
Studying oodles/pitzDailyDirectMapped was helpful to understand the structure.

My case is slightly different from pitzDailyDirectMapped in the sense that the upperWall is not a wall but a symmetryPlane. (I’m simulating a half of a double-sided expansion).
So I need to use either “symmetryPlane” or “slip” condition on the upperWall rather than “wall” condition.

My question is
  • Does the use of “symmetryPlane” or “slip” conditions on the upperWall can damage the use of directMappedPatch?
  • Between “symmetryPlane” and “slip” condition, what is more suitable to simulate a half of a double-sided expansion? In fact, I don’t understand clearly the difference between “symmetryPlane” and “slip”.
Many thanks,
Sungho

lakeat May 6, 2009 21:14

Hi men, you asked the problem for many times, still no answer? :)
Well, I think this helps:
Many CFD-codes assumes a zero flux of all quantities across a symmetry boundary. There is no convective flux across a symmetry plane: the normal velocity component at the symmetry plane is thus zero. There is no di ffusion flux across a symmetry plane: the normal gradients of all flow variables are thus zero at the symmetry plane. The symmetry boundary condition can therefore be summarized as follows:
  •  zero normal velocity at a symmetry plane
  •  zero normal gradients of all variables at a symmetry plane
As stated above, these conditions determine a zero flux across the symmetry plane, which is required by the defi nition of symmetry. Since the shear stress is zero at a symmetry boundary, it can also be interpreted as a "slip" wall when used in viscous flow calculations.

braennstroem May 8, 2009 06:29

Hi,

so a symmetryPlane is not suitable for your kind of flow using LES. The fluctuations will cancel out...

Fabian

lakeat May 8, 2009 06:35

Quote:

Originally Posted by braennstroem (Post 215501)
Hi,

so a symmetryPlane is not suitable for your kind of flow using LES. The fluctuations will cancel out...

Fabian

Agree......

syoon May 8, 2009 09:03

Thank you, Daniel and Fabian.

In fact, I'm not much worried about cancelled turbulence near the symmetry plane as long as I can generate turbulent flows near the lower wall which seems to affect the reattachment point after the step.

One big problem I encountered with symmetry (or slip) condition was that the u velocity becomes smaller and smaller near the symmetry wall as the time step increases. On the other hand the velocity near the lower wall (slightly away from the wall) becomes higher and higher with the time step.
I couldn't get a well developed turbulent (mean) velocity profile at all.
So I wondered and still wonder whether directMapped patch should be applied to a type of channel flows (I mean, wall conditions at both upper and lower boundaries)

Now, it seems to me I need to simulate a full double expansion with wall conditions at both upper and lower boundaries (rather than a half expansion with symmetry condition) and I'm currently doing it although it requires more cells and computation time.

Many thanks,

Sungho

gaby July 13, 2009 10:58

Hi! :)

Do you know if the 'directMappedPatch' utility uses the method proposed by Lund, T.S. et al (Generation of turbulent inflow data for spatially-developing boundary layer simulations. J.Comp.Phys.140, 233-258. 1998). Or it is something different?

I read Eugene's thesis, but I didn't find any previous reference about it...

Could anybody (or Eugene) give me some clue?

Thanks in advance!

Gaby

panda60 January 29, 2010 04:11

Quote:

Originally Posted by gaby (Post 222496)
Hi! :)

Do you know if the 'directMappedPatch' utility uses the method proposed by Lund, T.S. et al (Generation of turbulent inflow data for spatially-developing boundary layer simulations. J.Comp.Phys.140, 233-258. 1998). Or it is something different?

I read Eugene's thesis, but I didn't find any previous reference about it...

Could anybody (or Eugene) give me some clue?

Thanks in advance!

Gaby

Dear Gaby, I am also interseting in this kind of mapping.
I want to use like this:

u(inlet)=<u>(z)+f(z)*(u-<u>)recy

That meas not simplely mapping, but use mapping position's value subtract mean value and then multiply an function f(z), to the inlet. which z is coordinate.

I think this is not so difficult to modify 'directMappedPatch' to this, but I am beginer, so I can't modify, If somebody can do this work, it will be appreciate.

Fransje November 14, 2010 12:03

Hello Gaby,

Contrary to the method of Lund etal, there is no rescaling of velocity implemented in the directMapped boundary conditions. It is merely a copy from a plane in the domain to the inlet. If you want a rescaling of some kind, you will have to implement it yourself starting from one of the directMapped* boundary conditions.

Kind regards,

Francois.

Quote:

Originally Posted by gaby (Post 222496)
Hi! :)

Do you know if the 'directMappedPatch' utility uses the method proposed by Lund, T.S. et al (Generation of turbulent inflow data for spatially-developing boundary layer simulations. J.Comp.Phys.140, 233-258. 1998). Or it is something different?

I read Eugene's thesis, but I didn't find any previous reference about it...

Could anybody (or Eugene) give me some clue?

Thanks in advance!

Gaby



All times are GMT -4. The time now is 14:21.