understanding kEpsilon model source code

jet · November 8, 2009, 13:55

Hi,

I have a question about the implementation of the kEpsilon source code in the file kEpsilon.C (.../src/turbulenceModels/incompressible/RAS/kEpsilon). In line 181 the divDevReff()-function is declared. Can anybody explain the code in line 185 and 186 to me?

Thanks!

chingyinh · November 11, 2009, 23:25

Can you post that lines out?

jet · November 14, 2009, 17:22

Hi, I don't understand the following lines:

return
(
-laplacian(nuEff(), U)
-div(nuEff()*dev(grad(U)().T()))
);

The whole expression is defined as divDevReff() and should return the source term for the momentum equation. I tried to write it down mathematically but I failed!

Thanks for your help!

sega · November 15, 2009, 05:14

Some people don't know that dev() is a transposition!
Well - at least it was me who doesn't know until I was told.

But what is this T() ?

jet · November 15, 2009, 09:21

Hmm, are you sure that dev() is a transposition? I think that .T() is a transposition and dev() ist the deviatoric component ( tensor -1/3*(trace tensor)*I ) of the tensor grad(U)().T()?

But I wonder if dev() shouldn't be dev2()?

sega · November 15, 2009, 09:52

Actually your statement makes more sense.
I haven't thought about the fact that grad(U) is a tensor and not a vector like U itself!

November 8, 2009, 13:55	understanding kEpsilon model source code	#1
jet New Member Join Date: Nov 2009 Posts: 12 Rep Power: 17	Hi, I have a question about the implementation of the kEpsilon source code in the file kEpsilon.C (.../src/turbulenceModels/incompressible/RAS/kEpsilon). In line 181 the divDevReff()-function is declared. Can anybody explain the code in line 185 and 186 to me? Thanks!

November 15, 2009, 05:14		#4
sega Senior Member Sebastian Gatzka Join Date: Mar 2009 Location: Frankfurt, Germany Posts: 729 Rep Power: 20	Some people don't know that dev() is a transposition! Well - at least it was me who doesn't know until I was told. But what is this T() ? __________________ Schrödingers wife: "What did you do to the cat? It's half dead!"

November 15, 2009, 09:52		#6
sega Senior Member Sebastian Gatzka Join Date: Mar 2009 Location: Frankfurt, Germany Posts: 729 Rep Power: 20	Actually your statement makes more sense. I haven't thought about the fact that grad(U) is a tensor and not a vector like U itself! __________________ Schrödingers wife: "What did you do to the cat? It's half dead!"

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
add a source term in kEpsilon model	maurice	OpenFOAM Running, Solving & CFD	0	May 18, 2009 08:54
add a source term in incompressible k-epsilon model	maurice	OpenFOAM	0	May 18, 2009 02:41
Kepsilon code for flow through pipe	rakesh1434	OpenFOAM	2	August 9, 2007 09:23
DecomposePar links against liblamso0 with OpenMPI	jens_klostermann	OpenFOAM Bugs	11	June 28, 2007 18:51
public CFD Code development	Heinz Wilkening	Main CFD Forum	38	March 5, 1999 12:44

November 11, 2009, 23:25		#2
chingyinh New Member Nelson HO Join Date: Jul 2009 Posts: 12 Rep Power: 17	Can you post that lines out?

November 14, 2009, 17:22		#3
jet New Member Join Date: Nov 2009 Posts: 12 Rep Power: 17	Hi, I don't understand the following lines: return ( -laplacian(nuEff(), U) -div(nuEff()*dev(grad(U)().T())) ); The whole expression is defined as divDevReff() and should return the source term for the momentum equation. I tried to write it down mathematically but I failed! Thanks for your help!

November 15, 2009, 09:21		#5
jet New Member Join Date: Nov 2009 Posts: 12 Rep Power: 17	Hmm, are you sure that dev() is a transposition? I think that .T() is a transposition and dev() ist the deviatoric component ( tensor -1/3(trace tensor)I ) of the tensor grad(U)().T()? But I wonder if dev() shouldn't be dev2()?