Initial Turbulence Conditions
 

I am doing an automotive analysis using the k-omega SST turbulence model and I have a question on initial conditions.

The two values:
turbulentKE
turbulentOmega

So to calculate these first I calculated Reynolds Number:
Re = U * L / nu = 45 * 4.16 / 1.5e-5 = 12.5 E6

Then initial turbulence %:
I = 0.16 Re ^-1/8 = 0.0208

Now to turbulentKE:
k = 1.5 * (UI)^2 = 1.5 * (45 * 0.0208)^2 = 1.309

Now is the real question. I found 2 equations for epsilon with one being an approximation. After calculating both a get a big difference in numbers and wonder which one I should use to calculate omega.

epsilon ~ (k^1.5) / L = (1.309^1.5) / 4.16 = 0.360

epsilon = (cmu^.75) * (k^1.5) * (l^-1)
cmu = 0.09 (seems to be most commonly used)
l = 0.07 * L = 0.07 * 4.16 = 0.2912
epsilon = (0.09^0.75) * (1.309^1.5) * (0.2912^-1)
epsilon = 0.845

::::
::::
::::

epsilon = 0.360
epsilon = 0.845

So am I doing this right? If so, which epsilon value should I be using?

::::
::::
::::

turbulentOmega = k / epsilon

Anybody know? :o

I am also interested in some good input for this question. Although the openFoam forum has more traction, and this question is extremely relevant for openFoam, it might be appropriate for it to be in this forum:
http://www.cfd-online.com/Forums/main/

If you do repost the question there, let me know as I am very interested.

I thought about posting there but since this part of the forum (OpenFOAM) gets more foot traffic, I decided to post it here. I just made a new post however in the main.

http://www.cfd-online.com/Forums/mai...onditions.html

Hi,
I assume that your are doing external aerodynamics, in that case your problem is similar to motorBike case in OF.

According to OF training manual, you need to use the value corresponds to the following formula epsilon = (Cmu^0.75 K^1.5)/ L


Thanks,
Sivakumar

Hi !

Could you tell us where you found this relation ? I've always seen the other one you talked about.


For the turbulent length scale value, you choosed the same value for L as the characteristic length used for the Reynolds. If you are doing external aerodynamics, those two lengths could be different.

The characteristic length used for turbulent length scale is a length which depends on your wind tunnel's dimensions.

Could you tell us the meaning of the length L you used for the reynolds and the turbulent length scale ?

I guess it was a dimension of your obstacle. If it is right, then your calculations may not be right : the epsilon values you are calculating are the values of the freestream, before the fluid meets the obstacle : this is a boundary condition. You should have chosen a domain large enough so that this freestream value is not influenced by your obstacle. Then the epsilon value for the free stream should not be influenced by a dimension of your obstacle (the 4.16 value you showed us)


Have a nice day !

In the user guide epsilon = (Cmu^0.75 K^1.5)/ l ，l is different from L