CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Pre-Processing (https://www.cfd-online.com/Forums/openfoam-pre-processing/)
-   -   nut nutilda SpalartAllmarasDDES revisited (https://www.cfd-online.com/Forums/openfoam-pre-processing/190566-nut-nutilda-spalartallmarasddes-revisited.html)

 hulli July 17, 2017 04:56

nut nutilda SpalartAllmarasDDES revisited

Dear Foamers,

I have a question about the nutilda value in the motorbike tutorial, which LES capabilities (SpalartAllmarasDDES) I like to adapt for my case. Because I kind of gave up on the whale model which always gave me a checkerboard pattern.

Anyway, before starting my own case, I thought it might be a good idea to understand all values in the motorbike LES case first. Basically it is using apart from U and p also nut (in here are the wallfunct.), and the mysterious nutilda. What I found in the forums is that nutilda is a SpalartAllmarasDDES typical variable for the turbulent viscosity (nut). Which one can assign to e.g. an inlet boundary to induce some turbulence. So I hacked all parameters in the online calculator https://www.cfd-online.com/Tools/turbulence.php and into my excel file to double check all equations and parameters therein (see below). After some lit. research I found all equations and I get similar values as the online calculator which is nice.

So here comes my problem or question. Why is in the motorbike case the nutilda value equal to 0.05 ???? After the calculations below nut is equal to 0.16??? From where comes this difference? I found in a prior thread by rob3rt 0ng (https://www.cfd-online.com/Forums/op...t-nutilda.html) or on the cfd online page (https://www.cfd-online.com/Wiki/Spalart-Allmaras_model) that: nut=nutilda*fv1, where as rob3rt 0ng mentioned, fv1 is defined as x^3/(x^3 + cv1^3), where x is nut/nu and cv1 is a closure coefficient of 7.1 (CFD book by D. Wilcox). However, this relation gives me a conversion parameter from nut to nutilda of around 1.0.
On the other hand I also found that others suggest that nutilda should be chosen 3*nut or even other suggested nut should be 5 times higher than nut to initialize a turbulent flow.
So why is the nutilda value (0.05) in the Motorbike case lower than the nut value, if it is suggested to choose it either equal or somewhat higher than nut???
What am I missing here???:confused: Any help is appreciated!!! Thank you very much

Online calculator results:
U = 20 m/s
Re = 1.07E+07
kappa = 0.2671259999999999 m^2/s^2 which is around ~ 0.24 as in the tutorial
epsilon =0.04050541255196353 m^2/s^3
Turbulence intensity level = 2.11 %
Turbulent length scale = 0.0306
Kinematic viscosity (nu) = 1.5e-05 m^2/s
Dynamic viscosity (mu) = 2.25E-05 Ns/m^2
Turbulent viscosity (mut) = 0.24 (dens*0.09*sqrt(kappa)*turblength)
Turbulent viscosity (nut) = 0.16 (0.09*sqrt(kappa)*turblength)
The equation for nut/mut that the calculator seems to use I finally found in (CFD Calculations Of the Flow Around a Wind Turbine Nacelle by J. Varela and D. Bercebal)
Eddy viscosity ratio mut (mut/mu)= 10569.8910906474
Eddy viscosity ratio nut (nut/nu)= 10569.8910906474

 hulli July 20, 2017 04:01

---

Any ideas ??? nobody ???:eek: Best H

 All times are GMT -4. The time now is 11:16.