# How to initialize k and epsilon values?

 Register Blogs Members List Search Today's Posts Mark Forums Read

 April 20, 2019, 11:32 How to initialize k and epsilon values? #1 Member   Joe lee Join Date: Nov 2018 Posts: 59 Rep Power: 3 I have been using OpenFoam and ANSYS Fluent simultaneously, and it seems the initial values of the k-epsilon model calculated by ANSYS are different from what I calculated, so I do not know whether I got the calculation correctly or not. Here is my settings: I am trying to do an external flow like the motorBike tutorials, just by replacing the motorBike geometry by my own model. The parameters of my simulation: Freestream = 12.5m/s characteristic length (L) = 1.638m (I use the total length of my geometry as the characteristic length. Not sure if it is correct) Fluid : air Reynolds number = uL/v = (12.5)(1.638)/(1.5e-5) = 1365000 kinematic viscosity = 1.5e-5m^2/s Turbulence intensity level = 0.16*(Re^(-1/8)) = 2.7637% Turbulence length scale = 0.07L = 0.11466m By substituting the values into the calculation tools in cfd online (https://www.cfd-online.com/Tools/turbulence.php), I got: k = 0.17554J/kg e = 0.057727J/kgs However, from ANSYS Fluent: k = 0.24954J/kg e = 2.2404J/kgs It seems that either set of values are desperately wrong. May I know if I got my own calculation correctly? Help tell me if I did anything wrong! Thank you very much!!

 April 22, 2019, 09:15 #2 Senior Member   Paulo Vatavuk Join Date: Mar 2009 Location: Campinas, Brasil Posts: 188 Rep Power: 12 Hi Joe, You could run the simulation without the obstacle and check the values of k and e at the outlet. Doing that you will know equilibrium values of k and e. This kind of verification is recommended for CFD wind studies. Best Regards, Paulo

April 22, 2019, 11:10
#3
Member

Joe lee
Join Date: Nov 2018
Posts: 59
Rep Power: 3
Quote:
 Originally Posted by vatavuk Hi Joe, You could run the simulation without the obstacle and check the values of k and e at the outlet. Doing that you will know equilibrium values of k and e. This kind of verification is recommended for CFD wind studies. Best Regards, Paulo
Thanks for your reply. However, shouldn't the value of k and epsilon depend on the obstacle? K and epsilon could be obtained using turbulence length scale and intensity level, and these two values take the characteristic length as input. So if we remove the obstacle and capture the value of k and epilson at the outlet, how is that relevant to the case that the obstscle is presented?
I do not have a solid background on CFD, so please forgive for me ignorance. Thanks.

 August 2, 2019, 04:45 #4 New Member   WJ Join Date: Feb 2016 Location: MyHome Posts: 11 Rep Power: 5 The values from k, e calculation you entered in BCs are initial guess. And usually L is height (orthogonal to freestream direction).