[rizhang]

Hi, everyone:

I have a question about turbulence length scale and integral length scale.

I am right now doing a CFD simulation about Wind Turbine by using CFX with kappa-Omega 2-Equation turbulence model. In order to specify the turbulence level at the inlet (about 20% turbulence intensity ), i need to give the value of turbulence length scale and turbulence kinetic energy. But, i just know the intergral length scale at the inlet which is equal the chord length of the blade. How can i convert the integral length scale into turbulence length scale.

Another method is that i can give the value of kinetic energy (kappa) and turbulence frequency (Omega). But, in order to calculate the omega, i need to know eddy viscosity ratio. Dose anyone know the eddy viscosity ratio of water at temperature 20° C?

Thank you very much

regards

yang

PS: i am using kappa-omega model and with very high turbulence level (20%)

[ghorrocks]

Originally Posted by rizhang

I am right now doing a CFD simulation about Wind Turbine by using CFX with kappa-Omega 2-Equation turbulence model. In order to specify the turbulence level at the inlet (about 20% turbulence intensity ), i need to give the value of turbulence length scale and turbulence kinetic energy. But, i just know the intergral length scale at the inlet which is equal the chord length of the blade. How can i convert the integral length scale into turbulence length scale.

Get yourself a copy of "Turbulence modelling for CFD" by Wilcox. That will explain where all these terms come from.

Originally Posted by rizhang

Dose anyone know the eddy viscosity ratio of water at temperature 20° C?

This is a function of the local turbulence conditions, not a material property. A value of 10 or 100 is often used when you have no idea but that is nothing more than a guess of a value in the typical range.

[ahmed.enit]

The filter size is set to values which are smaller than the length scales returned by the conventional RANS models. Specifically, the level of the turbulent viscosity is corrected by comparing the turbulence length scale and the filter size Δ, which is based on the local meshing spacing:

μt_FBM= ((Cμ*ρm*k^2)/ε)*f_FBM with f_FBM= min(1,(∆.ε)/k^(3/2) )

with ∆=max(∆_present,∆_grid)

my question is about the variables in CFX that can be introduced in place of ∆_present and ∆_grid

thank you