CFD Online Discussion Forums - Inlet boundary conditions for turbulence changing

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- FLUENT (https://www.cfd-online.com/Forums/fluent/)

- - Inlet boundary conditions for turbulence changing (https://www.cfd-online.com/Forums/fluent/123909-inlet-boundary-conditions-turbulence-changing.html)

Inlet boundary conditions for turbulence changing

hi all,

I noticed a thing in fluent which I don't understand and I'll be happy if somebody could explain it to me.

I was working on a CFD simulation for a turbine vane with transitional SST turb model and I was setting the inlet turbulence boundary conditions in the form of turbulence intensity and length scale.

My idea was that these values were used in fluent and converted to the corresponding values of k and w which were actually imposed as boundary conditions at the inlet.
However, after my CFD has converged I found out, by plotting the contours of k and w at the inlet that their value is not even close to the value they should have from the turbulence intensity and length scale.

If I specify directly k and w as boundary conditions then they are kept constant at the inlet.

So my question is: how are the turbulence intensity and length scale information at the inlet handled in fluent? Why do they give way to different values of k and w?

I hope I've been clear enough.
Thanks for your answer.

Cheers,

Rob

Do the measured values match what it calculates if you use "Calculate from..." on the initialization tab?

How are you calculating k and omega from intensity and length scale? I believe $k=\frac{3}{2}(UI)^2$ and $\omega=\frac{\sqrt{k}}{\lambda_T}$ where

is the intensity and $\lambda_T$ is the length scale. Are you using a moving reference frame? If so, try it with both the absolute and relative velocity for

; it's probably absolute, but I'm not certain.

Quote:

Originally Posted by Kokemoor (Post 453334)

Do the measured values match what it calculates if you use "Calculate from..." on the initialization tab?

Nope, they do not match them.

Quote:

Originally Posted by How are you calculating k and omega from intensity and length scale? I believe [math

k=\frac{1}{2}(UI)^2[/math] and $\omega=\frac{\sqrt{k}}{\lambda_T}$ where

is the intensity and $\lambda_T$ is the length scale. Are you using a moving reference frame? If so, try it with both the absolute and relative velocity for

; it's probably absolute, but I'm not certain.

I didn't use any formula myself, I just referred to the calculated values from Fluent in the "calculate from inlet" function in the initialize tab.
I am not using a moving reference frame, everything is stationary.

If you use those formulas, do you get results matching either the data or the initialization computation?

nope...they don't match any of them

also I did notice that, after the calculations are completed the inlet values of turbulent intensity are completely different from the values I specified at the inlet.

Anybody has any idea on why is this happening?

Hey all,
as an answer to this problem, you should be mindful of two things:
1- your turbulent properties can change when you standard initialize the domain. make sure that you enter the right K and Omega (I am using the SST K-Omega model) values in the panel.
2- make sure that you have the right values in the "reference values" panel (accessible through the physics tab on top). Intensity is calculated based on K and turbulence length scale and hydraulic diameter so you should have correct length values set in that panel.