CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Turbulent Kinetic Energy (http://www.cfd-online.com/Forums/main/113427-turbulent-kinetic-energy.html)

 rietuk February 19, 2013 13:25

Turbulent Kinetic Energy

Hi,
I am conducting a CFD validation against some experimental data.
I am trying to compare turbulent kinetic energy as I am given the RMS of the fluctuating components for the 3 directions in an experimental report.
Am I correct in using the formula k=0.5*(Urms^2+Vrms^2+Wrms^2) to calculate the experimental TKE?
When I have done this the experimental and CFD TKE results are orders of magnitude different but the pressure profile is similar

 mp121209 February 19, 2013 15:16

k' = u'2 + v'2 + w'2

 rietuk February 19, 2013 15:45

Hi,
So if I have the RMS of the fluctuating components does this equal the u'2 etc. as you mentioned or is it or does it need squaring to equal that?

 FMDenaro February 19, 2013 16:32

The most important fact in the comparison, is that you must use the same fluctuations that means you must use the same average in the experimental measure and in the numerical simulation!

 rietuk February 19, 2013 16:39

Ok
From CFD I have Turbulent Kinetic Energy in J/Kg

From experimental I have RMS of fluctuating velocitities u~, v~ and w~, all squared and multiplied the sum by 0.5 to give k.

Is the k comparable to the Turbulent Kinetic Energy from the CFD?

Thanks for your help, much appreciated!

 FMDenaro February 19, 2013 16:59

Quote:
 Originally Posted by rietuk (Post 408822) Ok From CFD I have Turbulent Kinetic Energy in J/Kg From experimental I have RMS of fluctuating velocitities u~, v~ and w~, all squared and multiplied the sum by 0.5 to give k. Is the k comparable to the Turbulent Kinetic Energy from the CFD? Thanks for your help, much appreciated!
the problem is not in the unit but in the way the measure is done and in the way the CFD is formulated.

When you decompose your velocity in an averaged and a residual component, such a decomposition says you nothing if you don't define which is the averaged velocity... Generally, I suppose that you have in the experiment a Reynolds average, therefore your CFD formulation must reproduce the same average

 rietuk February 19, 2013 18:32

That makes sense, thank you

The experiment has an average velocity of 1m/s to give the Reynolds number of 40000

To match this Reynolds number the CFD average velocity is 15m/s

How do I scale the CFD TKE or experimental fluctuating velocity components so when I calculate TKE the experimental and CFD TKE are comparable?
Thank you!

 All times are GMT -4. The time now is 08:37.