Checking the Accuracy of Cold Flow LES Analysis

[can.sumeyye]

I'm performing a LES analysis. I currently have cold flow outputs and want to check the accuracy of this LES before burning. I have experimental data for this setup, including u', axial velocity, turbulent kinetic energy, turbulent intensity, and integral length scale. I also have standard k-epsilon analysis results. I plan to use these for comparison, but I couldn't find RMS in Fluent. Under the "Unsteady statistics" heading, there's "rmse," but it contains the term "error," so I think it's something different. I also couldn't find integral length scale.

 

Thank you in advance.

[Essence]

Hello,

What "error" you are referring to? Can you please share the screenshot of it?

[can.sumeyye]

This one is named as "root mean square error" (as far as I know)

[SRP]

Hi,

Integral length scale is not directly provided by Fluent. It is typically estimated based on geometry or flow characteristics. For jets or pipe flows, it is often taken as a fraction (1/4 to 1/2) of a characteristic dimension (e.g., inlet diameter, bluff body size)

[SamW]

If you look into the definition of RMSE, it is probably actually consistent with what you expect. Specifically, the "error" refers to the deviation from the statistical mean for each data point, which is how rms velocity is usually considered - in regards to the velocity fluctuations (deviation from the mean), rather than the total velocity.

If you want to build an RMS velocity variable based on "total" velocity values, you would need to do that yourself. You can also look into using the TKE to approximate the RMS values based on something like sqrt((2/3)*TKE) if that sounds appropriate for your application. This can be especially helpful for RANS.