CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Turbulence intensity BC (http://www.cfd-online.com/Forums/fluent/41550-turbulence-intensity-bc.html)

 Ralf Schmidt June 28, 2006 10:11

Turbulence intensity BC

Hi!

I have a simple geometry (free stream around a cylinder). The inlet BC is vel.inlet with turbulence parameters are turbulence intensity and hydraulic diameter (diameter of the cylinder).

Now when I plot the values of turbulence intensity at the inlet, they are 1. not all the same over the face (they vary about +- 5%) and 2. slightly lower (about 10%) than the values I set as BC. What happend?? The reference velocity is the free stream vel.

Any Idea?

Ralf

 freeman June 29, 2006 18:16

Re: Turbulence intensity BC

you might want to look up in the USER manual in that regards. Remember that, the turb. intensity at the inlet depends on Ub or tke. so if the vel. is not uniformaly distributed then you wouldn't expect the intensity to be uniform. what I'll suggest is that plot the vel, tke and based on that you tell what's going on.

all the best

 Ralf Schmidt June 30, 2006 01:02

Re: Turbulence intensity BC

I checked the points you mentioned:

there is a constant inlet velocity BC of 21 m/s BUT in the plot, the vel. is NOT uniformaly distributed, it is fluctuating about +-0.005% of the inlet vel.

How could that happend?? Because of mass balance? I do a incompressible calculation.

The Turbulence kinetic energy, turb. viscosity, omega and the turbulence intensity have the same fluctuations. BUT they are much larger!!

How could a vel. fluctuation of 0.009% cause a Tu.Intensity fluctuation of about 5% ???

What happend?

Ralf

 freeman June 30, 2006 04:17

Re: Turbulence intensity BC

if I may ask, what is the shape of the inlet, circular or rect?

anyway, to check if you're doing right thing, check your mass balance to see if mass_in=mass_out through the report menu. I think what happened was tha if you specified normal velocity at surface, Fluent interpolate it over the surface by dividing it according to the number of nodes so you wouldn't expect all the nodes to have the same values, however the mass-flow will be correct and the total integration of the vel will be equal to what you specified. You could cross check that through that through the 'REPORT menu' and also the user manual.

as for higher variations in the turb quantities are obvious, since they are in higher order of the vel. eg. tke=(...)vel^2. That's not a problem.

I hope that works. if I may ask, how are the general results look like? outlet plane results? are you comparing with expt or what? if you do, how did it deviate from it?

 All times are GMT -4. The time now is 11:30.