[scoufield]

Hello Guys,

i try to prepare a cfd-Simulation with FloEFD using the k--Model.

When i calculate the Intensity [ I=0,16*Re^- ]
i get a small Intensity for a huge Re-Number.
Normaly i should use higher Intensity for a higher Re-Number ?
https://www.simscale.com/forum/uploa...fd9f6cc954.png

Thanks

Sco

[LuckyTran]

Turbulence intensity tends to decrease with increasing Reynolds number, yes. That doesn't mean the turbulence is less. The turbulent kinetic goes k~(UI)^2

[scoufield]

Originally Posted by LuckyTran

Turbulence intensity tends to decrease with increasing Reynolds number, yes. That doesn't mean the turbulence is less. The turbulent kinetic goes k~(UI)^2

I calculate the turbulent kinetic with k~ 2/3 *(u*I)^2
so we get also a lower Turbulence kinetic when the Re-Number rise.

I mean the approximation Formula doesnt fit with the picture in the link.
Whats the reason for that ? In the pic we get higher Intensity when we choose a higher Re-Number. The Formula results in the opposite ..

Thanks

[FMDenaro]

The function I decreases with Re but I think one should also consider the increasing in the reference velocity U.

[BlnPhoenix]

Originally Posted by FMDenaro

The function I decreases with Re but I think one should also consider the increasing in the reference velocity U.

From this equation [ I=0,16*Re^(-1/8) ] high turbulent cases exist for lower Reynolds Numbers! If we say I = 20% means high turbulent cases. So this is very confusing, i agree with OP.

[FMDenaro]

Originally Posted by BlnPhoenix

From this equation [ I=0,16*Re^(-1/8) ] high turbulent cases exist for lower Reynolds Numbers! If we say I = 20% means high turbulent cases. So this is a very confusing, i agree with OP.

I know this formula appears in Fluent, being cited as an experimental correlation valid in the core of a fully developped duct flow. Whitout knowing the assumptions use to express the correlation I cannot say more. Maybe it is valid only in a specific range of Re numbers, as you can see for Re->+Inf and Re->0.

[LuckyTran]

There already is a clear measure of how turbulent something is and that's Reynolds number.

Keep in mind you need to specify two things (out of three). Usually you take a velocity scale (and often people choose the turbulence intensity, implicitly assuming the velocity scale is the mean velocity) and a length scale. You need to consider what the is happening to the length scale and/or time-scale with increasing Reynolds number and remember that turbulence intensity needs to be paired with the right velocity scale.

The turbulence intensity alone doesn't give the complete picture (plus it only makes sense if you specify also the mean velocity associated with it). If you think turbulence intensity is a measure of "high turbulence" then consider the low velocity (or low Reynolds number) limit where any velocity fluctuation results in I=>Inf. E.g. a butterfly slowing moving its wings or a falling leaf can generate Inf % I. Clearly that tends towards laminar flow and there isn't any turbulence.

[scoufield]

Originally Posted by LuckyTran

There already is a clear measure of how turbulent something is and that's Reynolds number.

Keep in mind you need to specify two things (out of three). Usually you take a velocity scale (and often people choose the turbulence intensity, implicitly assuming the velocity scale is the mean velocity) and a length scale. You need to consider what the is happening to the length scale and/or time-scale with increasing Reynolds number and remember that turbulence intensity needs to be paired with the right velocity scale.

The turbulence intensity alone doesn't give the complete picture (plus it only makes sense if you specify also the mean velocity associated with it). If you think turbulence intensity is a measure of "high turbulence" then consider the low velocity (or low Reynolds number) limit where any velocity fluctuation results in I=>Inf. E.g. a butterfly slowing moving its wings or a falling leaf can generate Inf % I. Clearly that tends towards laminar flow and there isn't any turbulence.

I also use the intensity & the length scale. But I thought the length scale is a parameter which i calculate out of the geometry. (L=0.07*D) Where D is the Pipe Diameter. So the length scale should not depending on Re or velocity ?
I already used the mean velocity for calculate the Re-Number.

The example with the butterfly and the leaf is right. And i think there should be a specific range of re numbers which you can use for the approximation formulae. Maybe FMDenaro is right. Thanks for your answers guys..

Sco

[Simbelmynë]

Just want to add that there is a Wiki on the topic here at CFD-Online:

https://www.cfd-online.com/Wiki/Turbulence_intensity

It is noted that the expression you use is from the Ansys manual and that no reference to it's origin is given.

Also, (@LuckyTran) please have a look at the definition of High- to low-turbulence cases.

For OP, I suggest that you do a small parametric study to test the dependence on the k and eps boundary conditions for your solution. In many cases there is only a weak dependence if the flow has time to develop inside the domain.

[BlnPhoenix]

I once read somewhere on this forum that turbulence intensity gets smaller with higher Re number, because the increased mean flow velocity somehow damps out the velocity fluctuations. So, i don't know if this is a general experimental observation for turbulence..

[Simbelmynë]

To me that sounds counter intuitive. Also, higher Re does not necessarily mean higher mean flow velocity.

[BlnPhoenix]

Yes, i agree it sounds counter intuitive. But it's the logic behind the posted equation for fully developed pipe flow. In an adiabatic pipe flow, increased mean velocity is equal to decreased pipe diameter, which results in higher Re number. The obeservation is: velocity fluctuations get smaller, so what else than some damping behavior can explain this..?

[scoufield]

Originally Posted by BlnPhoenix

Yes, i agree it sounds counter intuitive. But it's the logic behind the posted equation for fully developed pipe flow. In an adiabatic pipe flow, increased mean velocity is equal to decreased pipe diameter, which results in higher Re number. The obeservation is: velocity fluctuations get smaller, so what else than some damping behavior can explain this..?

I read something interesting:

"The issue here is that the Reynolds number is not a measure of "how turbulent" the flow is or will become. It is a ratio representing the relative importance of inertial forces to viscous forces in the flow."

Reference https://www.physicsforums.com/thread...ensity.676226/

Sco