Learning Forum

[abd fs]

Hi to all,

As it is known that the root mean square of the turbulent velocity fluctuation can be calculated in RANS as follows (with the turbulent kinetic energy k is known );

u'=(2k/3)^0.5

I am doing simulation in Fluent using one equation RANS and i want to plot  the rms of turbulent velocity for this RANS model.

I tried to plot, but I got a wrong curve

How can I plot  the rms of turbulent velocity in RANS?

 

Thank you.

 

 

EDIT - moving to Fluids. Please post in the correct topics. 

 

[DrAmine]

Why do you think it is wrong? That is the only way to get root mean square in Boussinesq apporach based turbulence models. 

[abd fs]

Hello Amine,

Because  the real curve is :

Best regards,

SERRARE

[DrAmine]

That is what we call validation: so the plot is correct but you are underestimating the fluctuation.

[abd fs]

Thank you for sharing these information here.

In your opinion how can I fix this problem?

Best  regards?

SERRARE

[Rob]

How much mesh have you got in the near wall region? 

[abd fs]

Hello,

this is my mesh

I carried out a mesh by adaptation region

Best  regards,

SERRARE

[DrAmine]

1/Use Low-Reynolds mesh with yplus 1 to 5 and at least 15 cells in boundary layers

2/Make a grid sensitivity analysis and work with high order schemes

3/Comapre with other RANS turbulence models (SST, Realizable with EWT or Menter Lechner) and especially the one which might account for anisotropy: RSM

3/Are you plotting the value directly after the expansion region / step  (weak recirculation zone)

4/Judge your results describe them and document them

 

there is no bad correct or wrong results: only bad or good comments on results. 

[abd fs]

Ok, I will try this and post here the results

Best regards,

SERRARE

[abd fs]

hello Amine,

I'm plotting  in this section, designated by the yellow line in the picture. 

velocity profile

Best regards,

SERRARE

 

[abd fs]

Hello Sir Amine,

(1/Use Low-Reynolds mesh with yplus 1 to 5 and at least 15 cells in boundary layers)

I tried implementing it , but I can't do it if you can help me with pictures that carry the steps to do it.

Best regards,

SERRARE

[DrAmine]

Just make rough estimation for yplus and use conservative cell growth rate for the first cells near the walls. Run once. Have a look into the turbulent viscosity ratio. The Max should of will indicate approximately the middle if boundary layer. If the resolution not enough remesh and rerun...

[abd fs]

Thank you so much Amine.

Ok, I will try this and post here the results

Best regards,

SERRARE