Learning Forum

[mutianzi1]

Hello!

 

I would like to predict the decay of turbulence intensity downstream of an inlet. In this way, I could apply a correct turbulence intensity at the inlet in order to get the expected value at the location of interest. The following equation is provided in ANSYS Custom Portal under Solution #2042913.

 

 

I have the following questions about this equation:

1. The unit of the equation is not included. Could you please let me know the unit of this equation?

 

2. Is ρ the air density and can be set as 1.225 kg/m^3 at standard atmospheric pressure assuming the equation is in SI unit?

 

3. Is μ the dynamic viscosity and can be set as 1.81e-5 Pa•s?

 

4. In the denominator, does μ(μ_t/μ) mean μ multiply μ_t/μ? μ_t/μ represents eddy viscosity ratio?

 

I used the equation and above parameters to do the calculation, but I get different results from the provided solution, which is also attached. That's why I made these questions. Could you please give me some suggestions?

 

This is what I got. The turbulence intensity at inlet is set to 0.5%. The V at freestream is 5 m/s and viscosity ratio is 10.

 

This is the comparing case from ANSYS Custom portal.

 For your convenience, the MatLab code I used for the calculation is copied here.

 

 

 

Thank you very much! Any help is appreciated.

  

Bests!

[DrAmine]

Do you have access to customer portal?

[mutianzi1]

Hello Amine,

 

I do not have the access to customer portal.

 

My advisor can access the customer portal and she gave me this information.

 

Thank you!

[DrAmine]

Ti is dimensless. The material properties are specified at ambient conditions as highlighted in the solution. You need to adjust them to your local conditions. Mu_t to Mu is the viscosity ratio.

[mutianzi1]

Hello Amine,

 

Thank you for your reply!

 

I understood TI is dimensionless. I am confused about the units of air density ρ, velocity V, mu, etc. Because the coefficients, e.g. 2 and 3, are related to the units of these parameters. I was wondering the unit is in metric system or imperial system? 

 

Also, does mu represent dynamic viscosity? In the denominator, it should multiply viscosity ratio?

 

Bests!

[DrAmine]

The term in the bracket is dimensionsless and all viscosities are dynamic ones.

[mutianzi1]

Hello Amine,

 

Thank you for your explanation. I have one more question. β and β* in the equation are two constants used in the Menter k-w SST (shear stress transport) turbulence model. In my CFD simulation, I use Large Eddy Simulation (LES) turbulence model to simulate the turbulence. Is this equation also applicable to LES model?

 

Best Regards!

[DrAmine]

There is no eddy viscosity in LES bur the unresolved stress obeys the same boussinesq formalism deploying a SGS viscosity. Moreover in LES you need to account for resolved and unresolved TKE. You can use it as crude estimation but for LES you will need other type of inlet boundary condition. I have seen some other correlation for TKE decay which can be found in the WWW.

[mutianzi1]

Hello Amine,

 

Thank you for your detailed explanation. It helps a lot. You mentioned some other correlation for TKE decay, which can be found in the WWW. Here, what do you mean by saying 'WWW'. Could you please share a little bit more info about this with me. I do appreciate your help.

 

Best Regards!

[Rob]

I suspect Amine means World Wide Web, but probably doesn't have a specific reference. 

[mutianzi1]

Hello rwoolhou,

 

Thank you for the info!

 

Bests!

[DrAmine]

Yeah. Just look after decay of tke.

[mutianzi1]

Hello Amine,

 

Thank you and sorry to bother you so long. May I ask one more question about the application of turbulence at the velocity inlet in Fluent.

 

Does the turbulence has any directions when it is applied at the velocity inlet? I used LES and turbulence is applied as turbulence intensity and turbulence length scale. I only applied one value as turbulence intensity. But I don't know whether it is applied in along-wind direction, or in all three directions, i.e. along-wind, across-wind, and vertical directions. 

 

Is TI only applied at the velocity inlet and it will decay with distance, or it is applied to the entire computational domain?

 

If it is applied to the entire computational domain, I do not have to consider the decay of TKE.

 

Best Regards!

[DrAmine]

Check inlet boundary condition for scale resolving simulation. Vortex method is recommended. Do not worry about decay but rather focus on establishing a good LES run with precursor URANS 

[mutianzi1]

Hello Amine, 

I studied the section of inlet boundary condition for scale resolving simulation. I am still confused about one thing. I only input one value for turbulence intensity, as well as turbulence length scale. How are these values actually applied to the inlet? Do they apply to he entire face zone of inlet, or only apply to curtain height? In reality, turbulence intensity and length scale may vary with height. Can I achieve this in Fluent and how? Could you please throw some light on this?

 

Also, could you please explain a little bit more about "establish a good LES with precursor URANS"? I don't quite understand the meaning of " with precursor URANS". Do you mean running a few time steps using URANS before using LES?

 

Thank you! HAPPY NEW YEAR!

[DrAmine]

You assess a LES spatial and temporal resolution based on precursor RANS run and start if possible from it after making the velocity field bit instantaneous. For thé LES inlet and assuming perturbation ANSYS FLUENT expects your input of TKE or TI to prescribe the fluctuations.

[mutianzi1]

Hello Amine,

 

I am not familiar with what you mentioned about precursor RANS run. Do you have an example, or reference, or video that I can follow? 

 

Thanks!

[DrAmine]

Check best practice guidelines for scale resolving simulation by Florian Menter as well as this tutorial produced by Raef.

https://youtu.be/NOmrcarpYn8