[lostking18]

Hi, there

I am trying to use Spalart–Allmaras turbulence model in CFX but haven’t directly found it in the list of turbulence models. On the internet it seems to suggest that SA model is the ‘Eddy Viscosity Transport Equation Model’. But there is no official documents to state that. Moreover the paper that referred by ‘Eddy Viscosity Transport Equation Model’ is from Menter instead of Spalart and Allmaras. Could anyone tell me how can I implement SA model in CFX?

Thank you very much!

[karachun]

You can enable SA model by switching beta features ON.
CFX Spalart Allmaras turbulence models
https://imgur.com/LBiJlw3
https://imgur.com/ipH5Nlm

[lostking18]

Originally Posted by karachun

You can enable SA model by switching beta features ON.
CFX Spalart Allmaras turbulence models
https://imgur.com/LBiJlw3
https://imgur.com/ipH5Nlm

Hi, Karachun:

Thanks! And I also got some additional info from Ansys technicians about the boundary condition set-up of SA model:

The S-A-Model is still in Beta-Stadium in CFX.
The GUI is not consistent. For this particular boundary condition, the value of TI is simply ignored.

The other options require two values to calculate the turbulent viscosity nu_t at the inlet:
Intensity and Length Scale: nu_t = Cmu * sqrt(k) * length_scale, k = 1.5 * Intensity^2 * Velocity^2
k and epsilon: nu_t = Cmu * k^2 / epsilon
k and omega: nu_t = k / omega
k and Length Scale: nu_t = Cmu * sqrt(k) * length_scale

Hope this will also be helpful.

[karachun]

Thanks for reply!

[modest_may]

Originally Posted by lostking18

Hi, Karachun:

Thanks! And I also got some additional info from Ansys technicians about the boundary condition set-up of SA model:

The S-A-Model is still in Beta-Stadium in CFX.
The GUI is not consistent. For this particular boundary condition, the value of TI is simply ignored.

The other options require two values to calculate the turbulent viscosity nu_t at the inlet:
Intensity and Length Scale: nu_t = Cmu * sqrt(k) * length_scale, k = 1.5 * Intensity^2 * Velocity^2
k and epsilon: nu_t = Cmu * k^2 / epsilon
k and omega: nu_t = k / omega
k and Length Scale: nu_t = Cmu * sqrt(k) * length_scale

Hope this will also be helpful.

Thank you for your kind and helpful message.
Do you know the requirement for the yplus value near the wall? Should the yplus be smaller than 1? There are two options, including default and scalable, for the wall function term.