[Ryan.]

Hello everyone,

I am using SST turbulence model for my simulations and I would like to know what type of boundary conditions for "K" and "Omega" are applied at different markers. In the config file the boundary conditions for pressure and velocity can be explicitly defined but I could not find any information about the type of boundary conditions used for "K" and "Omega" at the inlet, outlet or at the walls. Can someone please give me some information about the default settings of these BCs?

Thanks in advance.

[talbring]

Hi Ryan,

at the inlet k and Omega are computed using the values set with FREESTREAM_TURBULENCEINTENSITY (default 0.05) and FREESTREAM_TURB2LAMVISCRATIO (default 10) in the config file.

At the wall k is set to zero and omega is defined as described here:
https://turbmodels.larc.nasa.gov/sst.html

At the outlet the values are extrapolated from the interior.

I hope that helps.

Tim

[Ryan.]

Originally Posted by talbring

Hi Ryan,

at the inlet k and Omega are computed using the values set with FREESTREAM_TURBULENCEINTENSITY (default 0.05) and FREESTREAM_TURB2LAMVISCRATIO (default 10) in the config file.

At the wall k is set to zero and omega is defined as described here:
https://turbmodels.larc.nasa.gov/sst.html

At the outlet the values are extrapolated from the interior.

I hope that helps.

Tim

Perfect, thanks for the explanations

[sphani4391]

Thanks for the explanations. I have a few questions regarding the k omega SST turbulence model initialization in a cylindrical nozzle flow with recirculating flow at the outlet caused due to kelvin helmholtz instability at the nozzle exit. I am doing 2D axisymmetric simulations with a wedge BC.



1. How do I set the k, omega BC at outlet to take care of the recirculating flow? I could use inletOutlet with inletValue of 0 for k but that's unphysical.


2. What value of y+ should I ensure near the walls. I am using a wall function for k and omega. My Re=25000, which is a high Re case and I heard that k omega is for Low Re.


3. If I go to a fine mesh with y+ <1, should I switch off the wall function?

[bigfootedrockmidget]

Originally Posted by sphani4391

Thanks for the explanations. I have a few questions regarding the k omega SST turbulence model initialization in a cylindrical nozzle flow with recirculating flow at the outlet caused due to kelvin helmholtz instability at the nozzle exit. I am doing 2D axisymmetric simulations with a wedge BC.

Originally Posted by sphani4391

1. How do I set the k, omega BC at outlet to take care of the recirculating flow? I could use inletOutlet with inletValue of 0 for k but that's unphysical.

recirculating flow at the outlet should be avoided, as there is no way of knowing what kind of flow enters the domain from the outlet into the domain. Currently you cannot set any values for k,omega in case of backflow at the outlet.

Originally Posted by sphani4391

2. What value of y+ should I ensure near the walls. I am using a wall function for k and omega. My Re=25000, which is a high Re case and I heard that k omega is for Low Re.

"Low Reynolds turbulence models" usually mean that you can (or must) resolve the wall boundary layer. In SU2, this nomenclature is not used. When you use the wall functions, you can use a y+ >5, with best results between y+=[30 - 100]. If you do not use wall functions, y+ should preferably be y+ < 1

Originally Posted by sphani4391

3. If I go to a fine mesh with y+ <1, should I switch off the wall function?

yes, but this is done automatically when y+<5. That reminds me that I will add a config option to control this parameter.