[Sören Sander]

Hello everyone,

I try to figure out, how the heat transfer through a wall is calculated in fluent when using the kOmegaSST model. Unfortunately, there is nothing given in the fluent manuel. Is the laminar flow equation used, because yPlus is around 1?

[LuckyTran]

Originally Posted by Sören Sander

Hello everyone,

I try to figure out, how the heat transfer through a wall is calculated in fluent when using the kOmegaSST model. Unfortunately, there is nothing given in the fluent manuel. Is the laminar flow equation used, because yPlus is around 1?

Fluent uses an enhanced wall function appropriate for kOmegaSST model for all equations (including energy). If your mesh is sufficiently fine near the wall (y+ < ~10), then it is equivalent to using the the linear (laminar) model.

[Far]

Yes, with Y+1 you are able to solve the laminar sublayer.

[Sören Sander]

Thanks for your replys. However, now I have found another problem on this topic:

I want to calculate the heat transfer coefficient. fluent can derive it automatically, but again I am not able to find the formulation.

It should be h=(T_wall-T_inf)/q_dot, where q_dot=(T_wall-T_p)/dn
where T_p is the temperature derived in the first cell from the wall and dn is the distance normal from the wall. What value does T_inf have? Is it the same as T_p? Is there any reference to validate how it is calculated?

[LuckyTran]

Originally Posted by Sören Sander

It should be h=(T_wall-T_inf)/q_dot, where q_dot=(T_wall-T_p)/dn
where T_p is the temperature derived in the first cell from the wall and dn is the distance normal from the wall. What value does T_inf have? Is it the same as T_p? Is there any reference to validate how it is calculated?

T_inf should be the driving fluid temperature consistent with Newton's law of cooling. Things like freestream temperature, bulk temperature, recovery temperature are used to get the most physical value of h. T_inf is using nomenclature from low-speed external flows where the freestream temperature is the driving temperature. Otherwise the choice of T_inf is arbitrary as much as length scale is arbitrary in Reynolds number.

[Sören Sander]

So I can define T_inf in fluent?

[LuckyTran]

Originally Posted by Sören Sander

So I can define T_inf in fluent?

Oh yes. T_inf is a user supplied input. It is the reference temperature that is user specified in the reference values. It is specified as a constant. If you want to use a variable T_inf you will need to modify the fluent code (use a UDF).

The wall heat flux and surface temperature are determined from the solution. The heat transfer coefficient is then calculated by applying Newton's law of cooling and using Tref (the reference temperature specified in the reference values).