## Fluids

#### High roughness setting with R = 1000 µm for a flow between two cylinders

• Chandae Jeong
Subscriber

Dear FLUENT Users,

I am trying to simulate a helium flow of p = 8 MPa and Temperature range is about 573-660 K.

Therm. Cond. (W/m*K) ~ 0.27
Viscosity (Pa*s) ~ 3.5e-5
Density (kg/m3) ~ 5.6
Cp (J/g*K) ~ 5.1887

Re_Dh = 4000 - 10000

Let's think that the red inner cylinder is a solid heat source.

Problem is that I am dealing with a surface of heat source which has a high roughness up to 1000 µm.

Only inner surface will have roughness and outer surface is smooth.

I am planing to use Transition-SST model (gamma-Re-model), because results with very low roughness and low Re showed quite large laminar zone in the geometry.

If I put 1000 µm in the boundary condition of wall, I can see some results but I am doubt if the results are reliable. Our plot Nusselt number over Reynolds number shows quite huge different in compared to experiments by our cooperative research group.

1. Is such high roughness 1000 µm available with Transition-SST model?
2. I know that the constants of Transition-SST model were set for a plate flow. Can the default contants for transition-SST be used for such flow between two cylinders?

Thank you so much for reading.

With kind regards
C. Jeong

• Rob
Ansys Employee

1) How big is the near wall cell? That's the question you need to start with. From there check the documentation regarding how it's handled in the solver.

2) The coefficients tend to be fairly general so are good for a range of applications.

• Chandae Jeong
Subscriber

Hello,

thank you very much for the answer. I am trying to make near wall cell size locally maximum y+ less than five in the whole cfd domain for each reynolds numbers, because it is recommended y+ max. value < 5 for the transition-sst model.

• Rob
Ansys Employee

True, but how does that relate to the roughness height?

• Chandae Jeong
Subscriber

I guess y_p (distance between wall and first cell point) should be bigger than Ks. For my cases, if maximum y+ should be around five, the first cell size is about 2e-5 m, it means y_p is about 1e-5 m which is much smaller than Ks. I think y_p should bigger than Ks. It means I cannot use transition sst but other models like k-ep model which I can adapt standard wall functions with y+>30?

• Rob
Ansys Employee

It means you need to review all of the information and figure out what to do. It's an early introduction into why blindly following y+ guidelines isn't recommended. Have a more careful look at the k-w models & theory - they're also OK with higher y+ with care.

• Chandae Jeong
Subscriber

Thank you so much for the detailed guide!