[p_kmb]

I'm trying to simulate 2D flow over a rectangular obstacle.
It's an incompressible air flow with Re=10e6.

The problem is that I get the lowest Cp value equal to 1.15. I know that sometimes Cp can exceed 1 but I think 15% is too much.

Parameters of simulations:
- pressure based solver,
- k-e standard model (also tried RNG and realizable),
- enhanced wall treatment (y+ is lower than 5),
- velocity inlet, pressure outlet, symmetry for lateral walls of domain, and no-slip walls for obstacle,
- reference value of velocity is equal to inlet velocity.

Did anyone try to simulate flow over a bluffbody with similar Re? And got Cp=1?

I'll appreciate any help.

Paulina

[p_kmb]

What can be wrong in my simulation?

[p_kmb]

Maybe it will be more clear if I show an example.

I'm trying to recreate in Fluent a simulation from this article.
I chose the case with R=0.4 (R=B/H, B-wall along the direction of flow).

My parameters:
H=3,126 m
B=1.25 m

velocity inlet and pressure outlet:
v=0.1 m/s
k=6e-6 m2/s2
omega=0.02 1/s

upper and lower edges of the domain:
symmetry

walls of the building:
no slip conditions, roughness height=0 m

The calculation scheme is PISO, second order, and the time step=0.188 s.

The Cd mean value I got is around 3,8 and the reference article is 3.4.
Also my Cp value is 1,2.

I attached the pictures of grid, reference value and Cp.
Clearly I don't understand something and do it wrong. I can post more details if someone would like to help.

[metmet]

Hi Paulina
Have you done any validation process? What's the validation results?

[p_kmb]

Hi Metmet.

What do you mean by the "validation process"?
I compared my results of Cd and Cp with the results from the article.
They got mean Cd=3,4 and max value on the front edge Cp=1, whereas I have 3,7 and 1,2 respectively.

[metmet]

Originally Posted by p_kmb

Hi Metmet.

What do you mean by the "validation process"?
I compared my results of Cd and Cp with the results from the article.
They got mean Cd=3,4 and max value on the front edge Cp=1, whereas I have 3,7 and 1,2 respectively.

Yeap.
That's kind of validation. You need to compare your achieved results with the reference which you have done.
In your case was the grid independence study (GIS) done?
(GIS is necessary to find out your obtained results are not dependent to the grids number)
In the reference article which you have linked I found that authors have mentioned the Reynolds number equals to 21,400 (6th line of abstract) while you have performed your simulation at the Re=10e6. (1st comment)
Ans also authors have used K-Omega SST as a turbulence model while you have used K-Epsilon model.
I don't remember if your turbulence model has good agreement in viscous sub layer (due to Y+<5), to check this I suggest an article entitled "Wall y+ Strategy for Dealing with Wall-bounded Turbulent Flows BY Salim .M. Salim"