Learning Forum

[omidFSI]

Hi everyone,

 

I am simulating a simple flow over a cylinder with a circular domain shape in 3D with Fluent.
For continuity I am not getting a good convergence and the residual is close to 0.1 while I expect it to be much lower.

The drag coefficient is also not accurate especially for high Re. I have tried different RANS models , mesh is fine enough (O-type) though y+ getting from ANSYS is somehow high more than 100 which don't know why since my mesh is structured and very fine!
Re: 2e4-1e5
Transient solver with Time step~0.005-0.05 (s)

Any idea in this regard?

Thanks in advance!

[vganore]

How is your orientation of cylinder with respect to flow direction? Could you please show your geometry?

Are you doing similar to following tutorial but extending your domain to 3D? I would keep Y+ value close to 30 and use k-e standard wall function model.

https://www.youtube.com/watch?v=DXWO-X16UIo&t=686s

[omidFSI]

Thanks for the reply!

I have a similar flow simulation where the cylinder is perpendicular to the fluid flow. The only difference is using circular fluid domain.
( diameter of cylinder, flow left to right, domain diameter:40D, domain depth=3D-5D)

My y+ as I calculated is good and around 5. I am using Spalart-Allmaras and SST- k-omega as later I wanna use these models for more complex geometry.

For high Re (like 88000) and domain depth of around 5D (D-diameter of cylinder), the values of drag coefficient I get is not accurate at all and for all the cases the continuity residual is around 0.1. My previous simulation for a coarser mesh or 2d case had really good result which I don't understand at all!

Any help is appreciated!