[bryant_k]

hello everyone
I have made a simulation of Laminar Flow Around a Cylinder.

My computation domain is 7mx1mx0.025m and the diameter of the cylinder is D = 0.2m.the Reynolds number is defined by Re = UD/nu=40. so I set the inlet velocity U= 1m/s and the density rho=1kg/m3,the kinematic viscosity nu=0.005m2/s. The center of the cylinder is 1.5m away from the left inlet.The third attached file is the reference data of Re=40.In general ,Lw/D is about 2.20~2.29,but my result appeared that Lw/D= 0.3/0.2=1.5

I used the icoFoam solver to simulate the problem. but my result has big differences with the reference paper. I don't know why. Can any body help me? I have attached my case here.


Thank you very much
Best Regards

kob

[bryant_k]

I found that i can't attach my case here because the mesh files are a little big.
so i attached the other files here.

[MartinB]

Hi kob,

I suppose it's your mesh that makes the trouble...

You can try the attached cases: the boundary conditions are yours, I only changed the solvers in fvSolution and added nonOrthogonal correctors.

To run the cases use "blockMesh", than "icoFoam" or "simpleFoam".

Good luck

Martin

[bryant_k]

Thank you for your attention.I have run your case of case_icofoam1.7 and case_simplefoam1.7.Your result is better than me in large measure but it doesn't match the reference paper very well.Because the Lw is calculated from the right vertex of the cyclinder to the eddy wake not calculated by the way you have used.You can look into the second picture I attached .

[MartinB]

You can try a fully developed U inlet profile.

Unpack the attached changeDictionaryDict into the sytem directory.
Usage: "blockMesh", "changeDictionary", "simpleFoam".

Martin