Sectional Drag and lift coefficient?
I know how to get an overall drag and lift over whole domain,
What I need is also a sectional drag& lift output, in example:
a 3D Cylinder is 5 heigth units high. OF gives me
CD for the surface "cyl". And what I want to achieve is also a drag and lift for Z=1,2,3
I have an idea yet have not put into practice.
The drag and lift forces are calculated from pressure and friction forces around the cylinder, one can easily get the pressure distribution around the cylinder at desire Z by utilizing the sample utility. As for the friction force, with my immature knowledge about fluid dynamics I think one can calculate from the wall shear force since the utility wallGradU is available in OpenFoam.
These are just some rough thoughts, I'm not sure if it could work.
by the way I'm also doing LES of flow around cylinders at Re=1000 by pimpleFoam, however, as hard as I try, I could not obtain a satisfactory result in terms of drag coefficient. My LES model is smagorinsky with Cs=0.1 and y+ near the cylinder wall is smaller than 1. The Cd in my simulation is about 1.15 but from references it should be around 1.05.
I have slight troubles understanding what you want, but is this it: http://www.openfoam.org/version2.2.0...processing.php (see img.)?
If yes, then look at the appropriate motorBike tutorial (in the incompressible/simpleFoam folder).
Thanks for that idea, I found a workaround with paraview and got my sectional lift
LES Drag Coefficient Problem?
I am doing kind of a DNS and my results are matching perfectly. Used a modified icoFoam Solver 2nd order spatial and 4th order in time.
Are you using a structured mesh?
Is your case compressible or incompressible?
Did you set the right projected reference Length and area in your controldict for calculation of cd?
Did you also reconstructed CD value with paraFoam and compared it with OpenFoam Output?
Thanks for your answer, but this is not the answer for my problem. I solved by a workaround.
Later I found out Re=1000 could not be a good validation for LES, I tried Re=3900 and the results worked out well. There are some discussion about Re=1000 in this thread
Could you please elaborate how did you get your Cd and Cl from paraview? I'm very interested in that.
These are my paraviewsteps in short:
2)Generate Surface Normals
cp = normals*p; ->Pressure Drag
cf=normals*mu*wallGradU_y -> Friction Drag
4) Build Integral Drag=S(cp_X+cf_X)dA*2/rho*v*v*A
Hope this helps, but there are some differences to OpenFoam Output -> could be caused by building derivates or integration failures
thanks very much! I will try your method!
|All times are GMT -4. The time now is 23:27.|