CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   3D FLOW OVER A CIRCULAR CYLINDER (http://www.cfd-online.com/Forums/fluent/36066-3d-flow-over-circular-cylinder.html)

Srinivas Mettu March 18, 2005 13:36

3D FLOW OVER A CIRCULAR CYLINDER
 
hi, i am using FLUENT 6.1 to solve 3D flow over a circular cylinder in unbounded domain.i could validate drag(CD) and St for unsteady flow past a circular cylinder in 2D. But in my 3D calculations, the drag values for Re =100,200,300 are coming to be higher if i use a boundary layer on cylinder and lower if i dont use a Blayer. What could be the fault in my computations using FLUENT ??? please suggest me the efficient way to use FLUENT for 3D computations.


Chetan Kadakia March 22, 2005 19:56

Re: 3D FLOW OVER A CIRCULAR CYLINDER
 
Several variables are at play. What is your choice of turbulence model? What about near wall treatments?

Did you guess at your boundary layer mesh, or did you calculate it? I suggest looking into law of the wall. This also depends on near wall treatments.

If you are using LES, you will need more cells than if you are using k-epsilon, but it will require more memory/ time.

Other items to check: quality of mesh, skewness of cell.

You can email me, as I think I have done this before and also had problems getting accuracy in drag.


cfdcfd123 April 4, 2010 22:11

Why are we taliking about turbulent models for Re=100, 200 and 300. The flow on the cylinder is laminar for these flows. There is a small spanwise unsteadiness in the flow. Just calculate the y+ value first and get the first point of the wall for y+=0.06. Then stretch the grid at the rate of 1.08.
Keep the outer boundary based on how good the BC that you use is.
Good luck. Use explict solver. The stable time step is usually less than the time step you need to resolve the flow.
Good Luck.


All times are GMT -4. The time now is 19:03.