[Tsr63]

Hi,

I am relatively new to CFD and Fluent and I am trying to determine the Drag Force and Drag Coefficient for flow over a 2D cylinder, but I am having some problems with fluent producing the correct value (based on experimental results). First I ran a case for a Reynolds Number of 1,000 and obtained a drag force of .0008 N which is close the the experimental value of .0011 N, but the drag coefficient which should be around 1 is off by a magnitude of 10.

To measure drag force/coefficient I go to Report>Forces>Set X=1,Y=0 and chose the edge which represents the cylinder wall.

When I run a case for a Reynolds number of 100,000 I get a drag force of 2 (should be around 11 N) and a drag coefficient of 314 (which should be around 1). Since the flow is laminar until approximately Re=400,000 for flow around a cylinder, I am using the laminar solver model.

The problem also continues when I try to run a turbulent case using the k-epsilon model.

I created my geometry in gambit. I subtracted a .125 m diameter circle from a larger 2 m diameter circle. I created a sizing function from the edge (or cylinder) to the face. Next I meshed the face using a quad map scheme. This all seems correct to do, but I figured I'd just let you all know in case.

I've tried almost everything I can think of, and even made sure my reference made sense. Any help would be greatly appreciated.

- Jon

[ojha.mayank485]

Originally Posted by Tsr63

Hi,

Since the flow is laminar until approximately Re=400,000 for flow around a cylinder

- Jon

Am very much sure that at such high Re of 400,000, the flow cannot be laminar.
For flow which has circulations, K-e shows delayed separation and hence gives incorrect answers. According to the literature, SST should be what you should be using.
I am trying to simulate a flow over a cylinder at Re=10,000.

Remember: when you are using a turbulent model, make sure your y+ value is acceptable.

Regards,

Mayank

[skullCFD]

Did you try a transient simulation? Vortex problems are inherently transient and it helps a lot of times with the solution and convergence issues if you perform a transient simulation. I agree with ojha.mayank485. k-w SST might be good for your case.

[rishitosh]

hiii..

i m also new in dis CFD area...
i m working on flow over two side by side cylinders... @ Re= 100 and 200..
i got good results..

but i want to find out Strouhal number for all cases...

i came to know by finding shedding frequency with help of FFT, using cofficient of lift data, in fluent we can calculate Strouhal number...

but i m getting helpess..

how to use FFT to find shedding frequency....?????
-Rishi

[cfd seeker]

Yes try with SST kw and if problem still persists think of trying Transition model.
Note: Search on youtube or in the threads on this forum, I guess "far" has uploaded some videos for the flow simulation on cylinder but i guess those simulations are 3D

[huxxx088]

To use an FFT to calculate Shedding Frequency:

Place a numerical probe somewhere downstream of the cylinder where the flow looks stabilized. Measure the x and y velocity at each timestep and calculate the velocity spectra from that. In the spectra you should see a peak and that will be where your shedding frequency is.

Originally Posted by rishitosh

hiii..

i m also new in dis CFD area...
i m working on flow over two side by side cylinders... @ Re= 100 and 200..
i got good results..

but i want to find out Strouhal number for all cases...

i came to know by finding shedding frequency with help of FFT, using cofficient of lift data, in fluent we can calculate Strouhal number...

but i m getting helpess..

how to use FFT to find shedding frequency....?????
-Rishi