CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   weird Cd of CFX result (http://www.cfd-online.com/Forums/cfx/95648-weird-cd-cfx-result.html)

colopolo December 27, 2011 07:33

weird Cd of CFX result
 
2 Attachment(s)
Dear All,

I simulated steady 3D cylinder to compute drag coefficient.

I attached two figures - one is grid and another is Pressure coefficient.

Boundary condition is like as;
Diameter :0.08 m
Inlet velocity 2.0 m/s
turbulence model : k-e model

When I compute Cd using CCL, the value is too low, it is around 0.25.
I also check force_x@cylinder, the value of force_x@cylinder is 0.04N.

I thought the value of Cd should be around 1.0 but the result is too low.

Do you have any idea or comments on this problem?

Best Regards,

paul83 December 28, 2011 05:36

Hello,

I assume your fluid is air? Then your Reynolds-Nummer is about Re=11000 and your flow is still laminar (but you use a turbulence model). At this Re-Number you also get "vortex-street instabilities", so i think a transient run would catch your drag more accurate.

And maybe check the area of your cylinder in z-direction or other parameters of the cd determination.

Furthermore your grid seams a little bit coarse to me. Maybe you could create a 2-dimensional model if you not have enough hardware power to run with a fine mesh in 3d.

This problem is a well known example in cfd though. You should find a lot of further information on the internet or the common literature.

Best Regards

Paul

colopolo December 28, 2011 06:15

Quote:

Originally Posted by paul83 (Post 337165)
Hello,

I assume your fluid is air? Then your Reynolds-Nummer is about Re=11000 and your flow is still laminar (but you use a turbulence model). At this Re-Number you also get "vortex-street instabilities", so i think a transient run would catch your drag more accurate.

And maybe check the area of your cylinder in z-direction or other parameters of the cd determination.

Furthermore your grid seams a little bit coarse to me. Maybe you could create a 2-dimensional model if you not have enough hardware power to run with a fine mesh in 3d.

This problem is a well known example in cfd though. You should find a lot of further information on the internet or the common literature.

Best Regards

Paul

Thanks for your opinion.
I will do that .

ghorrocks December 28, 2011 06:21

"Why is my simulation inaccurate" has been asked thousands of times, so have a look at the FAQ:

http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

And yes, your mesh is almost certainly too coarse.

If the fluid is air then Re=1e5 is in the transitional region. This is the hardest region to simulate accurately and will require great attention to detail to get accurate results.

colopolo December 28, 2011 06:22

Quote:

Originally Posted by ghorrocks (Post 337168)
"Why is my simulation inaccurate" has been asked thousands of times, so have a look at the FAQ:

http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

And yes, your mesh is almost certainly too coarse.

If the fluid is air then Re=1e5 is in the transitional region. This is the hardest region to simulate accurately and will require great attention to detail to get accurate results.

YOu are right. I just simply checked the drag coefficient without considering something. that is why I got an wrong value of CD
Thank you !!

ghorrocks December 28, 2011 06:36

Quote:

I just simply check the drag coefficient in the transitional region.
What are you talking about? I mean the flow is transitional between laminar and turbulent and the choice between laminar or turbulent is highly dependant on upstream conditions, surface quality and other small effects. It is also likely a significant section of the flow is laminar and a significant portion is turbulent.

So I have no idea what you comment is referring to.

colopolo December 28, 2011 06:43

Quote:

Originally Posted by ghorrocks (Post 337170)
What are you talking about? I mean the flow is transitional between laminar and turbulent and the choice between laminar or turbulent is highly dependant on upstream conditions, surface quality and other small effects. It is also likely a significant section of the flow is laminar and a significant portion is turbulent.

So I have no idea what you comment is referring to.

You are misunderstand my english. when I did this simulation, I just quickly check a drag coefficient and I got an weird value of Cd, so I just posted in here without any literature review and study.

At this time, I mean I fully agreed your opinion because I finally understood what I was wrong through literature review and your comments.

Thanks you for your quick response.


All times are GMT -4. The time now is 16:14.