I am using SA and LES to simulate flow pass a sphere (both Fluent 5 and Fluent 6). I found that drag coefficients are not right. But

( drag coefficient ) / (density * U-infin. *U-infin.)

is very close to the experimental data. Is it possible that Fluent has a bug in the drag coefficient calculation?

Thanks.

The drag coefficient is calculated using the values in: report > reference values So its possible that your reference area is set wrong.

Dear Glenn:

Your message is not clear for me. I didn't set the reference area. I got the drag coefficient from:

Solve - > Monitors - > Force - > drag coefficient

and that value matches with

Report - > Forces - > total coefficient for vector (1,0,0).

Thanks,

-- Jean

Hi Glenn,

I understand what you mean now. In Report -> Referance Values, the Area = 1 m^2 and it should be Pi*D^2. Since D=1, so it will be Cd_new=Cd/Pi.

The value is still not right!

Thanks,

-- Jean

Guys, Sorry to interject, but I have a converged flow field with no objects in it and was wondering if I can report the drag coefficient on an imaginary sphere where I would set its properties (i.e. density, diameter)?

Rex

Im not too sure about this, but I think that you can only calculate the drag on a 'wall' boundary condition. However maybe someone knows different????

Rex, what do you mean by an imaginary sphere? How do you model a sphere as being imaginary?

I mean that if there was a sphere in the flow field (there isn't one really), could you report, for instance, the drag force being experienced at that point?

Rex

If you imagine its there, then the model will not realize it and it wont give you anything. But if you incorporate it in your model, you can set the reference values, and monitor the drag coefficient.

Your answer is yes. Report->reference values: set the frontal area, the velocity, density, etc Solve->monitors->force: select drag and the sphere wall

Hey Rex, surely the flow would be perturbed by the sphere, so changing its characteristics and hence the Drag ? unless you can some how take into account the effect the sphere has upon the flow ? I guess though if the shere were very small then its influence on the flow would be negligable, but then so would its drag ??? What is the interest of calculating the drag this way round ?? bob

That's an interesting study.

What's is the status of your project? For who are you doing this?

For wich values of Re are you calculating? How is your mesh?

are you willing to share some results with us?

greetings, Laika, still orbiting

The so called mechanistic models for particulate transport I am using in conjunction with Fluent flow data assume that the spheres do not effect the characteristices of the flow field (velocity profile, etc.), so how the particles modify the flow is not an issue I'm looking at. I just want to know the forces being experienced over a certain (pre-defined by me) area are.

Dear Glenn,

As you have mentioned that the drag coefficient is calculated using the values in: report > reference values. I have found that in my case, the problem is not reference area but reference velocity. For both pressure coefficient and drag coefficient, correction should be made by divided (V_inf * V_inf).

Many thanks,

-- Jean

Hi Laika,

Thanks for the email. I am willing to share some my results with you. I will email you later.

-- Jean

thanks!!

Laika, still orbiting