[neprendo]

Hey guys.

No matter what I do I can't seem to get stop the coefficient of drag becoming negative when plotted using the monitors in fluent v.14. I'm modelling a 2-D NACA 0012 aerofoil in a pressure far field at a velocity of 10m/s and angle of attack of 9 degrees using a steady pressure based solver. Mesh quality is 0.83 and refined around the aerofoil. The URF are 0.1 for pressure and momentum and default for the others. So far I have been adjusting the boundary conditions at the outlet to reflect what is seen at the inlet and playing around with the URF.
Any help on the topic would be appreciated as I'm kinda at a lose of what to try next.
The prj file I used in Icem is attached and I'll happily attach anything else that is requested.

Thanks in advance
Nick

[flotus1]

Originally Posted by neprendo

I'll happily attach anything else that is requested.

How about sketches or screenshots of the computational domain where we can see the boundary conditions, the coordinate system and the mesh.

I have been adjusting the boundary conditions at the outlet to reflect what is seen at the inlet

I dont fully understand that. Are you using a pressure far field boundary condition or not?

[neprendo]

Hi,

I think I have attached what you asked for (over 2 posts). Sorry if some of the screen shots are a little crap, I struggled with getting some of them within the file size limit. The word doc has a labelled diagram of the domain without the mesh and the coordinate system with the origin at the leading edge of the aerofoil.

In regard to the boundary conditions, I was initially using a pressure far field at the outlet, but that was forcing me to have the fluid as an ideal gas (for density) and was defining a mach number of 0.6 and component directions for the flow out. I thought that by matching the mach number and component directions of the flow at the outlet to the inlet it might help. It didn't really help so I changed the outlet boundary condition to a pressure outlet.

Nick

[neprendo]

these are the last 2 screen shots

thanks
Nick

[flotus1]

There are some minor issues with the setup, but none of them could explain a negative drag if you are using velocity inlet and pressure outlet.
Is there any particular reason why you used under-relaxation factors of 0.1 for pressure and momentum? Unless there is something seriously wrong with the setup, the default values of 0.3 and 0.7 respectively should work much better.
I hate to bring this up again and again, but as we dont know what your background in CFD is, did you make sure that the solution we are talking about is converged?
These two links should help you with this.
Once you covered this topic, we can have a closer look at the simulation.

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

http://www.cfd-online.com/Forums/flu...nvergence.html

BTW: please dont post images as microsoft word documents. You can always save them as an image file. This way, even people without microsoft software can have a look at them.

[neprendo]

Thank you for your help. My background in CFD is pretty limited, I'm a engineering masters student. This my first major project using it. I'll check out those links and try sort the problem. And thanks for the heads up about word docs.

Nick