CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)

 Bogey Jammer September 29, 2009 17:06

[FloWorks] Request advice for an airfoil calculation problem

Hi all, my first post here :D

As a hobby (and my job is not related with aircrafts), I'm trying to create a custom designed aircraft, in order to introduce myself to aerodynamics.

The WIP:

http://img39.imageshack.us/img39/8743/t02wip01.jpg

http://img3.imageshack.us/img3/5785/t02wip02.jpg

I've put it on a hold because it's time to design the wing. And since I want to understand the much info as possible, I decided to use the embedded CFD software to the CAD software that I use, respectively CosmosFloWorks and SolidWorks.

CosmosFloWorks is excellent for simple hydraulic problems, but in aerodynamics, it seems to bring weird results and behavior according to my past experiences.
So I decided to give it a chance again and started to compute a know airfoil detailed in:

http://img3.imageshack.us/img3/1900/...uinnasa.th.gif

My goal is to reach the real results in order to thrust this software and continue to use it for aerodynamics.
I've chosen to compute the airfoil at Reynolds=2.9 x 10^6 and alpha=4°, quite a simple situation, but as I expected the results are still wrong.. I'm currently trying to get a Cd between 0.01 and 0.015, and a Cl of 0.78, but the results say Cl=0.55 almost every time...

How I did:

I've set up a quite large computational domain but 0.024m thick (2D study situation):

http://img196.imageshack.us/img196/7...mputdomain.jpg

Note that the airfoil chord measures 0.6 meter.

Then the mesh:

http://img3.imageshack.us/img3/5302/maillage.th.jpg

Composed with parallelized rectangles, smaller near the airfoil, larger at corners, detailed in rounded surface, and automatically detailed during calculation with the flow (in the leading edge area and trail).

The coefficient calculation:

CosmosFloWorks computes many values and parameters but I still need to enter equations to get the coefficients.
I got them from calculated force components (Cl=Yforce/(q0*0.6*cos(4°)*0.024) and from pressure coefficients (equation is more complicated but the result is quite the same). So unfortunately both of these methods get a wrong result.

The speed setting and conditions:

I entered the windspeed according to the Reynolds number with standard atmospheric conditions. I'm sure it's good because I get a calculated Mach number of 0.2 like the real results. The material is set to adiabatic and without roughness.

Comparison with XFLR5:

Prior using CosmosFloWorks, I tried the airfoil in XFLR5 software using XFoil engine to analyze airfoils.

http://img29.imageshack.us/img29/941...atxflr5.th.gif

And surprisingly, it managed almost instantly to find the correct results...
I noted that the boundary layer just after the trailing edge is different in XFLR5, making me seriously doubting about CosmosFloWorks behavior.

http://img200.imageshack.us/img200/9...araison.th.jpg

For information, static pressure:

http://img200.imageshack.us/img200/5...tatique.th.gif

So, the questions are:
1. Am I wrong somewhere?
2. Can someone who has experience with this software comment this problem?
3. Can we really thrust this software?
4. Do I seriously need something to learn before going any further?