CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FloEFD, FloWorks & FloTHERM (http://www.cfd-online.com/Forums/floefd-floworks-flotherm/)
-   -   Simulation of a complex wing in solidworks flow simulation (http://www.cfd-online.com/Forums/floefd-floworks-flotherm/85329-simulation-complex-wing-solidworks-flow-simulation.html)

niels1900 February 23, 2011 07:07

Simulation of a complex wing in solidworks flow simulation
 
Hi all,

Im trying to simulate an aircraft wing in flow simulation and retreive information such as the lift coefficient and the drag coefficient of the total wing.
So i started to do a tutorial of an airflow over a wing in a 2D situation described in (http://books.google.com/books?id=w_T...page&q&f=false) Chapter 4.
When i followed the tutorial i get quite the same results as the tutorial.
Then i tried to simulate a Clark V profile with exactly the same settings, even the same model size. But the results are not satisfying. The lift coefficient we calculate from the given Y-force is about 0.057, but it must be about 0.2259 according to the Cl/alfa table. Here im working with the same reynoldsnumber.
I also tried other wing profiles and played with some settings of the simulation but i dont get any realistic results. I think the turbulence intensity and the turbulence lenght are an important factor, but I dont know what the exact meaning is of these variables and how to determine their correct values.
So im stuck at a relative simple 2D simulation of a wing, and i also want to do a 3D simualtion (in air and also in water) which I think is more complex to perform.

Can anyone help me and tell me what im doing wrong?
And is the flow simulation option in solidworks accurate enough to determine the lift and drag of a wing? I want to use the values for a school project, so it must be quite accurate.

niels1900 March 1, 2011 05:44

Really nobody with some suggestions?:(

Kiat110616 April 5, 2011 13:10

I think the geometry is effect with the results. You should be create the geometry by reference from origin. Can you show your model picture?

niels1900 April 7, 2011 05:59

Pictures
 
Hi,

I do not really understand your comment, but here are some pics anyway ;)
To draw the profile i used curve through points (imported from profili). I had to split it into two curves because otherwise the function made the trailing edge a round edge, to split it up into two at the trailing edge it becomes a sharp edge.

http://img849.imageshack.us/img849/9207/clarkvlarge.pnghttp://img121.imageshack.us/img121/4416/clarkvsmall.pnghttp://img857.imageshack.us/img857/1565/clarkv3d.png
Btw, i also made a model where a large plate on the YX front plane is attached to the wing. but then the wing visually shows a very bad curve, the upper part is made up of 1 line from half of the chord to the trailing edge. Is this pure a visual problem in solidworks, or does it physicly changes my object?

Thanks for responding btw!

Kiat110616 April 7, 2011 10:30

2 Attachment(s)
The geometry is important to obtain the result accuracy. The picture below shows different coordinate. The center of mass of the first case is started at the origin but offset from the origin for the second case. The boundary conditions of both cases were same condition. So, the results are different.

Attachment 7174

Attachment 7175

Turbulent intensity is the ratio of the root mean square of the velocity fluctuations to the mean flow velocity. In your case, the low turbulence intensity should be apply on it.

Good luck

Kevin De Smet April 9, 2011 09:30

Quote:

Originally Posted by niels1900 (Post 296542)
I want to use the values for a school project, so it must be quite accurate.

Wrong answer? ;)

This kind of problem is fairly theoretical and I think its known that general purpose solvers fair less than optimally for these kinds of problems such as 2d cross sections of airfoils for lift and drag.

alainF April 20, 2011 10:44

IMO Your discrepancy between the simulation and theorical results are just too large to be explained by turbulence farfiel values.

You should check that you use the correct coordinate system. Moreover, as you made a 2D calculation your actual domain thickness is usually far more smaller than the modelled geometry.

So even if you modelled a 1 m thick airfoil the integration made by flowsimulation can be made on a smaller lenght that would explain why your lift force is so small.


All times are GMT -4. The time now is 23:27.