Accuracy of results

Jefferson_dhv · July 25, 2014, 16:22

Hello,

I'm trying to run a case in sonicFoam of a NACA 0012 airfoil, but I'm not being able to get accurate results, not even close to those expected regarding the CD. I tried to change some parameters, even the turbulence model, but I could not improve really much the results since I'm a recent new user of OpenFOAM. So I ask you what can I do to have more accurate results.

Here is the case:
/0
U
P
T
k
epsilon
mut
alphat

/system
controlDict
fvSchemes
fvSolution

The expected value for CD is around 0.008.

Thanks in advance,
J. Vieira

miladrakhsha · August 31, 2014, 13:17

Dear Jefferson,

I believe there was a very good topic in this page where people discussed the same thing. However, in short, you can never predict drag coefficient with a high Reynolds turbulent models. In order to accurately predict the drag you have to accurately resolve the flow close to the airfoil which needs y+<1.
Nevertheless, I would like to know whether or not you were able to predict the lift coefficient in the stall region for that particular airfoil.

Jefferson_dhv · August 31, 2014, 19:27

Hi miladrakhsha,

I appreciate your help. Since I've posted I improved a lot my results, the main problem was the quality of the mesh, the cells at the airfoil wall weren't thin enough and some discontinuities were causing numerical instabilities. At this moment I'm obtaining reasonably good results, but in my study I'm not watching the stall region, the interest relies on low angles of attack.

Best,
Jefferson Vieira

August 31, 2014, 13:17		#2
miladrakhsha New Member miladrakhsha Join Date: Aug 2012 Posts: 29 Rep Power: 14	Dear Jefferson, I believe there was a very good topic in this page where people discussed the same thing. However, in short, you can never predict drag coefficient with a high Reynolds turbulent models. In order to accurately predict the drag you have to accurately resolve the flow close to the airfoil which needs y+<1. Nevertheless, I would like to know whether or not you were able to predict the lift coefficient in the stall region for that particular airfoil. Last edited by wyldckat; August 31, 2014 at 19:02. Reason: removed bold formatting from the whole text, to make it easier to read

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Transient Run - Output "Time" in partial results?	evcelica	CFX	2	May 16, 2012 22:36
Varying time-step and Results freequency	jstaylor89	CFX	1	April 17, 2012 03:00
CFX Varying time-step and Results freequency	jstaylor89	Main CFD Forum	0	April 16, 2012 12:14
Lost Scenario Results..??	Fugacity	Autodesk Simulation CFD	1	February 24, 2012 11:48
Problems with repeating results	Lee	Siemens	4	May 26, 2006 04:39

August 31, 2014, 19:27		#3
Jefferson_dhv New Member Jefferson Vieira Join Date: Jul 2014 Location: Brazil/Scotland Posts: 12 Rep Power: 12	Hi miladrakhsha, I appreciate your help. Since I've posted I improved a lot my results, the main problem was the quality of the mesh, the cells at the airfoil wall weren't thin enough and some discontinuities were causing numerical instabilities. At this moment I'm obtaining reasonably good results, but in my study I'm not watching the stall region, the interest relies on low angles of attack. Best, Jefferson Vieira