|
[Sponsors] |
May 5, 2011, 08:15 |
"Unefficient" wall boundary
|
#1 |
Member
Charles
Join Date: May 2011
Location: France
Posts: 77
Rep Power: 14 |
Hi,
I'm working on an NACA 0012 airfoil equipped with an airbrake. But when running the calculations, I have some pretty high residus (10^-5 for the velocities and 10^-2 for the continuity) though Cl and Cd are more or less converging. I've done tons of tests and I noticed that each times, I had some strange behaviours around my airfoil : Here's my mesh, Quad for the entire area except around the airfoil where I couldn't use Quad Elements, so it's "Tri Elements" (I got the boundary layer refinement by meshing the edges first). And there is a close look of my transparent wall in Fluent : The vectors looks very strange close to the Wall. It's easier to spot with high angles of attack (20° in that case). My Wall boundary is attached to the airfoil edges, and I double checked for double edges, faces etc.... My idea was to attach the Boundary condition to the foil Face but when I create one Gambit can't export it to Fluent because it's not meshed, a problem I couldn't solve... Someone got an idea of my problem ? Thanks, Charles |
|
May 5, 2011, 11:04 |
|
#2 |
Member
Kailash
Join Date: May 2011
Location: London, UK
Posts: 45
Rep Power: 14 |
hi,
I didnt quite understand ur problem. But some initial comments on ur grid, 1. The transiiton between the tri to quad cells at the trailing edge is not quite good. try clustering near the trailing edge. 2. what is the y+ of the grid? If you dont specify the BC to the aerofoil in Gambit, you can specify it as noslip wall in Fluent. Cheers |
|
May 5, 2011, 19:45 |
|
#3 |
Member
Charles
Join Date: May 2011
Location: France
Posts: 77
Rep Power: 14 |
Hi,
Thanks for your advice, I've add some extra mesh just behind the trailing edge. I change my strategy and made the entire mesh with Quad Elements, hoping that it could solve my problem. Murphy's law, it didn't solve anything. I get some amazing results with low pressure under the foil and high pressure over the foil. I can understand the high pressure zone above the foil due of the airbrake, but low pression under, it's impossible.... I also noticed that the air is hitting the foil quite higher that excepted. I rotated the grid by 5° and set up 30m/s on the X axis, so I should have a 5° angle of attack but if you take a close look, I get a negative one instead. The vectors are straight when exiting the inlet but their direction is changing and they are progressively heading down when approching the foil. k_k, you advised to check the Yplus but I'm using a laminar model, so I can't check Yplus right ? After a week of work, I'm iteratively losing all hope in life... I really need some help. If someone is okay with that, I can send him or her my .msh file. There are my screens : Mesh My unrealistic pressure And the velocity, note the strange direction of the vectors before hitting the foil |
|
May 6, 2011, 03:19 |
|
#4 |
Member
Kailash
Join Date: May 2011
Location: London, UK
Posts: 45
Rep Power: 14 |
hi,
Are you running a incompressible, laminar case? How far are the boundaries from the aerofoil? What is the shape or your inlet boundary? straight or curved? if it is curved, in the inlet BC check if you have specified velocity components rather than normal to boundary. If possible, try to create a domain with straight inlet boundary. For an incompressible case, it is better to rotate the aerofoil alone to the required angle of attack (not the complete domain). cheers |
|
May 6, 2011, 03:22 |
|
#5 |
Member
Kailash
Join Date: May 2011
Location: London, UK
Posts: 45
Rep Power: 14 |
your new grid, the lower surface of the aerofoil is very coarse to resolve the boundary layer. Make it more dense.
|
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
no-slip wall boundary condition | Atit Koonsrisuk | CFX | 3 | July 10, 2014 17:10 |
natural convection | mehrdadeng | CFX | 10 | February 25, 2011 05:25 |
CFX does not continue | Shafiul | CFX | 10 | February 17, 2011 07:57 |
CFX doesn't continue calculation... | mactech001 | CFX | 6 | November 15, 2009 21:25 |
Wall Boundary Condition in k - epsilon model | abhijeet vaidya | Main CFD Forum | 0 | July 14, 2002 03:18 |