CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   need help desperately (http://www.cfd-online.com/Forums/main/69063-need-help-desperately.html)

pixie October 10, 2009 19:54

need help desperately
 
Hi all
hope you can help me with this problem...
my project i due soon and I am totally stuck for the past few days!!!
I am trying to determine flow over an airfoil with plain as well as external flaps at different gaps and deflections angles- with and without leading edge slat . got the geometry set up in gambit fine, meshing is unstructured (need to run my different grids to determine grid tests and grid dependency). The problem is when i read the mesh file in fluent it does NOT see my domain boundary! it sees the airfoil, flap and slat as wall, but the domain boundaries are seen as wall boundaries, even though i have set them at pressure farfield [i have a CH mesh with domain size 12.5c (vertical)and 21c (horizontal) from airfoil] I've even tried changing the boundary condition from farfield to velocity inlet and outflow, i've redrawn the entire geometries and mesh twice already, but still have the same problem.] Also I posted a thread regardig boundary layer meshing, and no matter what i tried the direction of the boundary layer is always into the airfoil! I've figured that I can adapt the boundary layer in fluent, but i've gotta get my boundary conditions right!!!
PS If I change the outer domain from wall to anything else in fluent it gives me errors when initializing the solution and does not calculate anything...

PPS thanx in advance for the help!

Jullie

Ahmed October 11, 2009 13:21

Quote:

Originally Posted by pixie (Post 232120)
Hi all
hope you can help me with this problem...
my project i due soon and I am totally stuck for the past few days!!!
I am trying to determine flow over an airfoil with plain as well as external flaps at different gaps and deflections angles- with and without leading edge slat . got the geometry set up in gambit fine, meshing is unstructured (need to run my different grids to determine grid tests and grid dependency). The problem is when i read the mesh file in fluent it does NOT see my domain boundary! it sees the airfoil, flap and slat as wall, but the domain boundaries are seen as wall boundaries, even though i have set them at pressure farfield [i have a CH mesh with domain size 12.5c (vertical)and 21c (horizontal) from airfoil] I've even tried changing the boundary condition from farfield to velocity inlet and outflow, i've redrawn the entire geometries and mesh twice already, but still have the same problem.] Also I posted a thread regardig boundary layer meshing, and no matter what i tried the direction of the boundary layer is always into the airfoil! I've figured that I can adapt the boundary layer in fluent, but i've gotta get my boundary conditions right!!!
PS If I change the outer domain from wall to anything else in fluent it gives me errors when initializing the solution and does not calculate anything...

PPS thanx in advance for the help!

Jullie

Relax pilgrim, by definition, all boundaries are assigned the label wall unless modified. do you have your geometry correctly defined?
If that situation persists, your only help is your support engineer, your installation might be incomplete or corrupted!
Take a color screenshot of your geometry and attach it to your reply
Good luck

...................... Open Source for ever.............................................. ......

pixie October 13, 2009 16:43

Thanx so much Ahmed
It turns out that when subtracting the airfoil face from the domain, new edges were created and i had to specify my BC based on these new edges...took me 2 days to figure this out!!!
phew, thank goodness thatz figured out!
but, alas just when u think its smooth sailing (i don't think such a thing exists with cfd, and keep asking myself WHY i'm doing this project in cfd) i have another problem, non of my incompressible flow (inviscid solver) solutions converge, only at very low angles of attack (<2deg) and like i said my project is due soon - need to hand in a draft on friday....hope you can help
:)
Thanx

Ahmed October 13, 2009 19:27

Quote:

Originally Posted by pixie (Post 232483)
Thanx so much Ahmed
It turns out that when subtracting the airfoil face from the domain, new edges were created and i had to specify my BC based on these new edges...took me 2 days to figure this out!!!
phew, thank goodness thatz figured out!
but, alas just when u think its smooth sailing (i don't think such a thing exists with cfd, and keep asking myself WHY i'm doing this project in cfd) i have another problem, non of my incompressible flow (inviscid solver) solutions converge, only at very low angles of attack (<2deg) and like i said my project is due soon - need to hand in a draft on friday....hope you can help
:)
Thanx

Pilgrim
One more time, relax , relax and relax.
No need to put extra pressure on yourself, at your age, you ought to be looking for birds not sitting before your computer and doing CFD.
Here are few observations based on your description
1- incompressible has nothing to do with inviscid, incompressible is a constant properties flow while inviscid is a zero viscosity flow.
2- if using the inviscid solver why are you insisting on adding a boundary layer to your domain?
Boundary layers are the result of viscosity!!!!!!!!
Here is a little piece of advice, define clearly the objectives of your analysis, then select the correct solver accordingly.
finally log on fluent users center and download a tutorial about your analysis. There are two of them.
Good luck

NB: Computers are not smart or intelligent at all, they have no way of knowing what is on your mind, they just execute your instructions, so pay attention to the details of your analysis.



................................................. Open Source for ever....................................

pixie October 13, 2009 23:44

Heehhehe Ahmed ;) At this moment in time, relaxing is for slackers, 3 weeks left of semester with that much time to complete 3 major assignments...the least of my worries being the thesis on cfd! And at my age i should probably not still be studying (or rather tudying again), but be progressing in my career and life but thanx for the compliment. :)
ok so maybe i didn't make myself clear in previous posts, cos i totally know both points you have made ;) my problem seems to be one of execution mostly.
I am running a range of cases- plain airfoil as well as flapped foil over a range of speeds (airfoil at M=0.2 to 0.8), and it just seems every step of the way that there is at leat days of 'figuring out' to do... no free lunch here.
I've manages to run transonic and subsonic cases. now am running low M (incompressible cases) At the moment am trying to get comparable results for the airfoil alone so that when i run the flapped cases I have a basis to say why my results are reasonable. had major issues trying to do grid for flapped cases and now none of my airfoil alone results are converging.... i still them have to optimise this results!


All times are GMT -4. The time now is 11:33.