
[Sponsors] 
February 16, 2010, 15:24 
Wrong orientation of boundary layer in 3D

#1 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Hello everybody,
I want to do a 3dimensional analysis in Fluent and the most important part of my assembly is a cylinder. The problem is in Gambit when I want to create a boundary layer around my cylinder, that Gambit creates it just inside the cylinder which is a solid body and for my problem of an external flow I need the boundary layer obviously on the outer surface of the body. Maybe someone could help me and knows what can I do that my boundary layer is on the outer surface. Thanks for your help... 

February 17, 2010, 02:04 

#2 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
Except if you need keeping the solid domain, you may delete it. Then your problem disappear, because only the fluid domain remains.
Else you can click on middle button of your mouse for switching BLdirection
__________________
In memory of my friend Hervé: CFD engineer & freerider 

February 17, 2010, 05:25 

#3 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Hey Max,
thanks for your help but i tried this already. I subtracted my cylinder from the rest of the domain and specified it as a wall the rest of the domain is a fluid but it still doesn´t work, my boundary layer ist still inside the cylinder and i can´t choose the direction. 

February 17, 2010, 05:56 

#4 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
If your cylinder has a thickness, and you substracted it, then your cylinder may have one adjacent domain: the fluid.
You can post a picture
__________________
In memory of my friend Hervé: CFD engineer & freerider 

February 17, 2010, 06:26 

#5 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Yes, and i thought then if there is just one adjacent domain, the boundary layer should be in the fluid domain. on the picture there are two cylinders, just consider the inner one (smaller) the other one is just an interior.


February 17, 2010, 07:01 

#6 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
if your problem is symmetric, work with one quater of your domain.
if it is not symmetric, split your dommain into 4 domains. The split planes will resolve your issue Hope it helps
__________________
In memory of my friend Hervé: CFD engineer & freerider 

February 17, 2010, 14:58 

#7 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Unfortunately it isn´t a symmetric problem. And also to split the domain into 4 domains wasn´t successful...
Is there maybe an alternative to create the boundary layer manual? 

February 17, 2010, 15:10 

#8  
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
Quote:
If you still have problem you can provide me your geometry, I will described you how to proceed
__________________
In memory of my friend Hervé: CFD engineer & freerider 

February 17, 2010, 15:20 

#9 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
This would be great...
I created my geometry in mm, so the ellipse has a length of 10000mm (x) a height of 6000mm (y) and a depth of 12000mm (z). The cylinder which is used as a interior has a diameter of 600mm and a depth of 12000mm. The cylinder which is the wall has a diameter of 300mm and a depth of 9000mm, the cylinder is always moved into the middle of the domain (1500mm in zdirection) thanks for your help 

February 17, 2010, 15:28 

#10 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
we are agree that the 300mm diametercylinder is considered as a hole and you want to attach the BL around its outter surfaces, right?
__________________
In memory of my friend Hervé: CFD engineer & freerider 

February 17, 2010, 15:35 

#11 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Yes exactly, thats right.


February 18, 2010, 02:37 

#12 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
Ok as I said yesterday, you can create a plane (yz) (20000x20000) and a plane (xz) (20000x200000)
Split all the volumes with those planes > your domain will be splitted into quarter. Then you can create your BL.
__________________
In memory of my friend Hervé: CFD engineer & freerider 

February 19, 2010, 10:56 

#13 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
I tried your method and it works finally. Thank you very much for your help, without that I would have much more trouble because of this problem.
Thank you 

February 19, 2010, 12:53 

#14 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Now I have the next problem
The orientation of the boundary layer and anything works but when I mesh my volume around the boundary layer the following error occurs: ERROR: Currently, 3D boundary layers are limited to flat face groups with END conditions around the boundary. The problems occur in volume... along the following edges: Edge... You can turn the MESH.VOLUME. USE_3D_Boundary_Layers default off and try again. I tried already to turn the default off, but the error still occurs. And as I saw from your picture you had not this problem. Maybe you have an idea to solve also this problem? Sorry for bothering you 

February 19, 2010, 14:57 

#15 
New Member
Join Date: Dec 2009
Posts: 26
Rep Power: 8 
Okay, I solved this problem now, but anyway thank you.


March 9, 2010, 10:55 

#16 
New Member
stiko
Join Date: Mar 2010
Posts: 2
Rep Power: 0 
hello every body
the problem consists of a cylindre and an interior sphere, i can't create the boundary layer at the exterior side of the sphere, it also created at the interio side. I substract the 2 volumes and i make a split with a plane and also the same probleme 

March 10, 2010, 02:22 

#17 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
Do you need to keep the sphere (solid conduction?)?
If not delete the sphere, the outter surface should remain, and retry applying a BL.
__________________
In memory of my friend Hervé: CFD engineer & freerider 

March 10, 2010, 10:13 

#18  
New Member
stiko
Join Date: Mar 2010
Posts: 2
Rep Power: 0 
Thanks MAX
But the problem that the sphere and the cylinder are one volume(because i subtracted the shere from the cylinder) Quote:


March 10, 2010, 10:49 

#19 
Super Moderator
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,149
Rep Power: 32 
post a picture with your BL
__________________
In memory of my friend Hervé: CFD engineer & freerider 

April 2, 2011, 07:11 

#20  
New Member
rohitjvbibin
Join Date: Sep 2010
Posts: 13
Rep Power: 8 
Quote:
 Hey can you tell me how to turn it off default....Where can i find the option..I also tried to volume mesh my helical geometry but this 3d boundary layer problem persists....No matter what size i provide... 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Turbulent Boundary Layer on a Flat Plate  Hoshang Garda  FLUENT  1  November 27, 2013 11:24 
Combining BCs: wall  outlet. Boundary layer disappears  MartinaF  OpenFOAM Running, Solving & CFD  1  July 20, 2009 18:14 
Boundary Layer Meshing of a nosecone  Chris  FLUENT  1  March 15, 2005 13:42 
Multicomponent fluid  Andrea  CFX  2  October 11, 2004 05:12 
Wall turbulence, viscosity, boundary layer  Patrick Godon  Main CFD Forum  1  November 5, 2003 16:39 