CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Meshing pipe intersection in an exhaust system (http://www.cfd-online.com/Forums/fluent/48720-meshing-pipe-intersection-exhaust-system.html)

Fabio July 9, 2008 02:58

Meshing pipe intersection in an exhaust system
 
Hi all, i'm trying to mesh a volume composed by two pipes intersecting with an angle of 21 degrees between their axes. I'm doing the GAMBIT tutorial, in several examples Cooper scheme is used, at this point i'm in bubt if the effort of dividing the volume in order to use Cooper to achieve some advantage in calculation or use a Tet\Hybrid scheme. For the acuueracy of the results i need to apply boundary layers, i have tryed to use Cooper but there i a proble with the faces to project along because the one where the two pipes intersect is seen as a hole, i'm tyring to solve with a virtual surface but this on is not well shaped. Any advice with a better scheme or a smart volume subdivision is welcome.

Thanks in advance.

CDE July 9, 2008 20:02

Re: Meshing pipe intersection in an exhaust system
 
You can mesh this with cooper if you split the volumes correctly. Its is difficult to descibe how to do this here!

An alternative is to use 3D face boundary layers and mesh with Tet mesher.

Fabio July 10, 2008 02:50

Re: Meshing pipe intersection in an exhaust system
 
Thank you for the answer,i have tryed to split the volume by means of a surface at the intersection that would follow the generatrix of the main cylinder, but instead GAMBIT creates a virtual surface that partially follow the cylinder generatrix and partially the intersection curve generatrix, so it's shape is strange and i can't project along, i'm looking to understand how to use the external face of the main cylinder as Host. Even if i success in that i don't know if i should use Cooper on the smallest cilinder because it has a flat end and another with a sharp edge due to the 21 degrees angle. Btw is Cooper scheme so important to avoid convergence problem instead of using Tet/Hybryd? About the 3D boundary layers, to use them i was forced to change a default variable because one of the end faces wasn't planar, i don't remeber exactly which option.

Fabio July 14, 2008 11:54

Re: Meshing pipe intersection in an exhaust system
 
Ok i've divided the volume in 2 parts. But due to the shape of the pipe with 21 degrees i get an high skewness with Copper scheme, is ok if i use 2 different schemes for the 2 parts?

Fabio July 17, 2008 04:33

Re: Meshing pipe intersection in an exhaust system
 
Hi all again, i have solved my meshing prolems by using a 3D boundary layer on the wall surfaces and filing the volume with a Tetrahedral\TGrid scheme. Now i have completed a short steady simulation to obtain the field in which the droplets will be injected. Even with a very corse grid (only 12300 cells) i got the expected results for the velocity field, but instead the temperature field (both Static and Total) are the same in magnitude and contours and almost constant everywhere even if the velocity varies greatly, i got an inlet boumdary condition with 0.054 Kg/sec and 337 celsius at the outlet i have used Outflow maybe Pressure would be more appropriate?I have supposed no wall heat flux due to lack of data. Any ideas why the temperature field is like that? Thanks in advance.


All times are GMT -4. The time now is 04:43.