How many geometry domains?

davek · September 29, 2010, 09:06

I'm wondering what is good practice for relatively complex geometries. If the direction and velocity of flow is very different throughout the geometry, should I split it into a bunch of domains with interfaces so that I can set their initializations seperately? Or should I be keeping it as simple as possible?

Particularly with slow multiphase flows, CFX seemed to have a lot of difficulty producing convincing results if I didn't split it up and set initializations for velocity, direction, and mass or volume fraction seperately.

ghorrocks · September 29, 2010, 18:53

You don't need to split and join with interfaces just to set initial conditions. Initial conditions can be set on sub-domains, interpolation functions or CEL expressions. The simulation will partition better if it is one domain with no interfaces.

Obviously you need to set realistic initial conditions. This is standard CFD practice.

davek · September 30, 2010, 08:50

I will try subdomains again. I think I tried to use them before but I couldn't get them working.

September 29, 2010, 09:06	How many geometry domains?	#1
davek New Member David Join Date: Sep 2010 Posts: 7 Rep Power: 15	I'm wondering what is good practice for relatively complex geometries. If the direction and velocity of flow is very different throughout the geometry, should I split it into a bunch of domains with interfaces so that I can set their initializations seperately? Or should I be keeping it as simple as possible? Particularly with slow multiphase flows, CFX seemed to have a lot of difficulty producing convincing results if I didn't split it up and set initializations for velocity, direction, and mass or volume fraction seperately.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] O-grid around wing geometry	Anorky	ANSYS Meshing & Geometry	15	November 27, 2017 13:43
[Gmsh] Problem meshing an imported geometry	Alicia	OpenFOAM Meshing & Mesh Conversion	0	March 30, 2010 04:53
Simulation of Flow through Complex 3D Geometry	EmersonKB	CFX	5	July 2, 2009 08:17
Creating a flow geometry with domains with different porosity	arkasshka	OpenFOAM Running, Solving & CFD	15	December 7, 2007 02:11
vitual _ real	deneb	FLUENT	3	January 22, 2007 04:31

September 29, 2010, 18:53		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	You don't need to split and join with interfaces just to set initial conditions. Initial conditions can be set on sub-domains, interpolation functions or CEL expressions. The simulation will partition better if it is one domain with no interfaces. Obviously you need to set realistic initial conditions. This is standard CFD practice.

September 30, 2010, 08:50		#3
davek New Member David Join Date: Sep 2010 Posts: 7 Rep Power: 15	I will try subdomains again. I think I tried to use them before but I couldn't get them working.