Mass flow through cyclic boundary face

jose_rodrig · March 21, 2011, 04:31

Hi

I am simulation a section of a combustor using cyclic BC. When I was calculating mass balance and energy balance I realized that the mass flow through the cyclic boundary faces does not cancel as it should (or shouldnt??).

Can somebody shed me some light on this issue?

THX

Jose

kalle · March 22, 2011, 08:38

I have a faint memory of someone telling me this patch integration is not parallel (the cyclic patch crosses processor boundaries). If you have the possibility, you could try to do this on a single machine to find out if that works...

As said, a faint memory, but might be worth to check.

Regards,
Kalle

jose_rodrig · March 22, 2011, 11:00

Hi!

Thanks for your reply!

I solved the problem and here is how:

In the configuration I had before, I had set the cyclic patch manually in the constant/polyMesh/boundary, changing the patch type from wall to cyclic as follows (I first exported my mesh from gambit and then coupled the 2 faces with createPatch utility):

Cyclic
{
type cyclic;
nFaces 29022;
startFace 775347;
transform rotational;
rotationAxis ( 0 0 1 );
rotationCentre ( 0 -0.04 0 );
}

Then, if I eliminate the transformation entries ("transform", "rotationAxis" and "rotationCentre"), OF configures the patch automatically and now the mass balance is correct as it should be, SOLVING THE PROBLEM.

Anyway, I still dont understand what this parameters mean/do (I though I knew before

). I intend to look at the source for their meaning. However if somebody has a hint or answer ready to write here, it would be most appreciated.

Regards

José

March 21, 2011, 04:31	Mass flow through cyclic boundary face	#1
jose_rodrig Member José Rodrigues Join Date: Jun 2010 Location: IN+/IST Lisbon Posts: 53 Rep Power: 15	Hi I am simulation a section of a combustor using cyclic BC. When I was calculating mass balance and energy balance I realized that the mass flow through the cyclic boundary faces does not cancel as it should (or shouldnt??). Can somebody shed me some light on this issue? THX Jose

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Netgen] Import netgen mesh to OpenFOAM	hsieh	OpenFOAM Meshing & Mesh Conversion	32	September 13, 2011 05:50
RPM in Wind Turbine	Pankaj	CFX	9	November 23, 2009 04:05
How can apply mass flow rate boundary condition?	Sima	Phoenics	1	December 1, 2007 18:55
mass flow inlet boundary condition	vivian	FLUENT	5	April 21, 2006 06:23
How to get specie mass fraction at a boundary face	shuqin	FLUENT	0	February 24, 2006 01:52

March 22, 2011, 08:38		#2
kalle Senior Member Karl-Johan Nogenmyr Join Date: Mar 2009 Location: Linköping Posts: 279 Rep Power: 21	I have a faint memory of someone telling me this patch integration is not parallel (the cyclic patch crosses processor boundaries). If you have the possibility, you could try to do this on a single machine to find out if that works... As said, a faint memory, but might be worth to check. Regards, Kalle

March 22, 2011, 11:00		#3
jose_rodrig Member José Rodrigues Join Date: Jun 2010 Location: IN+/IST Lisbon Posts: 53 Rep Power: 15	Hi! Thanks for your reply! I solved the problem and here is how: In the configuration I had before, I had set the cyclic patch manually in the constant/polyMesh/boundary, changing the patch type from wall to cyclic as follows (I first exported my mesh from gambit and then coupled the 2 faces with createPatch utility): Cyclic { type cyclic; nFaces 29022; startFace 775347; transform rotational; rotationAxis ( 0 0 1 ); rotationCentre ( 0 -0.04 0 ); } Then, if I eliminate the transformation entries ("transform", "rotationAxis" and "rotationCentre"), OF configures the patch automatically and now the mass balance is correct as it should be, SOLVING THE PROBLEM. Anyway, I still dont understand what this parameters mean/do (I though I knew before ). I intend to look at the source for their meaning. However if somebody has a hint or answer ready to write here, it would be most appreciated. Regards José