Problem with Permit No Intersection on Parallel runs

Paul87 · July 31, 2015, 15:49

Hi all! This is my first post on the forum, after having lurked for quite a while

I'm currently having an issue with a simulation on CFX, simulating a heatsink subjected to some flow, with heat generating components making contact with the heatsink. In my CFX model though, there is a small gap between some of my components and my heatsink (in real life, there is a gap pad there, which I model using a thermal contact resistances). I compensate for this by ticking Permit No Intersection under Meshing Connection.

When I run this model in Serial mode, everything works 100%. However, when I switch it to run on Platform MPI Local Parallel (4 partitions, as I have a 4-core CPU), I get the following error:

Quote:

ERROR #001100279 has occurred in subroutine ErrAction.
Message:
The domain interface 'Heatsink PSU B' does not intersect at all.
This is only supported for dynamically re-intersected interfaces
(e.g. transient rotor stator interfaces or non-stationary interfaces
in moving mesh cases). The partitioner can not handle this
scenario for static interfaces.

+--------------------------------------------------------------------+
ERROR #001100279 has occurred in subroutine ErrAction.
Message:
Stopped in routine SU_MESH_GGI

I assume that this is an issue with the partitioner (as mentioned in the error message). Is there a setting or a way to circumvent this, or do I need to physically model my gap pads in order to avoid this issue?

Thanks in advance for the assistance!

fresty · July 31, 2015, 18:26

I may be wrong at this and would look forward for reassurance from the more experienced chaps but IMO if your interface is static as mentioned in the error message, you would not need to use 'permit no intersection'.. the interface meshes would not intersect anyway... apart from this, I have a feeling that 'permit no intersection' is not an equivalent for a gap pad in fact doesn't work as a separation between the domains..i believe a separate/new domain to model the gap is required with it's corresponding material properties..
Being said all this, when you mention that it all works 100%.. what exactly does that mean? The results you expect (gap pad interface behaving as you expect shown by results) or simply an error free routine?

Any images or related details would evoke precise responses...

cheers..

ghorrocks · July 31, 2015, 20:15

In a static analysis you should not need to use permit no intersection. It means you geometry and mesh are poorly formed. It means that sometimes the mesh will intersect and sometimes it will not. This is not reliable and not recommended.

I would generate a new geometry where the intersection is clear. Rather than putting in a small gap to do this a better way would be to extend one or other of the domains to achieve the intersection.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
problem during mpi in server: expected Scalar, found on line 0 the word 'nan'	muth	OpenFOAM Running, Solving & CFD	3	August 27, 2018 05:18
problem for parallel processing	minaret	OpenFOAM Running, Solving & CFD	14	January 20, 2015 00:41
problem about running parallel on cluster	killsecond	OpenFOAM Running, Solving & CFD	3	July 23, 2014 22:13
Problem with parallel computation (case inviscid onera M6)	Combas	SU2	11	January 30, 2014 02:20
problem in the CFX12.1 parallel computation	BalanceChen	ANSYS	2	July 7, 2011 11:26

July 31, 2015, 18:26		#2
fresty Senior Member Join Date: Aug 2014 Location: UK Posts: 213 Rep Power: 13	I may be wrong at this and would look forward for reassurance from the more experienced chaps but IMO if your interface is static as mentioned in the error message, you would not need to use 'permit no intersection'.. the interface meshes would not intersect anyway... apart from this, I have a feeling that 'permit no intersection' is not an equivalent for a gap pad in fact doesn't work as a separation between the domains..i believe a separate/new domain to model the gap is required with it's corresponding material properties.. Being said all this, when you mention that it all works 100%.. what exactly does that mean? The results you expect (gap pad interface behaving as you expect shown by results) or simply an error free routine? Any images or related details would evoke precise responses... cheers..

July 31, 2015, 20:15		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,872 Rep Power: 144	In a static analysis you should not need to use permit no intersection. It means you geometry and mesh are poorly formed. It means that sometimes the mesh will intersect and sometimes it will not. This is not reliable and not recommended. I would generate a new geometry where the intersection is clear. Rather than putting in a small gap to do this a better way would be to extend one or other of the domains to achieve the intersection.