Jumps In hybrid Mesh

realanony87 · May 6, 2009, 22:53

Hello Guys,
I have been testing out the simple TUBE hybrid hexa,tetra hybrid meshing tutorial in ICEM CFD.
I do a quick run in ansys CFX , and at the interfaces there appears to be a pressure jump at each mesh interface. The velocities seem fine though (although the boundaries are better with the hexa mesh)
Does anyone have experience with hybrid meshes? are they safe to use for more complex cases ?

ghorrocks · May 7, 2009, 20:00

Hi,

Hybrid meshes are fine. You can get some apparent jumps at interfaces when the mesh is not fine enough. Also I assume you have all the pitch ratio stuff turned off - that can cause these effects.

Glenn Horrocks

realanony87 · May 7, 2009, 22:53

Do you mean pitch ratio of the hexa mesh at the interface ? So the ratio there should be 1 ?

realanony87 · May 8, 2009, 00:02

Do you mean pitch ratio of the hexa mesh at the interface ? So the ratio there should be 1 ?

Roland R · May 8, 2009, 05:35

Hi,

I think that if there is pressure jump at interface than hexa and tetra mesh don’t connect exactly at the interface. The nodes of interface 1 have to fit to the nodes at the interface 2. (1:1 connect) If the nodes don’t fit to each other than there is an interpolation at the interface, this can cause some inexactness. (The size of the inexactness = How many cells do connect to one cell at the interfaces?).
Roland

ghorrocks · May 10, 2009, 22:43

Hi,

Pitch ratio is used in turbo machinery so blade rows with different numbers of blades can be modelled using periodic boundaries. That is a blade row with 100 blades can be connected to a row with 149 blades, where you are only modelling a single blade of each row. You use the pitch ratio to scale the flow across the interface to keep the relationship between the rows correct.

If you have pressure jumps at an interface then something is wrong. Is your solution properly converged? Is you mesh quality adequate? Have you got mesh size jumps at the interface?

Glenn Horrocks

May 6, 2009, 22:53	Jumps In hybrid Mesh	#1
realanony87 New Member Join Date: May 2009 Posts: 21 Rep Power: 17	Hello Guys, I have been testing out the simple TUBE hybrid hexa,tetra hybrid meshing tutorial in ICEM CFD. I do a quick run in ansys CFX , and at the interfaces there appears to be a pressure jump at each mesh interface. The velocities seem fine though (although the boundaries are better with the hexa mesh) Does anyone have experience with hybrid meshes? are they safe to use for more complex cases ?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
zonal hybrid mesh	Maolin Tu	Main CFD Forum	5	May 16, 2020 13:19
Gambit problems	Althea	FLUENT	22	January 4, 2017 04:19
[Commercial meshers] fluentMeshToFoam with Hybrid Mesh	-mAx-	OpenFOAM Meshing & Mesh Conversion	2	May 7, 2009 02:43
Hybrid Mesh	Marc	FLUENT	4	July 18, 2007 21:38
hybrid mesh around a stator	Michele Cagna	FLUENT	2	October 25, 2000 10:12

May 7, 2009, 20:00		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Hi, Hybrid meshes are fine. You can get some apparent jumps at interfaces when the mesh is not fine enough. Also I assume you have all the pitch ratio stuff turned off - that can cause these effects. Glenn Horrocks

May 7, 2009, 22:53		#3
realanony87 New Member Join Date: May 2009 Posts: 21 Rep Power: 17	Do you mean pitch ratio of the hexa mesh at the interface ? So the ratio there should be 1 ?

May 8, 2009, 00:02		#4
realanony87 New Member Join Date: May 2009 Posts: 21 Rep Power: 17	Do you mean pitch ratio of the hexa mesh at the interface ? So the ratio there should be 1 ?

May 8, 2009, 05:35		#5
Roland R Senior Member Roland Rakos Join Date: Mar 2009 Posts: 131 Rep Power: 17	Hi, I think that if there is pressure jump at interface than hexa and tetra mesh don’t connect exactly at the interface. The nodes of interface 1 have to fit to the nodes at the interface 2. (1:1 connect) If the nodes don’t fit to each other than there is an interpolation at the interface, this can cause some inexactness. (The size of the inexactness = How many cells do connect to one cell at the interfaces?). Roland

May 10, 2009, 22:43		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Hi, Pitch ratio is used in turbo machinery so blade rows with different numbers of blades can be modelled using periodic boundaries. That is a blade row with 100 blades can be connected to a row with 149 blades, where you are only modelling a single blade of each row. You use the pitch ratio to scale the flow across the interface to keep the relationship between the rows correct. If you have pressure jumps at an interface then something is wrong. Is your solution properly converged? Is you mesh quality adequate? Have you got mesh size jumps at the interface? Glenn Horrocks