[ANSYS Meshing] Are there any drawbacks to a hybrid mesh?

JuPa · April 11, 2014, 10:35

I've got quite a complicated geometry to mesh, using Ansys Meshing. I've decomposed my geometry into simple parts (in Design Modeller).

Because the geometry is very complicated, I really don't believe that I'd achieve a good quality hexahedral mesh.

However what I can try to achieve is a good quality hybrid mesh. My plan is to:
1. Use tetrahedral elements where the geometry is very complicated
2. Use hexahedral elements where the geometry is simple, sweepable and can be resolved using the multizone method
3. Use prism inflation layers at some boundaries to capture flow close to the surface.

This type of mesh will leave me with hex, tet and prism elements. Except from the increased element count due to tet and prism elements, is there anything wrong with this kind of mesh?

The solver that will be used will be Ansys CFX v15. The application is strong buoyancy driven flow, heat transfer and phase change, using the volume of fluid method.

I've had a look at the theory guide for the models I will be using in CFX. It doesn't say anything in the theory guide about mesh restrictions (i.e. it doesn't say I must use a hex mesh for a particular model I'm using).

I will remember to keep good mesh statistics. I.e. mesh quality, orthogonality, skewness, aspect ratio.

Is there anything I should be aware of before I commit to a hybrid mesh? Any information you can provide will be helpful I'm sure.

Thank you.

Far · April 12, 2014, 12:44

Eventually hexa mesh and hybrid should give same results if you have achieved mesh independence.

siw · April 12, 2014, 12:55

One advantage of a hexa mesh over a tetra/prism mesh is the quantity of cells for a given quantity of nodes. A hexa mesh will have about the same amount of cells to nodes. Whereas, a tetra/prism will have a greater amount of cells than nodes. So if you make a mesh for a given node count the tetra/prism will have more cells, so it'll take longer to solve and require more memory/storage.

JuPa · April 13, 2014, 07:51

Quote:

Originally Posted by Far

Eventually hexa mesh and hybrid should give same results if you have achieved mesh independence.

Thanks Far. Do you have any examples from past experiences about your experiences regarding hybrid meshes?

Quote:

Originally Posted by siw

One advantage of a hexa mesh over a tetra/prism mesh is the quantity of cells for a given quantity of nodes. A hexa mesh will have about the same amount of cells to nodes. Whereas, a tetra/prism will have a greater amount of cells than nodes. So if you make a mesh for a given node count the tetra/prism will have more cells, so it'll take longer to solve and require more memory/storage.

Absolutely. I even mentioned in my opening post that I fully accept tet/prism cells will increase the amount of cells in the mesh. I accept this (and implicitly accept a longer running time).

I just want to know if there are any other negative connotations to having a hybrid mesh.

I don't believe having a structured well blocked hexa mesh is possible for every CFD geometry.

LiamH · April 22, 2014, 10:33

I am familiar with that strategy and I find it the best. The rules I have picked up from personal experience as well as from ANSYS Instructors was:

Aspect Ratio < 100
Max Skewness < 0.98
Min Orthogonal Quality > 0.02

Those are rules of thumbs but you may want to be careful with the Skewness most of all. That is what will cause you grief in your simulations. The lower the max, the better.

April 11, 2014, 10:35	Are there any drawbacks to a hybrid mesh?	#1
JuPa Senior Member Mr CFD Join Date: Jun 2012 Location: Britain Posts: 361 Rep Power: 14	I've got quite a complicated geometry to mesh, using Ansys Meshing. I've decomposed my geometry into simple parts (in Design Modeller). Because the geometry is very complicated, I really don't believe that I'd achieve a good quality hexahedral mesh. However what I can try to achieve is a good quality hybrid mesh. My plan is to: 1. Use tetrahedral elements where the geometry is very complicated 2. Use hexahedral elements where the geometry is simple, sweepable and can be resolved using the multizone method 3. Use prism inflation layers at some boundaries to capture flow close to the surface. This type of mesh will leave me with hex, tet and prism elements. Except from the increased element count due to tet and prism elements, is there anything wrong with this kind of mesh? The solver that will be used will be Ansys CFX v15. The application is strong buoyancy driven flow, heat transfer and phase change, using the volume of fluid method. I've had a look at the theory guide for the models I will be using in CFX. It doesn't say anything in the theory guide about mesh restrictions (i.e. it doesn't say I must use a hex mesh for a particular model I'm using). I will remember to keep good mesh statistics. I.e. mesh quality, orthogonality, skewness, aspect ratio. Is there anything I should be aware of before I commit to a hybrid mesh? Any information you can provide will be helpful I'm sure. Thank you.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
how to set periodic boundary conditions	Ganesh	FLUENT	15	November 18, 2020 06:09
[ICEM] Hybrid mesh for 2D boundary layer	Bigio	ANSYS Meshing & Geometry	33	November 18, 2019 09:15
Gambit problems	Althea	FLUENT	22	January 4, 2017 03:19
Mesh motion with Translation & Rotation	Doginal	CFX	2	January 12, 2014 06:21
Icemcfd 11: Loss of mesh from surface mesh option?	Joe	CFX	2	March 26, 2007 18:10

April 12, 2014, 12:44		#2
Far Super Moderator Sijal Join Date: Mar 2009 Location: Islamabad Posts: 4,553 Blog Entries: 6 Rep Power: 54	Eventually hexa mesh and hybrid should give same results if you have achieved mesh independence.

April 12, 2014, 12:55		#3
siw Senior Member Stuart Join Date: Jul 2009 Location: Portsmouth, England Posts: 733 Rep Power: 25	One advantage of a hexa mesh over a tetra/prism mesh is the quantity of cells for a given quantity of nodes. A hexa mesh will have about the same amount of cells to nodes. Whereas, a tetra/prism will have a greater amount of cells than nodes. So if you make a mesh for a given node count the tetra/prism will have more cells, so it'll take longer to solve and require more memory/storage.

April 22, 2014, 10:33		#5
LiamH New Member Liam Join Date: Sep 2013 Location: Calgary, Alberta Posts: 19 Rep Power: 12	I am familiar with that strategy and I find it the best. The rules I have picked up from personal experience as well as from ANSYS Instructors was: Aspect Ratio < 100 Max Skewness < 0.98 Min Orthogonal Quality > 0.02 Those are rules of thumbs but you may want to be careful with the Skewness most of all. That is what will cause you grief in your simulations. The lower the max, the better.