Meshing problems

Fededea

Hello,

I am trying to mesh a very complex and big model, with duct with very different dimensions (i.e. an inlet duct is a rectangular inlet of 6x2 cm and another duct is a circular tube with 4 mm of diameter). 

I use a surfaces control, a BLs controls and then I start the volume mesh. I made some test with different size of the mesh and everytime the volume mesh process fails with many warnings and errors such the followings:

• WARNING: 1 duplicted elems excluded. It may cause association problem later.

• **Warning: no BL/fluid volumes meshed, BL contouring skipped.

• topo err: elem882024->side0: wings=3, non-baffle-wingss=3 (no error highlighted) Meshing failed.

• Volume Mesh Failed

But if I don't use a BL controls the mesh process will be fine and the model can start a simulation. How can I solve the problem?

I attach the mesh log.

Thanks a lot in advance.

acupro

It can sometimes be tricky to get appropriate volume/surface/BL meshing parameters when there are large length scale differences.  You need to make sure there is fine enough mesh to fit boundary layers, maybe in small passages.  If there are tight angles inside which the volume mesh resides, that can also inhibit boundary layers - so it's best to try to avoid those in the geometry, if possible, which can involve some simplifications/modifications to the geometry.  As a first attempt, I often use first boundary layer height type as fraction of surface mesh - typically something like 0.2, with 5 layers, and a constant growth rate of 1.2 or 1.3 - in order to get at least a mesh.  And make sure 'modify surface mesh' is active.

acupro

It can sometimes be tricky to get appropriate volume/surface/BL meshing parameters when there are large length scale differences.  You need to make sure there is fine enough mesh to fit boundary layers, maybe in small passages.  If there are tight angles inside which the volume mesh resides, that can also inhibit boundary layers - so it's best to try to avoid those in the geometry, if possible, which can involve some simplifications/modifications to the geometry.  As a first attempt, I often use first boundary layer height type as fraction of surface mesh - typically something like 0.2, with 5 layers, and a constant growth rate of 1.2 or 1.3 - in order to get at least a mesh.  And make sure 'modify surface mesh' is active.

Fededea

Thanks a lot, I solved the problems with some modification of the cad model where there were "strange" geometry. 

Just one more question, regarding the dimension of the volume mesh for the first attempt, will be ok to consider the volume mesh dimension almost the same of the surface mesh?

Thanks

Jagan Adithya Elango

Fededea said:

Thanks a lot, I solved the problems with some modification of the cad model where there were "strange" geometry. 

Just one more question, regarding the dimension of the volume mesh for the first attempt, will be ok to consider the volume mesh dimension almost the same of the surface mesh?

Thanks

It is ok as long as the mesh is good enough to capture flow characteristics, temperature gradients, etc. Just make sure that the mesh count isn't unreasonably high.