[FJSJ]

Hi,

I would like to know if "merge mesh" works with hex/hex mesh or at least one of this must be tetra/tri mesh?

[Far]

one of them must be tri. Hexa- hexa should work well if you have one to one interface I guess.

[BrolY]

If you want to merge Hex mesh with another Hex mesh, you need a conformal interface between the 2 meshes (i.e. the exact same interface with the exact same node repartition), and then use the "merge nodes" option rather than "merge meshes".

[FJSJ]

I´ve tried to do something similar. I have my entire domain split into 3 parts. First, I´ve done the first blocking and first mesh. I´ve saved blocking and mesh. Second, I´ve done the second blocking (merging blocks) and the second mesh and I get this. So..I think something wrong because the second mesh is (only a part) like the color of first one. Do you know why is it?

[diamondx]

FJSJ, sorry I'm not sure i understood what you said. please can you re-formulate.

[FJSJ]

Sure. I have the entire domain divided into 3 parts. First, I do the blocking and mesh in first part. I save the blocking and mesh then close the block and mesh. Then I do the same with the second part of the domain.
But as you can see, in the first part in "Interface 1", no problem. But, on the second mesh, "Interface 1" is a sector only. The rest is associated to another part, instead of associating the "Interface 1" (as in the mesh 1)... I´m not sure if I explain well.
I´ve tried to follow similar approach like "Merged Tetra-Hexa Mesh in a Hybrid Tube" tutorial. Maybe the problem is the curves association...

[diamondx]

i have more questions: What do you mean by sector only ?
Why did you separate the block in three part ?
Did you try merge blocks ?
Can you share your project using the public dropbox file ?

[FJSJ]

I separated the blocks because is easier to do the block on transition part of my blade. For this reason I´ve not tryed merge blocks. I´ve tryed to do something similar to BrolY told me. Yes, I´m going to share the project but right now dropbox isn´t work well.

[FJSJ]

at first, you can view the premesh in part 1. Then, you can open the block "bloque 2" and merge it.
Then view the premesh in this second block.

https://www.dropbox.com/s/yofo3a4vwe...MergeMesh1.rar

[diamondx]

Are you aware that you have overlapped surfaces... for one block it may be part of interface_1 and for the other one it may be part of FLOW_DOMAIN_TRANSITION



you can fix that, by doing a doing a build topology in the repair geometry...

[diamondx]

take a look at my blocking without splits:




Project file is here:

https://dl.dropboxusercontent.com/u/...MergeMesh1.zip

[FJSJ]

Thank Diamondx, you´re right. I have overlapped surface. I´ve delete them and now I think it´s ok.

Understood. It is better not split. I guess you get with 2 O-grid through the ends and then with a horizontally split? It is?

[diamondx]

ok; i could make another blocking similar to the one in youtube (by simon). minimum quality is 0.4, it can be improved more



link for the project is here:

https://dl.dropboxusercontent.com/u/...MergeMesh1.rar

[Far]

How you will manage sharp trailing edge and circular pipe beneath blade

[FJSJ]

That is the problem area. Because for a cylindric section O-grid is the best. However, for the other part of the blade not .. so ... how to make a block support the other... I mean how to pass from one topology to other one

[Far]

Solution would be to give a finite thickness to trailing edge, if it does not cause in- accuracy...

[FJSJ]

But I need a sharp trailing edge so... I think here is the problem

But... I think Diamondx managed it!

[Far]

Yes. Strategy by diamondx is good.

[FJSJ]

Thank you dimaondx!
You get a good quality in all blade. But why do you create top and down part?
I supposed that is because this part is hard to adjust but i´m not sure...
Thanks again!

[Far]

can you share bigger picture of your project. I mean blade with hub and farfield volume...