[Daniel_Khazaei]

Hello to all

I am trying to mesh a bifurcation geometry. I have managed to generate an unstructured mesh using Tetra/Mixed method.
However I am wondering if it possible to get a mesh like this one in ICEM using usual blocking?




Thanks in advance

D. Khazaei

[kad]

Check out Simon's Tipps and Tricks documentation. It has a detailed description for blocking a very similar geometry (fork).

[Daniel_Khazaei]

Originally Posted by kad

Check out Simon's Tipps and Tricks documentation. It has a detailed description for blocking a very similar geometry (fork).

Would please provide the link?

Best wishes

[FJSJ]

Hi Daniel,

I wondered the same thing few months ago, and as Kad said, simon'tips has two approaches.

Here the link (see 85 & 86 slides)

Simon's tip

[Daniel_Khazaei]

Originally Posted by FJSJ

Hi Daniel,

I wondered the same thing few months ago, and as Kad said, simon'tips has two approaches.

Here the link (see 85 & 86 slides)

Simon's tip

Thank you both for your suggestions.

I will read this PDF and come back.

[Daniel_Khazaei]

I don't think that blocking produce high quality mesh for circular tube at bifurcation region.
This is how I want the mesh look like:

[Daniel_Khazaei]

I have tried to do the blocking as Simon suggested in his tips and tricks.

However, I don't know how to fix it:

[FJSJ]

Hi Daniel,

I´ve tried with a geometry similar to yours. The problem was the same, just the junction surface. Please, can you share your geom?

[Daniel_Khazaei]

Originally Posted by FJSJ

Hi Daniel,

I´ve tried with a geometry similar to yours. The problem was the same, just the junction surface. Please, can you share your geom?

Of course...

Were you able to fix your problem?

Best wishes

[kad]

Originally Posted by FJSJ

Please, can you share your geom (and blocking)?

Looks like a problem with associations. Check vertex and edge association. If these are correct associate faces manually to their corresponding surfaces.

[Daniel_Khazaei]

Originally Posted by kad

Looks like a problem with associations. Check vertex and edge association. If these are correct associate faces manually to their corresponding surfaces.

I have tried to manually associate faces to corresponding surfaces, but the problem is still there.
However, about vertex and edge association:

It is a CAD geometry and there are to many curves there.

Do you think that the sharp angle in my blocking strategy is responsible for this problem?

[kad]

The sharp angle could be a minor problem, but in general blocking should be doable. Your blocking looks not 100% okay in your last attachment. It looks like there are degenerated blocks near the sharp angle. But this could also be a matter of perspective.

Why don't you share your project? Would be much easier.

[Daniel_Khazaei]

Originally Posted by kad

The sharp angle could be a minor problem, but in general blocking should be doable. Your blocking looks not 100% okay in your last attachment. It looks like there are degenerated blocks near the sharp angle. But this could also be a matter of perspective.

Why don't you share your project? Would be much easier.

I am trying to attach it here but the size is larger than the limit.
Now I am uploading the project on a another host.

[Daniel_Khazaei]

Originally Posted by kad

The sharp angle could be a minor problem, but in general blocking should be doable. Your blocking looks not 100% okay in your last attachment. It looks like there are degenerated blocks near the sharp angle. But this could also be a matter of perspective.

Why don't you share your project? Would be much easier.

here is the full project:

http://www.rodfile.com/r86znlonj5zq

[FJSJ]

What is the password?

[kad]

After having a look at your project I must admit that our suggested blocking strategy is probably wrong. You can do it, but you will end up with some bad elements and run into trouble with ogrid creation (see att which is without ogrids).

Meanwhile I think I have understood the strategy in your posted pics. I will explain this later/tomorrow as I am going home now.

[Daniel_Khazaei]

Originally Posted by kad

After having a look at your project I must admit that our suggested blocking strategy is probably wrong. You can do it, but you will end up with some bad elements and run into trouble with ogrid creation (see att which is without ogrids).

Meanwhile I think I have understood the strategy in your posted pics. I will explain this later/tomorrow as I am going home now.

Thank you very much.
I am eagerly waiting to hear from you.

[kad]

I have attached the generic blocking structure for your geometry. Now you have to apply to your geometry. If possible I would strongly suggest a geometry clean up in CAD. I tried it myself in Designmodeler. It somehow worked but destroyed the surface structure.

If blocking is to hard you can also try ANSYS Workbench CutCell meshing. I just started to work with it this week and it gave decent results in a short time. It is really not hard to handle if you are familiar with other meshing tools.

[kad]

Blocking attached. The .blk* (e.g. .blk3) show the progress of work.

[Daniel_Khazaei]

Originally Posted by kad

I have attached the generic blocking structure for your geometry. Now you have to apply to your geometry. If possible I would strongly suggest a geometry clean up in CAD. I tried it myself in Designmodeler. It somehow worked but destroyed the surface structure.

If blocking is to hard you can also try ANSYS Workbench CutCell meshing. I just started to work with it this week and it gave decent results in a short time. It is really not hard to handle if you are familiar with other meshing tools.

Originally Posted by kad

Blocking attached. The .blk* (e.g. .blk3) show the progress of work.

Thank you very much for the help man.

I will try to work this out.