[Catthan]

Hi fellow ansys users.

I am trying to Hexa mesh a stirling engine burner with ICEM 12.0.
Initially i tried tetra but i had some problems which i didn't manage to solve.

The model is quite complex and, being a beginner, i get lost with all those edges.. Followin the pipe - blade tutorial, i created a blocking for a solid part and then added the blocks to that part so the fluid mesh will not develop through it.
I would like to ask if it is possible to block all parts individually, add the blocks to those solid parts, and then load them all together.
The fluid volume comprises of the gaps between the solid parts.

What I have tried i so far is:

create a 3d bounding box around a solid feature, capture the geomerty with blocks, add those blocks to the associated part, save blocking, and load that blocking together with the blocking i saved for another solid part.
I select to merge the 2 blockings and a new feature is added under ''topology'' of the blocking display tree.
The resulting mesh is however distorted,,


is there a way to do that correctly??
unfortunately i am not allowed to display the model.


Kind regards,
Thanos

[PSYMN]

Yes, you can merge topologies, but the verts must have the same topology.

In other words, if the end of one pipe has two "i" splits and an ogrid, the other part you merge it with must have the same. If the other part only has the Ogrid and one "i" split, then you must split "i" again to get the same topology and then you can merge.

The merge topologies option (right click on topology in the model tree) does a merge verts by tolerance, but you can also merge the verts interactively if they were just too far out of alignment.

Generally speaking, merging topologies causes havoc with the index control and makes other commands, such as Ogrid or scan planes difficult. So make sure your sub models are pretty much done before you merge.

Also, I would recommend this approach for something that you connect end to end (like a piping assembly), not for something where you are putting one blocking inside another or to join fluid and solid regions. The merge must propagate edge distributions, etc. and it just becomes too hard to manage in some situations.

====================

For your example, I recommend a single blocking file and a top down approach. your model is essentially a CAN with 4 tubes inside (holes, tubes, etc are blocked the same way). I would block it by starting with a box that covered most of it, then create an OGrid with faces on the flat ends. You would split that Ogrid to give you the solid and fluid regions. You would then split the HGrid block in the middle into 4 blocks that run the length of the model. Within each of those blocks, you put an OGrid. Then you split this Ogrid such that the "HGrid" center is in the middle of the hole with space around it for fluid flow thru the OGrid. You would split the OGrid again to form the outer edges of the 4 circles, etc.

I understand that you can't send out your original model, but if you make a topologically similar one with similar challenges (or maybe even just adjust the parametric model), I could help by showing how to block that... I am guessing that it would be worth your effort.

As for too much screen clutter, yes, this is a common problem for newish users. Although experts have developed some skills at keeping the model moving to understand what is going on, the main thing experts do is cut down on screen clutter as much as possible. If you don't need to be looking at surfaces, turn them off. If you are looking at one side of a model, use the index control to reduce what you see down to just what you need to see. Index control and the display tree are your friends. use them to keep things as simple on the screen as possible. It also helps to do things int he right order. For instance, if you associate, move and align things into place as you go, then future splits and Ogrids will be created int he right place and not need to be moved (this just comes with experience). New users often insert ogrids, etc. before earlier steps have been tidied up and then have twice as many verts to move into place, etc.

[Catthan]

Thank you so much PSYMN.

I ve been trying several possibilities with merging since i posted the thread and i think it won't give me what i want after all..
I think I can manage to edit a single blocking up to a certain point for my model. I am very happy because with your instructions i will end up with a better result than i would on my own.

In the mean time I will ask my supervisor if i can at least show you the 2 most complicated features and how ''illustrative'' may these be, and maybe you can suggest smth, if this is ok with you.

Again, thank you very much for your suggestions and willingness to help.

All the best,
Thanos