[lnk]

Hi,

Here is my problem. The Y-block doesn't work well. May I ask how to deal with this geometry?

Best regards and many thanks,
lnk

[energy382]

you have to extend split or to modify your edge params like linked bunching

[lnk]

Originally Posted by energy382

you have to extend split or to modify your edge params like linked bunching

Hi,

Thank you for your answer. I really don't understand which split should I extend, to which place, and how could it solve my problem?

Best regards,
lnk

[Far]

why you need Y block?

[lnk]

Originally Posted by Far

why you need Y block?

Because there is a triangle over there. So I'm thinking about using Y-block to solve that problem. If there's any other way to solve this problem, I'd also like to know.

Best regards,
lnk

[Far]

Here are two methods. First method uses files you just posted with extended split and then another blocking without y-block. See the difference.

[PSYMN]

I had downloaded this earlier and just came to post my answer. I see that Far already beat me to it with the correct answer.

Actually your first blocking is fine. All I did was update the mesh sizes (Blocking (tab) => Pre-Mesh Params => Update Sizes => Update All) and recomputed the premesh to get this...

Wedge_Fine.jpg

You could go with the simpler blocking as Far showed, but there are situations where that quarter Ogrid blocking would be better... For instance, if you had an inner baffle that went part way up the side (between the Ogrid and the side block) or if you had a lot of flow gradients in that region which needed increased resolution (independent of the rest of the model)...

It it is really one volume with no special mesh focus needed, then just stretching the Ogrid to fit the shape is probably easiest and best.

Hexa is very flexible, keep at it and you will do well. Just remember that the blocking topology doesn't need to match the surface topology.