Edge parameters issue
 

Hi there!

I am trying to generate a mesh for a swirler configuration with ICEM 12.0.1. I have created a block structure and the only thing missing right now is the specification of the edge parameters. Unfortunately I have been encountering some issues at this final step.

The mesh consists of 4 periodic sections, where each of them consists of a large block that splits into several smaller ones at the face in radial direction. As soon as I try to specify the # of nodes for one edge (lets say RED in the attachment), the number of nodes in either BLUE or GREEN is changed as well. If I then try to change back e.g. BLUE, RED is set to a different # of nodes again.

I assume that obviously the sum of the nodes of all the radial edges has to match the number of nodes of the YELLOW edge (for each quarter). Is there a way to remove the dependency of the radial edges on YELLOW? In other words, how can I "freeze" my node distribution at the radial edges, and ask for YELLOW to adapt to it?

I would be very thankful for any kind of suggestion!

Regards,
Matthias

It does get tricky...

You could get into "Linked Bunching", look it up in the help. Basically, it lets you link the bunching of a longer edge to match the distribution (proportionally) of the series of parallel smaller edges.

But new users find it tricky some times, so perhaps an easier method would be to simply extend the splits. Once you have the splits, you can control the distribution more directly.

If you do neither, you could eventually get what you want by going over those three shorter edges and resetting the sizes (without touching the yellow edge) and it should figure it out. There is some order to it, even if it doesn't appear that way at first ;^)

Best regards,

Simon

mesh a swirler
 

I am trying to mesh a swirler using ICEM CFD as well. I tried tetra/prism first, which seemed out of control sometime. Should hexa be better? There are 2 radial swirlers and each with 16 vanes. I can be very hectic using hexa. I am new with ICEM and waiting for your suggestions. Thanks.