CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   [ICEM] How to create J-type grid for turbine blade (

Far November 8, 2011 13:08

How to create J-type grid for turbine blade
2 Attachment(s)
Hi every one
Just started to learn ICEM and completed all hexa tut. Now trying meshing on high curvature blade provided by skopa

Can any suggest me how to make the J-type or L-type topology at leading and trailing edge instead of simple H-type grid.

Far November 9, 2011 03:49

1 Attachment(s)
To create the blocking (J-type) in attached Figure, Do I need to use the collapse command? I use many schemes to create such blocking but I am not getting how to create blocks with common vertex?

Far November 9, 2011 08:35

1 Attachment(s)
refined quality further, still not getting to make J-type. Any help shall be highly appreciated.

Far November 9, 2011 09:36

1 Attachment(s)
Final grid

Far November 10, 2011 02:56

I guess it is not possible in ICEM, on the other hand GridPro provides this feature and it is way easy to manipulate the topology in GridPro than in ICEM CFD. I hope ICEM shall be up with these features soon !!!

PSYMN November 10, 2011 13:26

Sure it can be done in ICEM CFD, do a search on CFD online for "shifted periodic", I have talked about this a few times...

Basically, you use a quarter Ogrid on opposite corners to shift the periodic by one index. You can create this lots of ways, including a split block and merge vertex operation...

Far November 10, 2011 13:56

Can you please illustrate it with some pics on available geometry?

Far November 10, 2011 14:20

surely got your idea, now I think I can make the J-type topology. But I am not getting what does it mean by shifted periodic. There is no tutorial on this, and where to find this function in ICEM CFD

PSYMN November 10, 2011 14:26

Sorry, I don't have time to show it right now (I have shown steps before and you can find those)...

Basically, you need the sides to be periodic, and if you continue these across the model, you will get high skewness... SO we add one quarter ogrid on one side and another opposite one on the other side so we can shift the index between the two sides... In other words, if you follow the edge from one periodic side, it doesn't go right to its periodic twin. Instead, it is shifted... But in such a way that the whole model is still periodic...

All times are GMT -4. The time now is 03:46.