CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   [ICEM] Prisms Intersection (

sandeep_tu March 18, 2013 10:51

Prisms Intersection
1 Attachment(s)
Dear All,

I have got a problem while generating prisms (intersecting prisms), please see attached figure.
How can I overcome this problem (like Shrinkage or ignoring prisms there)? Any suggestions or tips could be helpful for me.

Thank you

diamondx March 18, 2013 11:01

you need to use "define thin cut" and select the two surfaces intersects. that's a very sharp angle there, not sure if that feature can help tough...

sandeep_tu March 19, 2013 04:19

1 Attachment(s)
I tried with specifing "thin cut" surfaces unfortunately it didnīt help.

Any suggestions how to apply shrinkage or ignoring prisms there (Ex: see attachment)


Far March 19, 2013 09:43

Try Tgrid. There is more control on prisms behaviour

kad March 19, 2013 10:54

You could try one of the following steps:
- set prism height "0" for floating height in problematic regions
- make use of the auto-reduction function which can be found under advanced prism settings.
- I would rather go back to geometry and just close the critical gaps if your geometry does not have to be 100% accurate

diamondx March 19, 2013 11:17

use simon's tips and trick to find a solution for your problem. worst come to worst, you will need to evaluate the importance of that region (remove the sharp angle if you don't need information there):

PSYMN March 19, 2013 22:09

I was asked by two people to come and comment here, but I think the right things have already been said.

1) refining the mesh size, setting the initial height to 0 (float) and turning on options like "auto reduction" should help.

2) If that fails, then altering the geometry so you don't have such a tight space will help. I have worked on actual industry examples and they usually place a box behind the body (intersecting the body and the seat). In real life the driver wouldn't be leaning forward from the seat anyway, closing off much of that gap will not hurt your accuracy, but it will make the meshing/solving easier.

3) If all else fails and you are interested in trying TGrid Meshing, TGrid Tetra really has advanced over the past few years and its "auto reduction" and handling of situations like this is really superior... Of course, it has its own learning curve, so it may not be for everyone.

All times are GMT -4. The time now is 03:09.