CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   [ICEM] triangle 86218 297 88224 seems to be in the wrong family (

Sushilkumar November 4, 2011 01:25

triangle 86218 297 88224 seems to be in the wrong family
What this error indicates..?

I want to create the prism layers, but it is not generating. The log file shows this message frequently.

PSYMN November 4, 2011 22:10

The shell elements inherit the part name of the surface they are generated on. This is actually done by ray tracing from the centroid of the shell in a normal direction. Sometimes this hits the wrong part and you can find some shells (usually near the edge of a surface) that are in a different part. Sometimes I will get this if I have overlapping surfaces in different parts; some shells will be in one part and adjacent interspersed shells will be in another depending on which surface the ray hit first.

When Prism runs, it extrudes shells based on their Part name. If you pick a part that has various other shells scattered in it that are not being extruded, this is a lot harder on the prism algorithm. In some cases, the part your extruding is the one that is scattered (or perhaps there is just a single shell in that part surounded by shells in an un-prismed part). The prismed shells try to form columns of prisms while trying to connect to the surrounding parts... Often it can handle it and you just have some quality issues (pyramid elements fit to the side of the column, particularly the thinnest base elements, are not usually good quality), other times the quality is so bad that the process fails and you get this message.

There is a check for these isolated shells (I forget the exact name of it but it is mentioned in the prism prep section of the help), or you could do a quick visual inspection. Either way, if you find any shells in the wrong part, right click on the part they should be in (in the model tree) and add those to the correct part... then run prism.

Best regards,


Sushilkumar November 8, 2011 11:52

Thanks Simon.

All times are GMT -4. The time now is 07:38.