CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   Uniform mesh in ICEM - again (

harerton June 29, 2011 13:25

[ICEM] Uniform mesh in ICEM - again
Hi everybody,

I again need to create an uniform mesh with hexa elements of specific size. Again the geometry was created with this in mind.

Following the steps in the femur tutorial I could setup the pre mesh parameters easily. The picture below shows part of the geometry with the pre-mesh showing. This is exactly what I was expecting to get.

Then after computing the new mesh using BFCart I get the following result (see picture below) where the elements are smaller and consequently not aligned with the pre-mesh (see the edge bunching along the edges).

Any suggestion?



harerton July 4, 2011 14:20

Anyone? thanks!

PSYMN July 5, 2011 09:21


As part of the algorithm, BFCart actually ends up working with its inverse mesh (this is how it steps back from the faces so it can insert a boundary layer).

You can use the Method option to get what you want.

When you select the Cartesian file to start from, you need to specify if that Cartesian file is the initial or final location of the splits...

My guess is you picked initial, so the inversion happens afterward. But Pick Final (Set "Enforce a Split" "Method" to "Final") and those will be the final split locations... (actually, behind the scenes, it just inverts it first, then the algorithm inverts it back again, which is not exactly the same if your sizes are changing, but it should be the same for a uniform distribution like you showed.)

harerton July 5, 2011 13:04


Thanks for answering again. I think I understood the process. So I need to generate a cartesian file after blocking. But when I go to file > blocking > write cartesian grid I get something that resemble an error dialog, but without any info. Am I missing something? (see picture below)


harerton July 5, 2011 17:15

I tried all the steps using a simple cube geometry and didn't get the above error message. Could this indicate a problem with my original geometry?

Edit: Anyway, tried again with original geometry but using a different filename (without any spaces) and it worked.

Thank you!

PSYMN July 6, 2011 15:24

I submitted a defect report on your behalf...

Best regards,


All times are GMT -4. The time now is 17:16.