assembly with more than 2 parts

Hi

I am new at using code_aster. I can do a calculation with 2 parts, but I can not do it with more than 2. I am using fuse the 2 solids together. Can anyone help me with making a calulation a assembly with more than 3 og 4 parts.

Ib Khater
ik@ik-e.dk

I need also to solve this problem so I created a tee-shaped block on 2 rubber supports, each support having a different property. The shape was defined using a python script t020.py which creates a shape called partition1.

I worked in the following stages:

1. Exploded partition1 into solids 1, 2 and 3. Renamed Solid_3 as TeeBlock.
2. Created groups of faces TopLoad (from TeeBlock), BaseFix1 (from Solid_1) and BaseFix2 (from Solid_2). Saved the file as t020a.hdf.
3. Meshed using auto tetrahedralisation; set 1D parameters average length 25 and quadratic; set 3D parameter max volume as 10000. Computed mesh.
4. Created groups on volumes Solid_1 as Part_1, Solid_2 as Part_2 and TeeBlock as Part_3.
5. Opened Aster module and used wizard. Saved .comm file.
6. Edited .comm file to define the three materials and allocate them to the three parts. Opened Eficas to check the file - OK.
7. Returned to Salome-Mech and ran the solver.

I thought all was going well but it ended with a copy error. The script file, the .hdf file and the .comm file are included in the attached .zip file. Any idea what's wrong?

After writing the previous message I had an idea to change the sequence around, to create the mesh groups Part_1, Part_2 and Part_3 (stage 4) before computing the mesh (end of stage 3).

It worked!!!!!

The result shows an asymmetrical deflection of the two rubber blocks which are the same size but different E values.

The key seems to be in the geompy.MakePartition line where two box shapes are partitioned from the main tee-shaped block.<br /><br />Post edited by: Andy Foan, at: 2008/05/22 19:22

See image attached.

[img]<br /><br />Post edited by: Andy Foan, at: 2008/05/22 19:24

Would it be possible to post the new mesh file as i am experiencing the same problem and can't seem to get around it.

Thanks

Warren