SubModeling

1st Step

Create the whole model and define it jobname, we say COARSE_jobname then perform, as usual the structural analysis holding this file setting. Running the analysis of the whole model It should be advisable:

1. to integrater in it, if possible, a solid model of the "submodel"
2. the model database ( jobname.db ) and the result file ( jobname.rst ) shall be saved in the working directory where will be run the submodel analysis

2nd Step

The sub model analysis shall have a different jobname, i.e. FINER_jobname, then

1. create the submodel using the same global reference system used for the coarse model;
2. the submodel shall be "a part" of the coarse model and for this reason should be useful to clear the mesh of the coarse model and to start from its "solid model" deleting all volumes/areas but the part to be analysed
3. create, obviously, a finer mesh of the part of the model to be analysed
4. selecting (make them the only active nodes usig for example NSEL command) only the nodes on the boundary of the submodel that has to take the displacement condition from the coarse model write them to a file using the ANSYS command NWRITE
5. save the database containing the finer model in the FINER_jobname.db file

3rd Step

Now the task is to determine, from the coarse solution primary unknowns, the boundary condition to be applied on the previously selected elements of the FINER model.

The list and the coordinates of the node are written in the file created by the NWRITE command (see the previous step); the data necessary for the interpolation are present in the two files, previously saved:

• COARSE_jobname.db
• COARSE_jobname.rst

The result of the interpolation is a file with the imposed displacements (for a structural analysis) of the boundary nodes of the finer model.

The command to perform the task is CBDOF that activates cut boundary interpolation (for submodeling). :

CBDOF, Fname1, Ext1, --, Fname2, Ext2, --, KPOS, Clab, KSHS, TOLOUT, TOLHGT, TOLTHK 

the correct sequence is:

1. set the new jobname
2. create the submodel
3. select the boundary node of the finer mesh
4. write them using NWRITE
5. save the model with SAVE
6. load the COARSE_jobname.db without changing the jobname ( "resume from" menu item )
7. enter /POST1 and load result file COARSE_jobname.rst
8. read relevant set of data
9. issue the command CBDOF

-- RobertoBernetti - 24 Feb 2013