QUESTION:

A user had a simple drop test analysis using ANSYS/LS-DYNA (either Rev 9.0 or 10.0). They only intended to test the technical capability that allow to switch parts from deformable to rigid and back to deformable during an analysis.

- 1st step : drop test up to 0.01 sec (new analysis)
- 2nd step : the structure is switched from deformable to rigid from t=0.01 to t=0.02 s (first restart)
- 3rd step : the structure is switched back to deformable (second restart) from t=0.02 to t=0.04 s.

Problem : The time-step at the end of step 2 (first restart) which is 10-5 s (as requested by user) is not recomputed by LS-DYNA during the 3rd step when the structure is switched back to deformable. Its value should normally be in accordance with the Courant condition for deformable solids (roughly 10-6) and not 10-5 s.

It appears that the source of the problem is that the LS-DYNA ".r" restart file that has been automatically generated by the ANSYS interface for the 2rd step contains part deletion commands that have never been requested by the user. If you run the analysis using a corrected ".r" file (i.e. without part deletion commands) it works fine.



ANSWER:

For this particular model, the problem was with three lines of data in drop.cdb.

*SET,_PARTSEL, 1.000000000000
*DIM,_PSEL ,ARRAY, 6, 1, 1,
*SET,_PSEL ( 6, 1, 1), 1.000000000000

These three lines of data are un-selecting parts 1-5.
That's why *DELETE_PARTS are written in file.r.

By including PARTSEL,'ALL' command in drop.run, the third step (after the rigid to deformable switch) works as documented.


QUESTION:

A user had a simple drop test analysis using ANSYS/LS-DYNA (either Rev 9.0 or 10.0). They only intended to test the technical capability that allow to switch parts from deformable to rigid and back to deformable during an analysis.

- 1st step : drop test up to 0.01 sec (new analysis)
- 2nd step : the structure is switched from deformable to rigid from t=0.01 to t=0.02 s (first restart)
- 3rd step : the structure is switched back to deformable (second restart) from t=0.02 to t=0.04 s.

Problem : The time-step at the end of step 2 (first restart) which is 10-5 s (as requested by user) is not recomputed by LS-DYNA during the 3rd step when the structure is switched back to deformable. Its value should normally be in accordance with the Courant condition for deformable solids (roughly 10-6) and not 10-5 s.

It appears that the source of the problem is that the LS-DYNA ".r" restart file that has been automatically generated by the ANSYS interface for the 2rd step contains part deletion commands that have never been requested by the user. If you run the analysis using a corrected ".r" file (i.e. without part deletion commands) it works fine.



ANSWER:

For this particular model, the problem was with three lines of data in drop.cdb.

*SET,_PARTSEL, 1.000000000000
*DIM,_PSEL ,ARRAY, 6, 1, 1,
*SET,_PSEL ( 6, 1, 1), 1.000000000000

These three lines of data are un-selecting parts 1-5.
That`s why *DELETE_PARTS are written in file.r.

By including PARTSEL,`ALL` command in drop.run, the third step (after the rigid to deformable switch) works as documented.





Show Form
No comments yet. Be the first to add a comment!