# Rotating Nodal coordinate frames

Finite Elements – Modify – Node – Edit: Analysis Coord system = Coord 1 –
Reference Coord System = Coord 1

– Analysis – solution type – normal modes – solution parameters – enable: cyclic symmetry
– write out a nastran input deck
– in
Product Family: MSC.PATRAN (0501) MSC/PATRAN R9

cyjoin side entry in cyclic symmetry
Last Reviewed: 03/16/01
Article ID: 4693
DESCRIPTION:
Changing cyjoin entry side 1 and side 2 changes the results. Why does it matter for rotation symmetry?

**********************************************************************************
In the MSC/NASTRAN V70.5 Quick Reference Guide, definition of the CYAX bulk data entry, Remark 1 states:

“1. The displacement coordinate system (see CD field on GRID entry) for a grid point lying on the axis of symmetry must be a rectangular system with the z-component of motion aligned with the axis of symmetry. The positive axis of symmetry is defined so that the azimuthal angle from positive side 1 to side 2 of a segment is in the same direction as the angle from T1 to T2 for the axis point. This is consistent with left- or right-hand rule.”

This means that the Z-axis if the GRIDs on the axis of symmetry must have a CD z-axis direction based on the locations of side 1 and side 2. By changing side 1 and side 2, this basis is violated.

If you want to reverse side 1 and side 2, you must reverse the direction of the CD z-axis on the GRIDs on the axis of symmetry.

**********************************************************************************

The problem is with CYJOIN not CYAX. Under CYJOIN in the QRG we suggest that cylindrical or spherical CD for rotational symmetry. There is no mention of a requirement for a rectangular system. If there is such a requirement then it should documented in the QRG. Also I tried a run with side 1 and 2 reversed but with z-axis of the CD reversed and it still fails to give the correct answer.
Could you give this some more thought. It may be unclear documentation but if that’s the case then we need to correct it.

**********************************************************************************

Bob Harder writes:

Here is how I would explain the cyclic rules. If there are no CYAX points, the SIDE-1 and SIDE-2 data can be interch