I have a simple large deflection model that has undergone a rotation. When I list the results using a cylindrical RSYS the nodal displacements appear to be in a Cartesian system. What's happening?


This is something that can be confusing in ANSYS. For a good illustration of what's happening, look in section 5.4.1 of the Basic Analysis Guide (figure 5.22).

Using a simple beam model comprised of 3 beam elements with nodes at (0,0), (2,0), and (0,2), we can show the effect of RSYS. The model forms a triangle which we will rotate about the origin by ROTZ = 45 degrees. If we look at the results in "RSYS,0" and "RSYS,1" here's what ANSYS reports:

RSYS KEY SET TO 0
NODE UX UY UZ USUM
1 0.0000 0.0000 0.0000 0.0000
2 -0.58579 1.4142 -0.49359E-05 1.5307
3 -1.4142 -0.58579 -0.12665E-04 1.5307

NODE ROTX ROTY ROTZ RSUM
1 -0.65037E-05-0.13278E-05 0.78540 0.78540
2 -0.65054E-05-0.13289E-05 0.78540 0.78540
3 -0.65053E-05-0.13291E-05 0.78540 0.78540

RSYS KEY SET TO 1
NODE UX UY UZ USUM
1 0.0000 0.0000 0.0000 0.0000
2 -0.58579 1.4142 -0.49359E-05 1.5307
3 -0.58579 1.4142 -0.12665E-04 1.5307

NODE ROTX ROTY ROTZ RSUM
1 -0.65037E-05-0.13278E-05 0.78540 0.78540
2 -0.65054E-05-0.13289E-05 0.78540 0.78540
3 -0.13291E-05 0.65053E-05 0.78540 0.78540


The Cartesian results (RSYS,0) are easy to see since ux and uy are merely transposed for nodes 2 and 3. In the cylindrical case (RSYS,1), you can see that the translational results are identical for nodes 2 and 3. If you look at figure 5.22, you can see what's happening; each node is reporting displacements for a local nodal system that has been rotated into a global cylindrical orientation.

Using RSYS, ANSYS rotates the nodal coordinate system accordingly and the results are interpreted with respect to that local nodal system's orientation. The nodalsystem, however, is always a Cartesian system. What typically confuses ANSYS users is the fact that, although RSYS is set to a cylindrical system, this only means the beginning orientation is aligned with the cylindrical system. The reported degree of freedom results, however, are Cartesian with respect to the now rotated nodal coordinate system.





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