Spring connector - No or small K6 yeilds large displacement

[mattjTD8AS]

I have been doing an analysis with two spring connectors with values representing the shaft support bearings. The stiffness matrix was provided by the bearing manufacturer, and does not have a Rz/K6 factor, as bearings do not exhibit a stiffness (ignoring frictional losses) about the bearing axis.

When no value, or a very small value is entered into K6, we observed very large displacements, specifically in the X-direction. When one, or both bearings got a high value for K6, this displacement ceased.

This displacement does not make sense to me or my team, and was wondering if this is an expected behavior.

[John_Holtz]

Hi @mattjTD8AS 

In general, a stiffness of 0 is bad in a numerical solution. The analysis is solving for the displacements = force/stiffness in all six directions. Force/0 can give any number between 0 and infinity when a numerical computer is involved.

The details of the model may make a difference between the results being expected or unexpected. For example,

1. What is the orientation of the model? (You implied the axis of the shaft is parallel to the global Z direction. If so, then I agree that K1 and K2 are the stiffness in the shear directions, k3 is the stiffness in the axial direction, and ideally K6 could be zero.)
2. What prevents the shaft from rotating freely about its axis in the analysis? Is there a constraint somewhere else to prevent it? If the shaft is free to rotate, then any imbalance in the load will cause the shaft to try to rotate, and force/0 will lead to the X and Y displacement being due to the shaft rotation.
3. What are the springs connected to? If connected directly to a node on a solid element, then no directions should be left as blank or 0. Other elements that do support rotations (such as rigid connectors or beams) may run okay if a stiffness is left blank.

John