Modal analysis using spring connectors

[Anonymous]

I would like to perform a modal analysis on a shaft that is subjected to rotational force around its axis. There are three radial forces that act on the shaft at specific areas, two of which are restoring forces (top and bottom) and one is an unbalancing force (middle). 

The current model replicates the restoring and unbalancing forces in the form of stiffness that act radially on the specific shaft surface by using rigid and spring connectors in conjunction. The unbalancing stiffness should theoretically act as a negative stiffness.

The restoring stiffness has been modelled by using a rigid connector to the part of the shaft surface where the restoring force acts, with the independent point of the rigid connector at the centre of the selected area (i.e., on the axis of the shaft in this case). A spring connector is then used at this independent point to add the stiffness. The spring is constrained using the “Grounded spring” function.

The unbalancing stiffness has been modelled by having the independent point of the respective rigid connector connected to a spring connector which is grounded at an offset (100mm in this case) from the axis of the shaft on a cylindrical coordinate system.

The idea of having the unbalancing stiffness is that the spring does not bring the shaft to rest at its equilibrium position. I would like to know if this setup does act as a restoring/unbalancing stiffness in my model.

How can I use spring connectors correctly to have them behave in this manner?

• Am I interpreting the function of rigid and spring connectors incorrectly?
• How does the stiffness value entered for the spring connector act in a radial system?
• If the shaft does start to move off-centre, does the spring connector modelled on the radial system pull it further away (Increases the unbalancing force as the shaft travels off-centre. This is what I expect the model to do) or push it towards the axis of rotation?

Can the model be created using differently (eg: forces that can be set using vectors but act in a radial manner) instead of stiffness? The magnitude of this force would need to vary depending on the offset it experiences.

I use “Prestress normal modes” analysis for this simulation.

@John_Holtz thank you for all the help. Perhaps you have some suggestions regarding this problem as well