Even though the role of damping in rotordynamics has been known for more than half a century, some incorrect statements on the effects of hysteretic damping, particularly in the subcritical range, can be found still in even the most authoritative journals. Although clarified in the 1970s, this issue recently resurfaced with the incorrect statement that hysteretic damping of rotating elements is destabilizing at any speed (even subcritical). The aim of the present paper is to clarify this issue in terms of the practical applications of the theory in an unequivocal way.

The centrifugal softening effect is an alleged and elusive reduction of the natural frequencies of a rotating system with increasing speed which is sometimes found in finite element rotordynamics. This reduction may, in some instances, be large enough to cause some of the natural frequencies to vanish, leading to a sort of elastic instability. Some doubts can, however, be cast on the phenomenon itself and on the mathematical models causing it to appear. The aim of the present work is to shed some light on centrifugal softening and to discuss the assumptions that are at the basis of three-dimensional FEM modeling in rotordynamics. One and two degrees of freedom models, such as the ones introduced by Rankine and Jeffcott, are first studied and then the classical rotating beam, ring, disk, and membrane are addressed. Some numerical models, built using the FEM, are then solved using both dedicated and general purpose codes. In all cases no strong centrifugal softening is found.