Page 190 - Steam Turbines Design, Applications, and Rerating

Rotor Dynamics Technology 169

Figure 10.4 Pairs of critical speeds exist for each mode whenever both the
vertical and horizontal bearing support stiffness are different. (General
Electric Company, Fitchburg, Mass.)

arate vertical and horizontal criticals that are reasonably close in
speed, the rotor will whirl in an orbit that exhibits one broad response
peak instead of two sharp peaks. Depending on the mode shape, par-
ticularly the second critical, a heavily damped response peak may not
be discernible at all. This often leads to a misinterpretation of the test
data and needless efforts to establish operating speed margins for the-
oretical, undamped criticals.

10.3 Effects of Damping on Critical
Speed Prediction
The damping forces within the bearing oil film are functions of vibra-
tion velocity and are normal to the journal deflection. The effect of
dashpot damping on a simple spring-mass system is to depress the nat-
ural frequency as well as to reduce the peak amplitude. However, bear-
ing oil film damping increases the equivalent stiffness and tends to
increase the resonant speeds. Since the journal whirls in an orbit, the
effect of damping on the equivalent stiffness will also depend on the
shape of the orbit.
The prediction of critical speeds, using the dynamic stiffness curves,
can be improved if an equivalent stiffness includes the approximate
effects of damping. Two examples are presented in Table 10.1 to show
the critical speeds predicted by the dynamic stiffness map as compared
to the response peaks calculated by the rotor response analysis. Exam-
ple A is a large, medium-speed rotor with four-pad, tilting-pad bear-