Page 245 - Steam Turbines Design, Applications, and Rerating

224 Chapter Twelve

critical and supercritical form, and no clear preference can be logically
spelled out for one or the other. Which to choose depends, among other
factors, on the design tradition of the manufacturer and the optimum
configuration of the machine.
About turbines for driving pumps and compressors, the issue is to
some extent influenced by past practice, since rotors for low steam
throughput capacities can be made subcritical without any serious
drawbacks. However, larger subcritical rotors may result in a design
that might be quite impractical. All large power station turbines in the
100 to 1000-MW range, for example, run at supercritical speed. Also,
knowledge in this field has now advanced so far that it is no longer suf-
ficient to design for rigid bearings, without taking into consideration
oil film, bearing and foundation stiffness.
In comparing the two kinds of machines, it can be assumed that the
efficiency penalty with subcritical operation is greater in the case of
the impulse turbine. The reason for this is that to raise the critical
speed of the rotor, the hub diameters have to be enlarged, whereupon
leakage losses, otherwise a point in favor of the impulse construction,
are greatly increased.
In addition, the following requirements are essential to the reliable
operation of any rotor:

■ Proper design of rotor and bearings
■ Correct choice of bearing type
■ Excellent balance

12.4.2 Blading
Apart from the rotor, the blading of impulse and reaction turbines show
the most marked differences. By its nature, the reaction machine
requires 75 to 85 percent more stages than an impulse turbine, for the
same heat drop. Despite this, the casings of both types are roughly the
same length because a single-reaction stage is much shorter in the axial
direction than an impulse stage, where the moving blades and the space
between the blades take up more length.
The fact that there are fewer stages often leads to the mistaken view
that on strength grounds or for operational reasons the reaction stage
cannot withstand such severe stresses, and therefore more stages are
needed. This is not the case at all; the number of stages is governed by
thermodynamic differences.

Vibration. If one wishes to examine the behavior of the blading in nor-
mal operation, it is sufficient to consider the dynamic stresses on the
blades. The static forces (centrifugal force and the force exerted by the