Page 236 - Compression Machinery for Oil and Gas

Reciprocating Compressors Chapter 5 225

signals and are often catastrophic. Typical package instrumentation is focused
on the monitoring of lateral vibration probes, and most often does not include
provisions for monitoring torsional oscillation. In addition, reciprocating
machinery tends to generate significantly large forcing torques at discrete
orders of operating speed, which can result in problematic excitation of critical
speeds. A proper torsional analysis is key to avoiding these types of problems.
The intent of this section is not to provide a listing of requirements for a
complete torsional analysis, as several notable resources are available that out-
line such information (e.g., see the GMRC publication Guideline and Recom-
mended Practice for Control of Torsional Vibrations in Direct-Driven
Separable Reciprocating Compressors [5], and API 684 [6], among others).
Rather, the following discussions provide a brief overview of the various com-
ponents involved in a typical torsional analysis, and some practical guidance for
dealing with common issues encountered during this type of work when recip-
rocating machinery is involved.
Torsional evaluations generally involve a steady-state analysis, preparation
of interference diagram(s), forced response analyses, and transient analyses (if
necessary and applicable). These analyses are most often conducted for trains
involving compressors, pumps, motors, turbines, or engines, and any associated
rotating components such as gearbox shafts, couplings, and viscous dampers.
One common question that arises when considering this type of effort is when
such an analysis is required, and what type of analysis is necessary. Generally
speaking, a complete torsional analysis is recommended when a nonduplicate
system is installed, or significant operating condition changes are planned.
Section 2.2 of the GMRC document [5] referenced above also provides a meth-
odology for determining what types of analyses are needed in specific instances.
The steady-state torsional analysis involves preparation of a torsional mass-
elastic model derived from manufacturer provided drawings and mass elastic
information data for the equipment involved. The steady-state torsional natural
frequencies (critical speeds) and mode shapes are then calculated. The mode
shapes are plotted to graphically show the deflection associated with the tor-
sional natural frequencies. This allows for investigation of controlling stiffness,
and relative model participation of the major inertias in the system.
An interference diagram (Campbell diagram) is prepared to graphically dis-
play prevalent excitation energy orders vs. speed and frequency. The specified
operating speed range(s) are superimposed on the interference diagram, and any
coincidences between calculated critical speeds and prevalent excitation energy
in the system are identified. The likelihood of exciting the critical speeds
involved in any such coincidences is assessed by studying the respective mode
shapes and potential excitation mechanisms.
A forced response analysis is usually conducted to determine anticipated
stress levels in the shafting during normal operation. For reciprocating equip-
ment, these calculations involve the characterization of dynamic torque content
for nominal operations, in addition to single cylinder engine misfire and locked