Part 2 of 5
Part 2 of 5
Part 2 of 5
In my previous article, covered the various API standard section references for rotor dynamic analysis of
different centrifugal machines (pumps , compressors , steam and gas turbines) and different rotating elements
connecting the driver and driven equipment (Coupling and gears). In this article further cover the fundamentals
of the how to perform the rotor dynamic analysis (Lateral Analysis) of rotating machine drive as per the API RP
684.
API RP 684 forms the basis of rotor dynamic analysis (lateral, stability and Torsional). The purpose of API RP
684 is to give the person who is performing and reviewing the calculations, 1) Provide guidance on the
requirements for analysis , 2) Aid in interpretation of rotor dynamic reports and 3) Provide guidance in judging
the acceptability of results presented.
The API RP 684 divided mainly into five different parts Introduction and overview , Lateral Rotor dynamic
analysis , Stability analysis, Torsional analysis and Balancing of Machinery. Three steps are important to
achieve the accurate results about the excitation frequency and amplitude values and unbalance location during
the analysis and design review of machine. They are 1) Modeling Criteria , 2) Analysis techniques and Results
and 3) Machine specific considerations.
The above figure shows the simple spring mass system. When the force is applied on the unbalance mass it
excites the system causing displacement . When the excitation frequency matches with natural frequency of
simple spring mass system the maximum displacement occurs. This state is called resonance. From the
governing mechanics equations of the above system is as shown below. From the governing equations the
natural frequency can be derived.
From the above equations both undamped and damped critical frequency is depend the stiffness and mass ,
damping co-efficient in case of damped system. The stiffness and damping co-efficients will be discussed in
details in the jeffcott rotor model. The simple rotor model is not sufficient for the actual rotor models. Because
the actual rotor is not a single degree of freedom system as shown in the simple spring mass system. The
stiffness characteristics varies widely in longitudinal and transverse directions. And Jeffcott rotor model brings
lot of additional characteristic values and understanding about the Rotor models with multiple degree of
freedom. The jeffcott rotor model is shown in the figure.
The stiffness used to calculate the natural frequency is combination of shaft stiffness , bearing stiffness and
bearing housing support stiffness , foundation stiffness. So change in any one of the stiffness has an influence
on the excitation frequency. The shaft stiffness can be calculated by using the solid mechanics formulas
depending on the support. Mainly it depends on whether the disc mass is overhung or simply supported between
the bearings .
The another important factor is damping coefficient. In simple terms damping force is the resistance offered by
the system against the excitation force and shaft displacement due to excitation force and frequency. The
damping is offered by the bearings and seals, diaphragms and other components in the rotor system. The
damping co-efficient is one of the most important factor which decides the amplification factor of the induced
excitation. The amplification factor is the ratio between the damping co-efficient of the system to the critical
damping co-efficient of the system. The various useful relations are given below.
From the above formula it is evident that the critical damping co-efficient also depends on the stiffness of the
rotor. Damping properties of the system can be modified by changing the bearing characteristics and clearances
between seals, diaphragms in the rotor systems. The above paragraphs are basic to understand the rotor system.
Now in the next paragraphs will see how the lateral critical speeds are calculated and analyzed and typical
outputs from lateral critical speed analysis and its inferences.
The advantages of the unbalanced response analysis is 1) It is planar and two dimensional , 2) It provides an
preliminary location of excitation points and where the unbalance mass need to be placed during the further
analysis to understand the machine behavior.
a) Adequate separation margin between the critical speeds and operating speeds
b) The probe vibration limit is not exceed with in the specified operating speed range even with twice the
maximum allowable residual unbalance present
c) No rubbing will occur even if the rotor's balance state degrades to the probe vibration limit
The unbalance mass location for this analysis is selected based on the output mode shape of Undamped critical
speed analysis. The distribution of this unbalance becomes as important as its amount. This is due to the fact
that to excite a natural frequency of any system, the forces must not be at the node points of the natural
frequency's mode shape. All modern day damped unbalance response analysis are performed by using the
algorithm developed by Lund and Orcutt.
The Output of the Damped unbalance response analysis is Bode Plot. Bode plot is the plot between the shaft
displacement from it is Centre line vs speed of the machine. The pecks in the bode plot's are the critical speeds
of the machine.The sample bode plot of charge gas compressor from API RP684 is shown below.
In general the vibration amplitude shall not exceed the 75% of clerance limit between the rotor and stationary
part as established in API Standard paragraphs.
Stability Analysis:
The stability analysis is performed to understand the behavior of shaft displacement with respect to time when it
is excited. It requested to understand whether the vibration is stable or unstable. Means amplitude of vibration
decreases with time it is stable vibration, if it increase with time then it is unstable vibration. I will try to cover
the stability analysis more in detail in the next article.
Machine Specific Considerations:
1) Steam Turbine: For Steam turbine bearing support flexibility and partial steam admission force are
specifically to be considered and impact on the outcome of the lateral critical speed analysis.
2) Electric Motors: In electric motors defining the winding mass and diameter in the rotor dynamic model
is difficult. This needs special consideration to get accurate results.
3) Gear Box: In gear boxes the gear loading pattern and tooth mesh frequency has a significant roles in the
outcome of the analysis.
4) Power Turbines and FCC Power recovery expanders: The inlet temperature are close to
1200 deg F. The stiffness and damping co-efficient's shall be considered at this temperature. Mostly the rotors
of these machines are stiff and their first undamped critical speeds are well above the operating speed.
5) Axial Compressors: The modelling techniques depends on the type of shaft construction. Typically
the rotors are constructed in four different ways. They are disc-on-shaft, shrink fit, stacked disk with tie bolts,
drum rotors with studs or tie-bolts and solid rotors. Generally rotor stability is not a problem for these machines.
6) Centrifugal Compressors:
a) Multistage Compressor: The damping of fluid films of gas seals, bearing oil seals and bearing
damping has important influence on the critical speed , stability of machine.
b) Overhung Compressor: The gyroscopic effects also known as "Morton's effect" to be considered in
the analysis of overhung compressor rotors.
c) Integrally Geared Compressor: It is like overhung machine analysis but the bull gear with
additional loading to be considered.
b) Only few choices exist for the addition of trial weight in the rotor at shop test bench. The one possible
location is coupling. If the coupling space and size is small cannot add more trial weight. At this condition it is
difficult to excite the first mode.
d) There are some risks which should be avoided during the testing. The elevated vibration levels in testing may
be due to in accurate testing or calculations. In order to prevent the machine damage test should be stopped
immediately.
e) The limitation in measurment capabilities at test bench , rotors balanced in high speed balancing facilities
adding weight at mid span, quarter span and other combination not possible in the test bench.
In Section 2.10 of the API RP 684 titled Standard paragraphs is used in most of API machine standards. This
section provide the formula's for calculating the Unbalance mass, Separation Margin and correction factor to be
applied against the calculated peak to peak amplitude.
There are numerous Journal references given in API RP684 under each section. Please refer to them on case by
case basis to discover further about the individual topics and deepen the
I will cover the Stability and Torsional analysis basics and methodology in next article.
References:
API RP 684 Second Edition