Proceedings of the Forty-First Turbomachinery Symposium

September 24-27, 2012, Houston, Texas

TORSIONAL NATURAL FREQUENCIES:

MEASUREMENT VS. PREDICTION

by
Qingyu Wang
Mechanical Engineer
Elliott Group
Jeannette, Pennsylvania

Troy D. Feese
Senior Project Engineer
Engineering Dynamics Inc.
San Antonio, Texas

and
Brian C. Pettinato
Manager, Product Development
Elliott Group
Jeannette, Pennsylvania

Qingyu Wang is a Mechanical Engineer at Brian C. Pettinato is Manager of Product

Elliott Group in Jeannette, Pennsylvania. Development at Elliott Group in Jeannette,
He has been with Elliott Group since 2005. Pennsylvania. He has been with Elliott
His areas of expertise include lateral and Group since 1995. His areas of expertise
torsional rotordynamics. He received his include lateral and torsional
B.S. and M.S. both in Mechanical rotordynamics, vibration analysis, and the
Engineering from Tsinghua University, testing and evaluation of fluid film journal
1997 and 2000 respectively. He received bearings. He currently manages a group
his Ph.D in Mechanical Engineering from responsible for compressor and expander
the University of Virginia, 2008. He serves on the API 684 technology development. Prior to joining Elliott Group, Mr.
rotordynamics task force, and is a member of ASME. Pettinato worked as a project engineer for an aftermarket
bearing manufacturer. Mr. Pettinato received his B.S.
(Mechanical Engineering, 1989) and M.S. (Mechanical
Troy Feese is a Senior Project Engineer at Engineering, 1992) degrees from the University of Virginia. He
Engineering Dynamics Incorporated (EDI) has coauthored over ten technical papers, and holds one U.S.
in San Antonio, Texas. He has 21 years of patent. He is a registered Professional Engineer in the State of
experience performing torsional vibration, Pennsylvania. He serves on the API 684 rotordynamics task
lateral critical speed, and rotordynamic force, and is a member of both ASME and STLE.
stability analyses as well as evaluating
structures using finite element methods.
He conducts field studies of rotating
machinery, reciprocating equipment,
piping, structures, and foundations. He is a lecturer at the
annual EDI seminar and has published technical papers on
torsional vibration, lateral critical speeds, and balancing. He
received a BSME from The University of Texas at Austin and
has a MSME from UTSA. He is a member of ASME,
Vibration Institute, contributed to API 684, was a speaker at
NPRA, and is a licensed Professional Engineer in Texas.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

ABSTRACT accuracy has not been benchmarked on a massive scale.
Whereas torsional measurement during complete string test has
Excessive torsional vibration can cause damage or failure been an option within API 617 since 1988, it has rarely been
to rotating equipment trains thereby resulting in costly selected by end users. This is due in part to the expense and
shutdowns. A comprehensive torsional vibration analysis is difficulty of a full string test.
the typical method of designing a torsional system that avoids
such problems. Despite the lack of benchmarking, there are relatively few
torsional failures related to insufficient separation margin from
Requirements of a torsional system design are commonly running speed. For example, Elliott Group has over 40 units
based on the API Standards (API 617, 2002 and API 684, in operation with insufficient margin from 2 compressor
2005). These standards require torsional natural frequencies running speed, all justified by stress analysis, and all having no
(TNFs) to have at least 10% separation margin (SM) from any problems. Insufficient margin from 1 running speed could
excitation frequency. If the recommended SM cannot be be of greater concern especially for certain torsional mode
achieved, then the torsional system design must be shown to be shapes. High risk cases that can be improved by separation
acceptable by stress analysis. The validity of the predicted margin include avoidance of pressure pulsation excitation
TNF and any stress analysis is dependent on the accuracy of the (Feese and Hill, 2009), and avoidance of transient startup and
model. fault excitation due to 1 and 2 line frequency (Vance et al.,
1984) such as occurs due to motor or generator applications,
Some degree of uncertainty is always present within the each of which can be improved by having good separation
analytical data, the modeling techniques, and the assumptions margin from these excitation frequencies, which are
for excitation and damping. This paper provides an non-synchronous.
uncertainty study of more than ten (10) torsional systems.
Major sources of uncertainty in torsional modeling are A torsional analysis that is performed utilizing good
identified. The effect of variation in mass-elastic data is modeling techniques, a proven computer program, and accurate
examined, and a comparison between measured and predicted mass-elastic data is generally considered to have accuracy for
TNFs for numerous cases is presented. Based on the studies the primary TNFs of 5%. There have been only a few
and measurements, a reasonable SM range is presented, and 5% comparisons made between predictions and measurements.
SM for measured TNFs is proposed.
LITERATURE REVIEW
INTRODUCTION
Lateral Natural Frequency Accuracy
The design of safe and reliable turbomachinery requires
proper vibration isolation by considering the lateral and Lateral critical speed predictions are required by API
torsional rotordynamics. Torsional vibration is the focus of standards to be within +/- 5% of the measured values. Lateral
this paper. systems include significant damping, and complicated
hydrodynamic effects, and would therefore seem to be more
API standards have required a torsional analysis since 1973 difficult to predict than a torsional system. In the case of
(API 617, 1973). These torsional requirements have evolved lateral dynamics of rotor-bearing systems, Vance, et al., (1984)
over time such as discussed by Pettinato, et al., 2011. The investigated rotor natural frequencies in both the free-free state
primary requirement of the API standard is a separation margin and the fluid film bearing supported state. Major sources of
of 10% from excitation frequencies, especially running speeds, error were attributed to modeling the fits of shrunk-on
2 running speeds, and one and two times electrical line components. The prediction error of turbomachinery lateral
frequency when applicable. The method of achieving natural frequencies was cited as being within 7% for rotors
acceptable torsional separation margin is generally limited to when considering bearing and foundation properties as well as
coupling selection and tuning, and in some cases modification the stiffening effect from shrunk-on disks. An examination of
of system inertia. There are limits to how far a torsional the free-free impact data would tend to indicate a 5% error limit
natural frequency can be shifted, and the required 10% SM for the test results.
cannot always be achieved, and the API standard allows for
torsional system justification by stress analysis in these Torsional Natural Frequency Accuracy
situations. Unavoidable violation of 10% SM requirement is
not uncommon for certain systems, such as variable speed By contrast, the torsional system is closer to a pure
systems with large operating speed range or reciprocating structural model with low levels of damping commonly cited as
systems with closely spaced excitation frequencies. It is being 2.5% modal damping or less (Wachel and Szenasi, 1993).
assumed that these API recommendations refer to calculated Given an assumed accuracy of +/-5%, if a TNF was predicted
results performed in the design stage. (Other criteria related to to have 10% separation margin, in reality the true separation
synchronous motor startup analysis and transient events such as margin would be between 5% and 15%. This assumed
electrical faults are not within the scope of this paper.) accuracy is backed by only a small number of comparisons
between predictions and measurements of torsional systems
Unlike lateral critical speeds, torsional critical speeds do within the literature. A sampling of the literature is presented
not require verification. Consequently, torsional system herein.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 After experiencing a torsional failure on a synchronous Hill (2009). The standard deviation of Steiners presented
motor driven string (Sohre, 1965), Elliott Group performed results ranged from 1.9% to 12.1%.
torsional measurements to develop its first generation of criteria
for torsional stress analysis. In this effort, torsional system Torsional Measurement
excitation and damping were characterized and reported by
Pollard, (1967). Some critical speed results from this earlier As shown within the literature, torsional testing is certainly
period of testing are presented herein (Pollard, 1972 and De not new and it has certain advantages (Feese and Hill, 2009).
Choudhury, 1979). In general, measurements should be more accurate than
calculations especially when conducted under full load where
Mondy and Mirro (1982) performed torsiograph actual response can be determined such as dynamic torque,
measurements on three equipment strings. Overall, the authors alternating stresses, torsional oscillation, current pulsation, etc.
concluded the error of using an undamped model to calculate These values can then be compared to allowable limits, and if
the torsional natural frequencies of turbomachinery strings can determined to be acceptable under all possible load conditions
be limited to less than +/-5% of the actual frequency. It including unloaded case, the SM would be secondary as the
should be noted that one string initially had computed errors on stress levels would be validated. Unfortunately, testing all
the order of 18%, but this was eventually attributed to poor load conditions may be difficult in a shop environment, during
coupling stiffness data that was later corrected. an outage, or while operating a plant. Some shop tests are
unloaded mechanical runs, and do not use the contract driver.
Murray, et al., (1996) examined three different motor
driven reciprocating compressor strings. Torsional prediction Even with strain gage telemetry system, laser vibrometer,
and measurement were conducted on two of those strings of or shaft encoder only a few points can be measured within the
equipment. The prediction error was 7.5% for the first string train. Therefore, a torsional analysis is still needed to infer
and 1.8% for the second string. The authors advocated an amplitudes at other locations that cannot be accessed for
uncertainty analysis for torsional natural frequencies, assigning measurement. By normalizing the computer model to match
different uncertainties to the stiffness and inertia of each the measured data (both TNFs and amplitudes) reasonable
component. One specific study resulted in an uncertainty of analytical accuracy can be assured such that conclusions and
-4% to +2% for the TNFs of that system. inferences can be drawn at the points of the system that are not
directly measured.
Coupling Models
The systems, which would benefit the most from torsional
Since couplings are often the most torsionally flexible parts measurement, and the ones with the highest risk for torsional
of the train, the torsional critical speed prediction can be failure seem to be related to Variable Frequency Drive (VFD)
particularly sensitive to their modeling. Methods for applications (Feese and Maxfield, 2008, Kerkman et al. 2008,
modeling couplings are discussed by Ker Wilson, (1956), Piergiovanni et al. 2010, Kocur and Muench, 2011, and Bosin
Nestorides, (1958), and AGMA 9004, (2008). The 1/3 et al. 1999). The causes of these failures cited in recent
penetration rule is cited where the shaft is assumed to twist literature were independent of separation margin and
independently of the total length of the coupling (as shown in unpredictable in the design stage using normal methods. Each
Figure 7). This one-third rule is applied whether the of these cases required torsional testing to solve the problem.
connection is keyed, splined, or interference fitted.
Accordingly, Elliott Group has performed torsional
Calistrat and Leaseburge, (1972) examined the effect of measurements both independently and with EDI. The primary
shaft-to-hub connection and flange-to-flange connection. reason for these measurements was to ensure a safe reliable
Their findings indicated that the torsional stiffness of system for the end user. Other benefits are to investigate
hub-to-shaft connection is substantially influenced by the discrepancies between measurements and predictions:
transmitted torque, the hub-to-shaft interference fit, and the
ratio between the shaft and hub outside diameters. Rotational 1. Different modeling methods were examined to better
speed was determined to have a lesser effect. Likewise, the correlate with measurements.
torsional stiffness of flanged connections can also be variable 2. Job coupling static stiffness was measured.
depending on the amount of bolt torque and number of bolts. 3. Other factors that might affect the coupling stiffness,
such as the shrink-fit tolerance and spinning speed
The conclusion that the modeling of fits and interfaces are influences, were considered.
critical to attaining accuracy for the torsional model is quite 4. Coupling vendors were surveyed for stiffness
similar to the conclusion reached by Vance, et al. for lateral accuracies and modeling techniques.
modeling (Vance et al., 1984). Steiner (2007) reported on
stiffness measurements of seven couplings manufactured by This paper discusses torsional error and uncertainty, and
three different vendors. His research suggests that examines over ten case histories to demonstrate where
contribution of disc pack and bolting to coupling stiffness may variability can occur. Studies of prediction accuracy and
be difficult to predict. The prediction error for coupling measurement repeatability of identical units are provided to
stiffness ranged from -10.6% to +12.1%, which is within the further a discussion in regard to a reasonable SM of both
20% coupling stiffness accuracy range discussed by Feese and analysis and measurements.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

TORSIONAL ANALYSIS Szenasi, 1993). Some notable exceptions that are not covered
in this paper include: various elastomeric couplings and fluid
Modeling and Natural Frequency Prediction couplings, which have some considerable damping and
non-linear behavior.
An analytical method can be used to predict the TNFs of a
machinery train. One modeling method that is accepted These systems of equations can be solved to determine the
within the industry is to use lumped inertias and equivalent eigenvalues and eigenvectors (i.e., the TNFs and their
torsional springs (mass-elastic model) as shown in Figures 1 corresponding mode shapes). Methods of solution include use
and 2. of commercially available Eigen-solvers, such as MATLAB, or
use of the Holzer or transfer matrix method (Dawson and
Davies, 1975). The inertia/stiffness model can be used to
calculate the vibrations if proper damping is added. The
uncertainty related to different equation solvers along with the
problem of root finding is not considered in this paper.

The calculated TNFs and modeshapes are typically plotted

as part of the torsional report, such as shown in Figure 3.

Figure 1. Motor-gear-compressor train. Mode 1: 865 cpm

Mode 2: 2098 cpm
Mode 3: 22962 cpm
1

Normalized Amplitude
0
Figure 2. Lumped Inertia-Stiffness (Mass-Elastic) System.

An accurate model can only be obtained from accurate

-1
representations of the components. Guidelines for modeling 1 2 3 4 5 6 7 8 9 10
each component including the couplings can be found in
literature (API 684, 2005, Ker Wilson, 1956, and Nestorides, Figure 3. Torsional Mode Shapes.
1958). The component models can be assembled into system
inertia and stiffness matrices as shown in Equation 1 Amplification Factor (AF) and Separation Margin (SM)

The calculation of AF and SM are illustrated in Figure 4,

J1 which is similar to the API lateral analysis (API 617, 2002 and
J API 684, 2005).
J 2


Jm (1)
K1 K1
K K K K2
K 1 1 2


Km2 K m 1

where m is the total number of inertias. Note that for trains

with different component speeds, the inertia and stiffness Figure 4. Torsional Response Plot.
values are usually converted to the same equivalent speed
before building the system matrices. However, more
Nc
advanced torsional analysis programs use full-matrix AF (2)
representation and can handle all types of gearboxes including N 2 N1
epicyclic gears.

An undamped analysis is typically sufficient to predict N mc N c

TNFs of most systems since torsional damping is usually small SM 100% (3)
and on the order of 2.5% modal damping or less (Wachel and N mc

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

where Nmc is the speed in the operating range that is closest to Note that the half power point (0.707 peak amplitude) method
Nc; usually the maximum or minimum continuous operating provides a good approximation of the maximum gain when the
speed. damping is small (see Ewins, 2000 for derivation/proof).

A typical representation of SM is the interference or Campbell Feese and Hill, (2009) show a measured AF, and the AF
diagram as shown in Figure 5. for common equipment trains is between 10 and 50 (damping
ratio 0.01 and 0.05). The system gain can be calculated using
Equation 5 for different speeds (different SMs), and different

Compressor Speed
2X Line Frequency
Line Frequency
damping ratios/AFs, as shown in Figure 6.
Motor Speed

25000
50

22962 =0.01

40

20000 2*RS

30
=0.0167

|G (i )|
15000
Frequency (cpm)

20
=0.025

=0.05
10

10000 1*RS

0
0.8 0.85 0.9 0.95 1 1.05 1.1 1.15 1.2
n

5000

Figure 6. System Gain (AF) for Different Damping Ratios.

2098

865 The numerical results in Table 1 show that for the API
0
0 2000 4000 6000 8000
recommended separation margin of 10%, amplification would
Speed (rpm) be approximately 5 for most cases, and amplitudes would
almost double if the separation margin was 5% instead of the
Figure 5. Interference Diagram. recommended 10%. Low separation of 1% or less will
increase the amplitudes significantly.
Calculation of the AF is not a required part of the API 617
torsional analysis nor is it related to the SM requirement (10%). Table 1. Magnification at Different SMs and AFs.
However, SM and AF are related to vibration amplitude, which
can be demonstrated using a simple model. Damping Separation Margin (SM)
Ratio 0% 1% 5% 10% 15%
A single-degree-of-freedom mass-damper-spring system is
depicted having equation of motion as shown in Equation 4, 0.05 AF = 10 10 7 5 3
and the gain/vibration amplitude as shown in Equation 5 (see 0.025 20 19 9 5 4
Meirovitch, 1986 for details). 0.0167 30 26 10 5 4
0.01 50 36 10 5 4
x(t ) 2 n x (t ) n2 x(t ) n2 Ae it (4)
Uncertainties in Modeling
where n is the natural frequency, and is the damping ratio.
If the lumped inertia-stiffness model truly represents the
x(t ) A G(i ) ei (t ) ,where equipment train, then the calculated TNFs will match the real
systems TNFs. All factors that cause a difference between the
1
G(i ) calculated TNFs and the true TNFs of the system are

1 / 2 /
(5) considered uncertainties in this paper. There are two types of
2 2 2
n n uncertainties in the modeling: errors and variations.

1 Error type uncertainties include:

AF is the maximum gain, i.e. AF G (i ) for Mass-elastic model contains insufficient number of
max
2 stations to predict natural frequencies of interest.
light damping. AF is the maximum amplification of the system Inaccurate dimensions.
gain (|G| = 1, when = 0), and AF is related to damping ratio. Inaccurate material properties.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Improper modeling of complex shapes and However, different motor/gear vendors quite often provide the
connections, such as motor core. torsional stiffness differently. Figure 7 (a) shows the correct use
Improper modeling of components, such as not of torsional stiffness: from the last inertia location to the start of
including entrained water in wet impeller inertia. the coupling. Figure 7 (b)-(e) show other ways that torsional
Misapplying the 1/3 penetration rule or neglecting it. stiffness is provided.
Assuming too much damping thus making predicted
torsional response unrealistically low. Since the shaft ends are the weakest part of the body, the
Unit conversion English versus SI. stiffness provided in different ways could have large
WR2 versus GD2. GD2 is based on diameter such that differences. When conducting the torsional analysis, the
GD2 = 4WR2. penetration factor needs to be carefully considered.
Manufacturing and assembly tolerances.
If correctly modeled, the uncertainties of the inertia of all
bodies including couplings are considered within 5% (see
Variation type uncertainties include:
Murray et al. 1996 for example). The stiffness of
Dynamic vs. static stiffness.
motor/gear/compressor is also considered accurate for most
Wear or deposit induced inertia change.
applications, usually within 5%. The coupling stiffness is the
Temperature induced stiffness change. least accurate and also most influential to the TNFs since they
Working condition/load induced stiffness change. are the weakest parts in the equipment trains.
Aging of material induced stiffness change.
Wearing, cracking or damage induced stiffness For coupling uncertainties, both error type and variation
change. type uncertainties are significant. Comparing to a rotor shaft,
Variation of shaft positioned in coupling hub. the coupling structure is much more difficult to be accurately
Variation in shrink fit. modeled. Besides following API Standard 671, (2010) and
Variation in bolt torque. AGMA 9004, (2008) coupling vendors also use FEA analysis
to optimize the design and calculate the stiffness. Some
Among all the error type uncertainties, one common manufacturers conduct tests to verify their own empirical
concern is the shaft end modeling for penetration type of stiffness calculations. However, even with all the standards,
couplings. Most coupling vendors consider the shaft end, which FEA analysis, large amount of tests, and strict rules for
is inside of the coupling, as part of coupling when calculating installation etc., for most vendors the uncertainty/accuracy of
the coupling stiffness, i.e., the 1/3 penetration rule in Figure 7. the couplings is still about 15-25%, and the estimation is based
on static test only (see Table 2).

Figure 7. Vendor-provided Torsional Stiffness.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 The reality could be even worse for the coupling stiffness are listed as -0.35% and 0.00%. Similar calculations were
accuracy because of the variations. For instance, load condition carried out for all other numbers.
can be a key factor for the torsional stiffness of disc type of
couplings because of the buckling effect (Steiner, 2007). The Several points can be drawn from Table 4:
buckling effect makes the stiffness vary due to the load
condition and history (see Figure 8). Vendors usually provide 1. Increasing the inertias always decreases the TNFs.
the coupling stiffness based on the rated/normal torque, so if Usually they affect the coupling mode which is next
the working condition is different, the torsional stiffness could to them, e.g., the motor affects the motor-gear
be 20-30% different from the designed value, which already coupling mode.
has 15-25% uncertainty. An example of TNF measurement can 2. Increasing the coupling stiffness always increases the
be seen in Figure 9 (the first TNF shifted 3%). TNFs. Usually the first coupling mode is the one
between the large inertias, e.g. motor and gear.
Rotation is known to affect the coupling stiffness in the 3. When the coupling sensitivity is within 0.3%, the
axial direction, but the influence for the torsional stiffness is not discrepancies between the predictions and
well established in the literature. An FEA study was conducted measurements (prediction errors) are within 5%,
on a case for which the axial stiffness changes 1/3 due to the which is a very weak relationship though.
rotation, and the results show that the influence on the torsional
stiffness is negligible. The first two points are straightforward, i.e., the systems
might be complicated but the basic physics works the same.
Aging of materials, including erosion/corrosion etc., could The third point, relationship between sensitivity and prediction
be a major problem for certain type of couplings, such as errors, needs more data to prove or disprove.
elastomeric type, so it needs to be considered at the design
stage. Second Study Worst Case Scenario

Influence of Uncertainties S.O.1-7 (M-G-C) and S.O.9 (T-C-M, i.e. no gear) are used
for the second study, estimating the worst case scenario.
With the estimation of the uncertainties for each Assume the uncertainties of all inertias are within 5%, and
component, it is possible to evaluate uncertainties in the uncertainties of all stiffness are within 5% except for the
predicted/calculated TNFs. coupling stiffness. Two sets of uncertainties of the coupling
stiffness are used, 15% and 25% (see Table 2 for reference),
Motor-gear-compressor trains are a common torsionally and the results are shown in Table 5.
concerned category. This paper conducted a study on TNFs
using seven shop orders of motor-gear-compressor trains whose The first five columns in Table 5 are the same as in
TNFs have all been measured (all with partial or no load). Table 4. The last two columns are the differences (in absolute
Some parameters are listed in Table 3. Two studies were values) between column 3 and the calculated TNFs with the
performed with these shop orders. assumed uncertainties, i.e., they are the uncertainties of the
predictions. The prediction uncertainties can be considered as
First Study - Sensitivity the overall sensitivity.

A sensitivity study was performed by varying each Since Table 5 shows the worst case scenario, it means if
parameter sequentially, such as the inertia of the motor or the the coupling stiffness has accuracy within 15%, and all other
stiffness of a coupling, to see the change in the calculated stiffness and inertias are within 5% of the real values, then
TNFs. The purpose is to find whether discrepancies between predicted TNFs should be within 10% of the true values. The
prediction and measurement are related to parameter variation. prediction uncertainty approaches 15%, when the coupling
stiffness accuracy is within 25%.
The simulation is done by applying 1% variation to the
inertia of the motor, gear, compressor, and the stiffness of the To investigate the discrepancies, some of the torsional
two couplings. The results are shown in Table 4, where I1-4 analyses were thoroughly checked to make sure there were no
stands for the inertia of the motor/gear/compressor/turbine, and human errors present. For some of the shop orders, the motor
K1-3 represents the torsional stiffness of the couplings. The and gear vendors were requested to confirm/verify their
predicted and measured values are listed for the coupling numbers. One coupling vendor even performed a static test to
modes of each shop order. In some cases, the 2 nd TNF could verify their predicted coupling stiffness. All checks indicated
not be measured. no major errors/mistakes, yet the differences are more than
10% from the measurements for some shop orders. One
The sensitivity study results, such as -0.35% and 0.00% in explanation would be the variation-type uncertainties for the
the table for S.O.1 in column I1, are calculated as: use +1% couplings, especially since all tests were performed with only
inertia for the motor and keep all other inertias and stiffness partial or no load. Another explanation would be that there
unchanged, calculate the TNFs (new TNFs), and the difference were still human errors, since some vendors may not be fully
between the new TNFs and the predicted for 1st and 2nd TNFs aware of the torsional modeling techniques.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Table 2. Coupling Vendor Survey Result.

Vendor A Vendor B Vendor C Vendor D Vendor E

Short: 10-25% Regular: 5-10%
Accuracy of torsional stiffness 5% 10% Elastomer 25%
Long: 5-15% block: 15%
Torsional stiffness test Static testing No in-house test Static testing Static testing Static testing
Penetration model (keyed & hydraulic) 1/3 penetration
Hydraulic fit 0.002-0.0025 0.002-0.0025 0.002-0.0028 0.002-0.0028 0.002
Interference
(inch/inch diameter) Straight 0.0005-0.00075 0.0005 0.0005-0.00075
Keyed 0.0005 0.0005-0.00075
Taper 0.001 0.001 0.001

Table 3. Parameters of the Geared-compressor Trains.

Shop
Start Year Driver Rated Power (HP) Driver speed (RPM) Gear Ratio Train Type1
Order
S.O.1 2010 11000 1000-1550 8.03 M-G-C
S.O.2 2009 4000 1200-1600 6.55 M-G-C
S.O.3 2009 4000 1200-1600 8.19 M-G-C
S.O.4 2009 9300 1450-1900 5.32 M-G-C
S.O.5 2009 13000 1400-1900 4.93 M-G-C
S.O.6 2009 12000 1400-1900 2.86 M-G-C
S.O.7 2009 1770 1400-1880 3.35 M-G-C
S.O.8 2005 1350 1040-1559 7.5588 M-G-C
S.O.9 2003 16200 3420-3636 1 T-C-M
S.O.10 2003 16200 3420-3636 1 T-C-C-M
S.O.11 1971 39793 3465-4851 1 T-C-C-C
S.O.12 1979 3485 5266-7899 1.4544 T-G-C
S.O.13 1977 6000 1800 4.2644 M-G-C

1 Trains are listed from driver to driven side. M means motor if on driver side (such as M-G-C), and means motor/generator if on
driven side (such as T-C-M/G). G means gear, C means compressor, and T means turbine.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Table 4. Sensitivity of TNFs to the Change of Inertia Stiffness.

Predicted Measured Difference Sensitivity of +1 percent change (%)

Shop TNF TNF
Order No. (Hz) TNF (Hz) (%) I1 I2 I3 I4 K1 K2 K3
1st 15.1 17.2 -12.21 -0.35 -0.12 -0.02 0.37 0
S.O.1
2nd 58.7 57.5 2.09 0.00 -0.07 -0.42 0.00 0.38
1st 15.8 17.2 -8.14 -0.35 -0.04 -0.1 0.38 0.04
S.O.2
2nd 32.7 32.2 1.55 -0.01 -0.25 -0.23 0.04 0.41
1st 15.2 16.2 -6.17 -0.34 -0.14 -0.02 0.40 0
S.O.3
2nd 64.9 0.00 -0.06 -0.41 0.00 0.43
1st 19.4 21 -7.62 -0.19 -0.26 -0.04 0.41 0
S.O.4
2nd 72.5 0.00 -0.07 -0.4 0.00 0.41
1st 20.8 23.6 -11.86 -0.25 -0.20 -0.04 0.38 0
S.O.5
2nd 71.8 0.00 -0.07 -0.41 0.00 0.33
1st 16.1 15.5 3.87 -0.29 -0.01 -0.2 0.24 0.13
S.O.6
2nd 37.6 36 4.44 -0.02 -0.39 -0.08 0.10 0.3
1st 11.8 12 -1.67 -0.35 0.00 -0.15 0.28 0.16
S.O.7
2nd 43.6 46 -5.22 -0.01 -0.44 -0.02 0.15 0.31
1st 12 11.6 -9.43 -0.33 -0.06 -0.11 0.39 0.01
S.O.8
2nd 53.4 0.00 -0.30 -0.19 0.01 0.39
1st 11.5 11.6 -1.01 -0.07 -0.08 -0.34 0.25 0.09
S.O.9
2nd 23.6 25.1 -5.84 -0.01 -0.35 -0.12 0.17 0.14
1st 8.6 8.6 -0.77 -0.12 -0.08 -0.04 -0.25 0.24 0.07 0.04
S.O.10 2nd 13.3 13.8 -3.98 -0.04 -0.27 0.00 -0.18 0.17 0.08 0.06
3rd 34.8 0.00 -0.02 -0.43 -0.04 0.00 0.19 0.09
1st 13.2 12.4 6.72 -0.16 0.00 -0.30 -0.04 0.12 0.38 0.01
S.O.11 2nd 34.0 32 6.25 -0.04 -0.20 -0.01 -0.26 0.20 0.03 0.27
3rd 36 35.1 2.66 -0.03 -0.21 -0.06 -0.19 0.18 0.09 0.22
1st 47.7 50 -4.57 -0.24 -0.06 -0.19 0.00 0.00
S.O.122
2nd -0.10 -0.38 0.00 0.43 0.00
1st 26.4 27.1 -2.58 -0.17 -0.04 -0.29 0.12 0.04
S.O.13
2nd 109 92.2 18.2 0.00 -0.37 -0.08 0.03 0.17

2 This train has a variable speed drive, and its internal low stiffness rather than the couplings dominates the first critical.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Table 5. Worst Cases of the Predicted TNFs Due to the Modeling Uncertainties.

Predicted Values Measured Difference Coupling Uncertainty

Shop
Order TNF No. TNF (Hz) TNF (Hz) (%) 15% 25%
1st 15.1 17.2 -12.21 8.78 13.09
S.O.1
2nd 58.7 57.5 2.09 8.91 13.36
1st 15.8 17.2 -8.14 9.22 13.99
S.O.2
2nd 32.7 32.2 1.55 9.49 14.55
1st 15.2 16.2 -6.17 9.12 13.75
S.O.3
2nd 64.9 9.29 14.13
1st 19.4 21 -7.62 9.18 13.86
S.O.4
2nd 72.5 9.14 13.87
1st 20.8 23.6 -11.86 8.89 13.34
S.O.5
2nd 71.8 8.40 12.39
1st 16.1 15.5 3.87 8.72 12.99
S.O.6
2nd 37.6 36 4.44 9.03 13.61
1st 11.8 12 -1.67 9.46 14.41
S.O.7
2nd 43.6 46 -5.22 9.54 14.63
1st 11.5 11.6 -1.01 8.13 12.16
S.O.9
2nd 23.6 25.1 -5.84 8.05 11.57

Table 6. Comparison of Measured and Calculated TNFs.

First TNF Second TNF

System
Measured Calculated Percent Error Measured Calculated Percent Error
1600 kW 720 CPM 686 CPM -4.7% 2760 CPM 2570 CPM -6.9%
7000 kW 1260 CPM 1165 CPM -7.5% -- 4347 CPM --
11,500 kW 930 CPM 968 CPM +4.1% 2,160 CPM 2,257 CPM +4.5%

Torsional Measurement to lack of shaft drawings and technical information,

testing should be performed to verify the analytical
According to API 684, 2005, a torsional analysis should results.
always be performed for new designs. However, other
industries do not require this and unfortunately, torsional Product development Newly designed systems that
vibration might not be considered to be a problem until after a will be mass produced should be tested under load. It
failure has occurred. Situations where torsional testing may is much easier to correct a problem with an initial unit
be required in addition to a computer analysis include: at the factory than it is to retrofit many units that have
already been shipped to customers and installed in the
field.
Troubleshooting torsional failures If an
unexpected or premature failure of a component Used systems Compressor systems that have been
occurs, testing of the repaired system is often the best modified or put into a different service, such as
method to fully investigate the cause(s) of the failure. re-staging and/or changing operating conditions,
should be re-analyzed or tested.
Critical applications If a system poses unusually
high risks to life, other machinery, or plant processes, Contractual Requirement Some municipalities
testing should be performed to ensure reliable have specifications that require torsional vibration
operation. This could include a prototype machine or testing of new units by a professional engineer during
a re-rated model operating at higher speeds or commissioning. Some end users also require
pressures than previously designed. full-load, full-string tests of their equipment at the
shop if possible or during commissioning at the plant.
High chance of variability This could include
Torsional testing is a bulleted item within the API
systems with torsionally soft couplings and/or wide
specifications indicating that it is an option that can be
speed ranges or operating conditions. If many
selected.
assumptions had to be made in the analysis phase due

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Torsional vibration is referred to as silent because it was shown that the TNFs of the low-speed side and high-speed
occurs in the axis of rotation, which conventional vibration side would match the modeling if analyzed decoupled (zero
monitoring equipment, such as accelerometers and proximity stiffness at the gear mesh).
probes, may not normally detect. Therefore, special test
equipment is needed to measure torsional vibration as discussed Waterfall plots of frequency spectra (i.e. Figure 21)
in the APPENDIX. One exception is gearboxes where are good for displaying multiple harmonics over a speed range.
dynamic torque produces varying tangential and separating Peaks often indicate resonant conditions. Order tracking can
forces between the gears and fluctuating loads on the bearings, also be used to separate responses at the various harmonics.
which can produce noise. Coupling chatter, disc pack Phase angles can be determined from a once-per-revolution
deformation, damaged rubber elements, damper failures, and tach signal and used to confirm a phase shift through a
compressor oil pump failures are possible symptoms of a suspected resonance as shown in Bode plots.
torsional vibration problem.
For systems without a gearbox, the torsional natural
Test Conditions frequencies and damping can sometimes be measured from the
time wave forms after a trip event. Suddenly removing the
Under different load conditions, the TNF might change. A power to the motor can cause a torque impulse and ring the
major factor is the stiffness change of the couplings. Figure 8 system.
shows the test results of static load versus angle of a disc-type
coupling. Under different loads, the stiffness could change An example is shown in Figure 10. By examining the
20-30%. Figure 9 shows the measured torsional vibration, and period of vibration, the first TNF was approximately 15 Hz.
the first peak shifts approximately 3%. The damping can be determined from the logarithmic
decrement, , which was approximately 0.44. This converted
For unloaded shop tests (centrifugal compressor operating to a damping ratio of 0.07 and an amplification factor AF of
under vacuum), it is best to collect data during a continuous approximately 7 (see Randall, 1990 for detailed explanations).
ramp up in speed. This helps to ensure that the gear teeth This system had a VFD motor and MVR (mechanical vapor
remain in contact due to torque required to accelerate the inertia recompression) fan with a rubber coupling, which had a
of the high-speed compressor. When the gear teeth reported dynamic magnifier of 8. Therefore, the
momentarily separate or backlash, the TNFs will change and measurements confirmed the analytical results.
the steady-state model for the entire train will be invalid. It

70000

60000

50000
Static Torque (lb-in)

40000

30000

20000

10000

0
-0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1

-10000
Angle (deg)

Figure 8. Static Load Test of a Disc Coupling.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

15.25 Hz
Vs.
15.75 Hz

Figure 9. TNFs under Different Load/Torque Conditions.

Figure 10. Free Torsional Vibration after Trip Event.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

ACCURACY OF PREDICTIONS The motor speed was controlled by a VFD and the
compressor was operated with a vacuum (no load). By
Motor Gearbox Compressor Systems accelerating the unit, the gear teeth would maintain sufficient
contact.
The systems consisted of a motor, speed increaser, and
centrifugal compressor as shown in Figure 11. The end-user The first and sometimes second TNFs could be determined
requested that torsional measurements be taken on several from the measured data. Table 6 shows how the torsional
compressor systems to verify the torsional natural frequencies. analyses agreed with the actual TNFs that were measured for
each system.
These measurements were performed during mechanical
runs at the shop. The strain gage telemetry system was Differences are attributed to coupling stiffness and/or
installed on the LS coupling as shown in Figures 12 and 13. simple motor model provided. It is believed that the accuracy
of the predictions could be improved with coupling test data
and multiple-station model for the motor core (Hudson and
Feese, 2006). Load may also change the torsional stiffness of
the shim pack couplings slightly.

Note that the 1600 kW system originally was found with a

much lower first TNF. Upon inspection, the motor shaft
diameter was undersized. This was corrected by
manufacturing a new motor shaft. Test results presented are
for the correctly sized motor shaft, which was later re-tested.

Effluent Pump System

The incoming water to the facility is primarily from storm

sewer drainage. The volume of water handled varies widely
Figure 11. Compressor System. from storm conditions to dry periods. The treated water is
discharged to the Pacific Ocean; therefore, the pump head
requirements depend on the tide as well as desired flow. The
effluent pump has a maximum capacity of 68.5 million gallons
per day (MGD), and 90 feet of head.

As shown in Figure 14, the system uses a Lufkin double

input, right-angle gear reducer with a Caterpillar diesel engine
with clutch on one input shaft and a Toshiba VFD motor with
clutch on the other input. The pump system is designed such
that only one clutch is engaged at a time. The vertical pump is
driven by the gear output shaft. The torsional natural
frequencies of the system change due to different drivers being
engaged or disengaged.

Figure 12. Low-Speed Coupling

Figure 14. Pump System.

Figure 13. Strain Gage Telemetry System.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 VFD Motor Operation It is interesting to note that the second author was recently
involved with another engine gearbox vertical pump system
With the VFD motor in operation, a torsional natural that experienced multiple line shaft failures at the keyway.
frequency (TNF) was measured at 12 Hz, which compares with These failures are believed to be caused by sudden engagement
9.5 Hz from the original torsional analysis. Another TNF was of the clutch between the engine and gearbox when the engine
measured at 170 Hz with the pump strain gages, which is already at a high idle speed. Although not typical torsional
compares with a calculated mode at 143 Hz. From strain vibration, it is important that sudden torque spikes not be
gages on the motor shaft, a TNF was found at 110 Hz, which introduced into the system that could yield the shaft material at
compares to a predicted value of 124 Hz. geometric discontinuities such as keyways and /or damage gear
components.
Engine Operation
Sewer District Pumping Station
With engine operation, the initial torsional response was
8.5 Hz, which compares with predicted 7.3 Hz (engine The system consisted of an induction motor, rigid
warm-up / clutched disengaged). During the speed sweeps, a coupling, and vertical centrifugal pump as shown in Figure 15.
TNF was measured at 10 Hz, which compares with 11.8 Hz The motor was rated for 1750 HP at 400 RPM. A new VFD
from the original torsional analysis. Two other TNFs were was being installed in order to operate the pump from 270 RPM
measured at 65 Hz and 85 Hz. These compare with the to 400 RPM and needed to be evaluated.
predicted values of 70 Hz and 89 Hz.

There were many possible variations in the system. For

example, the coupling mounted to the engine flywheel contains
rubber elements in shear. The stiffness properties of the rubber
elements can vary by up to 25% from the published values due
to age hardening or other factors. For example, temperature of
the blocks could affect the first TNF; colder elements will be
stiffer than warmer elements. In addition, the stiffness of the
clutch is affected by the air pressure, which clamps the element
and drum together.

It is interesting to note that the specification for this project

required three independent torsional analyses be performed in
the design stage and that all three analyses had to agree within a
few percent on the TNFs to be accepted. Even after all of that
effort, the measured results were substantially different from
the predicted frequencies. However, several important
upgrades were recommended in the design stage as a result of
the torsional studies, which included using higher strength shaft
materials for some parts.

This was an extremely complicated system with various

operating conditions and wide speed ranges. Therefore, it was
impossible to satisfy the contract specification, which required
that no torsional critical speeds be within the range of 20% Figure 15. Vertical Motor - Pump System.
below the minimum speed to 30% above the maximum speed.
This was a much larger SM than the 10% required by API. The first TNF was predicted to be 31.8 Hz with the
Therefore, it was recommended that one pump system be tested torsional analysis. A simple lumped parameter mass-elastic
during commissioning to confirm the resonant speeds and model was used. Later, the first TNF was verified to be 31.9
stresses by measurement. Hz using strain gage telemetry system in the field. This was
nearly perfect agreement with less than 1% error.
Both the engine and VFD motor were tested. Switching
back and forth between these two drivers while operating the To achieve good agreement, it is important to use the
pump was performed smoothly due to the excellent control correct pump impeller inertia. Entrained fluid will add to the
system. Therefore, the system was verified to be acceptable. WR2 value and is commonly referred to as the wet inertia value.
It was possible to have acceptable torsional vibration without Solid modeling can be used to calculate the inertia of the water
satisfying the SM required because of the rubber coupling within the impeller. Charts are available to estimate the extra
mounted on the engine flywheel, which provided damping, and inertia depending upon the impeller type and diameter. If
the smooth VFD with low torque ripple (less than 1.5%). The actual dimensions are unavailable, then a general rule of thumb
gearbox was rated up to 40% dynamic torque. is to assume an additional 25% inertia to account for entrained
water in the impeller.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

Mixed Refrigerant Compressor Train for LNG Project used in the analysis during the design stage. To understand
what may have happened, the torsional mass-elastic model
A torsional analysis of the system was performed by the must be normalized to match the measured field data.
compressor manufacturer. Measurements were later taken.
A comparison is shown in Table 7 below. Parametric studies were performed one at a time to
determine the sensitivity of each component on the first TNF.
Table 7. Calculated and Measured TNFs. In general, the torsional stiffness must be higher or the inertia
must be less than used in the analysis in order for the first TNF
Mode Calculated Measured % Diff to be greater than predicted (59 Hz instead of 55 Hz). Results
1 605 CPM 618 CPM 2.1% are summarized in Table 8.
2 1159 CPM 1320 CPM 12.2%
3 1632 CPM 1638 CPM 0.4% Table 8. Results of Parametric Study.
4 2581 CPM 2268 CPM 13.8% Component Percent of Given Value
Motor Inertia 75%
As shown in the table, the first and third TNFs had good Motor Core Stiffness Insensitive
agreement. However, the second and fourth TNFs had more Motor Ext. Stiffness 125%
than 12% difference. If the torsional mode shapes were
Coupling Stiffness 180%
examined in greater detail, it may have been possible to
Flywheel Inertia 75%
determine, which components had erroneous mass-elastic data.
Compressor Inertia Insensitive
Motor Driven Compressor System Compressor Stiffness Insensitive

The compressor system consisted of the following: As shown, the system is most sensitive to motor and
flywheel inertia as well as motor shaft extension stiffness.
There is some sensitivity to the coupling stiffness and no
Induction motor, rated 400 HP (298 kW) at 595 RPM
sensitivity to motor core stiffness, and mass-elastic properties
Disc pack coupling - hub on motor side and adapter on
of the compressor within a 30% to 300% range of the provided
compressor side values.
Flywheel mounted to compressor crankshaft
Reciprocating compressor with two throws To explain this, the percent strain energy and kinetic
energy for the first torsional mode were calculated. Results
A torsional analysis was performed in the design stage. show that 65% of the strain energy is in the motor shaft
The coupling size was selected based on proper service factor. extension and 30% is in the coupling. Therefore, these are the
The flywheel was sized to tune the first TNF between most sensitive components in terms of stiffness. Results show
compressor orders. For this case, the first TNF was predicted that the motor core and flywheel each have approximately 47%
to be 55 Hz, which would be between 5 and 6 compressor kinetic energy for the first torsional mode. This makes both of
speed. The second TNF was predicted to be above 30 these components the most sensitive from an inertia standpoint.
running speed and therefore not considered a problem.
The flywheel inertia calculation is straight forward and
Because of previous concerns with coupling stiffness should therefore be accurate. The compressor inertia and
values provided by the manufacturer and the criticality of stiffness were shown not to affect the first TNF. So for this
tuning the first TNF between harmonics to avoid high torsional system the uncertainty is most likely with the motor inertia and
vibration, measurements were recommended. A strain gage stiffness. Unfortunately, since the system was installed, there
telemetry system was mounted to the motor shaft extension. was no practical way to verify motor dimensions, weights, etc.
The first TNF was determined to be 59 Hz instead of 55 Hz as
predicted. The coupling stiffness could also affect the results, but
would need to have a larger percentage of error (80% for the
The percent error is approximately 7%, which is not that coupling versus only 25% for the motor). It was unknown if
far from the expected range of 5%. However, the first TNF the coupling manufacturer had actual factory test data to
is now only 1% from the 6 harmonic and electrical frequency substantiate the provided value.
of 60 Hz. This is much less than the API recommended
separation margin of 10%. For this particular system, it would This particular system had a heavy shrink fit on the
not be possible to achieve a 10% separation margin from both coupling hub, small diameter motor shaft, and no keyway. It
5 and 6 compressor harmonics even if the actual TNF had is interesting to note that if the 1/3 rule is omitted, then the
been 55 Hz as predicted. calculated first TNF would match the measured value.
Ignoring the penetration effect would be equivalent to modeling
The compressor system was tested over all load conditions a stepped shaft.
and found to still have acceptable stress levels in the motor
shaft and dynamic torque in the coupling. However, the
safety factor was reduced to less than two, which is normally

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

Example of Error Due to Torsional Stiffness of Coupling a medium level of uncertainty was placed on the 1/3 rule.

Multiple disc pack coupling failures were caused by high Based on this simple assessment, it would seem that when
torsional vibration due to insufficient separation margin from a the coupling spacer is torsionally much softer than the disc
significant compressor harmonic. All of the compressors were pack, uncertainty of the overall coupling stiffness will be
driven by 8000 HP (5966 kW) synchronous motors at 720 relatively low (example shown in Figure 12). However, when
RPM. The same flexible disc couplings were used on all units the spacer is torsionally stiff relative to the disc pack and no
between the motor shaft and compressor flywheel. hub (flanged directly to flywheel therefore no 1/3 rule),
uncertainty could be much higher than normal as shown in
Units that experience failures had a Siemens motor driving Figure 16. In the second example, the disc packs control the
an Ariel KBV/6 compressor. The first TNF was predicted to overall torsional stiffness of the coupling and are the most
be 3203 CPM (53.4 Hz). Measurements showed the actual difficult to predict. Due to the high torsional stiffness of the
frequency to be 58.5 60.5 Hz on the three units tested. larger coupling, it may be difficult for the manufacturer to test.
Analysis error was approximately 9% to 12% and the
separation margin was non-existent.

Other units had an Ideal motor driving the same

compressor. The first TNF was predicted to be 3186 CPM
(53.1 Hz). Measurements showed the actual TNF to be 57 Hz,
with 7% analysis error. The SM was 5% instead of 11%.
However, no coupling failures were reported on these units.

Originally, an independent consultant was requested to

perform a torsional analysis of reciprocating compressor trains.
The flywheel was sized for each motor to tune the TNFs.
However for reciprocating equipment, it is difficult to avoid all
harmonics by the recommended 10% separation margin. For
example, the first five compressor orders are: 12 Hz, 24 Hz,
36 Hz, 48 Hz, and 60 Hz. Perfect tuning between 4 and 5
would be 54 Hz or +/- 11%.
Figure 16. Disc Pack Coupling with Stiff Spacer.
Because there is not a sufficient separation margin from 5
For critical systems that are sensitive to coupling stiffness,
running speed, the dynamic torque in the coupling was much
the value needs to be accurate. Vendors with shop test data
higher than predicted. Measured values exceeded the coupling
will have less uncertainty associated with this value. For large
manufacturers allowable level. This is considered the root
couplings, the foundation in a test facility may not be stiff
cause of the shim pack failures.
enough to apply significant torque and remain relatively rigid.
Others have proposed dynamic testing of couplings to improve
By comparing the two torsional models, it was concluded
accuracy of torsional predictions.
that the most likely cause of these errors was the torsional
stiffness value provided for the coupling, which was common
Variation Among Identical Units
for all units. Both torsional models were normalized by
increasing the coupling stiffness by approximately 50%. A
From the previous example, the compressor crankshafts
simple hand-calculation showed that the stiffness of the
were retrofitted with additional inertia (sometimes called
coupling could be higher and that the shim packs could be the
donuts) to compensate for the higher coupling stiffness. Table 9
controlling factor. It was later discovered that static test data
summarizes the first two measured torsional natural frequencies
was unavailable for the shim packs used for this size coupling.
(TNFs) for multiple units tested. Note there was very little
variation among the units (average value of first TNF was
Uncertainty Associated with Disc Pack Coupling
53.75 Hz +/- 1% and having a standard deviation of 1%).
A typical coupling would consist of: hubs, disc packs and Table 9. Summary of Measured Results
spacer. The coupling spacer should be the easiest to model Unit First TNF Second TNF
(low uncertainty), followed by the hub and shaft penetration, Station C
and the most difficult calculation are the disc packs, which are 1 54.5 Hz 80 Hz
non-linear with load and difficult to model even with finite
2 53.5 Hz 80 Hz
element analysis (high uncertainty).
3 54.0 Hz 80 Hz
Station W
The 1/3 penetration rule is commonly used. Nestorides
(1958) discusses this assumption, but does not provide an 1 53.5 Hz 80 Hz
estimated accuracy. Various geometries such as shrink fit, 2 53.0 Hz 80 Hz
keyed shafts, and tapered shaft could affect results. Therefore 3 54.0 Hz 80 Hz

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Table 10. Calculated Torsional Natural Frequencies
Torsional Original Analysis EDI Torsional Analysis Amplification
Mode 1957 Report w/ Disc Cplg Factor Description
1 -- 2.65 Hz 159 CPM 5 Damper Mode
2 872 CPM 14.6 Hz 875 CPM 27 Twisting Through Quill Shaft
3 2,152 CPM 35.2 Hz 2,112 CPM 10 First Engine Crankshaft Mode
4 3,020 CPM 49.7 Hz 2,981 CPM 30 Quill Shaft, Gears, Compressor

The target for the first TNF was 54 Hz, which is half-way With the engine damper functioning properly, operating
between 4 (48 Hz) and 5 (60 Hz) compressor harmonics. throughout the entire speed range does not pose a problem for
The second TNF was measured at 80 Hz, which is 6.7 the system; therefore, it is very important to check the viscous
compressor speed. The separation margins of 10% for the dampers on a periodic basis. When the dampers are located
first TNF and 5% for the second TNF are considered inside the engine frames, they can be subjected to elevated
acceptable. By properly tuning the TNFs, the units have been temperatures and sometimes overlooked during routine
running satisfactorily. maintenance (Feese and Hill, 2009).

Engine Gearbox Compressor Measurement vs. Predictions

Torsional measurements were performed on engine

As a summary of the comparison between measured and
gearbox compressor units for a gas pipeline transmission
predicted TNFs, 41 cases are presented in Figure 17. The
station. The engine speed could vary from 430 RPM to 500
X-axis is the order of the TNF, i.e., 1st TNF, 2nd TNF, etc. The
RPM. The original torsional analysis was performed by
Y-axis is the difference in percentage of the predicted TNFs
Nordberg in 1957 and did not include the separate viscous
versus measured TNFs. Due to measurement difficulties, 2nd
damper ring on the engine. The mass-elastic model was
and 3rd TNFs are sometimes not measured.
modified to include the new couplings and viscous damper.
The first four TNFs are listed in Table 10.
25.00%
EDI performed a field study to investigate the cause for
Difference Between Predictions & Measurements

crankshaft failures that occurred in the Nordberg engines. 20.00%

From the torsional measurements and calculations, it was 15.00%

determined that the crankshaft failures were due to a torsional 10.00%
resonance caused by the third torsional natural frequency
5.00%
(engine crankshaft mode) being excited by the 4.5 engine
order combined with a failed engine damper. 0.00%
0 1 2 3 4
-5.00%
Without the engine damper operating properly, the stress
-10.00%
levels in the engine crankshaft were excessive when operating
near resonance. The torsional natural frequency of concern is -15.00%

the third mode, excited by 4.5 engine speed. -20.00%

-25.00%
Table 11 shows a variation of 33.8 Hz 35.6 Hz for the TNFs
first TNF. The average TNF was 34.9 Hz with a standard
deviation of 1.8% and having total variation of 2% to 3%. Figure 17. Discrepancies between Prediction and Measurement

Table 11. Comparison of TNFs Measured The distribution of the discrepancies is summarized in
Table 12, where for instance, 90% of all data points for the first
Unit Frequency (Hz) Speed (RPM)
TNF are within 15%.
Station H
1 34.5 460
2 34.5 460 Table 12. Percentage of Discrepancies
3 35.3 470
4 33.8 450 TNF within within within within within
Station D 25% 20% 15% 10% 5%
1 35.6 475
1st 100% 97.5% 90% 70% 30%
2 35.2 469
nd
3 35.3 471 2 100% 100% 93% 87% 53%

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

CASE HISTORIES VFD TORSIONAL PROBLEMS

To avoid torsional problems, API requires a 10% SM.

There are cases with wide operating speed ranges where this is
impractical. Sometimes special couplings are needed or skip
frequencies must be programmed into the VFD. There are
still other cases where the equipment trains had adequate
separation margins, but still suffered torsional failures.

Case History 1: Torsional Vibration Problem with VFD

Motor / ID Fan at an Oil Refinery

An induced draft (ID) fan system at an oil refinery

experienced a failure of the spool piece in the flexible disc
coupling between the motor and fan shafts. The fan system is
part of an atmospheric furnace that heats approximately
152,000 barrels of crude oil per day. The ID fan is driven by a
350 HP induction motor (Figure 18). To improve efficiency, Figure 20. Cracked Coupling (Center Spacer).
the fan speed is varied instead of using inlet dampers to control
the exhaust flow from the furnace. The motor speed is The failure of the flexible disc coupling consisted of
controlled by a low-voltage variable frequency drive (VFD) cracked space piece, which appeared to originate at a bolt hole
from 0 to 1200 RPM. Any unscheduled downtime would be (Figure 19). Initially, the plant maintenance was blamed for
costly and could quickly outweigh the energy savings from possibly over tightening the coupling bolts. However, the 45
using the VFD instead of inlet dampers. Therefore, it is degree angle of the crack through the coupling spacer as shown
imperative that the system have high reliability. in Figure 20 is typical of high torsional vibration.

To help diagnose the actual cause of the coupling failure,

field tests were performed. The transmitted torque was
measured using a wireless strain gage telemetry system
mounted on the motor shaft extension near the coupling hub.
See APPENDIX for various methods of measuring torsional
vibration. The waterfall plot in Figure 21 shows the measured
frequency spectra of the alternating torque versus speed. The
first torsional natural frequency (TNF) of the system was
identified at 58 Hz.

Torsional resonances occur when energy at multiples of

mechanical running speed and electrical harmonics from the
VFD intersect a TNF. Because the motor has 6 poles (3 pole
pairs), the mechanical frequency will be approximately 20 Hz
(1200 RPM) when the fundamental VFD frequency is 60 Hz,
Figure 18. ID Fan and Motor. neglecting slip of the induction motor. High dynamic torque
in the coupling was found when operating the fan from 1000 to
1200 RPM due to excitation at the 1 electrical frequency of
the VFD.

VFDs control motor speed by varying the electrical

frequency. In the United States, electrical power is supplied at
60 Hz. The VFD first rectifies the input AC power to the DC
bus, then inverts from DC back to AC power at the required
electrical frequency to drive the motor at the desired speed.
The output frequency from the drive can range from 0 to 60 Hz.
Because the output wave form is no longer a pure sine wave,
torque ripple can be produced. Some newer VFD
technologies such as pulse width modulation (PWM), produce
smoother wave forms and thus reduced torque excitation at
electrical harmonics.

Figure 19. Cracked Coupling (bolt hole).

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Figure 21. Waterfall Plot of Dynamic Torque.

Because the fan normally operates 1000 to 1200 RPM, with the mechanical engineers that this change was being made
which was the speed range where excessive dynamic torque and did not realize how it could impact the mass-elastic
occurred, this is believed to be the reason for the coupling torsional system.
failure. For example, the VFD excitation was approximately
5% of the full load torque (FLT) and at torsional resonance the American Petroleum Institute (API) recommends that a
dynamic torque is amplified by a factor of 30. Therefore, the torsional analysis be performed in the design stage to prevent
maximum alternating torque was approximately 150% of the failures. A separation margin (SM) of at least 10% between
transmitted torque, which exceeded the rating of the coupling. the torsional natural frequencies and the excitation frequencies
is recommended to avoid running at a torsional resonance
Due to the large diameter and weight involved, the inertia unless shown to be safe. Many times satisfying the 10% SM
of the fan is many times greater than the inertia of the motor. is impractical for VFD motor systems that operate over a large
For the first torsional mode, the motor is typically near an speed range.
anti-node and acts like a torsional pendulum. The fan is
usually near the node and acts as an anchor. The VFD infers A torsional analysis of the fan system was never performed
load changes by monitoring motor current, which could also with either motor. After the coupling failure, the motor inertia
contain variation from the first TNF. In a torsionally stiff, values were compared. It was found that the replacement
lightly damped system, the first TNF is very sensitive to any motor had a much lower inertia (WR2) value than the original
harmonic excitation or sudden speed adjustments from the VFD motor. Reducing the motor inertia caused the first TNF of the
motor (Feese and Maxfield, 2008). system, which was originally below the minimum speed, to
increase into the normal operating speed range.
After further discussion with plant personnel, it was
determined that the fan was originally driven by another motor Since switching back to the original motor was not an
from a different manufacturer. Motor repairs were needed and option; a temporary solution was recommended where the
would take longer than acceptable. Therefore, an alternate running speed should be limited to a maximum motor speed of
motor was acquired and installed. The new motor from a 1000 RPM (VFD frequency of 50 Hz) to avoid exciting the first
different manufacturer was similar in electrical performance, TNF at 58 Hz. This provided a SM of 13% between the VFD
but was vastly different in physical size and inertia. excitation frequency and the first TNF of the system.
Unfortunately, the electrical engineers did not communicate

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 A torsional analysis of the system was performed and VFD Case History 2 - VFD Motor Driving a High-Inertia Fan
normalized to match the measured field data. Based on the System with Sufficient Separation Margin
results of the computer analysis, an alternate coupling was
selected to detune the TNF away from the 1 electrical VFD motors are commonly used to drive induced draft
frequency of the VFD. A coupling with elastomer blocks in (ID) fans and mechanical vapor recompression (MVR) fans.
compression (Figure 22) generally has a lower torsional As previously discussed, API recommends a 10% separation
stiffness than a flexible disc coupling and provides additional margin from torsional resonance. However, VFDs are used to
damping. The damping reduces the dynamic torque when achieve wide speed range or to operate the motor above
operating near resonance. Elastomeric couplings in synchronous frequency. Older VFDs had significant torque
compression are often used on large VFD motor / fan systems excitation at multiples of electrical frequency (referred to as
found at power plants. torque ripple). Avoidance using skip frequencies and/or
rubber element couplings was often required to obtain a reliable
system.

Newer pulse-width-modulation (PWM) drives are much

smoother and can often operate at torsional resonance without
creating high response. A general rule of thumb would be that
drives which produce 1% torque ripple or less are considered
smooth. Although electrical harmonics have been greatly
reduced by these modern drives due to more sinusoidal wave
forces with less distortion, increased complexity in the control
system has introduced other problems. Trying to control
speed to a very precise value through virtual feedback loops has
been shown to cause instability in several models of
low-voltage and medium-voltage drives.
Figure 22. Flexible Coupling [Holset Catalog].
The problem occurs because the high-inertia fan acts like a
The torsional stiffness of the rubber blocks is non-linear
flywheel that does not suddenly change speeds. In fact,
and sensitive to shore durometer (hardness) of the rubber.
flywheels are commonly used on reciprocating equipment to
Therefore when using a rubber block coupling, it is important
smooth the dynamic torque in the system. Because of the
to compute the TNFs using various rubber durometers (SM60,
large diameter size and weight involved (see reference Feese
SM70, SM80) over the entire operating range. The
and Maxfield, 2008 for details), the inertia of the fan could be
interference or Campbell diagram shown in Figure 23
more than 20 the inertia of the motor. For the first torsional
illustrates how the first TNF varies with speed / load for various
mode, the motor core is typically near an anti-node and acts
rubber durometers. With SM60 blocks, the torsional
like a torsional pendulum. The fan on the hand is usually near
resonance was predicted well below the normal operating speed
the node and acts as an anchor. The drive calculates load
range. Fortunately, a suitable coupling with proper size and
changes by monitoring motor current, which could also contain
durometer blocks was located with a short delivery time.
variation from the first TNF. In torsionally stiff, lightly
damped systems, this effect can be more pronounced.
The coupling was installed and the fan system has been
operating satisfactorily for five years. This case study shows
The instability can be worse at low motor current (low
the importance of performing a torsional analysis on a new
load). It is believed this is because any disturbance from the
system in the design stage and whenever the system is
TNF is misinterpreted as load change. The drive then quickly
modified. Variables in the system include motor inertia and
over-reacts, thus feeding the torsional amplitude. When the
coupling torsional stiffness. The inertia was significantly
exhaust air from the boiler is hot, the air density will decrease,
different between the old and new motor models.
and so will the fan load of an ID fan.
The VFD excitation was slightly higher than reported by
In the case study from Alexander, et al., (2010), the
the manufacturer. For a smooth VFD producing only 1%
instability occurred at a VFD operating frequency 20% - 30%
torque ripple, this rubber element coupling may not have been
above the first TNF, which was well outside the API separation
required. To achieve a reliable design, adequate safety factors
margin of concern; therefore, a steady-state analysis performed
must be considered to account for possible variation in the
according to API would not predict or prevent this type of
supplied information.
torsional problem. Therefore, the VFD must be properly
tuned during commissioning.
This case history demonstrates how important it is to
perform a torsional analysis in the design stage and to have
proper separation margins or to show that operating at
resonance will not damage the machinery. When changes are
made to the train, the torsional analysis should be revised to
ensure these changes do not negatively impact the reliability.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Figure 23. Campbell Diagram.

Case History 3: VFD Motor / High-Speed Compressor at

Chemical Plant with Sufficient Separation Margin

Because equivalent inertia is related to speed ratio squared,

a high-speed centrifugal compressor with a speed increasing
gearbox could also behave similarly to a VFD motor / fan
system. A steady-state torsional analysis was performed in
the design stage. Results showed that the TNFs should be
sufficiently away from the exciting frequencies with the
exception of 50 Hz electrical line frequency. However, the
plant still experienced three failures of the low speed coupling
between the motor and gear in the compressor train:

First failure after 121 days operation.

Second failure after 319 days operation.
Third failure after 207 days operation. Figure 24. Failed Diaphragm Coupling.

During the first event, the compressor unit tripped on high In general, induction motors, gearboxes, and centrifugal
vibration due to a failure of the low speed coupling. A picture compressors do not normally produce significant torsional
of the failed diaphragm coupling is shown in Figure 24. The excitation. For these types of rotating systems, the transmitted
coupling manufacturer concluded that the failure was due to torque should be smooth (less than 10% dynamic torque);
torsional fatigue (Corcoran, et al., 2010). Furthermore, the however, as noted in several reference papers, VFDs have been
manufacturer confirmed that the coupling was properly sized shown to create significant torsional excitation in systems and
for this application, speed, and load. Metallurgical study caused torsional vibration and failures of the system.
indicated fatigue cracks. All of these coupling failures
exhibited characteristics typical of high torsional vibration. Various tests were performed by sweeping the speed of the
VFD motor. The first TNF was measured at 13.5 Hz. This
compared to the predicted TNF of 12 Hz or 720 CPM;
therefore, the actual first TNF was 12.5% higher than predicted.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

One area of the mass-elastic model that could be improved was The motor speed is not accurately controlled in the Scalar
the motor core. The motor should be modeled using multiple Mode, but this was not required for the process. Scalar Mode
inertia and stiffness values instead of a single lump. is used in the factory to test motors, but could not be used in the
field because of loss of protections. During one test run, it
Some load was applied to the compressor, but since the was found that the current spiked, and the motor had to be
plant was not in operation, a true full-load evaluation was not tripped. It was later concluded that the VFD may have run out
possible. It was found that the first TNF was continually of voltage indicating a possible problem with the line power
excited by the VFD when in Direct Torque Control Mode as supply at the plant.
shown in Figure 25. Based on the modeshapes, it is believed
that the first TNF and not the second TNF was responsible for One parameter in the VFD control system, Torque Limit,
the coupling failures. was found to suppress the power fluctuations when activated.
When the Torque Limit was reached, the torque fluctuation was
The action of the drive control system is to continuously reduced. It was later found that the torque limiter is contained
adjust the power to precisely maintain the speed of the motor. in the speed control loop. Therefore, as a short-term solution,
These adjustments can excite the first TNF and put high the Torque Limit was manually lowered just above the plant
stresses on the low speed coupling. It was shown that the operating conditions. With this modification to the VFD
control system was very sensitive to the time delay parameter. controls, the unit has been operating for nearly two years
When the time constant was changed to 50 milliseconds, the without another coupling failure.
dynamic torque increased dramatically, although the system
was operating well away from the first TNF. In conclusion, it is believed that many of these VFD
torsional problems could be avoided by using a less
When switched to the basic control method of Scalar Mode sophisticated control scheme that would allow for a small
without any virtual feedback loops, the excitation was greatly amount of speed variation. Precise speed control from the
reduced as shown in Figure 26. Note that nothing in the VFD is unnecessary for fans and compressors in most plants.
mechanical system was changed, only the operating mode of Also, sufficient time for fine tuning the VFD should be
the VFD was switched. allocated during commissioning.

Figure 25. Waterfall Plot of Dynamic Torque During Speed Sweep.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

Figure 26. Comparison of Various VFD Operating Modes.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

DISCUSSIONS AND CONCLUSIONS This difference in opinion of how API should be
interpreted causes particular difficulties for strings with wide
This paper has examined torsional prediction accuracy and operating speed ranges having numerous potential excitations,
presented the following: many of which may not be of real concern. The safety and
reliability of a torsional system is determined by stress, not
Torsional natural frequency error, comparing merely by separation margin from excitation sources that may
measured and predicted TNFs for numerous cases. or may not be of any importance.
Torsional natural frequency repeatability, comparing
measured and predicted TNFs for multiple strings of API allows the SM requirement to be waived if torsional
the same design. vibration response is shown to not damage the system when
Major sources of uncertainty in torsional modeling. running at resonance. For example, one of the cases described
Sensitivity to variation in mass-elastic data. wherein the predicted SM of 11% passed API, but the measured
Methods for conducting torsional tests. SM of 6% was interpreted to be a violation. The stress
analysis indicated that the system was acceptable. For another
Multiple torsional systems were examined in detail with example, the results of a torsional analysis may show that SM
regard to torsional natural frequency error. For the first is not required for a modern VFD with low torque ripple (1% or
natural frequency, the standard deviation between the measured less). This would therefore allow continuous operation over a
and predicted TNFs was approximately 5.5% with worst case wide speed range without skip frequencies.
errors of -12% and +7%. This is outside of the assumed
+/-5% error previously cited by Mondy and Mirro (1982), and In contrast with SM requirements, recent torsional failures
assumed by many rotordynamicists. involving VFDs that have been documented in the literature
each had sufficient separation margin from running speeds and
Uncertainty in coupling stiffness was shown to be a electrical harmonic frequencies, but still failed due to high
primary contributor to the TNF error. Other contributors to non-synchronous excitation from the VFD. A case was
uncertainty include the following: presented where extremely high torsional vibration occurred
due to instability when operating more than 20% to 30% above
the first TNF. Such instability issues with VFD motors seem
Coupling Model
to occur more often in high inertia systems (i.e. fans) with
Stepped Shaft Model
diaphragm or disc pack couplings that have very low damping.
Motor Core Model
Gear Mesh Model Such results indicate that a torsional test is valuable not
Crankshaft Model only for verification of separation margin, but also verification
Proper Consideration of the 1/3 Penetration Rule of low stress amplitudes and low levels of excitation. In
certain cases, verification of separation margins could be
Two torsional system designs were examined for considered of secondary importance, and validation of the
repeatability. The first system had a design comprised of a amplitudes the primary factor in determining acceptance.
motor and reciprocating compressor. Measurements were Additional benefits of torsional testing can include assistance
made on six strings of equipment. The standard deviation of with proper tuning of the drive during commissioning along
the first TNF was 1%. A second system had a design with failure prevention and troubleshooting.
comprised of an engine, gearbox, and centrifugal compressor.
Measurements were made on seven strings of equipment. The Torsional analytical accuracy is only as good as the
standard deviation of the first TNF was 1.8% for these strings. accuracy of the data, the programs, and procedures available.
The variation of TNF for identical units is therefore expected The design of reliable systems on the other hand is also
to be 1% to 2%. dependent on design standards. The API standard
requirements provide conservatism in most cases due to the
Results indicate that TNF prediction accuracy better than 10% analytical SM requirement as evidenced by the overall
5% should not be assumed. In addition, the results would lack of torsional failures in API machinery strings where SM
indicate that TNF error in excess of 5% is not particularly from running speed was determined to be the root cause.
uncommon. This would suggest that either separation margins
higher than 10%, or uncertainty modeling of components In those cases where additional conservatism is required,
should be considered when performing the torsional analysis. further system verification can be performed in the form of
parametric analysis (considering the errors and uncertainties
At a minimum, these results indicate that the API identified within this paper), stress analysis, torsional
specification margin could use some clarification with regard to measurement, or combination thereof.
its intent for torsional prediction and measurement. Many
rotordynamicists assume the intent is 10% predicted separation
margin with an uncertainty of 5%. Many engineering
procurement contractors, on the other hand, believe the intent is
10% actual separation margin including coupling reselection
and remanufacture if the margin is not met.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

NOMENCLATURE APPENDIX TORSIONAL MEASUREMENTS

AF = Amplification Factor Strain Gage Telemetry System

AGMA = American Gear Manufacturers Association
API = American Petroleum Institute A strain gage telemetry system can be used to directly
CPM = Cycles Per Minute measure transmitted and dynamic torque in a shaft or coupling
FEA = Finite Element Analysis spool piece. The full bridge arrangement with four gages
G = System Gain shown in Figure A.1 can measure torsional strain while
GD2 = Mass Moment of Inertia Based on Diameter minimizing the effects of bending strain, axial strain, and
Hz = Hertz (Cycles Per Second) temperature changes. The voltage signal produced by the
ID = Induced Draft (Fan) bridge is proportional to strain, but can also be calibrated to
J = Lumped Inertia output units of shear stress or torque. The measured values
K = Torsional Stiffness can then be compared to allowable limits.
MVR = Mechanical Vapor Recompression (Fan)
Nc = Torsional critical speed.
N1 = Initial (lesser) speed at 0.707 peak amplitude
of critical speed.
N2 = Final (greater) speed at 0.707 peak amplitude
of critical speed.
PWM = Pulse Width Modulation
RPM = Revolutions Per Minute
SM = Separation Margin
TNF = Torsional Natural Frequency
VFD = Variable Frequency Drive
WR2 = Mass Moment of Inertia Based on Radius
= logarithmic decrement
= frequency in rad/sec
= damping ratio
2, 3, = harmonics or multiples of fundamental frequency

Figure A.1. Full Bridge Arrangement with 4-gages.

Since the strain gages are mounted directly on the rotating

shaft, it is necessary to use a wireless telemetry system to
transfer the measured strain signals to the recording equipment.
Various systems are commercially available. Some require
batteries while others use induced power via a radio frequency
signal that is inductively coupled from the receiving to the
transmitting antenna. The resulting signal is digitized and the
digital data stream is reconstructed into an analog signal at the
receiver. Although resolution and frequency ranges may vary,
typical telemetry systems can have 16-bit resolution and a
0-500 Hz frequency range. Custom systems can have
frequency ranges up to 10,000 Hz.

The strain gages should preferably be located where

maximum twist occurs, which requires advance knowledge of
the torsional mode shapes of the system. Depending on the
particular mode shape, the optimum installation location may
not be feasible. Installing the gages on the shaft near the
coupling hub (but away from any keyways) is usually adequate,
or directly on the coupling spool piece if the dimensions are
known for the inside / outside diameters and the material shear
modulus (Figure A.2). Close-up view of the chevron type
strain gage is shown in Figure A.3. Installation of the gages
and telemetry system requires several hours with the unit down.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 While at rest, the zero and span can be adjusted on the receiver
to obtain a scale factor. Note that the shunt calibration does
not take into account possible deviation from the ideal
strain/torque relationship due to variation in gage factors and
exact gage placement. The shaft diameter should always be
verified.

Normally, using a shunt calibration will result in

amplitudes within 3% of actual values. If more accuracy is
required, such as for measuring efficiency, then a level arm
calibration may be needed (see examples in Feese and Hill,
2009). Therefore, a second mechanical calibration could be
performed where a known moment was applied directly to the
shaft; however, for accurately measuring performance or
efficiency, a mechanical calibration test may be necessary using
a lever arm arrangement. This requires additional time and
advanced planning.

Rotary Shaft Encoder / Gear Teeth / Printed Bar Pattern

A series of pulses can be used to measure the angular

position of a shaft. These pulses could be generated by a
rotary encoder, gear teeth, or a printed bar pattern depending on
the layout of the machinery to be tested.

Rotary encoders are probably the most accurate device for

Figure A.2. Strain Gage Telemetry System Installed on this type of measurement and are normally attached to the
Coupling Spool Piece. free-end of a rotating shaft using an adaptor as shown in Figure
A.4. Setup and use of the encoder is typically simple and
straightforward; however, the shaft must have proper
accommodations for the encoder, such as a tapped hole in the
shaft end.

Figure A.3. Close-up View of Chevron Style Strain Gage.

Calibration of Strain Gage System

Shunt calibration is usually satisfactory for calibration of Figure A.4. Encoder on End of Compressor Crankshaft.
the strain gage system. For this method, it is assumed that the
strain gages have been properly affixed to the shaft surface and An alternate way to install an encoder is shown in Figure
orientated. If so, it is generally thought that the shunt A.5. In this case, a jack shaft from the front of the engine was
calibration will be within 3% of the actual value, which is more driving a cooling fan, which prevented installing the encoder
than adequate for most trouble-shooting field studies. directly to the engine crankshaft; therefore, a special mounting
bracket was designed with a spring to apply a constant force to
By applying a known unbalance in the bridge circuit, the the encoder wheel to prevent slippage relative to the jack shaft.
telemetry system can be calibrated with the shunt resistor. With The measurements were adjusted by the diameter ratio of the
some telemetry systems, the shunt resistor may be built-in. jack shaft and the rider wheel.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Printed bar patterns can also be used as shown in Figure
Rider Encoder A.7. However, it is important that the spacing of the pattern
be even. Overlap at the ends can cause a problem. Also, the
Wheel optical device used to trigger the pulses must be able to react
quickly enough. Not all optical tachometers have high enough
frequency response for this application.

Figure A.5. Encoder with Rider Wheel.

Similar to shaft encoders, the frequency modulation system

uses proximity probes or magnetic pick-ups to measure the
pulse rate or gear tooth passing frequency (Figure A.6).
Assuming the gear teeth are equally spaced and the lateral
vibration is low, variations in time between tooth-passing will
indicate torsional vibration. The signal can be demodulated
and converted to angular velocity or integrated to angular
displacement. Lateral vibration or non-uniform profile could
affect results. Two probes, 180 degrees apart, are preferred to
cancel the effects of lateral vibration.

Figure A.7. Printed Black/White Pattern on Coupling Flange.

Regardless of the method used to generate the pulses,

additional processing is needed to determine rotational speed
and torsional vibration. There are several hardware options
available. These perform satisfactorily for steady-state
conditions, but may create low-frequency noise (ski-slope)
during speed changes due to analog integrators. Software
utilizing the Hilbert transform can also be used to measure
torsional vibration (Randall, 1990). The Hilbert transform
does not require as high of sampling rate and produces good
waterfall plots during speed ramps. This technique can also
be used with two pulse signals and a known torsional stiffness
to measure torque. A comparison between hardware and
software options is shown in Figure A.8 using real test data.

Figure A.6. Proximity Probe Pick-up on Engine Flywheel.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

 Figure A.8. Comparison of Processing Techniques for Torsional Vibration.

Laser Vibrometer

A laser vibrometer can be used to measure angular

displacement in degrees. The laser is non-contacting and can
be pointed at the rotating surface of interest. No downtime
is normally required for installation; however, the laser
vibrometer may not be capable of accurately evaluating
transient conditions, such as start-ups, speed sweeps, or
shutdowns. This could limit its ability to measure natural
frequencies of a system if speed changes occur too quickly.
Newer laser vibrometers have a frequency range down to 0.5
Hz and a dynamic range of 0.01 to 12 degrees.

Laser vibrometer are sometime mounted on a flexible stand

or deck. Excessive vibration of the laser head could affect
torsional vibration readings. Likewise, vibration of the
torsiograph or non-centered application could cause erroneous
readings. Figure A.9 shows a typical laser setup.

Figure A.9. Laser Aimed at Reflective Tape on Jack Shaft.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

Example of Laser Measurement on Engine Unfortunately, HBM torsiographs are no longer
manufactured. Some specifications, particularly from water
A Polytec laser was used to measure torsional vibration on districts and municipalities, may still call for a torsiograph
the front end of a gas engine (Feese and Smith, 2009). As test; however, today shaft encoders and torsional lasers are
shown in Figure A.9, two red dots can be seen from the laser more commonly used to measure torsional vibration instead of
beams on the reflective tape wrapped around the jack shaft. a torsiograph.
The laser is powered by a separate computer module, which
processes the signal, applies filters, and produces a voltage Inductive Torque Measurement
proportional to torsional oscillation in degrees.
A new method for measuring torque is a contactless
inductive sensor that measures the magnetic permeability
Torsiograph variation of a shaft material (Figure A.11). The concept is
based on the anisotropic magnetostrictive effect in
A torsiograph is an instrument that rotates with the shaft ferromagnetic shaft surfaces (Fraunhofer, 2007). The
and is used to measure angular velocity (deg/sec) or magnetic permeability of the shaft material differs in the tensile
displacement (degrees). For example, an HBM torsiograph versus compressive directions and is proportional to the
operates on the seismometer principle, with a mass retained by torsional stress. The sensor is able to measure the
springs whose relative motion compared to the stator is permeability variation at the surface over a wide torque range.
converted into an electrical signal by inductive proximity The sensor requires a narrow air gap (similar to a proximity
detectors (Figure A.10). The frequency range is probe), and typically has a frequency range of 0 200 Hz.
approximately 3 - 1,000 Hz. At the lower frequency range, the
internal masses and springs have a resonance near 3 Hz.

Figure A.10. Torsiograph on Oil Pump End of Compressor.

The device must be mounted on a free end of the shaft,

preferably near an anti-node or point of maximum torsional
oscillation for best results. While the instrument is easy to Figure A.11. Inductive Torque Measurement.
install once an adaptor has been fabricated, it is sensitive to
lateral vibration and will require that the shaft end be true and Inferring Torque from Vibration of Gears
drilled and tapped such that the torsiograph is centered on the
shaft. Downtime of the compressor system will be required Recall that torsional vibration is normally silent except for
for installation. a gearbox. Therefore, it may be possible to infer torque from
vibration measurements on a gearbox. On a recent job
The amount of oscillation may not be an indicator of shaft involving a VFD motor gearbox centrifugal compressor,
stresses. For example, high oscillation can occur in a system strain gage measurements were made near the low speed
with a soft coupling, but the stresses may be low. If only coupling as well as shaft vibration readings from proximity
torsional oscillations are measured, the torsional analysis probes for the LS gear and HS pinion shafts.
should be normalized based on the measurements to evaluate
the stresses in the system. Torsiographs have been used to The most significant dynamic torque occurred at the first
evaluate oil pumps on reciprocating compressors and fan blade TNF of 13 Hz. Transfer functions were calculated for each
failures on a gearbox. The gear pump would be removed and probe. It was found that the best probes were in the Y
an electric oil pump temporarily used for the test. direction for the LS gear and in the X direction for the HS

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

pinion. At the frequency of interest, the coefficients were Stator currents can be measured using a flexible head that
approximately 185,000 to 230,000 in-lb of torque per mil of mounts around the phase cables. The current probes are based
lateral shaft vibration. For example, 0.185 mils p-p of lateral on the Rogowski principle. Figure A.13 shows the installation
shaft vibration indicated 34,000 in-lb p-p of dynamic torque at of the stator current probes around each of the three phases in
13 Hz. Note that the transfer functions would be different for the motor junction box. If DC current values are needed, then
other frequencies and other gearboxes. a Hall Effect transducer must be used.

This approach could be used for simple long-term

monitoring when a torsional device is difficult to install. For
example, the proposed method was sensitive enough to detect
sidebands with 13 Hz spacing in the HS pinion vibration
readings when the dynamic torque reached a high level.
Estimation of shaft torque from shaft vibration was also
discussed by Tanaka, et al., (2009).

Electrical Measurements

It may be necessary to measure voltage and current to help

diagnose the torsional vibration problems with systems
involving electric motors. For example, VFDs can produce
torque ripple at multiples of electrical frequency, which can
excite torsional natural frequencies of the system (Feese and
Maxfield, 2008). Electrical measurements can be used to
determine electrical power and the amount of torque fluctuation
potentially being applied to the motor by the drive. Another Figure A.13. Current Probes.
example is when reciprocating compressors cause current
pulsation in motors that experience high torsional oscillation.
The National Electrical Manufacturers Association (NEMA
2011) specification states that current pulsations should not
exceed 66% in order to prevent flickering of lights and other
electrical problems.

The stator voltages can be measured with voltage dividers,

which have a large frequency response well in excess of
10,000 Hz. Figure A.12 shows three stator voltage probes
(one for each phase) inside a motor cabinet. The output signal
from the voltage probes requires impedance matching (1 MW)
and in some cases amplification.

Figure A.12. Voltage Probes.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

REFERENCES Feese, T.D. and Maxfield, R., 2008, Torsional Vibration
Problem with Motor/ID Fan System Due to PWM Variable
AGMA 9004, Flexible Couplings Mass Elastic Properties Frequency Drive, Proceedings of the Thirty-Seventh
and Other Characteristics, December 2008. Turbomachinery Symposium, Turbomachinery Laboratory,
Texas A&M University, College Station, Texas, pp 45
Alexander, K., Donohue, B., Feese, T., Vanderlinden, G., Kral, 56.
M., Failure Analysis of MVR (Mechanical Vapor
Recompressor) Impeller, Engineering Failure Analysis, Feese, T.D. and Smith, D.R., 2009, Critical Equipment
Volume 17, Issue 6, Elsevier, September 2010. Measured in the Field, Polytec InFocus, Issue 01.

API Standard 617, 1973, Centrifugal Compressors for General Fraunhofer ITWM, A Contactless Torque Sensor for Online
Refinery Services, Third Edition, American Petroleum Monitoring of Torsional Oscillations, 2007.
Institute, Washington, D.C.
Holset Flexible Couplings Catalog, 510/3.90/F.H., Application
API Standard 617, 2002, Axial and Centrifugal Compressor Information Type PM Industrial, Cincinnati, Ohio.
and Expander-Compressors for Petroleum, Chemical, and
Gas Refinery Services, Seventh Edition, American Hudson, J., Feese, T., Torsional Vibration A Segment of
Petroleum Institute, Washington, D.C. API 684, Proceedings of the 35th Turbomachinery
Symposium, Texas A&M University, College Station,
API Standard 671, Special Purpose Couplings for Texas, 2006.
Petroleum, Chemical and Gas Industry Services,
September 2010. Ker Wilson, W., Practical Solution of Torsional Vibration
Problems, Volume 1, New York, New York: John Wiley
API Standard 684, API Standard Paragraphs Rotordynamic and Sons, 3rd Ed., 1956.
Tutorial: Lateral Critical Speeds, Unbalance Response,
Stability, Train Torsionals, and Rotor Balancing, 2nd Kerkman, R.J., Theisen, J. and Shah, Kirti, 2008, PWM
Edition, August 2005. Inverters Producing Torsional Components in AC Motors,
Petroleum and Chemical Industry Technical Conference,
Bosin, D., Ehrich, R., and Stark, M., 1999, Torsional PCIC 2008. 55th IEEE.
Instabilities of Motor Driven Turbomachinery,
Turbomachinery International, pp 18-20. Kocur, J.A. and Muench, M.G., Impact of Electrical Noise on
the Torsional Response of VFD Compressor Trains,
Calistrat, M. M., and Leaseburge, G. G., 1972, Torsional Proceedings of the First Middle East Turbomachinery
Stiffness of Interference Fit Connections, ASME Paper Symposium, 2011.
72-PTG-37.
Meirovitch, L., 1986, Elements of Vibration Analysis, New
Corcoran, Kocur, and Mitsingas, Preventing Undetected Train York, New York: McGraw-Hill Inc.
Torsional Oscillations, Proceedings of the 39th
Turbomachinery Symposium, Houston, Texas 2010. Mondy, R. E. and Mirro, J., 1982, The Calculation and
Verification of Torsional Natural Frequencies for
Dawson, B. and Davies, M., 1975, An Improved Holzer Turbomachinery Equipment Strings, Proceedings of the
Procedure for Torsional Vibration Analysis, The Journal Twelfth Turbomachinery Symposium, Turbomachinery
of Mechanical Engineering Science, 17, 1, pp 26-30. Laboratory, Texas A&M University, College Station,
Texas, pp. 151-156.
De Choudhury, P., Torsional Analysis of Unit 59-K-400A
Comparative Study with Comments, Elliott Company Murray, B.D., Howes, B.C., Zacharias, V., Chui, J.,
Inter-Office Letter, April 3, 1979. Sensitivity of Torsional Analyses to Uncertainty in
System Mass-Elastic Properties, Presented at the
Ewins, D.J., 2000, Modal Testing: Theroy, Practice and International Pipeline Conference, Calgary, Alberta, June
Application, Second Edition, Research Studies Press Ltd. 1996.

Feese, T.D., and Hill, C.H., 2009, Preventing Torsional NEMA Standards Publication MG 1-2011, Motors and
Vibration Problems in Reciprocating Machinery, Generators, Rosslyn, Virginia, 2011.
Proceedings of the Thirty-Eighth Turbomachinery
Symposium, Turbomachinery Laboratory, Texas A&M Nestorides, E.J., A Handbook on Torsional Vibration,
University, College Station, Texas, pp. 213-238. Cambridge, 1958.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University

Pettinato, B. C., Kocur, J. A., and Swanson, E., 2011, ACKNOWLEDGMENTS
Evolution and Trend of API 617 Compressor
Rotordynamic Criteria, Turbomachinery Society of Japan, The authors would like to thank their respective companies
39, 5, pp 36-47. (Elliott Group and EDI) for permission to publish this paper.
In addition, we wish to thank the following specific individuals:
Piergiovanni, L. S., Lerch, E., Zurowski, R., Kumar, S. B., Yusuke Watanabe (EBARA Corporation Fluid Machinery
Osman, R., Deo, B., Bahr, B., Sakaguchi, J., Okazaki, Y., Division), Koichi Hayama (Elliott Group), and Frank Kushner
and Saito, T., 2010, Investigation of Subsynchronous (Elliott Group Consultant) who performed torsional
Torsional Interaction on LNG Power Plants, Proceedings measurements on several of the systems reported within this
of the 16th International Conference & Exhibition on paper. Finally, we would like to thank Charles Hill (EDI),
Liquefied Natrual Gas, Oran, Algeria. Don Smith (EDI / Advisory Board), and our monitor Lisa Ford
(Lufkin / Advisory Board) for their assistance in proof reading
Pollard, E. I., 1967, Torsional Response of Systems, ASME this tutorial.
Journal of Engineering for Power, 89, 3, pp 316-324.

Pollard, E. I., Torsional Critical Speed Test Report: Turbine

Driven Centrifugal Compressors, Elliott Company Inter-
Office Letter, June 15, 1972.

Randall, R.B., Hilbert Transform Techniques for Torsional

Vibration, The Institution of Engineers Australia
Vibration and Noise Conference, 1990.

Steiner, S., 2007, Revisiting Torsional Stiffness, Gas

Machinery Research Council Conference, Dallas, TX.

Sohre, J.C., 1965, Transient Torsional Criticals of

Synchronous Motor-Driven, High-Speed Compressor
Units, ASME Paper No. 65-FE-22.

Tanaka, Adachi, Takahashi, and Fukushima, Torsional-Lateral

Coupled Vibration of Centrifugal Compressor System at
Interharmonic Frequencies Related to Control Loop
Frequencies in Voltage Source Inverter, Proceedings of
38th Turbomachinery Symposium, Case Study,

Vance, J. M., Murphy, B. T., and Tripp, H. A., 1984, Critical

Speeds of Turbomachinery: Computer Predictions vs
Experimental Measurements, Proceedings of the
Thirteenth Turbomachinery Symposium, Turbomachinery
Laboratory, Texas A&M University, College Station,
Texas, pp. 105-130.

Wachel, J. C. and Szenasi, F. R., 1993, Analysis of Torsional

Vibrations in Rotating Machinery, Proceedings of the
Thirteenth Turbomachinery Symposium, Turbomachinery
Laboratory, Texas A&M University, College Station,
Texas, pp. 127-151.

Copyright 2012 by Turbomachinery Laboratory, Texas A&M University