National Conference on Power Transmission, 1975

ISOLATION OF TORSIONAL VIBRATIONS IN ROTATING MACHINERY

F. R. S Z E N A S I , P. E.
Senior Research Engineer
L. E. BLODGETT
Senior Research P h y s i c i s t

Engineering Dynamics Inc.

San Antonio, TX
ABSTRACT
T o r s i o n a l v i b r a t i o n response o f r o t a t i n g machinery must
be determiced when d e s i g n i n g a n equipment t r a i n .
Accur a t e response p r e d i c t i ~ nr s q u i r e s s o p h i s t i c a t s d a n a l y s i s
t e c h n i q u e s which i n c l u d e c o n s i d e r t i o n of a l l f o r c i n g
f u n c t i o n s i n t h e system i n a d d i t i o n t o t h e mass - e l a s t i c
p r o p e r t i e a o f t h e s h a f t s snd e o u p l i s g s . Matrix methods
u t i l i z i n g t h e e i g e n v a l u e s o l u t i o n p r o v i d e an qdequate
s o l u t i o n technique. A proper understanding of t h e
assumptions r e l a t e d t o t h e s o l u t i o n t e c h n i q u e i s essent i a l . Such a r e a s a s lumping, branched system i n t e r p r e t a t i o n , g e a r t o o t h f l e x i b i l i t y , and coupling parameters a r e b u t a few examples where assumptions must be
made which a r e e s s e n t i a l t o a proper s o l u t i o n . I s o l a t i o n
of t c r i o n a l v i b r a t i o n s can be acconqli4hed once t h e
system t o r s i o n a l response i s p r o p e r l y d e f i n e d . Case
h i s t o r i e s u t i l i z i n g t h i s technique a r e presented t o
i l l u s t r a t e methods of t o r s i o n a l i s o l a t i o n .
INTRODUCTION
+

T o r s i o n a l v i b r a t i o n problasts are being encountered i n an

e v e r i n c r e a s i n g numher due t o t h e expanded u s e of high
speed r o t a r y equipment. These problems a r e d i f f i c u l t
t o recognize i n t h e i r primary s t a g e s , and many times
t h e y f i r s t m a n i f e s t themselves when a f a i l u r e occurs.
T y p i c a l l y , t h i s t y p e of f a i l u r e r e s u l t s i n a s u b s t a n t i a l
p e n a l t y i n terms of p l a n t downtime, i n v o l v i n g f i r s t an

a n a l y s i s o f t h e s o u r c e o f t h e problem, and t h e n t h e
d e f i n i t i o n and implementation o f a s u i t a b l e t r e a t m e n t .
To diagnose and minimize t h e o c c u r r e n c e o f p o s s i b l e
f a i l u r e s due t o t o r s i o n a l v i b r a t i o n , s e v e r a l t y p e s of
a n a l y s i s may be employed.
The o b j e c t of t h i s paper i s t o p r e s e n t t h e most p r a c t i c a l and a c c u r a t e method f o r c a l c u l a t i n g b o t h t h e t o r s i o n a l r e s o n a n t frequency and t h e f o r c e d v i b r a t i o n response. With such a n a l y s e s , stress l e v e l s c a n be c a l c u l a t e d and compared t o a f a i l u r e c r i t e r i o n .
I n addit i o n , t h i s gaper p r e s e n t s one method which has been
s u c c e s s f u l l y used t o e v a l u a t e t h e e f f e c t of g e a r s t i f f n e s s on t o r s i o n a l response. Although g e a r s t i f f n e s s
is o f t e n n e g l e c t e d o r improperly s i m u l a t e d i n some ins t a n c e s , f i e l d e x p e r i e n c e shows its e f f e c t a r e o f t e n
critical.
Case h i s t o r i e s a r e i n c l u d e d t o g i v e example d a t a and t o
i l l u s t r a t e how a f o r c e d v i b r a t i o n a n a l y s i s c a n be used
t o p r e d i c t equipment r e l i a b i l i t y o r t o a i d i n f a i l u r e
solution.
HOLZER ANALYSIS

The method most commonly employed t o c a l c u l a t e t h e t o r s i o n a l r e s o n a n t frequency of s h a f t i n g systems i s t h e

Holzer Analysis, c h i e f l y because it i s c o n v e n i e n t l y
a d a p t a b l e t o hand c a l c u l a t i o n s . The method r e q u i r e s
s u c c e s s i v e e s t i m a t e s of t h e s h a f t r e s o n a n t frequency a s
i n p u t , and it i s p o s s i b l e t o overlook s o l u t i o n s i f t h e
s e l e c t e d frequency increment i s t o o l a r g e . The Holzer
t a b u l a t i o n method i s u s e f u l f o r a c u r s o r y check of most
s h a f t systems; however, it becomes q u i t e cumbersome and
t e d i o u s when a f o r c e d v i b r a t i o n response of complex syst e m s i s r e q u i r e d . Forced v i b r a t i o n problems c a n be made
with t h e Holzer A n a l y s i s ; however, t h i s method d o e s n o t
account f c r t h e phese o f t h e f o r c i n g t o r q u e s i n a camp l e x f o r c i n g f u n c t i o n ; t h e r e f o r , a n error c a n e x i s t .
This e r r o r i s s l i g h t f o r r e l a t i v e l y c c n s t a n t t o r q u e
machines; however, i n r e c i p r c c a t i n g compressors and int e r n a l combustian e z g i n e s , t h e t o r q u e i s d e c i d e d l y
unsteady d a e t o p i s t o n s t r o k e , and s u b s t a n t i a l errors
can t h e r e f o r e be i n c u r r e d .
Another l i m i t a t i o n c f t h e Holzer W - a l y s i s i s t h e i n a b i l i t y t o p r o p e r l y s i m u l a t e m u l t i p l e branced geared
systems. The dynamically e q u i v a l e n t system presupposes
t h a t t h e g e a r assembly i s t o r s i o n a l l y r i g i d , which i s

approximately the case in larse industrial gear boxes

with generous factors of safety. For less conservatively designed systems, gear tooth flexibility has a definite influence on the torsional frequencies and should
be considered.
EIGENVECTOR

EIGENVALUE MATRIX SOLUTION

A more powerful and versatile method for solving torsional

resonant frequencies is the eigenvector-eigenvalue matrix
solution. While this method insures the calculation of
all the possible modes 05 vibration, it requires the use
of a digital computer as the computations are numerous
and complex. To calculate the tarsional resonant frequencies of a system, a mathematical model must be synthesized, which will respond in the same manner as the
actual system. All of the elastic, mass and damping
properties of the system are necessary to set up the
mathematical model. Usually, these elastic properties
and the masss inertia can be calculated, measured or obtained from the manufacturer of the element.

Sample differential equations of motion have been written

and &re included in Table I. These differential equations can be converted into a matrix equation for simplicity of solution, and the general form of the equation
would be:

where [Jl is the diagonalized mass matrix and [K] is the

stiffness matrix. These matrixes are shown in Table 11.
Complex periodic motion may be reduced to individual
harmonics which can be handled easier without compromising the rigorousness of the solution. The equation
for simple harmonic motion can then be assumed for the
general solution.

0 = A sin at

The following relationship for

'6 can be obtained by differentiation

8 = Aw c o s w t

3 =

- ~ w 2sin o t

-4

-0

By substitution, the matrix equation can be rewritten

where the [ w

] represents the diagonalized eigenvalue matrix which will

be called [I
] This yields:

By multiplying both sides of the matrix equation by

[ J] l , the following

equation results:

and

This form of the matrix is the eigenvalue equation. The

values for [ A ] for which the equation is soluble are
known as the characteristic values, or eigenvalues, of
the matrix. The vector so utions for f e ] are the eigenvectors of the matrix [JI-' IK1 which shall be referred
to as the stiffness-mass natrix.
Physically, the eigenvector represents the mode shape of
the vibration, The corresponding eigenvalue represents
the vibrational frequency squared. In general, the
characteristic equation will have "n u roots with "nu
corresponding eigenvectors for a system with nnn rotating
masses.

The s o l u t i o n f o r t h e e i g e n v a l u e s of a problem w i t h f i v e
masses w i l l be a f i f t h o r d e r e q u a t i o n whose s o l u t i o n w i l l
g i v e f i v e r o o t s of t h e c h a r a c t e r i s t i c equation. Each of
t h e s e f i v e r o o t s w i l l t h e n r e p r e s e n t a r e s o n a n t frequency
squared. The c h a r a c t e r i s t i c e q u a t i o n of t h i s example i s
a s follows:

(where A, B. C , D , E, and F r e p r e s e n t complex f u n c t i o n s

of mass and s t i f f n e s s .
The r o o t s of t h i s e q u a t i o n a r e
t h e e i g e n v a l u e s from t h e e i g e n v a l u e problem).
The e i g e n v e c t o r s a r e o b t a i n e d from t h e o r i g i n a l e i g e n v a l u e
e q u a t i o s . A set of e q u a t i o n s c a n be d e r i v e d by u s i n g t h e
s t i f f n e s s - m a s s m a t r i x , m u l t i p l y i n g by an unknown eigenv e c t o r , and f o r c i n g t h i s t o e q u a l t o t h e product of t h e
eigenvalue and t h e unknokn e i g e n v e c t o r . T h i s s e t 3f
e q u a t i o n s can be solved f o r e i g e n v e c t o r c , which r e p r e s e n t
t h e v i b r a t i o n mode shape corresponding t o t h e eigenvalue
o r r e s o n a n t frequency.
The e i g e n v e c t o r method of s o l v i n g f o r t h e t o r s i o l ~ a lresonant frequency and mode shapes e n a b l e s t h e c a l c u l a t i o n
of t h e f o r c e d v i b r a t i o n a l r e s p o n s e of t h e system due t o
various forcing functions a t d i f f e r e n t m a s s lociititns,
i n c l u d i n g t h e phasing of a l l f o r c e s . The damping of t h e
system must a l s o b e i n c l u d e d t o i n s u r e proper r e s o n a n t
f r e q u e n c i e s . An harmonic F o u r i e r expansion of t h e f o r c i n g f u n c t i o n can be a p p l i e d a t any mass l o c a t i o n .
In
t h i s manner, a complex f o r c i n g f u n c t i o n w i t h an a r b i t r a r y
number of harmonics can b e simulated.
Systems c o n t a i n i n g s e v e r a l p i n i o n s d r i v e n by one g e a r
which cannot be s i m u l a t e d r e a d i l y by simple hand c a l c u l a t i o n methods c a n b e s o l v e d by t h e m a t r i x method. The
s t i f f n e s s m a t r i x can b e modified by a d d i t i o n a l o f f
d i a g o n a l terms which i n f l u e n c e t h e d e f l e c t i o n of t h e branch
p o i n t mass. These a d d i t i o n a l t e r m s a r e b e s t determined
by w r i t i n g t h e t o r s i o n a l e q u a t i o n s of motion i n a s y s t e matic method. The m a t r i x e q u a t i o n s c o n t a i n i n g t h e s t i f f n e s s m a t r i x f o r a branched g e a r system w i l l be solved i n
t h e same manner.
Once t h e system h a s been modeled p r o p e r l y and f o r c i n g
f u n c t i o n s a p p l i e d , t h e a m p l i t u d e s a t each mass l o c a t i o n
a r e known. Since r e l a t i v e d e f l e c t i o n between t h e masses
determines t h e stresses i n t h e s h a f t , stresses c a n be
c a l c u l a t e d f o r each harmonic, and w i t h proper phasing,
t h e complex peak-to-peak stress wave c a n be generated
which a l l o w s f o r t h e c a l c u l a t i o n of t h e maximum o v e r a l l
peak-to-peak stress.
266

GEAR TORSIONAL STIFFNESS

Theory
In many geared systems the effect of gear stiffness can
exhibit a critical influence on calculating the torsional
natural frequencies of the system. The effect of gear
stiffness is usually increasingly important as the dynamic load ncreases above the normal transmitted static
torque load , particularly when the gear is located at a
dynamic torque maximum in the torsional mode shape.

Gear stiffness can only be approximated, since there are

several parameters varying during rotation. These include:
Instantaneous point of contact
Direction of applied force
Number of teeth in contact
The following discussion will be limited to spur and
bevel gears, since they represent a large majority of
cases where gear stiffness is critical. The tooth profile is approximated by an isoceles triangle, and Figure
1 illustrates the approximation and notation used. The
linear flexibility is calculated from the equation2.

This value must be calculated for each gearwheel and a

correction factor must be applied to account for the
following second order effects:

(a)

D e p r e s s i o n of t h e t o o t h s u r f a c e at t h e l i n e of contact.

(b)

F l e x i b i l i t y of w h e e l body a d j a c e n t t o t h e tooth.

(c)

D e f o r m a t i o n of wheel body.

E x p e r i e n c e h a s shown t h e following c o r r e c t i o n f a c t o r s a r e applicable:

(a)

R = 1 . 3 f o r plain s p u r gears.

(b)

R = 1.25 for bevel gears.

(c)

R = 1.0 for internal spur gears.

Using t h e s e v a l u e s , t h e l i n e a r flexibility a s r e l a t e d t o e a c h g e a r w h e e l b e comes:

The torsional stiffness of the gear system must be

to the torque on the pinion or bullgear. Assuming
the load is shared equally between the two teeth,
following expression can relate the flexibility to
appropriate torque.

related
that
the
the

The torque related stiffness is in reference to the gear

wheel with radius r. This stiffness can thereby be applied as a stiffness between the two gear wheel mass inertias. The following sample analysis indicates the influence of gear stiffness upon the torsional frequency.
FIELD CASE

In a recent startup at a chemical processing plant there

were repeated cooling fan gear box failures. The failures
occurred after various lengths of operation ranging from
3-48 hours. A torsiograph was installed on one of the
fan systems, and the unit was started and tripped several

times. The data identified a torsional resonant frequency

at approximately 23 Hz. This resonant frequency was found
to be extremely close to the blade-passing frequency of
the fan system (24 Hz.), indicating the possibility that
the torsional resonance excited by the blade-passing perturbations was a contributing cause of the shaft failure.
In an effort to solve the problem, the system was simulated on the eigenvector-eigenvalue computer program. The
gear stiffness was computed as outlined, and the calculated torsional resonant frequency was in good agreement with
the field data. To demonstrate the system sensitivity to
gear stiffness, a parametric analysis was made by varying
the gear stiffness. This data is presented in Table 111.
It should be noted that the gear stiffness is important
in the design stage where coincidence of resonant frequency and excitation is to be avoided.
RECIPROCATING ENGINES
Theory
comprehensive torsional analysis of a reciprocating
engine driven system must properly simulate the masselastic properties of the crankshaft and the force produced by the power cylinders. Simulation of the engine
crankshaft stiffnesses and masses, although tedious, can
be accomplished with careful calculations lr2
The
forced vibration response, however, is less straightforward since it must include cylinder phasing (firing
order), gas torques (harmonic content), and reciprocating
mass inertia in combination with the dynamic response
(mode shape). Gas torque curves are usually available
for new engines in harmonic content form which can be
directly used in the analysis3
A

The harmonic torque amplitudes for each cylinder are

applied to the rotating mass at each crank thrown by
expressing the torque function as a series of sine and
cosine terms including phase angles. In this form, the
forcing functions are easily used in the eigenvector
matrix method which intrinsically combines the forcing
function with the dynamic response to produce the vibration amplitudes and stresses.
By utilizing the matrix generated from the equation of
motion, in combination with the applied forces, the
torsional deflections can be calculated. A damping term
must be included to calculate a dynamic magnification
factor ( Q ) which provides a relationship between vibraThe
tional amplitude and frequency ration

v i b r a t i o n a l amplitude can be simply expressed by t h e

following expression:

This technique has been used t o determine r e l i a b i l i t y of

r e c i p r o c a t i n g engine d r i v e n systems i n t h e d e s i g n s t a g e
a s well a s i n o p e r a t i n g systems. The following operating u n i t , plagued by s h a f t f a i l u r e s , was analyzed by t h i s
technique t o determine t h e cause o f f a i l u r e s .
F i e l d Case
p l a n t a i r compressor system experienced s e v e r a l s h a f t
f a i l u r e s because of o p e r a t i n g near system t o r s i o n a l
n a t u r a l frequencies. The system c o n s i s t s of a fourc y l i n e r a i r compressor d r i v e n by a 16-cylinder r e c i p r o c a t i n g engine through a gear box. The mass- elastic
p r o p e r t i e s of t h e system were c a l c u l a t e d , and t h e t o r s i o n a l system was simulated w i t h t h e eigenvector-eigenvalue method. The c a l c u l a t e d t o r s i o n a l n a t u r a l frequenc i e s of i n t e r e s t a r e a s follows:
A

f l = 2688 cpm
f2 = 4212 cpm
f 3 = 6180 cpm
An i n i t i a l a n a l y s i s was made t o study t h e e f f e c t of t h e

engine torques on t h e system. The forced v i b r a t i o n

a n a l y s i s included both t h e engine and compressor loading
torques which w e r e obtained from t h e equipment manufact u r e r s.
The maximum stresses f o r t h i s system occurred i n t h e
p i n i o n s h a f t and a r e p l o t t e d i n Figure 2 f o r t h e engine
harmonics and t h e second, t h i r d and f o u r t h compressor
harmonics i n t h e speed range of t h e u n i t . The f o u r t h
compressor harmonic e x c i t e s t h e f i r s t t o r s i o n a l n a t u r a l
frequency a t an engine speed of 1206 cpm. The t i r d harmonic e x c i t e s t h e f i r s t t o r s i o n a l n a t u r a l frequency a t
an engine speed of 1608 cpm. The second compressor
harmonic is on t h e f l a n k of t h e reasonance curve, and
t h e s e stresses i n c r e a s e from 1700 t o 3000 p s i peak-topeak. The combined maximum s t r e s s e s w e r e caused by

compressor l o a d i n g t o r q u e s a t 1608 cpm and w e r e s l i g h t l y

i n e x c e s s of 14800 p s i peak-to-peak which exceeds t h e
U. S. MIL STD* c r i t e r i a f o r t o r s i o n a l stress l e v e l s . The
maximum stresses produced by t h e e n g i n e w e r e 10,000 p s i
peak-to-peak when t h e f o u r t h o r d e r r e a c t e d w i t h t h e
second t o r s i o n a l n a t u r a l frequency.
The Campbell diagram o r i n t e r f e r e n c e diagram shown i n
F i g u r e 3 i n d i c a t e s which e n g i n e and compressor harmonics
e x c i t e t h e t o r s i o n a l n a t u r a l f r e q u e n c i e s through t h e
speed range. The second e n g i n e o r d e r and t h e t h i r d and
f o u r t h compressor harmonics e x c i t e t h e f i r s t t o r s i o n a l
n a t u r a l frequency ( F i g u r e 2 ) . These produce t h e major
The
c o n t r i b u t i n g stresses which a c t on t h e system.
second t o r s i o n a l n a t u r a l frequency (4212 cpm), however,
can be e x c i t e d by t h e f i f t h , s i x t h , and seventh comp r e s s o r harmonics a s w e l l a s t h e t h i r d and f o u r t h engine
o r d e r s . The e f f e c t of t h e h i g h e r harmonics e x c i t i n g t h e
second t o r s i o n a l n a t u r a l frequency s e r v e s t o i n c r e a s e
The t h i r d t o r t h e stresses of t h e p i n i o n s h a f t a l s o .
s i o n a l n a t u r a l frequency (6180 cpm) c a n be e x c i t e d by t h e
seventh through e l e v e n t h compressor harmonics a s w e l l a s
t h e f o u r t h through s i x t h e n g i n e o r d e r s .
These h i g h e r
harmonic t o r q u e s w i l l a l s o c o n t r i b u t e a d d i t i o n a l stress;
however, t h e s e harmonics produce a s m a l l percentage of
energy compared t o t h e f i r s t few harmonics and conseq u e n t l y do n o t c a u s e a s i g n i f i c a n t i n c r e a s e i n stress
level.
DISCUSSION AND CONCLUSION
Proper d e s i g n a n a l y s i s should b e implemented on r o t a r y
equipment t o i n s u r e i n c r e a s e d r e l i a b i l i t y and minimal
downtime due t o e x c e s s i v e maintenance o r f a i l u r e s . One
a r e a of major concern i s t h e l o c a t i o n of t o r s i o n a l
resonant frequencies r e l a t i v e t o t h e e x c i t a t i o n forces
i n t h e system. The eigenvector- eigenvalue method prov i d e s t h e f o l l o w i n g information:

*The a l l o w a b l e t o r s i o n a l stress f o r t h i s system

based upon U.S. Navy MIL STD 167 i s 4000 p s i zero peak
o r 8000 p s i peak-to-peak

Calculation of a l l possible resonant frequencies directly.

Exact simulation of multiple-branched s y s t e m s i s possible
along with branch-on-branch capabilities.
Forcing functions can be applied t o any m a s s location while
maintaining proper phase relationships.
Calculation of torsional s t r e s e e s produced by unbalanced
torques for complex s y s t e m s .

Since all calculations are dependent on the initial

values, it is imperative that rotary mass, stiffness,
and damping be properly calculated and applied. Gear
stiffness has classically plagued the accuracy of torsional calculations. However, the eigenvector-eigenvalue
technique along with proper simulation of stiffness, mass
and damping has been used to advantage in solving problems in operating systems.

NOMENCLATURE

A, B, C, D, E. F

Complex functions of m a s s and stiffness.

F o r c e v e c t o r (torques i n l l b s ) .
T o r s i o n a l Modulus (lb/in 2 ).
Height of t r i a n g l e i n F i g u r e I (in).
Height f r o m chord t o pitch c i r c l e i n F i g u r e I (in).
Diagonalized m a s s m a t r i x .
Stiffness m a t r i x .
L i n e a r stiffness of g e a r tooth.
Rotational stiffness r e l a t e d t o a g e a r wheel.
T o t a l l i n e a r stiffness.
L i n e a r stiffness of d r i v e r and d r i v e r g e a r .
L i n e a r stiffness a f t e r t h e c o r r e c t i o n s f a c t o r s a r e applied.
F a c e width of g e a r tooth (in).
Dynamic magnification factor.
C o r r e c t i o n f a c t o r f o r s e c o n d a r y effects.
G e a r wheel r a d i u s (in).
Length of chord a t root of tooth (in).
2
Modulus of e l a s t i c i t y ( l b l i n ).
Angular displacement m a t r i x .
Angular a c c e l e r a t i o n m a t r i x .
Diagonalized eigenvalue m a t r i x .
Operating frequency, r a d i a n s p e r second.
T o r s i o n a l n a t u r a l frequency, radian0 p e r second.
Diagonalized eigenvalue m a t r i x .

REFERENCES
1.

Ker Wilson, W., Practical Solution of Torsional Vibration

Problems,

Vol. I,

(1956),

Vol. 11,

(1963)(London: Chapman

and Hall).
2.

Nestoridee, E. J, , (Editor) A Handbook on Torsional Vibration

(London: Cambridge University P r e s s ) , 1958.

3.

Porter, F r e d e r i c P. , ttHarmonic Coefficients of Engine Torque

Curves tt , Journal of Applied Mechanics, March 1943, pp A- 33.

TABLE I
TORSIONAL EQUATIONS OF MOTION

TABLE I I I
RESONANT FREQUENCY PARAMETRIC ANALYSIS
GEAR STIFFNESS
(in-lblrad)

TORSIONAL RESONANT FREQUENCY

Hz

6
0.1 x10

4.94

1.0 x lo6

13*85

3.8 x lo6
6.24 x lo6

22.99

10.0 x lo6

24.62

100.0 x lo6

27.77

1,000.0 x lo6

28.13

10,000.0 x lo6

28.16

20,8

LEASECIRCLE
PITCH CIRCLE

GEOMETRIC NOMENCLATURE

FIGURE I

ENGINE SPEED, RPM

CALCULATED PINION SHAFT STRESSES.

FIGURE 2

INTERFERENCE DlAORAM OF ENOINE ORDERS (E) AND

COMPRESSOR HARMONICS (C) WITH TORSIONAL RESONANCES.

FIGURE 3