S.

DUBOWSKY
Senior Engineer,
Dynamic Analysis of Mechanical Systems
With Clearances3
O p t i c a l Technology Division,
Perkin-Elmer Corp.,
Danbury, Conn.

F. FREUDENSTEIN Part 2: Dynamic Response

Professor,
Department o f Mechanical Engineering, The theory developed in Part I has been applied to the determination of the dynamic re-
Columbia University,
sponse of the Impact Pair under various operating conditions. Simulation techniques,
New York, N. Y.
Mem. ASME as well as approximate methods used in control theory, have been used for this purpose.

Introduction
X2-X|
hEKHAPS the most significant result of this investiga-
tion is t h a t a simple clearance model, the impact pair, will ex-
hibit the nonlinear behavior common to complex systems con- JPERIOD OF I t h PERIOD OF i
taining clearances. It is hoped t h a t insight into the behavior CYCLE CYCLE
of such systems can be obtained by studying the characteristics
Fig. 5 Free-vibration motion of Impact Pair
of the impact pair. The following sections contain a discussion
of the dynamic response of the impact pair in the case of free
vibrations, constant-load operation, and under displacement-
forced excitation. required excessive computer time to yield sufficient accuracy,
We proceed directly to the integration of the equations of while the less-direct method would result in an efficient software
motion derived in P a r t I. package.

Numerical Simulation Methods—General Impact Pair Dynamic Behavior

The equations of motion, (6), for the impact pair, as shown in
Part I, may be solved directly using standard numerical tech-
niques on the digital computer. In this study both the analytical
solution and a digital solution have been applied to a number of
cases. The objective was to determine the accuracy and capabil-
ity of the numerical solutions, because the simulation approach
can be used also in the study of the Impact Pair in combination
with complex systems. For these systems an analytical approach
is not practical. The comparison of the numerical and ana-
lytical solution shows that the numerical approach, using a pre-
dictor-corrector technique, would yield the correct dynamic re-
sponse of the impact pair. The comparison showed also t h a t
simple, direct integration methods, such as an Euler algorithm,
3
Based on the doctoral dissertation of S. Dubowsky in partial ful-
fillment of the requirements of the degree of Doctor of Engineering
Science in the School of Engineering and Applied Science, Columbia
University, New York.
Contributed by the Design Engineering Division and presented
at the Mechanisms Coiiferenee, Columbus, Ohio, November 2-4,
1970, of THE AMERICAN SOCIETY OP MECHANICAL
Manuscript received at ASME Headquarters, July 27, 1970.
No. 70-Mech-65.
(a) Free Vibrations. Fig. 5 contains the time-displacement plot
for relative motion of the free vibration of the Impact Pair. The
important characteristics of this motion are the maximum surface
compression, which determines the contact forces and thus the
contact stresses, and the frequency of the vibration. Both of
these quantities are functions of the energy content of the pair,
which is a function of its free relative velocity. The maximum
stress corresponding to a given amplitude is found by multiply-
ing the compression, A',,, by the value of k. The resultant load
is then substituted into the expression for the stress as a function
of static load for the given connection configuration. This ap-
proach assumes that the frequency of elastic waves in the body
is much higher than the free-vibration frequency.
For the free-vibration case, it can be shown t h a t the introduc-
tion of clearances in the pair does not increase the maximum
compression or stresses, over the stresses found in a model with
surface compliance. I t will be seen that this is not the case for
forced vibration, which is discussed later, in which significant
stress amplification due to the clearance is found.
Fig. 6 contains the phase-plane diagram of the motion shown
ENGINEERS.
Paper in Fig. 5 with its decaying spiral, characteristic of a passively
damped system. The value of f, the damping ratio, used for this

310 / FEBRUARY 1 971 Transactions of the ASME

Copyright © 1971 by ASME
Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms
 £ = .025
Un = 16,000
r =,001

*-(-». - . -
STATIC
COMPRESSION^
x 2 -x,

>r^
L-^

Fig. 6 Free-vibrotion phase-plane plot of Impact Pair

Fig. 8 Phase-plane plot for constant-load case

.006
U n = 16,000 RAD/SEC
r =.001 INCHES
£=.025 X 2 -X| IN INCHES PER SECOND

P = 0. POUNDS MODE I
BEHAVIOR

.003 "2AI
2 3 4 5 IN INCH,
TIME IN MILLISECONDS

Fig. 7 Relative motion of impact pair under constant load

case is less than ten percent. In most cases damping ratios of

between one and five percent can be expected from metallic ma-
terials used for mechanism connections. T h e phase plane for
this case without damping is a simple closed curve.
The equations of motion show the existence of successive im-
pacts on alternate sides of the connection for the case of free
Fig. 9 Phase-plane plot of impact pair subject to various loads under
vibrations. I t can be seen in Fig. 5 t h a t the period of these im-
given initial relative velocity
pacts increases, or the frequency decreases, as the amplitude of
the motion decreases. Thus the system with clearance coupling
does not possess a single characteristic or natural frequency, but
a continuum of oscillating frequencies, which depend on time and
initial conditions. Systems with relatively high relative veloci- The model which is used to describe this behavior is shown in
ties or amplitudes will oscillate at frequencies very close to, but Fig. 1(6) where /<\(£) and Fi{t) are of equal and constant magni-
below, the natural frequency of the pair without clearances. tudes, P, but of opposite signs. This results in the pair elements
This shows that the behavior of the system becomes more linear being forced together toward one side, such as might result from
for amplitudes which are large compared to the clearance. Such the effect of gravity on a system. The resulting general motion
behavior, which is fundamental to the impact pair, is found also can be one of three different and distinct forms. The first mode,
in more complex systems with clearances. For this reason com- which corresponds to a large initial relative velocity compared to
plex control systems with clearances will exhibit very small un- the constant applied forces, has successive impacts on alternate
controlled, but bounded, oscillations, called limit cycles, for small sides of the clearance and looks much like the free-vibration
inputs, but are not unstable in the global sense. motion. This mode will decay into a second mode, in which the
As the amplitude of the motion decays, the frequency decreases effects of the initial velocity are of the same order of magnitude
and in the limit approaches zero. This corresponds to a lowering as those due to the applied loads and the pair exhibits impacts
of the effective spring rate as the amplitude decreases and the pair only on one side of the clearance. In the lowest-energy mode the
is in the free region a greater percentage of the time. relative velocity of the elements is not sufficient to separate the
pair elements and there is continuous contact on one side of the
(b) Constant-Load Operation. The constant-load case indicates
clearance. In this last condition the relative motion is described
some interesting aspects of the dynamic behavior of mechanisms
by a simple linear system. Fig. 7, which is the sample of the
with clearances. One is that it demonstrates clearly with a
relative motion for a mechanism under constant load, shows the
simple model the occurrence of multiple impacts on one side of
decay of the motion through the three modes. Fig. 8 shows the
the clearance (unilateral impact). This behavior, which is ex-
phase-plane plot corresponding to this motion. Fig. 9 contains
cluded by the Kobrinskii model, is exhibited also by the har-
the phase-plane plots for an undamped pair under several dif-
monically forced pair.

Journal of Engineering for Industry FEBRUARY 1971 / 311

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms

 MODE 3

16,000 Un = 15,000 rad/sec

r =.001 INCHES
14,000
p =1,0 POUND
12,000
C =0.
10,000
8000
6000
4000
2000

0 5.0 10.0
IMPACT VELOCITY IN INCHES PER SECOND
Fig. 10 Vibration frequency versus impact velocity at surface (— r),
constant-load operation

8 12 16 20
TIME IN MILLISECONDS

Fig. 12 Motion of displacement-forced impact pair for 0.001 in. clearance

suit permits us to calculate the increase in the internal forces

10 15 20 25
TIME IN MILLISECONDS
Fig. 11 Motion of displacement-forced impact pair for 0 . 1 0 in. clearance
ferent loads. The maximum amplitude of the surface compres-
sion is always greater than the static compression. I t can also
be shown t h a t for mode 1 and large amplitudes, the element be-
haves as in the free-vibration case, i.e., having the linear-system
frequency and amplitude for large velocities with these values
decreasing as the velocity decays. However, as shown in Fig. 10,
the frequency sharply increases back to the linear-system fre-
quency as mode 3 is entered. The maximum compression always
decreases, approaching the static compression, as the relative
velocity decays.
(c) Displacement-Forced Motion. The final and probably most
important case to be discussed here is the displacement-forced
case. In this case element 1 is assumed to follow a simple har-
monic motion. I t was found that, in the steady state, the output
will respond at the driving frequency. T h e major characteristic
of interest of this motion is the amplitude of the relative dis-
placement of the elements as a function of frequency. This re-
due to the clearance. This force or stress amplification factor is
of significance for the mechanism designer because it determines
the required tolerances of high-speed equipment. The transient
behavior of this case is in many ways similar to the transient re-
sponse of the constant-load case.
Before proceeding with the discussion, we note several charac-
teristics which are independent of clearance. First, in instrument
and high-speed-mechanism applications, the nominal range of
tolerances leads to very high spring rates, on the order of 10s
lb /in. When these spring rates are combined with the relatively
low masses usually found in these devices, resulting linear natural
frequencies in the range of 10,000 rad/sec (95,000 rpm) to 100,000
rad/sec (960,000 rpm) are not uncommon. Such devices are
usually driven at only fraction of these speeds. This is not a coin-
cidence, since elementary mechanics will predict t h a t for speeds
well below the natural frequency the output element will follow
the driving element faithfully; however, as the speed increases
the output element is not capable of following the input motion
and a large relative motion occurs. At very high speeds, well
past the resonance of the system, the output element will stand
still and the input motion will be transformed completely into
compression and extension of the compliant members resulting
in extremely high forces and stresses. Thus, for practical
reasons, mechanical systems are rarely driven at speeds more
than a few percent of any compliant resonances within the moving
elements. Thus, the frequency of the transient response is so
high compared to the forcing frequency, t h a t even for the light
damping assumed in this analysis, its effect usually decays within
a single cycle of the forced motion. Thus, particular attention in
this study was directed to the steady-state responses at fre-
quencies well below the linear resonance point of the pair.
The steady-state solution does exhibit high-frequency motion
when clearances are considered. However, this motion is inde-
pendent of the initial conditions; it is a periodic excitation of the
natural frequency of the pair due to the nonlinearity. Fig. 11
shows the nature of this motion quite clearly. For this plot, the
clearance to forcing amplitude, r/x, was taken as 0.1 (this is much
greater than would be expected for typical mechanisms, as are the
frequency and damping ratios), in order to illustrate the nature
of the motion. Fig. 12 illustrates the motion for a more reason-

312 / FEBRUARY 1 971 Transactions of the ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms


ill
Fig. 13 Surface compression in db as a function of normalized forcing frequency for
various values of clearance ( 0 . 0 0 0 1 in. < r $$ 0 . 0 5 0 in.)
0 I 2 3 4 5 6 7 8 9
CLEARANCE r, IN MILS (IO- 3 inches)
Fig. 14 Force-amplification versus clearance, case of
excitation
able clearance-to-driving-amplitude ratio of 0.001. In the top
curve of this figure, the three curves representing the driven
member and the two impact boundaries are indistinguishable,
while the relative-motion curve clearly shows the nature of the
motion. This curve also shows that even for relatively low
damping (zeta equal to 0.025) and a high frequency ratio (10
percent), the motion strongly resembles a series of constant-load
cases of alternating direction. The importance of permitting
multiple impacts on one surface is clear from these figures. I t
was found t h a t solutions could be obtained for all driving fre-
quencies and that the fundamental-frequency component of
these solutions was always the driving frequency. The forbidden
regions predicted by the Kobrinskii analysis were not in evidence.
It can be seen from Fig. 11 that the maximum surface com-
pression, X cml , x , is significantly higher than the static surface
compression; this results in significant stress amplification. Fig.
13 represents a plot of the surface compression as a function of
frequency for several different values of clearances. These curves
show t h a t the surface compression, and therefore the internal
forces, increase with frequency. In addition, it can be seen t h a t
increasing the clearance will greatly increase the dynamic forces,
Journal of Engineering for Industry
Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms
1 1—I I i i r ~i—i—i M 11
U n = 16,000 RAD/SEC
£ =.025
X,= 1.00 INCH
io!ooo (7
Ui -
- . 0 0 0 ( N 0 CLEARANCE)
1 I J l l i i i in J I
especially for low frequency ratios at which many devices operate.
Fig. 14 shows a plot of the dynamic force of couplings with
clearances above the zero, as a function of r. This graph was
constructed directly from Fig. 13 for «/&>„ equal to 0.01. As the
clearance increases the spring rate and natural frequency will
change somewhat depending upon the connection configuration,
and thus such a curve may not hold over very wide ranges of
clearances, Fig. 15 shows the effect of damping ratio within the
expected range of one-half to five percent of critical effect on the
force amplification of the impact pair is quite minor.
Approximate Solutions
T h e applicability of the approximate nonlinear techniques of
control engineering to the solution of the impact pair equations
was determined next. The most promising method appeared
to be the well-known describing-function technique.
The describing-function method can be used in two ways to
predict the dynamic response of nonlinear systems. One, it can
10
be used in conjunction with the Nyquist criterion to determine
the stability of a freely vibrating system. This includes predict-
forced-motion ing the existence of limit cycles, their frequencies, and ampli-
tudes. Two, the describing function can be used in very simple
systems to determine the frequency response to sinusoidal forcing
functions. The former application is the more common one,
and will be used here for the free-vibration case, while the second
method is applied to the displacement-forced case. Both of these
techniques are limited by the assumption associated with the
describing function, namely that when the input of the nonlinear
element is a sine wave the output is assumed to be sinusoidal,
but amplified by the gain factor, 0D, with a phase shift, <j>.
The magnitude of the gain, GD, and phase shift are functions
of the frequency and amplitude of the input sine wave and these
two variables constitute the describing function. This assump-
tion results in all of the higher harmonics of the nonlinear element
response being taken as zero. Such a system can be represented
by the block diagram shown in Fig. 16, where G(s) and H(s) are
the transfer functions of the linear forward and feedback elements
expressed as functions of the Laplacian variables.
The expressions for GD and </> for a clearance or dead zone are
well-known [11] and are given by:
R
GD{e) = - - V l - R*/e (23)
IT t e
<$> = o
FEBRUARY 1 9 7 1 / 313
 <:=5%

u.

LINEAR SYSTEM RESPONSE (£= 1/2 OR 5 % )

5 Damping versus frequency ratio, case of forced-motion excitation

ZERO CLEARANCE CASE

*l(s)
^O Gn*! G (s) -X2(s)

H (s)

Fig. 16 General block diagram

k + cS
K> 7^£GO.*0
mS 2
-X? (s)

Fig. 17 Impact-Pair system, block diagram

1.0 10.0 100.0

E/r VIBRATION AMPLITUDE OVER CLEARANCE r

Fig. 18 Free-vibration frequency versus relative-motion amplitude ^ r

where e is the magnitude of the sine-wave input to the non-
linearity, and R is one-half the total dead zone or clearance.
The block diagram corresponding to Fig. 16 for the one impact
pair is shown in Fig. 17, in which Xi(s) is equal to zero for the
free-vibration case, k and c denote the spring rate and damping
coefficient of the surface compliance, and m is the effective mass
of the pair.
(a) Free Vibrations. For the free-vibration case it can be shown
by standard Nyquist methods that no values of the initial con-
ditions result in a sustained oscillation. This is exactly the con-
clusion one would expect for an unforced, passive system with
dissipation.
Setting the damping equal to zero in this case will result in an
undamped limit cycle. The describing function then yields the
frequency as a function of the initial amplitude of the limit cycle.
This relationship is obtained from the Nyquist method, in which
limit cycles are predicted for certain intersections of the G(w)H(o>)
curve with the — l/GD(e) curve in the complex plane. The
G(o>)H(oi) curve is obtained from the transfer functions of the
linear elements by settings equal to ju(j = -\/—l). In this case
the curves G(co) and H(co) are coincident. Thus, there are an in-
finite number of crossings representing a continuum of limit-
cycle frequencies. This same result is given by the exact solu-
tion, equation (10). An expression for the frequency as a func-
tion of amplitude can thus be obtained by equating G(o>)H(o>) to
- l/0D, or
-1 -k
-i (24)
2 TV 1
- sin" - - -e V l -- 2
R /e 2
TV _2 " e
Fig. 18 contains a plot of free-vibration frequency for an un-
damped case, as a function of the ratio of relative-motion ampli-
tude to clearance, r. Both the exact solution, equation (10), and
the describing-function solution, equation (24), are shown. It
can be seen that for very small amplitudes not only are the pre-
dicted frequencies well below the linear-system frequency, but the
describing function overestimates the frequency. It is also
shown in the dissertation that this error in the limit-cycle fre-
quency was also found when the describing function was applied
to more complex electromechanical systems containing clearances.
(fc) Displacement-Forced Motion. The frequency response of the
impact pair to a displacement-forcing function may also be
analyzed using the describing-function technique. In the usual
manner, the response is considered after all the transients due to
the initial conditions have disappeared. The block diagram for
this case is obtained from Fig. 17 with Xi(s) being a harmonic
function and m infinite.
The transfer function of the relative motion may be written
from the block diagram as:

314 / FEBRUARY 1 971 Transactions of the AS ME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms

 - n U= 16,000
-
- ll £=.025 -
- / \ X, =1.000
-
1 1 • ' "J l 1 1 1 1 1 1 l| 1 1 1 1 I . / ' l[ [ 1 1 1 1 1 1 l| 1 i 11
1 I 1
001 .010 .100 / 1.000 10.000
? o -10
<->l-
05 -. / 4 _
-20
1,1 ^

S Q -30 - X| =
-
IE =>
=> 1 - '°'°'X^^V^// DESCRIBING FUNCTION
« | -40 - y?j/ SOLUTION -
5?
5
-50

-
AX

-60

-70 /y^~-— 'x,=.ooi -

X
-80
X / ^ N O CLEARANCE (SOLUTIONS COINCIDENT)
<^ -90 / -
/ /
Fig. 19 Displacement versus frequency ratio for displacement-forced Impact Pair, com-
parison of describing-function solution w i t h exact solution

(J n = l 6 , 0 0 0 RAD./SEC.
£ =.025
X, = 1.000

I I I I 1111 I—I I I 11 H
.000 10.000 (j

APPROXIMATE
DESCRIBING
FUNCTION
SOLUTION
EXACT SOLUTION
NO CLEARANCE (SOLUTIONS
COINCIDENT)

X=-00l
"I
r=.000l
X|

Fig. 2 0 Displacement versus frequency ratio for displacement-forced Impact Pair,

comparison of approximate describing-function solution with exact solution

Af2s2 due the high-frequency harmonics present in the solution. I t is

(25)
M^ + GD(e)(k + cs) a basic property of the describing function to consider only the
fundamental component of the response. An interesting, and
Setting s to ju and solving for e results in the frequency response

Equation (26) is a transcendental expression for 6 which can be possibly useful result was obtained in an attempt to find an ap-
solved both numerically and approximately to yield e as a func- proximate solution to the transcendental equation (26) for the
tion r, the clearance; w„ and f; and the dynamic parameters, u results presented in Fig. 19. In this approximation, the de-
and Xi, the forcing frequency and amplitude. scribing-function value given by equation (23) was approximated
The accuracy of the describing-function approach to this prob- by:
lem, equation 26, was determined by comparing its response to
the exact solution. The results are shown in Fig. 19. Clearly 4£ (27)
OB 1
the describing function does not adequately predict the force 67T
amplification due the clearance effects. The reason can be seen
for the inadequacy of the describing function from Fig. 13, in With this approximation equation (26) becomes a quadratic equa-
which it is apparent t h a t a good part of the stress amplification is tion and thus solvable directly for the relative-motion amplitude:

Journal of Engineering for Industry FEBRUARY 1 971 / 315

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms

 CLEARANCE
'COUPLING Acknowledgments
S. Dubowsky would like to express his thanks to the Perkin-
Elmer Corporation for its cooperation and generous financial
support of his efforts during this investigation, especially to P. J.
(^CONSTANT \\\\\\\\Y^
Convertito and F . S. Holzer.
The authors are grateful to Columbia University for the use of
computational facilities.

HIGH INERTIA DISCUSSION

DRIVING ELEMENT

Fig. 21 Illustrative e x a m p l e A. G. Enlman'and G. N. Sandor6

€[(a>r2 - l ) 2 + (2wr)2] = - {1 - av*(l - 4 2 )}
7T
16r2
^ V'Qi where a)r = j/co„ and Q =
IT2
2 2 2 s 2 2
X {1 - a>r (l - 4 )} - {(wr - l ) + (2u r ) }
16r2
X (42<V + 1) - z,2cor4
_ '""2
The result given by this approximation is compared to the exact
solution in Fig. 20, and it can be seen to be a poor approximation
to the describing-function solution. However, it does yield a
very fast solution which is not far different from the exact solu-
tion for some cases and is conservative for all cases. I t is possible
that equation (28) might be a useful design tool.
A brief example to illustrate how the results of this study
might be used to determine the maximum permitted tolerance for
a design is the specification of the clearances in the pin connection
of a slider-crank mechanism with a long connecting rod. If the
driving elements (see Fig. 21) are of high inertia compared to the
slider, then the position-forced impact pair is a good dynamic
model of this problem. Assume t h a t the pin connection shown
in Fig. 3(6), for which the maximum allowable cyclical deflection,
X cm ai is approximately 0.6 X 10 ~4 in., is to be used. Taking
the stroke as one inch and assuming the slide weighs one pound,
then from equation (16a)
co„ = 14,600 rad/sec
and
6 10
X^/Xt = - * ~* = - 8 4 d b
From Fig. 13 it follows that for the zero-clearance case the system
may be run at 0.008w/«„ or 115 rad/sec or approximately 1230
rpm with no fatigue-induced wear of the pin. A clearance of
±0.001 in. would reduce the permitted speed to 0.0015 CO/OJ„ or
230 rpm. The problem could be just as readily worked for the
case where the speed is given and the required tolerance is cal-
culated.
Conclusion
The dynamic response of the impact pair under a variety of
operating conditions has been analyzed. I t is shown that the
simple, elastic, impact-pair model used exhibits a variety of dy-
namic characteristics which are believed representative of me-
chanical systems with clearances and backlash.
These papers well deserved the Proctor and Gamble Award
which they won at the 11th ASME Mechanisms Conference.
The authors have laid the ground work for additional contribu-
tions which involve the complete, real mechanism system. At high
speeds, the varying inertia forces in the mechanism, as well as
the strains due to elasticity of links, will certainly affect the forces
—
(28) on the impact pair. In other words, there will not only be dis-
placement-forced motion which is quasi-harmonic, but the forces
Fi(t) and F2(t) will be functions of parameters such as length
and stiffness of links, configuration, distribution of mass, and
input speed of the mechanism. Although the solutions for these
force functions may not be derived in closed form, they may
be obtained by energy balance in the mechanism. Once these
functions are found, and a new criterion developed for deriving
new initial conditions for successive impact regions including the
above mentioned parameters in addition to previous exit condi-
tions, the same type of solution as outlined in these papers would
yield results to many real problems. A further possibility for
the extension of the present model could include a representation
of the liquid or gaseous lubricant film filling the clearance.
4
Doctoral Candidate/Graduate Research Assistant.
6
Alcoa Foundation Professor of Mechanical Design/Chairman,
Machines and Structures Division, respectively, Rensselaer Poly-
technic Institute, Troy, N. Y.
Authors' Closure
The authors wish to thank Mr. Arthur Erdman and Professor
George N . Sandor for their kind comments.
With regard to the forces acting on the impact pair as a result
of other elements and effects within a complex mechanical
system, the discussers have pointed out appropriately that such
forces will have some effect on the impact-pair characteristics.
The subject of system interaction, or dynamic coupling, has
been considered by the authors to some extent in their subsequent
work, which is part of a general studj' of clearances in electro-
mechanical systems, including plane mechanisms. I t is hoped
that the results of this study will be published in due course.
In this study, the response of a simple impact pair is compared to
that of a complex system containing clearance elements. In the
light of the above discussion, two conclusions, which have •
emerged from this stud}', may be worth mentioning. The first is
that the simple displacement impact pair is capable of predicting
closely the response of complex systems containing clearance
coupling. The second is t h a t when considering complex me-
chanical sj'stems, a numerical approach using the digital computer
appears to offer a practical method of solution.
The inclusion of both inertia and elastic forces in the dynamic
analysis of clearance effects by means of an energy balance, as
suggested by the discussers, would constitute a valuable contribu-
tion. The authors look forward to learning about the results of
such research on the part of our esteemed colleagues.

316 /FEBRUARY 1971 Transactions of the ASME

Downloaded From: http://manufacturingscience.asmedigitalcollection.asme.org/ on 04/07/2014 Terms of Use: http://asme.org/terms