bud98209_fm_i-xxii_1.

indd Page 1 12/9/13 4:52 PM f-494 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

Shigley’s
Mechanical
Engineering
Design
bud98209_ch16_817-870.indd Page 858 22/11/13 3:02 PM user-f-w-198 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

858 Mechanical Engineering Design

16–12 Flywheels
The equation of motion for the flywheel represented in Fig. 16–1b is
# # $
^ M 5 Ti (ui, ui ) 2 To (uo, uo ) 2 Iu 5 0
or
$
I u 5 Ti (ui, vi ) 2 To (uo, vo ) (a)
# $
where Ti is considered positive and To negative, and where u and u are the first and
second time derivatives of u, respectively. Note that both Ti and To may depend for
their values on the angular displacements ui and uo as well as their angular velocities
vi and vo. In many cases the torque characteristic depends upon only one of these.
Thus, the torque delivered by an induction motor depends upon the speed of the motor.
In fact, motor manufacturers publish charts detailing the torque-speed characteristics
of their various motors.
When the input and output torque functions are given, Eq. (a) can be solved for
the motion of the flywheel using well-known techniques for solving linear and non-
linear differential equations. We can dispense with this here by assuming a rigid shaft,
giving ui 5 u 5 uo and vi 5 v 5 vo. Thus, Eq. (a) becomes
$
I u 5 Ti (u, v) 2 To (u, v) (b)

When the two torque functions are known and the starting values$ of the displacement u
and velocity v are given, Eq. (b) can be solved for u, v, and u as functions of time.
However, we are not really interested in the instantaneous values of these terms at all.
Primarily we want to know the overall performance of the flywheel. What should its
moment of inertia be? How do we match the power source to the load? And what are
the resulting performance characteristics of the system that we have selected?
To gain insight into the problem, a hypothetical situation is diagrammed in Fig. 16–27.
An input power source subjects a flywheel to a constant torque Ti while the shaft rotates
from u1 to u2. This is a positive
$ torque and is plotted upward. Equation (b) indicates
that a positive acceleration u will be the result, and so the shaft velocity increases from
v1 to v2. As shown, the shaft now rotates from u2 to u3 with zero torque and hence,
from Eq. (b), with zero acceleration. Therefore v3 5 v2. From u3 to u4 a load, or
output torque, of constant magnitude is applied, causing the shaft to slow down from
v3 to v4. Note that the output torque is plotted in the negative direction in accordance
with Eq. (b).
The work input to the flywheel is the area of the rectangle between u1 and u2, or

Ui 5 Ti (u2 2 u1 ) (c)

Figure 16–27 T, ␻
Ti ␻2 ␻3
␻

␻1 Ui ␻4
␪3 ␪4
␪
␪1 ␪2
Uo
To

1 cycle
bud98209_ch16_817-870.indd Page 859 22/11/13 3:02 PM user-f-w-198 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

Clutches, Brakes, Couplings, and Flywheels 859

The work output of the flywheel is the area of the rectangle from u3 to u4, or
Uo 5 To (u4 2 u3 ) (d)

If Uo is greater than Ui, the load uses more energy than has been delivered to the
flywheel and so v4 will be less than v1. If Uo 5 Ui, v4 will be equal to v1 because
the gains and losses are equal; we are assuming no friction losses. And finally, v4
will be greater than v1 if Ui . Uo.
We can also write these relations in terms of kinetic energy. At u 5 u1 the fly-
wheel has a velocity of v1 rad/s, and so its kinetic energy is
1 2
E1 5 Iv1 (e)
2
At u 5 u2 the velocity is v2, and so
1 2
E2 5 Iv2 (f )
2
Thus the change in kinetic energy is
1
E2 2 E1 5 I(v22 2 v21 ) (16–61)
2
Many of the torque displacement functions encountered in practical engineering
situations are so complicated that they must be integrated by numerical methods.
Figure 16–28, for example, is a typical plot of the engine torque for one cycle of
motion of a single-cylinder internal combustion engine. Since a part of the torque
curve is negative, the flywheel must return part of the energy back to the engine.
Integrating this curve from u 5 0 to 4p and dividing the result by 4p yields the mean
torque Tm available to drive a load during the cycle.
It is convenient to define a coefficient of speed fluctuation as
v 2 2 v1
Cs 5 (16–62)
v

Figure 16–28
Relation between torque and
crank angle for a one-cylinder,
four-stroke–cycle internal
Crank torque T

combustion engine.

Tm

180° 360° 540° 720°

Crank angle ␪
bud98209_ch16_817-870.indd Page 860 22/11/13 3:02 PM user-f-w-198 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

860 Mechanical Engineering Design

where v is the nominal angular velocity, given by

v2 1 v1
v5 (16–63)
2
Equation (16–61) can be factored to give
I
E2 2 E1 5 (v2 2 v1 )(v2 1 v1 )
2
Since v2 2 v1 5 Csv and v2 1 v1 5 2v, we have
E2 2 E1 5 Cs Iv2 (16–64)

Equation (16–64) can be used to obtain an appropriate flywheel inertia corresponding

to the energy change E2 2 E1.

EXAMPLE 16–6 Table 16–6 lists values of the torque used to plot Fig. 16–28. The nominal speed of
the engine is to be 250 rad/s.
(a) Integrate the torque-displacement function for one cycle and find the energy that
can be delivered to a load during the cycle.
(b) Determine the mean torque Tm (see Fig. 16–28).
(c) The greatest energy fluctuation is approximately between u 5 15° and u 5 150°
on the torque diagram; see Fig. 16–28 and note that To 5 2Tm. Using a coefficient
of speed fluctuation Cs 5 0.1, find a suitable value for the flywheel inertia.
(d) Find v2 and v1.

Solution (a) Using n 5 48 intervals of Du 5 4py48, numerical integration of the data of

Table 16–6 yields E 5 3368 in ? lbf. This is the energy that can be delivered to the load.

Table 16–6 U, T, U, T, U, T, U, T,
deg lbf • in deg lbf • in deg lbf • in deg lbf • in
Plotting Data for
Fig. 16–28 0 0 195 2107 375 285 555 2107
15 2800 210 2206 390 2125 570 2206
30 2090 225 2260 405 289 585 2292
45 2430 240 2323 420 8 600 2355
60 2160 255 2310 435 126 615 2371
75 1840 270 2242 450 242 630 2362
90 1590 285 2126 465 310 645 2312
105 1210 300 28 480 323 660 2272
120 1066 315 89 495 280 675 2274
135 803 330 125 510 206 690 2548
150 532 345 85 525 107 705 2760
165 184 360 0 540 0 720 0
180 0
bud98209_ch16_817-870.indd Page 861 22/11/13 3:02 PM user-f-w-198 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

Clutches, Brakes, Couplings, and Flywheels 861

3368
Answer (b) Tm 5 5 268 lbf ? in
4p
(c) The largest positive loop on the torque-displacement diagram occurs between
u 5 0° and u 5 180°. We select this loop as yielding the largest speed change.
Subtracting 268 lbf ? in from the values in Table 16–6 for this loop gives, respectively,
2268, 2532, 1822, 2162, 1892, 1572, 1322, 942, 798, 535, 264, 284, and 2268 lbf ? in.
Numerically integrating T 2 Tm with respect to u yields E2 2 E1 5 3531 lbf ? in. We
now solve Eq. (16–64) for I. This gives

E2 2 E1 3531
Answer I5 5 5 0.565 lbf ? s2 in
Cs v2 0.1(250) 2

(d) Equations (16–62) and (16–63) can be solved simultaneously for v2 and v1.
Substituting appropriate values in these two equations yields

v 250
Answer v2 5 (2 1 Cs ) 5 (2 1 0.1) 5 262.5 rad/s
2 2

Answer v1 5 2v 2 v2 5 2(250) 2 262.5 5 237.5 rad/s

These two speeds occur at u 5 180° and u 5 0°, respectively.

Punch-press torque demand often takes the form of a severe impulse and the
running friction of the drive train. The motor overcomes the minor task of overcom-
ing friction while attending to the major task of restoring the flywheel’s angular speed.
The situation can be idealized as shown in Fig. 16–29. Neglecting the running friction,
Euler’s equation can be written as
1
T(u1 2 0) 5 I(v21 2 v22 ) 5 E2 2 E1
2
where the only significant inertia is that of the flywheel. Punch presses can have the
motor and flywheel on one shaft, then, through a gear reduction, drive a slider-crank
mechanism that carries the punching tool. The motor can be connected to the punch

Figure 16–29 Torque T Torque TM

(a) Punch-press torque demand

during punching. (b) Squirrel-
cage electric motor torque-
speed characteristic.

Tr

0 0
0 ␪1 0 ␻r ␻s
Rotation ␪ Angular velocity ␻

(a) (b)
bud98209_ch16_817-870.indd Page 862 22/11/13 3:02 PM user-f-w-198 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

862 Mechanical Engineering Design

continuously, creating a punching rhythm, or it can be connected on command through

a clutch that allows one punch and a disconnect. The motor and flywheel must be sized
for the most demanding service, which is steady punching. The work done is given by
u2

#
1
W5 [T (u) 2 T ]du 5 I(v2max 2 v2min )
u1
2
This equation can be arranged to include the coefficient of speed fluctuation Cs as
follows:
1 I
W5 I(v2max 2 v2min ) 5 (v max 2 v min )(v max 1 v min )
2 2
I
5 (Cs v )(2v0 ) 5 ICs vv0
2
When the speed fluctuation is low, v 0 < v, and
W
I5
Cs v 2
An induction motor has a linear torque characteristic T 5 av 1 b in the range
of operation. The constants a and b can be found from the nameplate speed vr and
the synchronous speed vs:
T r 2 Ts Tr Tr
a5 5 52
vr 2 vs vr 2 vs vs 2 vr
(16–65)
Tr vs 2 Ts vr Tr vs
b5 5
vs 2 vr vs 2 vr
For example, a 3-hp three-phase squirrel-cage ac motor rated at 1125 rev/min has
a torque of 63 025(3)y1125 5 168.1 lbf ? in. The rated angular velocity is vr 5
2pnry60 5 2p(1125)y60 5 117.81 rad/s, and the synchronous angular velocity vs 5
2p(1200)y60 5 125.66 rad/s. Thus a 5 221.41 lbf ? in ? s/rad, and b 5 2690.9 lbf ? in,
and we can express T(v) as av 1 b. $During the interval from t1 to t2 the motor
accelerates the flywheel according to I u 5 TM (i.e., Tdvydt 5 TM). Separating the
equation TM 5 Idvydt we have
t2 v2 v2
I av2 1 b
# # #
Idv dv I T2
dt 5 5I 5 ln 5 ln
t1 vr
TM vr
av 1 b a avr 1 b a Tr

or
I T2
t2 2 t1 5 ln (16–66)
a Tr
For the deceleration interval when the motor and flywheel feel the punch torque on
the shaft as TL, (TM 2 TL) 5 Idvydt, or
t1 vr vr
I avr 1 b 2 TL
# # #
dv dv
dt 5 I 5I 5 ln
0 v2
TM 2 TL v2
av 1 b 2 TL a av2 1 b 2 TL

or
I Tr 2 TL
t1 5 ln (16–67)
a T2 2 TL
bud98209_ch16_817-870.indd Page 863 22/11/13 3:02 PM user-f-w-198 /204/MH01996/bud98209_disk1of1/0073398209/bud98209_pagefiles

Clutches, Brakes, Couplings, and Flywheels 863

We can divide Eq. (16–66) by Eq. (16–67) to obtain

T2 TL 2 Tr (t2 2t1)yt1
5a b (16–68)
Tr TL 2 T2
Equation (16–68) can be solved for T2 numerically. Having T2 the flywheel inertia is,
from Eq. (16–66),
a(t2 2 t1 )
I5 (16–69)
ln (T2 yTr )
It is important that a be in units of lbf ? in ? s/rad so that I has proper units. The
constant a should not be in lbf ? in per rev/min or lbf ? in per rev/s.

PROBLEMS
16–1 The figure shows an internal rim-type brake having an inside rim diameter of 300 mm and a
dimension R 5 125 mm. The shoes have a face width of 40 mm and are both actuated by a
force of 2.2 kN. The drum rotates clockwise. The mean coefficient of friction is 0.28.
(a) Find the maximum pressure and indicate the shoe on which it occurs.
(b) Estimate the braking torque effected by each shoe, and find the total braking torque.
(c) Estimate the resulting hinge-pin reactions.

30° 30°

F F

Problem 16–1 120° 120°

Pin Pin

30° 30°

16–2 For the brake in Prob. 16–1, consider the pin and actuator locations to be the same. However,
instead of 120°, let the friction surface of the brake shoes be 90° and centrally located. Find
the maximum pressure and the total braking torque.

16–3 In the figure for Prob. 16–1, the inside rim diameter is 11 in and the dimension R is 3.5 in.
The shoes have a face width of 1.25 in. Find the braking torque and the maximum pressure
for each shoe if the actuating force is 225 lbf, the drum rotation is counterclockwise, and
f 5 0.30.

16–4 The figure shows a 400-mm-diameter brake drum with four internally expanding shoes. Each
of the hinge pins A and B supports a pair of shoes. The actuating mechanism is to be arranged
to produce the same force F on each shoe. The face width of the shoes is 75 mm. The material
used permits a coefficient of friction of 0.24 and a maximum pressure of 1000 kPa.