CHAPTER

Axisymmetric
Problems

8.1

INTRODUCTION

Many problems of practical importance are concerned with solids of revolution

which are deformed symmetrically with respect to the axis of revolution.
Examples of such solids are circular cylinders subjected to uniform internal and
external pressures, rotating circular disks, spherical shells subjected to uniform
internal and external pressures, etc. In this chapter, a few of these problems will
be investigated. Let the axis of revolution be the z-axis. The deformation being
symmetrical with respect to the z-axis, it is convenient to use cylindrical coordinates.
Since the deformation is symmetrical about the axis, the stress components do
not depend on q. Further, trq and tq z do not exist. Consequently, the differential
equations of equilibrium [Eqs (1.67)(1.69)] can be reduced to our special case.
However, it is instructive to derive the relevant equations applicable to
axisymmetric problems from first principles. Consider an axisymmetric body shown
in Fig. 8.1. Let an elementary radial element be isolated. The stress vectors
acting on its faces are as shown.

z
sz + Dsz
H
Dz

sr

Chapter_08.pmd

269

sr + Dsr
D F

E
A

Fig. 8.1

sz
Dr

C
B
sq

An axisymmetric body

6/18/2008, 2:22 PM

270 Advanced Mechanics of Solids

On faces ABCD and EFGH, the normal stresses are sq and there are no shear
stresses. On face ABFE, the stresses are sz and tzr. On face CDHG, the normal
and shear stresses are
sz + Dsz = sz +

z
Dz
z

trz + Dtrz = trz +

rz
Dz
z

On face AEHD, the normal and shear stresses are sr and trz. On face BCGF,
r
rz
the stresses are sr +
Dr and trz +
Dr
r
r
For equilibrium in z direction

z
rz
r

z + z z r + 2 r + rz + r r (r + Dr) Dq Dz

trzr Dq Dz sz r +

r
Dq Dr + gz
2

r +
Dq Dr Dz = 0
2

where gz is the body force per unit volume in z direction. Hence,

z
rz
r
(r + Dr) Dr Dq Dz
r +
r z +
z
2
r

+ trz Dr Dq Dz + gz r + Dr Dq Dz = 0

2
Cancelling Dr Dq Dz and going to limits

z rz rz
+ gz = 0
+
+
r
z
r

(8.1)

Similarly, for equilibrium in r direction we get

r rz r
+ gr = 0
+
+
r
r
z

(8.2)

where gr is the body force per unit volume in r direction. Since the stress components
are independent of q, the equilibrium equation for q direction is identically satisfied.
For the problems that we are going to discuss in this chapter, we need expressions for the circumferential strain eq and the radial strain er.
Referring to Fig. 8.2(a), consider an arc AE at distance r, subtending an angle
Dq at the centre. The arc length is rDq. The radial displacement is ur. Consequently, the length of the arc becomes (r + ur) Dq. Hence, the circumferential
strain is
eq =

( r + ur ) r = ur

(8.3)

ur
r

(8.4)

r
The radial strain is, from Fig. 8.2(b),

er =

Chapter_08.pmd

270

6/18/2008, 2:25 PM

Axisymmetric Problems

Dq E
A
r Dr B

ur +

ur
A

A
Dr

271

ur
r
r

(b)
(a)

Fig. 8.2

Displacements along a radius

The axial strain is

ez =

uz
z

(8.5a)

where uz is the axial displacement. In subsequent sections we shall consider the

following problems:
Circular cylinder subjected to internal or external pressure
Sphere subjected to internal or external pressure
Sphere subjected to mutual gravitational attraction
Rotating disk of uniform thickness
Rotating disk of variable thickness
Rotating shaft and cylinder

8.2 THICK-WALLED CYLINDER SUBJECTED TO

INTERNAL AND EXTERNAL PRESSURESLAME'S
PROBLEM
Consider a cylinder of inner radius a and outer radius b (Fig. 8.3). Let the cylinder
be subjected to an internal pressure pa and an external pressure pb. It is possible
to treat this problem either as a plane stress case (sz = 0) or as a plane strain case
(ez = 0). Appropriate solutions will be obtained for each case.
pb
pb
b
pa

pa

(b)

(a)

Fig. 8.3 Thick-walled cylinder under internal and external pressures

Chapter_08.pmd

271

6/18/2008, 2:22 PM

272 Advanced Mechanics of Solids

Case (a) Plane Stress

Let the ends of the cylinder be free to expand. We shall assume that sz = 0 and
our results will justify this assumption. Owing to uniform radial deformation, trz = 0.
Neglecting body forces, Eq. (8.2) reduces to

r r
+
=0
r
r

(8.5b)

Since r is the only independent variable, the above equation can be written as
d
( r r ) = 0
dr
Equation (8.1) is identically satisfied. From Hookes law

r = 1 ( r ) ,

E
or the stresses in terms of strains are

(8.5c)

= 1 ( r )
E

E
( r + ) = E 2 ( + r )
1 2
1
Substituting for er and eq from Eqs (8.3) and (8.4)

r =

r =

u
E dur
+ r

r
1 2 dr

(8.6a)

dur
E ur
+ dr
1 2 r
Substituting these in the equation of equilibrium given by Eq. (8.5c)

d
dr

or

(8.6b)

dur
dur
ur
r dr + ur r + dr = 0

dur
d 2 ur
du
u
du
+r
+ r r r =0
2
dr
dr
r
dr
dr
d 2 ur

1 dur ur
2 =0
r dr
dr
r
This can be reduced to

i.e.

d dur ur
+ =0
dr dr
r
d 1 d
(8.7)
(u r ) = 0
dr r dr r
If the function ur is found from this equation, the stresses are then determined
from Eqs. (8.6a) and (8.6b).
The solution to Eq. (8.7) is

or

ur = C1 r +

Chapter_08.pmd

272

C2
r

(8.8)

6/18/2008, 2:22 PM

Axisymmetric Problems

273

where C1 and C2 are constants of integration. Substituting this function in

Eqs. (8.6a) and (8.6b)

r =

E
1 2

C1 (1 + ) C2 (1 ) 12
r

(8.9a)

E
1 2

1
C1 (1 + ) + C2 (1 ) r 2

(8.9b)

The constants C1 and C2 are determined from the boundary conditions.

When
r = a,
sr = pa
when
r = b,
sr = pb
Hence,

whence,

E
1 2

1
C1 (1 + ) C2 (1 ) a 2 = pa

E
1 2

1
C1 (1 + ) C2 (1 ) b2 = pb

C1 =

1 pa a 2 pb b 2
E
b2 a2

C2 =

1 + a 2b2
( p pb )
E b2 a2 a

Substituting these in Eqs (8.8) and (8.9) we get

ur =

r =
=

1 pa a 2 pb b 2
1 + a 2 b2 pa pb
r+
2
2
E
E
r b2 a2
b a
pa a 2 pb b 2
2

b a

pa a 2 pb b 2
2

b a

a 2 b 2 pa pb

a 2 b 2 pa pb

r2

r2

(8.10)

(8.11)

b2 a 2

(8.12)

b2 a 2

It is interesting to observe that the sum sr + sq is constant through the thickness of the wall of the cylinder, i.e. independent of r. Hence, according to
Hookes Law, the stresses sr and sq produce a uniform extension or contraction
in z direction, and cross-sections perpendicular to the axis of the cylinder remain
plane. If we consider two adjacent cross-sections, the deformation undergone
by the element does not interfere with the deformation of the neighbouring
element. Hence, the elements can be considered to be in a state of plane stress,
i.e. sz = 0, as we assumed at the beginning of the discussion. It is important to
note that in Eqs (8.10)(8.12), pa and pb are the numerical values of the compressive pressures applied.

Chapter_08.pmd

273

6/18/2008, 2:22 PM

274 Advanced Mechanics of Solids

Cylinder Subjected to Internal Pressure
Eqs (8.11) and (8.12) become

In this case pb = 0 and pa = p. Then

r =

pa 2
b2

b2 a 2
r2

(8.13)

pa 2
b2
+
1

b2 a 2
r2

(8.14)

These equations show that sr is always a compressive stress and sq a tensile

stress. Figure 8.4 shows the variation of radial and circumferential stresses across
the thickness of the cylinder under internal pressure. The circumferential stress is
greatest at the inner surface of the cylinder, where

( )max =

p a2 + b2
2

b a

(8.15)

sq

sr

b 2 + a2
b 2 a2

p
(a)

Fig. 8.4

(b)

Cylinder subjected to internal pressure

Hence, (sq)max is always greater than the internal pressure and approaches this
value as b increases so that it can never be reduced below pa irrespective of the
amount of material added on the outside.
Cylinder Subjected to External Pressure
(8.11) and (8.12) reduce to

In this case, pa = 0 and pb = p. Equations

sr

sq

Fig. 8.5 Cylinder subjected to

external pressure

Chapter_08.pmd

a2

b2 a2
r2

a2
1 + 2
b a
r

(8.16)

pb 2

r =

274

pb 2

(8.17)

The variations of these stresses across the

thickness are shown in Fig. 8.5. If there is
no inner hole, i.e. if a = 0, the stresses are
uniformly distributed in the cylinder with
s r = s q = p.

6/18/2008, 2:22 PM

Axisymmetric Problems

275

Example 8.1 Select the outer radius b for a cylinder subjected to an internal
pressure p = 500 atm with a factor of safety 2. The yield point for the material (in
tension as well as in compression) is syp = 5000 kgf/cm2 (490000 kPa). The inner
radius is 5 cm. Assume that the ends of the cylinder are closed.

Solution The critical point lies on the inner surface of the cylinder, where
= p

s r = p,

b2 + a 2
2

b a

z = p

a2
b a2
2

(assumed)

In the above expressions, it is assumed that away from the ends, sz caused by p
is uniformly distributed across the thickness. The maximum and minimum principal
stresses are s1 = sq and s3 = sr. Hence,
1 = p b2
(
3)
2 1
b2 a2

tmax =

Substituting the numerical values (1 atm = 98.07 kPa),

5
a = 6.45 cm
3

b=

Example 8.2 A thick-walled steel cylinder with radii a = 5 cm and b = 10 cm is

subjected to an internal pressure p. The yield stress in tension for the material is
350 MPa. Using a factor of safety of 1.5, determine the maximum working pressure
p according to the major theories of failure. E = 207 106 kPa, v = 0.25.

Solution
(i) Maximum normal stress theory
Maximum normal stress = sq at r = a
= p

(b 2 + a 2 )
2

(b a )

or

p=

(b 2 + a 2 )
(b 2 a 2 )

y
N
350 106 100 25

= 140 103 kPa

1.5
100 + 25

(ii) Maximum shear stress theory

Maximum shear stress = 1 ( r ) at r = a
2
2

= 1 p 22b 2
2 b a

Chapter_08.pmd

y
b2
=1
2
2
N
b a
2

275

6/18/2008, 2:22 PM

276 Advanced Mechanics of Solids

350 106 100 25

= 87.5 103 kPa

3
100
(iii) Maximum strain theory
Maximum strain = eq = 1 ( r ) at r = a
E

or

p=

p
a2
2
E (b a 2 )

b2
b2
1 + 2 1 2
a
a

p
( a 2 + b 2 ) (a 2 b 2 ) = y
NE
E (b2 a 2 )

or

350 106
p
125 + ( 0.25 75) =
(100 25)
1.5

350 106 75
= 121.7 103 kPa
1.5 143.75

p=

(iv) Octahedral shear stress theory

1/ 2
t oct = 1 2 + r2 + ( r )2
at r = a

3
1/ 2
1
2 ( r ) 2 + 2 r
3

1/ 2

= 2
3

2
2
p (b 2 + a 2 )

2 (b + a )

p
p

(b 2 a 2 )
(b 2 a 2 )

2 y
3 N

1/ 2

or

(b 2 + a 2 )
2 4b 4

p 2

2 2
3
(b 2 a 2 )
(b a )
p

40000 125

5625
75

1/ 2

2 y
3 N

350
1.5

\
p = 100 103 kPa
(v) Energy of distortion theory
This will give a value identical to that obtained based on octahedral shear
stress theory, i.e. p = 100 103 kPa.
Example 8.3 A pipe made of steel has a tensile elastic limit sy = 275 MPa and
E = 207 106 kPa. If the pipe has an internal radius a = 5 cm and is subjected to an
internal pressure p = 70 103 kPa, determine the proper thickness for the pipe wall
4
according to the major theories of failure. Use a factor of safety N = .
3

Chapter_08.pmd

276

6/18/2008, 2:22 PM

Axisymmetric Problems

Solution
(i) Maximum principal stress theory
Maximum principal stress = sq at r = a
=p

(b 2 + a 2 )
2

(b a )

y
N

70 106 (25 10 4 ) + b 2 275 106 3

=
4
b 2 (25 104 )

or

1750 104 + 70b 2 = 825b 2

20625
104
4

or
136.25b2 = 6906.25 104
\
b = 7.12 102 m = 7.12 cm
\ Wall thickness t = 2.12 cm
(ii) Maximum shear stress theory
tmax = 1 ( r ) at r = a
2
=
\

y
pb 2
=
(b 2 a 2 ) N
70 106 b 2

[b 2 (25 10 4 )]

3
275 106
8

or
70b2 = 103.13 b2 2578.1 104
\
b = 8.82 102 m = 8.82 cm
\ Wall thickness t = 3.82 cm
(iii) Maximum strain theory (with n = 0.25)
emax = 1 ( r ) at r = a
E

p
(a 2 + b 2 ) (a 2 b 2 ) = y
2

NE
E (b a )
2

70 106

(0.75 25 10 4 )
[b 2 (25 10 4 )]
3
275 106
4
1312.5 104 + 87.5b2 = 206.25b2 5156.25 104
+ (1.25 b 2 ) =

or

\
b = 7.38 102 m = 7.38 cm
\ Wall thickness t = 2.38 cm
(iv) Maximum distortion energy theory
From Eq. (4.12) with
s 1 = sq , s2 = 0, s3 = sr = p

Chapter_08.pmd

277

6/18/2008, 2:22 PM

277

278 Advanced Mechanics of Solids

U* =
=

(1 + )
(22 + 2 r2 2 r )
6E

2
(1 + ) 2
1+ y
( + r2 r ) =
3E
E N2
2

\
i.e.

1
[ 2 + r2 + ( r )2 ]
12G

r2

r =

2y
N2

2
(b 2 + a 2 ) 2
(b 2 a 2 ) y
+1+ 2
p 2
=
2 2
(b + a 2 ) N 2
(b a )
2

y
Putting
= fy and simplifying one gets
p N
(3 f y2 ) b 4 + 2a 2 f y2 b 2 + (1 f y2 ) a 4 = 0

b2 =

=
With

2a 2 f y2

4a 4 f y4 4a 4 (1 f y2 ) (3 f y2 )

2 (3 f y2 )

a 2 f y2 (4 f y4 3)

2 (3 f y2 )

a = 5 102
fy =

275 106 3
70 106 4

= 2.946

\
b 2 = (63 or 13.4) 104 or b = 7.9 102 m = 7.9 cm
Wall thickness t = 2.9 cm

Case (b) Plane Strain

When the cylinder is fairly long, sections that are far from the ends can be considered to be in a state of plane strain and we can assume that sz does not vary
along the z-axis. As in the case of plane stress, the equation is
d
(r r ) = 0
dr
From Hooke's law
e r = 1 [ r ( + z ) ]
E
e q = 1 [ ( r + z ) ]
E
e z = 1 [ z ( r + ) ]
E

Chapter_08.pmd

278

6/18/2008, 2:22 PM

Axisymmetric Problems

279

Since ez = 0 in this case, one has from the last equation

s z = n (sr + sq)
er =

1+
[(1 ) r ]
E

(8.18)

1+
[ (1 ) r ]
E
Solving for sq and sr
eq =

sq =

E
[ r + (1 ) ]
(1 2 ) (1 + )

(8.19a)

sr =

E
[(1 ) r + ]
(1 2 ) (1 + )

(8.19b)

On substituting for er and eq from Eqs (8.3) and (8.4), the above equations become
sq =

E
(1 2 ) (1 + )

ur
dur
dr + (1 ) r

(8.20)

sr =

E
(1 2 ) (1 + )

dur
ur

(1 ) dr + r

(8.21)

Substituting these in the equation of equilibrium, Eq. (8.5c)

d
dr

dur
dur
ur

(1 ) r dr + u r dr (1 ) r = 0

or

dur
d 2 ur ur
+r

=0
dr
r
dr 2

i.e.

d dur ur
+ =0
dr dr
r

This is the same as Eq. (8.7) for the plane stress case. The solution is the same as
in Eq. (8.8).

C2
r
where C1 and C2 are constants of integration. From Eqs (8.20) and (8.21)
u r = C1 r +

sq =

E
(1 2 ) (1 + )

C2

C1 + (1 2 ) r 2

(8.22a)

sr =

E
(1 2 ) (1 + )

C2

C1 (1 2 ) r 2

(8.22b)

Once again, we observe that sr + sq is a constant independent of r . Further, the

axial stress from Eq. (8.18) is
sz =

Chapter_08.pmd

2 E
C
(1 2 ) (1 + ) 1

279

(8.22c)

6/18/2008, 2:22 PM

280 Advanced Mechanics of Solids

Applying the boundary conditions
s r = pa when r = a,

s r = pb

when r = b

E
(1 2 ) (1 + )

C2

C1 (1 2 ) a 2 = pa

E
(1 2 ) (1 + )

C2

C1 (1 2 ) b 2 = pb

Solving,

C1 =

(1 2 ) (1 + ) pb b 2 pa a 2
E
a 2 b2

and

C2 =

1 + ( pb pa ) a 2 b2
E
a2 b2

Substituting these, the stress components become

sr =

sq=

pa a 2 pb b 2
b2 a 2
pa a 2 pb b 2
b2 a 2

s z = 2

pa pb a 2 b 2
b2 a 2

(8.23)

r2

pa pb a 2 b 2
b2 a 2

(8.24)

r2

pb a 2 pa b 2

(8.25)

b2 a2

It is observed that the values of sr and sq are identical to those of the plane
stress case. But in the plane stress case, sz = 0, whereas in the plane strain
case,sz has a constant value given by Eq. (8.25).

8.3

STRESSES IN COMPOSITE TUBESSHRINK FITS

The problem which will be considered now, involves two cylinders made of two
different materials and fitted one inside the other. Before assembling, the inner
cylinder has an internal radius a and an external radius c. The internal radius of
the outer cylinder is less than c by D, i.e. its internal radius is c D. Its external
radius is b. If the inner cylinder is cooled and the outer cylinder is heated, then
the two cylinders can be assembled, one fitting inside the other. When the cylinders come to room temperature, a shrink fit is obtained. The problem lies in determining the contact pressure pc between the two cylinders.
The above construction is often used to obtain thick-walled vessels to withstand high pressures. For example, if we need a vessel to withstand a pressure of
say 15000 atm, the yield point of the material must be at least 30000 kgf/cm2
(2940000 kPa). Since no such high-strength material exists, shrink-fitted composite
tubes are designed.
The contact pressure pc acting on the outer surface of the inner cylinder reduces
its outer radius by u1. On the other hand, the same contact pressure increases the
inner radius of the outer cylinder by u 2 . The sum of these two quantities,

Chapter_08.pmd

280

6/18/2008, 2:22 PM

Axisymmetric Problems

281

i.e. (u 1 + u2) must be equal to D, the difference in the radii of the cylinders. To
determine u1 and u2, we make use of Eq. (8.10), assuming a plane stress case.
For the inner tube
u1 =

or

1 1
E1

c 2 c + 1 + 1 a 2 c 2
pc 2

E1
c
c a2

pc

c a2

cpc
(1 1 ) c 2 + (1 + 1 ) a 2

E1 (c 2 a 2 )

u1 =

For the outer tube

u2 =
or

1 2
1 + 2 c 2b2
c2
+
p
c
c

E2
E2
c
b2 c 2

pc
2

b c2

cpc

(1 2 ) c 2 + (1 + 2 )b 2

E2 (b 2 c 2 )

u2 =

In calculating u2, we have neglected D since it is very small as compared to c.

Noting that u1 is negative and u2 is positive, we should have
u1 + u2 = D
cpc

(1 1 ) c 2 + (1 + 1 ) a 2

E1 ( c 2 a 2 )

i.e.

cpc

(1 2 )c2 + (1 + 2 )b2 =

E2 (b2 c 2 )

(8.26a)

Regrouping, the contact pressure pc is given by

pc =

/c
1 c + a + 1
1
E1 c 2 a 2
E2
2

b2 + c2

+ 2
2
2
b c

(8.26b)

If the two cylinders are made of the same material, then E1 = E2 and n1 = n2.
Equation (8.26) will then reduce to
pc =

2
2
2
2
E (c a ) (b c )
2c 3
(b 2 a 2 )

(8.27)

It is important to note that in Eqs (8.26) and (8.27), D is the difference in radii
between the inner cylinder and the outer jacket. Because of shrink fitting, therefore, the inner cylinder is under external pressure pc. The stress distribution in the
assembled cylinders is shown in Fig. 8.6.
If the composite cylinder made up of the same material is now subjected to an
internal pressure p, then the two parts will act as a single unit and the additional
stresses induced in the composite can be determined from Eqs (8.13) and (8.14).
At the inner surface of the inner cylinder, the internal pressure p causes a tensile
tangential stress sq , Eq. (8.14), but, the contact pressure pc causes at the same
points a compressive tangential stress, Eq (8.17). Hence, a composite cylinder can

Chapter_08.pmd

281

6/18/2008, 2:26 PM

282 Advanced Mechanics of Solids

sq
a

Fig. 8.6

sr
D

pc

Streses in composite tubes

support greater internal pressure than an ordinary one. However, at the inner
points of the jacket or the outer cylinder, the internal pressure p and the contact
pressure pc both will induce tensile tangential (i.e. circumferential) stresses sq .
For design purposes, one can choose the shrink-fit allowance D such that the
strengths of the two cylinders are equal. To determine this value of D, one can
proceed as follows.
Let a and c be the radii of the inner cylinder, and
c and b the radii of the jacket (see Fig. 8.7) c is
the common radius of the two cylinders at the
contact surface when the composite cylinder is
sq
a
+
experiencing an internal pressure p and the shrinkA B
sr
fit pressure pc. If the strengths of the two cylinc
b
ders are the same, then according to the maximum
shear stress theory, (s1 s3) at point A of the
inner cylinder should be equal to (s1 s3) at
point B of the outer cylinder. s1 and s3 are the
Fig. 8.7 Equal strength
maximum
and minimum normal stresses, which are
composite tube
respectively equal to sq and sr.
At point A, due to internal pressure p, from Eqs (8.13) and (8.14),
(sq sr)A = p

b2 + a 2
b2 a 2

= 2p

( p)

b2
b a2
2

Because of shrink-fitting pressure pc, at the same point, from Eqs (8.16) and (8.17),
(sq sr)A = 2 pc

c2
c a2
2

Hence, the resultant value of (sq sr) at A is

(sq sr)A = 2 p

b2 2 p
c2
c
b2 a2
c2 a2

(8.28)

At point B of the outer cylinder, since the composite involves the same material, due to
the pressure p, from Eqs (8.13) and (8.14), and observing that r = c in these equations,

Chapter_08.pmd

282

6/18/2008, 2:26 PM

Axisymmetric Problems

283

a 2 (c 2 + b 2 ) a 2 (c 2 b 2 )
(sq sr)B = p 2 2
2 2
2
2
c (b a ) c (b a )

= 2p

a2b2
c (b 2 a 2 )
2

At the same point B, due to the contact pressure pc, from Eqs (8.13) and (8.14),
with internal radius equal to c and external radius b,
c2 + b2 c2 b2
(sq sr)B = pc 2
2

2
b c2
b c
b2
b c2

= 2 pc

The resultant value of (sq sr) at B is therefore

(sq sr)B = 2 p

a 2b2
b2
+ 2 pc 2
2
2
(b c 2 )
c (b a )

(8.29)

For equal strength, equating Eqs (8.28) and (8.29)

2p

b2
c2
a 2b2
b2
p
p
p

2
=
2
+
2
c
c
c 2 (b 2 a 2 )
(b 2 a 2 )
(c 2 a 2 )
(b 2 c 2 )

b2

b2

c2
a 2b2
(8.30)
pc 2
p
+
=

2
2
2
2
2
2
2
2
(c a )
b c
(b a ) c (b a )
The shrink-fitting pressure pc is related to the negative allowance D through Eq.
(8.27) and it is this value of D that is required now for equal strength. Hence,
substituting for pc from Eq. (8.27), Eq. (8.30) becomes

or

2
2
2
2

b2

E (c a ) (b c ) b 2
c2
a 2b2
+
=
p

2
2
2 2
2
3
2
2
2
2
2
2
2c
(b a )
(b c ) (c a )
(b a ) c (b a )

or

2 2
2 2
4
pb 2 (c 2 a 2 )
E (2b c b a c )
=
2c 3
(b 2 a 2 )
c 2 (b2 a 2 )

D=

or

b 2 c (c 2 a 2 )
2p
E b2 (c 2 a 2 ) c 2 (b 2 c 2 )

(8.31a)

Also, from Eq. (8.30),

b 2 (c 2 a 2 )2 (b 2 c 2 )
c 2 (b 2 a 2 ) b 2 (c 2 a 2 ) + c 2 (b 2 c 2 )
The value of (sq sr) either at A or at B, from Eqs (8.28) and (8.31), is

pc = p

sq sr = p

Chapter_08.pmd

2b 2
(b 2 a 2 )

283

(c 2 a 2 ) (b 2 c 2 )
1 2 2
2
2
2
2
b ( c a ) + c (b c )

6/18/2008, 2:26 PM

(8.31b)

284 Advanced Mechanics of Solids

or

1
sq sr = p 22b 2 1
2
2
b
(b a )
+ 2c 2

2
2
b c
c a

(8.32)

Therefore, for composites made of the same material, in order to have equal strength
according to the shear stress theory, the shrink-fit allowance D that is necessary
is given by Eq. (8.31a), and this depends on the internal pressure p. Further, D
depends upon the difference between the external radius of the inner cylinder and
the internal radius of the jacket. In other words, this depends on c(+) and c().
With a, b and p fixed, one can determine the optimum value of c for minimum
(sq sr) at A and B. From Eq. (8.32), the minimum value of (sq sr) is obtained
when the denominator of the second expression within the square brackets is a
maximum, i.e. when D is a maximum, where
b2 + c2
b2 c 2 c 2 a 2
Differentiating with respect to c and equating the differential to zero,

D=

2c (c 2 a 2 ) 2c3
2cb2
dD =
+
=0
dc
(b2 c 2 )2
(c 2 a 2 )2

Simplifying, one gets

c = ab
The corresponding value of (sq sr), from Eq. (8.32), is

1
1

(sq sr)min = p 2b
(b 2 a 2 )
b2
ab
+

b(b a) a(b a)

= p
or

(sq sr)min = p

2b2
(b a 2 )
2

(b a )
1 2b

b
(b a)

(8.33)

Also, the optimum value of D is from Eq. (8.31a),

D opt =

p
1
pc =
E
E

(8.34)

ab

Example 8.4 Determine the diameters 2c and 2b and the negative allowance D
for a two-layer barrel of inner diameter 2a = 100 mm. The maximum pressure
the barrel is to withstand is pmax = 2000 kgf/cm2 (196000 kPa). The material
is steel with E = 2(10)6 kgf/cm2 (196 105 kPa); syp in tension or compression is
6000 kgf/cm2 (588 103 kPa). The factor of safety is 2.

Chapter_08.pmd

284

6/18/2008, 2:26 PM

Axisymmetric Problems

285

Solution From Eq. (8.33),

Therefore,

6000 = 2000 b
ba
2
b = 3a

Since c = ab , c = 3 a. The numerical values are therefore, 2a = 100 mm,

2b = 300 mm, 2c = 173 mm. With c = ab, the value of D is, from Eq. (8.34),
p
2000
ab =
(50 150) = 0.866 mm
D=
E
2 106
Example 8.5 A steel shaft of 10 cm diameter is shrunk inside a bronze cylinder of
25 cm outer diameter. The shrink allowance is 1 part per 1000 (i.e. 0.005 cm
difference between the radii). Find the circumferential stresses in the bronze cylinder
at the inside and outer radii and the stress in the shaft.

and

E steel = 2.18 106 kgf/cm2 (214 106 kPa)

E bronze = 1.09 106 kgf/cm2 (107 106 kPa)
n = 0.3 for both metals.

Solution In Eq. (8.26),

a = 0, c = 5, b = 12.5, D = 0.005, n1 = n2 = 0.3
Substituting in Eq. (8.26a),
5 pc
5 pc
(0.7 25) +
6
6
2.18 10 25
1.09 10 (156.25 25)

( 0.7 25 + 1.3 156.25) = 0.005

2

or,
pc = 610 kgf/cm (59780 kPa)
For the bronze tube, the circumferential stress is, from Eq. (8.14),
610 25 156.25
1+
sq =

(156.25 25)
r2
When r = 5 cm and r = 12.5 cm
sq = 842.4 kgf/cm2 (82555 kPa)
sq = 232.4 kgf/cm2 (22775 kPa)
The shaft experiences equal sr and sq at every point, from Eqs (8.16) and (8.17).
Hence,
s r = sq = 610 kgf/cm2 (59780 kPa)
Example 8.6 A compound cylinder made of copper inner tube of radii a = 10 cm
and c = 20 cm is snug fitted (D = 0) inside a steel jacket of external radius
b = 40 cm. If the compound cylinder is subjected to an internal pressure
p = 1500 kgf/cm2 (147009 kPa), determine the contact pressure pc and the values of
sr and sq at the inner and external points of the inner cylinder and of the jacket. Use
the following data:
E st = 2 106 kgf/cm2 (196 106 kPa),
Ecu = 1 106 kgf/cm2 (98 106 kPa),
nst = 0.3,

Chapter_08.pmd

285

ncu = 0.34

6/18/2008, 2:26 PM

286 Advanced Mechanics of Solids

Solution Since the initial shrink-fit allowance D is zero, the initial contact pressure is zero. When the compound cylinder is subjected to an internal pressure p,
the increase in the external radius of the copper cylinder under p and contact
pressure pc should be equal to the increase in the internal radius of the jacket
under the contact pressure pc,
i.e.

(ur ) p + (ur ) p at r = c = (u r ) p at r = c
c
c

cu
st

For copper cylinder, from Eq. (8.10),

1 cu pa 2 c
1 + cu a 2 c 2
p
+
(ur)p = E
2
2
Ecu
c (c 2 a 2 )
(c a )
cu

(u ) pc =

(ur)total =

1 cu pc c3
1 + cu a 2 c 2
pc

2
2
2
Ecu (c a )
Ecu
c (c a 2 )

pc c
2 pa 2 c
c 2 (1 cu ) + a 2 (1 + cu )

2
2

Ecu (c a ) Ecu (c 2 a 2 )

For steel jacket, from Eq. (8.10),

(u ) pc =
=

1 st
pc c3
1 + st c 2 b 2
pc
+
Est (b 2 c 2 )
Est
c (b 2 c 2 )
pc c

c 2 (1 st ) + b 2 (1 + st )

Est (b 2 c 2 )

Equating the (ur)s

pc c
2 pa 2 c
c 2 (1 cu ) + a 2 (1 + cu )

2
2

Ecu (c a ) Ecu (c 2 a 2 )
=

or

pc c

c 2 (1 st ) + b 2 (1 + st )

Est (b 2 c 2 )

(c 2 + a 2 ) cu (c 2 a 2 ) (b 2 + c 2 ) + st (b 2 c 2 )
+
pc

Ecu (c 2 a 2 )
Est (b 2 c 2 )

= p

Ecu

2a 2
(c 2 a 2 )

With p = 1500 kgf/cm , a = 10, c = 20, b = 40, nst = 0.3, ncu = 0.34,
500 300 0.34 2000 + 1200 0.3 3000 100
pc
+
=
300 106
2 1200 106 300 106

Chapter_08.pmd

pc = 433 kgf/cm2 (42453 kPa)

286

6/18/2008, 2:26 PM

Axisymmetric Problems

287

Now, pc will act as an external pressure on the copper tube and as an internal
pressure on the steel jacket. For copper tube, from Eqs (8.11) and (8.12),
(i) Inner surface:,
sr at r = a is 1500 kgf/cm2 and,
sq at r = a is
=

(1500 100) (433 400) 100 400 (1500 433)

+

(400 100)
100
(400 100)

= 1357 kgf/cm2 (compressive)

(ii) Outer surface:
sr at r = c is 433 kgf/cm2 and,
sq at r = c is
=

(1500 100) (433 400) 100 400 (1500 433)

+
(400 100)
400
(400 100)

= 279 kgf/cm2
For steel jacket, from Eqs (8.11) and (8.12),
(i) Inner surface:
sr at r = c is 433 kgf/cm2 and,
sq at r = c is
=

(433 400) 400 1600

433
=

(1600 400)
400
(1600 400)

= 722 kgf/cm2
(ii) Outer surface:
sr at r = b is zero and,
sq at r = b is
=

(433 400) 400 1600

433
+

(1600 400)
1600
(1600 400)

= 289 kgf/cm2

8.4

SPHERE WITH PURELY RADIAL DISPLACEMENTS

Consider a uniform sphere or spherical shell subjected to radial forces only, such
as internal or external pressures. The sphere or the spherical shell will then
undergo radial displacements only. Consider a particle situated at radius r before
deformation. After deformation, the spherical surface of radius r becomes a
surface of radius (r + ur) and the particle undergoes a displacement ur. Similarly,
another particle at distance (r + Dr) along the same radial line will undergo a
u

displacement ur + r r .
r

Hence, the radial strain is

er =

Chapter_08.pmd

ur
r

287

6/18/2008, 2:26 PM

288 Advanced Mechanics of Solids

Before deformation, the circumference of any great circle on the surface of radius
r is 2p r. After deformation, the radius becomes (r + ur) and the circumference of
the great circle is 2p (r + ur) Hence, the circumferential strain is

2 ( r + ur ) 2 r ur
=
2 r
r
This is the strain in every direction perpendicular to the radius r. Because of
complete symmetry, we can choose a frame of reference, as shown in Fig. 8.8.
ef =

sf
sr + Dsr

q
sr

f (ef)

r
(c)

r (er)

sf

sq

2q

q (eq)

s r + Dsr

sr
(a)

sf

2q
sq

sf

(b)

Fig. 8.8

Sphere with purely radial displacement

Thus, the three extensional strains along the three axes are

ur
,
r

ur
u
(8.35)
,
= r
r
r
Because of symmetry, there are no shear stresses and shear strains. Let gr be the
body force per unit volume in the radial direction.
The stress equations of equilibrium can also be derived easily. Consider a
spherical element of thickness Dr at distance r, subtending a small angle 2q at
the centre. Because of spherical symmetry, s q = s f. For equilibrium in the
radial direction,
er =

sr (2q r)(2q r) + (sr + Dsr) (r + Dr)2q (r + Dr)2q

2 r +

Putting sq = sf and Dsr =

r2

r
r
2r sin 2 r +
2 r sin + r 4 2 r 2 r = 0
2
2

r
r

r , the above equation reduces in the limit to

r
+ 2r r 2r + r 2 r = 0
r

Since r is the only independent variable, the above equation can be rewritten as
d 2
(r r ) 2r + r 2 r = 0
dr

Chapter_08.pmd

288

(8.36)

6/18/2008, 2:26 PM

Axisymmetric Problems

289

If body force is ignored,

1 d (r 2 ) = 2
r
r
r 2 dr
From Hookes law

(8.37)

e r = 1 r ( + )
E
or

dur
= 1 ( r 2 )
E
dr

(8.38)

and
e f = 1 ( + r )
E
or

ur
= 1 (1 ) r
E
r
Equations (8.37)(8.39) can be solved. From Eq. (8.39)

(8.39)

u r = 1 (1 )r r r
E
Differentiating with respect to r

dur
1 (1 ) d (r ) d (r r )
=
E
dr
dr
dr
Subtracting the above equation from Eq. (8.38)
d ( r )
d ( r r )
+
+ r 2
dr
dr
Substituting for sf from Eq. (8.36)

0 = (1 )

2
d 2 (r 2 r )
d (r r )
1
d (r r ) = 0
(1 )

+
r
2
dr
r
dr
dr 2
2
If r sr = y,

(8.40)

d (r ) = d y = 1 dy 1 y
r
dr
dr r r dr r 2
Therefore, Eq. (8.40) becomes
d 2 y dy y
y dy
1
1 ) 2
+ 2 2 +
=0
(
2
r
dr
r dr
dr
r
r
d2y
y
2 2 =0
2
dr
r
This is a homogeneous linear equation with the solution

or

(8.41)

y = Ar 2 + B
r

Chapter_08.pmd

289

6/18/2008, 2:26 PM

290 Advanced Mechanics of Solids

where A and B are constants. Hence,
sr = A +

B
r3

(8.42)

And from Eq. (8.37)

sf = 1 d Ar 2 + B = A B3
2r dr
r
2r
The constants A and B are determined from the boundary conditions.

(8.43)

Problem of Thick Hollow Sphere

Consider a spherical body formed by the boundaries of two spherical surfaces of
radii a and b respectively. Let the hollow sphere be subjected to an internal pressure pa and an external pressure pb . The boundary conditions are therefore
s r = pa when r = a, and sr = pb when r = b
From Eq. (8.42)
B
B
p a = A + 3 and pb = A + 3
a
b
Solving,
b3 pb a 3 pa

a 3b 3
( pb pa )
b a
b a3
Thus, the general expressions for sr and sf are

A=

sr =

1
b a3
3

B=

a 3b 3
3
b pb + a pa + 3 ( pb pa )
r

3
1
a 3b 3 p p
3

b
p
a
p
( b a )
b
a

b3 a 3
2r 3

If the sphere is subjected to internal pressure only, pb = 0, and

sf = =

s r = pa

a3
b3
1

b3 a 3
r3

(8.45)

(8.46)

a3
b3
1
+

b3 a 3
2r 3
The above two equations can also be written as
sf = = pa

s r = pa

(8.44)

(8.47)

a3
13
3 3
3

1 (a / b ) b
r

sf = = pa

1
a3
1
+

1 ( a 3 / b3 ) b3 2r 3

In the case of a cavity inside an infinite or a large medium, b and the above
equations reduce to
a3
s r = pa 3
r

Chapter_08.pmd

290

(8.48)

6/18/2008, 2:26 PM

Axisymmetric Problems

291

a3
(8.49)
2r 3
The above equations can also be used to calculate stresses in a body of any
shape with a spherical hole under an internal pressure pa, provided the outer
surface of the body is free from pressure and provided that every point of this
outer surface is at a distance greater than four or five times the diameter of the
hole from its centre.

s f = = + pa

Example 8.7 Calculate the thickness of the shell of a bomb calorimeter of

spherical form of 10 cm inside diameter if the working stress is s kgf/cm2
(98 s kPa) and the internal pressure is s/2 kgf/cm2 (49 s kPa).

Solution From equations (8.46) and (8.47), the maximum tensile stress is due to
sf, which occurs at r = a. Hence,
3
3

sf = 3 5 3 1 + b 3
2 b 5
25
Equating this to the working stress s

53
b 53
3

b3
1 +
=2
2 53

\
b 6.3 cm
Hence, the thickness of the shell is 1.3 cm.
Example 8.8

Express the stress equation of equilibrium, i.e.

1 d (r 2 ) 2 = 0
r
r
r 2 dr
given by Eq. (8.37), in terms of the displacement component ur, using Hooke's
law and straindisplacement relations.

Solution We have er = 1 ( r 2 )
E

e f = 1 (1 ) r
E
Solving for sr and sf,
sr =

E
(1 ) + 2

(1 + )(1 2 ) r

(8.50)

sf =

E
( r + )
(1 + )(1 2 )

(8.51)

Using the straindisplacement relations

er =

Chapter_08.pmd

dur
dr

291

and =

ur
r

6/18/2008, 2:26 PM

292 Advanced Mechanics of Solids

and substituting in the equilibrium equation, we get

E (1 )
(1 + )(1 2 )

u
E (1 )
d dur
+2 r =0
r
(1 + )(1 2 ) dr dr

or

8.5

d 2 ur 2 dur
2
2 + r dr 2 ur = 0
r
dr

(8.52)

STRESSES DUE TO GRAVITATION

When body forces are operative, the stress equation of equilibrium is, from
Eq. (8.36),
1 d r 2 2 + = 0
(8.53)
r
r
r
r 2 dr
where gr is the body force per unit volume. The problem of a sphere strained by
the mutual gravitation of its parts will now be considered . It is known from the
theory of attractions

gr = g r
a
where a is the radius of the sphere, r is the mass density, r is the radius of any
point from the centre and g is the acceleration due to gravity. Expressing the
equations of equilibrium in terms of displacement ur [Eq. (8.52)], we have
u
E (1 )
d dur
r
+ 2 r g = 0
r
a
(1 + ) (1 2 ) dr dr

(8.54)

The complementary solution is

u r = Cr +

C1
r2

and the particular solution is

(1 + ) (1 2 )
ur = 1
gr 3
10 E (1 ) a

Hence, the complete solution is

u r = Cr +

C1
r

1 (1 + ) (1 2)
gr 3
10 E (1 )a

For a solid sphere, C1 should be equal to zero as otherwise the displacement will
become infinite at r = 0. The remaining constant is determined from the boundary
condition sr = 0 at r = a. From the general solution

dur
3(1 + ) (1 2 )
= C+
gr 2 ,
10 E (1 ) a
dr
and from Eq. (8.50)
sr =

Chapter_08.pmd

ur
(1 + ) (1 2 )
= C+
gr 2
10 E (1 ) a
r

E
(1 ) + 2

(1 + ) (1 2 ) r

292

6/18/2008, 2:26 PM

Axisymmetric Problems

u
dur
E
(1 ) + 2 r
r
(1 + ) (1 2 ) dr

E
(1 + ) (1 2 )

293

3(1 + ) (1 2 )

gr 2 + 2C
C (1 ) +
10 Ea
+

(1 + ) (1 2 )

gr 2
5E (1 )a

(3 ) (1 + ) (1 2 )

gr 2
C (1 + ) +
10 Ea (1 )

From the boundary condition sr = 0 at r = a,

=

E
(1 + ) (1 2 )

C=
Hence,

(3 ) (1 2 )
ga
10 E (1 )

(3 ) 2
g
sr = 1
(a r 2 )
10 (1 )
a

(8.55)

and from Eq. (8.53)

(3 )a 2 (1 + 3 ) r 2 g
sf = = 1
10
(1 )
a

(8.56)

It will be observed that both stress components sr and sq are compressive at

every point. At the centre (r = 0), they are equal and have a magnitude
3
sr = = = 1
ga (compressive)
10 1
Further,
3(1 + ) (1 2 )
dur
= C+ 1
gr 2
10
E (1 )a
dr
(3 ) (1 2 )
3(1 + ) (1 2 )
= 1
ga + 1
gr 2
10
10
E (1 )
E (1 )a

(1 + ) (1 2 ) g 2 (3 ) 2
= 1
3r
a
10
E (1 )
a
(1 + )
The above value is zero when
r2 =

(3 ) 2
a
3(1 + )

Hence, if n is positive (which is true for all known materials), there is a definite
surface outside which the radial strain is an extension. In other words, for
1/ 2

(1 + )
r > a
3(1 + )
the radial strain er is positive though the radial stress sr is compressive everywhere. This result is due, of course, to the Poisson effect of the large circumferential stress, i.e. hoop stress, which is compressive.

Chapter_08.pmd

293

6/18/2008, 2:26 PM

294 Advanced Mechanics of Solids

8.6

ROTATING DISKS OF UNIFORM THICKNESS

We shall now consider the stress distribution in rotating circular disks which are
thin. We assume that over the thickness, the radial and circumferential stresses
do not vary and that the stress sz in the axial direction is zero. The equation of
equilibrium given by Eq. (8.5b) can be used, provided we add the inertia force
term rw 2r, i.e. in the general equation of equilibrium [Eq. (8.2)] we put the body
force term equal to the inertia term rw 2r, where w is the angular velocity of the
rotating disk and r is the density of the disk material. The z-axis is the axis of
rotation. Then;

r r
+
+ 2 r = 0
r
r

(8.57a)

d r + 2 r 2 = 0
( r)
dr
The strain components are, as before,

or

(8.57b)

dur
u
and = r
dr
r
From Hookes law, with sz = 0,

(8.58)

er =

e r = 1 ( r )
E
e q = 1 ( r )
E
From Eq. (8.58)
e r = d ( r )
dr
From Hookes law
1 ( ) = e = d (r ) = 1 d (r r )
r

r
E dr
dr
E r

(8.59)

Let
rsr = y
Then, from Eq. (8.57b)

(8.60a)

dy
+ 2 r 2
dr
Substituting these in Eq. (8.59) and rearranging

(8.60b)

d2y
dy
+r
y + (3 + ) 2 r3 = 0
2
dr
dr
The solution of the above differential equation is
r2

(8.61)

(3 + )
y = Cr + C1 1
2 r3
r
8

(8.62)

(3 + )
s r = C + C1 12
2 r3
8
r

(8.63)

From Eq. (8.60)

Chapter_08.pmd

294

6/18/2008, 2:26 PM

Axisymmetric Problems

(1 + 3 )
sq = C C1 12
2 r3
8
r
The integration constants are determined from boundary conditions.

295
(8.64)

Solid Disk
For a solid disk, we must take C1 = 0, since otherwise the stresses sr and
sq become infinite at the centre. The constant C is determined from the condition
at the periphery (r = b) of the disk. If there are no forces applied there, then,
(sr)r = b = C

3+
2b 2 = 0
8

Hence,
3+
2b 2
8
and the stress components become

C=

sr =

3+
2 (b 2 r 2 )
8

(8.65a)

3+
1 + 3
(8.65b)
2b 2
2r 2
8
8
These stresses attain their maximum values at the centre of the disk, where

sq =

sr = =

3+
2b 2
8

(8.66)

Circular Disk with a Hole of Radius a

If there are no forces applied at the boundaries a and b, then
(sr)r=a = 0,
(sr)r=b = 0
from which we find that
3+
2 b 2 + a 2 ,
8
Substituting these in Eqs (8.63) and (8.64)

C=

sr =

sq =

C1 =

3+
2a 2b 2
8

2 2
3+

2 b2 + a 2 a b2 r 2
8

(8.67)

2 2
3+

1 + 3 2
r
2 b 2 + a 2 + a b2
8
3+

The radial stress sr reaches its maximum at r =

ab where

3+
2 (b a) 2
8
The maximum circumferential stress is at the inner boundary, where
3+
1 2

(sq)max =
2 b 2 +
a
4
3 +

It can be seen that (sq)max is greater than (sr)max.

(sr)max =

Chapter_08.pmd

295

(8.68)

6/18/2008, 2:26 PM

(8.69)

(8.70)

296 Advanced Mechanics of Solids

When the radius a of the hole approaches zero, the maximum circumferential
stress approaches a value twice as great as that for a solid disk [Eq. (8.66)]. In
other words, by making a small circular hole at the centre of a solid rotating disk,
we double the maximum stress.
The displacement ur for all the cases considered above can be calculated from
Eq. (8.58), i.e.
u r = r = r ( r )
(8.71)
E
Example 8.9 A flat steel disk of 75 cm outside diameter with a 15 cm diameter
hole is shrunk around a solid steel shaft. The shrink-fit allowance is 1 part in 1000
(i.e. an allowance of 0.0075 cm in radius). E = 2.18 106 kgf/cm2 (214 106 kPa).
(i) What are the stresses due to shrink-fit?
(ii) At what rpm will the shrink-fit loosen up as a result of rotation?
(iii) What is the circumferential stress in the disk when spinning at the above
speed?
Assume that the same equations as for the disk are applicable to the solid
rotating shaft also.

Solution
(i) To calculate the shrink-fit pressure, we have from Eq. (8.27)
pc =

2.18 106 0.0075 (7.52 0) (37.52 7.52 )

2 7.53
(37.52 0)

or
pc = 1044 kgf/cm2 (102312 kPa)
The tangential stress at the hole will be the largest stress in the system
and from Eq. (8.24)
sq =

1044 7.52
37.52
1
+

(37.52 7.52 )
7.52

= 1131 kgf/cm2 (110838 kPa)

(ii) When the shrink-fit loosens up as a result of rotation, there will be no
radial pressure on any boundary. When the shaft and the disk are rotating,
the radial displacement of the disk at the hole will be greater
than the radial displacement of the shaft at its boundary. The difference
between these two radial displacements should equal D = 0.0075 cm at 7.5
cm radius. From Eqs (8.71), (8.67) and (8.68)
u disk = r ( r )
E
2 2
3+

1 + 3 2
= r
2 b 2 + a 2 + a b2
r
E 8
3+
r

2 2

b2 + a 2 a b2 r 2

Chapter_08.pmd

296

6/18/2008, 2:28 PM

Axisymmetric Problems

297

( 3 + ) (1 ) 2 b 2 + a 2 + 1 + b 2 a 2 1 + r 2
= r

1 r2
3 +
E
8
3.3 0.7
7.5
2
=

6
8
2.18 10
2
2

1.3 37.5 7.5 1.3 7.52

37.52 + 7.52 +

0.7
3.3
7.52

= 4052 106 rw 2
From equations (8.71), (8.65a) and (8.65b)
ushaft = r ( r )
E
=

1
2 r (3 + ) b 2 (1 + ) r 2
8E

0.7
2 7.5 (3.3 7.52 1.3 7.52 )
8 2.18 106

= 34 106 rw 2
Therefore,
(4052 34) 106 rw 2 = 0.0075
or
w 2 = 0.0075 106

1 981
4018 0.0081

= 226066 (rad/s)2
Therefore,
w = 475 rad/s or 4536 rpm
(iii) The stresses in the disk can be calculated from Eq. (8.68)

s r = 3.3 2 37.52 + 7.52 + 37.52 1.9 7.52

8
3.3

= 1170 rw 2
= 1170 0.0081 226066
981
= 2184 kgf/cm2 (214024 kPa)

Example 8.10 A flat steel turbine disk of 75 cm outside diameter and 15 cm inside
diameter rotates at 3000 rpm, at which speed the blades and shrouding cause a
tensile rim loading of 44kgf/cm2 (4312 kPa). The maximum stress at this speed is to
be 1164 kgf/cm2 (114072 kPa). Find the maximum shrinkage allowance on the
diameter when the disk and the shift are rotating.

Solution Let c be the radius of the shaft and b that of the disk. From Eq. (8.70),
the maximum circumferential stress due to rotation alone is
(sq)1 =

Chapter_08.pmd

3+
1 2

2 b 2 +
c
4
3 +

297

6/18/2008, 2:28 PM

298 Advanced Mechanics of Solids

= 3.3 2 37.52 + 0.7 7.52

4
3.3

= 1170 rw 2
Owing to shrinkage pressure pc, and the tensile rim loading pb, from Eq. (8.12)
(sq)2 =

2
pc c 2
b2 + 2 pb b
+
1

b2 c2
c2
b2 c2

= pc

7.52
37.52 + 2 44 37.5
+
1

37.52 7.52
7.52
37.52 7.52

= 1.08pc + 91.7
Hence, the combined stress at 7.5 cm radius is
sq = 1170rw 2 + 1.08pc + 91.7
This should be equal to 1164 kgf/cm2. Hence,
1.08 pc = 1164 1170rw 2 91.7
2
= 1164 1170 (100 ) 0.0081 91.7
981
= 1164 953.5 91.7
= 118.8
Hence,
pc = 110 kgf/cm2
The corresponding shrink-fit allowance is obtained from Eq. (8.27), i.e.

7.52 37.52 7.52

E

110 =
2 7.53
37.52

= 0.064 ED
or

D=

110 106
= 0.0008 cm
0.064 2.18

8.7 DISKS OF VARIABLE THICKNESS

Assuming that the stresses do not vary over the thickness of the disk,
the method of analysis developed in the previous section for thin disks of
constant thickness can be extended also to disks of variable thickness. Let h be
the thickness of the disk, varying with radius r. The equation of equilibrium can
be obtained by referring to Fig. 8.9.
For equilibrium in the radial direction

) (

) (

( h r )
r
h 2
r
h r + r r ( r + r ) + r + 2 h + 2 r + 2

h r r 2 h + h r sin = 0
2
2

Chapter_08.pmd

298

6/18/2008, 2:28 PM

Axisymmetric Problems

299

Dr

r
Dq

Fig 8.9 Rotating disk of variable thickness

Simplifying and going to the limit
r d (h r ) + h r + 2 r 2 h h = 0
dr
d (r h ) h + 2 r 2 h = 0
r

dr
Putting
y = rhsr

or

(8.72)
(8.73a)

dy
+ h 2 r 2
dr
The strain components remain as in Eq. (8.58), i.e.
dur
u
er =
and = r
dr
r
d
Hence,
er =
(r )
dr
From Hookes law and Eq. (8.59)

hs q =

(8.73b)

1 ( ) = 1 d (r r )

r
E r
E dr
Substituting for sr and sq from Eqs (8.73a) and (8.73b)
d2y
dy
dy

+r
y + (3 + ) 2 hr 3 r dh r
y = 0
(8.74)

dr
h dr dr
dr 2
In the particular case where the thickness varies according to the equation
y = Crn
(8.75)
in which C is a constant and n any number, Eq. (8.74) can easily be integrated.
The general solution has the form
y = mrn + 2 + Ara + Brb
r2

in which

m=

(3 + ) 2 c
( n + 3n + 8)

and a and b are the roots of the quadratic equation

x2 nx + rn 1 = 0
A and B are constants which are determined from the boundary conditions.

Chapter_08.pmd

299

6/18/2008, 2:28 PM

300 Advanced Mechanics of Solids

Example 8.11

Determine the shape for a disk with uniform stress, i.e. sr = sq .

Solution From Hooke's law

e r = 1 ( r ) ,
= 1 ( r )
E
E
if sr = sq then er = eq . From straindisplacement relations
er =

dur
dr

and =

ur
r

e r = d ( r ) = d ( r r )
dr
dr
Since er = eq , the above equation gives
we get

i.e.

dr
=0
dr
e r = constant

Hence, from Hooke's law, sr and sq are not only equal but also constant throughout the disk. Let sr = sq = s. Equilibrium Eq. (8.73) gives
hs = d (rh ) + h 2 r 2
dr
= h + r dh + 2 hr 2
dr
1
dh
1
2
or
= r
h dr

which upon integration gives

log h =
or

8.8

2 2
r + C1
2

2
2 2

r + C1 = C exp 2 r
h = exp
2

ROTATING SHAFTS AND CYLINDERS

In Sec. 8.5 and 8.6, we assumed that the disk was thin and that it was in a state
of plane stress with sz = 0. It is also possible to treat the problem as a plane
strain problem as in the case of a uniformly rotating long circular shaft or a
cylinder. Let the z-axis be the axis of rotation. The equation of equilibrium is the
same as in Eq. (8.57):
d (r ) + 2 r 2 = 0
r

dr
The strain components are, as before,

er =

Chapter_08.pmd

dur
,
dr

300

ur
,
r

(8.76)

z =

uz
=0
z

(8.77)

6/18/2008, 2:28 PM

Axisymmetric Problems

301

From Hooke's law

e r = 1 [ r ( + z )]
E
e q = 1 [ ( r + z )]
E
e z = 1 [ z ( r + ) ]
E
Since ez = 0 (plane strain),
s z = ( r + )
and hence, substituting in equations for er and eq
er =

1+
(1 ) r ]
E [

1+
(1 ) r ]
E [
From straindisplacement relations given in Eq. (8.77)

eq =

e r = d (r )
dr
and using the above expressions for er and eq , we get
d
[(1 )r r r ]
dr
With rsr = y, Eq. (8.76) gives for sq
(1 ) r =

(8.78)

dy
+ 2 r 2
dr
Substituting for sr and sq in Eq. (8.78)

sq =

(1 )

y
dy

2 r 2 = d
r
dr
dr

dy
2 3
(1 ) r dr + r y

3 2
d2y
dy
+r
y+
2 r 3 = 0
2
1

dr
dr
The solution for this differential equation is
or

r2

(3 2 )
y = Cr + C1 1
2 r 3
r 8(1 )

Hence,

and

Chapter_08.pmd

(3 2 )
s r = C + C1 12
2 r 2
8(1

(8.79a)

(1 + 2 )
sq = C C1 12
2 r 2
8(1

(8.79b)

1
s z = 2C
2 r 2
2(1 )

(8.79c)

301

6/18/2008, 2:28 PM

302 Advanced Mechanics of Solids

(i) For a hollow shaft or a long cylinder, sr = 0 at r = a and r = b which are the
inner and outer radii. From these
C = K (a 2 + b 2 ) and C1 = Ka 2b 2
where

K=

(3 2 )
2
8 (1 )

Hence, from Eqs (8.79ac)

sr =

(3 2 )
8 (1 )

2
a 2b 2 r 2 2
2
(
a
+
b
)

r2

(8.80)

sq =

(3 2 )
8 (1 )

2
a 2b 2 1 + 2 2
2
2
( a + b ) + 2 3 2 r
r

(8.81)

sz =

(a 2 + b 2 ) (3 2 ) 2r 2 2

4 (1 )

(8.82)

sq assumes a maximum value at r = a and its value is

(sq)max =

(3 2 ) 2
1 + 2 2
2b + a 2
a 2
8 (1 )
3 2

If a2/b2 is very small, we find that

(sq)max

(3 2 ) 2 2
b
4 (1 )

(8.83)

(ii) For a long solid shaft, the constant C1 must be equal to zero, since otherwise
the stresses would become infinite at r = 0. Using the other boundary condition
that sr = 0 when r = b, the radius of the shaft, we find that
C=

(3 2 )
2b 2
8 (1 )

Hence, the stresses are

sr =

(3 2 ) 2
(b r 2 ) 2
8 (1 )

(8.84)

sq =

(3 2 ) 2 1 + 2 2
b
r 2
8 (1 )
3 2

(8.85)

sz =

b 2 (3 2 ) 2r 2 2

4 (1 )

(8.86)

The value of sq at r = 0 is
(sq)max =

(3 2 ) 2 2
b
8 (1 )

(8.87)

Comparing Eq. (8.87) with Eq. (8.83), we find that by drilling a small hole
along the axis in a solid shaft, the maximum circumferential stress is doubled
in its magnitude.

Chapter_08.pmd

302

6/18/2008, 2:28 PM

Axisymmetric Problems

303

Example 8.12 A solid steel propeller shaft, 60 cm in diameter, is rotating at a speed

of 300 rpm. If the shaft is constrained at its ends so that it cannot expand or contract
longitudinally, calculate the total longitudinal thrust over a cross-section due to
rotational stresses. Poissons ratio may be taken as 0.3. The weight of steel may be taken
as 0.0081 kgf/cm3 (0.07938 N/cm3).

Solution The total axial force is

b

Fz = z 2 r dr
0

and from Eq. (8.86), substituting for sz,

Fz =

b2 (3 2 ) b 2 b4 2

4(1 )

= b4 2
2
Substituting the numerical values
0.3 30 4 0.0081 300 2 42
2
981
602
= 3120 kgf (31576 N) Tensile force

Fz =

8.9 SUMMARY OF RESULTS FOR USE IN PROBLEMS

(i) For a tube of internal radius a and external radius b subjected to an internal
pressure pa and an external pressure pb, the radial and circumferential stresses are
given by (according to plane stress theory)
sr =
sq =

pa a 2 pbb 2
2

b a

pa a 2 pbb 2
2

b a

2 2 p p
a
b
a b2
r b2 a 2

2 2 p p
a
b
+ a b2
r b2 a 2

sz = 0
The stress sr < 0 for all values of pa and pb, whereas sq can be greater or less
than zero depending on the values of pa and pb. sq is greater than zero if
pa >

pb b 2

+ 1
2 a 2

The maximum and minimum stresses are

(sr)max = r (at r = b) = pb
(sr)min = r (at r = a ) = pa
(sq)max = (at r = a) =

Chapter_08.pmd

303

pa (a 2 + b 2 ) 2 pb b 2
b2 a 2

6/18/2008, 2:28 PM

304 Advanced Mechanics of Solids

(sq)min = (at r = b) =
If
then,

pa =

2 pa a 2 pb (a 2 + b 2 )
b2 a 2

1 p b 2 + 1

2 b a 2

(sq)min = 0
2
b2 + a 2

(ur)r = a = a pa 2
+ 2 pb 2 b 2
2
E b a
b a

b2 + a 2

(ur)r = b = b 2 pa 2 a 2 pb 2

2
E
b a
b a

(ii) Built-up cylinders: When the cylinders are of equal length, the contact pressure pc due to difference D between the outer radius of the inner tube and the
inner radius of the outer tube is given by
pc =

/c
1 c + a

1 + 1
2
2
E2
E1 c a
2

b2 + c2

b2 c 2 2

where E1, n1, a and c refer to the inner tubes modulus, Poissons ratio, inner
radius and outer radius respectively. E2,n2, c and b are the corresponding values
for the outer tube.
If E1 = E2 and n1 = n2, then

(b 2 c 2 )(c 2 a 2 )
pc = E3
b2 a 2
2c
(iii) For a sphere subjected to an internal pressure pa and an external pressure pb,
the radial and circumferential stresses are given by
sr =

1
b a3
3

a 3b 3
3
b pb + a pa + 3 ( pb pa )
r

3
a 3b 3 ( p p )
3
b
a
b pb + a pa
2r 3

(iv) For a thin solid disk of radius b rotating with an angular velocity w, the
stresses are given by

sq = =

sr =

1
b a3
3

3+
2 (b2 r 2 )
8

3+
1 + 3
2b 2
2 r 2
8
8
These stresses attain their maximum values at the centre r = 0, where

sq =

sq = r =

Chapter_08.pmd

304

3+
2b 2
8

6/18/2008, 2:28 PM

Axisymmetric Problems

305

The radial outward displacement at r = b is

1
2b3
4E
(v) For a thin disk with a hole of radius a, rotating with an angular velocity w, the
stresses are

(ur)r = b =

sr =

2 2
3+

2 b 2 + a 2 a b2 r 2
8

sq =

3+
2
8

2
a 2b 2 1 + 3 2
2
b + a + 2 3 + r
r

(sr)max = r at r = ab =
(sq)max = (at r = a) =
and

3+
2 (b a )2
8

3+
2
4

2 1 + 2
b + 3 + a

(sq)max > (sr)max

The radial displacements are

(ur)r = a =

3+
1 2

2 a b 2 +
a
4E
3 +

(ur)r = b =

3+
1 2

2b a 2 +
b
4E
3 +

8.1 A thick-walled tube has an internal radius of 4 cm and an external radius of

8 cm. It is subjected to an external pressure of 1000 kPa (10.24 kgf/cm2). If
E = 1.2 108 kPa (1.23 106 kgf/cm2) and n = 0.24, determine the internal
pressure according to Mohrs theory of failure, which says that

( )max n ( ) min tenslie strength

where n is the ratio of s-tensile strength to s-compressive strength. For the
present problem, assume s-tensile strength = 30000 kPa (307.2 kgf/cm2) and
s-compressive strength = 120000 kPa (1228.8 kgf/cm2).
[Ans. p = 17000 kPa (174 kgf/cm2)]
8.2 In the above problem, determine the changes in the radii.

Ans. Dr1 = 0.01 mm

Dr2 = 0.007 mm
8.3 In Example 8.1, if one uses the energy of distortion theory, what will be the
external radius of the cylinder? The rest of the data remain the same.
[Ans. = 6.05 cm]
8.4 A thick-walled tube with an internal radius of 10 cm is subjected to an
internal pressure of 2000 kgf/cm2 (196000 kPa). E = 2 106 kgf/cm2

Chapter_08.pmd

305

6/18/2008, 2:28 PM

306 Advanced Mechanics of Solids

(196 106 kPa) and n = 0.3. Determine the value of the external radius if the
maximum shear stress developed is limited to 3000 kgf/cm2 (294 106 kPa).
Calculate the change in the internal radius due to the pressure.
Ans. r2 = 17.3 cm

Dr1 = 0.023 cm
8.5 A thick-walled tube is subjected to an external pressure p2. Its internal and
external radii are 10 cm and 15 cm respectively, n = 0.3 and E = 200000 MPa
(2041 103 kgf/cm2). If the maximum shear stress is limited to 200000 kPa
(2041 kgf/cm2), determine the value of p2 and also the change in the external
radius.
Ans. p 2 = 111 MPa (1133 kgf/cm2)

Dr2 = 0.19 mm
8.6 Determine the pressure p0 between the concrete tube and the perfectly rigid
core. Assume Ec = 2 106 kgf/cm2, rc = 0.16. Take r1/r2 = 0.5 (Fig. 8.10).
[Ans. p0 = 17.4 kgf/cm2]

r1
p = 12 kgf/cm 2 (1126 kPa)

r2

Fig. 8.10 Problem 8.6

8.7 Determine the dimensions of a two-piece composite tube of optimum
dimensions if the internal pressure is 2000 kgf /cm 2 (196000 kPa),
external pressure p2 = 0, internal radius r1 = 8 cm and E = 2 106 kgf/cm2
(196 106 kPa). The maximum shear stress is to be limited to 1500 kgf/cm2
(147 10 6 kPa). Check the strength according to the maximum shear
theory.
Ans. r2 14 cm; r3 = 24 cm

D = 0.014 cm

pc = 500 kgf/cm2

(49030 kPa)

8.8 Determine the radial and circumferential stresses due to the internal pressure p = 2000 kgf/cm2 (196,000 kPa) in a composite tube consisting of an
inner copper tube of radii 10 cm and 20 cm and an outer steel tube of
external radius 40 cm.nst = 0.3, ncu = 0.34, Est = 2 106 kgf/cm2 (196 106 kPa)
and Ecu = 106 kgf/cm2 (98 106 kPa). Calculate the stresses at the inner and
outer radius points of each tube. Determine the contact pressure also.

Chapter_08.pmd

306

6/18/2008, 2:28 PM

Axisymmetric Problems

307

Ans. For inner tube:

2
s r = 2000 kgf/cm (196000 kPa)

s r = 577 kgf/cm2 (56546 kPa)

s t = 1800 kgf/cm2 (176400 kPa)

s t = 371 kgf/cm2 (36358 kPa)

For outer tube:

s r = 577 kgf/cm2 (56546 kPa)

sr = 0
2

s t = 962 kgf/cm (94276 kPa)

2
s t = 385 kgf /cm (37730 kPa)

pc = 577 kgf/cm2 (56546)

8.9 In problem 8.7, if the inner tube is made of steel (radii 10 cm and 20 cm) and
the outer tube is of copper (outer radius 40 cm), determine the circumferential and radial stresses at the inner and outer radii points of each tube.
Ans. For inner tube:

s r = 2000 kgf/cm2 (196000 kPa)

s r = 248 kgf/cm2 (24304 kPa)

s t = 2672 kgf/cm2 (262032 kPa)

s t = 920 kgf/cm2 (90221 kPa)

For outer tube:

2
s r = 248 kgf/cm (24304 kPa)

sr = 0

s t = 413 kgf/cm2 (40474 kPa)

2

s t = 165 kgf/cm (16170 kPa)

2
pc = 248 kgf/cm (24304 kPa)

8.10 A composite tube is made of an inner copper tube of radii 10 cm and 20 cm

and an outer steel tube of external radius 40 cm. If the temperature of the
assembly is raised by 100C, determine the radial and tangential stresses at the
inner and outer radius points of each tube. acu = 16.5 10 6; ast = 12.5 10 6;
Est, nst, Ecu and ncu are as in Problem 8.

Ans. For inner tube:

sr = 0

2
s r = 173 kgf/cm (16954 kPa)

s t = 461 kgf/cm2 ( 45080 kPa)

s = 288 kgf/cm2 (28243 kPa)

For outer tube:

2
s r = 173 kgf/cm (16954 kPa)

sr = 0

s t = 288 kgf/cm2 (28243 kPa)

s t = 115 kgf/cm2 (11270 kPa)

8.11 Determine for the composite three-piece tube (Fig. 8.11):
(a) Stresses due to the heavy-force fits with interferences of D1 = 0.06 mm and
D2 = 0.12 mm in diameters
(b) Stresses due to the internal pressure p = 2400 kgf/cm2
r1 = 80 mm, r2 = 100 mm, r3 = 140 mm, r4 = 200 mm,
E = 2.2 106 kgf/cm2.

Chapter_08.pmd

307

6/18/2008, 2:28 PM

308 Advanced Mechanics of Solids

r4

r3

r1 r2

Fig 8.11 Problem 8.11

8.12 The radial displacement at the outside of a thick cylinder subjected to an
internal pressure pa is

(2 p r r )
E (r r )
2
a b a

2
b

2
a

By Maxwell's reciprocal theorem, find the inward radial displacement at the

inside of a thick cylinder subjected to external pressure.

2 pb ab 2
Ans. ur =

E (b 2 a 2 )

8.13 A thin spherical shell of thickness h and radius R is subjected to an internal

pressure p. Determine the mean radial stress, the circumferential stress and
the radial displacement.
Ans. u = pR 2 (1 )/(2 Eh)
r

pR

sq = =

2h

(sr) average = 1 p

8.14 An infinite elastic medium with a spherical cavity of radius R is subjected to

hydrostatic compression p at the outside. Determine the radial and circumferential stresses at point r. Show that the circumferential stress at the surface of the cavity exceeds the pressure at infinity.

R2
Ans. s r = p 1 3

R3
sq = sf = p 1 + 3

2r

(sq) at cavity = 3 p
2

Chapter_08.pmd

308

6/18/2008, 2:28 PM

Axisymmetric Problems

309

8.15 A perfectly rigid spherical body of radius a is surrounded by a thick spherical shell of thickness h. If the shell is subjected to an external pressure p,
determine the radial and circumferential stresses at the inner surface of the
shell (b = a + h).

3(1 )b3 p
Ans. r =
3
3
2(1 2 )a + (1 + )b

8.16 A steel disk of 50 cm outside diameter and 10 cm inside diameter is shrunk

on a steel shaft so that the pressure between the shaft and disk at standstill
is 364 kgf/cm2 (3562 kPa). Take r = 0.0081/g kgm/cm3.
(a) Assuming that the shaft does not change its dimensions because of
its own centrifugal force, find the speed at which the disk is just free
on the shaft.
(b) Solve the problem without making assumption (a).
Ans. (a) 4013 rpm

(b) 4028 rpm.

8.17 A steel disk of 75 cm diameter is shrunk on a steel shaft of 7.5 cm diameter.
The interference on the diameter is 0.0045 cm
(a) Find the maximum tangential stress in the disk when it is at a standstill.
(b) Find the rotation speed at which the contact pressure is zero.
(c) What is the maximum tangential stress at the above speed.

Ans. (a) 647 kgf/cm2 (6349 kPa)

(b) 4990 rpm

2
(c)
2622
kgf/cm
(257129
kPa)

8.18 A disk of thickness t and outside diameter 2b is shrunk on to a shaft of

diameter 2a, producing a radial interface pressure p in the non-rotating condition. It is then rotated with an angular velocity w rad/s. If f is the coefficient of friction between disk and shaft and w0 is the value of the angular
velocity for which the interface pressure falls to zero, show that

(a) the maximum horsepower is transmitted when w = w 0 / 3 and

(b) this maximum horsepower is equal to 0.000366 a2t fpw0, where dimensions are in inches and pounds.
8.19 A steel shaft of 7.5 cm diameter has an aluminium disk of 25 cm outside
diameter shrunk on it. The shrink allowance is 0.001 cm/cm. Calculate the
rpm of rotation at which the shrink-fit loosens up. Neglect the expansion of
the shaft caused by rotation. nal = 0.3, Est = 7.3 105 kgf/cm2 (175 105 kPa);
g = 2.76 103 kgf/cm3.
[Ans. 13420 rpm]

Chapter_08.pmd

309

6/18/2008, 2:28 PM