Strength of Materials: An Undergraduate Text

Graham M. Seed

Saxe-Coburg Publications, 2001

ISBN: 1-874672-12-1

Chapter Solutions

For more information about this book contact:

Saxe-Coburg Publications
10A Saxe-Coburg Place
Edinburgh, EH3 5BR, UK

Tel: +44(0)131 315 3222

Fax: +44(0)131 315 3444
Email: buro@saxe-coburg.co.uk
URL: http://www.saxe-coburg.co.uk

 G.M. Seed, 2001. All rights reserved. No part of this document may be transmitted,
in any form or by any means whatsoever, without the written permission of the
copyright owner.
Chapter 1 Solutions
1.1 The cross-sectional area, A, of the bar is
( )
2
A = π 50x10 −3 = 7.85x10 −3 m2
The axial stress, σ, due to the axial load P=30kN is
P 30x103
σ= = = 3.82MPa
A 7.85x10−3

1.2 From (1.6) the axial strain, e, is

σ 3.82x106
e= = 9
= 18.2 x10−6
E 210x10
From (1.3) the axial extension, δ, is
δ = eL = 18.2 x10−6 x0.5 = 9.1x10 −6 m

1.3 From (1.16) the in-plane strains are

ε xx =
1
E
[ ]
σ xx − νσ yy , ε yy =
1
σ − νσ xx
E yy
[ ]
Re-arranging the first equation for σxx
σ xx = Eε xx + νσ yy
and substituting into the second equation then σyy is found to be

σ yy =
E
1− ν2
[ ε yy + νε ]
xx =
70x109
1 − 0.282
[60 + 0.28x40]x10−6 = 5.4 MPa
Substituting σyy into εxx and re-arranging for σxx

σ xx =
E
1− ν2
[ ε xx + νε ]
yy =
70x109
1 − 0.282
[40 + 0.28x60]x10−6 = 4.3MPa

1.4 From (1.23) the bulk modulus is

E 210x109
K= = = 175GPa
3(1 − 2ν ) 3(1 − 2 x0.3)

1.5 With E=σ/ε, σ=P/A, ε=δ/L and δ=α(∆T) then P is given by

EAα ( ∆T ) 120x109 x7.1x10−4 x11x10−6 (100)
P= = = 125kN
L 0.75
illustrating that a compressive axial load of 125kN is required to cancel out the
extension due to thermal expansion.

1.6 The compressive stresses of each bar are

P 150x103
σ copper = = = 76MPa
( )
2
Acopper π 25x10 −3

P 150x103
σ steel = = = 34 MPa
( )
2
Asteel π 37.5x10−3
The contractions of each bar are

1
 PLcopper 150x103 x0.5
δ copper = = = 318x10−6 m
E copper Acopper (
120x109 xπ 25x10 −3 2
)
PLsteel 150x103 x0.6
δ steel = = = 102x10−6 m
( )
2
E steel Asteel 200x109 xπ 37.5x10−3
The total contraction of the composite bar is δ=δcopper+δsteel=420x10-6m.

1.7 From (1.31,1.32,1.36) the in-plane strains εxx, εyy and γxy are given by
∂u 1 ∂v 1
ε xx = =
∂x E
(
σ xx − νσ yy , ε yy = ) ∂y E
= σ yy − νσ xx ( )
1  ∂u ∂v 
γ xy =  + =0
2  ∂y ∂x 

2
Chapter 2 Solutions
2.1 Figure Sol2.1 illustrates a circle of radius R with an elemental strip of thickness dy
at distance y from the x-axis. The area, dA, of the elemental strip is therefore
dA = 2 xdy = 2 R 2 − y 2 dy
The area of the strip is therefore given by, (2.1)
+R

A = ∫∫ dA = 2 ∫ R 2 − y 2 dy
A
−R

The integration is assisted by making the substitution y=Rsinθ, dy=Rcosθdθ

π /2 π /2 π /2
1  ϑ sin 2ϑ 
A = 4 R ∫ cos dϑ = 4 R ∫  (1 + cos 2ϑ )dϑ = 4 R 2  +
2 2 2
 πR 2
0    0
0
2 2 2
The first moment of area Qx is, (2.2)

( ) [( ]
+R +R
3/ 2 + R
Qx = ∫∫ ydA = ∫ y 2 R − y dy = 2 ∫ y R 2 − y 2 dy = −
A
2 2 2
3
R2 − y 2
−R
)
=0
−R −R
Similarly, it is found that Qy=0. Therefore, as expected the coordinates of the centroid
are xc=Qy/A=0 and yc=Qx/A=0.
y

dx

2x
dy
R

θ x
2y

Figure Sol2.1. A circle of radius R and elemental strip dy.

2.2 The derivation of the area, first moments of area and centroid are analogous to
those outlined in Exercise 2.1 except that the range of integration is now [0:R]. Thus,
the area of the semicircle is
π /2 π /2
ϑ sin 2ϑ  πR 2
R

A = ∫∫ dA = 2∫ R 2 − y 2 dy = 2 R 2 ∫ cos2 ϑdϑ = 2 R 2  + =
A
0 0 2 2  0 2
The first moment of area Qx is, using the elemental strip dy shown in Figure Sol2.1

3
 [( ]
R

) 2R3
2 3/ 2 R
Qx = ∫∫ dA = 2 ∫ y R 2 − y 2 dy = − R2 − y 2 =
A
0
3 0 3
The y-coordinate of the centroid is therefore
Q 2R3 / 3 4R
yc = x = =
A πR 2 / 2 3π
The first moment of area Qy is, using the elemental strip dx shown in Figure Sol2.1 in
which dA=ydx

[( ]
+R
3/ 2 + R
Qy = ∫∫ xdA = ∫ x R 2 − x 2 dx = −
A
1
3
R2 − x 2
−R
=0 )
−R
and hence xc=Qy/A=0.

2.3 Figure Sol2.3 illustrates an ellipse with elemental strip of length 2x and width dy.
The area A of the ellipse is therefore, noting that the equation of an ellipse is
x2/a2+y2/b2=1
b b y2 2a b
A= ∫∫ dA = ∫ 2 xdy = 2a ∫
A −b −b
1−
b 2
dy =
b ∫−b
b 2 − y 2 dy
Using the standard indefinite integral
a2  x 1
∫ a 2 − x 2 dx =
2
sin −1   + x a 2 − x 2
 a 2
then the area is found to be
b
2a  b 2  y 1 
A=  sin −1   + y b 2 − y 2  =
b 2  b 2  −b
2 a  b 2  b 2   2a  πb 2 πb 2 
 sin (1) −  sin ( −1)  =
−1 −1
=  + = πab
b  2  2  b  4 4 
The second moment of area Ix is, with dA=2xdy
b b y2 b y2
I x = ∫∫ y 2 dA = ∫ y 2 2 xdy = ∫ y 2 2a 1 − dy = 2 a ∫−b y 2
1 − dy
A −b −b b2 b2
Using the standard indefinite integral
 −a 
x
∫ x ax + cdx = 4a (ax ) cx c2
3
2 2 2
+c − ax + c −
2
sin −1  x  ; a<0
8a 8a − a  c 
then Ix is
b
 
3
 1 / b 2  
 y y2  y y 2
1 πab 3
I x = 2a  1 − 2  − 1− 2 − −1

sin  y   =
(
 − 4 / b
2
)
 b  8 −1 / b 2 ( ) b (
−8 / b 2 1 / b 2 )  1  
 −b
4

Similarly, it can be shown that Iy=πa3b/4.

4
 y

+b dy

2x x
-a +a

-b

Figure Sol2.3. An ellipse with half major and minor axes a and b.

2.4 The centroidal cordinates for rectangles 1 and 2 are (xc1,yc1)=(27.5,2.5) and
(xc2,yc2)=(2.5,25). The areas of rectangles 1 and 2 are A1=225mm2 and A2=250mm2
with the total cross-sectional area equal to A=A1+A2=475mm2. From (2.2) and (2.3)
Qy and xc are
Qy
Q y = ∑ x ci Ai = x c1 A1 + x c2 A2 = 6,812 mm 3 , x c = = 14.34 mm
i A
and similarly with Qx and yc given by
Q
Qx = ∑ y ci Ai = y c1 A1 + y c 2 A2 = 6,812 mm 3 , y c = x = 14.34 mm
i A
with xc=yc and Qx=Qy due to the symmetry of the bracket about the (x,y) axes.
From (2.5) and (2.6) the second moments of area of rectangles 1 and 2 are
given by, with respect to the centroidal axes
45(5) 3 5(45) 3
I xc1 = = 468mm 4 , I yc1 = = 37,968mm 4
12 12
3
5(50) 50(5) 3
I xc2 = = 52,083mm 4 , I yc2 = = 520mm 4
12 12
Use of the parallel-axis theorem (2.8) and (2.9) for rectangle 1 gives Ix and Iy with
respect to axes (x,y)
I x1 = I xc1 + A1 y c21 = 1,875mm 4 , I y1 = I yc1 + A1 x c21 = 208,125mm 4
and similarly for rectangle 2
I x 2 = I xc 2 + A2 y c22 = 208,333mm 4 , I y 2 = I yc2 + A2 x c22 = 2,083mm 4
Finally, Ix and Iy are given by
I x = I x1 + I x 2 = 210,208mm 4 , I y = I y1 + I y 2 = 210,208mm 4
with Ix=Iy as expected.

2.5 From (2.10) with dA=bdy then Ixy is given by

h/2
h/ 2  y2 
I xy = ∫∫ xydA = bx ∫ ydy = bx   =0
−h/ 2
A
 2  −h/2
and is seen to be equal to zero due to the symmetry about the coordinate axes.

2.6 From (2.13) the polar second moment of area for the ellipse of Exercise 2.3 is

5
 πa 3b πa 3b πab 2
I p = Ix + I y =
4
+
4
=
4
(
a + b2 )
2.7 From (2.15) the radii of gyration rx and ry for the ellipse of Exercise 2.3 are
Ix πab 3 / 4 b Iy πa 3b / 4 a
rx = = = , ry = = =
A πab 2 A πab 2

6
Chapter 3 Solutions
3.1 From (3.12) the maximum shear stress is given by
16T 16(10x103 )
τ max = = = 407 MPa
πD 3 π 50x10−3 3 ( )
3.2 From (3.1) the polar moment of area, J, is
( )
−3 4
πD 4 π 50x10
J= = . x10−6 m4
= 6136
32 32
and from (3.11) the angle of twist is
TL 10x103 x125
.
θ= = = 2.55x10−3 radians = 015
. o
GJ 80x10 61369
(
. x10
−6

)
3.3 From Example 3.1 then the maximum shear stress occurs at the smallest diameter
of d1=50mm
16T
τ max = 3 =
16 12x103
= 489MPa
( )
πd1 π 50x10−3 3 ( )
and the angle of twist is
32TL 1 1
θ=  − 3
3πG(d 2 − d1 )  d1 d 2 
3

=
( )
32 12 x103 x1 
 1
−
1

 = 01147
. radians = 6.57 o
3πx80x10 (75x10 − 50x10 )  (50x10 )
−3 −3
(75x10 ) −3 3 
9 −3 3

 

3.4 From (3.16) the mean shear stress is

T 100
τm = = = 163MPa
2πRm t 2π 3125 ( )( )
2 −3 2 −3
. x10 . x10
01

3.5 Re-arranging (3.16) for wall-thickness t

T . x103
15
t= = = 2.3mm
2πRm2τ m 2π 40x10−3 2 65x106 ( )( )
3.6 From (3.1) and (3.2) the polar moments of area of the inner solid bar, JA, and
outer tube, JB, are
( )
−3 4
πRA4 π 12.5x10
JA = = . x10−8 m4
= 383
2 2
JB =
π 4
2
(RB − RA4 =
π
)2 [( 4
25x10−3 − 12.5x10−3 ) ( ) ] = 5.75x10
4 −7
m4
From (3.21) the angle of twist is

7
 TL 5x103 (0.75)
θ= = = 01976 radians = 11324
. o
( ) ( )
.
G A J A + GB J B 45x109 383 −8
. x10 + 30x10 5.75x10
9 −7

3.7 From (3.20) the torques in the inner solid bar, TA, and outer tube, TB, are

TA = 
GA J A  

T = 
46x109 383 (
. x10 −8 ) 
 5x103 = 454 Nm
 G A J A + GB J B  9
 46x10 383 ( −8
) (
. x10 + 30x10 5.75x10 
9
)
−7 

 GB J B   30x10 (5.75x10 )
9 −7

TB =   T =   5x10 = 4.55kNm
3

 G A J A + GB J B   46x10 ( 383
. x10 ) + 30x10 (5.75x10 ) 
9 −8 9 −7

and from (3.22) the corresponding maximum shear stresses are

τ A,max =
TA R A 454 12.5x10
=
(−3

= 148MPa
)
JA . x10 −8
383

τ B ,max =
3
TB RB 4.55x10 25x10
=
(
−3

= 198MPa
)
JB 5.75x10 −7

8
Chapter 4 Solutions
4.1 From (4.2) and (4.4) the principal stresses in the pipe are
pr 50x6.895x10 3 x0.8 σ
σ1 = = −3
= 18.4 MPa , σ 2 = 1 = 9.2 MPa
t 15x10 2

4.2 From the Hookian equations (1.16) we have

ε zz = [σ zz − νσ θθ ] , ε θθ = [σ θθ − νσ zz ]
1 1
E E
Solving these for the in-plane stresses then we have
70x10 9
2 [ zz θθ ]
E
2 [
σ zz = ε + νε = 429 + 0.3x1,821]x10 −6 = 75MPa
1−ν 1 − 0.3
70x10 9
[ ]
E
σ θθ =
1−ν 2
ε θθ + νε zz =
1 − 0.32
[1821 + 0.3x429]x10−6 = 150MPa
The maximum in-plane shear stress is, (4.7)
σ 1 − σ 2 σ θθ − σ zz (150 − 75) x106
τ max = = = = 37.5MPa
2 2 2

4.3 With the pressure p equal to ρgh and σθθ=σ2 not exceeding the maximum
allowable stress σallow=300MPa/S where S(=10) is the safety factor then from (4.11)
we find the maximum permissible depth of water to be

h=
(−3
2tσ allow 2 25x10 300x10
=
)6

= 153m
ρgrS 1000x9.81x1x10

4.4 From (4.5) and (4.12) the circumferential strains for the cylinder, εc, and hemi-
spherical end caps, εs, are
pr  ν  pr
εc = 1 −  , ε s =
 
(1 − ν )
Et c 2 2 Et s
where E and ν denote Young’s modulus and Poisson’s ratio respectively, p is the
internal pressure, r is the radius of the vessel and tc and ts are the wall thicknesses of
the cylinder and hemi-spherical ends respectively. Equivalence of the circumferential
strains εc and εs yields
 1− ν 
ts =  t
2−ν c
and with tc=1mm then ts=0.4mm.

4.5 From (4.9) the absolute maximum shear is equal to

σ pr
τ max = 1 = σ 2 =
2 2t
The cylinder pressure, p, is
P P
p= = 2
A πr
where P is the force acting on the piston and r is the piston radius. Substituting p into
τmax then

9
 τY P
τ max = =
S 2πrt
where τY is the yield stress in pure shear and S is the safety factor. Re-arranging for
the cylinder wall thickness t
SP 2(50x103 )
t= = = 2.65mm
2πrτ Y 2π (40x10−3 )150x106

4.6 Re-arranging (4.11) for internal pressure p with σUTS denoting the ultimate tensile
stress then
2σ UTS t 2 x737 x106 (10−3 )
p= = = 227 MPa
r 65x10−3

4.7 From (4.19) Ix and Iy are given by

I x = I y = πr 3 t = πx253 x1 = 49,087 mm 4

10
Chapter 5 Solutions
5.1 From the equations of equilibrium the two unknown reactions RA and RB are found
∑ F v = 0 : R A + RB − 30(2) − 40 = 0 ⇒ R A + RB = 100kN
∑M = 0: RB (7) − 60(4) − 40(8) = 0 ⇒ R A = 80kN and ∴ R A = 20kN
The shear force, Vx, and bending moment, Mx, for the four intervals i) 0<x<3, ii)
3<x<5, iii) 5<x<7 and iv) 7<x<8 are:

i) 0<x<3
Vx = RA = 20kN
M x = R A x = 20 xkNm
ii) 3<x<5
Vx = RA − 30( x − 3) = 20 − 30( x − 3)kN

M x = R A x − 30
( x − 3)2 = 20 x − 30
( x − 3)2 kNm
2 2
iii) 5<x<7
Vx = R A − 60 = −20kN
M x = R A x − 60( x − 4) = 20 x − 60( x − 4)kNm
iv) 7<x<8
Vx = R A + RB − 60 = 40kN
M x = R A x + RB ( x − 7) − 6( x − 4) = 20 x + 80( x − 7) − 6( x − 4)kNm
with Vx and Mx illustrated in Figure Sol5.1.

11
 30kN/m 40kN

A B
RA RB

Vx

40
20
0
-20
-40

Mx
60
40
20
0
-20
-40

Figure Sol5.1. Shear force and bending moment diagrams for the simply supported
beam of Exercise 5.1.

5.2 From the equations of equilibrium the two unknown reactions are
WL 5x10 3 x5
R A = RB = = = 12.5kN
2 2
Taking moments at an arbitrary cut at a distance x from the left hand end
x Wx 2 WLx Wx 2
M xx + Wx − R A x = 0 ⇒ M xx = R A x − = −
2 2 2 2
Differentiating Mxx with respect to x and setting dMxx/dx=0 we find that the maximum
bending moment occurs at x=L/2. Substituting x=L/2 into Mxx then the maximum
bending moment is given by
WL2 5x10 3 x52
M max = = = 15,625Nm
8 8
For the rectangular section the second moment of area is given by, Example 2.3

12
 I=
(50x10 )(75x10 )
−3 −3 3

= 17578
. x10 −6 m 4
12
From (5.30) the maximum tensile and compressive stresses occur at
y=±h/2=±37.5mm

σ max = ±
M max ( h / 2)
=±
(
15,625 37.5x10 −3 ) = ±333MPa
−6
I 17578
. x10
5.3 From Example 5.2 the shear force, Vx, at a distance x from the left hand end of the
beam is
L 
V x = W  − x
2 
and attains maximum and minimum values at x=0 and x=L respectively
WL 5x10 3 x5
Vmax = = = 12.5kN , Vmin = −Vmax = −12.5kN
2 2
From the beam shear formula (5.40) and the rectangular cross-section examined in
Example 5.4 then the maximum shear stress at a given section is
Vx h 2
τ max =
8I
and occurs at the neutral axis of the beam. Substituting Vmax and Vmin we have
( )
2
12.5x10 3 75x10 −3
τ max,min = ± = ±5MPa
(
8 17578
. x10 −6 )
and are seen to be considerably less than the maximum tensile and compressive
bending stresses of Exercise 5.2.

5.4 From the equations of equilibrium the two unknown reactions RA and RB are found
WL
∑ F v = 0 : R A + RB = 2
WL  2 L  WL WL
∑ M = 0 : RB L − 2  3  = 0 ⇒ RB = 3 ⇒ RA = 6
where the centre of gravity of the distributed load acts at x=2L/3. From the equations
of equilibrium for the free body diagram of Figure Sol5.4 the bending moment is
given by
Wx 2  x  WLx Wx 3
∑ M = 0 : M x +  
2L  3 
− R A x = 0 ⇒ M x =
6
−
6L

13
 Wx/L

Mx
A
Vx
x
RA
2

Wx /2L
Figure Sol5.4. Free body diagram for the beam of Figure 5.27 cut at a distance x from
support A.

Substituting the bending moment into (5.41) we have

WLx Wx 3
EIv' ' = −
6 6L
Integrating with respect to x then the slope of the beam is
WLx 2 Wx 4
EIv' = − + C1
12 24 L
and integrating once more then the deflection is
WLx 3 Wx 5
EIv = − + C1 x + C2
36 120 L
The boundary condition v=0 at x=0 reveals that C2=0 and the boundary condition v=0
at x=L gives C1=-21WL3/1080. Substituting C1 and C2 into the above expressions for v
and v′ and re-arranging we have
v=−
Wx
(
360 LEI
)
3 x 4 − 10 L2 x 2 + 7 L4

v' = −
W
360 LEI
( )
7 L4 − 30 L2 x 2 + 15 x 4
The maximum deflection, δmax, can be determined from the condition that the slope of
the beam will be horizontal at the point of maximum deflection, xmax. Substituting
v′=0 in the above and re-arranging we have the following quadratic equation for
unknown x2
15 x 4 − 30 L2 x 2 + 7 L4 = 0
2
Solving for x
 480  2
x 2 = 1 ± L
 30 
which gives the two solutions x=1.3154L and x=0.5193L. Since x≤L then the
maximum deflection occurs at xmax=0.5193L. Finally, substituting xmax into v we
arrive at δmax
WL4
δ max = −v(0.5193L ) = 0.00652
EI

5.5 From the equations of equilibrium for the entire beam

WL WL 3 L
R A + RB − =0 , − RA L = 0
2 2 4
we find that the reactions RA and RB are given by

14
 3WL WL
RA = , RB =
8 8
As noted in Example 5.8, when using Macaulay's method, if a distributed load does
not extend to the right hand end of the beam then we need to extend and
counterbalance the distributed load to the right hand end of the beam. The bending
moment at a cut x-x (L/2<x<L) is, Figure 5.27b)
x {x − L / 2}
M xx + Wx − R A x − W {x − L / 2} =0
2 2
3WLx W {x − L / 2} 2 Wx 2
⇒ M xx = + −
8 2 2
Substituting Mxx into the flexure formula (5.41)
3WLx W {x − L / 2}2 Wx 2
EIv' ' = + −
8 2 2
and integrating twice
3WLx 2 W {x − L / 2} 3 Wx 3
EIv' = + − + C1
16 6 6
3WLx 3 W {x − L / 2}4 Wx 4
EIv = + − + C1 x + C2
48 24 24
From the two boundary conditions v=0 at x=0 and x=L we find that the constants of
integration C1 and C2 are
9WL3
C1 = − , C2 = 0
384
and upon substitution into v we arrive at the following expression for the deflection of
the entire beam
WLx 3 W {x − L / 2} 4 Wx 4 9WL3 x
EIv = + − −
16 24 24 384
Macaulay's method informs us that a term in curly brackets is ignored if it is either
zero or negative. Therefore, when x≤L/2 we have
v=−
Wx
(
384 EI
)
16 x 3 − 24 Lx 2 + 9 L3 ; 0 ≤ x ≤
L
2

5.6 The beam of Figure 5.28 has three unknown reactions (RA, RB and MA) but only
two independent equations equilibrium can be written and as a result the beam is
statically indeterminate to the first degree. Let reaction RB be the redundant reaction.
From the equations of equilibrium
R A + RB − WL = 0 ⇒ R A = WL − RB
L WL2
M A + R B L − WL = 0 ⇒ MA = − RB L
2 2
The bending moment at an arbitrary cut x-x is
x Wx 2
M xx + M A + Wx − R A x = 0 ⇒ M xx = R A x − M A −
2 2
Substituting RA and MA into Mxx and using the flexure formula (5.41) we have
WL2 Wx 2
EIv' ' = WLx − RB x − + RB L −
2 2
Performing two integrations we have

15
 WLx 2 RB x 2 WL2 x Wx 3
EIv' = − − + RB Lx − + C1
2 2 2 6
WLx 3 RB x 3 WL2 x 2 RB Lx 2 Wx 4
EIv = − − + − + C1 x + C2
6 6 4 2 24
Applying the boundary conditions v'=0 at x=0 and v=0 at x=0 we find that C1=0 and
C2=0. Applying the boundary condition v=0 at x=L gives the redundant reaction
3WL
RB =
8
The unknown reactions RA and MA are now given by
5WL WL2
RA = , MA =
8 8
Substituting RB, C1 and C2 into v and collecting terms then the deflection for the beam
is given by

v=−
Wx 2
48 EI
( )
3 L2 + 2 x 2 − 5 Lx

5.7 The beam has three unknown reactions (RA, RB and MA) but only two independent
equations of equilibrium and is therefore statically indeterminate to the first degree.
Letting RB be the redundant reaction then from the equations of equilibrium
WL2
R A = WL − RB , M A = − RB L
2
Using the method of superposition we now proceed to apply the distributed load W to
the released beam, Figure Sol5.7b), and the redundant reaction RA to the released
beam, Figure Sol5.7c). Since the deflection of the original beam is zero at end B then
we have the following compatibility equation
δ B = δ B1 − δ B 2 = 0
The force-displacement relations give the deflections δB1 and δB2
WL4 R L3
δ B1 = , δ B2 = B
8 EI 3EI
which upon substituting into the compatibility equation gives
WL4 RB L3
δB = − =0
8 EI 3EI
and solving for RB
3WL
RB =
8
The remaining reactions RA and MA are found from the above equilibrium equations
5WL WL2
RA = , MA =
8 8

16
 W

a) B
A
MA
RA
RB
L

b) δB 1

δB 2
c)
RB

Figure Sol5.7. Propped cantilever beam of Exercise 5.7. a) The propped cantilever
beam subject to a uniformly distributed load. b) The released beam with redundant
reaction RB and load W applied. c) The released beam with RB applied.

17
Chapter 6 Solutions

6.1 With I=πd4/64 for a solid circular cross-section of diameter d and substituting I
into (6.10) and re-arranging for d we have
1/ 4 1/ 4
 64 Pcr L2   64 x200x10 3 x4 2 
d =  =  = 75mm
 π E   π 3 x210x10 9 
3

6.2 The second moments of area with respect to centroidal coordinates (x,y) for the
square, circle and equilateral triangle shown in Figure Sol6.2 are
Ab 2 I A
square: I = I xx = I yy = , A = b2 , r 2 = = ≈ 0.083& A
12 A 12
Ab 2 I A
circle: I = I xx = I yy = , A = πb 2 , r 2 = = ≈ 0.0797 A
4 A 4π
Ab 2 3 2 I A
equilateral triangle: I = I xx = I yy = , A= b , r2 = = ≈ 0.0722 A
36 4 A 8 3

From (6.14) σcr is given by

π 2 Er 2
σ cr =
L2
with σcr seen to be proportional to r2. Therefore, the struts from largest to smallest σcr
are the square, circular and equilateral triangle cross-sections.

y y
y

x x
b b b x
b
b b
Figure Sol6.2. Square, circle and equilateral triangle.

6.3 With reference to Figure Sol6.3 the second moments of area Ix and Iy are, (2.5)
and (2.6)
π /2
I x = ∫∫ y 2 dA = 4 ∫ (a cosθ ) (tds) = 4 ∫ (a cosθ ) 2 (tadθ ) = πa 3 t
2
A 0
π /2
I y = ∫∫ x 2 dA = 4 ∫ (a sin θ ) ( tds) = 4 ∫ (a sin θ ) 2 (tadθ ) = πa 3 t
2
A 0
with Ix=Iy due to symmetry. The area of the section is
π /2
A= ∫∫ dA = 4∫ tds = 4∫ t ( adθ ) = 2πat
A 0
Finally, the radius of gyration is, (6.12)
I πa 3 t a
r= = =
A 2πat 2

18
 y

ds
a

dθ
θ x
Figure Sol6.3. One quadrant of a circular tubular cross-section.

6.4 From the equations of equilibrium the compressive load acting on the strut is
equal to √3W. The cross-sectional area and second moment of area of the strut are
( )
2
A = 50x10 −3 = 2.5x10 −3 m 2
1
( )( )
3
I= 50x10 −3 50x10 −3 = 5.2083& x10 −7 m 4
12
From (6.12) the radius of gyration is
I 5.2083x10 −7
r= = = 0.0144 m
A 2.5x10 -3
The strut is built-in at one end and pin-jointed at the other end where the load W acts
so that from (6.42) the effective length of the strut is Le=0.7L=1.75m and the critical
buckling load is, (6.31)
π 2 EA π 2 x210x10 9 x2.5x10 −3
Pcr = = = 351kN
( Le / r ) 2 (175
. / 0.0144)
2

Thus, the strut will fail due to buckling when W exceeds 351kN.

6.5 The area and second moment of area of the rectangular tube are
A = (0.2 x01
. ) − (018
. x0.08) = 5.6x10 −3 m 2

I=
1
12
[ . 3 − 018
0.2 x01 ]
. x0.08 3 = 8.986& x10 −6 m 4
From (6.32) the critical buckling load is
4π 2 EI 4π 2 x210x10 9 x8.986x10 −6
Pcr = = = 993.3kN
L2 52
and the critical stress is
Pcr 993.3x10 3
σ cr = = = 177.4 MPa
A 5.6x10 −3

6.6 The stress and eccentricity ratio are

19
 P 350x10 3
= = 107 MPa
A 3270x10 −6
ey eA 0.025x3270x10 −6
= = = 0.5677
r2 S 144 x10 −6
The critical buckling load is, (6.10)
π 2 EI π 2 x210x10 9 x11x10 −6
Pcr = = = 912 kN
L2 52
From (6.52) and (6.57) the maximum deflection and stress are
 π P  
δ = e sec  − 1 = 19.5mm
  2 Pcr  
P  ey π P 
σ max = 1 + 2 sec   = 215MPa
A  r  2 Pcr  

6.7 From Table 6.2 the constant a is equal to 1/7500 and the slenderness ratio is
L 2
l= = = 50.505
r 39.6x10 −3
Therefore, from (6.61) the critical stress according to the Rankine-Gordon formula is
σY 300x10 6
σR = = = 224 MPa
1 + al 2 50.5052
1+
7500

20
Chapter 7 Solutions

7.1 The σxx stress is

P 75x103
σ xx = =− = −50MPa
A 1500x10−6
From (7.8) and (7.12) the local direct and shear stresses on the cut plane ab are
1
σ x 'x ' = σ xx (1 + cos 2ϑ ) = σ xx cos2 ϑ = −37.5MPa
2
1
σ y'y' = σ xx (1 − cos 2ϑ ) = −12.5MPa
2
1
τ x ' y ' = −σ xx sin 2ϑ = −σ xx sin ϑ cos ϑ = 2165
. MPa
2

7.2 From the stress transformation equations we have

σ x 'x ' = 195MPa , σ y ' y ' = 105MPa , τ x ' y ' = 50MPa
observing that the sum of the global and local stresses are equal, (7.13)
σ xx + σ yy = σ x 'x ' + σ y ' y '

7.3 The centre C, point A, point B and radius R of Mohr’s circle are, see §7.5.1
 σ + σ yy 
C =  xx ,0 = [ −25,0]
 2 
( ) (
A ϑ = 0o = σ xx , τ xy = (50,−50) )
B(ϑ = 90 ) = (σo
yy )
,−τ xy = ( −100,50)

 σ − σ yy 
2

R =  xx  + τ xy = 90.14
2

 2 
Mohr’s circle can now be plotted and is shown in Figure Sol7.3.

θ(+ve.)
S2
τm
D’ in

A(θ=0 )
2θs 2

P2 2θp 2 -2θ
P1 σx ’ x ’
C O p1
R
β

o 2θs 1
B(θ=90 ) o

D(θ=45 )
τm a x
S1

τx ’ y ’
Figure Sol7.3. Mohr’s circle for the stress element of Exercise 7.3.

21
The principal stresses and associated planes are
σ 1 = OC + R = −25 + 9014. = 6514.
σ 2 = OC − R = −25 − 90.14 = −11514 .
 2
2ϑ p 2 = 180o − ∠ACP1 = 180o − tan −1   = 146.31o ; ϑ p 2 = 7316
. o
 3
2ϑ p1 = 180o + 2ϑ p 2 ; ϑ p1 = 16316
. o or − 16.84o
The maximum and minimum shear stresses and associated planes are
τ max = R = 90.14
τ min = − R = −9014
.
 3
2ϑ s2 = ∠ACS2 = tan −1   = 56.31o ; ϑ s2 = 2815
. o
 2
2ϑ s1 = 180o + 2ϑ s 2 ; ϑ s1 = 11815
. o
The stresses on the plane θ=-45° are represented by points D and D′ on Mohr’s circle
shown in Figure Sol7.3. With angle β given by
 2
β = 2ϑ − ∠ACP1 = 90o − tan −1   = 90o − 33.69o = 56.31o
 3
σ x 'x ' ( D) = OC + R cos β = −25 + 9014
. cos 56.31o = 25
τ x ' y ' ( D) = R sin β = 90.14 sin 56.31o = 75
σ y ' y ' = σ x 'x ' ( D' ) = OC − R cos β = −25 − 90.14 cos 56.31o = −75

7.4 Mohr’s circle can be schematically constructed by positioning the centre, C, of the
circle and determining its radius, R, as follows for the three requested cases:

σxx=σ
a) Uniaxial (σ σ, σyy=ττxy=0)

 σ + σ yy   σ 
C =  xx ,0 =  ,0
 0  2 
 σ − σ yy 
2
σ
R =  xx  + τ xy =
2

 2  2

σxx=σ
b) Equi-biaxial (σ σyy=σ
σ, τ xy=0)

C = [σ ,0]
R=0
noting that Mohr’s circle reduces to a point.

σxx=σ
c) Pure Shear (σ σyy=0, τ xy=ττ)

C = [0,0]
R=τ

22
Each of the above cases is correspondingly illustrated in Figure Sol7.4.

C[σ/2,0] σx ’ x ’ C[σ,0] σx ’ x ’ C[0,0] σx ’ x ’

O O

σ/2 τ

τx ’ y ’ τx ’ y ’ τx ’ y ’

a) b) c)
Figure Sol7.4. Mohr’s circle for the cases of a) uniaxial, b) equi-biaxial and c) pure
shear loadings.

7.5 (i) With I1, I2 and θp given by

Ix + Iy  I − Iy 
2
2 I xy
I1,2 = ±  x  + I xy , tan2ϑ p = −
2

2  2  Ix − Iy
and since Ix=Iy then
π 4 1 4  4  9π 2 + 18π − 128  4
I1 = I x + I xy =  − + − r =  r
 16 9π 8 9π   144π 
π 4 1 4  4  π − 2 4
I 2 = I x − I xy =  − + + r =  r
 16 9π 8 9π   16 
tan 2ϑ p = ∞ ; 2ϑ p = 90o ; ϑ p = 45o
(ii) With r=10mm then from the Ix, Iy and Ixy expressions given
I x = I y = 549 mm4 , I xy = −165mm4
and from the transformation equations
 I + Iy   Ix − I 
I x ', y ' =  x  ±  cos 2ϑ m I xy sin 2ϑ = I x m I xy sin 2ϑ = 692 mm and 406mm
4 4

 2   2 
I x ' y ' = I xy cos 2ϑ = −82.5mm4

7.6 With θ=30° then cos2θ=1/2 and sin2θ=√3/2 and substituting Ix, Iy and Ixy into
(7.39) and (7.41) we find that Ix′=Iy, Iy′=Ix and Ix′y′=Ixy as required.

7.7 From (7.54) the shear modulus, G, for the aluminium alloy and steel are
Eal 70
Gal = = = 26GPa
2(1 + ν al ) 2(1 + 0.33)
Esteel 210
Gsteel = = = 81GPa
2(1 + ν steel ) 2(1 + 0.3)

23
Chapter 8 Solutions

8.1 Since θ is taken as positive for an anticlockwise rotation then θ=-30° in the
present case. From (8.10) the local strains are
 500 + (-300)  −6  500 − ( −300)  −6 200x10 −6
ε x ' x ', y ' y ' =   x10 ±   x10 cos( − 60 o
) ± sin( −60 o )
 2   2  2
= 213x10 −6 ,−13.4 x10 −6
γ  500 − ( −300)  −6 200x10 −6
= −  x10 sin( −60 ) + cos( −60 o ) = 396.4 x10 −6
x' y ' o

2  2  2
so that γx′y′ is equal to 793x10 .
-6

8.2 From (8.11) the principal strains ε1 and ε2 are

ε xx + ε yy
 ε − ε yy   γ xy 
2 2
−6 −6
ε 1,2 = ±  xx  +  = −190x10 ,−360x10
2  2   2 
From (8.11) we obtain the planes of the principal strains
 80  −1  8 
2ϑ p = tan −1   = tan  
 −350 − ( −200)  15
which has the two roots of θp=-14° and 76° for θp in the range 0≤θp≤180°. To
establish which angle is associated with either ε1 and ε2 then let us examine, say, θp=-
14° for εx′x′, (8.10)
−350 + ( −200) −350 − ( −200) 80
ε x ' x ' ( −14 o ) = x10 −6 + x10 −6 cos( −28 o ) + x10 −6 sin( −28 o )
2 2 2
−6
= −360x10
observing that εx′x′ (-14°)=ε2 so we conclude that θp2=-14° and θp1=76°. The principal
strains are illustrated graphically in Figure Sol8.2a).
y y
y’ y’
o

θp 2 =76

ε2 dy’ εa v g dy’
x’
o

θs 1 =31
x εa v g dx’ x
o

θp 1 =14
x’
ε1 dx’

a) b)
Figure Sol8.2. Schematic illustration of strains. a) Principal stress element. b)
Maximum shear strain element.

The maximum shear strain is, (8.12)

24
 γ x 'y'   ε xx − ε yy   γ xy 
2 2
 −350 − ( −200) 
2 2
 80  −6
  =   +  =   +   = 85x10
 2  max  2   2   2   2
and therefore γmax=170x10-6 and is illustrated graphicaly in Figure Sol8.2b).

8.3 With θa=0°, θb=60° and θc=120° we find that the system of equations (8.18)
reduces to
60x10−6 = ε xx
135x10−6 = 0.25ε xx + 0.75ε yy + 0.433γ xy
−6
264x10 = 0.25ε xx + 0.75ε yy − 0.433γ xy
which result in the global strains εxx=60x10-6, εyy=246x10-6 and γxy=-149x10-6.
We will use the transformations equations to determine the principal strains
and their associated planes. From (8.11) the principal strains are
 60 + 246   6 − 246   −149 
2 2
−6 −6
ε 1,2 =  ±   +  = 272 x10 , 34 x10
 2   2   2 
and with principal planes
 γ xy  1 −149  1
 = tan −1 
1
θ p = tan −1   = tan (0.8) = 19.33
−1 o

2  ε xx − ε yy  2  60 − 246  2
Inserting 19.33° into (8.10) we find that θp2=19.33° and therefore
θp1=90°+θp2=109.33°. These results can be compared with Example 8.2 which
alternatively determined the principal strains and planes using Mohr's circle.

8.4 Points A and B, centre C and radius R of Mohr’s circle for the in-plane strains
εxx=250x10-6, εyy=-150x10-6 and γxy=120x10-6 are as follows
( ) ( ) (
A ϑ = 0o = ε xx , γ xy / 2 = 250x10−6 ,60x10−6 )
B(ϑ = 90 ) = (ε ,−γ / 2) = ( −150x10
o
yy xy
−6
,−60x10 −6 )
C = [ε ,0] = [50x10 ,0]
avg
−6

 ε − ε yy   γ xy 
2 2
−6
R =  xx  +  = 209 x10
 2   2 
with Mohr's strain circle illustrated in Figure Sol8.4. From Mohr’s circle we find the
principal strains and planes
ε 1 = OC + R = (50 + 209) x10−6 = 259 x10−6
ε 2 = OC − R = (50 − 209) x10−6 = −159 x10−6
 60 
2ϑ p1 = tan −1   = 16.7 ; ϑ p1 = 8.35
o o
 250 − 50 
2ϑ p 2 = 180o + 2ϑ p1 ; ϑ p 2 = 98.35o
The maximum shear strain (γ/2)max=R so that γmax=418x10-6. From Mohr’s circle the
planes of the maximum and minimum shear strains are
( )
2ϑ s1 = − 90o − 2ϑ p1 = −73.3o ; ϑ s1 = −36.65o
2ϑ s2 = 180o − 2ϑ s1 = 106.7 o ; ϑ s 2 = 53.35o

25
 o

B(θ=90 )
(-150,-60)

- 6

ε2 2θ p 2 C[50,0] ε1 εx ’ x ’ x10
O
2θ p 1

-2θs 1
R=209
o

A(θ =0 )
(250,60)

-6

(γx ’ y ’ /2)x10
Figure Sol8.4. Mohr’s circle for the in-plane strains εxx=250x10-6, εyy=-150x10-6 and
γxy=120x10-6.

8.5 The average normal strain εavg, centre C, points A and B and radius R of Mohr’s
strain circle are
ε + ε yy
ε avg = xx = −200x10−6
2
[ ] [
C = ε avg ,0 = −200 x10−6 ,0 ]
A(ϑ = 0 ) = (ε , γ / 2) = ( −300x10 ,50x10 )
o
xx xy
−6 −6

B(ϑ = 90 ) = (ε ,−γ / 2) = ( −100x10 ,−50x10 )

o
yy xy
−6 −6

 ε − ε yy   γ xy 
2 2

R =  xx  + . x10 −6
 = 1118
 2   2 
Mohr’s circle can now be constructed and is shown in Figure Sol8.5. Since we are
required to determine the strain components on an element that is rotated by 20° in a
clockwise direction then θ=-20°. From Figure Sol8.5 the principal plane θp2 is given
by
1  50  1 −1 1
ϑ p2 = − tan −1   = − tan = −13.28o
2  300 − 200  2 2
so that β=2θ-2θp2=40°-26.57°=13.43°. From triangle DEC we find the local strains
ε x 'x ' = −(OC + R cos β ) = − 200 + 1118 (
. cos 13.43o x10−6 = −309 x10−6 )
γ x 'y'
. x10−6 sin 13.43o = −25.97x10−6 ; γ x ' y ' = 5193
= − R sin β = −1118 . x10 −6
2
and from triangle D′FC we find the remaining local strain εy′y′

26
 . cos 13.43o ) x10−6 = −9126
ε y ' y ' = OC + R cos β = (−200 + 1118 . x10−6

B(θ=90 )
D(εx ’ x ’ ,-γx ’ y ’ /2) (-100,-50)

-2θ
C[-200,0] -6

β εx ’ x ’ x10
F O
E
-2θp 2

o
D’(εy ’ y ’ ,γx ’ y ’ /2)
A(θ=0 )
(-300,50)

(γ x ’ y ’ /2)m a x
=R=111.8

-6

(γ x ’ y ’ /2)x10
Figure Sol8.5. Mohr’s circle for the in-plane strains εxx=300x10-6, εyy=-100x10-6 and
γxy=100x10-6.

8.6 From (8.10) the local strains are

 200 + 400  −6  200 − 400  −6 100x10 −6
ε x ' x ', y ' y ' =   x10 ±   x10 cos 80 ±
o
sin 80 o
 2   2  2
= 332 x10 −6 , 268x10 −6
γ x' y '  200 − 400  −6 100x10 −6
= −  x10 sin 80 +
o
cos 80 o = 107 x10 −6
2  2  2
so that γx′y′ is equal to 214x10 . -6

From the Hookian equations (1.16) we have

τ x'y'
ε x'x' =
1
E
[ ]
σ x ' x ' − νσ y ' y ' , ε y ' y ' =
1
E
[ ]
σ y ' y ' − νσ x ' x ' , γ x ' y ' =
G
Solving these for the in-plane stresses then we have

σ x ' x' =
E
1−ν 2
ε [
x' x ' + νε y' y ' = ]
210x10 9
1 − 0.32
[332 + 0.3x268]x10 −6 = 95MPa
σ y'y' =
1−ν
E
2
ε y'y'[ + νε x ' x' = ]
210x10 9
1 − 0.32 [
268 + 0.3x332]x10 −6 = 85MPa

E 210x10 9
τ x ' y ' = Gγ x ' y ' = γ x' y ' = 214 x10 −6 = 17 MPa
2(1 + ν ) 1(1 + 0.3)

8.7 With θa=0°, θb=45° and θc=90° then the system of equations (8.18) reduce to
65x10 −6 = ε xx
95x10 −6 = 0.5ε xx + 0.5ε yy + 0.5γ xy
−6
25x10 = ε yy

27
which result in the global strains εxx=65x10-6, εyy=25x10-6 and γxy=100x10-6. From
(8.11) the principal strains are
ε xx + ε yy  ε xx − ε yy   γ xy 
2 2
−6 −6
ε 1,2 = ±   +  = 98.85x10 , − 8.85x10
2  2   2 
and from (8.11) we obtain the planes of the principal strains
 100 
2ϑ p = tan −1   = tan (2.5) = 68.2
−1 o
 65 − 25
which has the two roots of θp=34.1° and 124.1° for θp in the range 0≤θp≤180°.

28
Chapter 9 Solutions
9.1 From (9.10) the in-plane strains are
ε xx =
E
[
1
]
σ xx − νσ yy , ε yy = [1
σ − νσ xx
E yy
]
and substituting into the strain energy density, (9.9), we have

[
1
]
U 0 = σ xx ε xx + σ yy ε yy =
2 2E
[
1
]
σ 2xx − 2νσ xxσ yy + σ 2yy
Integrating U0 throughout the entire volume of the plate then the strain energy is given
by
U = ∫ U 0 dV =
V
V
( )
σ 2 − 2νσ xxσ yy + σ 2yy
2 E xx

9.2 From (9.4) the strain energy density, U0, is equal to σε/2 so that the strain energy
is given by
L σ ( x)
2
U = ∫ U 0 dV = ∫ A( x )dx
V 0 2E
where A is the cross-sectional area of the bar and is a function of x. With σ(x)=W/A(x)
then U is
W 2 L dx
2 E ∫0 A( x)
U=

It remains to find A(x). Linearly interpolating across the length of the bar from d1 to d2
then
2
πd12   d 2  x 
2
 d1  d 2 d1  x 
A( x ) = πy = π  +  −   =
2
1 +  − 1 
2  2 2  L 4   d1  L 
and substituting A(x) into U
2W 2 L dx
2 ∫0
U=
Eπd1   d2  x 
2

1 +  − 1 
  d1  L 
The integral is seen to be of the following general form
dx 1
∫ (ax + b)2 = ∫ (ax + b) dx = − a (ax + b)
−2 −1

Performing the integration then U is found to be

2W 2    2W 2 L
L
L  1
U= −   =
Eπd12  d 2 / d1 − 1  1 + ( d 2 / d1 − 1)( x / L)  0  Eπd1d 2
 
as required.
From Castigiliano's second theorem then the displacement of the bar is, (9.59)
∂U 4WL
δ= =
∂W Eπd1d 2
When d1=d2=d then the bar is of constant cross-section and δ is given by
4WL
δ=
Eπd 2
and is seen to agree with δ=WL/AE where A=πd2/4.

29
9.3 Resolving forces vertically at joint B then the force, F, in each member is
P
F=
2 cosϑ
and the length, L, of each member is d/cosθ. Considering member BC then the strain
energy is, (9.12)
σ 2  σ 2 AL
L

U BC = ∫ U odV = ∫  xx  Adx = xx
V 0
 2E  2E
with x taken along the member axis. Substituting for σxx and L then
P2d
U BC =
8EA cos3 ϑ
Due to symmetry UAB=UBC so that the total strain energy, U, of the frame is
P 2d
U = U AB + U BC = 2U BC =
4 EA cos3 ϑ
From Castigilano’s second theorem (9.59) the displacement, δB, at joint B is
∂U ∂  P2d  Pd
δB = =  =
∂P ∂P  4 EA cos ϑ  2 EA cos3 ϑ
3

9.4 Re-arranging (9.49) for applied load W

δd 4 cos α 10 x54 cos 25o
W= = = 50.4 N
3  cos α 2 sin 2 α   cos 2 25o 2 sin 2 25o 
2
8D n + 3
+
E 
8x50 x10 3 
 G  85x10 . x10 3 
2261
The shear stress is given by (9.51)
M (d / 2) (WD / 2) cos α (d / 2) 50.4(50 / 2) cos 25o (5 / 2)
τ = x4 = = = 46.53MPa
πd / 32 πd 4 / 32 π 54 / 32
and the bending stress is given by (9.52)
M y (d / 2) (WD / 2) sin α (d / 2) 50.4(50 / 2) sin 25o (5 / 2)
σ= = = = 43.39 MPa
πd 4 / 64 πd 4 / 64 π 54 / 64

9.5 Letting ls(=nd) denote the solid length then from (9.40) we have
W Gd 4 45x10 9 d 4
k= = ⇒ 2 .5 = ⇒ D 3 = 45x10 6 d 5
δ 8D n 3 3
8 D (l s / d )
From (9.42) with τ≤120MPa and re-arranging for D
τπd 3 120x10 6 πd 3
D= = = 1346
. x10 6 d 3
8W 8x35
Eliminating D from these two equations we have
45x10 6
d= 4 = 2.07 mm
2.439 x1018
It follows that the mean coil diameter is
( )
3
. x10 6 2.07 x10 −3
D = 1346 = 1194
. mm
Finally, the number of coils is equal to
ls 50
n= = ≈ 24
d 2.07

30
9.6 Considering the right hand side of the ring shown in Figure 9.19 then the bending
moment, M, at angle θ is seen to be
M = ( R − R cos ϑ ) P = RP(1 − cos ϑ )
For pure bending the strain energy, U, of a beam is given by (9.35) which in the
present case is
π
M 2 Rdϑ
L
M 2 dx
U=∫ =∫
0
2 EI 0
2 EI
substituting M we have
π π

U=
1
[
2 EI ∫0
]
RP(1 − cos ϑ ) Rdϑ =
2
(R3 P 2
2 EI ∫0
)
1 − 2 cos ϑ + cos2 ϑ dϑ

Performing the integration

π
R3 P2  ϑ sin 2ϑ  3πR 3 P 2
U= ϑ − 2 sin ϑ + + =
2 EI  2 4  0 4 EI
An application of Castigliano’s second theorem, (9.59), gives the displacement, δu, at
the point at which the point load P acts
∂U 3πR 3 P
δu = =
∂P 2 EI
Since the total gap, δ, between the two opposing point forces is equal to twice δu then
3πR 3 P
δ = 2δ u =
EI
as required.
To determine the value of P required to produce a total gap of δ=10mm we
can re-arrange the above expression for P with I=w4/12, where w=2.5mm is the width
of the square section
EIδ 210x109 x39.0625x10 −12 x10x10−3
P= = = 955 .N
3πR 3 3x(45x10 −3 ) 3
Inspection of the expression for bending moment M we observe that M obtains a
maximum at θ=180° (position perpendicular to the applied forces P, as expected) and
is equal to Mmax=2RP=2x45x10-3x95.5=8.595N. Thus, the maximum bending stress
is, (5.30)
 2.5x10 −3 
8.595x 
M max y  2 
σ max = − =− = −275MPa
I 39.0625x10 −12
which is less in magnitude than the tensile yield stress of σY=300MPa.

9.7 From the given beam deflection equation the maximum static deflection, δst,
occurring at x=L is
WL3 1xgx13
δ st = = = 0.2307 mm
3EI 3x210x109 x6.75x10−8
The impact factor F is, (9.89)
2h 2x0.5
F = 1+ 1+ = 1+ 1+ = 66.85
δ st 2.307 x10 −4
Therefore, the maximum displacement, δmax, due to the impact of the falling mass is

31
δ max = Fδ st = 15.42mm

32
Chapter 10 Solutions
10.1 Refer to §10.2.

10.2 Refer to §10.3.

10.3 Refer to §10.4.

10.4 To determine the slope at the free end of the cantilever beam then we can apply a
virtual unit couple moment 1Nm at the free end of the beam, as shown in Figure
Sol10.4. The virtual bending moment, Mv, is
Mv =1 0≤ x≤ L
with the bending moment due to the applied load P
M =0 0≤ x ≤b
M = P( x − b ) b≤ x≤ L
An analogous equation to (10.36) can be written for the angle of rotation, θ
1
θ=
EI ∫L
M v Mdx
In the present example we have
P ( L − b)
L 2
1 P  x2  P  L2 − 2bL + b 2  Pa 2
P( x − b)dx =
L
θ=
EI ∫
b EI  2
− bx  = 
 b EI  2
=
 EI 2
=
2 EI

P
1Nm

a b θ

x
Figure Sol10.4. A cantilever beam with a virtual unit couple moment applied at the
free end.

10.5 With a virtual unit load applied at the free end, Figure Sol10.5, then the
associated bending moment is
M v = −( R − R sinθ )
whereas the bending moment due to the real applied point force P is
M = − PR cosθ
Thus, from (10.36) the horizontal deflection, δ, is
1 π /2 v 1 π /2
− R(1 − sin θ )( − PR cosθ ) Rdθ =
EI ∫0 EI ∫0
δ= M Mds =
π /2
PR 3  1  PR 3
= sin θ + 4 cos 2θ  =
EI 0 2 EI

33
 P
R-Rsinθ
ds 1
R δ

dθ Rsinθ

R-Rcosθ Rcosθ
Figure Sol10.5. A quarter circle beam subject to a concentrated force P and virtual
unit force at the free end.

10.6 To determine the vertical deflection at the free end of the beam we add a virtual
unit load at the free end, see Figure Sol10.6. For the virtual unit load the bending
moment is
Mv = x 0≤ x≤7
For the real applied loading
M =0 0 ≤ x ≤ x3
M = 10( x − 3) 3≤ x ≤ 4
M = 10( x − 3) + 20( x − 4) 4≤x≤7
From (10.36) we have
[ ]
EIδ = ∫ M v Mdx = ∫ 10( x − 3) xdx + ∫ 10( x − 3) + 20( x − 4) xdx =
4 7

L 3 4
4
 x 3 3x 2 
[ ]
7
= 10 −  + 10 x 3 − 55x 2 4 = 9,768.3&
3 2 3
Re-arranging for δ we have
9,768.3&
δ= = 9.53mm
(
205x10 9 5x10 −3 )

34
 10kN 20kN 1kN

δ
3m 1m 3m

x
Figure Sol10.6. A cantilever beam subject to real concentrated loads of 10kN and
20kN and a virtual concentrated unit load at the free end of the beam.

10.7 To determine the deflection at the free end of the beam we apply a unit virtual
point force at this point. With x measured from the free end of the beam then for the
virtual unit load the virtual bending moment, Mv, is
M v = x 0 ≤ x ≤ 3a
For the real applied loading system we have the following bending moments, M
M = 0 0≤ x ≤a
M = W ( x − a ) a ≤ x ≤ 2a
M = W ( x − a ) + W ( x − 2a) 2a ≤ x ≤ 3a
From (10.36) the displacement, δ, at the free end is
2a 3a

EIδ = ∫ W ( x − a) xdx +
a
∫ [W ( x − a) + W ( x − 2a)]xdx
2a

with no integral in the interval 0≤x≤a because M=0. Performing the integrations we
find the desired solution
6Wa 3
δ=
EI

35
Chapter 11 Solutions
11.1 From (11.27) we have
p  (b / r ) + 1 (b / r ) − 1 2 p  (b / r ) 
2 2 2
σ θθ − σ rr
=  +  =  
r r  (b / a) 2 − 1 (b / a ) 2 − 1 r  (b / a) 2 − 1
Differentiating the radial stress with respect to r we have
2 p  (b / r ) 
2
∂σ rr
=−
p
−(2b 2
/ r 3
) =  
∂r (b / a) 2 − 1 r  (b / a) 2 − 1
which is equivalent to the above equation and therefore satisfies the equilibrium
equation (11.3).

11.2 With a=75mm, b=250mm and p=75MPa then from (11.27) the radial and
circumferential stresses on the inner surface, r=75mm, are
 (b / r ) 2 − 1   (250 / 75) 2 − 1
σ rr = − p   = −75  = −75MPa
 (b / a ) − 1  (250 / 75) − 1
2 2

 (b / r ) 2 + 1   (250 / 75) 2 + 1
σ θθ = p   = 75  = 90MPa
 (b / a ) − 1  (250 / 75) − 1
2 2

Assuming closed ends then from (11.33) the axial stress is

p 75
σ zz = = = 7.4MPa
(b / a) − 1 (250 / 75) 2 − 1
2

11.3 With a=0.5m, b=1m, pi=5MPa, po=100kPa=0.1MPa then the constants A and B
in (11.20) are

A=
(b / a ) po − pi (1 / 0.5) x01
2

=
2
. x106 − 5x106
. & x106
= 153
1 − (b / a ) 1 − (1 / 0.5)
2 2

B=
b 2 ( po − pi )
=
( . x106 − 5x106
12 01 ) = 163
. & x10 6

1 − (b / a ) 1 − (1 / 0.5)
2 2

From (11.21) the radial and circumferential stresses at the radius r=0.75m are
B & . & x106
163
σ rr = A − 2 = 153 . x10 − 6
== −137
. MPa
r 0.752
B . & x10 6
163
σ θθ = A + 2 = 153 . & x106 + = 4.44 MPa
r 0.752
From (11.23) the axial stress is
po (b / a) − pi 0.1(1 / 0.5) − 5
2 2

σ zz = = . &
= 153MPa
1 − (b / a) 1 − (1 / 0.5)
2 2

11.4 Consider the first vessel with the boundary conditions

σ rr = −45MPa at r = 75mm
σ rr = 0 at r = 100mm
From Lame’s equations, (11.16), we find the constants A and B to be given by
A = 586
. x106 , B = 578.6x103

36
For the inner surface, in which σθθ will be maximum, then the circumferential stress is
B 578.6x106
σ θθ = A + 2 = 57.86x106 + = 160.72 MPa
r 75x10 (
−3 2
)
For a safety factor of 2 then the maximum allowable cylindrical stress for the second
cylindrical pressure vessel will be 160.72/2=80.36MPa.
For the second pressure vessel then our boundary condition (σrr=0 at the inner
surface) combined with the maximum design circumferential stress give the two
simultaneous equations, (11.16)
B' B'
σ θθ = A'+ 2 ⇒ 80.36x10 6 = A'+
( )
2
ri 75x10 −3
B' B'
σ rr = A'− ⇒ 0 = A '−
(75x10 )
2
r2 −3
+ 50x10 −3
noting the two new constants A´ and B´. Solving for A´ and B´ we have
A' = 2127
. x106 , B' = 332.34x103
Re-arranging the radial stress component of Lame’s equations (11.16) for applied
internal pressure p then we have at the inner surface, r=75mm
 
 B'  332.34 x103 
p = −  A'− 2  = − 2127
. x106 − = 37.8MPa = 378bar
 r  
 75 x(
10 −3 2 
 )
Thus, the maximum safe working pressure for the second pressure vessel is 378bar.

11.5 We first need to determine the interference pressure, p, so that the maximum
stress (σθθ) at the sleeve-collar interface does not exceed 300x106. Therefore, from
Lame's equations
σ θθ ( Rci ) = 300x106 = A + , σ rr ( Rco ) = 0 = A −
B B
( −3 2
) ( )
2
49.5x10 100x10−3
Solving for A and B we find
A = 59 x10 6 , B = 59 x10 4
The radial stress at r=Rci is
B 59x104
σ rr ( Rci ) = A − = 59 x10 6
− = −182MPa
( )
2
Rci2 49.5x10−3
Therefore, the interference pressure is 182MPa.
From (11.40) the radial compression on the shaft is
pR 182 x106 (50 x10 −3 )
us = − s (1 − ν ) = − 9 (1 − 0.3) = −30.3& x10−6 m
E 210x10
From (11.44) the radial expansion of the collar is
Rci  pRci2    Rco  
2

uc = 1 − ν + (1 + ν )  
E  Rco2 − Rci2    Rci  
 
49.5x10−3  182 x106 (49.5x10 −3 ) 2    100  
2
−6
= 1 − 0.2 + (1 + 0.3)   = 85x10 m
210x109  (100x10 −3 ) 2 − (49.5x10 −3 ) 2   49.5 
Therefore, the total radial interference is, (11.45)

37
 δ = uc + us = 115.3& x10 −6 m

11.6 Let the inner and outer cylinders be denoted by vessels 1 and 2 as in §11.10.
From (11.52), (11.51), (11.48) and (11.27) the total radial interference is given by
δ = u2 − u1 =
b
E
( [
σ θθ ,2 − νσ rr ,2 ) − (σ θθ ,1 − νσ rr ,1 ) = ]
1 2
Eb
[
b (1 − ν )( A2 − A1 ) + (1 + ν )( B2 − B1 ) ]
where

A2 − A1 = 2
pb 2 a 2 − c 2( )
, B2 − B1 = 2
pb 4 a 2 − c 2 ( )
= b 2 ( A2 − A1 )
(
a −b c −b
2 2
)(
2
) (
a −b c −b
2 2 2
)( )
Substituting (A2-A1) and (B2-B1) into δ we finally arrive at

δ=
b( A2 − A1 ) pb 3
=
a 2 − c2 ( )
E E a 2 − b2 c2 − b2 ( )( )
Re-arranging for p then the interference pressure is

p= 3
(
Eδ a − b c − b
2 2 2
)(2

=
) 
( ) (
210x109 100x10−6  502 − 752 1002 − 752 )( ) x10 −6
= 91MPa
b (
a2 − c2 ) 75x10 −3
3
(
 )
502 − 1002 ( ) 

11.7 From (11.72) with p=100MPa, a=100mm, b=175mm and r=100mm then the
circumferential stress is
p  2 + (b / r )  100  2 + (175 / 100) 
3 3

σ θθ =  =   = 68.66MPa
2  (b / a) 3 − 1 2  (175 / 100) 3 − 1 

38
Chapter 12 Solutions

12.1 Letting the total strain, ε, be the sum of the elastic, εe, and plastic, εp, strains and
with εp=εe/5 then ε is
ε 6ε
ε = εe + ε p = εe + e = e
5 5
From Hooke's law εe=σe/E and at the point of yielding then σe=σY and the total strain
is
6σ Y
ε=
5E
Substituting this total strain into the constitutive equation
E  6σ Y 
1/ 5

σY =  
200  5E 
Solving for σY then we arrive at
E
. x10 −3 E =
σ Y = 139
719

12.2 From (12.9) the mean yield stress, σm, is

1 1 ε  σ  ε σ 
σ m = ∫ σ Y dε = ∫ σ Y + B ε − Y  dε = σ Y + B − Y 
ε ε 0  E  2 E 

12.3 To determine the empirical constants C and n of Ludwik's power law from the
given engineering stress and strain data then we require relations (12.11) and (12.14)
which relate the true and engineering components, that is
σ = σ 0 (1 + e0 ) = 340(1 + 0.3) = 442 MPa
ε = ln(1 + e0 ) = ln(1 + 0.3) = 0.2624
Inserting these into the Ludwik power law, (12.1), we have
442 = C (0.2624)
n

At the point of plastic instability we know from (12.15) that σ=dσ/dε, or from
Example 12.2 that σ=Cnn and ε=n for Ludwik's power law. Therefore, n=ε=0.2624
and upon substitution into the previous equation and solving for C we have
442 = C( 0.2624) ⇒ C = 627.9
0.2624

12.4 The maximum bending stress, σ, and shear stress, τmax, for the circular cross-
section bar are, (5.30) and (3.11)
My M ( d / 2) 32 M TR Td 16T
σ= = = , τ max = = =
I I πd 3
J 2 J πd 3
and substituting into (12.61)
2 2
 32 M   16T 
 3 
+ 3 3  = σ Y2
 πd   πd 
Substituting M=cT and dividing by τmax we arrive at the required result
σY
= 3 + 4c 2
τ max

39
12.5 Tresca's yield criterion is given by (12.29) with the difference in principal
stresses obtained from (12.31). Therefore, the yield stress is

(σ ) (500 − 100) 2 + 4(100) 2

2
σY = xx − σ yy + 4τ 2xy = = 447MPa
To determine the value of the yield stress according to the Huber-von Mises yield
criterion we will first evaluate the principal stresses, (12.30)
 σ xx + σ yy   σ xx − σ yy 
2

σ 1,2 = ±   + τ xy
2

 2   2 

 500 + 100   500 − 100 

2

= ±   + 100 = 523MPa , 76MPa

2
 2   2 
From (12.46) the yield stress is
σ Y = σ 12 − σ 1σ 2 + σ 22 = 5232 − (523)(76) + 76 2 = 489MPa

12.6 From (12.66) the total torque, T, consisting of the elastic torque, TE, and plastic
torque, TP, is
 1  Rp  3  2  1  16  3 
2
T = TE + TP = πkR 1 −    = πx175x10 25x10
3 6 −3 3
(
1 −    = 5.352kNm )
3  4  R   3  4  25 
with R=50/2=25mm and Rp=25-9=16mm. At first yield Rp=R and the torque is equal
to
π
TY = kR 3 = 4.295kNm
2
and when the entire section is fully plastic then Rp=0 and the torque is equal to
2
TFP = πkR 3 = 5.727kNm
3

12.7 From (12.90) the applied bending moment, MY, at first yield is
( )
−2 −2 2
bh 2 σ Y 2.5x10 4 x10 x250x10 6 .
MY = = = 166
. kNm
6 6
From (12.96) the value of applied moment, M, to cause the plasticity to extend to a
depth of 1cm is
 1. 6. x10 3 
3 
 1  1x10 −2  2 
3 MY  1  y0    
2

M= 1 −   = 1 −  −2 
 = 2.29 kNm
2  3  h / 2   2  3  2 x10  
For the plasticity to spread throughout the entire cross-section then y0=0 at which
point the bending moment is
.
3 1. 6 x10 3 
3MY  
M FP = = = 2.5kNm
2 2

40
Chapter 13 Solutions
13.1 At the nodes i and j we have, (13.23)
φ i = α 1 + α 2 xi , φ j = α 1 + α 2 x j
Solving for α1 and α2 we find
φ i x j − φ j xi φ j − φi
α1 = , α2 =
L L
and substituting into the interpolation function (13.23) then φ is
φ x − φ j xi
φ= i j +
( )
φ j − φi x  xj − x
=
 x − xi 
φ i +   φ = Niφ i + N jφ j
L L  L   L  j
where Ni and Nj are the shape functions of the element and have the following
properties:

• Ni=1 at x=xi and Ni=0 at x=xj.

• Nj=1 at x=xj and Nj=0 at x=xi.
• The sum of Ni and Nj is always equal to unity for x within the range xi≤x≤xj.
• The shape functions are of the same order as the interpolation function.

With xi=2 and xj=6 then the L=xj-xi=4. With φi=10 at xi=2 and φj=20 at xj=6 then from
the above interpolation function at x=3 we have Ni=3/4 and Nj=1/4 with φ=12.5. The
value of φ at x=3 is seen to be a linear interpolation of the nodal values.

13.2 From (13.33) the D matrix for plane stress is

1 ν 0   1 0.28 0 
E   9
[ D] = 1 − ν 2 ν 1 0  = 75.95x10 0.28 1 0 

 0 0 (1 − ν ) / 2   0 0 0.36
and from (13.36) the D matrix for plane strain is
1 − ν ν 0  0.72 0.28 0 
  9
0 
E
[ D] = 1 + ν 1 − 2ν  ν 1 − ν 0  = 124.29 x10 0.28 0.72
( )( ) 0
 0 (1 − 2ν ) / 2  0 0 0.22 
From (13.32) the stress vector for plane stress is
σ xx  ε xx   1 0.28 0  60 6.26
    9     
σ yy  = [ D]ε yy  = 75.95x10 0.28 1
−6
0  80x10 = 7.35x10 6
τ  γ   0 0 0.36 55  15 
 xy   xy   . 
and similarly for plane strain
σ xx  ε xx  0.72 0.28 0  60 815
. 
    9     
σ yy  = [ D]ε yy  = 124.29 x10 0.28 0.721 0  80x10 = 9.25x10
−6 6

τ  γ   0 0.22 55  15 
 xy   xy  0  . 

13.3 The cross-sectional areas of elements 1 and 2 are

πD12 π 80 2 πD22 π 50 2
A1 = = = 5,027 mm 2 , A2 = = = 1,964 mm 2
4 4 4 4
The element stiffness matrices are, (13.44)

41
 (
5,027 200x10 3  1 −1)  1 −1
[K ]
1
=
500  
 −1 1 
= 2010.8x10 3   N / mm
 −1 1 
(
1964 120x10 3  1 −1) 3
1 −1
[ ]2
K =
700  −1 1 
 
= 336.7 x10 −1 1  N / mm
 
The structure stiffness matrix and force vector are, using node ordering (1,2,3)
 2010.8 −2010.8 0   0 
[ K ] s = −2010.8 2018. + 336.6 −336.7 N / mm , {F } s =  0 x103 N
 
 0 −336.7 336.7  185
 
Incorporating the boundary condition u1=0 then the structure system of equations to
be solved for U is
 2010.8 −2010.8 0   0  0 + R1 
   
10 −2010.8 2347.5 −336.7 u2  =  0 x10 3
3

 0 −336.7 336.7  u3   −185 

where R1 is the reaction at node 1. Performing row multiplications we have
10 3 ( −2010.8u2 ) = R1
10 3 ( 2347.5u2 − 336.7u3 ) = 0
10 3 ( −336.7u2 + 336.7u3 ) = −185x10 3
Solving these equations we find u1=0, u2=-0.092mm and u3=-0.6415mm. Both u2 and
u3 are negative and u3<u2 as expected. An additional check illustrates that R1 is equal
and opposite to the applied force of -185kN. From (13.30) the element strains are
1 1
ε 1xx = ( − u1 + u2 ) = (0 − 0.092) = −184x10−6
L 500
1 1
ε 2xx = ( − u2 + u3 ) = (0.092 − 0.6415) = −785x10 −6
L 700
From (13.32) the element stresses are
( )
σ 1xx = E1ε 1xx = 200x10 3 −184 x10 −6 = −36.8N / mm 2
σ 2xx = E 2 ε 2xx = 120x10 ( −785x10 ) = −94.28N / mm
3 −6 2

and are found to agree exactly with the theoretical estimates

P 185x10 3
σ 1xx = =− = −36.8N / mm 2
A1 5,027
P 185x10 3
σ 2
xx = =− = −94.28N / mm 2
A2 1,964

13.4 The stiffness matrices of elements 1 and 2 are, (13.81)

 0.25   0.25 
 −0.433 0.75   
[ ] [ ]
sym.  , K 2 =  0.433 0.75 sym.
K1 =  
 −0.25 0.433 0.25   −0.25 −0.433 0.25 
   
 0.433 −0.75 −0.433 0.75 −0.433 −0.75 0.433 0.75
Inserting into the global stiffness matrix we find the structure system of equations

42
  0.25  0   0 
−0.433 0.75  0   0 
    
 −0 .25 0.433 0 .25 sym .   u   0 
  = 
2
346.41x106  
 0.433 −0.75 0 15
.  v2  −1000
 0 0 −0.25 −0.433 0.25  0   0 
    
 0 0 −0.433 −0.75 0.433 0.75  0   0 
with node ordering 1, 2 and 3. Multiplying rows 3 and 4 we find
0.25u2 = 0
346.41x106 (15 . v2 ) = −1000
-6
leading to u2=0 and v2=-1.9245x10 m; with u2=0 due to symmetry. Similarly, the
axial strain, stress and force components in elements 1 and 2 are equivalent and so we
will consider only element 1. The axial strain is, from (13.83)
u1 
v 
 1
{ }ε = [ − C − S C S ] 
1 1
L1 u2 
v 2 
 0 
 0 
 
[ -0.5 0.866 0.5 -0.866]
1 −6
= x10 = 14433
. x10 −6
.
11547  0 
−19245
. 
The axial stress and force are
σ 1 = Eε 1 = 0.2887MPa , F 1 = σ 1 A = 577.32N
From the free-body diagram in Figure 13.28 we observe that 2Tcos30°=1kN
and therefore T=577.35N which agrees exactly with the finite element prediction. The
displacement v2 is found by an application of Castigliano’s second theorem as in
Exercise 9.4
Pd
v2 = − = −19245
. x10 −6 m
2 EA cos ϑ
3

which, again, is exact agreement with the finite element estimate.

13.5 The cross-sectional area, A, and perimeter, P, for both elements are 25x10-6m2
and 20x10-3m respectively. Letting elements 1 and 2 have nodes (1,2) and (2,3)
respectively then the stiffness matrix for element 1 has contributions due to
conduction and perimeter convection, (13.118)
k A  1 −1 hPL1 2 1  516 −4.9167
[ ]
. −3
K 1 = xx   +   =  x10
L1 −1 1  6 1 2 −4.9167 .
516 
with the force vector due to {Fh} only, (13.119)
hT PL 1 0
{ } F1 = ∞ 1   =  
2 1 0
Element 2 is identical to element 1 except that it also experiences end-convection
through node 2 so that the stiffness matrix of element 3 is, (13.118)

43
 −4.9167  −4.9167
[ K ] = −4.9167 0 0  516
516
. .
3
 x10 −3 + hA  =  x10
−3

516
.  0 1   − 4 .9167 5185
. 
Similarly, adding the end-convection term to the force vector of element 2 we have
0 0 0
{ } F 3 =   + hT∞ A  =  
0 1 0
Assembling the element contributions into the structure stiffness matrix and force
vector we find
 516 . −4.9167 0  T1  0
−3     
10 −4.9167 516 . + 516 . −4.9167 T2  = 0
   
 0 −4.9167 5185
.  T3  0
with node ordering 1 to 3. We have a prescribed temperature of 100oC at node 1
which results in a non-homogeneous boundary condition. The stiffness matrix and
force vector are modified by first setting all non-diagonal terms in the first row and
column of the stiffness matrix to zero. Also, the term (-4.9167)x100oC=-491.67 on the
left hand side of the second equation is transposed to the right hand side as +491.67.
The resulting systems of equations is now given by
1 0 0  T1   100 
−3     
10 0 10.32 −4.9167 T2  = 49167 . 
0 −4.9167 .
5185  T3   0 
The second through to third equations are now solved in the usual manner, with the
solution vector given by {T}={100,86.87,82.34}.

13.6 All three elements experience no perimeter convection (P=0) with element 3
experiencing convection at node 4. Assuming a unit cross-sectional area for all three
elements then the stiffness matrix and force vector for element 1 are, (13.118) and
(13.119)
k A  1 −1  1 −1 0
[ ]
K 1 = xx 
L1  −1 1  = 01.
 −1 1 
 { }
, F1 =  
0
and similarly for element 2
k A  1 −1  1 −1 0
[ ]
K 2 = xx 
L2 −1 1   = 01
.
 −1 1 
 { }
, F2 =  
0
Element 3 consists of an additional end-convection term
k A  1 −1 0 0  01 −01 . 0  0 
[ ] { }
.
K 3 = xx   + hA  =  , F 3 = hT∞ A  =  
L3  − 1 1  0 1 −01 . 501 . 1 −250
Assembling the elements into the structure stiffness matrix and force vector gives,
with node ordering 1 to 4
 01. −01
. 0 0   T1   0 
 −01 0  T2   0 
 . 0.2 −01 .
 = 
 0 −01 . 0.2 −01 .  T3   0 
 
 0 0 −01 . 501 .  T4  −250
Incorporating the prescribed boundary condition T1=25oC then the system of
equations is modified as follows

44
 1 0 0 0  T1   21 
0 0.2 −01 . 0  T2   2.5 
  = 
0 −01 . 0.2 −01 .  T3   0 
 
0 0 −01 . 501 .  T4  −250
where the right hand side of the first equation is set to 25oC. The term (-0.1)x25oC=-
2.5 on the left hand side of the second equation is transposed to the right hand side as
+2.5. Solution of the system of equations yields {T}={25,15,5,-5}. From Fourier's law
the heat flux for an element of length L and nodes i and j is
dT  1 1   Ti 
q x = − k xx = k xx [ B]{T} = − k xx −  
dx  L L  Tj 
Evaluation of qx for all three elements reveals that the heat flux is constant and equal
to Qx=qxA=qx=1 for all three elements.

13.7 The stiffness matrices for all three elements are equivalent and given by (13.136)
Ak xx  1 −1
[ ] [ ] [ ]
K1 = K2 = K 3 =
L  −1 1 

=
(31416
. x10 )1x10
02 −2
 1 −1
= 9.425x10 −4 
 1 −1
0.33&   
−1 1   −1 1 
Since there are no sources or sinks and no applied surface flow rates then both Q and
q are equal to zero in (13.138) so that the element force vectors are
0
{ } { } { }
F1 = F 2 = F 3 =  
0
Assembling the element components we have the following structure system of
equations for unknown fluid heads p1, …, p4
 1 −1 0 0   p1  0
−1 2 −1 0   p  0
9.425x10 −4    2  =  
 0 −1 2 −1  p3  0
 
 0 0 −1 1   p4  0
Incorporating the non-homogeneous boundary conditions of p1=0.2m and p4=0.1m in
a similar manner to that discussed in Example 13.4 we arrive at
1 0 0 0  p1  1885 . x10 −4 
0 2 −1 0  p   
−4   . x10 −4 
 2  1885
9.425x10  = 
0 −1 2 0  p3  9.425x10 −5 
   
  p4  9.425x10 
−5
0 0 0
Solving the second and third equations for p2 and p3 then the solution vector is
{P}={0.2,0.16,0.13,0.2}. From (13.130) the element velocity for an element of length
L with nodes i and j is
 1 1   pi 
v x = − k xx { g} = − k xx [ B]{ P} = − k xx  −  
 L L   p j 
For example, for element 1 we have
 1 1   0.2 
v x = −1x10 −2  − 
−3
 &  = 1x10 m / s
 0.33& &
0.33  016
. 

45
with equivalent velocities for elements 2 and 3.

46
Chapter 14 Solutions
14.1 From §1.11 the in-plane strains are
∂u ∂v 1  ∂u ∂v  1
ε xx =
∂x
= 2axy 2 , ε yy =
∂y
= bx 3 , γ xy = (
 +  = 2ax y + 3bx y
2  ∂y ∂x  2
2 2
)
Differentiating the strains
∂ 2 ε xx ∂ ∂ 2 ε yy ∂
∂y 2
= (4axy ) = 4ax ,
∂y ∂x 2
=
∂x
( )
3bx 2 = 6bx

∂ 2γ ∂ 1
(4axy + 6bxy) = (4ax + 6bx)
1
=
xy

∂x∂y 
∂y  2  2
Upon substituting these into the compatibility equation, (14.21), we observe that u
and v are compatible.

14.2 Differentiating φ we find

∂φ ∂ 2φ ∂ 3φ
= 2 Ax + By , = 2 A , =0 , L
∂x ∂x 2 ∂x 3
∂φ ∂ 2φ ∂ 3φ
= 2Cy + Bx , = 2C , =0 , L
∂y ∂y 2 ∂y 3
∂ 2φ
=0 , L
∂x 2 ∂y
Thus, we observe that φ satisfies ∇4φ=0 since all terms of φ are less than power 4.
The stresses follow immediately from the Airy stresses, (14.46)
∂ 2φ ∂
σ xx = 2 = ( Bx + 2Cy ) = 2C
∂y ∂y
∂ 2φ ∂
σ yy = = (2 Ax + Bx ) = 2 A
∂x 2 ∂x
∂ 2φ ∂
τ xy = − = − (2 Ax + By ) = − B
∂x∂y ∂x
It follows that φ provides a solution for a plate subject to uniform stresses along its
sides of σxx=2C, σyy=2A and τxy=-B.
From the Hookian equations for a state of plane stress, (14.42), the in-plane
strains are
ε xx =
1
E
[ ] 2
σ xx − νσ yy = (C − νA)
E
ε yy =
1
E
[ ] 2
σ yy − νσ xx = ( A − νC )
E
2(1 + ν ) 2(1 + ν )
γ xy = τ xy = − B
E E
Integrating the strains we find the displacements
2 2
u = ∫ ε xx dx = (C − νA)dx = ( C − νA) x + f ( y )
E E
2 2
v = ∫ ε yy dy = ( A − νC )dy = ( A − νC ) y + g ( x )
E E

47
where the functions f(y) and g(x) are determined from the boundary conditions.

14.3 Let the resultant stress be S(Sx,Sy,Sz) with

S 2 = S x2 + S y2 + S z2
This resultant stress consists of both normal, SN, and shear, SS, components
S 2 = S N2 + S S2
If the direction cosines of ABC are l=cosα, m=cosβ and n=cosγ then
S N = lS x + mS y + nS z
where (Sx,Sy,Sz) in terms of the coordinate components are
S x = lσ xx + mτ xy + nτ xz
S y = lτ yx + mσ yy + nτ yz
S z = lτ zx + mτ zy + nσ zz
Substituting Sx, Sy and Sz into SN we have
(
S N = l 2σ xx + m 2 σ yy + n 2 σ zz + 2 lmτ xy + mnτ yz + lnτ zx ) , S S = S 2 − S N2
Substituting σij and (l,m,n) into Sx, Sy and Sz we have
1 30
Sx = (10 + 5 + 15) = = 17.32
3 3
1 50
Sy = (5 + 20 + 25) = = 28.87
3 3
1 70
Sz = (15 + 25 + 30) = = 40.41
3 3
with S equal to
2 2 2
 30   50   70 
S =   +   +   = 52.6
 3  3  3
The normal and shear stresses are
1  30  1  50  1  70 
SN =  +  +   = 50 , S S = 52.6 − 50 = 16.33
2 2

3  3 3  3 3  3
The direction of SN acts normal to plane ABC and is therefore defined by
(l,m,n)=(1/√3,1/√3,1/√3) with α=β=γ=cos-1(1/√3)=54.74°. A useful check is to ensure
that l2+m2+n2=1 which is the case. The direction of SS acts parallel to the plane ABC
and let it be defined by (ls,ms,ns) where ls=cosαs, ms=cosβs and ns=cosγs. The
components (Sx,Sy,Sz) can now alternatively be defined as, resolving SN and SS
S x = S N cos α + S S cos α s = lS N + l s S S
S y = S N cos β + S S cos β s = mS N + ms S S
S z = S N cos γ + S S cos γ s = nS N + n s S S
from which it follows
1
ls = ( S x − lS N ) / S S = −
2
( )
ms = S y − mS N / S S = 0
1
ns = ( S z − nS N ) / S S =
2

48
confirming that ls2 + ms2 + ns2 = 1 . The angles αs, βs and γs are αs=cos-1(-1/√2)=135°,
βs=cos-1(0)=90° and γs=cos-1(1/√2)=45°. Finally, we can check that the two direction
vectors of SN and SS are orthogonal by ensuring that the dot product of (l,m,n) and
(ls,ms,ns) is equal to zero, that is (1/√3,1/√3,1/√3).( -1/√2,0, 1/√2)=0.

14.4 From (14.100) the C matrix is

1 − 0.3 0.3 0  0.7 0.3 0 
210x10 9   9 
[C] = 1 + 0.3 1 − 2x0.3  0.3 1 − 0.3 0  = 404 x10  0.3 0.7 0 
( )( ) 0
 0 (1 − 2x0.3) / 2  0 0 0.4 
T -6
With [ε] =[-19,64,3]x10 then the stress vector is
σ xx  0.7 0.3 0  −19  2.38
  9    −6  
σ yy  = 404 x10  0.3 0.7 0   64 x10 = 158 . MPa
τ   0 0 0.4  3  0.48
 xy   

14.5 From (14.190) the shear stresses τxz and τyz are
∂φ ∂φ 3Gθ  2ax 
τ xz = = − Gθy[1 − x ] , τ yz = − =  − x2 + y2 
∂y ∂x 2a  3 
The centroid of the triangle is at (x,y)=(0,0) and the three corners are at (2a/3,0), (-
a/3, a / 3 ) and (-a/3,- a / 3 ). Substituting these coordinates into τyz above we
observe that τyz is equal to zero at the centroid and three corners.

14.6 Since the hole in the plate is circular, a/b=1, then from (14.212) the stress
concentration factor is Kt=3. The maximum of σP that can be applied to the plate is
therefore
σ 300MPa
σ P (max) = Y = = 100MPa
Kt 3
From (14.210) the σyy stress at a distance x=a/2=6.25mm from the notch root (notch
root radius of ρ=b2/a=a=12.5mm) is
ρ 12.5
σ yy = σ P K t = 100x10 6 x3 = 173MPa
ρ + 4x 12.5 + 4(6.25)

14.7 From (7.19) the principal stresses are given by

 σ xx + σ yy   σ xx − σ yy 
2

σ 1,2 = ±   + τ xy
2

 2   2 
With the stresses given by (14.223) for the semi-infinite Boussinesq wedge then the
three main terms in the above expression are
σ xx + σ yy
2
f
(
= − 4 x 3 + xy 2
πr
)
 σ xx − σ yy 
2 2
 f 
( )
2
  =  4  xy − x
2 3

 2   πr 
2
 f 
τ 2xy =  4  4 x 4 y 2
 πr 

49
Noting that the square root term reduces to
 σ xx − σ yy 
2 2
 f  2 2

 2   πr 
(
 + τ xy =  2  x x + y
2 2
)
and substituting into the principal stresses we finally arrive at the required result
2 fx
σ1 = 0 , σ2 = − 2
πr

50
Chapter 15 Solutions
15.1 Refer to §15.2 for a discussion on equivalent stress and strain.

15.2 Refer to §15.2 for a discussion on the constancy of volume condition and its
application to illustrate that Poission’s ratio is equal to ½ for incompressible
materials.

15.3 For a thin-walled pressure vessel with no shear stresses then the circumferential,
axial and radial coordinate stresses σθθ(=pd/2t), σzz(=σθθ/2) and σrr are equivalent to
the principal stresses σ1, σ2 and σ3; where p is the internal pressure, d is the mean
diameter and t is the wall thickness. From (15.1) the equivalent stress is

(σ θθ − σ zz )2 + σ 2zz + σ θθ2 = 2 σ θθ
1 3
σ=
2
The equivalent plastic strain is given by (15.2) with the plastic strain increments dε 1p ,
dε 2p and dε 3p determined from the Levy-Mises flow rule. The deviatoric component
of σθθ is
 σ + σ zz + σ rr  2  
 = σ θθ − (σ zz + σ rr )
1
σ θθ
'
= σ θθ −  θθ
 3  3 2 
The plastic strain increments are, from (15.22)
2λ  1  λ
dε 1p = dε θθ
p
= σ − (σ + σ ) = 2 σ θθ
3 
θθ zz rr
2
2λ  
σ zz − (σ θθ + σ rr ) = 0
1
dε 2p = dε zzp = 
3  2 
2λ   λ
σ rr − (σ θθ + σ zz ) = − σ θθ
1
dε 3p = dε rrp = 
3  2  2
with the equivalent plastic strain given by, (15.2)
2
( ) ( ) ( )
2 2 2
dε p = dε θθ
p
− dε zzp + dε zzp − dε rrp + dε θθp
− dε rrp
3

15.4 In the absence of shearing stresses, τrz, then the principal stresses are equal to the
coordinate stresses σθθ, σrr and σzz. With the external pressure po=0 and the internal
presure pi=p then we have from Lame's equations, (15.34)
σ rr = − p
 k 2 + 1
σ θθ =  2
p  ; k =b/a
 k − 1
σ zz = 0 (open ends)
Let us examine both the Tresca and Huber-von Mises yield criteria starting with
Tresca's criterion.

Tresca's Yield Criterion

Since the coordinate stresses are equivalent to the principal stresses, (15.33)
σ θθ − σ rr = σ Y

51
Substituting the radial and circumferential stresses we find that first yield occurs when
σ  1
p = Y 1 − 2 
2  k 
Re-arranging for k
 (σ Y / p) 
1/ 2

k= 
 (σ Y / p) − 2 
With p=100MPa and σY=250MPa then k=√5=2.24.

Huber-von Mises Yield Criterion

From the Huber-von Mises yield criterion

(σ rr − σ θθ ) + (σ θθ − σ zz ) + (σ zz − σ rr )2 = 2σ Y2
2 2

With σzz=0 for open ends then this equation reduces to

σ rr2 − σ rr σ θθ + σ θθ
2
= σ Y2
Substituting the radial and circumferential stresses and re-arranging for p we arrive at
k2 − 1
p =σY
3k 4 + 1
Re-arranging for k we arrive at the following quadratic equation with unknown k2
[ ] ( )
3 p 2 − σ Y2 k 4 + 2σ Y2 k 2 + p 2 − σ Y2 = 0
Solving, then k is given by

( )( )
1/ 2
 −σ 2 ± σ 4 − 3 p 2 − σ 2 p 2 − σ 2 
k= 
Y Y Y Y

 3p − σ Y
2 2

 
With p=100MPa and σY=250MPa then the two solutions of k are k=0.69 and k=1.83.
Since k>1 then k=1.83 and is approximately 22% less than the Tresca prediction and
results in approximately 70% difference in cross-sectional area.

15.5 The required applied internal pressure, p, to produce an elastic-plastic boundary

to a depth of c=70mm can be found by setting r=a and p=-σrr in (15.40)
  c  1   c  2    70  1   70  2  
p = −σ rr = σ Y ln  + 1 −     = 300ln  + 1 −     = 177.45MPa
  a  2   b      50  2   100   
The fully plastic condition is reached when c=b and the required pressure, pY, is
 b  100 
pY = σ Y ln  = 300 ln  = 207.94MPa
 a  50 

15.6 From the material power law

dσ
( )
n −1
= nC ε P
dε P

and upon substitution into the plastic instability condition (15.48) we have
( ) ( ) n
n −1 n
nC ε P = 3σ = 3C ε P ⇒ εP = = 0.26
3
From (15.49) the mean radius and wall thickness at the point of plastic instability are
r = r0 e 3ε / 2 = 0.45e 3 (0.26)/ 2 = 0.56m , t = t 0 e − 3ε / 2 = 1e − 3 ( 0.26)/ 2 = 0.8mm
P P

52
15.7 From Exercise 15.6 the equivalent plastic strain remains the same. From (15.57)
the mean radius and wall thickness at the point of plastic instability are
r = r0 e ε / 2 = 0.45e 0.26 / 2 = 0.51m , t = t 0 e − ε / 2 = 1e −0.26/ 2 = 0.88mm
P P

53
Chapter 16 Solutions
16.1 From (16.5) and (16.6) the constants a, b and c and area A are
ai = 8 , bi = −2 , ci = −1
a j = −2 , b j = 3 , c j = −1
a k = −1 , bk = −1 , ck = 2
A = 2.5
From (16.8) the shape functions at point p are
1 1 3
Ni = , N j = , Nk =
5 5 5
From (16.11) the displacement vector at point p is
 ui 
v 
 i
u  N i 0 N j 0 Nk 0  u j 
 =  =
v   0 N i 0 N j 0 N k  v j 
u k 
 
v k 
 1
 1
 
1 / 5 0 1 / 5 0 3 / 5 0   3  2 
=   =  
 0 1 / 5 0 1 / 5 0 3 / 5 4 2.2
2
 
2

16.2 Denoting nodes (1,2,3) by (i,j,k) to assist in the use of the required formulae then
from (16.5) and (16.6) the constants a, b and c and area A are
ai = 9 , bi = −2 , ci = −1
a j = −3 , b j = 2 , c j = −1
a k = −2 , bk = 0 , c k = 2
A=2
From (16.8) the shape functions are
1 1 1
N i = [9 − 2 x − y ] , N j = [ −3 + 2 x − y ] , N k = [ −2 + 2 y ]
4 4 4
From (16.11) the [N] matrix is
N 0 N 0 N 0 
[ N ] =  0i N 0 j N 0k N 
 i j k

From (16.17) the [B] matrix is

−2 0 2 0 0 0
[ B] = 4  0 −1 0 −1 0 2
1

 −1 −2 −1 2 2 0
From (13.33) the [D] matrix is

54
  1 0.3 0 
[ D] = 219 x10 0.3 1 0  N / mm2
3

 0 0 0.35
From (16.17) the strain vector is
1 
1 
ε xx  −2 0 2 0 0 0    0.5 
2 
{ε } = ε yy  = [ B]{U } =  0 −1 0 −1 0 2  =  1.5 x10−3
  1
γ  4
 − − −  3 2.25
 xy  1 2 1 2 2 0  4   
 
5 
From (16.18) the stress vector is
σ xx   1 0.3 0   0.5   209 
  3   
{σ } = σ yy  = [ D]{ε } = 220x10 0.3 1 0   1.5 x10 =  363 N / mm 2
−3

τ   0 0 0.35 2.25 173.25

 xy   

16.3 With xi=3, yi=3, xj=4, yj=1, xk=5 and yk=3 then the constants a, b and c and the
area, A, of the element are
ai = 7 x10 −6 m 2 , bi = −2 x10 −3 m , ci = 1x10 −3 m
a j = 6x10 −6 m 2 , b j = 0m , c j = −2 x10 −3 m
a k = −9 x10 −6 m 2 , bk = 2 x10 −3 m , ck = 1x10 −3 m
A = 2x10 -6 m 2
From (16.17) the [B] matrix is
bi 0 b j 0 bk 0   −2 0 0 0 2 0 
1  
[ B] = 2 A  0 ci 0 c j 0 ck  = 250 0 1 0 −2 0 1
ci bi c j b j c k bk   1 2 −2 0 1 2 
 
and from (13.33) the [D] matrix is
1 ν 0   1 0.3 0 
E   11 
[ D] = 1 − ν 2 ν 1 0 
 = 2 1978 10 
0 . x 0.3 1

 0 0 (1 − ν ) / 2  0 0 0.35
The stiffness matrix is, (16.34)
[ K ] = [ B]T [ D][ B]tA =
 −2 0 1
0 1 2 
  1 0.3 0   −2 0 0 0 2 0
0 0 −2  
= 250 11
(2.1978x10 ) 0.3 1 0 250 0 1 0 −2 0 1(2 x10 −3 )(2 x10 −6 )
 
 0 −2 0   0 0 0.35  1 2 −2 0 1 2 
2 0 1
 
0 1 2
Performing the multiplications we find

55
  4.35 
 01. 2.4 
 
 0.7 14. 14
. sym. 
[ K ] = 5.4945x107  12 −2

 . 0 4 
−4.35 −01 . −0.7 −12 . 4.35 
 
 −13. −0.4 −0.4 −2 13
. 2.4
noting that the matrix is symmetric (Kij=Kji), the principal terms are all positive and
non-zero (Kii>0) and the sum of all terms in either a row or column are zero
( ∑ j =1 Kij = 0 ).
6

To determine the force vector we require just the contribution due to the edge
pressure. From (16.42) the normal pressure term for edge (i,k) is
 px   0 
p   
 y  −200
L t0  0 
{ F } = {F } pressure = ik  0  =  
2    0 
 px   0 
   
 p y  −200

16.4 Measuring the length coordinate ξ from node 1 then ξ=3/4 for point p. From
(16.60) the shape functions are
2
 3  3 1
N 1 = 2ξ − 2ξ + 1 = 2  − 2  + 1 = −
2
 4  4 8
2
 3 3 3
N 2 = 2ξ − ξ = 2  − =
2
 4 4 8
2
 3  3 3
N 3 = 4ξ − 4ξ 2 = 4  − 4  =
 4  4 4
The displacement u at point p is, (16.59)
 1  3  3
u = N 1 u1 + N 2 u2 + N 3 u3 =  −  2 +   2.25 +   2.55 = 2.51
 8  8  4

16.5 To determine the element shape functions (16.84) we first require the area
coordinates of point p. From (16.6) we find
5 1 1 3
A= , A1 = , A2 = , A3 =
2 2 2 2
From (16.43) the area coordinates of p are
A 1 A 1 A 3
ξ1 = 1 = , ξ2 = 2 = , ξ3 = 3 =
A 5 A 5 A 5
confirming that ξ3=1-ξ1-ξ2. Substituting ξ1, ξ2 and ξ3 into (16.84)

56
  1 1 3
N 1 = (2ξ 1 − 1)ξ 1 =  2 − 1 = −
 5 5 25
 1 1 3
N 2 = (2ξ 2 − 1)ξ 2 =  2 − 1 = −
 5 5 25
 3 1 3
N 3 = (2ξ 3 − 1)ξ 3 =  2 − 1 =
 5  5 25
1 1 4
N 4 = 4ξ 1ξ 2 = 4 ⋅ =
5 5 25
1 3 12
N 5 = 4ξ 2 ξ 3 = 4 ⋅ =
5 5 25
1 3 12
N 4 = 4ξ 1ξ 3 = 4 ⋅ =
5 5 25
confirming that ΣNi=1. The displacement (u,v) at point p is
u1 
v 
u  N 1 0 L N 6 0    315 . 
1

=
     =
M 
v   0 N 1 L 0 N 6  u   38 . 
 6
v 6 
16.6 From (16.60) the element shape functions are
N 1 = 2ξ 2 − 3ξ + 1 , N 2 = 2ξ 2 − ξ , N 3 = 4ξ − 4ξ 2
so that the derivatives with respect to ξ are
∂N 1 ∂N 2 ∂N 3
= 4ξ − 3 , = 4ξ − 1 , = 4 − 8ξ
∂ξ ∂ξ ∂ξ
From (16.70) the Jacobian matrix is
∂x ∂N ∂N ∂N
[ J ] = ∂ξ = ∂ξ1 x1 + ∂ξ2 x2 + ∂ξ3 x3 = (4ξ − 3)1 + (4ξ − 1)5 + (4 − 8ξ )3 = 4

16.7 From (16.53)

1 1 n
I y = ∫∫ x dA = ∫ ∫ f (ξ 1 , ξ 2 ) J dξ 1dξ 2 = 2 A∑ wi f (ξ 1 , ξ 2 )
2
A
0 0 i =1

noting that |J|=2A. With reference to Table 16.2 the x coordinate at the three
integration points a, b and c are
 1  1  2 1
3
x a = ∑ ξ ai xi = 0  + 3  + 0  =
i =1
 6   6  3 2
 2  1  1 1
3
xb = ∑ ξ bi xi = 0  + 3  + 0  =
i =1
 3  6  6 2
 1  2  1
3
x c = ∑ ξ ic x i = 0  + 3  + 0  = 2
i =1
 6  3  6
With A=6 and wi=1/6 for all three integration points then Iy is
 1  1  2 1  1  2 1 2 
I y = 2 x 6   +   + 2  = 9
 6  2  6  2 6 

57
An exact evaluation of Iy can be performed by referring to Example 2.1. Considering
an elemental strip dy at a distance y from the x-axis then
hx
y =h−
b
With dA=ydx then Iy is
 hx 
(h b
) hb 3
b b
I y = ∫∫ x 2 dA = ∫ x 2 ydx = ∫ x 2  h −  dx = ∫ bx 2 − x 3 dx =
A 0 0  b b 0 12
3
With b=3 and h=4 then Iy=4(3) /12=9 which agrees exactly with the evaluation of Iy
using Gaussian integration.

58
Chapter 17 Solutions
17.1 Refer to §17.2 for a discussion of the stress intensity factor, §17.3 for a
discussion of the T-stress and §17.5 for a discussion of several well-known stress
intensity and T-stress expressions.

17.2 For a symmetrically cracked circular hole in a plate with far-field uniform
loading consider the two limits of a→0 and a→∞ in the given expression. For a→0
we have
2 .4
 R 
  → 1 and K I → 3.365σ πa
 R + a
and for the limit a→∞ we have
2 .4
 R 
  → 0 and K I → σ πa
 R + a
Thus, as the cracks grow beyond the influence of the circular hole then KI tends to the
case of a centrally cracked plate.
The case of a→0 requires further analysis. When a crack is short (a<<R) then
the crack is approximately equivalent to an edge crack in a semi-infinite plate but with
the applied stress modified by the circular hole stress concentration factor of 3σ. In
this case KI is given by
K I = 11215
. (3σ ) πa = 3.3645σ πa
which is essentially equivalent to the given equation by letting a→0. Similarly, for the
T-stress as a→0 we have
T = −0.5258( 3σ ) = −15774
. σ

17.3 Refer to §17.8 for a discussion of plane strain fracture toughness and its
experimental determination.

17.4 The total crack length is 2a=25mm so that a=12.5mm. Neglecting finite plate
effects with Y=1 then for case i) we have
( )
K IC = σ f πa = 220 π 12.5x10−3 = 43.6MNm−3/ 2
For case ii) with α=1/π for Irwin's model then from (17.91) KIC is
 σ f  
2

K IC = σ f πa 1 + απ    = 49.1MNm−3/ 2
  σ Y  
The discrepancy between the elastic and plasticity correction estimates increases as
σf/σY increases although the small-scale yielding assumption becomes increasingly
invalid.

17.5 The rotor rotational speed in radians per second is

 2π 
ω = 12 x10 3   = 1,257 radians / second
 60 
At the critical crack length then KI=KIC and a=acritical

59
 ω 2 R 2  3 − 2ν 
K IC = Yρ   πa critical
8  1− ν 
1,257 2 (0.35)  3 − 2(0.33) 
2

⇒ 85x10 = 0.55 8x10

6
( 3

8
) 
 1 − 0.33 
 πa critical

Solving for acritical we have

a critical = 16.7 mm
Comparing KI with ∆K = Y∆σ πa then we observe that the cyclic stress is
3 1,257 (0.35)  3
− 2( 0.33) 
2
ω 2 R 2  3 − 2ν 
2

∆σ = ρ   = 8 x10   = 6755
. x10 6 MPa
8  1− ν  8  1 − 0.33 
Since when operating the rotational speed is constant then the only way in which the
stress cycles is from the starting-stopping of the rotor. Using this assumption then ∆N
will provide us with the number of times that the rotor can be run up to speed. From
(17.159) the number of cycles is
1  ai1− m/ 2 − a 1f− m/ 2 
∆N =  =
C∆σ mπ m/ 2 Y m  m / 2 − 1 
1  0.012 1−3/ 2 − 0.017 1− 3/ 2 
= 3   = 250 cycles
4.11x10 −11 ( 6755. ) π 3/ 2 (0.55) 
3
0.5 

17.6 Refer to §17.11 for a discussion on long and short fatigue cracks. From Hobson’s
growth law (17.164) with α=0 and re-arranging for the number of cycles we have
∆N = ∫
1 a f da 1
[
= − ln(d − a ) a ]
af

C ai d − a C i

With ∆σ=638MPa then C is found to be

1
. x10 −34 (638)
C = 164 = 2.887 x10 −3 ; = 346
11.14

C
With the long crack condition ∆Kth=6MPa√m and Y=2/π then from ∆K = Y∆σ πa
the total threshold crack length is
2(6)
2
∆K 2
a th = 2 2 th 2 = = 138x10 −6 m = 138µm ≡ d
πY ∆σ  2
2

π   (638)
2

π
With ai=10µm and af=84% of d which is equal to 116µm then the number of cycles is
∆N = −346[ln(138 − 116) − ln(138 − 10)] = 609 cycles

17.7 For plane strain conditions then (κ+1)/8µ=(1-ν2)/E=1/E′. Therefore, from

(17.183) the mode I stress intensity factor is
KI = ( )
E ' J = 231x10 9 2.4 x10 3 = 2355
. MPa m

60
Chapter 18 Solutions
18.1 Refer to sections 18.1 to 18.8.

18.2 From (18.7) and taking logarithms we have

ln ε& s = ln A + n ln σ
From a ln ε& s versus lnσ plot with two 1 (σ=100MPa) and 2 (σ=200MPa) then the
slope is

n=
( −4
ln ε& s 2 − ln ε& s1 ln 7.6x10 − ln 4.4 x10
=
)
−6

= 7.4328
( )
ln σ 2 − ln σ 1 ln 200 − ln 100
Assuming n to be an integer and equal to 7 then the constant A is, (18.7)
ε& s 4.4 x10 −6
A= n = = 4.4 x10 −20
σ 100 7

in which the first test point has been used.

18.3 From (18.10) and taking logarithms we have

Q
ln ε& s = ln D −
RT
From a ln ε& s versus 1/T plot with two points 1 (T=160K) and 2 (T=200K) then the
slope is

− = =
(
. x10 −3 − ln 7.24 x10 −6
Q ln ε& s 2 − ln ε& s1 ln 819
= −5,625
) ( )
R 1 / T2 − 1 / T1 1 / 200 − 1 / 160
from which Q is found to be
Q = 198
. x5,625 = 11kcal / mol
Finally, solving for D at T=160K we have

ln D = ln ε& s +
Q
RT
(
= ln 7.24 x10 −6 +
11x10 3
. (160)
198
)
⇒ D = 8.7 x10 9

18.4 For the equivalents strains in the two tests then

t1 e − Q / RT1 = t 2 e − Q / RT2
Taking logarithms we have
ln t1 + ( − Q / RT1 ) = ln t 2 + ( − Q / RT2 )
and re-arranging for Q we arrive at the required result
RT1T2
Q=
T2 − T1
[ln t1 − ln t 2 ]

18.5 Re-arranging (18.29) for t/tf we have

n
 A
3
t  3
= 1 −   = 1 −   = 0.5781
tf  A0   4

18.6 From (18.43) the skeletal radius is

n / ( n −1) 4 /( 4 −1)
 (1 + 3n) R 4   (1 + 3x4)50 4 
n / ( n −1)
 J
rsk =   = (1+ 3n )/ n 
= (1+ 3 x 4 )/ 4 
= 38mm
 Jc   4nR   4 x4 x50 

61
18.7 The elastic second moment of area for the rectangular beam is
b( 2h)
3
50x753
I= =
= 1,757,812 mm 4
12 12
The creep second moment of area is, (18.47)
( 2 x 5+1)/ 5
2bn ( 2n +1)/ n 2 x50x5  75
Ic = h =   = 131,961
2n + 1 2 x5 + 1  2 
Thus, the skeletal depth is given by, (18.52)
n /( n −1) 5/ ( 5−1)
 I  1,757,812 
y sk =  =  = 25.45mm
 Ic   131,961 

62
Chapter 19 Solutions
19.1 From (19.6) the strain after 100s using the Maxwell model is
 t 1  100 1  −6
ε =  + σ 0 =  +  5x10 = 517 x10
3

µ E  1000x10 6
0.3x10 9 
and according to the Voigt model, (9.16)
σ
( ) ( [
ε = 1 − e − Et / µ 0 = 1 − exp −100x10 3 x100 / 10x10 6
E
]) 5x10 3
100x10 3
= 32 x10 −6

19.2 The total strain, ε, is the sum of the strains due to the Maxwell, εm, and Voigt, εv,
models from (19.6) and (19.16)
 t 1  σ
ε = εm + εv =  +
 µ m Em 
( − E t /µ
σ 0 + 1 − e v v 0 )Ev
At t=0 the instantaneous elastic strain is ε=σ0/Em. Differentiating ε with respect to
time then the steady creep rate is
σ  µ 
ε& = 0 1 + m e − Ev t / µ v 
µm  µv 
As t→∞ then ε& → σ 0 / µ m . The variation of ε against t is shown in Figure Sol19.2.
Clearly, when µm=0 then the model is equivalent to the standard linear solid model.
ε
σ 0 / µm
1
((Em +Ev )/Em Ev )σ0

σ0 /Em

t
Figure Sol19.2. Creep response for the model of Exercise 19.2.

19.3 From the solution of Exercise 19.2 the strain is given by

 t 1  σ
ε = εm + εv =  +
 µ m Em 
( − E t /µ
σ 0 + 1 − e v v 0 )
Ev
After the removal of stress σ0 at time t=τ then the elastic strain σ0/E is instantaneously
recovered followed by the recovery strain of, from (19.6) and (19.22)
σ t' σ
( )
ε = 0 + 0 1 − e − Evτ / µ v e − Ev t ' /µ v
µm Ev
As t→∞ then ε→σ0t′/µm. The strain-time curve is illustrated in Figure Sol19.3.

63
 ε

σ0 /Em
((Em +Ev )/Em Ev )σ0

σ0 /Em
σ0 t’/µm
t
τ
Figure Sol19.3. Strain-time curve for the model of Exercise 19.2 with removal of
stress σ0 at time t=τ.

19.4 From (9.33) the relaxation stress at time t=10s is

0.8x10 9 (10x10 −6 )
σ (10s) =
(0.8 + 0.2) x10 9
[ ]
0.2 x10 9 + 0.8x10 9 exp[ − (0.8 + 0.2) x10 9 x10 / 5x10 9 = 2.466kPa

19.5 From (19.6) and (19.10) the creep compliance, C(t), and relaxation modulus,
G(t), for the Maxwell model are
t 1
C(t ) = + , G (t ) = Ee − Et / µ
µ E
and from (19.16) the creep compliance function for the Voigt model is
1 − e − Et / µ
C(t ) =
E
with G(t)=0.

19.6 Following a similar procedure as used in the derivation of (19.6) then the total
strain rate is
1
ε& = ε& s + ε& d = σ& + Aσ n
E
For constant stress σ=σ0 then
ε& = Aσ n0 or ε = Aσ n0 + B
where B is a constant of integration. If at time t=0 the instantaneous elastic strain is
σ0/E then B=σ0/E and ε is given by
σ
ε = 0 + Aσ no t
E

19.7 From (19.42) the strain at time t=3500s is

. x10 −9 )( 3500) 0.15 +
ε (3500) = 0.5x10 6 (152
. x10 −9 )( 3500 − 1000) 0.15 +
0.25x10 6 (152
. x 10 −9 )( 3500 − 2000) 0.15 −
0.25x10 6 (152
. x10 −9 )(3500 − 3000) = 115
1x10 6 x(152 . x10 −3

64
Chapter 20 Solutions
20.1 Refer to §20.2 for a summary of different types of damage.

20.2 From (20.1) the damage parameter is

A 826
ω= = = 0.81
A0 1,018

20.3 From (20.2) the continuity parameter is

ψ = 1 − ω = 1 − 0.81 = 019
.
20.4 From (20.2) and (20.9)
E' 67
ω = 1− ψ = 1− = 1− = 0.65
E 190

20.5 From (20.15)

σ 0− n = A(n + 1)t f = Bt f
Letting m=-n and taking natural logarithms we have
ln t f = m ln σ 0 − ln B
From a (lnσ0, lntf) plot then m is given by, using the first and last data points
ln 2.17 − ln 97.11
m= = −13.32 ; n = − m = 13.32
ln 14 − ln 10
and with B given by
( )
ln B = − n ln σ 0 + ln t f = −(13.32 ln 14 + ln 2.17) = −35.9269
from which B is found
B = exp( −35.9269) = 2.5x10 −16
The constant A now follows
B 2.5x10 −16
A= = . x10 −17
= 175
n + 1 13.32 + 1

20.6 From (20.17) ψ is given by

ψ = (1 − 0.8)
1/(13.32 +1)
= 0.8937
and ω is equal to ω=1-ψ=1-0.8937=0.1063.

20.7 From (20.30) λ is

εf
λ=
Aσ n t f
At the start of tertiary creep t=0 and ω=0 then from (20.23)
ε& s = Aσ n
Substituting ε& s then λ is given by
εf
λ=
ε& s t f
as required.

65
Chapter 21 Solutions
21.1 From (21.1) and (21.3) the maximum volume fractions (r=R) for square and
hexagonal fibre configurations are
πr π
2

V f ,max ( sq ) =   = ≈ 0.785
4  R 4
π r π
2

V f ,max (hex ) =   = ≈ 0.907

2 3  R 2 3
From (21.13) and (21.18) E|| and E⊥ for the square and hexagonal configurations at
Vf,max are found to be
E || (hex ,V f ,max ) 0.907 + 0.093n E ⊥ (hex ,V f ,max ) 0.215 + 0.785n
= , =
E || ( sq ,V f ,max ) 0.785 + 0.215n E ⊥ ( sq ,V f ,max ) 0.093 + 0.907n
where n=Em/Ef. For example, when n=0.1 then E||(hex,Vf,max)/E||(sq,Vf,max)=1.14 and
E⊥(hex,Vf,max)/E⊥(sq,Vf,max)=1.6.

21.2 From the rule of mixtures (21.13)

( )
E|| = E f V f + E m 1 − V f = 76(0.45) + 4(1 − 0.45) = 36.4GPa

21.3 From (21.19)

E f Em 76x4
E⊥ = = = 6.97GPa
( )
E f 1 − V f + Em E f 76(1 − 0.45) + 4( 0.45)

21.4 With E1=E||=36.4GPa from Exercise 21.2 and E2=E⊥=6.97GPa from Exercise
21.3 then E , G and ν from (21.34) are
3 5 3 5
E = E || + E ⊥ = (36.4) + (6.97) = 18GPa
8 8 8 8
1 1 1 1
G = E || + E ⊥ = (36.4) + (6.97) = 6.29GPa
8 4 8 4
E 18
ν= −1= − 1 = 0.43
2G 2(6.29)

21.5 From (21.38) the volume fraction at cross-over between low and high Vf is
σ *m 75
V f' = * = = 013
.
σ f − σ f + σ m 2,000 − 1,500 + 75
' *

If V f > V f' then fibre strength dominates.

21.6 From (21.43) the transverse failure strength, σ *⊥ , is

 Vf   
σ *⊥ = σ *m  1 − 2  = 65 1 − 2 0.3  = 24.83MPa
 π   π 

21.7 From (21.45) the critical embedded fibre length is

66
lce =
σ *f r
=
(
750x10 6 0.2 x10 −3 ) = 17. mm
2τ (
2 45x10 6 )

67
Chapter 22 Solutions

22.1 Resolving Pr we find that P=Prsinθ and Q=Prcosθ and upon substitution into
(22.7) and (22.18) and superimposing the stresses gives the stresses in the half-plane.

22.2 The stresses for a point force at x=b can be found by replacing x by (x-b) in
(22.7)
2 P( x − b ) y
2

σ xx = −
[
π ( x − b) + y 2 ]
2 2

3
2 Py
σ yy = −
[
π ( x − b) + y 2 ]
2 2

2 P( x − b ) y 2
τ xy = −
[
π ( x − b) + y 2 ]
2 2

To show that τxy is zero along the line x=b/2 write expressions for τxy when P is at x=0
and x=b, from (22.7) and the above equation
Pby 2 Pby 2
τ xy ( P at x = 0) = − τ ( = ) =
[ ]
, P at x b
[ ] π (b / 2) + y 2
xy
π (b / 2) + y 2
2 2 2

which cancel if superimposed.

22.3 Referring to Figure 22.26 then the applied pressure distribution is given by
p0 ( a − x )
p( x ) = ; x ≤a
a
Performing the integrations in (22.20) then the stresses are found to be
p   r r 
σ xx = 0 ( x − a )θ 1 + ( x + a )θ 2 − 2 xθ + 2 y ln 1 22  
πa   r 

{( x − a)θ 1 + ( x + a )θ 2 − 2 xθ }
p0
σ yy =
πa
p y
τ xy = − 0 (θ 1 + θ 2 − 2θ )
πa
where r, r1, r2, θ, θ1 and θ2 are given by
r = x 2 + y 2 , r12 = ( x − a ) + y 2 , r22 = ( x + a ) + y 2
2 2

y y y
tan θ = , tan θ 1 = , tan θ 2 =
x x−a x+a

22.4 For a uniform pressure (N=P and T=0) then Φ(z) and Ψ(z) are
p  z − a
[ ]
p a dt p
ln( z − t ) t =− a =
t =a
Φ( z ) = −
2πi ∫− a
=
z − t 2πi
ln 
2πi  z + a 
zp a dt paz
Ψ( z) = − ∫
2πi − a ( t − z ) 2
=−
(
πi z 2 − a 2 )

68
With reference to Figure Sol22.4 then let z-t=re-iθ where r=|z-t|. From (14.148) and
(14.152) the stress components are
2p
σ xx + σ yy = 4 Re Φ( z ) = − (θ 1 − θ 2 )
π
2 pa  z − z  4 pay
σ yy − σ xx + 2iτ xy = 2[ z Φ' ( z ) + Ψ ( z )] =  2 a
=−
πi  z − a  π z2 − a2 ( )
Solving for σxx, σyy and τxy we arrive at

σ xx
p
= − (θ 1 − θ 2 ) +
(
2 pay x 2 − y 2 − a 2 )
π
[( 2
π x 2 + y 2 − a 2 + 4a 2 y 2 ) ]
σ yy
p
= − (θ 1 − θ 2 ) −
(
2 pay x 2 − y 2 − a 2 )
π
[( 2
π x 2 + y 2 − a 2 + 4a 2 y 2 ) ]
4 paxy 2
τ xy =
[(
π x2 + y2 − a2 )
2
+ 4a 2 y 2 ]
y

p
-a O t a x
θ θ
θ
1

2 r
r 2
r 1

z
Figure Sol22.4. A half-plane subject to a uniform normal pressure.

22.5 From (22.55) the pressure beneath the centre of the punch is
P 5x10 3
p(0) = = = 0127
. MPa
(
πa π 12.5x10 −3 )
22.6 From (22.70) the radius of the contact radius, a, and total displacement, δ, are
1/ 3
 3 P R1 R2  1 − ν 12 1 − ν 22  
a=  + 
 4 R1 + R2  E1 E2  
1/ 3
9 P2 R + R  1 − ν 12 1 − ν 22  
2

δ= 1 2
 +  
 16 R1 R2  E1 E2  


69
Letting body 2 be the half-plane, R1=∞, and R=R2 then R1R2/(R1+R2) is found to
reduce to the following

R1 R2 R1 R2
= = R2 = R as R1 → ∞
R1 + R2 R1 (1 + R2 / R1 )
In addition, letting E=E1=E2 and ν=ν1=ν2 for both bodies having the same elastic
properties then the above expressions for a and δ are found to reduce to
1/ 3
9P2  1− ν 2  2 
1/ 3
 3 PR  1 − ν 2  
a=   , δ =   
 2  E   16 R  E  
as required.

22.7 We can obtain expressions for a, δ and p0 for the circular ball in a circular seat
from (22.70) by simply making R1 negative.

70
Chapter 23 Solutions
23.1 Table Sol23.1 summarises the failure strengths of the tested composite material.

failure strength ab. frequency rel. frequency cum. Ab. frequency cum. rel. frequency
650 1 0.1 1 0.1
680 2 0.2 3 0.3
700 2 0.2 5 0.5
710 3 0.3 8 0.8
740 1 0.1 9 0.9
750 1 0.1 10 1
Table Sol23.1. Sample of 10 values of the tensile failure strength (MPa) of glass-
polyester unidirectional laminae composite specimens.

23.2 From (23.4) the mean is

N
1
x=
N
∑x
i =1
i = 703MPa

From (23.5) the variance is

1 N
s2 = ∑ ( xi − x ) 2 = 84555
. MPa
N − 1 i =1
and the standard deviation is the square root of the variance
s = s 2 = 29.08MPa

23.3 Since a thrown die can result in a number which is both even and a multiple of 3
then both events can occur simultaneously and therefore the events are arbitrary.
Letting A represent an even number and B represent a number which is a multiple of 3
then A={2,4,6} and B={3,6}. The required probabilities for arbitrary events are
P(A)=3/6=1/2, P(B)=2/6=1/3 and P(A∩B)=1/6 so that from (23.13) we have
1 1 1 2
P ( A ∪ B) = P( A) + P( B) − P( A ∩ B) = + − =
2 3 6 3

23.4 From (23.28) the mean is

3
b 3  x  x3  9
x = ∫ xf ( x )dx = ∫ x   dx =   =
a 0  2  6 0 2
From (23.30) the variance is
3
 x 4 3x 3 81x 2 
2
 9  x 243
∫ (x − x)
b
σ = f ( x )dx = ∫ − = − + =
2
   
2

x dx  
a 2  2 8 2 16  0 16

23.5 For case i) then x =1 and σ2=1 so that we can obtain the probability directly
from Table 23.2, and is found to 0.9772. For case ii) with x =0.4 and σ2=4 then the
standardised variable is
x − x 2 − 0.4
z= = = 0.8
σ 2
From Table 23.2 we have Φ(0.8)=0.7881.

23.6 From (23.53) and Table 23.3 the mean strength is

71
 σ0  1 650  1
σ= Γ1 +  = Γ  1 +  = 718MPa
V 1/ m
 m  0.2251/10
 10 

23.7 Letting the specimen of Exercise 23.6 be denoted by specimen 1 and the new
specimen by 2 then from (23.65) the expected mean strength of the new specimen is,
with σ 1 =718MPa
σ1 718
σ2 = = = 670MPa
(V2 / V1 ) 1/ m
(0.45 / 0.225)1/10

72