Finite Elements in Engineering Chandrupatla PDF

CHAPTER 1
FUNDAMENTAL CONCEPTS
1.1 We use the first three steps of Eq. 1.11
σx σ σ
εx = −ν y −ν z
E E E
σ σ σ
ε y = −ν x + y − ν z
E E E
σ σ σ
ε z = −ν x − ν y + z
E E E
Adding the above, we get
1 − 2ν
εx + εy + εz = (σ x + σ y + σ z )
E
σx
Adding and subtracting ν from the first equation,
E
1+ ν ν
σ x − (σ x + σ y + σ z )
εx =
E E
Similar expressions can be obtained for εy, and ε z.
From the relationship for γyz and Eq. 1.12,
E
τ yz =
γ yz etc.
2(1 + ν )
Above relations can be written in the form
σ = Dε
where D is the material property matrix defined in Eq. 1.15. 
1.2 Note that u2(x) satisfies the zero slope boundary condition at the support.
Introduction to Finite Elements in Engineering, Fourth Edition, by T. R. Chandrupatla and A. D. Belegundu. ISBN 01-3-216274-1.
© 2012 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved. This publication is protected by Copyright
and written permission should be obtained from the publisher prior to any prohibited reproduction, storage in a retrieval
system, or transmission in any form or by any means, electronic, mechanical, photocopying, recording, or likewise. For information regarding permission(s),
write to: Rights and Permissions Department, Pearson Education, Inc., Upper Saddle River, NJ 07458.
1.3 Plane strain condition implies that
σ σy σz
ε z = 0 = −ν x − ν +
E E E
which gives
σ z = ν(σ x + σ y )
We have, σ x = 20000 psi σ y = −10000 psi E = 30 × 10 6 psi ν = 0.3 .
On substituting the values,
σ z = 3000 psi 
1.4 Displacement field

(
u = 10 −4 − x 2 + 2 y 2 + 6 xy )
v = 10 −4
(3x + 6 y − y )
2
∂u ∂u
= 10 − 4 (− 2 x + 6 y ) = 10 − 4 (4 y + 6 x )
∂x ∂y
∂v ∂v
= 3 × 10 − 4 = 10 − 4 (6 + 2 y )
∂x ∂y
 ∂u 
 
 ∂x 
 ∂v 
ε= 
 ∂y 
 ∂u + ∂v 
 ∂y ∂x 
 
at x = 1, y = 0
− 2 
 
−4
ε = 10  6 
9 
 

1.5 On inspection, we note that the displacements u and v are given by
u = 0.1 y + 4
v=0
It is then easy to see that
∂u
εx = =0
∂x
∂v
εy = =0
∂y
∂u ∂v
γ xy = + = 0.1
∂y ∂x

1.6 The displacement field is given as
u = 1 + 3x + 4x3 + 6xy2
v = xy − 7x2
(a) The strains are then given by

∂u
εx = = 3 + 12 x 2 + 6 y 2
∂x
∂v
εy = =x
∂y
∂u ∂v
γ xy = + = 12 xy + y − 14 x
∂y ∂x
(b) In order to draw the contours of the strain field using MATLAB, we need to create a
script file, which may be edited as a text file and save with “.m” extension. The file
for plotting εx is given below
file “prob1p5b.m”
[X,Y] = meshgrid(-1:.1:1,-1:.1:1);
Z = 3.+12.*X.^2+6.*Y.^2;
[C,h] = contour(X,Y,Z);
clabel(C,h);
On running the program, the contour map is shown as follows:
1
8
14
18
14
18
0.8
6
0.6
12
16
10
12
8
16
10
6
8
0.4
4
0.2
14
14
0
12
10
-0.2
12
8
10
-0.4 4
6
-0.6
16
14
14
16
6
8
-0.8
10 8
12 12
18
10
18
-1
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Contours of εx
Contours of εy and γxy are obtained by changing Z in the script file. The numbers on
the contours show the function values.
(c) The maximum value of εx is at any of the corners of the square region. The
maximum value is 21.

1.7
(x, y) (u, v)
0.2
a) =
u y ⇒ =
u 0.2 y =
v 0
1
∂u ∂v ∂u ∂v
b) εx = = 0 εy = = 0 γ xy = + = 0.2
∂x ∂y ∂y ∂x
1.8
σ x = 40 MPa σ y = 20 MPa σ z = 30 MPa
τ yz = −30 MPa τ xz = 15 MPa τ xy = 10 MPa
T
1 1 1 
n= 
2 2 2
From Eq. 1.8 we get
Tx = σ x n x + τ xy n y + τ xz n z
. = 35.607 MPa
T y = τ xy n x + σ y n y + τ yz n z
= −6.213 MPa
Tz = τ xz n x + τ yz n y + σ z n z
= 13.713 MPa
σ n = Tx n x + T y n y + Tz n z
= 24.393 MPa

1.9 From the derivation made in P1.1, we have
σx =
E
(1 + ν )(1 − 2ν )
[
(1 − ν )ε x + νε y + νε z ]
which can be written in the form
σx =
E
[(1 − 2ν )ε x + νε v ]
(1 + ν )(1 − 2ν )
and
E
τ yz = γ yz
2(1 + ν )
Lame’s constants λ and µ are defined in the expressions
σ x = λε v + 2µε x
τ yz = µγ yz
On inspection,
Eν
λ=
(1 + ν )(1 − 2ν )
E
µ=
2(1 + ν )
µ is same as the shear modulus G. 
1.10
ε = 1.2 × 10 −5
∆T = 30 0 C
E = 200 GPa
α = 12 × 10 -6 / 0 C
ε 0 = α∆T = 3.6 × 10 − 4
σ = E (ε − ε 0 ) = −69.6 MPa

1.11
du
εx = = 1 + 2x 2
dx
L du  2  L
δ=∫ dx =  x + x 3 
0 dx
 3  0
 2 
= L1 + L2 
 3 

1.12 Following the steps of Example 1.1, we have
(80 + 40 + 50 ) − 80  q1  60

  = 
 − 80 80  q 2  50
Above matrix form is same as the set of equations:
170 q1 − 80 q2 = 60
− 80 q1 + 80 q2 = 50
Solving for q1 and q2, we get
q1 = 1.222 mm
q2 = 1.847 mm

1.13
When the wall is smooth, σ x = 0 . ∆T is the temperature rise.

a) When the block is thin in the z direction, it corresponds to plane stress condition. The
rigid walls in the y direction require ε y = 0 . The generalized Hooke’s law yields the
equations
σy
ε x =−ν + α∆T
E
σy
ε y=
+ α∆T
E
− Eα∆T . ε x is then calculated
From the second equation, setting ε y = 0 , we get σ y =
using the first equation as (1 −ν ) α∆T .
b) When the block is very thick in the z direction, plain strain condition prevails. Now we
have ε z = 0 , in addition to ε y = 0 . σ z is not zero.
σy σz
ε x =−ν −ν + α∆T
E E
σy σz
ε y= −ν + α∆T= 0
E E
σy σz
ε z =−ν + α∆T =0
+
E E
From the last two equations, we get
− Eα∆T 1 + 2ν
σy = σz = − Eα∆T
1 +ν 1 +ν
ε x is now obtained from the first equation. 
1.14 For thin block, it is plane stress condition. Treating the nominal size as 1, we may set the
0.1
initial strain ε 0 = α∆T = in part (a) of problem 1.13. Thus σ y = −0.1E . 
1
1.15
The potential energy Π is given by

x=0
2
 du 
2 2
1
Π= ∫ EA  dx − ∫ ugAdx
2 0  dx  0
Consider the polynomial from Example 1.2, g=1

E=1
(
u = a3 − 2 x + x 2 ) A=1
= (− 2 + 2 x )a3 = 2(− 1 + x )a3

du
x=2
dx
On substituting the above expressions and integrating, the first term of becomes
2 2
2a 3  
 3
and the second term
2
2 2
 2 x3 
∫0 ugAdx = ∫0 udx = a 
3− x +


3  0
4
= − a3
3
Thus
Π=
4 2
3
(
a3 + a3 )
∂Π 1
=0⇒ a3 = −
∂a3 2
this gives u x =1 = −
1
(− 2 + 1) = 0.5 
2
1.16
E=1
A=1
x=0 f = x3 x=1
We use the displacement field defined by u = a0 + a1x + a2x2.
u = 0 at x = 0 ⇒ a0 = 0
u = 0 at x = 1 ⇒ a1 + a2 = 0 ⇒ a2 = − a1
We then have u = a1x(1 − x), and du/dx = a1(1 − x).
The potential energy is now written as
2
1  du 
1 1
2 ∫0  dx 
Π=   dx − ∫ fudx
0
1 1
= ∫ a1 (1 − 2 x ) dx − ∫ x 3 a1 x(1 − x )dx
1 2 2
20 0
( ) ( )
1 1
1
= ∫ a1 1 − 4 x + 4 x 2 dx − ∫ a1 x 4 − x 5 dx
2
20 0
1 2 4 4 1 1
= a1 1 − +  − a1  − 
2  2 3 5 6
2
a a
= 1 − 1
6 30
∂Π a1 1
=0 ⇒ − =0
∂a1 3 30
This yields, a1 = 0.1
Displacemen u = 0.1x(1 − x)
Stress σ =E du/dx = 0.1(1 − x) 
1.17 Let u1 be the displacement at x = 200 mm. Piecewise linear displacement that is
continuous in the interval 0 ≤ x ≤ 500 is represented as shown in the figure.
u = a1 + a2x u = a3 + a4x
u1
0 200 500
0 ≤ x ≤ 200
u = 0 at x = 0 ⇒ a1 = 0
u = u1 at x = 200 ⇒ a2 = u1/200
⇒ u = (u1/200)x du/dx = u1/200
200 ≤ x ≤ 500
u = 0 at x = 500 ⇒ a3 + 500 a4 = 0
u = u1 at x = 200 ⇒ a3 + 200 a4 = u1
⇒ a4 = −u1/300 a3 = (5/3)u1
⇒ u = (5/3)u1 − (u1/300)x du/dx = − u1/200
2 2
 du   du 
200 500
1 1
Π=
2 ∫
0
E al A1   dx + ∫ E st A2   dx − 10000u1
 dx  2 200  dx 
2 2
1  u  1  u 
Π= E al A1  1  200 + E st A2  − 1  300 − 10000u1
2  200  2  300 
1  E al A1 E st A2  2
=  + u1 − 10000u1
2  200 300 
∂Π E A E A 
= 0 ⇒  al 1 + st 2 u1 − 10000 = 0
∂u1  200 300 
Note that using the units MPa (N/mm2) for modulus of elasticity and mm2 for area and
mm for length will result in displacement in mm, and stress in MPa.
Thus, Eal = 70000 MPa, Est = 200000, and A1 = 900 mm2, A2 = 1200 mm2. On
substituting these values into the above equation, we get
u1 = 0.009 mm
This is precisely the solution obtained from strength of materials approach 
1.18
In the Galerkin method, we start from the equilibrium equation
d du
EA +g =0
dx dx
Following the steps of Example 1.3, we get
du dφ
2 2
∫ − EA
0
dx dx
dx + ∫ gφdx
0
Introducing
( )
u = 2 x − x 2 u1 , and
φ = (2 x − x )φ 2
1
where u1 and φ1 are the values of u and φ at x = 1 respectively,
 
( )
2 2
φ1 − u1 ∫ (1 − 2 x ) dx + ∫ 2 x − x 2 dx  = 0
2
 0 0 
On integrating, we get
 8 4
φ1  − u1 + =0
 3 3
This is to be satisfied for every φ1, which gives the solution
u1 = 0.5 
1.19 We use
u = a1 + a 2 x + a3 x 2 + a 4 x 3
u = 0 at x = 0
u = 0 at x = 2
This implies that
0 = a1
0 = a1 + 2a 2 + 4 3 + 8a 4
and
( ( )
u = a3 x 2 − 2 x + a 4 x 3 − 4 x )
= 2a (x − 1) + a (3 x − 4)
du 2
3 4
dx
a3 and a4 are considered as independent variables in
[ ( )]
2
Π = ∫ 2a3 ( x − 1) + a 4 3 x 2 − 4 dx − 2(− a3 − 3a 4 )
1 2
20
on expanding and integrating the terms, we get
Π = 1.333a3 + 12.8a 4 + 8a3 a 4 + 2a3 + 6a 4

2 2
We differentiate with respect to the variables and equate to zero.
∂Π
= 2.667 a3 + 8a 4 + 2 = 0
∂a3
∂Π
= 8a3 + 25.6a 4 + 6 = 0
∂a 4
On solving, we get
a3 = −0.74856 and a4 = −0.00045.
On substituting in the expression for u, at x = 1,
u1= 0.749
This approximation is close to the value obtained in the example problem. 
1.20
L L
Π = ∫ σ T εAdx − ∫ T ( x )udx
1
(a)
20 0
du
σ = Eε and ε =
dx
On substitution,
2
 du 
60 30 60
1
Π = ∫ EA  dx − ∫ T udx − ∫ T udx
2 0  dx  0 30
2
( )∫  du 
60 30 60
1
Π= 60 × 10 6  dx − ∫ 10 xudx − ∫ 300udx
2 0 
dx  0 30
(b)
Since u = 0 at x = 0 and x = 60, and u = a0 + a1x + a2x2, we have
u = a 2 x( x − 60 )
= a 2 (2 x − 60 )
du
dx
On substituting and integrating,
Π = 216 × 1010 a 2 + 8775000a 2

2
Setting dΠ/da2 = 0 gives
a 2 = −2.03125 × 10 −6
= −60.935(2 x − 60 )
du
σ=E
dx
Plots of displacement and stress are given below:
-3
x 10
2
1.8
1.6
1.4
1.2
0.8
0.6
0.4
0.2
0
0 10 20 30 40 50 60
Displacement u
4000
3000
2000
1000
-1000
-2000
-3000
-4000
0 10 20 30 40 50 60
Stress
. 
1.21 y = 20 at x = 60 implies that
20 = a 0 + 60a1 + 3600a 2 , which yields

a 0 = 20(1 − 3a1 − 180a 2 )
Substituting for k, h, L, and a0 in I, we get
60
I = ∫ 10(a1 + 2a 2 x ) dx +
2 1
(25)[20(1 − 3a1 − 180a 2 ) − 800]2
0
2
( )
60
I = ∫ 10 a1 + 4 xa1 a 2 + 4 x 2 a 2 dx + 5000(3a1 + 18a 2 + 39 )
2 2 2
I = 45600a1 + 612000a1 a 2 + 45 × 10 5 a 2 + 117 × 10 4 a1 + 702 × 10 4 a 2 + 7605000

2 2
dI
= 912000a1 + 612000a 2 + 117 × 10 4 = 0
da1
dI
= 612000a1 + 90 × 10 5 a 2 + 702 × 10 4 = 0
da 2
On solving,
a2 = 0.1699
a1 = −13.969
Substituting into the expression for a0, we get
a0 = 246.538
. 
1.22 Since u = 0 at x = 0, the displacement satisfying the boundary condition is u = a1x. Also
the coordinates are x2 = 1, and x3 = 3.
The potential energy for the problem is
2
1 3  du 
2 ∫0
=π EA   dx − P2u2 − P3u3
 dx 
du
We have u2 = a1, u3 = 3a1, E = 1, A = 1, and = a1 . Thus
dx
1 3 3
π= ∫ ( a1 ) dx − a1 − 3a1= a12 − 4a1 .
2
2 0 2
dπ
For stationary value, setting = 0 , we get
da1
3a1 – 4 = 0, which gives a1 = 0.75.
The approximate solution is u = 0.75x. 
1.23 Use Galerkin approach with approximation u =a + bx + cx 2 to solve
du
+ 3= u x 0 ≤ x ≤1
dx
u ( 0) = 1
The week form is obtained by multiplying by φ satisfying φ ( 0 ) = 0 .
1  du 
∫ φ  dx + 3u − x  dx =
0
0
We now set u =1 + bx + cx 2 satisfying u ( 0 ) = 1 and =

φ a1 x + a2 x 2 . On introducing these
into the above integral,
∫ ( a x + a x )( b + 2cx + 3 + 3bx + 3cx − x ) dx =0

1
2 2
0 1 2
a ∫ ( bx + 3 x − x + 3bx + 2cx + 3cx ) dx + a ∫ ( bx + 3x 2 − x 3 + 3bx 3 + 2cx 3 + 3cx 4 ) dx + =0

1 1
2 2 2 3 2
1 0 2 0
On integrating, we get
b 3 1 2c 3c  b 1 3b c 3c 
a1  + − + b + +  + a2  + 1 − + + +  = 0
2 2 3 3 4 3 4 4 2 5
3 17 7  13 11 3
a1  b + c +  + a2  b + c +  = 0
2 12 6  12 10 4
This must be satisfied for every a1 and a2. Thus the equations to be solved are
3 17 7
b+ c+ = 0
2 12 6
13 11 3
b+ c+ = 0
12 10 4
The solution is b = –1.9157, c = 1.2048. Thus u = 1 − 1.9157 x + 1.2048 x 2 . 
1.24
3
Pa Pa 2
The deflection and slope at a due to P1 are − 1
and − 1 . Using this the deflection
3EI 2 EI
and slope at L due to load P1 are
Pa 3 Pa ( L − a )
2
v1 = − 1 − 1
3EI 2 EI
2
Pa
v1′ = − 1
2 EI
The deflection and slope due to load P2 are

P2 L3
v2 = −
3EI
P L2
v2′ = − 2
2 EI
We then get
v= v1 + v2

v=′ v1′ + v2′
1.25
(0,1) (1,1)
(0,0) (1,0)
(a) The displacement of B is given by (–0.1, 0.1) and A, C, and D remain in their original
position. Consider a displacement field of the type
u =a1 + a2 x + a3 y + a4 xy
v =b1 + b2 x + b3 y + b4 xy
The four constants can be evaluated using the known displacements
a1 = 0
At A (0, 0)
b1 = 0
a1 + a2 =
−0.1
At B (1, 0)
b1 + b2 =0.1
a1 + a2 + a3 + a4 =
0
At C (1, 1)
b1 + b2 + b3 + b4 =
0
a1 + a3 = 0
At D (0, 1)
b1 + b3 =
0
The solution is
a1 = 0, a2 = –0.1, a3 = 0, a4 = 0.1
b1 = 0, b2 = 0.1, b3 = 0, b4
u= −0.1x + 0.1xy
This gives
= v 0.1x − 0.1xy
(b) The shear strain at B is
∂u ∂v
γ= + = 0.1x + 0.1 − 0.1 y
∂y ∂x 
γ=
B 0.1(1) + 0.1 − 0.1( 0=
) 0.2

Finite Elements in Engineering Chandrupatla PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Finite Elements in Engineering Chandrupatla PDF

Uploaded by

Copyright:

Available Formats

CHAPTER 1

1.1 We use the first three steps of Eq. 1.11

From the relationship for γyz and Eq. 1.12,

1.4 Displacement field

1.5 On inspection, we note that the displacements u and v are given by

It is then easy to see that

1.6 The displacement field is given as

(a) The strains are then given by

On running the program, the contour map is shown as follows:

1.12 Following the steps of Example 1.1, we have

(80 + 40 + 50 ) − 80  q1  60

Above matrix form is same as the set of equations:

Solving for q1 and q2, we get

When the wall is smooth, σ x = 0 . ∆T is the temperature rise.

ε x is now obtained from the first equation. 

The potential energy Π is given by

Consider the polynomial from Example 1.2, g=1

= (− 2 + 2 x )a3 = 2(− 1 + x )a3

We then have u = a1x(1 − x), and du/dx = a1(1 − x).

The potential energy is now written as

This yields, a1 = 0.1

Stress σ =E du/dx = 0.1(1 − x) 

This is precisely the solution obtained from strength of materials approach 

where u1 and φ1 are the values of u and φ at x = 1 respectively,

This is to be satisfied for every φ1, which gives the solution

Π = 1.333a3 + 12.8a 4 + 8a3 a 4 + 2a3 + 6a 4

We differentiate with respect to the variables and equate to zero.

On substituting in the expression for u, at x = 1,

This approximation is close to the value obtained in the example problem. 

On substituting and integrating,

Π = 216 × 1010 a 2 + 8775000a 2

Setting dΠ/da2 = 0 gives

1.21 y = 20 at x = 60 implies that

20 = a 0 + 60a1 + 3600a 2 , which yields

Substituting for k, h, L, and a0 in I, we get

I = 45600a1 + 612000a1 a 2 + 45 × 10 5 a 2 + 117 × 10 4 a1 + 702 × 10 4 a 2 + 7605000

Substituting into the expression for a0, we get

The potential energy for the problem is

The approximate solution is u = 0.75x. 

We now set u =1 + bx + cx 2 satisfying u ( 0 ) = 1 and =

∫ ( a x + a x )( b + 2cx + 3 + 3bx + 3cx − x ) dx =0

a ∫ ( bx + 3 x − x + 3bx + 2cx + 3cx ) dx + a ∫ ( bx + 3x 2 − x 3 + 3bx 3 + 2cx 3 + 3cx 4 ) dx + =0

The deflection and slope due to load P2 are

You might also like