LECTURE 2: Deformation analysis, Thermal stresses, Transient Axial

loading, Resonance

1. Deformation analysis

2. Axial loading

3. Thermal strain and thermal stress

4. Transient loading, Resonance

1
1. DEFORMATION AND STRAINS

Deformation in x direction ：Solid element: original shape; Dashed element: deformed shaped

u u+du Displacement along the x direction at position x of a

body is denoted as u
Displacement along the x direction at position
x+dx of the body is u+du
x+dx The elongation of the element in the x direction is
(u+du)u=du.
y The normal strain along the x direction, denoted as εx , is

dx (dx  u  du )  (dx  u ) (u  du )  u du
x εx   
dx dx dx
O x
Note: u is the function of x. This is, u=u(x)

d: differential operator

In Fundamentals of Mechanics, only rigid bodies are considered. Under the action of the external forces, the shape of
the body remains unchanged.

In Applied Mechanics, we study the elastic (even plastic) solids. The solids will deform when there are external forces
applied to them. The deformations of the solids can be described by the displacements and strains of the solids. 2
 Deformation in y direction ：Solid element: original shape; Dashed element: deformed shaped

Displacement along the y direction at position

y of a body is denoted as v
v+dv
y+dy
Displacement along the y direction at position
y+dy of the body is v+dv

dy The elongation of the element in the y direction is

y (v+dv)v=dv.
v

y
The normal strain along the y direction, denoted as εy , is

(dy  v  dv)  (dy  v) dv

O x εy  
dy dy

Note: v is the function of y. This is, v=v(y)

3
The displacements of the solid, u and v, can also vary with x and y . In this situation, shear strain develops

Solid element: original shape; Dashed element: deformed shaped

(u  du )  u du
u+du  
dy dy

(v  dv)  v dv
 
dx dx

dy The shear strain in x-y plane,
u
y
 v+dv denoted as  xy , is
v
du dv
dx  xy      
dy dx
O x

4
 Strains in three dimensional solids

If the components of the displacement vector along the x, y and z directions at point (x, y, z) of a body are
u(x, y, z), v(x, y, z) and w(x, y, z), respectively, the normal strains are:
u v w
εx  , ε y  , εz 
x y z

and the shear strains are:

u v v w u w
 xy   ,  yz   ,  xz  
y x z y z x

z,w
B F

Three are six strain components:

A E 3 normal strains and 3 shear strains.

y,v
C
G

x,u
D H
5
2. AXIAL LOAD
APPLICATIONS
Structural components subjected only to tension or compression are known as axially loaded
members. Solid bars with straight longitudinal axes are the most common type, although cables and
coil springs also carry axial loads. Examples of axially loaded bars are truss members, connecting
rods in engines, spokes in bicycle wheels, columns in buildings.

Application

Most concrete columns are reinforced with steel rods; and these two materials work together in
supporting the applied load. Are both subjected to axial stress?

6
Three fundamental equations:
 u is the axial displacement of the rod. It is the
P( x) du ( x) function of x. This is, u=u(x).
   Eε ε
A( x) dx
 Provided these quantities do not exceed the
dx proportional limit, we can relate them using
x
P1 P2 Hooke’s Law, i.e.  = E.

P( x)  du  du P( x)
L  E ( x)   
A( x)  dx  dx E ( x) A( x)

Can be used to determine the axial displacement

of the rod as a function of x.
du
dx
 Total elongation of the rod  is
d: differential operator  = u(L)  u(0)

The product EA is known as the cross- For constant axial force and PL
sectional axial rigidity of the bar. 
uniform cross section EA

7
Example 1 (Composite Materials): A system of Composite Rods under tension. The two rods are
jointed securely at their interfaces. Cross-sectional properties are: E1, A1 and E2, A2. Determine
(a) the axial strains and stresses in each rod;
(b) the equivalent axial Young’s modulus of the composite system.

E1 A1 1 E1 A1 1
N N
E2 A2 2 E2 A2 2

Solution:

(1) Equilibrium equation: 1A1 + 2A2 = N

(2) Compatibility equation: elongations and the axial strains of the two materials are equal, 1 = 2 = 0

1 
(3) Material property (constitutive) : 1  , 2  2
E1 E2

8
(4) From (1) - (3)，

N
(a) 1   2   0 
E1 A1  E2 A2

E1 N E2 N
The stresses in rods: 1  , 2 
E1 A1  E2 A2 E1 A1  E2 A2

E1 A1 1 E1 A1 1
N N
E2 A2 2 E2 A2 2

You can see:

The axial strains of the two rods (materials) are identical;
However, the stresses in the two rods (materials) are different.

9
(b) Equivalent Young’s Modulus of the composite material
E1 A1
* Average axial stress in the composite: N N
E2 A2
N
0 
A1  A2

* Equivalent axial Young’s modulus of the composite: E1 A1

0 0
E2 A2
 E A  E2 A2
E 0 E 1 1  E1 f1  E2 f 2
0 A1  A2

A1, 2 A1, 2 L V1, 2

where f1, 2   
A1  A2 A1L  A2 L V1  V2

V : volume ;
f : volume fraction. The subscripts 1 and 2 denote, respectively,
material #1 and material #2

10
 If you have a composite of many
single materials. The general form:
* Average axial strain in the composite:

N
0  M

E A
i 1
i i
1
2
* Axial stresses in the rods:

Ei N N i N
i  M

E A
i 1
i i
M

* Equivalent axial Young’s modulus of the composite: Fig. a. Composite

materials made of M
M different materials
General form: E   Ei f i Rule-of-Mixture
i 1

Vi
fi 
V1  V2  ...VM

V : volume ;
f : volume fraction. The subscript i denoted the material #i
11
3. Thermal strain and Thermal stress

3.1 Thermal Expansion and Contraction

A change in temperature can cause a body to change its

dimensions. Generally, if the temperature increases, the body
will expand, whereas if the temperature decreases, it will
contract. Usually this expansion or contraction is linearly
related to temperature change as:

T =  L ( T )

where L is the original length, T is the change in temperature , 

is the constant of proportionality, called the coefficient of
thermal expansion. It is a property of the material. In the unit of
( C )1.

Typical values of  are measured in 106 ( oC )1.

For structural steel,  = 12  106 ( oC )1

12
3.2 Thermal Strain

What is thermal strain?

T
T   T
L

 T , L 
For a bar under axial stress  in an changing
temperature environment, what is its total
length change? What is its total strain?

  T    TL  L
E
 
  T    T   
L E

=E    E  T 

Extremely important
13
 3.3 Thermal Stress

Example 2: A bar AB of length L is held between immovable supports. If the

A temperature of the bar is raised uniformly by an amount T, what thermal stress
 is developed in the bar? (Assume that the bar is made of linearly elastic
material.)
Solution:
L T
Strain due to temperature change: T =  ( T );
Strain due to the stress in the bar:  =  / E; =E
Resultant strain  = T +  =  ( T ) +  / E;
Compatibility condition:  = 0.
B
Thus, the thermal stress is: ( T ) +  / E = 0   =  E  ( T ).

Consequence of excessive thermal

stresses

14
Table 1 Density ρ, modulus of elasticity E, shear modulus G, Poisson’s ratio , and thermal
expansion coefficient α for some common materials

z
Three dimensional thermal strains in isotropic materials
 xT  T ,  yT  T ,  zT  T
 x   1      x  1
  1   
 y     1    y   T 1
  E      1    1
y  z   z  
T
 yz  xz  xy
 yz  ,  xz  ,  xy 
x G G G
Temperature change does not affect the shear strains (stresses) for
isotropic materials 15
 4. Axial transient loading in a straight bar

Below is a bar. Its left end is fixed. The right end is subjected to a transient
(dynamic) load P(t). As a result, the displacement, the strain and the stress
in the bar will vary with location x as well as time t.

In order to establish the equation governing dx

the displacement of the bar, we cut a free
  
body of length dx from the bar. We then A   dx  A
 x 
investigate the equilibrium of the free body.

   
 F    x dx  A  A  x Adx

  2u
 2
x t

16
Periodic applied load

Therefore, we have obtained the governing equation of the bar:

  2u u
 2 ,   E  E Applied force
x t x
Putting the second into the first. One gets the governing
equation of the bar in terms of the displacement:

 2u  2u
E 2  2 (1)
x t
Or

 2u  2u E
c2
 , c (2) c: is the so-called sound speed in the material
x 2 t 2 

The partial differential equation (2) is called a one-dimensional

wave equation, and the parameter c is called wave speed.

For steel with  = 0.3, E = 210 GPa, and ρ = 7.83  103 kg/m3,
we find c = 5179 m/s.
17
 Boundary conditions are:
u  0 at x  0
u P sin t
E  at x  L
x A
In this situation, we see the displacement everywhere A is the cross-sectional area of the bar
in bar also varies with time according to sin t. We
also see that the displacement is proportional to the To use the other (the stress) boundary condition at
load P. Thus, we assume the displacement is x=L, the stress associated with Eq. (3) is written as
u( x, t )  PU sin t (3) u ( x, t )  x
( x, t )  E  EPB cos sin t (7)
where U is an unknown function of x. It does not vary x c c
with time. Substituting Eq. (3) into Eq. (2) gives:
Compare Eq. (7) with the boundary condition at x=L,
2
dU we see that
c2 2
U2  0 (4)
dx
 L P sin t
This second order differential equation has the EPB cos sin t  (8)
following general solution: c c A
x x or
U  B sin  D cos (5)
c c c 1
B
Where B and D are constants to be determined from  A cos L
the boundary conditions of the problem. The
complete solution of the displacement is:
c
Finally, the exact solution of the displacement of this
 x x 
u ( x, t )  P B sin  D cos  sin t (6)
problem is
 c c  x
sin
P c c sin t
u ( x, t )  (10)
Since u=0 when x=0, we see D=0. EA  cos L
c 18
 Boundary conditions are:
u  0 at x  0
u
E  P sin t at x  L
x
Once the solution of the displacement has been established,
the strain and stress can be found:

x
cos
P c sin t
 ( x, t )  (11)
EA cos L
c

x
cos
P c sin t
( x , t )  (12)
A cos L
c

19
 Resonance

If you observe Eqs. (10)-(12), you can see that large (infinite)
displacement, strain and stresses develop in the bar if:

L L  1
cos 0   n   , n  integer : 1, 2, 3, ...
c c  2

This means that elastic field (displacement, strain and stress) are infinite and structural failure
occurs if the frequency of the applied load is such that :

 1 c
ω  ωn   n  π 
2n  1π E
(4)
 2 L 2L ρ

n: Natural frequencies of vibration of the fixed-free bar.

Resonance of a structure occurs when the frequency of applied force matches a natural
frequency of the structure, i.e.,   n

In most situations, resonance must be avoided. However, there are situations that resonance can be used.
For example: shattering of glass at certain high pitch; ultrasonic destruction of kidney stone; etc.
20
 Natural frequencies

ωn
fn 
2

n : the angular frequencies of the oscillation, measured in radians/second.

f n : the cyclic frequencies, number of oscillation per second

For the oscillation in the left figure:

f 1 = 0.5;
1
3 f 2 = 1.0;
2
f 3 = 2.0

(s)

21
 Example 3: A bar of length L = 2 m along its axial direction that is fixed at its left end. Determine the first 3
natural frequencies of the bar for the longitudinal vibration. Assume the bar has Young’s Modulus E = 68.95 GPa
and density  = 2700 kg /m 3.

In this example, we studied the natural frequencies

of a fixed-free bar. For other types of bar (e.g., fixed-
2m fixed), its natural frequencies are different. You will
see this in the Practical Class.

E
Solution: c  5053.42 m/s


ω  ωn 
2n  1π c
2 L

ω1 
2 1  1π c
 3968.95 rad/s f1 
1
 631.68 Hz
2 L 2

2
ω2 
2  2  1π c f2   1895.03 Hz
2
2 L

ω3 
2  3  1π c f3 
3
 3158.39 Hz
2 L 2

• 1 Hz means the system would oscillate one time per second.

22
Example 4: Finite element analysis of Example 3 by using only
one element. A is the cross-sectional area of the bar.

Solution: (the Ansys Finite Element Software gets the

frequency by solving the below equation – derivation of this
equation is beyond the scope of this subject )
E, 
AL 2 1 u1  EA  1  1 u1  0
L     
6 1 2 u2  L  1 1  u2  0

Since Node #1 is fixed, u1=0. Therefore

AL EA d 2u
u2  u2  0 u  2
3 L dt
1 2

Assuming u2  sin t

AL 2 EA
   0
3 L

1 E
 3
L 

1 E
f  3  696.52 Hz
2L  23
Example 5: Find the natural frequencies of the bar if its Boundary conditions are:
both ends are fixed.
u  0 at x  0
u  0 at x  L

L
It can be seen from Eq. (B) that:

Solution:
n n E
We need to find a solution of Eq. (2) – slide #17 :  c
L L 
 2u  2u E
c2
 , c (2)
x 2 t 2  Thus
n E
Since both ends of the bar are fixed, we can assume n  , n  1,2,3....
a solution as L 

nx
u ( x, t )  B sin sin t , n  1,2,3... (A) are natural frequencies of the fixed-fixed bar.
L

You can see the boundary conditions are identically

satisfied. The coefficient B is to determined from
the external loading conditions.
Putting (A) into Eq. (2) yields:

 n  nx nx
2

 c   B sin
2
sin t  2 B sin sin t (B)
 L L L
24
LECTURE 3
PLANE STRESS ASSUMPTION
3.1 Transformation of stresses
y y
3.2 Transformation of strains
3.3 Orthotropic materials
x z

3.1 STRESS TRANSFORMATION

in two dimensional solids
 zx  0 ,  zy  0 ,  zz  0
 Navigate between rectilinear co-
ordinate systems for stress  Both sides of the plate are free from external loads
components
 Determine principal stresses and
maximum in-plane shear stress Plane-stress : Only stress components are x , y and xy
(The state of plane stress at a point is uniquely represented
by three components acting on an element that has a
specific orientation at the point)

1
• Sign Convention:
– Positive normal stress acts outward from all faces，
– Positive shear stress acts upwards on the right-hand face of the element,
– Both the x-y and x-y system follow the right-hand rule,
– The orientation of an inclined plane (on which the normal and shear stress components are to
be determined) will be defined using the angle θ. The angle θ is measured from the positive x
to the positive x-axis. It is positive if it follows the curl of the right-hand fingers.

2
 Normal and shear stress components: consider the equilibrium of the free-body diagram

+ΣFx = 0; σx∆A – (τxy ∆A sinθ) cosθ – (σy ∆A sinθ) sinθ

– ( τxy ∆A cosθ) sinθ – (σx ∆A cosθ) cosθ = 0
 σx = σx cos2 θ + σy sin2θ + τxy (2 sinθcosθ)

+ΣFy = 0; τxy ∆A + (τxy ∆A sinθ) sinθ – (σy ∆A sinθ) cosθ

– ( τxy ∆A cosθ) cosθ + (σx ∆A cosθ) sinθ = 0
 τxy  = (σy – σx) sinθ cosθ + τxy (cos2θ – sin2θ)
3
 Finally, stress transformation in 2D：
  x  c 2  x  s 2  y  2sc xy

  y  s  x  c  y  2sc xy c  cos  , s  sin 
2 2
(1)
   sc  sc  (c 2  s 2 )
 xy x y xy

  x   x   c2 s2 2 sc 
     
OR 
 y   [T ( )] y  (2) where [T ()]   s 2 c 2  2sc 
     sc sc c 2  s 2 
 xy   xy   

The system of three linear algebraic equations (1) can be solved to get the stresses in the
(x-y) coordinate system in terms of the stresses in the (x-y) system.

  x  c 2  x  s 2  y  2 sc xy In fact, (x-y) inclines anti-clockwise  to (x-y)

 is equivalent to (x-y) inclines clockwise  to
   s 2
 x  c 2
 y  2sc xy (3)
(x-y) . Thus [t ()]  [T ()]
y
  sc  sc  (c 2  s 2 )
 xy x y xy 

 x    x  c 2 s 2  2sc 
     
OR 
 y  [t  ( )]  y  (4) where [t ()]   s 2 c 2 2 sc 
     sc  sc c 2  s 2 
 
xy  xy   
4
Example 1: The state of plane stress at a point on the surface of the airplane fuselage is
represented on the element oriented as shown in the Fig. a. Represent the state of stress at the
point on another element that is oriented 60°anti-clockwise from the position shown.

Solutions:

x=  80 MPa, y= 50 MPa,

xy=  25 MPa,

 = 60 (Fig. c)

 x   cos 2 60 sin 2 60 2 sin 60 cos 60   80   4.15

      
 y    sin 60  2 sin 60 cos 60   50    25.8 MPa
2
cos 2 60
   sin 60 cos 60 sin 60 cos 60 cos 2 60  sin 2 60  25  68.8 
 xy      

The results are shown on the

element as shown in Fig. d.

5
3.2 STRAIN TRANSFORMATION in two dimensional solids

 Navigate between rectilinear co-ordinate system for strain

components
 Define stress-strain relationship

z
 In 3D, the general state of strain at a point is
represented by a combination of 3
components of normal strain x , y , z , and 3
xdx ydy
components of shear strain xy , yz , xz.

 In plane-strain cases, z, xz and yz are zero.

zdz
 The state of plane strain at a point is uniquely
represented by 3 components (x , y and xy)
acting on an element that has a specific x x y y
orientation at the point.

Plane stresses, x , y , do not produce plane strains

Plane-strain : Only non-zero strain
components are x ,  y and xy in the x-y plane since z ≠0

Note: Plane-stress case ≠ plane-strain case

6
 Positive normal strain x and y cause elongation;
 Positive shear strain xy causes change (reduce) of angle AOB (Fig. a);
 Both the x-y and x-y system follow the right-hand rule;
 The orientation of an inclined plane (on which the normal and shear strain
components are to be determined) will be defined using the angle θ. The angle
is measured from the positive x- to positive x-axis. It is positive if it follows
the curl of the right-hand fingers (Fig. b).

y
xy y
y
ydy A
x
dy
x  x
O B
dx xdx Positive sign convention

(a) (b)

7
 y, v
y, v1
u: displacement component along the x direction
x, u1
v: displacement component along the y direction

 x, u
u1 : displacement component along the x direction
Positive sign convention
v1: displacement component along the y direction

We know that

u1 v1 u1 v1

 x  ,  y  ,  xy   (5a)
x y y x

u v u v
x  ,  y  ,  xy   (5b)
x y y x
You have seen above equations in Lecture 2

Therefore, we first need to know the relations betweens

the coordinates (xoy) and (xoy), and the relations between
the displacements in two different coordinate systems.
8
 y, v
y, v1 x'  x cos   y sin 
x, u1 (7)

y '   x sin   y cos 


 x, u
x  x' cos   y ' sin 
(8)

y  x' sin   y ' cos 

From the figure, you can see that


u1  u cos  v sin  dx dx
 cos    sin 
(6) dx' dy '
(9)
v1  u sin   v cos dy dy
 sin   cos 
dx ' dy '

9
 y, v
y, v1
x, u1

 x, u Substitution of Eqs. (6) and (9) into Eq. (5a) gives

u1 u v
 x   cos   sin 
x x x
 u x u y   v x v y 
    cos      sin 
 x x y x   x x y x 
 u u   v v 
  cos   sin   cos    cos   sin   sin 
 x y   x y 
u v  u v 
 cos 2   sin 2      sin  cos 
x y  y x 

Thus  x   x cos2    y sin 2    xy sin  cos

10
 y, v
y, v1
x, u1

 x, u

v1 u v
 y   sin   cos 
y  y  y 
 u x u y   v x v y 
     sin      cos 
 x y y y   x y y y 
 u u   v v 
  sin   cos   sin     sin   cos   cos 
 x y   x y 
u 2 v  u v 
 sin   cos      sin  cos 
2

x y  y x 

Thus  y   x sin 2    y cos 2    xy sin  cos

11
 y, v
y, v1
x, u1

 x, u

u1 v1
 xy  
y x

u1 u v
 cos   sin 

y y  y 
 u x u y   v x v y 
    cos      sin 
 x y y y   x y y y 
 u u   v v 
   sin   cos   cos     sin   cos   sin 
 x y   x y 
 v u  u v
    sin  cos   cos 2   sin 2 
 y x  y x
12
 y, v
y, v1
x, u1

 x, u

u1 v1
 xy  
y x

v1 u v
 sin   cos 
x x x
 u x u y   v x v y 
     sin      cos 
 x x y x   x x y x 
 u u   v v 
   cos   sin   sin    cos   sin   cos 
 x y   x y 
 v u  u v
    sin  cos   sin 2   cos 2 
 y x  y x
13
 y, v
y, v1
x, u1

 x, u

u1 v1
 xy  
y x
 v u  u v
    sin  cos   cos 2   sin 2 
 y x  y x
 v u  u v
    sin  cos   sin 2   cos 2 
 y x  y x

Thus  xy  2 x y sin  cos   xy cos2   sin 2  

14
 Finally, strain transformation in 2D：
  x  c 2  x  s 2  y  sc xy
 c  cos  , s  sin 
  y  s 2  x  c 2  y  sc xy (10)
  2sc  2 sc  (c 2  s 2 ) 
 xy x y xy

  x   x   c2 s2 sc 
     
  y   T ()  y  [T ()]   s 2 c2  sc 
OR (11) where
     2 sc 2 sc c 2  s 2 
 xy   xy   
The system of three linear three algebraic equations (11) can be solved to express the strains in
the (x-y) coordinate system in terms of the strains in the (x-y) system. This yields

  x  c 2  x  s 2  y  sc xy In fact, note that (x-y) inclines anti-clockwise 


   s 2
  c 2
 y  sc xy (12)
to (x-y) is equivalent to (x-y) inclines
y x
clockwise  to (x-y) . Thus [t ()]  [T ()]
  2sc  2sc  (c 2  s 2 ) 
 xy x y xy 

 x    x   c2 s2  sc 
     
OR   y   [t ()]  y  (13) where [t ()]   s 2 c2 sc 
    2 sc  2 sc c 2  s 2 
 xy  xy   
15
Example 2: A small element of material at a point is subjected to a state of
plane strain, x=500(106), y= 300(106), xy=200(106), which tends to distort
the element as shown in Fig. a. Determine the equivalent strains acting on an
element of the material oriented at the point, clockwise 30 from the original
position.

Solutions
Since θ is positive if it is counter-clockwise, here = 30 (see Fig. b)

  x   cos 2  sin 2  sin  cos    500 

  
  y    sin 
2
cos 2 
 
 sin  cos    300 10 6  
   2 sin  cos  2 sin  cos  cos 2   sin 2   200 
 xy    
 213 
 

  13.4 10 6 
 793 
 

The results are shown on the element as

shown in Fig. c.

16
 Example 3: Pure shear of an element shown in Fig. a. Obtain the normal strain in the 45°direction.

Solution
0   45o ,  x  0 ,  y  0 ,  xy   0
Thus  45o  sin  cos  0   0 / 2  0 / 2G (i)
0 0
Eq. (1) is the desired answer. You can also obtain the solution in the
following manner
0
 x  0 ,  y  0 ,  xy  0
(a)
 45o  2 sin  cos  0  0 ,   45o  2 sin  cos  0  0

min =   max =  If the material is linearly elastic and follows Hooke’s law, normal strain in the
45°direction is

 45o  45o 0  0 (1  )0

 45o      (ii)
E E E E E
From Eqs. (i) and (ii), we see

max =  min =  
E
G
2(1   )

17
Example 4: A thin disk is subject to a uniform lateral pressure p as shown in the figure. The material of the disk is
isotropic.
(i) Prove that the normal stress is p for any direction and the shear stress is zero everywhere inside the disk.
(ii) What is the normal strain in z direction (normal to the x-y plane) if the normal stress in z direction is zero.

y Solution
From the problem description, it can be seen that
m
 x   p ,  y   p ,  xy  0 (i)

The normal stress in direction  (along m) is
Disk x
m  c 2 x  s 2 y  2sc xy  c 2 p  s 2 p  2sc  0  (c 2  s 2 ) p   p (ii)
p
Thus, the normal stress has no dependence on. It is
constant in any other direction and is  p

The shear stress in the plane normal to m is

 m   sc ( p)  sc ( p)  (c 2  s 2 )  0  0 (ii)

Thus, there is no shear stress inside the disk.

The normal strain along the z direction is found from

the three-dimensional stress-strain relation as

 x  y  z ( p) ( p) 0 2p

z        
E E E E E E E

18
3.3 Orthotropic materials

* For orthotropic materials, their properties in different main directions are

different.

Reinforced concrete

Fibre reinforced composites (FRC)

Our human bones are

orthotropic materials

19
 1
A body has a stress in the 1-direction only.
3 1
The strain in the x direction will be: 1 
2 E1
The strains in the 2 and 3 directions are:
1 
1  2   υ121   υ12 ,  3   υ131   υ13 1
1 E1 E1
2

3 2
2 
2 E2
2 
1   υ 21 2   υ 21 ,  3   υ 23 2   υ 23 2
E2 E2
2
1

A stress in the 3-direction only

3 3
3
2 3 
E3
3 
1   υ31 3   υ31 ,  2   υ32 3   υ32 3
1 E3 E3
3

where Ei is the Young’s modulus along axis i. Gij is the shear modulus in direction j on the plane whose normal is
in direction i. ij is the Poisson’s ratio that corresponds to a contraction in direction j when an extension is applied
in direction i.
20
 1  
1   υ 21 2  υ 31 3
E1 E2 E3
By the superposition principle, 1  2   These are the strain-stress
 2   υ12   υ32 3
The total strains are obtained E1 E2 E3 relations for orthotropic materials

1  
 3   υ13  υ 23 2  3
E1 E2 E3

For orthotropic materials, the shear stresses do not produce normal strains,
the normal stresses do not produce any shear strains
 Use Hooke’s Law for shear stress and shear strain

23 13
12

1
12 
1
  23   23 1
G12 1 2 G23 13  13
G13

Gij is the shear modulus in direction i on the plane whose normal is in direction j.

ij is the shear stress in direction j on the plane whose normal is in direction i. 21
Example 5: Give the strain energy density expression
of the orthotropic body if there are normal strains Or
(normal stresses) only.
1 12 1 υ 211 2 1 υ3113
Solution U  
2 E1 2 E2 2 E3
1 υ12 2 1 1  22 1 υ32 2 3
In Mechanics of Materials, we have seen that the   
strain energy density for a linear and elastic body is 2 E1 2 E2 2 E3
1 1 υ1331 1 υ 233 2 1 32
U     
2 2 E1 2 E2 2 E3
This is for one dimensional strain and strain. In three
dimensional, it should be written as There are the strain energy density for 3D
orthotropic solids.
U
1
11   2 2  33 
2
Substituting the strain-stress relations for orthotropic
materials.

1  1   
U 1   υ 21 2  υ31 3 
2  E1 E2 E3 
1     
  2   υ12 1  2  υ32 3 
2  E1 E2 E3 
1     
 3   υ13 1  υ 23 2  3 
2  E1 E2 E3 

22
Example 6: If a normal stress 1 is applied to an orthotropic
body. A normal stress 2 is applied subsequently. Give the
strain energy density of the body.
Solution
(1) When 1 is applied, normal strains Below shaded area:
develop in directions 1 and 2. But only work done by 1
1 does the work because 2 is not when it is applied
applied. The work done by 1 is 1 1
1 1 
W1  111  1 1 1 1
2 2 E1 11 : strain in 21 : strain in
11
(2) Due to application of 2, direction 1 due to direction 2 due to
additional normal strains develop 2 application of 1 application of 1
in directions 1 and 2. These are
22
 
12   υ 21 2 ,  22  2
E2 E2
Remember: these are the extra 1 2
normal strains resulted in by the
12
application of 2.
2 11 22 : strain in 2
(3) Note that when 2 is applying, both 1 and 2 will do direction due to
12: strain in application of 2
extra works. These works are
direction 1 due to
1 application of 2 Above shaded area:
W2   2   22  1  12 work done by 2
2 Shaded area: Work done by when it is applied.
1     1 when 2 is applied and
  2  2  1    υ 21 2  12 is developed.
2 E2  E2 

(5) Note that the work done by the applications of 1 and 2

(4) The total work done by 1 and 2 to the body is (1)+(3) will store in the body as the strain energy. Thus, the strain
energy density of the body U is equal to W
1 12 1 2 1  22 1 12   1  22
W  21  U  21 1 2 
2 E1 E2 2 E2 2 E1 E2 2 E2 23
Example 7: Example 6 obtains the strain energy density when
the order of the loads is 1 first, then 2.

By direct analogy, if the order of the loads is 2 first, then 1,

the strain energy density is

1 12  2 1 1  22
U  12 
2 E1 E1 2 E2

Example 8: What conclusion can you draw from Examples 6 and 7?

Solution: The strain energy in a linear and elastic body does not

depend on the order of the loading. It only depend on the final

state of the loads. Thus the strain energy density U of example 6

and the strain energy density U of example 7 are equal.

 21 12

Accordingly, .
1  
E2 E1
1   υ12 2  υ13 3
E1 E1 E1

31 13 32 23 1  2 

This shows the symmetry of the material constants. You can also obtain:
E3
 ,
E1 E3

E2
 2   υ12   υ 23 3
E1 E2 E2

1  
 3   υ13  υ 23 2  3
E1 E2 E3
24
Strain-strain relation (constitutive equations) of 3D orthotropic solids are

1  12  13
 1   1   s11 s12 s13 0 0 0 s11  s12  s13 
s E1 E1 E1
 
  2
 
  2  12 s22 s23 0 0 0  1  23
 3   3   s13 s23 s33 0 0 0 s22  s23 
  where E2 E2
  [ S ]   , [ S ]  

  
  0 0 0 s44 0 0 1
23 23
s33 
 13   13  0 0 0 0 s55 0 E3
     
 12  12   0 0 0 0 0 s66  1 1 1
s 44  s 55  s 66 
G23 G13 G12

Where [S] is commonly known as the compliance matrix.

There are 9 independent material constants

Matrix material Steel reinforced concrete can be considered as a fiber

reinforcing composite. Steel: fiber; Concrete: matrix
Fiber cross-section
25
Finally, strain-strain relation (constitutive equations) of 3D orthotropic solids are

 1   1   s11 s12 s13 0 0 0

    s s22 s23 0 0 0 
  2   2  12
 3   3   s13 s23 s33 0 0 0
   [ S ]   , [ S ]   

  23 
  23 0 0 0 s44 0 0
 13   13  0 0 0 0 s55 0
     
 12  12   0 0 0 0 0 s66 

E
G
2(1  υ)

 1   1  c11 c12 c13 0 0 0

    c 0 
Stiffness matrix [C] for   2   2  12 c22 c23 0 0
orthotropic materials  3    3  c13 c23 c33 0 0 0 1
   [C ] , [C ]     [S ]
 23   23  0 0 0 c44 0 0
 13   13  0 0 0 0 c55 0
     
12   12   0 0 0 0 0 c66 

Where [C] is the inversion of [S] and is known as the stiffness matrix.
26
 L4 Transformation of stress-stress relations,
Anisotropic materials, Basic concept of Numerical method by
finite element method

 In Lecture 3. We have studied transformation of stresses. We have also studied

transformation of strains.

 4.1 In this lecture, we will study transformation of stress-strain relation between

different rectilinear co-ordinate systems.

 4.2 We will also introduce the concept of anisotropic material

 4.3 Equations of thermal strains for orthotropic and anisotropic materials will also be
given (already given in Lecture 3)

 Appendix A Matrix analysis by MATLAB is introduced.

 Appendix B Finite element numerical method (Ansys) is also introduced briefly.

1
4.1 Two dimensional stress-strain relationship for anisotropic materials y

• Material properties (the stress-strain relations) behavior 2 1

different reference coordinate systems may be very different. 
• Consider the right figure. A material is orthotropic in (1-2)
coordinate system with elastic properties, E1, E2, 12 and G12.

The material is orthotropic in (1-2) coordinate system and the

corresponding stress-strain relations are
 1  12 
 0 
 1   1   E1 E1 
      12
0 
1
  2   [ S ] 2 , where [ S ]   
    
E1 E2
 12   12  1 
 0 0 
 G12 
The matrix [S] is commonly known as the compliance matrix.
The inverse of [S] is known as the stiffness matrix and is
usually denoted as [Q].
  •The 1-2 coordinate system inclines angle 
 E 12 E2 0 
1 1  1  (anti-clockwise is positive) to the reference
[Q]  [ S ]  12 E2 E2 0 
12
2
E2 
1   E 
2 coordinate system x-y .
E1  0 0 G12 1  12 2 
  E1  •We will obtain the stress-strain relationship in
Thus, the stress-strain relationship is also usually expressed as the x-y coordinate system.
 1   1 
   
 2   [Q]  2  (A)
    2
 12   12 
 Procedure: y
1. The stress-strain relationship in 1-2
coordinate system is 2 1
 1  12 
 0  
 1   1   E1 E1 
    
  2   [ S ] 2 , [ S ]   12 0 
1

     1
E E2
 12   12  1 
 0 0  c  cos  , s  sin 
 G12 

2. Obtain the inverse of [S] is denoted as [Q]. Thus 3. Using Eq. (1) – slide #5 of Lecture 4, the stresses in (x-
y) coordinate system are expressed in terms of the stresses
in (1-2) coordinate system as
 1   1 
     x   1  c 2 s 2  2 sc 
 2   [Q]  2  (A)      2 
     y   [t ()] 2 , [t ()]   s c2 2sc  (B)
 12   12       sc  sc c 2  s 2 
 
xy  12   

You can obtain [Q] by using MATLAB

4. Using Eq. (13) – slide #15 of Lecture 3, the strains
software or from following equation
in (1-2) coordinate system are written in terms of the
  strains in (x-y) coordinate system as
 E 12 E2 0 
1 1  1   1   x   c2 s2 sc 
[Q]  [ S ]  12 E2 E2 0       
12
2
E2    2   [T ()]  y , [T ()]   s  sc  (C)
  E 
2
1 2 c2
E1  0 0 G12 1  12 2     
 E1   12   2sc 2 sc c 2  s 2 
  xy   

3
5. The stress-strain relations of the material in (x-
y
y) coordinate system are obtained by substituting
Eqs (A) and (C) into Eq. (B) . This gives
2 1
 x   1   1   x 
        
 y   [t ()] 2   [t ()][Q]  2   [t ()][Q][T ()]  y  (D)
       
 xy   12   12   xy 
6. Equation (D) is commonly written as

 x   x 
   
  y   [Q]  y  (E)
   
 xy   xy 
where

[Q]  [t ][Q][T ()] (F)

It is the stiffness matrix of the material in the (x-y)
coordinate system.

y
Example 1：Most composite materials and concrete structures are
orthotropic. The elastic properties of an orthotropic material in its main 2
directions (the 1-2 coordinate system) are E1= 200 MPa, E2= 100 MPa, 1
G12= 50 MPa, 12=0.6. The main directions incline 30 (anti-clockwise) to
the reference coordination system x-y. Obtain the constitutive equations
(stress-strain relations) of the material in (x-y) coordinate system.

4
 Solution:
y
1. First, the stress-strain relation ( the matrix [Q] ) in the main
directions (1 and 2) of the material should be given 2
1
 
 E 12 E2 0  243.9 73.17 0 
1  1   
[Q]  12 E2 E2 0   73.17 122.0 0  MPa
12
2
E2 
1  12 2
E2   0 0 50
E1  0 0 
G12 1  
  E1 

2. Also obtain the matrices for the transformations of c  cos 30 , s  sin 30
stresses and strains (find these matrices in Lecture 3)

c 2 s 2  2sc   0.75 0.25  0.866

 2  
[t ()]   s c 2
2 sc    0.25 0.75 0.866 
4. Therefore, the stress-strain relations of the
 sc  sc c 2  s 2  0.433  0.433 0.5 
  material in x-y coordinate system is

 c2 s2 sc   0.75 0.25 0.433    x  209.8 76.83 28.51   x 

 2 
[T ()]   s c2  sc    0.25 0.75  0.433     
 y   76.83 148.8 24.29   y 
 2 sc 2sc c 2  s 2   0.866 0.866 0.5     28.51 24.29 53.66  
 
 xy     xy 
3. Calculate the stiffness matrix in (x-y) coordinate
system You can also see that application of a normal
stress can result in the shear deformation of the
209.8 76.83 28.51 material, and a shear stress can also result in
[Q]  t [Q][T ()]  76.83 148.8 24.29 MPa the normal strains.
 28.51 24.29 53.66

You can see that all elements in the matrix are non-zero,
although there are some elements in matrix [Q] are zero. 5
4.2 Anisotropic materials • From example 1, we have seen that application of a normal stress may result in the
shear deformation of the material, and application of a shear stress may result in the
normal strains.
• Materials with such behaviour are anisotropic materials.

  * The behaviour is quite different from the

orthotropic materials.

Stiffness matrix [C] for anisotropic materials：

 1  c11 c12 c13 c14 c15 c16   1 

   c c22 c23 c24 c25 c26    2 
 2   12
 3  c13 c23 c33 c34 c35 c36    3 
   
 23  c14 c24 c34 c44 c45 c36   23 
 13  c15 c25 c35 c45 c55 c56   13 
    
12  c16 c26 c36 c46 c56 c66   12 
There are 21 independent material constants
3
2 • Material in one coordinate system may be orthotropic but can
become anisotropic if the coordinate system is changed.
6
1
 Compare with the orthotropic materials (Studied in Lecture 3)

* The number of independent elastic constants for

orthotropic materials is 9. 3
3

 1   1   1   1  2
        2
  2  2   2  2

 3  
 3  
 3  
 3 
   [S ]     [C ] , 1
 23   23   23   23  1
 13   13   13   13 
        Above figure: the material is orthotropic in 1-2, 2-3

 12 
 
12  
12  
 12 
 and 1-3 plans

 
 1 12 13 
 E   0 0 0 
E1 E1
 1 
 12 1 23
 0 0 0   s11 s12 s13 0 0 0
 E1 E2 E2 
     s12 s22 s23 0 0 0 
0  
1
 13  23 0 0
[S ]  
E1 E2 E3    s13 s23 s33 0 0 0
 1  0 0 0 s44 0 0

 0 0 0 0 0  
 G23  0 0 0 0 s55 0
 0  
0   0
1
0 0 0 0 0 0 0 s66 
 G13 
 1 
 0 0 0 0 0  •Application of a normal stress does not
 G12  result in the shear deformation.
•A shear stress does not result in any
normal strains
7
Compare with the isotropic materials (Studied in Lecture 1)

 1   1   1   1  3
       
  2  2   2  2 2
  3   3   3   3 
   [S ]      [C ] 

  23  23   23   23  1
 13   13   13   13 
       
 12  12  12   12 
* Two independent elastic
 1   0 0 0  constants for isotropic materials.
  1   0 0 0 
  E
G
1     1 0 0 0  2(1  )
[S ]   
E 0 0 0 2(1  ) 0 0 
 0 0 0 0 2(1  ) 0 
 
 0 0 0 0 0 2(1  )

2(1  ) 2 2 0 0 0 
 2 2(1  ) 2  0 0 0 
 
1 E  2 2 2(1  ) 0 0 0 
[C ]  [ S ]   
2(1  2)(1  )  0 0 0 1  2 0 0 
 0 0 0 0 1  2 0 
 
 0 0 0 0 0 1  2
8
Example 2：Consider the plane stress deformation of an isotropic material. Suppose y
E= 170 MPa and =0.3 in the 1-2 coordinate system which inclines =20 anti-
clockwise to the reference coordination system x-y. Obtain the constitutive equations 2
(stress-strain relations) of the material in x-y coordinate system. 1
Solution:
1. First, the stress-strain relation ( the matrix [Q] ) in the main
directions (1 and 2) of the material should be given
c  cos 20 , s  sin 20
 1   1   
1  0  186.8 56.04 0 
    E 
 2   [Q]  2 , [Q]   1 0   56.04 186.8 0  MPa
    1  
2
1 
 12   12   0 0   0 0 65.38
 2 

2. Also obtain the matrices for the transformations of 4. Therefore, the stress-strain relations of the
stresses and strains material in x-y coordinate system is

c 2 s 2  2sc  0.8830 0.1170  0.6428  x   x 

 2  186.8 56.04 0 
[t ()]   s 2
2 sc   0.1170 0.8830 0.6428        MPa
c
 y   [Q]  y , [Q]  56.04 186.8 0
 sc  sc c 2  s 2  0.3214  0.3214 0.7660       
   
 xy   xy   0 0 65.38

 c2 s2 sc   0.8830 0.1170 0.3214  You can also see that stiffness matrices in
 
[T ()]   s 2 c2  sc    0.1170 0.8830  0.3214 different coordinate systems are same.
 2 sc 2 sc c 2  s 2   0.6428 0.6428 0.7660 
 
If you try other values of E,  and , you can find
3. Find the stiffness matrix in x-y coordinate system the same rule

186.8 56.04 0  Therefore, the properties of isotropic materials do


[Q]  t [Q][T ()]  56.04 186.8 0  MPa not depend on the coordinate system
 0 0 65.38 9
 4.3 Thermal strains and stresses in 3-D orthotropic materials

 1   1   1   s11 s12 s13 0 0 0

3
       s s22 s23 0 0 0 
  2  2   2  12
  3   3   3  s s23 s33 0 0 0
   [ S ]     (T ), [ S ]   13  2

 23   23   0  0 0 0 s44 0 0
 13   13   0  0 0 0 0 s55 0
       
 12  12   0   0 0 0 0 0 s66 
1
Above figure: the material is
orthotropic in 1-2, 2-3 and 1-3
plans

 1   1   1   1   1 
          
 2  2   2  2  2
 3    3   3  1
 3   
1  3 
   [C ]     (T ), [C ]  [ S ] ,    [ S ]  
 23   23   0  0 0
 13   13   0  0 0
         
12   12   0   0   0 

1: coefficient of thermal expansion along direction 1. The coefficients of thermal expansion are property
2: coefficient of thermal expansion along direction 2. of the material. They are determined experimental by
applying a change in the temperature and measuring
3: coefficient of thermal expansion along direction 3.
the changes in dimensions of the specimen

10
Appendix A: Matrix analysis by MATLAB software
A1: Matrix multiplication

In this lecture, we frequently see the multiplication of matrices, for example: Q  t Q [T ]

Definition: [A] is an 3 × 3 matrix, [B] is another 3 × 3 matrix,

The subscripts indicate the location of the elements of the matrix.

 a11 a12 a13  b11 b12 b13  First index: row, second index: column .
[ A]  a21 a22 a23  , [ B]  b21 b22 b23  . For example, a21 is the element at the second row and first
a31 a32 a33  b31 b32 b33  column of matrix [A] , b32 is the element at the third row and
second column of matrix [B].

The matrix product [C] = [A][B] is defined to be the 3 × 3 matrix

 c11 c12 c13 

[C ]  c21 c22 c23  , Such that cij  ai1b1 j  ai 2b2 j  ai 3b3 j
c31 c32 c33 
This means that, cij is the product of the elements in the i-th row of matrix [A] and j-th column of matrix [B] .

If your calculator is functionally powerful, you can let it do this for you.
You can also use commercial software MATLAB. For example,

11
A2: Product of a matrix and a column

 a11 a12 a13  b1 

 
Definition: [A] is an 3 × 3 matrix, [B] is a column, [ A]  a21 a22 a23  , {b}  b2 .
a31 a32 a33  b 
 3
 c1 
 
The matrix product [C] = [A]{B} is a column, {c}  c2  , Such that ci  ai1b1  ai 2b2  ai 3b3
c 
 3

For example, if you use commercial software MATLAB,

12
A3: Matrix inverse

1 0 0
1 
If [A] is an 3 × 3 matrix, the inverse of [A] is [A]1 , such that [ A][ A]  0 1 0 .

 
0 0 1

Identity matrix

You can also use commercial software, for example, MATLAB do this for you. For example,

13
A4. Solution of linear algebraic equation You can easily use commercial software, for
example, MATLAB obtain the solution. For
The linear system algebraic equations example, if you want to solve below equations:
( xi are unknowns to solve, aij and bi are
known constants): 2 x1  3x2  5 x3  3,
4 x1  6 x2  3x3  1,
a11x1  a12 x2  a13 x3  b1 ,
7 x1  2 x2  4 x3  2.
a21x1  a22 x2  a23 x3  b2 ,
The MATLAB commands and results are:
a31x1  a32 x2  a33 x3  b3 .

They can be re-written as

[ A]x  {b}, (1)

where

 a11 a12 a13   x1  b1 

   
[ A]  a21 a22 a23  , [ x]   x2 , [b]  b2 .
a31 a32 a33  x  b 
 3  3
The solution of Eq. (1) is

x  [ A]1{b}

14
 Appendix B – Finite element method

* “Finite Element” means that we “divide a structural system (or a solid material) into a finite number of elements”;

* Element Method (FEM) is a must known numerical analysis tool for structural engineers, mechanical engineers and
materials scientists

* We have used FEM in the computer lab classes in the weeks 2 to 3 and we will continue using it in our future classes.

– only take a rod as an example

F1, u1 F2, u2 x

1 EA, L 2 This is the “Link” Element in Ansys

First: consider one rod element only

• The element has length L, elasticity modulus E, cross sectional area A.

• The element is subjected to external forces F1 and F2, at its grids 1 and 2, respectively. Positive
signs of the forces means that the forces are along the direction +x.

• The axial displacements of the grids 1 and 2 are u1 and u2, respectively.
15
 F1, u1 F2, u2 x

1 EA, L 2

Equation (B4) can be written as

The strain of the element is
 EA EA 
u u  L 
  2 1 (B1) L  u1   F1 
L  EA EA       (B5)
  u2   F2 
The (internal) axial force of the element is  L L 
(remember: N is positive if it is tensile)
u2  u1
N  EA (B2) Eq. (B5) is the equilibrium equation of the element.
L
The element must be in equilibrium state
(the external and internal forces must be •Once the external forces in the grids are given, the
balanced). Therefore, displacements of the element can be calculated

F1   N , F2  N (B3) •If the displacements in the grids are prescribed,

the forces in the elements can also be found.
From Eqs (B2) and (B3), we obtain

EA EA
u1  u2  F1 ,
L L (B4)
EA EA
 u1  u 2  F2
L L

16
 • The length of the element is L, elasticity modulus E,
How about two elements? cross sectional area A. The subscripts denote the
element number.

F1, u1 F2, u2 F3, u3 x • The external forces are F1, F2 and F3, at grids 1, 2 and
1 E1A1, L1 E2A2, L2 3, respectively. Positive signs of the external forces
2 3
means that the forces are along the direction +x.
The rod is divided into two elements
• The internal forces in elements 1 and 2 are N1 and N2,
respectively. Positive signs of the internal forces
F1, u1 N1, u2 N2, u2 F3, u3 means that the forces are tensile.

1 E1A1, L1 2 2 E2A2, L2 3 • The axial displacements of the grids 1, 2 and 3 are u1,
u2, and u3, respectively.
Element #1 Element #2
From Eqs (B6)- (B8), we obtain
The (internal) axial forces of the elements are
(remember: N is positive if it is tensile) E1 A1 EA
u1  1 1 u2  F1
u2  u1 u u L1 L1
N1  E1 A1 , N 2  E2 A2 3 2 (B6)
L1 L2 E2 A2 E A
 u3  2 2 u2  F3 (B9)
The element must be in equilibrium state L2 L2
(the external and internal forces must be
E1 A1 E A E A  E A
balanced). Therefore,  u1   1 1  2 2 u2  2 2 u3  F2
L1  L1 L2  L2
F1   N1 , F3  N 2 (B7)

The grid 2 must also be in equilibrium F2

state (see right figure). Therefore,
N1 N2
N1  N2  F2 (B8) Grid #2
17
 F1, u1 F2, u2 F3, u3 x Example: A non-homogeneous rod whose cross-sectional
area varies with x according to A=3x. The rod has elasticity
1 E1A1, L1 2 E2A2, L2 3 modulus 1 and length 2. It is fixed at the left end and is
subjected to an axial tension load P=10. Obtain the
The rod is divided into two elements displacement of the rod at the middle and the right end

Equation (B9) can be written as

 E1 A1 EA  P=10 x
  1 1 0 
 L1 L1  u1   F1 
 E1 A1 E1 A1 E2 A2 E2 A2     
  u2    F2  (B10)
 L1 L1 L2 L2     
 E A E2 A2  u3   F3  Solution
 0  2 2 
 L2 L2 
F1, u1 u2 P, u3 x
Eq. (B10) is the equilibrium equation of the rod 1 E1A1, L1 2 E2A2, L2 3
when it is divided into two elements.
Two rod elements
•Once the external forces in the grids are given, the
displacements of the element can be calculated We only divide the rod into two elements of equal length.
Element 1: E1=1, L1=1, A1=30.5=2.5 (use the cross
•If the displacements in the grids are prescribed, sectional area at the center of the element, i.e., at x=0.5) ;
the forces in the elements can also be found.
Element 2: E2=1, L2=1, A2=31.5=1.5 (use the cross
sectional area at the center of the element, i.e., at x=1.5).

The loading conditions of the problem are

u1  0 , F2  0, F3  10

18
The finite element formula, Eq. (B10) becomes
The exact solutions of the displacements
 2.5  2.5 0  u1   F1  at grids 2 and 3 are:
 2.5   u   F 
 4 1.5 2   2  u2  4.055,
 0  1.5 1.5     
u3   F3  u3  10.99 (E-3)

Applying the loading conditions to this equation gives

The error of the finite element method is
 2.5  2.5 0   0   F1 
 2.5  u    0  4.000  4.055
 4  1.5 2   
(E-1) At grid # 2:  1.4%,
 1.5 1.5      4.055
 0 u3  10
Solving 10.67  10.99
 2.9%,
10.99
u2  4.000
   (E-2)
u3  10.67 

19
 Lecture 5: Theory of beam of homogeneous and composite materials

 Forces in beam
 Strains in a beam under bending
 Stresses in a beam under bending
 Neutral plane, Neutral axis
 Bending stiffness of cross-section, Second moment of cross-section (moment of inertia)
 Beam of composite shapes
 Design of beams under an eccentric load

1
Beam: Supporting lateral (transverse) loads

Slotted hole Anchor

Beam bolt

Bearming Beam
plate

Concrete Roller Support

wall

Column

Base plate Pole

Beam
Hinged or
Pole Pinned Support

Fixed or Clamped or Beam

Built-in Support

Concrete pier

2
 1. Classification of Beams

A beam is a horizontal structural element that transmits the transverse loads to the supports by bending action.

Statically
Determinate
Beams
Simply supported beam Overhanging beam Cantilever beam

Statically
Indeterminate
Beams

Continuous beam Beam fixed at one end and simply Fixed beam
supported at the other end

• Simply supported beam

• Cantilever beam (one end fixed, one end free)
• Fixed end beam

• Simply supported end: no deflection, free rotation;

• Fixed end: no deflection, no rotation

3
 2. Types of loads on beams

Beams may be subjected to different types of loads

 Concentrated loads (point load) – discrete forces acting at a

specific point on the beam
Unit: N

 Distributed loads – act over a length of the beam and may be

uniform or non-uniform
Unit: N/m

 Moments or couples produced by two equal and opposite forces

acting on beam.
Unit: Nm

4
3. APPLICATIONS

Excessive shear forces and bending moments results in considerable

deformation and significant stresses

5
 4. Bending of beam

Geometry
z y z
A

o y
height o x

A
Axial view Cross-sectional view
Geometry and coordinates of the beam, size not in proportion

• Beam axis is parallel to the x-axis;

• Beam is symmetric about the x-z plane (i.e., everything on y =  and y =+ are identical)

• Beam bends in x-z plane (bends about the y-axis). Therefore, the displacement and forces
along the y-direction is not considered.

You have already studied theory of beams in Mechanics of Materials and/or Advanced Mechanics of
Materials but here we will use a different approach to demonstrate the theory. Once you understand
this, you can understand the theory of plates (which are 2-D) with no difficulty .

6
Internal force: Shear forces in the beam

* The out-of-plane transverse shearing force intensity Qx is shown acting on the sides of the element.
* These forces have dimensions of force (e.g., N).

z
y

Qx
o x

Qx

Sign Conventions :
* Forces are positive when they act in the directions shown.

* The positive shear force tends to rotate the beam anti-clockwise (*)

* Sign Conventions here may be different from what you used in Mechanics of Materials and/or Advanced
Mechanics of Materials. We use such approach because we need to make this to be coincident with our later
studies of 2-D plates

7
Internal force: Bending moments in the beam

* The bending moment Mx is shown acting on the sides of the element.

* These moments have dimensions of N.m.

z
y

o x
Mx Mx

Sign Conventions :
* Bending moments are positive when they act in the directions shown.

* Positive bending moment tends to compress the lower part of the beam and elongate the upper part (*).

* Again, Sign Conventions here may be different from what you used in Mechanics of Materials and/or Advanced
Mechanics of Materials. We use such approach so that the demonstrations of beams and plates are coincident with
each other.
* Note that M x on the positive face of the beam segment (i.e, the face normal to the positive x direction) makes the
beam rotates about the y axis.

8
External force: Distribution load

* Also shown is the transverse (lateral) external force, q(x) per unit length and arises
from, for example, an external pressure or gravitational field.
* The distribution load q has dimension of N/m.

Sign Conventions :
* The distribution load q is positive when it is along the positive z direction.

z y

o x

q(x)

9
Conclusion: Forces and moments in beam

Qx
dA
x o y

Mx
Qx Mx

Cross sectional view

q(x)

Acting of the bending moment Mx will result in normal

stresses on beam cross-section. Thus:

M x    x zdA A: the cross-section of the beam

A

Note: M x is the bending moment about the y axis. The

integration is performed on the entire cross section A of
the beam z
y

x
x
o

10
4.1 Geometry and deformation m n
c d
• In classic bean theory, an assumption is made that
a b
after deformation, the cross section of the beam is
m m
still a plane AND is still normal to the beam axis

• Beam of such nature is commonly known as the

Euler beam.

Plane assumption

z
x

11
We can see that:

1. There is no obvious displacement along the y direction.

2. There is a plane in the beam which (1) is normal to the z axis in the beam, and (2) has no displacement along
the axial direction. This plane is the so-called Neutral Plane (the white planes of the below figures).

z y

Original shape
o x

z y

x Deformed
o
shape

12
Assumptions:
1. The neutral plane is located at z = 0.
2. The beam is very thin. This means that the height (dimension in direction z) of the beam is considerably
smaller than the length of the beam.
3. The displacement along the z direction is denoted as w. w is same at any point on a cross section of the
beam. Therefore, w is only a function of x. This is, w = w(x).
4. The axial displacement is denoted as u. It varies with z and x. This is, u = u(x, z).
5. Because the beam is very thin, it can not hold any shear stress. Therefore, xz is negligible.

x
The beam is very thin. Compare to the
normal stress x, the shear stress xz in the
beam is negligible.

z y

o x

Deformed shape

13
4.2 Strain-displacement relation

* Based on the assumption #5 (last slide), we see that

0   xz  G xz   xz  0

u w u w
 xz    
z x z x
w
 u ( x, z )   z  u0 ( x )
x
w
* Neutral axis is located at z = 0. Therefore, 0  u ( x)  0   u0 ( x )  u0 ( x )  0
x
w
* As a result, u ( x, z )   z
x
x
* Normal strain of the beam :

u 2w
x    z 2 (1)
x x
z y

x Deformed
o
shape

14
4.3 Stress-displacement relation, moment-displacement relation
2w
* The normal stress, in terms of the displacement of the beam, is :  x  E x   Ez (2)
x 2

E : Modulus of elasticity of the material of the beam.

If the beam cross-section is non-homogeneous (e.g.,

layered materials), E will be a function of z. z
y
* The bending moment, in terms of the
displacement of the beam, is x
x
 o
2w   2w 
M x    x zdA   E   z 2 zdA  D  2  (3)
A A
 x   x 

* The quantity D is the bending stiffness of the beam cross-

section and is dependent on the material and shape of the z
beam cross-section: max
D   Ez dA
2
(4)
A

* From Eqs (2) and (3), one obtains the dependence x(z)
of the stress in the beam on the bending moment: x
Mxz
x  E (5) M M
D
* Thus, once the applied bending moment is given/obtained, the max
bending stress can be determined. The distribution of the normal stress
on the cross-section of the beam is shown in right figure.
15
4.4 Neutral axis, neutral plane
Axis of symmetry
• In this unit, the (x-y) plane is the neutral plane of the z
beam.

• The y axis of the beam is always located on the y

neutral plane. Therefore, the y axis is known as the
“neutral axis” of the beam. Neutral plane

Longitudinal axis Neutral axis

4.5 The bending stresses in Homogeneous Beams

If you study Eq. (4) of Slide #15 , D   Ez 2dA , you can see that if the cross-section
A
of the beam is homogeneous (i.e., beam made of a single material):

D  EI (6)
where,
I   z 2dA (7)
A

is the second moment (in some books /papers, this is also defined as moment of inertia) of
the cross-section about the neutral axis

* The stress for the beam of homogeneous material is

Mxz * To obtain the bending stress, we will need to know second

x  (8)
moment of area about the neutral axis of the cross-section.
I
16
 z
z
Location of neutral axis (neutral plane) r
y y
– for homogeneous cross-section h

* For cross-section with regular shape (such as circle or b

rectangular), its neutral axis passes through the geometry
center (i.e., centroid) of the cross-section. The centroid is
in the center of the cross-section.

* Symmetric shape, its centroid is also in the center.

Cross-sections with irregular shapes ?

17
* We consider an element of area dA in the cross section. The element is located at distance z from the neutral
axis, and therefore the stress x acting on the element is given by x = z Mx / I. The force acting on the element
is equal to x dA and is tensile when z is positive. Because there is no resultant force acting on the entire cross
section, the integral of xdA over the area of the entire cross section must vanish:
Mx
z   xdA   z
A A
I
dA  0 (9)

dA * Because Mx, and I are nonzero constants at any given cross section of a
bent beam, they are not involved in the integration over the cross-sectional
z area. Therefore, we can drop them from the equation and obtain
y
 zdA  0
A
(10)

* This equation states that the y axis must pass through the centroid of the cross
section. Since the y axis is also the neutral axis, we have arrived at the following
x (normal to the (y-z) plane)
important conclusion:
The neutral axis passes through the centroid of the cross-sectional area when
the material follows Hooke’s law.
z This observation makes it relatively simple to determine the position of the
max neutral axis.

x(z) How to determine the location of the centroid of

x the cross section ? - Next

M M

max 18
 Centroids of plane areas
zR
In order to obtain the position of centroid C (i.e., z ), we introduce a
z
reference coordinate system yRozR. From the previous slide ,we know that

dA  zdA   ( z
A A
R  z )dA  0
. y
C in which dA is a differential element of area having coordinates yR and zR.
zR
z Therefore, the coordinate z of the centroid C is equal to:
yR
z z
o dA dA
R R

z A
 A
(11)
• Plane area of arbitrary shape  dA
A
A
with centroid C
• yR is parallel to y where A is the total area of the figure (cross-section of the beam).
• zR is parallel to z

z  zR  z z
A
R dA : First moment of the cross section about yR axis

Once the location of the centroid is found, the second moment of

the cross-section can be calculated from Eq. (7). That is:

You should have seen this

I   z dA 2
formula in Mechanics of
A Materials.

or through Parallel-Axis Theorem (next slide

19
Using Parallel-Axis Theorem: The second moment of an area about an axis (e.g., the reference line of below figure) is
equal to the area’s second moment about a parallel axis passing through the “centroid” plus the product of the area and
the square of the perpendicular distance between the axes:

I ref line  I c  Ad 2 (12)

Where A is the area of the shape, d is the distance from the centroidal axis of the shape to the desired
parallel axis , Ic is the second moment of the area about its own neutral axis (which has a distance d
from the reference line).

dA The derivation:
z C
I ref line   zR2 dA   (d  z ) 2 dA  d 2  dA  2d  zdA   z 2dA
zR A A A A A

Since z=0 is at the centroid of the area, we know that

d
 zdA  0
Reference line A
We therefore get Eq. (12), where I c   z dA
2
.
About C
zR  z  d

You should have seen this in Mechanics of Materials.

20
• Composite shapes (same materials but different shapes) z

(1) Consider the shape as the combination of some A1

simple/regular elements: A1, A2,…. NA
y
y
z11
(2) Determine the neutral axis by Eq. (11) :
y
N z zy2
 zi Ai A2 2

z i 1
N

A
i 1
i Reference line

(3) Calculate Ii: the second moment of each simple (4) Second moment of the shape is the sum of all
element about the neutral axis (NA) by parallel axis simple/regular elements:
theorem:

I i  I ci  Ai di2  I ci  Ai ( zi  z ) 2 I total aboutNA  I1  I 2  

(5) Calculate the bending stress from:

This is the moment of the i-th simple element Mz
about its own neutral axis (e.g, the central point 
ci of the below rectangular element). I
In which z is the where you want to find the
stress. It is measured from the NA of the shape:
b
1 3
A  bh , I ci  bh
h 12
ci

di
NA 21
 Example 1: Find the maximum bending stress

Mz
120 mm

30 mm 5.5 kNm 5.5 kNm I
90 mm Mx= 5.5 kNm

zmax=? I=?

60 mm

Solution:
Step 1: Split the cross-section’s shape into two simple elements, A1 and A2

A1  120  30  3600 mm 2 z1  (30 / 2  90)  105 mm

A1
A2  90  60  5400 mm 2
z2  90 / 2  45 mm NP
105
Step 2: Locate the position of the neutral plane (NP) z
A2 45
z1 A1  z2 A2 105  3600  45  5400
z   69 mm
A1  A2 3600  5400

22
Step 3: Calculate the I values of each
element about NP: b

1 3 1 3
I bh  Ad 2 h
12 c
I c  bh
12
120 mm

30 mm A1
NP = NP
105
NP
90 mm
69 mm
+
69
A2 45

60 mm

1 1
I1  120  303  3600(105  69) 2 I2   60  903  5400(69  45) 2
12 12
 4.9356 106 mm 4  4.9356 10 6 m 4  6.7554 106 mm 4  6.7554 10 6 m 4

Step 4: Calculate the total I of the composite areas about NP:

I  I1  I 2  11.691106 m4

23
 120 mm
Step 5: Calculate the corresponding maximum value of bending stress 30 mm A1

NP
90 mm
A2 69 mm
Maximum stress in A1 is at the top of the beam:

Mztop (5500)  [(0.09  0.03  0.069))

 top  
I (11.69110 6 ) 60 mm

 23.993(106 ) Pa  23.993 MPa

compressive

Maximum stress in A2 is at the bottom of the beam:

Mzbottom (5500)  (0.069)

bottom   6
 32.461(10 6
) Pa  32.461 MPa
I (11.69110 )
tensile

24
4.6 Composite material beams
Composite Materials
Many types of composite beams have been developed in
recent years, primarily to save material and reduce weight.
For instance, sandwich beams are widely used in the
aviation and aerospace industries, where light weight plus
high strength and rigidity are required. Such familiar Wood Steel face-
core sheets
objects as skis, doors, wall panels, book shelves, and
cardboard boxes are also manufactured in sandwich style.

Reinforced Concrete E is not a constant on the cross-section

Concrete

Steel fiber

Composite structures
Concrete

Steel frame
Sandwich beams

25
 2w 2w
x  z 2  x   Ez
x x 2
z z

x x

M M

Normal strain variation (profile view): The

strain varies linearly with coordinate z. Normal stress variation (profile
However, the stress does not vary in this view): At the interface of materials,
way. This is due to the fact that the Young’s the strain is the same, but the stress
modulus also changes with z. changes suddenly.

2w 2w M
M  D 2  
x x 2 D Special case of homogeneous cross-section:

D   Ez 2dA : Bending stiffness of the cross section. 2w Mz


M 
x
A
2
EI I
EMz Mz D  EI I   z 2dA
  A
D I
26
 EMz
 Location of the neutral axis (neutral plane)
D
We will need to find the bending stiffness D
The position of the neutral axis is found from the condition that
about the neutral axis. Therefore, the first
the resultant force acting on the entire cross section along the
step is to locate the neutral axis.
beam axis vanishes. Therefore,
z
2w
A dA  0  A Ez x 2 dA  0  A EzdA  0 (13)
y For example, if the cross-section is made of two
simple materials,

 E zdA   E zdA  0
A1
1
A2
2

z
The first integral is evaluated over the cross-
sectional area of material 1; the second integral is
evaluated over the cross-sectional area of material 2.
y
Equation (13) is a generalized form of the analogous equation
for a beam of one material,  zdA  0 .
A
(a) Composite beam of two materials,
(b) Cross-section of beam, This observation makes it relatively simple to determine the
(c) Distribution of strains x throughout the position of the neutral axis for composite cross section.
height of the beam
(d) Distribution of stresses x in the beam for
the case where E2>E1 27
 Location of the neutral axis (neutral plane)

* In order to obtain the position of centroid C, we introduce a reference

zR coordinate system yROzR. From Eq. (13), we know that
z

dA
 E( z
A
R  z )dA  0

. y
C in which A is the total area of the figure, dA is a differential element of area
zR
z having coordinates ZR and y. Therefore, the coordinate z , where the neutral
axis passes through is equal to:
yR
o
• Plane area of arbitrary shape
with centroid C
 Ez R dA
• yR is parallel to y z A (14)
• zR is parallel to z  EdA
A

z  zR  z • This is a generalized form of the analogous equation for a beam of

one material,

z R dA
z A
Eq. (11), in Slide #19
 dAA

28
Composite material beams with regular cross section shapes z

(1) Determine the location of the neutral axis for the cross-section:
E1 AA
11

N
y
 Ez R dA E Az i i i
yz 1
z A
 i 1
(15) NP 1
 EdA
N

A
E A
i 1
i i
zy E2
AA zy
2
2 2 2
(2) Calculate Ii: the second moment of each simple
element about the neutral axis (NA) by parallel axis
theorem : Reference line
I i  I ci  Ai di2  I ci  Ai ( zi  z ) 2
(3) Find the bending stiffness of the cross-section
This is the moment of the i-th simple element about about the neutral axis of the shape:

D   Ez 2dA  ( EI ) total aboutNA

its own neutral axis (e.g, the central point ci of the
below rectangular element). (16)
A

 E1 I1  E2 I 2  

(4) Calculate the bending stress from:

b
EMz
1
A  bh , I ci  bh3 
12 D
h ci In which z is the where you want to find the
stress. It is measured from the NA of the shape:
di
NA
29
 Example 2: A composite beam is made of wood and reinforced with a
steel strip located on its bottom side. It has the cross-section area shown in
the figure. If the beam is subjected to a bending moment of M =2 kNm,
determine the normal stress at points B and C. Take Est = 200 GPa and Ew =
12 GPa.

1) Location of the neutral axis:

z
E z A
i i i

(200)(10)(20)(150)  (12)(95)(150)(150)
 36.38 mm
E A i i (200)(20)(150)  (12)(150)(150)

2) Second moments of areas about the neutral axis

(150)(20)3 NP
I1   (36.38  10) 2 (20)(150)  2.187(106 ) mm 4  2.187(106 ) m 4 z
12
(150)(150)3
I2   (36.38  95) 2 (150)(150)  119.5(106 ) mm 4  119.5(106 ) m 4
12
3) Bending stiffness of the entire cross section

D  E1I1  E2 I 2  200(109 )(2.187)(106 )  12(109 )(119.5)(106 )  1.8714(106 ) N  m2

4) Bending stress at points B and C

E1Mz 200(109 )(2000))(0.03638)

C    7.78 MPa
D 1.8714(106 )

E2 Mz 12(109 )(2000)(0.17  0.03638)

B    1.71 MPa
D 1.8714(106 )
30
5. Design of beams under an eccentric load
• An eccentric load P can be replaced by a centric load P and a
couple M = Pe.
P
e • Here the discussion is limited to small value of P. Higher value
P of P will result in nonlinear deformation of the beam and the
M = Pe discussion will be made in Lecture 6.

• For homogeneous cross-section, normal stresses

can be found from superposing the stresses due to
the centric load and couple,

  centric   bending
P Mzmax
 centric 
P
Mc  max   
A A I
bending I • For non-homogeneous cross-section , the problem
becomes more complicated. The example 4 will
demonstrate the solution procedure.

N N

x
e e = +
Ne Ne
N N

31
Example 3 (please study this example by yourself): The
Solutions
rectangular block of negligible weight in Fig. a is subjected to
a vertical force of 40 kN, which is applied to its corner. • For uniform normal-stress distribution the stress is
Determine the largest tensile normal stress and compressive
normal stress acting on a section through ABCD. P  40
   125 kPa Compressive
A 0.80.4

• For 8 kN.m, the maximum bending stress is

M x ymax 80.2
 max    375 kPa
12 0.80.4 
3
Ix 1

• For 16 kN.m, the maximum bending stress is

M y xmax 160.4
 max    375 kPa
Iy 1
12 0.4 0.8 3

• By inspection the normal compressive stress at point

EXAMPLE 5 (cont)
C is the largest since each loading creates a
compressive stress there

c  125  375  375  875 kPa

• By inspection the normal tensile stress at point A
is the largest since each bending loading creates
a tensile stress there

t  125  375  375  625 kPa

32
Example 4: A simply supported composite beam of 3 m Solution
long is under a compressive force P=100 kN at the
geometry centre of the cross-section at its right end. The 1) The location of the neutral axis y
beam is constructed of a wood member, 100 mm wide
by 150 mm deep, reinforced on its lower side by a steel Ew Aw z w  Es As z s
z
plate 8 mm thick and 100 mm wide. The moduli of Ew Aw  Es As
elasticity are Ew = 10 GPa for the wood and Es = 210 GPa
10(109 )(150)(100)(75  8)  210(109 )(8)(100)(4)
for the steel. Find the maximum stresses at the top of 
the beam, the bottom of the beam and the interface 10(109 )(150)(100)  210(109 )(8)(100)
between the wood and steel, respectively.  41.264 mm
z
2) Second moments of areas about the neutral axis

100 1503
e Iw   (75  8  41.264) 2  (150 100)
P 12
z  5.4253 107 mm 4  5.4253 10 5 m 4

P 100  83
Is   (41.264  4) 2  (100  8)
12
 1.1152 106 mm 4  1.1152 10 6 m 4

3) Bending stiffness of the entire cross section

The cross-sectional area of the beam is
D  Ew I w  Es I s
A  (0.1)(0.15  0.008)  0.0158 m2
 10(109 )  5.4253 10 5  210(109 ) 1.1152 106
 P  100,000
Is the stress equal to     6.329 MPa ?  7.7672 105 N  m 2
A 0.0158

33
 z 3) The eccentricity of loading point (distance
between the loading point and the neutral axis) :

The loading point is (150+8)/2 mm from the

P e bottom of the cross-section. Thus, the
z eccentricity of loading point is
e = (150+8)/2  41.264 = 37.736 mm
P
Loading point

4) The bending moment is:

M   Pe  100,000  0.037736
 3773.6 N  m

34
5) Stress due to the bending moment

The stress in the wood on the top of the wood (point A) q= 3.0 kN/m
A
. M

150 mm
Ew MzA 10(109 )  (3773.6)  [(0.158  0.041264)]
A  
D
  5.67 MPa Compressive
7.7672(105 ) 41.264 mm B
C
. 8 mm
3m

The stress in the wood at the interface of the wood and steel (point B+)

Ew MzB 10(109 )  (3773.6)  (0.041264  0.008)

B  
D 7.7672(105 )
 1.62 MPa Tensile

The stress in the steel at the interface of the wood and steel (point B-)

Es MzB 210(109 )  (3373.6)  (0.041264  0.008)

B   5
 33.9 MPa Tensile
D 7.7672(10 )

The stress in the steel on the bottom of steel (point C)

Es MzC 210(109 )  (3773.6)  (0.041264)

C   5
 42.1 MPa Tensile
D 7.7672(10 )
35
6) Compressive stress due to the compressive load itself (see example 1
of Lecture 2, or tutorial question 1 of Week 5)

In the wood q= 3.0 kN/m

A
.
E1 10(109 )(100,000) 150 mm
1  P  3.14 MPa
E1 A1  E2 A2 10(109 )(0.1)(0.15)  210(109 )(0.1)(0.008)

In the steel
B
C
. 8 mm
3m P
E2 210(109 )100,000
2  P  66.0 MPa
E1 A1  E2 A2 10(109 )(0.1)(0.15)  210(109 )(0.1)(0.008)

7) Finally, stresses due to combined bending moment and compressive load are 5)+6):

In the wood (point A) Awood  5.67  3.14  8.81 MPa Compressive

In the wood (point B) Bwood  1.62  3.14  1.52 MPa Compressive

In the steel (point B) Bsteel  33.9  66.0  32.1 MPa Compressive

In the wood (point C) Csteel  42.1  66.0  23.9 MPa Compressive

* These are the actual stresses in the composite beam due to eccentric
loading. You can see if the load is applied not at the neutral axis (point O)
of the cross-section, the stresses will be quite different. 36
 L6 Bending deformation of beams made of homogeneous materials and composite
materials

• Force equilibrium of beam

• Deformation (deflection) of beam - elastic curve- static loading

• Analytical solutions of the deflection of beam for typical loading and boundary conditions

• Deformation and stress of beams under an eccentric load

• Curved beam

•Effective of transverse shear stress on the deflection (displacement) of the beam

1
1. Equilibrium equation of the beam
* Small deflections (small slope), also noticed that
z, w

dx Qx 
Q x
dx
F z 0
y x
* Summing forces in the z direction yields the
x equation:
 Q 
 Qx  x dx   Qx  q( x)dx  0
Mx
M x
Qx Mx  dx  x 
x
Qx
q(x)  q 0 (1)
x
* Summing moments about the y-axis yields the
equation: * From Eqs (1) and (2), one obtains the equation:

 M x 
Mx  dx   M x  Qx dx  0
 x  2M x
q 0 (3)
 Qx 
M x
0 (2)
x 2
x

Equilibrium of beam in terms of displacements

• Putting this into Eq. (3) yields:
As mentioned, the x axis always on the neutral plane
of the beam (x-y is the neutral plane of the beam 4w
D 4 q 0 (5)
x
D: bending stiffness of the beam
 w
2
M x  D
x 2 D   Ez 2dA (4)
A
2
2. Engineers’ theory of bending

General equations: for composite material beams :

2w  2 w MEz 4w

M  D 2  x   Ez 2  D 4 q 0
x x D x

Beams made of one material (homogeneous cross-section):

2w  2 w Mz 4w
M   EI 2  x   Ez 2   EI 4  q  0
x x I x

D: bending stiffness of the beam I   z 2dA (6)

A

D   Ez 2dA (4) Second moment (or moment of inertia) of

A the cross section about the y axis.

3
3. Governing equation for bending beam
Q
In view of the relationship between lateral load (Q) moment (M) and
distributed loading q, we have obtained the governing equations for beam
bending
M
M
M Q  w 4
 w
4 Q
Q ,  q EI 4  q (i) D 4 q (ii)
x x x x
q(x)
This fourth-order ordinary differential equations (i) for homogeneous beam, and (ii) for
composite material beam are to be solved together with the boundary conditions (BCs).

For simply supported end or roller supported end:

Simply supported end (pin ended)
2w 2w
w  0 , M   D 2  0 or w  0, 2  0
x x

w
For clamped (fixed) end: w0 , 0
x Q0

For free end

2w 3w
M   D 2  M 0 , Q   D 3  Q0 M0
x x
fixed end Free end
 w
2
M0 w 3
Q0
or   ,  
x 2 D x 3 D
4
4. Elastic curve- static loading, deflection, bending moment
* The static bending equilibrium equation of the beam
and transverse shear force of the beam
is
4.1. Fixed-free (cantilever) beam
4w
A cantilever beam of length L is subjected to constant lateral D 4 q (ii)
loading q. (The beam is perfectly straight, its cross-section x
properties are constant along the x-axis) .
z L * The general solution of this equation is

w(x)
w( x) 
q
24 D

x 4  Ax 3  Bx 2  Cx  F  (A)
x

q
* The deflection of the beam at x is denoted as w(x).

1. The constants A, B, C and F should be determined from the BCs:

w
At x = 0: w  0 , 0 4. The bending moment is
x
2w
2w 3w M   D 2   x 2  2 Lx  L2 
q
At x = L:  0, 0 x
x 2 x 3 2

2. This four BCs can be used to obtain 5. The shear transverse force is
A  4L , B  6L2 , C  0 , F  0 3w
Q   D 3  q x  L 
x
3. Therefore, the deflection function of the beam is

w( x) 
q
24 D

x 4  4 Lx3  6 L2 x 2  (7)
5
 4.2. Simply supported (S-S) beam
* The static bending equilibrium equation of the beam
is
A simply-supported beam of length L is subjected to
4w
constant lateral loading q. (The beam is perfectly straight, D 4 q (ii)
its cross-section properties are constant along the x-axis) x
z L * The general solution of this equation is
w(x)
 
x q
w( x)  x 4  Ax 3  Bx 2  Cx  F (A)
24 D
q

2w
1. The boundary conditions are w(x)=0 and 2  0 at x=0 and x=L.
x

2. This four BCs can be used to obtain

4. The bending moment is
A  2L , B  0 , C  L3 , F  0
2w
M   D 2   x  L 
qx
3. Therefore, the deflection function of the beam is x 2

w( x) 
qx 3
24 D

x  2 Lx 2  L3  (8) 5. The shear transverse force is
3w
Q   D 3   2 x  L 
q
The deflection at the middle of the beam is maximum and is x 2
5qL4 qL4
 max  w( L / 2)   0.01302
384 D D

6
4.3. Bending of fixed-pinned beam
* The static bending equilibrium equation of the beam
is
A fixed-simply supported beam of length L is subjected to
constant lateral loading q. (The beam is perfectly straight,
4w
its cross-section properties are constant along the x-axis) D 4 q (ii)
x
z L
* The general solution of this equation is
w(x)

 
x q
w( x)  x 4  Ax 3  Bx 2  Cx  F (A)
24 D
q

1. The constants A, B, C and F should be determined from 4. The bending moment is

the BCs:
w 2w
w = 0 and
x
 0 at x = 0, M  D 2  
x
q
24 D

12 x 2  15Lx  3L2 
and
2w
w = 0 and  0 at x=L. 5. The shear transverse force is
x 2

3w
5 3 Q  D 3  
q
8x  5L
2. These yield A  L , B   L2 , C  0 , F  0 x 8D
2 2
?. Can you find the maximum bending moment
3. The deflection function of the beam is and the maximum shear force in the bean? The
maximum bending moment is at Q=0 (you know
q  4 5 3 3 2 2 why), or x  5L / 8 in this problem and is
w( x)   x  Lx  L x  (9)
24 D  2 2  5L 0.0703qL2
M max  M( ) 
8 D
7
 4.4. Simply supported (S-S) beam
* Due to symmetry, only the left half of the
A simply-supported beam of length L carries a point bean is examined.
force Q at its middle. (The beam is perfectly straight, its
cross-section properties are constant along the x-axis) z L/2
w(x)
z x
L
w(x) P/2
x

P * Note that q=0 for this problem. Thus, the

general solution of the deflection curve is
1. The boundary conditions are
w( x)  Ax 3  Bx 2  Cx  F (B)
2w
at x=0: w(x)=0 and 0
x 2
w P 3w P
at x=L/2:  0 and Q  or  D 
x 2 x 3 2
4. The bending moment is
2. These four BCs can be used to obtain
P PL2  2 w Px
A , B0 , C  , F 0 M  D 2 
12 D 16 D x 2
3. Therefore, the deflection function of the beam is
5. The transverse shear force is
P x 3
L x 2
w( x)      (10) 3w P
D  12 16  Q  D 3 
x 2
The deflection at the middle of the beam is maximum and is
PL3
 max  w( L / 2) 
48D
8
4.5. Fixed-free (cantilever) beam under a concentrated load at its free end

A cantilever beam of length L is subjected to point load Q.

(The beam is perfectly straight, its cross-section properties
are constant along the x-axis) .
The general solution of the deflection curve is
z L
w(x) w( x)  Ax 3  Bx 2  Cx  F (B)
x
P

* The deflection of the beam at x is denoted as w(x).

1. The constants A, B, C and F should be determined from the BCs: 4. The bending moment is
w 2w
At x = 0: w  0 , 0 M   D 2  P ( x  L)
x x
2w 3w
At x = L: 2  0 , Q  P or  D P 5. The shear transverse force is
x x 3
2. This four BCs can be used to obtain 3w
Q  D 3  P
P PL x
A , B , C  0, F  0
6D 2D
3. Therefore, the deflection function of the beam is The deflection at the free end of the beam is
maximum and is
P  x 3 Lx 2 
w( x)      PL3
2 
(12)  max  w( L) 
D 6 3D
9
 Example 1: A simply supported composite beam of 3 m long z
carries a concentrate load P = 9 kN at the middle of the beam. The
beam is constructed of a wood member, 100 mm wide by 150 mm
deep, reinforced on its lower side by a steel plate, 8 mm thick and
100 mm wide. Find the maximum deflection of the beam due to y
the point load if the moduli of elasticity are Ew = 10 GPa for the
wood and Es = 210 GPa for the steel. Also determine the
maximum stresses in the wood and steel.

Solution z
The location of the neutral axis y and the bending moment of the
cross-section have been obtained in Example 4 of Lecture 5. These
are:

z  41.264 mm , D  7.7672 105 N  m2 y

Thus, the maximum deflection of the beam is z

PL3 9,000(3)3
   0.006518 m  6.518 mm
48D 48  7.7672 105
For bending stress calculation: M max  4500 1.5  6,750 N  m

Ew M max [(0.15  0.008  0.041264)] Es M max (0.041264)

 wood   steel 
D D You will have obtained the solutions
10(109 )(6,750)(0.116736) 210(109 )(6,750)(0.041264) of this problem in your Computer
  Lab class. The analytical results of
7.7672 105 7.7672 105 here can be used to validate your
 10.14(106 ) Pa  10.14 M Pa  75.3(106 ) Pa  75.3 M Pa Ansys solutions.
(compressive) (Tensile)
10
5. Deformation of beams under an eccentric load (homogeneous beam only)

P
L
P P e
x NA
e e

z (a) Side view of the beam

z
M0=Pe M0=Pe
(b) Cross-sectional view of the
x
beam
P P
w(x)
z
(c) Equivalent problem

• In previous examples, we considered ideal beams, that is beams that are initially perfectly straight and whose
compressive load is applied through the centroid of the cross-section of the member. Such ideal conditions never
exist in reality, since perfectly straight structural members cannot be fabricated, and since the point of application
of the load seldom, if not never, lies exactly at the centroid of the cross-section.
• Figs. (a) and (b) show a beam that is compressed by a compressive load P that is applied with a small
eccentricity e measured from the Neutral Axis of the beam.
• The eccentric axial load is equivalent to a centric load P and a couple of moment M0=Pe (Fig. (c)).
• The moments exist from the instant the load is first applied therefore the beam begins to deflect at the beginning
of loading.
• The deflection increases steadily as the load increases.

11
We study the equivalent problem, (c). From the FBD of Fig. (d), we see that the bending
moment in the beam is
Both ends of the beam are simply supported.
The relevant boundary conditions are: M ( x)  Pw  M 0 (13)
w(0)  0 , w( L)  0 Relationship between the deflection and bending
moment has been established (Slide #3)
M0=Pe 2w
M   EI 2 (14)
L P x

w(x) Substitution of Eq. (14) into Eq. (13) yields

d2w
x EI  Pw  Pe  0 (15)
z M(x) dx 2

The general solution of the differential equation (15) is

(d) A free-body diagram (FBD) for beam segment Lx
w( x)  A sin x  B cos x  e (16)
where
 P 2 EI
 , Pcr  2 (17)
L Pcr L

The constant A and B are determined from the

boundary conditions at x = 0 and x = L as

B=e, A= e tan(L/2) (18)

Therefore, the equation of the deflection curve is

 L 
w( x)  e tan sin x  cos x  1 (19)
 2  12
 The maximum deflection of the beam (denoted
as ) is at the middle of the beam and is obtained
P from Eq. (19) as

Pcr e=0  
 
e=0  1 
e=e1   e  1 (20)
 
e=e2  cos  P  
 2 EI   2 Pcr  
Pcr  2
L
e2>e1>0 The deflection becomes infinitely large
when P approaches Pcr .

0

• The deflection increases as P increases, but the relationship is nonlinear. Therefore,

we cannot use the principle of superposition for calculating deflections for more than
one load. For example, the deflection due to an axial load 2P does not equal twice the
deflection caused by an axial load P.

•As the load P approaches the critical load, the deflection  increases without limit,
even for extremely small e (i.e., the load is applied at the neutral plane of the beam)

13
The maximum bending moment of the
beam is at x=L/2 and is:
Mmax
Pe
M max  P(e  )  (21)
 P 
cos 

 2 Pcr 
• When P is small, the maximum moment equals to Pe,
and the problem is reduced to that of Lecture 5.

• As P increases, the bending moment increases in a

Pe
nonlinear manner.

• Theoretically, the bending moment becomes infinite as 0

Pcr P
P approaches the critical load.

In beam designing, always let the centroid of the cross-section aligns with the loading
point. This is to avoid the extra bending moment/stress by eccentric load.

14
 Example 2: A steel bar having a square cross section (50 mm  50 mm) and length
L = 2.0 m is pinned supported at its ends (simply supported beam) and is
compressed by an axial load that has a resultant P = 30 kN acting at the midpoint
of one side of the cross section (see figure). The modulus of elasticity E is equal to
210 GPa and that the ends of the beam are pinned. Calculate the maximum
deflection , the maximum bending moment Mmax and the maximum bending
compressive stress max.

Solution:
0.05  0.0053
The moment of inertia and the area of the beam are A  0.05  0.05  0.0025 mm 2 , I   5.208 10- 7 m 4
12
The critical load is calculated from Eq. (17) as

2 EI 2  210(109 )  5.208 107

Pcr  2
 2
 2.699 105 N
L 2

The maximum deflection  is found from Eq. (20) as The maximum bending moment Mmax is found from
Eq. (21) as
   
    Pe 30000  0.025
  M max    866.1 N  m
 1  25  1  3.869 mm
1 1
  e  P   30000 
   30000  cos  cos 
 cos  P    cos    
 2 2.699 10
5
  2 Pcr     2 2.699 105   
 2 Pcr 

Note: use Radian for cosine calculation

zmax M max P 0.0025  866.1 30,000
The maximum compressive stress max is max       53.58 MPa
I A 5.208 10-7 0.0025

zmax M max P 0.025  866.1 30,000

The maximum tensile stress is      29.58 MPa
I A 5.208 107 0.0025
15
6. Curved beam (imperfection in beams) - homogeneous beam cross-sectional only

L The previous section indicated how the behavior of a straight

beam is affected if the load is applied eccentrically rather than at
the centroid of the cross-section of the member. The behavior of a
x (x) beam is also affected by any initial lack of straightness of the axis
of the beam, or initial imperfection of the beam. The ideal pin-
(a) The beam is not ideal straight. Instead, it is ended (simply-supported) of such beam is shown in Fig. a that
curved having original shape 0 varying along the has an initial deflection 0(x).
axis of the beam (the x axis)
Although 0(x) is usually small, its exact functional form differs
from beam to beam and is unknown. However, we can get an idea
of the effect of initial deflection by assuming that v0(x) can be
w(x)
replaced by
P   0sin(x/L) (21)
x (x) max
From the free-body diagram of Fig. (c) we get,

(b) Curved beam under compressive loading M ( x)  P[w( x)  ( x)] (22)

where w(x) is the deflection (transverse displacement) of the beam

caused by the load P. Combining Eqs. (21) and (22) with the
M(x) 2w
P moment-curvature equation, M   EI 2 gives the differential
w(x)+(x) x
P x equation:
x 2w  x 
EI 2  Pw   P0 sin  (23)
x  L
(c) A free body diagram
16
 The solution to Eq. (23) with the boundary conditions w(0)=w(L)=0 is
max
0  x 
w( x)  sin  (24)
1   L 
where

P 2 EI
 , Pcr  2 (25)
Pcr L

0 From Eq. (24) we can determine the maximum deflection, max, and the
Pcr maximum bending moment, Mmax, as follows:
P
(d) Load-deflection diagram L 
 max  w   0  0 (26)
2 1 

P0
Mmax M max  P max  (27)
1 
Then,
P M max z
x    (28)
A I

As can be seen from Eqs. (26) and (27) and Fig.s (c) and (d), the
deflection and moment are all nonlinear in the load P.
P0
Pcr
P
(d) Load-moment diagram 17
7: Effect of transverse shear on the deformation of the beam

7.1 Shear in a straight beam

• Transverse shear stress always has its associated longitudinal shear stress acting along
longitudinal planes of the beam.

Transverse shear
stress
Longitudinal
shear stress  The vertical shear stresses must vanish at the top and
bottom surfaces; in other words, =0 where y=h/2.

The vertical shear stresses are maximum at the neutral plane

x of the beam.

V
y

18
7.2 Distribution of transverse shear stress

2
6V h
z
Rectangular cross section : (y) = 3
bh
[( )  z ]
2
2

The maximum shear stress is at the middle of the cross-

section. This is, the y=0 plane
y

3V
 max =
2A

in which A=bh is the cross-sectional area. Thus, the maximum

shear stress in a beam of rectangular cross section is 50% larger
than the average shear stress V/A.

So far, we have limited our discussion to the situation that the

normal stresses in the beam is significantly higher than the shear
Distribution of shear stresses in a beam of stress. If the shear stress is obvious, the deformation of the beam will
rectangular cross section: (a) cross section of be different from that predicted by Engineer’s Theory of Bending.
beam, and (b) diagram showing the parabolic We will demonstrate this through an example – next page.
distribution of shear stresses over the height
of the beam.

19
Example 3：A cantilever beam under a point force at its free P
end. Find the deflection of the beam at its free end.
L

Solution：The deflection of the beam is due to two parts:


1. P makes a bending moment M=P(xL) over the length
M
span of the beam. The deflection due to this bending moment
PL3
is predicted by Engineer’s Theory of Bending as 1 
3EI 1
2. P makes shear stress/shear strains in the beam. The L
shearing deformation will result in the deflection of the beam.
This deflection can be studied as: P
L
 Broken line: deformation of the beam due to shear stress

The average shear stress in the beam is  xy 
P
， 2
A
where A is the cross-sectional area of the beam
P
 The shear strain is  xy 
GA
 The definition of the shear strain is such that:  xy  
PL
 Accordingly, displacement of the beam due to shear stress is  2  L 
GA
A constant shear strain across the cross-section is assumed !
PL
So a shear correction factor k is introduced: 2 
kGA 20
 L P

The real deformation of the beam is



PL3 PL PL3 3EI

  actual  1   2    (1  2
)
3EI kGA 3EI kGAL

b
Normal stress to shear stress ratio:

PL (h / 2) 6 P L 3P
max   , max  h
I A h 2A
 max L
4
 max h
• For slender beams with large L/h values, the bending
normal stress is significantly higher than the bending
shear stress

21
 actual
1 • Calculation is made for =0.25 and k=0.85;
• L/h= 4; difference is 4.6%;
• L/h=10; difference is 0.74%
b

h • L/h > 10, “long beam”, “slender beam”,

Beam theory can be used.
• 4 < L/h < 10, “short beam”, “deep beam”
transverse shear deformation should be
considered (e.g., Modified beam theory)
• L/h < 4, Beam theory can not be used, three
L/h dimensional elasticity analysis should be
carried out.

If the beam cross section is rectangular

PL3  E h 
2

actual  1    
3EI  4 kG  L  

actual
2
E h
 1  
1 4kG  L 
22
Shear correction factor
Choice of shear correction factor k largely depended on experience. There is no
universal value for k

for rectangular cross section *

for circular cross-section *

Many other values were used, e.g. , k=1, k=5/6, k=9/10, k=2, k=0.8.

* Stephen Timoshenko, James M. Gere. Mechanics of Materials. Van Nostrand

Reinhold Co., 1972. Pages 207.

23
Chapter 11. Un-symmetric bending • Cross-section of the beam is
symmetric in z (i.e., everything
11.1 Introduction on z =  and z = + are
In Lectures 5 and 6 and also in Mechanics of Materials,  identical).

we saw that when a beam (with a symmetrical cross- • Beam bends in the x-y plane.
section in the plane of the page, Figs a and b) has a
y
moment M applied to each end (as a vector, M is y
perpendicular to the page, and to the axis of symmetry, x
Fig. c). o z
• The beam will bend into an arc of radius , in the
plane of symmetry (here the plane of the page).
•The top face will be in compression and the bottom (b) Cross section
(a) Profile view
face will be in tension. A neutral surface with zero stress
will exist between the top and the bottom surfaces.
•The neutral surface will be perpendicular to the plane
of the page. The vector of M will be parallel to the
neutral surface.
The strain/stress varies linearly with distance y from the (c) The moment (couple) may be represented by
neutral surface and is related to the bending moment by a vector in the form of a double-headed arrow.
the Engineers’ Theory of Bending: The arrow is perpendicular to the plane
containing the couple. The direction of the
y My My Ey
x    , x    (11.1) moment is indicated by the right-hand rule for
 EI I  moment vectors, namely, using your right hand,
let your fingers curl in the direction of the
where  is the radius of curvature of bending. moment, and then your thumb will point in the
direction of the vector.
1
11.2 Moment arbitrarily applied on doubly symmetric beams

We now find what happens when the applied moment is not parallel to the axis
of symmetry of the beam.

If the beam’s cross-section has two axes of symmetry, we can split the
moment into two components perpendicular to each of the axes of symmetry,
and compute the bending stress due to each component separately. We can
treat the application of My and Mz as two separate problems.

M y  M sin  , M z  M cos  (11.2)

For each problem, the moment vector is perpendicular to an axis of

symmetry, so we can use the Engineer Theory of Bending to calculate the
resulting stress distributions.
M yz Mzy
x    (11.3)
Iy Iz
in which Iy and Iz are the moments of inertia of the cross-section area with
respect to the y and z axes, respectively. Using this equation, we can find the
normal stress at any point in the cross-section by substituting the appropriate
algebraic values of the moments and the coordinates.

Note that the moment pointing in the positive y direction will cause a tensile
stress at a point with a positive z value, hence the plus sign in Eq. (11.3).
However, a moment pointing in the positive z direction will cause a
compressive stress at a positive y value, hence the minus sign in Eq. (11.3).
2
 Neutral axis
The equation of the neutral axis can be determined by equating the normal
stress x (Eq. 11.3) to zero:
M yz Mzy
 0 (11.4)
Iy Iz
This equation shows that the neutral axis NA is a straight line passing through the
centroid C (right figure) . The angle  between the neutral axis and the z axis is
determined as follows:
y M yIz
tan    (11.5 )
z MzIy
Depending upon the magnitudes and directions of the bending moments, the angle  may
vary from 90 to +90. Knowing the orientation of the neutral axis is useful when
determining the points in the cross section where the normal stresses are the largest.

Since the stresses vary linearly with distance from the neutral axis, the maximum stresses
occur at points located farthest from the neutral axis.

Special cases are:

•If the only applied moment is Mz, z will be the neutral axis.

•If the only applied moment is My, y will be the neutral axis.

We have studied the stresses for these special cases in previous chapters.

3
Relationship between the neutral axis and the loads
We wish to determine the orientation of the neutral axis relative to the angle
of applied loads acting on the beam. The ratio of moments can be found from
Eq. (11.2) as tan = My /Mz , which shows that the resultant moment vector M
makes an angle  from the z axis (right figure). Consequently, the angle 
between the neutral axis NA and the z axis is obtained from Eq. (11.5) as:

y M yIz Iz
tan     tan  (11.6)
z MzIy Iy

which shows that the angle  is generally not equal to the angle . Thus, the neutral axis
is not parallel to the direction of the moment load vector, except in below special cases:
1. When the load lies in the xy plane ( = 0 or  =180), which means that the z axis is the
neutral axis.
2. When the load lies in the xz plane ( = 90 or  = 90), which means that the y axis is
the neutral axis.
3. When the principal moments of the inertia are equal (that is, when Iy = Iz ). In this case,
the neutral axis is always parallel to the direction of the moment load vector.

4
Example 1: A 180 Nm couple is applied to a wooden beam of rectangular
cross-section 40 mm  90 mm, in a plane forming an angle of 30 with the
vertical. Determine the maximum bending stress in the beam and locate the
neutral axis.
Solution
Loads and bending moments. Split the couple into y and z components.

M y  180  sin 30  90 N.m ,

M z  180  cos 30  155.9 N.m

Moments of inertia. The moments of inertia of the cross section with respect to
the y and z axes are:
0.09(0.043 )
Iy   0.48(10 6 ) m 4 ,
12
0.04(0.093 )
Iz   2.43(10 6 ) m 4
12
Bending stresses. The stress at any point in the cross section can be obtained from Eq. (11.3) by substituting the
coordinates y and z of the point. From the orientation of the cross section and the direction of the bending moments
(right figure), it is apparent that the maximum compressive stress occurs at point A (where y = 45 mm and z = 20 mm)
and the maximum tensile stress occurs at point B (where y =  45 mm and z = 20 mm). Substituting these coordinates
into Eq. (11.3) and then simplifying, we obtain expressions for the maximum and minimum stresses in the beam:

90(0.02) 155.9(0.045) 90(0.02) 155.9(0.045)

A    6.64 MPa ,     6.64 MPa
0.48(106 ) 2.43(106 ) 0.48(106 ) 2.43(106 )
B

5
Neutral axis. In addition to finding the stresses in the beam, it is
often useful to locate the neutral axis. The equation of this line is
obtained from Eq. (11.4) as:

90 z 155.9 y
 0
0.48(106 ) 2.43(106 )

Or

y  2.9225z

The neutral axis is shown in the right figure as line NA.

Example2: A cantilever beam is made of steel

(Y = 250 MPa ) with dimensions shown. It
carries a vertical force of 20 kN. To maximize the
load it can carry, the beam is arranged so that its
long edge is vertical. If the force is accidentally
applied at an angle , the beam will bend to the
side. Calculate the minimum value for the angle 
at which the beam will yield.

6
Solution
1. Loads and bending moments. Split the couple into 2. Moments of inertia. The moments of inertia of the
y and z components. cross section with respect to the y and z axes are:

M y  20,000  2  sin  , 0.2(0.053 )

Iy   2.0833(10 6 ) m 4 ,
M z  20,000  2  cos  12
0.05(0.23 )
Iz   33.333(10 6 ) m 4
12

3. Bending stresses. The stress at any point in the cross section can be obtained from Eq. (11.3) by substituting the
coordinates y and z of the point. From the orientation of the cross section and the direction of the bending moments
(right figure), it is apparent that the maximum stress occurs at y = 100 mm and z = 25 mm). Substituting these
coordinates into Eq. (11.3) and then simplifying, we obtain expressions for the maximum and minimum stresses in
the beam:

 20,000  2  sin (0.025)  20,000  2  cos (0.1)

   480(10 6
) sin   120(10 6
) cos 
2.0833(106 ) 33.333(106 )
Maximum angle. Comparing the above stress against the yielding stress of the beam gives

480.01(106 ) sin   120(106 ) cos   250(106 )    16.3

Therefore, if  is greater than 16.3, the beam yields.

7
11.3 Non-symmetrical bending – general case

Consider the general case of a prismatic beam subjected to bending-moment components My and Mz , as shown, when
the x, y, z axes pass through the centroid of the cross section. If the material is linear-elastic, the normal stress in the
beam is a linear function of position such that  = a + by + cz. The constants a , b, and c can be determined from the
equilibrium conditions:

 dA  0, M
A
y   zdA, M z   ( y)dA
A A

These yield:

a  dA  b  ydA  c  zdA  0
A A A

M y  a  zdA  b  yzdA  c  z 2 dA (a)

A A A

M z  a  ydA  b  y 2 dA  c  yzdA
A A A

Note that the x axis passes through the centroid of the cross-section. Thus,

A
ydA  0,  zdA  0
A

Accordingly, Eq. (a) gives

a0
M y  b  yzdA  c  z 2 dA (b)
A A

M z   b  y 2 dA  c  yzdA
A A

8
We know that I y   z dA, I z   y dA are the moments of inertia about the y axis and the z axis,
2 2
A A

respectively. We also introduce the product of inertia of an area with reference to a yz reference
frame in the plane of the area:

I yz   yzdA (11.7)
A

Now the constants b and c can be solved from Eq. (b), in terms of My and Mz , to give

I yz M y  I y M z I z M y  I yz M z
b ,c 
I y I z  I yz2 I y I z  I yz2

After substituting a = 0, b and c into  = a + by + cz, we find that the normal stress
can be determined from the equations:

 y ( I yz M y  I y M z ) z ( I z M y  I yz M z )
  (11.8)
I y I z  I yz2 I y I z  I yz2

If y or z is one of the centroidal principal axes, the product of inertia will be zero. In this case,
Eq. (11.8) reduces to Eq. (11.3), which is a special case when the y and z axes are parallel to the
centroidal principal axes.

9
 Example 3: An unequal-leg angle section has the dimensions shown
in Fig. a. At this cross section the moment is M = 10 kN.m and is
oriented parallel to the short leg of the angle, as shown.
(a) Determine the orientation of the neutral axis of the cross section,
and show this orientation on a sketch.
(b) Determine the maximum tensile stress and the maximum
compressive stress on the cross section.
Solution
Step 1: Identify the moments on the cross section

M y  0, M z  10 106 N.mm

Fig. a

Step 2: Find the location of centroid (Fig. b).

200  25 100  (100  25)  25 12.5

y0   76.14 mm
200  25  (100  25)  25

75  25  62.5  200  25 12.5

z0   26.14 mm
200  25  (100  25)  25

Fig. b
10
 Step 3: Find moments of inertia and the product of inertia (Fig. c).

( 26.14 25) 26.14

I y   z 2dA   25z 2dz   200 z 2dz  4.5484 106 mm 4
A (100 26.14) ( 26.14 25)

 ( 76.14 25) ( 20076.14)

I z   y dA  
2
75 y dy  
2
25 y 2dy  27.205 106 mm 4
A 76.14 76.14

( 26.14 25)  ( 76.14 25) 26.14 ( 20076.14)

I yz   z ydydz   z ydydz
(100 26.14) 76.14 ( 26.14 25) 76.14

1.14 51.14 26.14 123.86

Fig. c
 z ydydz   z ydydz
73.86 76.14 1.14 76.14

 4.3387 106  1.6273 106  5.9660 106 mm4

Step 4: Apply Eq. (11.8) to obtain the expression of the normal stress

I yz z  I y y 45.484 y  59.66 z
 Mz  106
I y I z  I yz2 88.146

Step 5: Determine the orientation of the neutral axis

of the cross section based on  = 0
45.484 y  59.66 z
The correct orientation of the NA is shown in Fig. d.
Fig. d 11
 Step 6: Identify the points with maximum stresses (Fig. e)

Points A and B are the two points that are farthest from the
neutral axis. To compute the stresses at these points, we need
their coordinates. From the figures, we get

y A  76.14 , z A  26.14
yB  (200  76.14)  123.86 , zB  (26.14  25)  1.14.

So

45.484(76.14)  59.66(26.14)
Fig. e A  106  56.98 MPa
88.146

45.484(123.86)  59.66(1.14)
B  106  63.14 MPa
88.146

Note: The maximum/minimum stress occurs at a corner point

of the cross section (Fig. f). Therefore, you can also calculate
the stresses for all corner points (i.e, A, B, E, F, G of the left
figure) and compare these to determine the largest tensile stress
and compressive stress. In this method, there is no need to find
the neutral axis.
Fig. f
12
Example 4: The moment M acts in a vertical plane and is applied to a
beam oriented as shown. Determine (a) the orientation of the principal
axes, and (b) the maximum tensile stress in the beam.
Solution
1. The cross-section has an axis of symmetry. It is the y axis of
Fig. b. Then, the y and z are principal axes and the product of
inertia about y and z is zero ( i.e., Iyz = 0).
2. By such definition of the coordinate system, the
moments about the y and z axes are:
M y  400  sin 30  200 N.m,
M z  400  cos 30  346.41 N.m Fig. a.

3. Calculate the location of the centroid and find the

y
moments of inertia, Iy and Iz. Here they have been given. Iy=281,000 mm4
4. Calculate the bending stress (using Eq. 11.3) at each Iz=176,900 mm4
corner and compare these to determine the maximum.
200,000(25) 346,410(25)
A    31.16 MPa
281,000 176,900
200,000(25) 346,410(25)
B    66.75 MPa
281,000 176,900
z
200,000(25) 346,410(18.57)
D    18.57 MPa
281,000 176,900
200,000(25) 346,410(18.57)
E    54.16 MPa
281,000 176,900 Fig. b. 13
Chapter 12. Shear center
12.1 Introduction

In Mechanics of Materials, we
considered combined bending and
torsion, as would be experienced,
for example, by a beam that is =
loaded parallel to, but not along,
an axis of symmetry, like the
double symmetric box beam in
Figs. a and b.

(a) Load parallel to the plane (b) Equivalent

of symmetry but not passing loading plus
through the centroid torque

In this Chapter, we will see that for an un-

symmetric beam, even if the load passes
through the centroid of the cross-section, the
beam twists. See right figure as an example.

1
Given any cross-sectional configuration, one point may be found in the plane of the cross section through which the
resultant of the transverse shearing stresses passes. A transverse load applied on the beam must act through this point,
called the shear center or flexural center, if no twisting is to occur. The center of shear is sometimes defined as the
point in the end section of a cantilever beam at which an applied load results in bending only. When the load does not
act through the shear center, in addition to bending, a twisting action results.

The location of the shear center is independent of the direction and magnitude of the transverse forces. For singly
symmetrical sections, the shear center lies on the axis of symmetry, while for a beam with two axes of symmetry,
the shear center coincides with their point of intersection (also the centroid). It is not necessary, in general, for the
shear center to lie on a principal axis, and it may be located outside the cross-section of the beam.

The shear center of a cross-section In this cross-section, the shear forces in In this picture, the horizontal
is defined as that point in the cross- the left and right parts occur in equal but shear stresses in the flanges
section through which the resultant opposite pairs. This means that the shear cause a torque.
shear force V must pass if the beam forces in the cross-section do not cause
is to bend without twisting. the beam to twist.

Looking the examples of above pictures. The applied shear force (V) cause shear stresses in the cross-section.

2
12.2 Torsion under a shear force
Let us investigate what happened in the right hand picture above.

As an example, we look at a beam with the cross-section shown, that

is carrying a vertical shear force of V (downward, in Newton). V is
applied through the centroid. The moment of inertia about the neutral
axis (here, the z-axis) is
40  (250  7)3 (40  4)  (250  7)3
I   13.54(106 ) mm 4
12 12
We start by looking at the distribution of the shear stress.

At point A, the shear stress is zero.

Fig. a. (Drawing not to scale)
At a distance s from point A, the shear stress is to the left, given by

VQ V  s  (7)(125)
 
It 13.54(106 )(7)
where Q is the first moment based on the shaded area in Fig. b. The shear
stress at B in the flange is
V  38  (7)(125)
B  6
 350.81(103 ) V N/mm 2
13.54(10 )(7)
The shear stress at B in the web is

V  38  (7)(125)
B  6
 613.92(106 ) V N/mm 2
13.54(10 )(4) Fig. b
3
Inside the vertical web, the shear stress is downward, given by

 125  y 
V  (38)  (7)(125)  (4)125  y 
VQ  2 
 
It 13.54(106 )(4)
(i)
32250.3  y 2
 V N/mm
27.08(106 )

where Q is based on the shaded area in Fig. c. The shear stress

accumulates from B to the neutral axis. The maximum shear stress is
found by letting y = 0 in the above equation: Fig. c
 y 0  1190.9(106 )V N/mm 2

The resultant shear force on the web BD is obtained as

125
Vw   tdy
125

where  has been obtained in Eq. (i). Thus Vw is.

 32250.3  y 2
125
2
Vw    6
V N/mm  (4)dy  0.9986 V N
125
 27.08(10 ) 

It can be seen that the resultant shear force in the web

Fig. d
is very close to the total shear force being carried.
4
Since shear stress in the flange AB increases linearly between A and B,
as illustrated in Fig. e, the flange shear force Vflange is.
1
V flange  [350.81(106 )](38)(7)  0.04666 V N
2
The direction of this force is parallel to the positive y axis.

By symmetry, the resultant shear force in the flange DE is the same but
the direction is parallel to the negative z axis, as shown in Fig. e.

Unlike the cross-section with a vertical axis of symmetry, the shear forces
on the flanges AB and DE do not have a mirror-image force to cancel them
Fig. e
out. They form a pair with equal magnitude and opposite directions.

This means that the shear forces on the flanges cause a net couple or torque
on the beam,

T  V flange  h  (0.04666 V N)  250  11.66 V N  mm (ii)

Therefore, if the vertical load is applied passing the centroid of the cross-
section, the entire beam twists, in the direction shown right (Fig. f).

This twisting is as if a torque with the magnitude T we calculated is applied

to the tip of the beam. Fig. f

5
12.3 Shear center

How do we prevent this torsion of the beam?

We could stop it by applying a couple or torque to the tip of the

beam to cancel the torsion caused by the force V.

Alternatively we could change the point of application of the

shear force at the tip, so that it provides both the shear force and
the required torque. Fig. a. If the load is not applied at
the shear center, the beam bends
Where would we have to apply the shear force to achieve this? and twists.

We require that the force and torque is equivalent to force alone.

V e  T

Substituting the expression for T we obtained earlier, Eq. (ii),

V  e  T V flangeh

V flangeh
e (12.1) Fig. b. If the load is applied at the
V
shear center, the beam bends only.
This is the point where we have to apply the shear force so that it cancels
the torsion of the non-symmetric beam. It is called the shear center.
6
Returning to the cross-section we looked at earlier.

The torque caused by the shear forces is

T  11.66 V N  mm

The location of the shear center is

T
e  11.66 mm
V
This is the distance from the shear
center to the center of the web. The Fig . g. (Drawing not to scale.)
distance from the shear center to the left
side of the web is 11.66  2 = 9.66 mm.

7
Example: Find the shear stress and the location of the shear center of the
cross-section shown. Assume the wall is very thin (i.e., t << a and t << b).

Solution
Step 1: Locate the neutral axis.
The cross section has a plane of symmetry. The neutral axis z is shown

Step 2: Calculate I about the neutral axis

 ta 3  
2
 t (2a)3 2ta 2
I  2 I AB  2 I BC  I CD 2
 12
a 
 ta    2t (2b)a 

2
 2a  6b
  2   12 3
Step 3: q and  will be zero at the slit. Calculate Q from
point A, where shear stress is zero.
Segment AB, point 1, at a distance y from the z axis
y ty 2
Q1  QA1  ty   , 0 ya
2 2
ta 2
At point B, y = a and QB is QB  QAB 
2
Segment BC, point 2, at a distance z from the y axis
Q2  QAB 2  QAB  QB 2
ta 2
 QAB  t  (b  z )  a   tab  taz ,  b  z  b
2
ta 2
At point C, z =  b and QC is QC  QABC   2tab
2 8
Segment CD, point 3, at a distance y from the z axis
az
Q3  QABC 3  QABC  QC 3  QABC  t (a  y )
2
ta 2 t
  2tab  (a 2  y 2 )
2 2
t
 ta 2  2tab  y 2 ,  a  y  a
2
ta 2
At point D, y =  a and QD is QD  QABCD   2tab
2
Segment DF, point 4, at a distance z from the y axis

Q4  QABCD 4  QABCD  QD 4
 QABCD  t (b  z )(a)
ta 2 ta 2
  2tab  ta (b  z )   tab  taz ,  b  z  b
2 2
ta 2
At point F, z = b and QF is QF  QABCDF 
2

Segment FG, point 5, at a distance y from the z axis

Q5  QABCDF 5  QABCDF  QF 5
 a  y  ty
2
 QABCDF  t (a  y )   ,a  y 0
 2  2

9
Step 4: Determine the distribution of the shear stresses and
resultant shear forces on each segment (flanges and webs)

Segment AB, point 1 at y

VQ1 Vy 2
1   , 0 ya
It 2I
a ta 3V
F1   t1dy 
0 6I
Segment BC, point 2 at z

VQ2 V  a 2 
2     ab  az  ,  b  z  b
It I 2 
b tabV
F2   t2 dz  (a  2b)
b I

Segment CD, point 3 at y

VQ3 V  2 1 
3    a  2ab  y 2  ,  a  y  a
It I 2 

a ta 2V
F3   t3dy  (12b  5a)
a 3I
a(a  2b)V
The maximum shear stress in segment CD is at y = 0 and is 3 y 0

I
10
Segment DF, point 4 at z

VQ4 V  a 2 
4     ab  az  ,  b  z  b
It I 2 
b tabV
F4   t4 dz  (a  2b)
b I
Segment FG, point 5 at y
VQ5 Vy 2
5   , a  y  0
It 2I
0 ta 3V
F5   t5 dy 
a 6I

Step 5: Identify any unbalanced shear forces, and calculated their

combined moment about a reference point (here the middle point of
CD). Use Eq. (12.1) to calculate the location of the shear center.

ta 3V tabV 2ba  3b 2
Ve  (2b)( F1  F5 )  (a)( F2  F4 )  2(2b)  ( 2a ) (a  2b)  V
6I I a  3b
Therefore, the shear center is at

2ba  3b 2
e Review the solution: As expected, F5 and F1 have the same magnitudes and
a  3b same directions. F4 and F2 have the same magnitudes but with opposite
directions. In addition, F3 F1F5 =V . This means that the resultant shear
stress in the cross-section is equal to the total applied shear force on the
section. 11
Chapter 13 Buckling of beam

13.1 Introduction

Load-carrying structures may fail in a variety of ways. For example, a machine shaft may fracture suddenly
from repeated cycles of loading, or a beam may deflect excessively, so that it can not perform its intended
functions. These kinds of failures can be avoided by designing structures so that the maximum stresses and
maximum displacements are within tolerable limits.

Another type of failure is buckling. If a compression member is

relatively slender, it may deflect significantly and fail by
bending.
An example of this is a beam under the action of an axial load,
Fig. 13.1. When lateral bending occurs, the bending moment in
the beam now increases as the deflection increases.
1. This makes the beam deflect more.
2. This makes the bending moment increases further.
3. This makes the beam deflect even further.
4. Eventually and suddenly, the column will collapse
completely.
In such situations bodies/structures are referred to as being
unstable.
Fig. 13.1 Buckling of a beam due to You can easily perform a buckling “experiment” by applying an
an axial compressive load P axial load to a thin plastic ruler or other slender object.

* Buckling is the sudden loss of stability of a structure

when the load reaches a critical value.
1
13.2 Stability of equilibrium

The fundamental concepts of buckling and stability can be demonstrated by considering the equilibrium of a ball
placed upon a smooth surface (Fig. 13.2). If the surface is concave upward, like the inside of a dish, the
equilibrium is stable and the ball always moves back toward the low point when it is slightly displaced to one
side and then released. If the surface is convex upward, like a dome, the ball can theoretically be in equilibrium
on top of the surface. But in reality, the ball will tend to move further from the equilibrium position at the top of
the surface if it is slightly displaced to either side. In this case, the equilibrium is unstable. If the surface is
perfectly flat, the ball is in neutral equilibrium and remains wherever it is placed.

Stable equilibrium Neutral equilibrium Unstable equilibrium

Fig. 13.2 Stability of equilibrium

2
13.3 Buckling of common beams

Beam buckles by compressive load

Beam buckling mostly (almost always) always occurs under compressive normal stress.
For instance, when a beam bends, the length of the bent axis of the beam remains the same as its original length.
However, the two ends of the beam move towards each other (Fig. 13.3).
Looking at the equilibrium of the beam shown in Fig. 13.4a. Due to deflection , free end of the beam moves a
distance  along the direction of the external force P. As a result, the applied force will tend to increase the
moment in the beam and bends it more. Accordingly, this system has lower stability. In Fig. 13.4b, the external
force P tends to reduce the moment in and the deflection of the beam. Accordingly, the system is more stable.

Fig. 13.3 When bending, the two ends of the beam moves toward each other

(a) This beam tends to bend more by the (b) This beam tends to move back by the
compressive/pushing load P. Therefore, this tensile/pulling load P. Therefore, this
system is less stable. system is more stable.

Fig. 13.4
3
•Buckling also occurs under shearing load – we will demonstrate this in the buckling of
plate in Week 12.
•Buckling can be global (beam buckles as a whole) or local (see below figure)

Local buckling
4
1. Buckling analysis of ideal pin-ended beam
The simply supported beam is subjected to a
compressive load P at both ends (Fig. a). Both
ends of the beam are pin-ended. The boundary
conditions are:
at x  0 : w  0
(a)
at x  L : w  0
The deformed beam is shown in Fig. b. A free-
body diagram is shown in Fig. c.

From the FBD, the bending moment at x is

M ( x)  Pw (b)

In Lectures 5 and 6, we known that the

moment-deflection relationship of the beam is

2w
 EI 2  M ( x) (c)
x
Therefore,
2w
EI 2  Pw  0 (d)
x
This is a differential equation that governs the
deflected shape of a pin-ended beam. It is a
homogeneous, linear, second-order, ordinary
differential equation.

5
The general solution of Eq. (d) satisfying the The function that presents the shape of the deflected
boundary conditions at x = 0 is beam is called a mode shape, or buckling mode. The
constant A is arbitrary but small.
w( x)  A sin(x) (e)
The value of P at which buckling will actually occur
where is obviously the smallest value given by Eq. (h) (i.e.,
n =1 ). Thus, the critical load is
P
 (f)  2 EI
EI Pcr  2 (13.1)
L
If the constant A in Eq. (e) is made to equal to zero,
the deflection w is zero everywhere, and the beam and the corresponding fundamental buckling mode is
will have the original straight configuration.  x 
If we want an alternative equilibrium configuration,
w( x)  A sin  (j)
 L
A should not be equal to zero. In this situation, to
In conclusion, if the compressive load P is small, the
meet the boundary conditions at x = L, the condition
value of  is also small and the function sin(L) can not
sin(L) = 0 must be satisfied. This yields
be equal to zero. In this case, the boundary condition at
sin(L)  0  L  n , n  1, 2,... (g) the right end of the beam, w(L) = 0 can be satisfied only
when A is zero (see Eq. e). Therefore, there is no
From Eqs. (d) and (g), we see that transverse deflection w associated with the beam. The
beam only deforms along its axial direction under the
n 2 2 EI compressive load.
P  Pn  , n  1, 2,... (h)
L2 If the compressive load P is increasing and reaches the
The deflection that corresponds to each load Pn is critical value given by Eq. (13.1), sin(L) becomes zero
obtained by combining Eqs. (e), (f) and (h) to get and the boundary conditions at the right end of the beam
can be satisfied for any value of A, the beam buckles.
 nx  (i)
w( x)  A sin 
 L 
6
2. Buckling analysis of ideal fixed-free beam
The cantilevered beam is subjected to a
compressive load P at the free end (Fig. a). One
end of the beam is fixed and the other end of the
beam is free. The boundary conditions are:
w
at x  0 : w  0, 0 (a)
x
The deformed beam is shown in Fig. b, where  is
the maximum deflection of the beam. A free-body
diagram is shown in Fig. c.

From the FBD, the bending moment at x is

M ( x)   P(  w) (b)

In Lectures 5 and 6, we known that the

moment-deflection relationship of the beam is

2w
 EI 2  M ( x) (c)
x
Therefore,
2w
EI 2  Pw  P (d)
x
This is a differential equation that governs the
deflected shape of a pin-ended beam. It is a
homogeneous, linear, second-order, ordinary
differential equation.

7
The general solution of Eq. (d) satisfying the
The function that presents the shape of the deflected
boundary conditions at x = 0 is
beam is called a mode shape, or buckling mode. The
w( x)  A sin(x)  B cos(x)   (e) constant  is arbitrary but small.
where A and B are any constants, and The value of P at which buckling will actually occur
is obviously the smallest value given by Eq. (i) (i.e., n
P
 (f) =1 ). Thus, the critical load is
EI
 2 EI
To satisfy the boundary conditions at x = 0, Eq. (a), Pcr  (13.2)
we must have A = 0 and B =  . Accordingly, the ( 2 L) 2
deflection of the beam, Eq. (e) reduces to
and the corresponding fundamental buckling mode is
w( x)     cos(x) (g)
  x 
w( x)  1  cos  (k)
If we want a non-zero deflection w,  should not   2 L 
be equal to zero. To meet the boundary conditions
at x = L, that is w(L) = , the condition cos(L) =
0 must be satisfied. This yields If the compressive load P is small, the value of  is also
n small and the quantity inside the square brackets of Eq. (j)
cos(L)  0  L  , n  1, 3, 5, ... (h) can not be equal to zero. In this case, the boundary
2 condition at the right end of the beam, w(L) =  can be
From Eqs. (f) and (h), we see that satisfied only when  is zero (see Eq. g). Therefore, there
is no transverse deflection w associated with the beam.
n 2 2 EI
P  Pn  , n  1, 3, 5,... (i) If the compressive load P is increasing and reaches the
( 2 L) 2 critical value given by Eq. (13.2), cos(L) becomes zero
The deflection that corresponds to each load Pn is and the boundary conditions at the right end of the beam
obtained by combining Eqs. (f), (g) and (i) to get can be satisfied for any value of , the beam buckles.
  nx 
w( x)  1  cos  (j)
  2 L 
8
3. Buckling analysis of ideal fixed-pin-ended beam One end of the beam is fixed and the other end
of the beam is pin-ended (Fig. a). The boundary
conditions are:
w
at x  0 : w  0, 0
x
(a)
at x  L : w  0
The beam is subjected to a compressive load P at the
pined end. The deformed beam is shown in Fig. b. A
free-body diagram is shown in Fig. c. Due to bending,
the roller at the right end of the beam provide a
transverse reaction to the beam, which is unknown and
is denoted as Q.
From the FBD, the bending moment at x is
M ( x)  Pw  Q( L  x) (b)

In Lectures 5 and 6, we known that the

moment-deflection relationship of the beam is

2w
 EI 2  M ( x) (c)
x
Therefore,
2w
EI 2  Pw  Q( L  x) (d)
x
This is a differential equation that governs the
deflected shape of a pin-ended beam. It is a
homogeneous, linear, second-order, ordinary
differential equation.
9
The general solution of Eq. (d) is The solution of Eq. (h) is

Q L  4.493 (i)
w( x)  A sin(x)  B cos(x)  ( L  x) (e)
P
From Eqs. (f) and (i), we get
where A and B are any constants, and
P 2 EI 2 EI
 (f) P  Pcr   (13.3)
EI (0.6992 L) 2 (0.7 L) 2
To satisfy the boundary conditions at the left end of
the beam, we must have A = Q/(P) and B = QL/P.
Accordingly, the deflection of the beam, Eq. (e) If the compressive load P is small, the value of 
reduces to is also small and Eq. (h) can not be hold. In this
Q QL Q case, the boundary condition at the right end of the
w( x)  sin(x)  cos(x)  ( L  x) (g) beam, w(L) = 0 can be satisfied only when Q is
P P P zero (see Eq. g). Therefore, there is no transverse
If we want a non-zero deflection w, the boundary deflection w associated with the beam.
condition at x = L, that is w(L) = 0 must be If the compressive load P is increasing and
satisfied. This yields reaches the critical value given by Eq. (13.3), the
boundary conditions at the right end of the beam
sin(L)  L cos(L)  0 (h) can be satisfied for any value of Q, the beam
buckles.

10
4. Buckling analysis of ideal fixed-fixed beam Both ends of the beam are restricted from rotation
and the lateral displacement (Fig. a). The boundary
conditions are:
w
at x  0 : w  0, 0
x
(a)
w
at x  L : w  0, 0
x
The beam is subjected to a compressive load P at
the pined end. The deformed beam is shown in Fig.
b. A free-body diagram is shown in Fig. c. Due to
bending, the supporter at the right end of the beam
provide a couple to the beam so that the beam does
not rotate at that end. This couple is unknown and
is denoted as M0.
From the FBD, the bending moment at x is
M ( x)  Pw  M 0 (b)

In Lectures 5 and 6, we known that the

moment-deflection relationship of the beam is

2w
 EI 2  M ( x) (c)
x
Therefore,
2w
EI 2  Pw   M 0 (d)
(c) A free-body diagram x

11
The general solution of the homogeneous, linear,
second-order, ordinary differential equation Eq.
(d) is The solution of Eq. (h) for lowest value of  is

M0 L  2 (i)
w( x)  A sin(x)  B cos(x)  (e)
P
where A and B are any constants, and From Eqs. (f) and (i), we get

P 2 EI
 (f) P  Pcr  (13.4)
EI (0.5L) 2
To satisfy the boundary conditions at the left end of
the beam, we must have A = 0 and B = M0/P.
Accordingly, the deflection of the beam, Eq. (e) If the compressive load P is small, the value of 
reduces to is also small and Eq. (h) can not be hold. In this
case, the boundary conditions at the right end of
M0 M
w( x)  cos x  0 (g) the beam can be satisfied only when M0 is zero
P P (see Eq. g). Therefore, there is no transverse
deflection w associated with the beam.
If we want a non-zero deflection w, the boundary
conditions at x = L, those are is w(L) = 0 and If the compressive load P is increasing and
w(L) = 0 must be satisfied. This yields reaches the critical value given by Eq. (13.4), the
boundary conditions at the right end of the beam
sin(L)  0 , cos(L)  1 (h) can be satisfied for any value of M0, the beam
buckles.

12
The effective length of various beams

 2 EI
Introduce Le, the effective length of the beam. The buckling load can be written as Pcr  2
Le

Buckling loads were established for homogeneous beam with

2 D
constant E. If the beam is made of composite materials, the Pcr  2
bending stiffness EI in the equation of buckling load should be Le
replaced by the bending stiffness D, that is, on the right
13
Example 1 – Buckling load can be obtained through some observations

L/2 Pcr

L/2
w(x)

Original beam (pinned-pinned-pinned)

= • The original beam is equivalent to double of the

lower beam
0.5L Pcr 2 EI
• The buckling load of the lower beam is Pcr 
(0.5L) 2
w(x) • Buckling load of the original beam is the same as
the lower beam

Equivalent beam (pinned-pinned )

14
 •The original beam is equivalent to half of the lower beam
2 D 2 D
•Buckling load of the lower beam is Pcr  
(0.5  2 L) 2 L2
•Buckling load of the original beam is the same as the lower
beam

Original beam (fixed-guided)

Equivalent beam (fixed-fixed)

15
5. How to enhance the stability of the structures ?

Buckling load of the beam can be written in general form as

 2 EI
Pcr  2
Le
where Le is defined as the effective length of the beam.

To improve the buckling load, you can

1. Increase the Young’s modulus of the material (structure),

2. Reduce the length of the structure,

3. Increase the cross-sectional area of the structure

4. Provide additional supports (constraints) to the structures

16
Example 2: A steel pipe column that
is fixed at the base and free at the Solution:
top supports an axial load P = 240 We use millimeter (mm), newton (N), MPa units. From the buckling
kN. The length of the pipe is L = 3.6 load calculation formula, we see that the minimum required moment of
m. The outer radius of the pipe wall inertia is
is R = 80 mm. Find the minimum
required thickness t min for the pipe 2 EI P cr (2 L) 2 n  240000  7200 2
P cr  I   12 . 606 10 6
mm 4
wall using a factor of safety n = 2. ( 2 L) 2  2E  2  200000
Use E = 200 GPa and Y = 250 MPa.
From the moment of inertia calculation formula, we see that the
minimum required pipe wall thickness can be calculated from:

 (R 4
 r 4) 4I 4  12 . 606  10 6
I  r  R   r  80 
4 4 4 4
 r  70 . 65 mm 2

4  

The required minimum pipe wall thickness is:

tmin = R  r = 9.35 mm.

Supplementary calculation. Since the calculations for the critical

load is valid only if the material follows Hooke’s law, we need to
verify that the critical stress does not exceed the yield limit of the
material. We can obtain the following critical stress:

n P 2  240000
cr    108 . 5 MPa
A  (80 2  70 . 65 2)

 ( R 4  r 4) 2EI
Hint : I  , P cr  This stress is less than the yield limit (Y = 250 MPa). Therefore, the
4 (2L) 2 minimum pipe wall thickness calculation is satisfactory.
17
Example 3: A steel beam of rectangular cross-section, pin-pin support at two
ends in xoy plane, fixed-fixed support in xoz plane. a = 40 mm, b = 60 mm. l
= 2.1 m, l1 = 2 m, E = 205 GPa, Y = 200 MPa. Find the buckling force Pcr .

Solution:
(1) If buckling occurs in xoy plane

ab3 2 EI z 2 EI z
Iz  , Pcr  2   330.3KN
12 l 21002
(2) If buckling occurs in xoz plane

ba 3 2 EI y
Iy  , Pcr  2
 647.5KN
12 (0.5l1 )

Therefore, if buckles, buckling will occur in xoy plane

The critical load for buckling is Pcr = 330.3 KN.

330.3(1000)
The stress at buckling is cr   137.6MPa  Y
ab

Hence, no compression crush. If fails, the beam collapses by buckling

18
Example 4: A sandwich beam having aluminium-alloy faces enclosing
a plastic core (right Fig.) has a length L = 3.0 m. The thickness of the
faces is t = 5 mm and their modulus of elasticity is E1 = 72 GPa. The
height of the plastic core is hc = 150 mm and its modulus of elasticity is
E2 = 800 MPa. The overall dimensions of the beam are h = 160 mm and
b = 200 mm. Determine the buckling load of the beam. Assume that the
beam is fixed-fixed for bending in x-z plane and simply supported for
bending in the x-y plane.

Solution
1. For x-z plane bending (i.e. about the y axis). The centroid of the
cross-section is at the middle of the shape.

5  2003 150  2003

Dy  2 E1I y  face  E2 I y core  2  72(1000)   800   5.6 1011 N  mm 2
12 12
 2 Dy2  5.6 1011
Thus, if buckling occurs in xoz plane, Pcr    2.456 106 N
(0.5L) 2
(0.5  3000) 2

2. For x-y plane bending (i.e. about the z axis). The centroid of the
cross-section is also at the middle of the shape.
 200  53  200 1503
Dz  2 E1 I z  face  E2 I z core  2  72(1000)    77.5  (5  200)   800 
2
 9.102 1011 N  mm 2
 12  12
2 Dz 2  9.102 1011
Thus, if buckling occurs in xoy plane, Pcr  2
 2
 9.98 105 N
L 3000

3. Therefore, buckling will occur in xoy plane (i.e. buckling about the z axis) and the
critical load for buckling is Pcr = 998 kN.
19
 Chapter 14 : Lateral vibration of a beam

14.1 Introduction

• Vibration is a mechanical phenomenon whereby oscillations occur about an

equilibrium point. The oscillations may be periodic, such as the motion of a
pendulum—or random, such as the movement of a tire on a gravel road.

Loads can be classified as static or dynamic depending upon whether they

remain constant or vary with time.

• A static load is applied slowly, so that it causes no vibrational or dynamic

effects in the structure. The load increases gradually from zero to its maximum
value, and thereafter it remains constant.

• A dynamic load may take many forms

* some loads are applied and removed suddenly (impact loads). Impact loads
are produced when two objects collide or when a falling object strikes a
structure.
* others persist for long periods of time and continuously vary in intensity
(fluctuating loads). Fluctuating loads are produced by rotating machinery,
traffic, wind gusts, water waves, earthquakes, and manufacturing processes.

1
 Structural vibration occurs when dynamic
forces cause the structures to vibrate.

Vibrations due to the structure being

mechanically resonant lead to structural failures,
poor structural reliability, and safety concerns.

It is not possible to cover every aspect of the http://www.strand7.com/html/archive/structural/building.htm

structure vibration in this lecture of just 2 hours.
There are a number of references (books) about the
Structural Dynamics where more complicate
structural/system analysis are introduced. 2
14.2 Free vibration of a spring/mass system (A single degree-of-freedom system)

In order to understand the more complicate vibration of elastic system such as a beam,
we first study the behavior of a simple mass/spring system. The basic rules of the
vibrations of different system are similar.

* Newton’s second law:

F = Ma F: the force applied to the object of mass M; a: acceleration.

* The spring force is opposite to the direction of x.

F = k x(t) k : spring coefficient, stiffness coefficient

* The acceleration of the object

2 x
a  2  x
t
* Collecting the above terms, the motion differential equation of the system is obtained as

Mx(t )  kx(t )  0 (14.1)

It is a differential equation of the second order

3
* Assume that, initially, the object is at x = 0, that is x(0) = 0, the
solution of Eq. (14.1) will be
k
x(t )  A sin(t ) ,   (14.2)
M
: has no dependence on the external force, there is known as the
natural frequency of the system,
A: a constant to determine from the initial condition.
* Denote the initial velocity of the object as V0. Therefore

x (t ) t 0  V0 (a)

* From Eqs. (14.2) and (a), one obtains

V0
A (b)

* Finally, the motion equation of the object is

V0
x(t )  sin(t ) (c)


V0
* The amplitude of the vibation is A0 


: natural frequency of the system, A0: amplitude of the vibration.

4
 2

A0
x(t)

x(t)
t

Displacement-time relationship, Eq. (c)

5
14.3 Force vibration of a spring/mass system
If, in addition to the spring force, there is an external force P applied to the object,
additional term should be added to the motion differential equation of Eq. (14-1).
From the free body diagram of the left bottom figure, we obtain

Mx(t )  kx(t )  P(t ) (14.3)

The solution of this equation is not easy. We consider two special cases of the
applied load and the initial conditions.

•Special case 1: P(t)=P0sin(pt) and the initial position of the object is zero, where p
is the frequency of the external load.

If the solution of x is as follows, the initial conditions are can be satisfied:

x(t )  A sin( pt ) (14.4)

where A is a constant. Substitute Eq. (14.4) into Eq. (14.3), A can be obtained.
As a result, the solution is

P0 k
x(t )  sin( pt ) , ω  (14.5)
M (2  p 2 ) M

From Eq. (14.5) we see that resonance occurs when the frequency of applied force

matches the natural frequency of the structure, i.e.,

k
p ω
M
6
* Special case 2: A constant P = P0 is applied to the object. We will try to find a solution to Eq. (14.3)
when the initial location is zero and the initial velocity is V0. We assume

 P0 cosωt   V0 sin ωt , ω 

P0 M k
x(t )  (14.6)
k k M
It can be seen that Eq. (14.3) is satisfied by the above equation, where  is the same
as that of Eq. (14.5).

Example 1: The impact of an object falling onto the upper end of a spring.
An object of mass M, being at rest, falls from a height h onto and attach to the top end of
the spring. When the object strikes the spring, the spring begins to shorten. The object
moves downward and reach its position of (negative) maximum displacement. Thereafter,
the object moves up, then down, then up again as the spring vibrates longitudinally.

Solution: When the object attaches to the spring, the vibration of the system begins
The initial velocity of the object are (when the object attaches to the spring) is

V0   2 gh (a)
where g is the gravity acceleration. Negative sign means the direction of the initial
velocity is downwards.
In addition to the force from the spring, the object is also subjected to a constant gravity
force P0 = Mg.

7
Thus, the motion of the object is [use Eq. (14.6)]:

 Mg Mg
cos  t   sin   t 
2Mgh
x(t )   (b)
k k k

Equation (b) is the desired result of the motion of the object with time.

The maximum deflection of the object A0 is the maximum absulate

value of x(t)

Mg
2
 Mg  2Mgh  2h 
A0  xmax        
st 1  1  (14.7)
k  k  k   st 

Mg
where the static deflection of the object is  st 
k
If the object is initially attached to the spring and is released
suddenly. From Eq. (14.7), one obtains

A0 = 2st (14.8)

Therefore, in this case, the dynamic deflection is twice of the

static deflection.

Note that Eqs. (14.7) and (14.8) also apply to the multi-degree-
of-freedom elastic system/structure
8
14.4 Lateral (transverse) vibration of beam (Multi-degree-of-freedom system)

•From the vibration analysis of a mass/spring system, we see that resonance occurs if the natural
frequency of the system matches the frequency of the applied force.
•For the continuous structure system, resonance also occur if the frequency of the applied load
is close to a frequency of the structure.
•Therefore, a major task for a structure engineer is to find the frequencies of the structure.

dx
M
M V M dx
x
z

We assume that the z = 0 axis is placed at the “neutral

V
V dx axis” of the beam and the rotary inertia of the beam is
x
ignored. The cross-section of the beam is A, second
q
moment of inertia is I, Young’s modulus is E, mass density

A free-body diagram for a beam segment, length dx is  (i.e., mass per unit volume in the unit of kg/m3 ).

9
Governing equations
If the beam is uniform, EI , A and  will be constants. Based
• Equilibrium (lateral) on Eq. (d), The vibration of the beam under the application of
 V   w 2
V  dx  Vdx  qdx  Adx 2 (a) external load q can be obtained from Eqs. (c) and (d) as
 x  t
4w 2w q
• Equilibrium (bending) a 2
  (14.9)
x 4 t 2 A
M
M dx  M  V (b) where a is a constant
x
• Motion equation is obtained by putting Eq. (b) EI
into Eq. (a) a (14.10)
A

 2 M ( x) 2w If there is no applied distribution load, the lateral

 A 2  q (c)
x 2 t free vibration equation of beam is

• Constitutive equation (Lecture #5) 4w 2w

a2
 0 (14.11)
 w( x)
2 x 4 t 2
M ( x)   EI (d)
x 2

10
Free vibration of a beam

Detailed procedure for obtaining the solution of Eq. (14.11) is beyond the scope of this unit. If at the
beginning the deflection of the beam is zero, that is w = 0 for t = 0, the solution can be expressed as

w( x, t )  [C1 sin(x)  C2 cos(x)  C3 sinh(x)  C4 cosh(x)] sin(t ) (14.12)

in which
Note that: sinh, cosh, tanh are
  / a (14.13) hyperbolic functions

and the constants , the constants C1, C2, C3 and C4 and  are
determined from the boundary condition (B.C.) of the beam.

It will be shown that the constant  is a property of the beam. It depends on the material of the beam,
the shape of the cross section, the length of the beam, and the boundary conditions. It has no
relationship with the external force. Therefore,  is known as the natural frequency of the beam.

11
(1) Ideal pin-ended (simply supported) beam
 2 w(0) Deflection and
w(0)  0,  0,
x 2 bending moment
B.C. : are zero at the
 2 w( L)
w( L)  0, 0 pin-ended ends
x 2

From Eq. (14.12) and the BCs, we see that the constants C1, C2, C3 and C4 must be such that
w(0)  0  C2  C4  0 w( L)  0  C1 sin(L)  C3 sinh(L)  0

 2 w(0)  2 w( L)
 0   C2  C4  0  0   C1 sin(L)  C3 sinh(L)  0
2 x 2 x

C3 sinh(L)  0 , C1 sin(L)  0 (a)

C2  0, C4  0
C3  0

• Since sinh(L) can not be zero, C3 = 0 is obtained. Therefore, the only non-zero solution is C1. If it is also made
to equal to zero, the deflection w is zero everywhere, and the beam will have the original straight configuration.
• If we want an alternative equilibrium configuration, the value of  must be such that:
sin(L)  0 (b)

As a result of Eq. (b), the vibration frequencies of the beam can be determined as
n
n  , n  1,2,.... (c)
L
12
  EI
Since   , a , the vibration frequencies are obtained as
a A

 n 
2
EI
n    , n  1,2,.. (14.14)
 L  A

n : natural frequencies of the vibration, 1 is the fundamental frequency of the vibration

The general solution to the vibration of the beam is the combination of all possible vibration modes:

w( x, t )   Cn  n ( x) sin(nt ) (e)
n 1

where Cn depend on the initial conditions, and n are the vibration modes (or natural modes of
vibrations):
 nx 
 n ( x)  sin  (f)
 L 

From Eqs. (14.14) and (f), you can see there are infinite number of natural frequencies and vibration
modes. This system is said to be multi-degree-of-freedom. This is a feature of all elastic systems.

13
Below figure depicts the first three vibration mode shapes

111 1st mode shape, one half-wave

2 EI
1  2
L A

L
2
2nd mode shape, two half-waves
4 2 EI
2  2
L A

3
3rd mode shape, three half-waves
9 2 EI
3  2
L A
L

14
1st mode shape,
one half-wave

2nd mode shape,

two half-waves

3rd mode shape,

three half-waves

15
 (2) Cantilever beam
w(0)
w(0)  0, 0 Bending moment
x and transverse shear
B.C. :
force are zero at the
 2 w( L)  3 w( L)
 0,  0 free end
x 2
x 3

Follow a same analysis as the simply-supported beam, we can obtain the natural frequencies of this problem. These are

3.516 EI 22.03 EI 61.70 EI

1  ,   ,   ,..., (14.15)
A A A
2 3
L2 L2 L2

 (2n  1)  1
2
EI
For n larger, n    2 1 is the fundamental frequency of the vibration
 2  L A
The equation of vibration is

w( x, t )   Cn  n ( x) sin(nt )
n 1

The vibration modes (or natural modes of vibrations) are:

 sinh(n L)  sin(n L)  n
sin(n x)  sinh(n x)   cosh(n x)  cos(n x)  , n 
EI
 n ( x)   , a
 cos(n L)  cosh(n L)  a A

16
Below figure depicts the first three vibration mode shapes of the fixed-free beam

2
3(x) 1(x)
2(x)
1.5

1
2(x)
0.5

-0.5 3(x)

-1

-1.5

-2
0 0.2 0.4 0.6 0.8 1

17
 (3) Fixed-pinned beam
w(0) Bending moment
w(0)  0, 0
x and deflection are
B.C. : zero at the right end
 2 w( L)
w( L)  0 , 0
x 2

Follow a same analysis as the simply-supported beam, we can obtain the natural frequencies of this problem. These are

15.4182 EI 49.9651 EI 104.2482 EI

1  ,   ,   (14.16)
A A A
2 3
L2 L2 L2

 (4n  1)  1
2
EI 1 is the fundamental frequency of the vibration
For n larger, n    2
 4  L A
The equation of vibration is

w( x, t )   Cn  n ( x) sin(nt )
n 1

The vibration modes (or natural modes of vibrations) are:

n
 n ( x)  sin(n x)  sinh(n x)  tan(n L)cos(n x)  cosh(n x) , n 
EI
, a
a A

18
Below figure depicts the first three vibration mode shapes of the fixed-free beam

19
 (4) Fixed-guided beam
w(0)
w(0)  0, 0 Transverse shear
x
B.C. : force is zero at the
w( L)  3 w( L) right end
L  0, 0
x x 3

Follow a same analysis as the simply-supported beam, we can obtain the natural frequencies of this problem. These are

5.5932 EI 30.2258 EI 74.6392 EI

1  ,   ,   ,..., (14.17)
A A A
2 3
L2 L2 L2

 (4n  1)  1
2
EI 1 is the fundamental frequency of the vibration
For n larger, n    2
 4  L A
The equation of vibration is

w( x, t )   Cn  n ( x) sin(nt )
n 1

The vibration modes (or natural modes of vibrations) are:

cos(n L)
 n ( x)  sin(n x)  sinh(n x)  cos(n x)  cosh(n x) , n  n , a  EI
sin(n L) a A

20
Below figure depicts the first three vibration mode shapes of the fixed-free beam

21
 (5) Fixed-fixed beam
w(0)
w(0)  0, 0
x Deflection and slope
B.C. :
w( L) are zero at both ends
w( L)  0, 0
x

Follow a same analysis as the simply-supported beam, we can obtain the natural frequencies of this problem. These are

22.3729 EI 61.6728 EI 120.9032 EI

1  ,   ,   (14.18)
A A A
2 3
L2 L2 L2

 (2n  1)  1
2
EI
For n larger, n    2 1 is the fundamental frequency of the vibration
 2  L A
The equation of vibration is

w( x, t )   Cn  n ( x) sin(nt )
n 1

The vibration modes (or natural modes of vibrations) are:

sinh(n L)  sin(n L)
 n ( x)  sin(n x)  sinh(n x)  cosh(n x)  cos(n x), n  n , a  EI
cosh(n L)  cos(n L) a A

22
Below figure depicts the first three vibration mode shapes of the fixed-free beam

23
Conclusions – natural frequencies

 n 
2
EI
n    , n  1,2,..
 L  A

3.516 EI 22.03 EI 61.70 EI

1  ,   ,   ,...
A A A
2 3
L2 L2 L2

15.4182 EI 49.9651 EI 104.2482 EI

1  ,   ,   ,...
A A A
2 3
L2 L2 L2

5.5932 EI 30.2258 EI 74.6392 EI

1  ,   ,   ,...
A A A
2 3
L2 L2 L2

22.3729 EI 61.6728 EI 120.9032 EI

1  ,   ,   ,...
A A A
2 3
L2 L2 L2

Results of Composite Material Beam can be obtained directly from the results of homogeneous
beam by replacing EI with D and A with 1A1+2A2+…. 24
Example 2: Find the first three natural frequencies of a steel bar that is simply
supported in x-z plane and cantilevered in x-y plane. The bar is 45 cm long and has a
rectangular cross-section of 0.3 cm high and 2 cm width. Take the density of the bar
as  = 7850 kg/m3 and E = 210 GPa.

Solution: Use N, m, kg, Pa units.

1. Due to symmetry, the centroid is at the middle of the shape.

2. For x-z plane vibration (i.e. bending about the y axis)

bh3 9 0.02  0.003

3
EI y  E  210 10  9.45 Nm2
12 12
A  7850  0.02  0.003  0.471 kg/m

    
2 2
EI y 9.45
1       218.3 rad/s , f1  1  34.75 Hz
 L A  0.45  0.471 2

 2  2
2
EI y
2     873.3 rad/s , f 2   139.0 Hz
 L  A 2  is the angular natural frequency

 3  3
2
EI y
3     1964.8 rad/s , f 3   312.7 Hz f is the circular natural frequency
 L A 2

25
3. For x-y plane vibration (i.e. bending about the z axis)

hb3 9 0.003  0.02

3
EI z  E  210 10  420 Nm2
12 12
A  7850  0.02  0.003  0.471 kg/m

3.516 EI z 3.516 420 1

1    518.5 rad/s , f   82.52 Hz
A 0.452 0.471 2
1
L2

22.03 EI z 
2   3249 rad/s , f 2  2  517.0 Hz
L 2
A 2

61.70 EI z 
3   9098 rad/s , f 3  3  1448 Hz
L2 A 2

 is the angular natural frequency

f is the circular natural frequency

26
Chapter 15. Theory of Plate

15. 1 Introduction

* A plate is a two dimensional structure: the size in one direction is

considerably smaller than the sizes in other two directions;

* The thickness of the plate is much smaller than its length and width;

* Many engineering structures are made of beams (columns) and plates.

1
 h

z
y A plate is a flat structure.
b
The thickness is small compared with the surface
x dimensions.

The thickness is measured normal to the middle

surface of the plate and is usually constant but may
a be variable.

Plates subjected only to in-plane loading can be

solved using two-dimensional plane stress theory.
On the other hand, plate theory is concerned mainly
with lateral loading.

h
The mid-plane of plate

2
15.2 Geometry of plate

h (*) The mid-plane of the plate is a reference surface. The

xoy coordinate describes the plate on the mid-plane. The
x-y-z coordinate system is space fixed.
z y
(*) The thickness of plate is denoted as h.
b
o x (*) Coordinate z is in the thickness direction, from h/2 to h/2.

(*) The plate is considerably longer and wider than its

a
thickness. Therefore

Fig. 15.1. Geometry h  a , h  b

Purpose of plate theory:

To eliminate the dependence on z of the equations.

3
15.3 Stresses and internal forces on the plate

z
y

z
The plate is under the plane stress state.
The non-zero stress are x, y and xy
 xz  yz

x  xy  xy  y y
x

4
Bending moments and torques on the plate

y
y
yx
z
y z x
x xy
o x x
xy

yx x

y

z z
One of the differences between 2-D plane stress and
plate theory is that in the plate theory the stress x
components are allowed to vary through the thickness
of the plate, so that there can be bending moments. z x
x
Mx

M x    xx dz z    xx zdz
z z

Mx: Bending moment about the y axis. It is moment

per unit length. Its unit is length (e.g., meter).
5
 y

yx
z
y
x xy
o x x
xy

yx

y

z z

z y yx
z
x x
Myx
x

yx M yx    yx dz z    yx zdz
z z

Myx: Torque about the y axis

Myx: Moment per unit length. Its unit is

length (e.g., meter).
6
 M x    xx zdz Bending moment about the y axis
z
My
Myx
M y    yy zdz Bending moment about the  x axis
z
z
Mx y
Mx
o x Mxy M xy    xy zdz Torque about the  x axis
Mxy Myx z

My
M yx    yx zdz Torque about the y axis
z
Fig. 3. Bending and torque moments on the plate

Myx = Mxy is obvious.

(*) The bending moments are shown acting on the sides of the element.

(*) The directions of the moments shown in the figure are positive.

(*) These bending moments have dimensions of force (i.e., N) , e.g., moment per unit length (i.e., N.m/m).

7
Out-of-plane (transverse) shear forces on the plate

(*) The out-of-plane (transverse shearing force) intensities Qx and Qy are shown acting on the
sides of the element, with positive forces acting in the positive directions on positive faces.

(*) These forces have dimensions of force per unit length (e.g., N/m).

Qy
(*) The force is positive if it is acting in the
positive directions on a positive face;
Qx z y Qx (*) The force is negative if it is acting in the
negative directions on a positive face;
o x
(*) The force is positive if it is acting in the
negative directions on a negative face;

(*) The force is negative if it is acting in the

Qy positive directions on a negative face.

8
Example: What are the values of Qx and Qy (the Example: What are the values of Qx and Qy (the
positive/negative sign). positive/negative sign).

50N/m 55N/m

20N/m z 40N/m z
y y
20N/m 40N/m
o x o x

50N/m 55N/m

Answer: Qx = 20 N/m, Qy = 50 N/m Answer: Qx = 40 N/m, Qy =  55 N/m

9
 In total, there are 5 internal forces:
My • Lateral shear forces: Qx, Qy
Myx
• Bending moments: Mx, My
z • Torque: Mxy
Mx y
Mx
x Mxy
o
Mxy Myx
My

Qy

Qx z y Qx

o x

Qy

10
 External pressure on the plate surface

z y

o x q(x, y)

q : Lateral pressure, q(x, y). It is positive if

directed to the positive z direction

(*) The transverse external pressure q arises from, for example, an

external pressure or gravitational field. The unit of q is force per
unit area (e.g., N/m2 ).

11
15.4 Relationships between the internal forces and external forces – the equilibrium of the plate

The lateral forces and their incremental changes are shown acting on the sides of the element.

Q y
Qy  dy
y

Qx z Q x
q(x, y) y Qx  dx Summing forces in the z-direction yields:
x
o x
Qx Qy 2w
dy   q  h 2 (15.1)
x y t

Qy
: volume density of the plate.
dx

12
 M y M yx
My  dy M yx  dy
y y
Summing moments about the y–axis
yields the equation:
z
y M x
Mx Mx  dx
dy Mxy x M x M yx
o x dxdy  dydx  Qx dydx  0
Mxy
Myx x y (15.2a)
M x M yx
My    Qx  0
x y
dx

Q y
Qy  dy
y In a same way, another moment equilibrium
equation (about the rotation of plate about the
x-axis):
Qx z Q x
q(x, y) y Qx  dx
x M xy M y
x   Qy  0 (15.2b)
o x y

dy

Qy

dx

13
Equilibrium equations of the plate.

Substituting Eqs. (15.2a) and (15.2b) into Eq. (15.1), one obtains

2M x  2 M xy  2 M y 2w
2   q  h 2 (15.3)
x 2 xy y 2 t

Special case: a one-dimensional beam

Nothing depends on y, thus

2M x
Eq. (15.3) reduces to q 0
x 2

which has been developed in Lecture 6

14
15.5 Deformation of plate

Deformed shape

(*) Plate theory is an approximate theory.

w It is very like the beam theory– only with
an extra dimension
z y
(*) Things are more complicated for plates
than for the beams. For one, the plate not
x only bends, but torsion may occur (it can
o twist).

Original position

(*) Deflection of the plate: it is the displacement of the plate normal to its plane (i.e, along
the thickness direction, the z direction). It is denoted as w.

(*) Due to deflection, the cross-section of the plate also rotates about the x and y axes.

15
Notations:
• The transverse deflection (displacement along the thickness of the plate)
does not vary with z. It is function of x and y. Hence w = w (x, y) ;
• u-displacement of any point of the plate in the x direction.
• v-displacement of any point of the plate in the x direction.

Assumptions:
• The transverse shear stress are small. This means xz and yz are
negligible, when comparing with the in-plane stress x, y and xy.

Based on these assumptions, we see that

u w w
 xz  0   xz  0    0  u  z (15.4a)
z x x

v w w
 yz  0   yz  0    0  v  z (15.4b)
z x y
From Eq. (15.4), the normal strains and the shear strain of the plate are

u  2 w( x, y )
 x ( x, y )   z
x x 2
v  2 w( x, y )
 y ( x, y )   z (15.5)
y y 2
u v  2 w( x, y )
 xy ( x, y )    2 z
y x xy
16
15.6 Moment – deflection relationships and stresses in the plate

From Eq. (15.5), the normal stress and the shear stress of the plate are obtained as

E Ez  2 w 2w
x  ( x   y ) x   (  2 )
1  2 1  2 x 2 y

y 
E
( y   x ) Ez 2w 2w
y   (  ) (15.6a)
1  2 1  2 x 2 y 2
E M xz
 xy   xy Ez 2 2 w x 
2(1  )  xy   h 3 / 12
2(1  ) xy
M yz
y  (15.7)
h 3 / 12
M xy z
h/2  2w 2w   xy 
Mx    x zdz   D 2   2  h 3 / 12
h / 2
 x y 
h/2  2w 2w  (15.6b)
My    y zdz   D  2  2 
h / 2
 x y 
h/2 2w
M xy    xy zdz  (1  ) D
h / 2 xy

h/2 E Eh 3
D z dz 
2 D: bending stiffness, bending moment
 h / 2 (1  2 ) 12(1  2 ) of inertia, : Poisson’s ratio

17
 15.7 Strain energy of the plate   2 w( x, y ) 
 
 x 2 
  2 w( x, y ) 
(Generalized) strains:  
My  y 2 
Myx   2 w( x, y ) 
 2 xy 
 
z
Mx y
Mx Mx 
x Mxy
o  
Mxy (Generalized) stresses: {M }   M y 
Myx M 
 xy 
My

Strain energy density:

1
e (Generalized stress)  (Generalized strain)
2

Strain energy of the entire plate, U

1  2w 2w 2w  1   2 w  2 2w 2w  2w 

2
 2w  
2

U      M x 2  M y 2  2M xy dxdy    D  2   2 2    2(1  )  dxdy (15.8)

2 x y x y xy  2 x y  x  x y 2  y 2   xy  


The special situation is beam (nothing depends on y), U reduces to

2
1  2w  1  2w 
U     M 2 dx   EI  2  dx (15.9)
2 x x  2 x  x 

Note that M in Eq. (15.9) has the unit of moment (i.e., N.m). 18
Chapter 16. Bending of Plate

16.1 Basic equation of plate

Equilibrium equations of the plate

(developed in Chapter 15).

2M x  2 M xy  2 M y 2w
2   q  h 2 (15.3)
x 2 xy y 2 t

Moment – deflection relationships (developed in Chapter 15).

 2w 2w 
M x   D 2   2 
 x y 

 2w 2w 
M y   D  2  2  (15.6b)
 x y 
2w
M xy  (1  ) D
xy

Static equilibrium equation , in terms of deflection, is

obtained by substituting Eq. (15.6b) into Eq. (15.3)
and ignore the inertia term,
Note that in static loading, the right hand side term of
 4w 4w 4w 
D 4  2 2 2  4   q( x, y ) (16.1) Eq. (15.3) vanishes.
 x x y y 

1
16.2. Boundary conditions
q

1. Clamped (fixed) boundary (at x = a): x

b

w( x, y ) a
w(x, y) =0 =0
x= a x x= a y

Clamped
2. Simply supported (at x = 0):

w(x, y) =0 Mx(x, y) =0
x=0 x=0

 2w 2w  2w

 D 2   2   0 0
 x y  x 0 x 2 x 0

3. Free boundary (at y = b):

2w
Myx(x, y) =0  (1  ) D 0
y=b xy y b

 2w 2w 
My(x, y) =0  D  2  2   0
y=b  x y  y b

2
Solution procedure:
1. Solve the plate bending equation for w(x, y)
2. Calculate moments (Mx, My, Mxy), using Eqs. (15.6b).
3. Calculate stresses (x, y, xy), using Eqs. (15.6).

16.3 Pure bending of an elastic plate.

Consider a plate subjected to bending moments Mx= M1 and Accordingly, we get the solution of w as
My= M2 , with no other loading, as shown in figure below.
1 M 1  M 2 2 1 M 2  M 1 2
M2 y w x  y (a)
2 D(1  2 ) 2 D(1  2 )

M1
x In the special case of equal bending
M1 moments, with M1=M2=M0, we obtain
1 M0
w ( x 2  y 2 ) (b)
M2 2 D(1  )

From Eq. (15.6b), one obtains The plate bends to a sphere

 2 w M 1  M 2

x 2 D(1  2 )

 2 w M 2  M 1

y 2 D(1  2 )
2w
0
xy
3
Note that when there is only one moment, My = 0 say, there is
still curvature in both directions. This is due to Poisson’s
effect. In this case, one can solve the moment-curvature
equation to get y

2w M1 2w 2w M1

 ,   2 x
x 2 D(1  2 ) y 2 x
M1

w
1 M1
2 D(1   )
2
x 2  y 2 

In order to get a pure cylindrical deformation (this means that

w does not vary with y), one needs to apply moments Mx and
My = Mx, in which case, from Eq. (a), one obtains

1 M1 2
w x
2 D

4
16.4. Pure torsion of an elastic plate.
M0
In pure torsion, one has the twisting moment Mxy = M0 with no
other loading. See right figure. From moment-curvature
equations, Eq. (15.6b), one obtains z
y
2w 2w 2w M0 x M0
 0,  0,  o
x 2 y 2 xy D(1  ) M0 M0

M0
w xy
D(1  )

The middle plane is deformed as below. Note that

there is no deflection along the lines x = 0 or y = 0.

5
16.5 Plate bending in rectangular coordinates

Example 1: A rectangular plate bends in one x

direction only (if b is either very small or very high b
(comparing with a), the deflection of the plate will
mainly depend on x.
a
y
This means that w = w(x). From the moment-
curvature relation, the deflection and moments
are

d2w d2w
M x   D 2 , M y  D 2
dx dx

The differential equation, Eq. (16.1) reads

d 4 w q ( x)

dx 4 D
It is the same as the beam bending equation.

6
Example 2: Deflection of a circular plate that is clamped at its edge by a uniform lateral pressure

Consider a circular plate with boundary

x2  y 2  a2

The plate is clamped at its edges and subjected to a

uniform lateral pressure q. Note that here q is downward.
The equilibrium equation of Eq. (16.1) becomes

4w 2w 4w q

 2   
x 4 x 2y 2 y 4 D

The boundary conditions are such that the slope and deflection are zero at the boundary:

w w
w  0,  0,  0 along x2  y 2  a2
x y
It can be seen that the solution of the deflection is

w
q
64 D

x2  y2  a2 
2

This is plotted in the right figure. The maximum w

deflection occurs at the plate centre and is

qa 4
wmax 
64 D
7
Since the deflection of the plate is obtained, from the
stress-moment relations, Eq. (15.6a), the stresses in
the plate are found to be

x  
3qz
4h 3

(3  ) x 2  (3  1) y 2  (1  )a 2 

y  
3qz
4h 3

(3  1) x 2  (3  ) y 2  (1  )a 2  The normal stresses on the top/bottom surface at
the plate centre are:
2
3qz 3q a h
 xy   (1  ) xy  x   y   (1  )  , z   , ( x, y)  (0,0)
2h 3 8 h 2
At the plate edge, for example, x = a and y = 0, the
normal stress on the top/bottom surface of the plate is:
Ez  2 w 2w
x   (  2 ) 3q  a 
2
h
1  2 x 2 y  x     , z   , ( x, y)  (a,0)
Ez 2w 2w
4 h 2
y   (   ) (15.6a)
1  2 x 2 y 2 The maximum shear stress is on the top/bottom
Ez 2 w 2
surface of the plate at where xy=0.5a2
 xy  
2(1  ) xy
2
3q a h 1
 xy   (1  )  , z   , xy  a 2
8 h 2 2

8
Example 3: Bending of a rectangular plate that is simply supported
at four edges and under a surface pressure q = q(x, y).
The boundary conditions are
q
w  0 , M x  0 , for x  0 and x  a;
w  0 , M y  0 , for y  0 and y  b
b


x

2w
a w  0,  0 , for x  0 and x  a;
y x 2
2w
w  0,  0 , for y  0 and y  b
y 2

The solution of the bending equation, Eq. (16.1) is expressed by the sum of a series:
This solution is due to Navier and is called Navier’s solution
M N
 mx   ny 
w( x, y )   Cmn sin  sin  (a) to the rectangular plate problem. If the terms of the series are
m 1 n 1  a   b  infinite, the result will converge, therefore, Eq. (a) is known
as the exact/theoretical solution of the problem.
where M and N are sufficiently large numbers to ensure the convergence of the solution, and:

4 a b  mx   ny 
Cmn 

2  0 q ( x , y ) sin   sin
 a   b 
dxdy (b)
n 
2 2 0
4 m
abD      
 a   b  
* The derivation of Eq. (a) needs the advanced analysis method such as,
separation-of-variable, energy method, or variational principle. It is beyond
the scope of this unit.
9
 If the load is uniform, q = q0 :

16q0
Cmn  , m  1,3,5...; n  1,3,5...

D6 mn (m / a) 2  (n / b) 2 2

 
 mx   ny 
w( x, y)   
m 1, 3,... n 1, 3,...
Cmn sin  sin
 a   b 


If plate is square and the applied is a uniform pressure q=q0 :

q0

16q0 a 4 1 16q0 a 4 1
x C11  , C13  C31  ,
D6 4 D6 300
a
16q0 a 4 1 16q0 a 4 1
C15  C51  , C35  C53  ,
D6 3380 D6 17340
a
16q0 a 4 1
y C55 
D6 62500

You can see Cmn decreases quickly with m and n.

The convergence is very fast.

10
Deflection is maximum at the center of the plate. If plate is square, it is :

16q0 a 4 M N
(1) ( m n ) / 21
wmax 
D6
 
m 1, 3,... n 1, 3,... mn( m  n )
2 2 2

Theoretically, if M and N are taken as infinite, the result is exact.

In reality, M and N do not need to be infinite.

For example,
0.004161q0 a 4
• If M = 1 and N = 1 is used: wmax 
D

0.004055q0 a 4
• If M = 3 and N = 3 is used: wmax 
D

0.0040624q0 a 4
• If M = 3 and N = 3 is used: wmax 
D
• Further increase of M and N does not change the result visibly.

? Can you find the stresses in the plate?

Answer: Since the deflection w(x, y) has been obtained, the stresses in the
plates at any point (x, y, z) can be found from Eq. (15.6a) of Chapter 15:

• The z =  h / 2 (the faces of the plate) will have the

maximum compressive/tensile/shear stresses.
11
Example 4: Bending of a rectangular plate under a
concentration load Q at the middle of the plate

Q
If the plate is square (a = b) :
x
b a 2Q
C11 
D4
a 0.04a 2Q
y C31  C13  
D4
a 2Q
C33  0.0123 
D4
If the plate is simply supported at four edges, the
solution is the same as Eq. (a) of Example 3. However, 0.00346a 2Q
C35  C53  
the coefficients Cmn are different and are: D4
0.00592a 2Q
C15 
 m   n  D4
4Q sin   sin  
Cmn   2   2 
, m  1,3,5...; n  1,3,5...
2 2
(c) You can see that the series
 2
m n  converges very fast.
abD 4      
 a   b  

12
Example 5: Find the maximum deflection and stresses Solution:

of a rectangular plate under a concentration load Q at Deflection is maximum at the center of the plate.
From Eqs. (a) and (c) of Examples 3 and 4, one gets
the middle of the plate for a = 2000 mm, b = 1000 mm,
h = 10 mm and Q = 1 kN. Assume the plate is made of a b M N
4Q
w( , )    2
aluminum with E = 72 GPa and  = 0.3. 2 2 m 1,3,... n 1,3,... 
4 m
2
n 
2

abD      
 a   b  
Q M N
3.114015
  
m 1, 3,... n 1, 3,... (0.25m  n )
2 2 2
x
b You can easily make a computer program to perform
the calculation of the summation of the sequences. If
a we take (M = 11 and N = 11), and (M = 99 and N = 99),
y respectively and compare the deflections calculated, we
get.

11 11
3.114015
w0     2.488 mm
m 1, 3,... n 1, 3,... 0.5m  n 
2 2 2

99 99
3.114015
w0     2.506 mm
m 1, 3,... n 1, 3,... 0.5m  n 
2 2 2

It is obvious that (M = 11 and N = 11) has reached a

solution with sufficient and satisfying precision.

13
 16.6 Transverse shear stresses in a plate z y
From the stress-moment relationship and based on the
equilibrium of three-dimensional theory of elasticity, we get
x
 x  xy  xz
  0 h
x y z

 xz  x  xy Ez  3 w  3 w
   (  )
z x y 1  2 x 3 xy 2

Integrating
Ez  3 w  3 w 3D   z    3 w  3 w 
2
z
 xz    (  )dz  1     3  
2 
 h / 2 1  2 x 3 xy 2 2h   h / 2    
 x x y 

Similarly:

3D   z     3 w  3 w 
2

 yz  1      
2h   h / 2   y 3 x 2y 

The maximum transverse stresses are at the mid-plane of the plate (i.e., at z = 0):

3D   3 w  3 w  3D   3 w  3 w 
 xz max      yz max     (d)
2h  x 3 xy 2  2h  y 3 x 2 y 

14
Example 6: Transverse shear stresses in an edge-clamped circular
plate by a uniform lateral pressure (Example 2).

Submit the moments obtained in Example 2, into Eq.

(d), we get

3qx   z  
2

 xz   1    
4h   h / 2  

3qy   z  
2

 yz   1    
4h   h / 2  

The maximum transverse stresses are at the edge (x2 + y2 =

a2) and on the mid-plane of the plate (i.e., for z = 0):

3aq 3aq
 xz max   ,  yz max  
4h 4h

Note that the maximum in-plane stress has been

obtained in Example (2) and is
2
3q  a 
x    
4 h
Thus the transverse shear stress is of an order h / a smaller than the
normal stress. For very thin plate, the transverse shear stress can
be ignored.

15
Chapter 17. Buckling of plates

17.1 Introduction

* Buckling is characterized by a sudden increase of deflection of

a structural member;

* As the applied load increases, the deflection will ultimately

become large enough to cause the structure member to become
unstable and the stresses in the structure member become very
high. The structure member is said to have buckled.

* Plate buckling may occur when a structure is subjected to

compressive or shear stress;

* Buckling may occur under the stresses that are well below
those needed to cause failure of the material of the structure.

1
17.2 In-plane external forces at the plate

Consider the plate under the application of external forces Nx, Ny and Nxy. The
unit of these force is Force per unit Length (e.g., N / m) . These forces are
parallel to the x axis or the y axis and perpendicular to the z axis.

Ny (*) The in-plane force intensities Nx, Ny and Nxy are

shown acting on the sides of the element, with positive
Nyx
forces acting in the positive directions on positive faces.
z
y (*) These forces are acting on the mid-plane of the plate
Nx Nxy
o x (*) These forces have dimensions of force per unit length
Nxy Nx (e.g., N / m).
Nyx

Ny

(b) In-plane external forces on the plate

The direction of forces in

this figure are all positive

2
17.3 Buckling equilibrium equation of the plate

Like buckling of beam, for the buckling of plate, the deformed state
must be considered when establishing the equilibrium equations. dx
Nx
w
x
Deformed state
Ny

After plate deformation, Nx produces a bending

Nx moment about the y axis (note that when studying
plate bending, the “moment” is actually the
moment per unit length).

From the free body diagram above, the moment is

found to be

dy (M x )0  N x  w  N x w (a)

Initial state Similarly, Ny produces a moment (per unit length)

dx about the negative x axis. It is

z y
(M y )0  N y w (b)
x
o
3
 Deformed state
Nxy
Nxy
w
y

After plate deformation, Nxy produces a torque moment

(per unit length) about the negative x axis.

From the free body diagram above, the torque moment is

dy found to be

(M xy )0  N xy  w  N xy w (c)
Initial state
dx

z y

x
o

4
Note that these moments are due to the application of the in-plane
forces Nx, Ny and Nxy. These should be added to the moments of the
plate due to the bending of the plate, to form the total moments in  2w 2w 
M x   D 2   2 
the plate.  x y 

Therefore, by adding Eqs. (a), (b) and (c) to Eq. (15.6b), one get the  2w 2w  (15.6b)
M y   D  2  2 
total moments in the plate as:  x y 

 2w 2w  2w

M x   D 2   2   N x w M xy  (1  ) D
xy
 x y 

 2w 2w  (17.1)

M y   D  2  2   N y w
 x y 

2w
M xy  (1  ) D  N xy w
xy 2M x  2 M xy  2 M y
2  0 (15.3)
x 2 xy y 2
Putting Eq. (17.1) into the equilibrium of the moments of the plate
(static load, without lateral distribution load), Eq. (15.3), we get

 4w 4w 4w  2w 2w 2w

D 4  2 2 2  4   N x 2  2 N xy  Ny 2  0 (17.2)
 x x y y  x xy y

5
17.4 Buckling of a rectangular plate

(A). A rectangular plate, four sides are simply supported

(pined), under compression on two opposite sides.

Note that Ny = 0, Nxy = 0 and Nx =  P.

Assume the solution is :

 
 mx   ny 
w( x, y )   Cmn sin  sin  (a)
m 1 n 1  a   b 

Submit Eq. (a) in Eq. (17.2), one obtains (can you obtain???):

2
  m  2  n  2 
  
 m 
2

D Cmn       P  Cmn  

m 1 n 1  a   b   m 1 n 1  a 

  m 2 2 2 2
 
    n    m  

m 1 n 1
Cmn  D        P
  a   b  
 0
 a  
(C)
 

 4w 4w 4w  2w 2w 2w

D 4  2 2 2  4   N x 2  2 N xy  Ny 2  0 (17.2)
 x x y y  x xy y
6
Non-vanishing solution of Cmn exists only if the load P
increases to a critical value Pcr so that:
2
 m  2  n  2   m 
2

D       Pcr   0
 a   b    a 

2
D 2 2   a n  
2

Pcr  2 m 1    
a   b m  

Obviously, when n = 1, Pcr has the minimum value and is:

2
D 2  1 a 
2

Pcr  2 m     (17.3)
a  m  b  

Notice: the units of P and Pcr are force per unit length (i.e., N / m). Hence the total load applied
on the edge of the plate is F = P  b, Fcr = Pcr  b.

7
First order buckling mode

Second order buckling mode

Third order buckling mode

8
 Example 1: A plate of aluminum (E = 72 GPa, =0.3) is
pin supported at all edges and is subjected to compressive
loads at two opposite edges (see left figure). The plate has
length a =1800 mm, width b =1000 mm and thickness h =
100 mm. Determine the buckling load of the plate.
2
D2  1 a 
2

Solution: We use millimeter (mm), Newton (N), MPa units. Pcr  2 m     (17.3)
a  m  b  

Eh 3 72(1000) 1003
D   6.5934 109 Nmm
12(1   )
2
12  (1  0.3 )
2

From Eq. (17.3), we get

2 2
D 2  1 a  6.5934 109  2  1  1800  
2 2 2
 3.24 
Pcr  2 m      m      20084.6m 
a  m  b   18002  m  1000    m 

If m = 1, Pcr  20084.61 
3.24 
 361072.9 N/mm
 1  The lowest value of Pcr for all values of
2 m is the critical buckling load
If m = 2, Pcr  20084.62  3.24   263196.6 N/mm
 2 
2
If m = 3, Pcr  20084.63  3.24   334336.3 N/mm
 3 

Therefore the critical buckling load is identified as Pbuckling=263196.6 N / mm.

9
Example 2: A steel plate that is simply supported at four
edges and subjected to a compressive load P = 400 kN / mm
at its two opposite edge. The length and the width of the plate
are a = 2.0 m and b = 0.7 m, respectively. Find the minimum
required thickness h. (see the right figure, use E = 200 GPa
and =0.3)
2
D2  1 a 
2

Pcr  2 m     (17.3)
a  m  b  
Solution: We use millimeter (mm),
Newton (N), MPa units. From Eq.
Therefore, Pcr  80.75 106 D is the buckling load
(17.3), one obtains

400,000  80.75 106 D

If m = 1, Pcr  207.2 106 D
D  4.95294 109 Nmm
If m = 2, Pcr  91.27 106 D
Note:

If m = 3, Pcr  80.75 106 D Eh 3 200，000  h3

D   18315.02  h3
12(1   ) 12  (1  0.3 )
2 2

If m = 4, Pcr  90.04 106 D We get

4.95294 109
h 3  64.7 mm
18315.02

10
(B). A rectangular plate, four edges are simply supported (pined), under compressive load of magnitude P on
two opposite sides and compressive load of magnitude fP on the other two opposite sides.

fP
Note that Nx =  P, Ny =  fP and Nxy = 0.

Assume the solution is

P P b
 
 mx   ny 
w( x, y )   Cmn sin  sin 
m 1 n 1  a   b 

Substitute this into Eq. (17.2), one obtains

fP
a 2
   m  2  n  2 
D Cmn     
m 1 n 1  a   b  
   m 
2  
 n  
2

 P  Cmn   f  Cmn   

 m 1 n 1  a  m 1 n 1  b  

  m 2 2 2 2 
 
     
n  m  2  n   
 Cmn  D        P 
  a   b   
  f     0
  b   
m 1 n 1
 
 a


 4w 4w 4w  2w 2w 2w

D 4  2 2 2  4   N x 2  2 N xy  Ny 2  0 (17.2)
 x x y y  x xy y
11
 Non-vanishing solution of w should be such that the load P
increases to a critical value Pcr so that:

2
 m  2  n  2   m  2  n  
2

D       Pcr    f  0
 a   b    a   b  

2 2
 m  2  n  2   m  2  n  2 
D      D      
2

Pcr   a   b     a   b  

 m 
2
 n 
2 2 2
m n
   f     f 
 a   b  a b

If f < 1, Pcr has the minimum value when n = 1 and is:

2

2 m
2
1 
2

D      
 a   b  
Pcr  2 2
(17.4)
m 1
   f 
a b

* The values of m for minimum critical load depend of the b/a value of the plate.

* The buckling load of the plate is the lowest value of Pcr calculated by Eq. (17.4)
for all values of m.

12
Example 3: A plate of aluminum (E = 72,000 MPa,  = 0.3) is pin supported at all edges.
The plate has length a = 1800 mm, width b = 1000 mm and thickness h = 100 mm. Suppose
the value of f is f = 0.5 (see below figure). Determine the buckling load of the plate Pcr.

0.5P
Solution: We use millimeter (mm), Newton
(N), MPa units.

Eh 3 72000 1003
D   6.5934 109 Nmm
P P b 12(1   ) 12  (1  0.3 )
2 2

Using Eq. (17.4):

2

9 2  m 
2
 1  
2
0.5P 6.5934 10      
 1800   1000  
a Pcr  2 2
 m   1 
   0.5 
 1800   1000 

If m = 1, Pcr = 137814.4 N/mm;

If m = 2, Pcr = 187329.0 N/mm

2

2 m
2
1 
2
It can be seen that the buckling load is
D      
 a   b   Pcr = 137814.4 N / mm, and buckling
Pcr  2 2
(17.4)
m 1 occurs for m = 1 and n = 1.
   f 
a b

13
17.5 Conclusions

In theoretical development, only simply supported plates and uniform compressive

load are considered. For complicated geometries (structures), loading (e.g., shearing)
and boundary conditions, numerical methods such as FEM have to be used.

2
D 2  1 a 
2

Pcr  2 m     (17.3)
a  m  b  

fP

2

2 m
2
1 
2

P P b D      
 a   b  
Pcr  2 2
(17.4)
m 1
   f 
a b
fP
a

14
Chapter 18 Free vibration of plate

18.1 Introduction

•Finding the natural frequencies of the plate is important in the design.

•The natural frequencies of the plate must not be equal to the frequency of the applied
force, to avoid the resonance.

Equilibrium equation of plate has been developed in Chapter 15 as Eq. (15.3). In the
absence of external load, it becomes

2M x  2 M xy  2 M y 2w
2   h 2 (18.1)
x 2 xy y 2 t
Inertia term
Constitutive equations were also developed in Chapter 15 as

 2w 2w 
M x   D 2   2 

 x y 

 2w 2w 
M y   D  2  2  (15.6b)
 x y 

2w
M xy  (1  ) D
xy

Putting Eq. (15.6b) into Eq. (18.1) results in

 4w 4w 4w  2w

D 4  2 2 2  4   h 2  0 (18.2)
 x x y y  t

1
18.2 Free vibration of a rectangular plate. The four edges of the plate are simply supported (pined)

x
b  4w 4w 4w  2w
D 4  2 2 2  4   h 2  0 (18.2)
 x x y y  t
a
y

Assume the solution is the sum of a series:

 
 mx   ny 
w( x, y )   Cmn sin  sin  A sin(t )  B cos(t )  (a)
m 1 n 1  a   b 

where Cmn, A and B are constants, determined by the boundary and initial conditions.
Putting this into Eq. (18.2) yields:

  m 2 2 2 
 
    n   2

m 1 n 1
Cmn  D        h   0
  a   b   
(b)
 

Non-vanishing solution of w should be such that

D  m   n  
2
 m  2  n  2  2 2

D        h  0
2
     
 a   b   h  a   b  

2
For different vibration wave shapes (i.e., w for different m and n), the values of
the vibration frequencies are different.

D  m   n  
2 2

mn       (18.3)
h  a   b  

The correspnding vibration modes are

 
 mx   ny 
( x, y )   Cmn sin  sin 
m 1 n 1  a   b 

It can be seen that the vibration frequency increases with m and n. The lowest
frequency is for m = 1 and n = 1, and is

D       
2 2

11       
h  a   b  

11 is the fundamental frequency.

3
For different vibration wave shapes (i.e., w for different m and n), the
corresponding vibration frequencies are different. 4
 Example 1:A plate of steel has E = 210 GPa, volume density
x
(mass per cubic metre) 7850 kg / m3 and  = 0.3. The plate is
b
pin supported at all edges. The plate has length a = 1800 mm,
width b = 1000 mm and thickness h = 25 mm. Obtain the first
a
four frequencies of the plate. What is the fundamental y
frequency of the plate?
 1  2  1  2  
Solution: We use N, m, kg, Pa units. 11  386.192       505.39 rad/s , f11  11  80.43 Hz
 1.8   1.0   2
Eh 3 210 109  0.0253
D 
12(1  2 ) 12  (1  0.32 )  1  2  2  2  
12  386.192       1663.96 rad/s , f12  12  264.83 Hz
 3.00481105 Nm  1.8   1.0   2

Using Eq. (18.3), one gets  2  2  1  2  

21  386.192       862.97 rad/s , f 21  21  137.35 Hz
 1.8   1.0   2
3.00481105  m  2  n  2 
mn      
7850  0.025    1.0    3  2  1  2  
 1.8
31  386.192       1458.95 rad/s , f 31  21  232.20 Hz
 m  2  n  2   1.8   1.0   2
 386.192     
 1.8   1.0    2  2  2  2  
22  772.384       2021.55 rad/s , f 22  22  321.74 Hz
2
.  1.8   1.0  
.
.
D  m   n  
2 2
The fundamental frequency is 11  505.39 rad/s , or f11  80.43 Hz
mn       (18.3)
h  a   b  
5
Ansys solution

6
Example 2: An aluminum plate has length a = 2 m and width b = 1.5 m. The design requires that the
fundamental frequency of the plate f11 must be lower than 200 Hz. Give the maximum thickness of the
plate h. Take the volume density of the plate as 2700 kg / m3 and E = 72 GPa, =0.3.

Solution: Use N, m, kg, Pa units. Note that the fundamental frequency is (based on Eq. (18.3))

11 1 D       
2 2
D
f11          0.02099
2 2 h  a   b   h
D  m   n  
2 2

mn       (18.3)
Since h  a   b  
f11  200Hz

We get
Remember: For natural frequency calculation,
0.02099 D / h  200
always use N, m, kg, Pa units.
D  9.079(107 )h

Eh 3 72 109  h3
D   6.593 109  h3
12(1   )
2
12(1   )
2

9.079(107 )h  6.593 109  h3

9.079(107 )
h  2

6.593 109
h  0.117 m  117 mm
7