CIVL 7/8117 Chapter 16 - Structural Dynamics 1/85

Chapter 16 – Structural Dynamics

Learning Objectives
• To discuss the dynamics of a single-degree-of
freedom spring-mass system.

• To derive the finite element equations for the time-

dependent stress analysis of the one-dimensional
bar, including derivation of the lumped and
consistent mass matrices.

• To introduce procedures for numerical integration in

time, including the central difference method,
Newmark's method, and Wilson's method.

• To describe how to determine the natural

frequencies of bars by the finite element method.

• To illustrate the finite element solution of a time-

dependent bar problem.

Chapter 16 – Structural Dynamics

Learning Objectives
• To develop the beam element lumped and
consistent mass matrices.

• To illustrate the determination of natural

frequencies for beams by the finite element
method.

• To develop the mass matrices for truss, plane

frame, plane stress, plane strain, axisymmetric, and
solid elements.

• To report some results of structural dynamics

problems solved using a computer program,
including a fixed-fixed beam for natural frequencies,
a bar, a fixed-fixed beam, a rigid frame, and a
gantry crane-all subjected to time-dependent
forcing functions.
CIVL 7/8117 Chapter 16 - Structural Dynamics 2/85

Structural Dynamics
Introduction
This chapter provides an elementary introduction to time-
dependent problems.

We will introduce the basic concepts using the single-

degree-of-freedom spring-mass system.

We will include discussion of the stress analysis of the one-

dimensional bar, beam, truss, and plane frame.

Structural Dynamics
Introduction
We will provide the basic equations necessary for structural
dynamic analysis and develop both the lumped- and the
consistent-mass matrices involved in the analyses of a bar,
beam, truss, and plane frame.

We will describe the assembly of the global mass matrix for

truss and plane frame analysis and then present numerical
integration methods for handling the time derivative.

We will provide longhand solutions for the determination of the

natural frequencies for bars and beams, and then illustrate the
time-step integration process involved with the stress analysis
of a bar subjected to a time dependent forcing function.
CIVL 7/8117 Chapter 16 - Structural Dynamics 3/85

Structural Dynamics
Dynamics of a Spring-Mass System
In this section, we will discuss the motion of a single-degree-of-
freedom spring-mass system as an introduction to the
dynamic behavior of bars, trusses, and frames.

Consider the single-degree-of-freedom spring-mass system

subjected to a time-dependent force F(t) as shown in the
figure below.

The term k is the stiffness of the spring and m is the mass of the
system.

Structural Dynamics
Dynamics of a Spring-Mass System
The free-body diagram of the mass is shown below.

The spring force T = kx and the applied force F(t) act on the
mass, and the mass-times-acceleration term is shown
separately.
Applying Newton’s second law of motion, f = ma, to the mass,
we obtain the equation of motion in the x direction:

F (t )  kx  mx
where a dot ( • ) over a variable indicates differentiation with
respect to time.
CIVL 7/8117 Chapter 16 - Structural Dynamics 4/85

Structural Dynamics
Dynamics of a Spring-Mass System
The standard form of the equation is: mx  kx  F (t )
The above equation is a second-order linear differential
equation whose solution for the displacement consists of a
homogeneous solution and a particular solution.

The homogeneous solution is the solution obtained when the

right-hand-side is set equal to zero.

A number of useful concepts regarding vibrations are available

when considering the free vibration of a mass; that is when
F(t) = 0.

Structural Dynamics
Dynamics of a Spring-Mass System
k
Let’s define the following term:  2 
m

The equation of motion becomes: x   2 x  0

where  is called the natural circular frequency of the free

vibration of the mass (radians per second).

Note that the natural frequency depends on the spring stiffness

k and the mass m of the body.
CIVL 7/8117 Chapter 16 - Structural Dynamics 5/85

Structural Dynamics
Dynamics of a Spring-Mass System

The motion described by the homogeneous equation of motion

is called simple harmonic motion. A typical displacement -
time curve is shown below.

where xm denotes the maximum displacement (or amplitude

of the vibration).

Structural Dynamics
Dynamics of a Spring-Mass System

The time interval required for the mass to complete one full
cycle of motion is called the period of the vibration  (in
seconds) and is defined as:
2



The frequency in hertz (Hz = 1/s) is f = 1/ =  /(2).

CIVL 7/8117 Chapter 16 - Structural Dynamics 6/85

Structural Dynamics
Dynamics of a Spring-Mass System

Structural Dynamics
Dynamics of a Spring-Mass System
CIVL 7/8117 Chapter 16 - Structural Dynamics 7/85

Structural Dynamics
Dynamics of a Spring-Mass System

Structural Dynamics
Dynamics of a Spring-Mass System
CIVL 7/8117 Chapter 16 - Structural Dynamics 8/85

Structural Dynamics
Dynamics of a Spring-Mass System

Structural Dynamics
Dynamics of a Spring-Mass System
CIVL 7/8117 Chapter 16 - Structural Dynamics 9/85

Structural Dynamics
Direct Derivation of the Bar Element

Let’s derive the finite element equations for a time-dependent

(dynamic) stress analysis of a one-dimensional bar.

Step 1 - Select Element Type

We will consider the linear bar element shown below.

where the bar is of length L, cross-sectional area A, and mass

density  (with typical units of lb-s2/in4), with nodes 1 and 2
subjected to external time-dependent loads: fx e (t )

Structural Dynamics
Direct Derivation of the Bar Element

Step 2 - Select a Displacement Function

A linear displacement function is assumed in the x direction.

u  a1  a2 x
The number of coefficients in the displacement function, ai, is
equal to the total number of degrees of freedom associated
with the element.
We can express the displacement function in terms of the shape
functions:
u  x x
u  N1 N2    1  N1  1  N2 
u2  L L
CIVL 7/8117 Chapter 16 - Structural Dynamics 10/85

Structural Dynamics
Direct Derivation of the Bar Element

Step 3 - Define the Strain/Displacement and Stress/Strain

Relationships
The stress-displacement relationship is:
du
 x    [B ] d 
dx

 1 1  u1 
where: [B ]    d    
 L L  u2 
The stress-strain relationship is given as:

 x   [D] x   [D][B ]d 

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

The bar element is typically not in equilibrium under a time-

dependent force; hence, f1x ≠ f2x.

We must apply Newton’s second law of motion, f = ma, to each

node.

Write the law of motion as the external force fxe minus the
internal force equal to the nodal mass times acceleration.
CIVL 7/8117 Chapter 16 - Structural Dynamics 11/85

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

Therefore:  2u1  2u2

f1x e  f1x  m1 f2 x e  f2 x  m2
t 2 t 2
where:
 AL  AL
m1  m2 
2 2
Mass lumped to the nodes

  2u1 
 f1x e   f1x  m1 0   t 2 
 e      
f2 x  f2 x   0 m2    2u2 
 t 2 

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

If we replace {f} with [k]{d} we get: f (t )  k d  m d

e

Where the elemental stiffness matrix is:

 1 1  2 d 
AE
k   L  1 1 
 
 
d 
t 2

and the lumped-mass matrix is:

 AL  1 0 
m   2 0 1
CIVL 7/8117 Chapter 16 - Structural Dynamics 12/85

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

Let’s derive the consistent-mass matrix for a bar element.

The typical method for deriving the consistent-mass matrix is

the principle of virtual work; however, an even simpler
approach is to use D’Alembert’s principle.

The effective body force is:  X     u

e

The nodal forces associated with {Xe} are found by using the
following:
fb    [N ]T { X } dV
V

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

Substituting {Xe} for {X} gives: fb      [N ]T u dV
V

The second derivative of the u with respect to time is:

u  [N ]d u  [N ]d

where u and u are the nodal velocities and accelerations,
respectively.
CIVL 7/8117 Chapter 16 - Structural Dynamics 13/85

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

fb      N  N d dV   m d
T
Therefore:
V

m     N  N  dV
T
where:
V

The mass matrix is called the consistent mass matrix because

it is derived using the same shape functions use to obtain the
stiffness matrix.

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

Substituting the shape functions in the above mass matrix
equations give:
 x
1  L   x x 
m      x  1  L L  dV
V   
 L 
 x
L 1  L   x x
m   A   x  1  L L 
dx
0 
 L 
CIVL 7/8117 Chapter 16 - Structural Dynamics 14/85

Structural Dynamics
Direct Derivation of the Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

Substituting the shape functions in the above mass matrix
equations give:
 x
2
 x  x
  1    1   
 L  L L
L

m    A  dx
0 
2 
 1  x  x  x 
L 
 L  L   
Evaluating the above integral gives:
 AL 2 1
m   6  1 2 

Structural Dynamics
Direct Derivation of the Bar Element

Step 5 - Assemble the Element Equations and Introduce

Boundary Conditions
The global stiffness matrix and the global force vector are
assembled using the nodal force equilibrium equations, and
force/deformation and compatibility equations.

F (t )  K  {d }  M d
where

F    f ( e ) 
N N N
K    k (e )  M    m( e ) 
e 1 e 1 e 1
CIVL 7/8117 Chapter 16 - Structural Dynamics 15/85

Structural Dynamics
Numerical Integration in Time
We now introduce procedures for the discretization of the
equations of motion with respect to time.

These procedures will allow the nodal displacements to be

determined at different time increments for a given dynamic
system.

The general method used is called direct integration. There

are two classifications of direct integration: explicit and implicit.

We will formulate the equations for two direct integration

methods.

Structural Dynamics
Numerical Integration in Time
The first, and simplest, is an explicit method known as the
central difference method.

The second more complicated but more versatile than the

central difference method, is an implicit method known as the
Newmark-Beta (or Newmark’s) method.

The versatility of Newmark’s method is evidenced by its

adaptation in many commercially available computer
programs.
CIVL 7/8117 Chapter 16 - Structural Dynamics 16/85

Structural Dynamics
Central Difference Method
The central difference method is based on finite difference
expressions for the derivatives in the equation of motion.

For example, consider the velocity and the acceleration

at time t:

d   
d i 1  d i 1
2( t )
i

d   d 
 
di 
i 1

2( t )
i 1

where the subscripts indicate the time step for a given time
increment of t.

Structural Dynamics
Central Difference Method
The acceleration can be expressed in terms of the
displacements (using a Taylor series expansion).

The first two terms of Taylor series of a function d(t) is:

d
d i 1  d i  di t  i  t   
2

2
Solving for the acceleration gives:


di 

2 d i 1  d i  di t 
d   
 t 
2
d i 1  d i 1
2( t )
i

Substitute the central difference for the first derivation

CIVL 7/8117 Chapter 16 - Structural Dynamics 17/85

Structural Dynamics
Central Difference Method
Therefore, accelerations can be expressed in terms of the
displacements as:

d   
d i 1  2 d i   d i 1
( t )2
i

We generally want to evaluate the nodal displacements;

therefore, we rewrite the above equation as:

di 1  2di   di 1  di  ( t )2

The acceleration can be expressed as:

d   M F   K d 

i
1
i i

Structural Dynamics
Central Difference Method
To develop an expression of di+1, first multiply the nodal
displacement equation by M and substitute the above equation
 
for di into this equation.

Mdi 1  2 Mdi   Mdi 1  Fi   K di   t 

2

Combining terms in the above equations gives:

Mdi 1   t  Fi   2 M   t  K  di   Mdi 1

2 2

 
To start the computation to determine d i 1 , di 1 , and di 1  
we need the displacement at time step i -1.
CIVL 7/8117 Chapter 16 - Structural Dynamics 18/85

Structural Dynamics
Central Difference Method
Using the central difference equations for the velocity and
acceleration:

d   
d i 1  d i 1
 di 1  2( t )di   di 1
2( t )
i

di 1  2di   di 1  di  ( t )2

Solving for {di-1} gives:
( t )2
d i 1  d i   ( t )di   di 
2

Structural Dynamics
Central Difference Method

Procedure for solution:

1. Given: d0 , d0 , and Fi (t )

2. If the acceleration is not given, solve for d0  

d   M F   K d 
0
1
0 0

3. Solve for {d-1} at t = -t

( t )2
    
d 1  d0   ( t ) d0  d0

2
CIVL 7/8117 Chapter 16 - Structural Dynamics 19/85

Structural Dynamics
Central Difference Method

Procedure for solution:

4. Solve for {d1} at t = t using the value of {d-1} from Step 3

Mdi 1   t  Fi   2 M   t  K  di   Mdi 1

2 2

1

d1  M  t  F0   2 M   t  K  d0   Md 1
2 2


Structural Dynamics
Central Difference Method

Procedure for solution:

5. With {d0} given and {d1} determined in Step 4 solve for {d2}
1

d 2   M  t  F1  2 M   t  K  d1  Md0 
2 2

 
6. Solve for d1 : d   M F   K d 
1
1
1 1

d   
d 2   d0 
 
7. Solve for d1 : 1
2( t )

8. Repeat Steps 5, 6, and 7 to obtain the displacement,

acceleration, and velocity for other time steps.
CIVL 7/8117 Chapter 16 - Structural Dynamics 20/85

Structural Dynamics
Central Difference Method

Procedure for solution:

Structural Dynamics
Central Difference Method – Example Problem
Determine the displacement, acceleration, and velocity at 0.05
second time intervals for up to 0.2 seconds for the one-
dimensional spring-mass system shown in the figure below.

Consider the above spring-mass system as a single degree of

freedom problem represented by the displacement d.
CIVL 7/8117 Chapter 16 - Structural Dynamics 21/85

Structural Dynamics
Central Difference Method – Example Problem

Procedure for solution:

1. At time t = 0: d0   0 d   0
0

2. If the acceleration is not given, solve for d0  

d   M F   K d 
0
1
0 0

 100(0)
d   2,00031.83
0  62.83 in
s 2

Structural Dynamics
Central Difference Method – Example Problem

Procedure for solution:

3. Solve for {d-1} at t = -t
( t )2
    
d 1  d0   ( t ) d0  d0

2
(0.05)2
d 1  0  (0.05)0  (62.83)  0.0785 in
2
4. Solve for {d1} at t = 0.05 s using the value of {d-1} from Step 3
1 2

d1  M  t  F0   2 M   t  K  d0   Md 1
2

d1 
1
31.82

 0.05  2,000  2  31.83    0.05  100   0   31.83  0.0785 
2 2

 0.0785 in
CIVL 7/8117 Chapter 16 - Structural Dynamics 22/85

Structural Dynamics
Central Difference Method – Example Problem

Procedure for solution:

5. With {d0} given and {d1} determined in Step 4 solve for {d2}
d 2   M  t  F1  2 M   t  K  d1  Md0 
1 2 2

d 2  
1
31.82
 
 0.05  1,500  2  31.83    0.05  100   0.0785    31.83  0 
2 2

 0.274 in

6. Solve for d1 :  d   M F   K d 

1
1
1 1

d   31.83
1
1
1,500  100  0.0785    46.88 in
s 2

Structural Dynamics
Central Difference Method – Example Problem

Procedure for solution:

d   
d 2   d 0 
7. Solve for d1 :   1
2( t )
0
d   0.274
1
2  0.05 
 2.74 in
s

8. Repeat Steps 5, 6, and 7 to obtain the displacement,

acceleration, and velocity for other time steps.
CIVL 7/8117 Chapter 16 - Structural Dynamics 23/85

Structural Dynamics
Central Difference Method – Example Problem

Procedure for solution:

5. With {d1} given and {d2} determined in Step 4 solve for {d3}
d3   M  t  F2   2 M   t  K  d2   Md1
1 2 2

d3  
1
31.82
 
 0.05  1,000  2  31.83    0.05  100   0.274    31.83  0.0785 
2 2

 0.546 in

6. Solve for d2 :   d   M F   K d 

2
1
2 2

d   31.83
2
1
1,000  100  0.274    30.56 in
s 2

Structural Dynamics
Central Difference Method – Example Problem

Procedure for solution:

d   
d3   d1
7. Solve for d2 :   2
2( t )
 0.0785
d   0.546
2
2  0.05 
 4.68 in
s

8. Repeat Steps 5, 6, and 7 to obtain the displacement,

acceleration, and velocity for other time steps.
CIVL 7/8117 Chapter 16 - Structural Dynamics 24/85

Structural Dynamics
Central Difference Method – Example Problem

The following table summarizes the results for the remaining

time steps as compared with the exact solution.

t (s) F(t) (lb) di (in/s2) di (in/s) di (in) di (exact)

0.00 2,000 62.8338 0.0000 0.0000 0.0000
0.05 1,500 46.8786 2.7428 0.0785 0.0718
0.10 1,000 30.5552 4.6787 0.2743 0.2603
0.15 500 13.9918 5.7923 0.5464 0.5252
0.20 0 -2.6815 6.0751 0.8535 0.8250
0.25 0 -3.6252 5.9174 1.1539 1.132

Structural Dynamics
Central Difference Method – Example Problem

Plotting the motion for about 4 s gives:

70.000

Velocity
60.000 Acceleration
Position
50.000

40.000

30.000

20.000

10.000

0.000
0 0.5 1 1.5 2 2.5 3 3.5 4

-10.000

-20.000
CIVL 7/8117 Chapter 16 - Structural Dynamics 25/85

Structural Dynamics
Newmark’s Method
Newmark’s equations are given as:

d   d   (t ) (1   )d    d 

i 1 i i i 1

d i 1  d i   ( t )di   ( t )2  21    di    di 1

where  and  are parameters.

The parameter  is typically between 0 and ¼, and  is often

taken to be ½.

For example, if  = 0 and  = ½ the above equation reduce to

the central difference method.

Structural Dynamics
Newmark’s Method
To find {di+1} first multiply the above equation by the mass matrix
[M] and substitute the result into this the expression for
acceleration. Recall the acceleration is:
 
d  M F   K d  
i
1
i i

The expression [M]{di+1} is:

Mdi 1  Mdi   ( t ) Mdi   ( t )2 M  21    di 
  ( t )2 Fi 1  K d i 1
Combining terms gives:
M   (t ) K d    ( t ) F   Md 
2
i 1
2
i 1 i

 
( t ) M di  ( t )2 M  21    di  
CIVL 7/8117 Chapter 16 - Structural Dynamics 26/85

Structural Dynamics
Newmark’s Method
Dividing the above equation by (Δt)2 gives: K'd i 1  F'i 1
where:
1
K '  K    ( t )2 M
 M
F 'i 1  Fi 1     
d i   ( t ) di   21    ( t )2 di 
 ( t ) 
2 

The advantages of using Newmark’s method over the central

difference method are that Newmark’s method can be made
unconditionally stable (if  = ¼ and  = ½) and that larger time
steps can be used with better results.

Structural Dynamics
Newmark’s Method

Procedure for solution:

CIVL 7/8117 Chapter 16 - Structural Dynamics 27/85

Structural Dynamics
Newmark’s Method

Procedure for solution:

0  
1. Given: d , d , and F (t )
0 i

2. If the acceleration is not given, solve for d0  

d   M F   K d 
0
1
0 0

3. Solve for {d1} at t = 0

K'd1  F'1

Structural Dynamics
Newmark’s Method

Procedure for solution:

 
4. Solve for d1 -- original Newmark equation ford i 1 rewritten
 
for di 1 :

d    (1t ) d   d   (t )d   (t ) 

1 2 1 0 0
2 1
2  
   d0 


 
5. Solve for d1

d   d   (t ) (1   )d    d 

1 0 0 1

6. Repeat Steps 3, 4, and 5 to obtain the displacement,

acceleration, and velocity for the next time step.
CIVL 7/8117 Chapter 16 - Structural Dynamics 28/85

Structural Dynamics
Newmark’s Method – Example Problem
Determine the displacement, acceleration, and velocity at 0.1
second time intervals for up to 0.5 seconds for the one-
dimensional spring-mass system shown in the figure below.

Consider the above spring-mass system as a single degree of

freedom problem represented by the displacement d.

Use Newmark’s method with  = 1/6 and  = ½.

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

 
1. Given: d 0 , d0 , and Fi (t )

2. If the acceleration is not given, solve for d0  

d   M F   K d 
0
1
0 0

1

1.77
100  70(0)  56.5 in / s 2
CIVL 7/8117 Chapter 16 - Structural Dynamics 29/85

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

3. Solve for {d1} at t = 0.1 s K'd1  F'1

1 1
K '  K    (t )2 M  70  1
(0.1)2
(1.77)  1,132 lb / in
6

M
F '1  F1 
 ( t ) 
2    
d0   ( t ) d0   21    ( t )2 d0 


1.77
 80  0  (0.1)0   21  61  (0.1)2  56.5    280 lb
2 
1
6 (0.1)
F '1  280 lb
d1   0.248 in
K ' 1,132 lb in

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

 
4. Solve for d1 at t = 0.1 s

d    (1t )
1 2     
d1  d0   ( t ) d0  ( t )2  21    d0 

1
 0.248  0  (0.1)0  (0.1)2  21  61  56.5   35.4 in s2
2 
1
6 (0.1)

 
5. Solve for d1

d   d   (t ) (1   )d    d 

1 0 0 1

 0  (0.1) (1  21 )56.5  21  35.4    4.59 in s

CIVL 7/8117 Chapter 16 - Structural Dynamics 30/85

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

6. Repeat Steps 3, 4, and 5 to obtain the displacement,
acceleration, and velocity for the next time step.
Repeating Steps 3, Solve for {d1} at t = 0.2 s
M
F '2   F2  
 ( t ) 2    
d1  ( t ) d1   21    ( t )2 d1 


1.77
 60  0.248  (0.1)4.59   21  61  (0.1)2  35.4    934 lb
1
6 (0.1)2 

F '2   934  0.825 in

d2  
K ' 1,132

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

Repeating Step 4: solve for d1 at t = 0.2 s  
d    (1t )
2 2     
d 2   d1  ( t ) d1  ( t )2  21    d1 


1
 0.825  0.248  (0.1)4.59  (0.1)2  21  61  35.4 
1
6 (0.1)2 

 1.27 in s2
CIVL 7/8117 Chapter 16 - Structural Dynamics 31/85

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

 
5. Solve for d2

d   d   (t ) (1   )d    d 

2 1 1 2

 4.59  (0.1) (1  21 )35.4  21 1.27  

 6.42 in s

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

The following table summarizes the results for the time steps
through t = 0.5 seconds.
t (s) F(t) lb di (in/s2) di (in/s) di (in)
0.0 100 56.4972 0.0000 0.0000
0.1 80 35.4155 4.5956 0.2473
0.2 60 1.1939 6.4261 0.8270
0.3 48.5714 -28.9276 5.0394 1.4253
0.4 45.7143 -43.7731 1.4044 1.7599
0.5 42.8571 -42.3776 -2.9031 1.6838
CIVL 7/8117 Chapter 16 - Structural Dynamics 32/85

Structural Dynamics
Newmark’s Method – Example Problem

Procedure for solution:

Plotting the motion for about 4 s gives:
80.000

Velocity
60.000
Acceleration
Position

40.000

20.000

0.000
0 0.5 1 1.5 2 2.5 3 3.5 4

-20.000

-40.000

-60.000

Structural Dynamics
Natural Frequencies of a One-Dimensional Bar
Before solving the structural stress dynamic analysis problem,
let’s consider how to determine the natural frequencies of
continuous elements.

Natural frequencies are necessary in vibration analysis and

important when choosing a proper time step for a structural
dynamics analysis.

Natural frequencies are obtained by solving the following

equation:

M d  K d   0
CIVL 7/8117 Chapter 16 - Structural Dynamics 33/85

Structural Dynamics
Natural Frequencies of a One-Dimensional Bar
The standard solution for {d} is given as: d (t )  d ' eit

where {d } is the part of the nodal displacement matrix called

natural modes that is assumed to independent of time, i is the
standard imaginary number, and  is a natural frequency.

Differentiating the above equation twice with respect to time

gives:

d  d '   2  e it


Substituting the above expressions for {d} and d into the
equation of motion gives:

 M  2 d ' e it  K d ' e it  0

Structural Dynamics
Natural Frequencies of a One-Dimensional Bar

Combining terms gives: e it K    2 M d '  0
Since eit is not zero, then: K   M  0
2

The above equations are a set of linear homogeneous

equations in terms of displacement mode {d }.

There exists a non-trivial solution if and only if the determinant

of the coefficient matrix of is zero.

K   2 M  0
CIVL 7/8117 Chapter 16 - Structural Dynamics 34/85

Structural Dynamics
One-Dimensional Bar - Example Problem
Determine the first two natural frequencies for the bar shown in
the figure below.

Assume the bar has a length 2L, modulus of elasticity E, mass

density , and cross-sectional area A.

Structural Dynamics
One-Dimensional Bar - Example Problem
Let’s discretize the bar into two elements each of length L as
shown below.

We need to develop the stiffness matrix and the mass matrix

(either the lumped- mass of the consistent-mass matrix).

In general, the consistent-mass matrix has resulted in solutions

that compare more closely to available analytical and
experimental results than those found using the lumped-mass
matrix.
CIVL 7/8117 Chapter 16 - Structural Dynamics 35/85

Structural Dynamics
One-Dimensional Bar - Example Problem
Let’s discretize the bar into two elements each of length L as
shown below.

However, when performing a long hand solution, the

consistent-mass matrix is more difficult and tedious to
compute; therefore, we will use the lumped-mass matrix
in this example.

Structural Dynamics
One-Dimensional Bar - Example Problem
Let’s discretize the bar into two elements each of length L as
shown below.

The elemental stiffness matrices are:

1 2 2 3

AE  1 1 AE  1 1
k(1)   k(2)  
L  1 1  L  1 1 
1 2 3

The global stiffness matrix is:  1 1 0 

K   L  1 2 1
AE 

 0 1 1 
CIVL 7/8117 Chapter 16 - Structural Dynamics 36/85

Structural Dynamics
One-Dimensional Bar - Example Problem
Let’s discretize the bar into two elements each of length L as
shown below.

The lumped-mass matrices are:

1 2 2 3

 AL  1 0   AL  1 0 
m(1)  
2 0 1 m(2)   2 0 1
1 2 3

The global lumped-mass matrix is: 1 0 0

 AL 
M   0 2 0
2  
0 0 1

Structural Dynamics
One-Dimensional Bar - Example Problem
Substituting the above stiffness and lumped-mass matrices into
the natural frequency equation:
 
K    2 M d '  0
and applying the boundary condition {u1} = 0 (or {d1} = 0)
gives:
 AE  2 1  AL 2 0   d '2  0 
    2     
 L  1 1  2 0 1  d '3  0 

Set the determinant of the coefficient matrix equal to zero as:

AE  2 1  AL 2 0 
 1 1    2 0 1  0 where  = 2
L    
CIVL 7/8117 Chapter 16 - Structural Dynamics 37/85

Structural Dynamics
One-Dimensional Bar - Example Problem
Dividing the above equation by AL and letting   E gives:
 L2
2   
2
 0  2  2  0
  2
2

Evaluating the determinant of the above equations gives:

  2   2 1  0.5858  2  3.4142

For comparison, the exact solution gives  = 0.616, whereas

the consistent-mass approach yields  = 0.648.

Structural Dynamics
One-Dimensional Bar - Example Problem

Therefore, for bar elements, the lumped-mass approach can

yield results as good as, or even better than, the results from
the consistent-mass approach.

However, the consistent-mass approach can be mathematically

proven to yield an upper bound on the frequencies, whereas
the lumped-mass approach has no mathematical proof of
boundedness.

The first and second natural frequencies are given as:

1  1  0.7654  2  2  1.8478 
CIVL 7/8117 Chapter 16 - Structural Dynamics 38/85

Structural Dynamics
One-Dimensional Bar - Example Problem

The term  may be computed as:

E 30  106 psi
   4.12  106 s-2
L 2 lb.s2
(0.00073 in.4 )(100in.)2

Therefore, first and second natural frequencies are:

1  1.56  103 rad / s 2  3.76  103 rad / s

Structural Dynamics
One-Dimensional Bar - Example Problem

In general, an n-degree-of-freedom discrete system has n

natural modes and frequencies.

A continuous system actually has an infinite number of natural

modes and frequencies.

The lowest modes and frequencies are approximated most

often; the higher frequencies are damped out more rapidly and
are usually less important.
CIVL 7/8117 Chapter 16 - Structural Dynamics 39/85

Structural Dynamics
One-Dimensional Bar - Example Problem
Substituting 1 into the following equation

 AE  2 1 2  AL  2 0   d '2  0 
  1 1    2 0 1  d '   0 
 L     3  

Gives: 1.4142 d '2(1)   d '3(1)  0

  d '2(1)  0.7071 d '3(1)  0
where the superscripts indicate the natural frequency.

It is customary to specify the value of one of the natural modes

{d } for a given i or i and solve for the remaining values.


For example, if d '(1)
3  
 1 than the solution for d '(1)
2 
 0.7071

Structural Dynamics
One-Dimensional Bar - Example Problem
 
Similarly, if we substitute 2 and let d '3  1 the solution of the
(2)

above equations gives d '(2)2 

 0.7071
The modal responses for the first and second natural
frequencies are shown in the figure below.

The first mode means that the bar is completely in tension or

compression, depending on the excitation direction.
CIVL 7/8117 Chapter 16 - Structural Dynamics 40/85

Structural Dynamics
One-Dimensional Bar - Example Problem
 
Similarly, if we substitute 2 and let d '3  1 the solution of the
(2)


above equations gives d '(2) 2  0.7071
The modal responses for the first and second natural
frequencies are shown in the figure below.

The second mode means that bar is in compression and tension

or in tension and compression.

Structural Dynamics
One-Dimensional Bar - Example Problem
 
Similarly, if we substitute 2 and let d '3  1 the solution of the
(2)


above equations gives d '(2) 2  0.7071
The modal responses for the first and second natural
frequencies are shown in the figure below.
CIVL 7/8117 Chapter 16 - Structural Dynamics 41/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
Consider the one-dimensional bar system shown in the figure
below.

Assume the boundary condition {d1x} = 0 and the initial

 
conditions {d0} = 0 and d0  0

Let  = 0.00073 lb-s2/in.4, A = 1 in.2, E = 30 x 106 psi, and

L = 100 in.

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
The bar will be discretized into two elements as shown below.

The elemental stiffness matrices are:

1 2 2 3

AE  1 1 AE  1 1
k(1)   k(2)  
L  1 1  L  1 1 
1 2 3

The global stiffness matrix is:  1 1 0 

K   L 1 2 1
AE 

 0 1 1 
CIVL 7/8117 Chapter 16 - Structural Dynamics 42/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
The bar will be discretized into two elements as shown below.

The lumped-mass matrices are:

1 2 2 3

 AL  1 0   AL  1 0 
m(1)   m(2)  
2 0 1 2 0 1
1 2 3

The global lumped-mass matrix is: 1 0 0

 AL 
M   0 2 0
2  
0 0 1

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
Substitute the global stiffness and mass matrices into the global
dynamic equations gives:

 1 1 0   u1   1 0 0   u1   R1 
AE     0 2 0  u    0 
1 2 1 u2  
AL
L   2   2  
 
 0 1 1  u3      
0 0 1 u3  F3 (t )

where R1 denotes the unknown reaction at node 1.

CIVL 7/8117 Chapter 16 - Structural Dynamics 43/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
For this example, we will use the central difference method,
because it is easier to apply, for the numerical time integration.

It has been mathematically shown that the time step t must be

less than or equal to two divided by the highest natural
frequency.
2
t 
max

For practical results, we should use a time step defined by:

3 2 
t   
4  max 

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
An alternative guide (used only for a bar) for choosing the
approximate time step is: L
t 
cx
Ex
where L is the element length, and c x 

is the longitudinal wave velocity.

Evaluating the time step estimates gives:

3 2  1.5
t     0.40  10 3 s
4  max  3.76  10 3

L 100in.
t    0.48  10 3 s
cx 30  10 psi
6

lb. s2
0.00073 in.4
CIVL 7/8117 Chapter 16 - Structural Dynamics 44/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
Guided by these estimates for time step, we will select
t = 0.25 x 10-3 s.

Procedure for solution:

1. At time t = 0: d0   0 d   0
0

2. If the acceleration is not given, solve for d0  

d   M F   K d 
0
1
0 0

u  2  12 0    0  AE  2 1 0  
 
d0   2         
u3 t 0  AL  0 1  1,000  L  1 1  0  

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
Applying the boundary conditions u1 = 0 and u1  0 and
simplifying gives:
u2  2000 0   0  in
d   u 
0    
 AL  1 27,400  s
2
 3 t  0
3. Solve for d-1 at t = -t
( t )2
d 1  d0   ( t )d0   d0 
2
u2  0  3 0  (0.25  10 3 )2  0 
      (0.25  10 )     
u3 1 0  0  2 27,400 
 0 
 3 
in
0.856  10 
CIVL 7/8117 Chapter 16 - Structural Dynamics 45/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
4. Solve for d1 at t = t using the value of d-1 from Step 3

d1  M  t  F0   2 M   t  K  d0   Md 1

1 2 2

u2  2  12 0    0   2(0.073) 2 0 
 
2
3
    0 1  0.25  10    
 3 1
u 0.073   1,000   2  0 1

1  0  0.073 2 0   
  30  10  1
2 0

2
 0.25  103 4
       3 
 1   0  2 0 1 0.856  10 

u2  2  12 0    0   0 
    0 1  0.0625  10 3   0.0312  10 3 
 3 1
u 0.073      
u2   0 
   3 
in
 3 1 
u 0.858  10 

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
5. With d0 given and d1 determined in Step 4 solve for d2

d2   M  t  F1  2 M   t  K  d1  Md0 

1 2 2

u2  2  12 0    0   2(0.073) 2 0 
 
2
3
      0.25  10     
u3 2 0.073  0 1  1000   2  0 1

1    0.073 2 0  0 
  30  10  1
2 0

2
 0.25  103 4
   
 1   0.858  10 3  2 0 1 0 

u2  2  12 0    0   0.0161 10 3 
       3   3  
u3 2 0.073  0 1  0.0625  10  0.0466  10 

u2  0.221 10 3 
   3 
in
u3 2  2.99  10 
CIVL 7/8117 Chapter 16 - Structural Dynamics 46/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

6. Solve for d1 

d   M F   Kd 
1
1
1 1

u2  2  12 0    0  4  2 1  0 
         (30  10 )    3  
u3 1 0.073  0 1  1000   1 1  0.858  10 

u2   3,526  in
      2
u3 1 20,345  s

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

7. Solve for d1  

d   
d 2   d0 
1
2( t )

 0.221 10 3  0 
 3 
  
u2    2.99  10  0  0.442 in
     
u3 1 
2 0.25  10 3   5.98  s
CIVL 7/8117 Chapter 16 - Structural Dynamics 47/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

8. Repeat Steps 5, 6, and 7 to obtain the displacement,

acceleration, and velocity for other time steps.
Repeating Step 5:
1

d3   M  t  F2   2 M   t  K  d2   Md1
2 2

u2  2  12 0    0   2(0.073) 2 0 
 
2
3
      0.25  10     
u3 3 0.073  0 1  1000   2  0 1

1  0.221 103  0.073 2 0   

  30  10  1
2 0

2
 0.25  103 4
  3 
    3 
 1    2.99  10  2 0 1 0.858  10 

u2  1.096  10 3 
   3 
in
u3 2 5.397  10 

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

Repeating Step 6. Solve for d2  

d   M F   Kd 
2
1
2 2

u2  2  12 0    0  4  2 1 0.221 10 3 

          (30  10 )   3  
u3 2 0.073  0 1  1000   1 1   2.99  10 

u2  10,500  in
      2
u3 1  4,600  s
CIVL 7/8117 Chapter 16 - Structural Dynamics 48/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

Repeat Step 7: Solve for d2  

d   
d3   d1
2
2( t )

 1.096  10 3   0 
 3 
 3  
u2   5.397  10  0.858  10   2.192 in
    
u3 2 
2 0.25  10 3 
9.078  s

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

The following table summarizes the results for the remaining

time steps as compared with the exact solution.

t (10-3 s) d2 d3 d2 (in/s) d3 (in/s) d2 (104in/s2) d3(104in/s2)

0.00 0.0000 0.0000 0.0000 0.0000 0.0000 2.7397
0.25 0.0000 0.0009 0.4398 5.9697 0.3518 2.0360
0.50 0.0002 0.0030 2.1870 9.0987 1.0459 0.4672
0.75 0.0011 0.0054 5.1477 8.6772 1.3227 -0.8044
1.00 0.0028 0.0073 7.6928 6.4426 0.7134 -0.9833
1.25 0.0049 0.0086 7.9409 4.8502 -0.5149 -0.2906
1.50 0.0068 0.0097 5.3555 4.8455 -1.5535 0.2868
CIVL 7/8117 Chapter 16 - Structural Dynamics 49/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

Plotting the displacements for 0.015 s gives:

10 -3
14

12

10

-2
0 0.005 0.01 0.015

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

Plotting the velocities for 0.015 s gives:

15

10

-5

-10

-15
0 0.005 0.01 0.015
CIVL 7/8117 Chapter 16 - Structural Dynamics 50/85

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example

Plotting the accelerations for 0.015 s gives:

10 4
3

-1

-2

-3
0 0.005 0.01 0.015

Structural Dynamics
Time-Dependent One-Dimensional Bar - Example
CIVL 7/8117 Chapter 16 - Structural Dynamics 51/85

Structural Dynamics
Beam Element Mass Matrices and Natural Frequencies

We will develop the lumped- and consistent-mass matrices for

time-dependent beam analysis.

Consider the beam element shown in the figure below.

The basic equations of motion are: F (t )  K  {d }  M d

Structural Dynamics
Beam Element Mass Matrices and Natural Frequencies

The stiffness matrix is: v1 1 v2 2

 12 6L 12 6L 
 2
6L 2L2 
EI  6L 4L
  L3  12
k 
6L 12

6L 
 
 6L 2L2 6L 4L2 
The lumped-mass matrix is: v  v 
1 1 2 2

1 0 0 0
 0 0 0
 AL 0
m   2 0 
0 1 0
 
0 0 0 0
CIVL 7/8117 Chapter 16 - Structural Dynamics 52/85

Structural Dynamics
Beam Element Mass Matrices and Natural Frequencies

The mass is lumped equally into each transitional degree of

freedom; however, the inertial effects associated with any
possible rotational degrees of freedom is assumed to be zero.

A value for these rotational degrees of freedom could be

assigned by calculating the mass moment of inertia about
each end node using basic dynamics as:
2
1 1   AL   L   AL3
I  mL2   
3 3  2   2  24

Structural Dynamics
Beam Element Mass Matrices and Natural Frequencies
The consistent-mass matrix can be obtained by applying
m     N  N  dV
T

 N1 
N 
m      N2  N1 N2 N3 N4  dA dx
L

0 A  3
N4 
where 1 1

N1  3 2 x 3  3 x 2L  L3
L
 
N2  3 x 3L  2 x 2L2  xL3
L

1 1 3
N3 
L3 
2 x 3  3 x 2 L  N4 
L3x L  x 2L2 
CIVL 7/8117 Chapter 16 - Structural Dynamics 53/85

Structural Dynamics
Beam Element Mass Matrices and Natural Frequencies
The shape functions are shown below:
1.2

1.0

0.8
N1
0.6 N2

0.4 N3

0.2 N4

0.0
0.00 1.00

-0.2

-0.4

Structural Dynamics
Beam Element Mass Matrices and Natural Frequencies
Substituting the shape functions into the above mass
expression and integrating gives:
 156 22L 54 13L 
 2
13L 3L2 
 AL  22L 4L 
[m ] 
420  54 13L 156 22L 
 
 13L 3L 22L 4L2 
2
CIVL 7/8117 Chapter 16 - Structural Dynamics 54/85

Structural Dynamics
Beam Element - Example 1
Determine the first natural frequency for the beam shown in the
figure below. Assume the bar has a length 2L, modulus of
elasticity E, mass density , and cross-sectional area A.

Let’s discretize the beam into two elements each of length L.

We will use the lumped-mass matrix.

Structural Dynamics
Beam Element - Example 1
We can obtain the natural frequencies by using the following
equation.
K   2 M  0
The boundary conditions are v1 = 1 = 0 and v3 = 3 = 0.

v1 1 v2 2 v2 2 v3 3

 12 6L 12 6L   12 6L 12 6L 
 2   2 
EI 6L 4L 6L 2L  EI 6L 4L 6L 2L 
2 2
k (1)   3  k (2)   3 
L  12 6L 12 6L  L  12 6L 12 6L 
 2   2 
 6L 2L 6L 4L   6L 2L 6L 4L 
2 2
CIVL 7/8117 Chapter 16 - Structural Dynamics 55/85

Structural Dynamics
Beam Element - Example 1
We can obtain the natural frequencies by using the following
equation.
K   2 M  0
The boundary conditions are v1 = 1 = 0 and v3 = 3 = 0.

v  v  v  v 
1 1 2 2 2 2 3 3

1 0 0 0 1 0 0 0
 0 0 0  0 0 0
 AL 0   AL 0 
m  
(1)
m  
(2)

2 0 0 1 0 2 0 0 1 0
   
0 0 0 0 0 0 0 0

Structural Dynamics
Beam Element - Example 1
We can obtain the natural frequencies by using the following
equation.
K   2 M  0
The boundary conditions are v1 = 1 = 0 and v3 = 3 = 0.

Therefore, the global stiffness and lumped-mass matrices are:

v2 2 v2 2

EI 24 0   AL 2 0 
K  L3  0 8L2 
M  2 0 0 

CIVL 7/8117 Chapter 16 - Structural Dynamics 56/85

Structural Dynamics
Beam Element - Example 1
Substituting the global stiffness and mass matrices into the
global dynamic equations gives:

EI 24 0  1 0
K   2 M  0 3  2
L  0 8L 
  2  AL   0
0 0 
Dividing by AL and simplify

24EI 4.8990 EI
2  
 AL4 L2 A
The exact solution for the first natural frequency is:
5.59 EI

L2 A

Structural Dynamics
Beam Element - Example 2
Determine the first natural frequency for the beam shown in the
figure below. Assume the bar has a length 3L, modulus of
elasticity E, mass density , and cross-sectional area A.

Let’s discretize the beam into three elements each of length L.

We will use the lumped-mass matrix.

CIVL 7/8117 Chapter 16 - Structural Dynamics 57/85

Structural Dynamics
Beam Element - Example 2
We can obtain the natural frequencies by using the following
equation.
K   2 M  0
The boundary conditions are v1 = 1 = 0 and v4 = 4 = 0.

Therefore the elements of the stiffness matrix for element 1 are:

v1 1 v2 2

 12 6L 12 6L 
 2 
EI  6L 4L 6L 2L 
2
k   3
(1)

L  12 6L 12 6L 

 2 
 6L 2L 6L 4L 
2

Structural Dynamics
Beam Element - Example 2
Element 2: v2 2 v3 3

 12 6L 12 6L 
 2 
EI 6L 4L 6L 2L 
2
k (2)   3 
L  12 6L 12 6L 
 2 
 6L 2L 6L 4L 
2

Element 3: v3 3 v4 4

 12 6L 12 6L 
 2 
EI 6L 4L 6L 2L 
2
k (3)   3 
L  12 6L 12 6L 
 2 
 6L 2L 6L 4L 
2
CIVL 7/8117 Chapter 16 - Structural Dynamics 58/85

Structural Dynamics
Beam Element - Example 2
Assembling the global stiffness matrix as:
v2 2 v3 3

 12  12 6L  6L 12 6L 
 6L  6L 4L2  2L2 6L 2L2 
EI 
K   L3  12 6L 12  12 6L  6L 

 
 6L 2L2 6L  6L 4L2  4L2 
v2 2 v3 3

 24 0 12 6L 
 0 8L2 6L 2L2 
EI 
  L3 12 6L 24 0 
K 
 2 
 6L 2L 0 8L2 

Structural Dynamics
Beam Element – Example 2
We can obtain the natural frequencies by using the following
equation.
K   2 M  0
Therefore the elements of the mass matrix for element 1 are:
v1 1 v2 2

1 0 0 0
 0 0 0
 AL 0 
m  
(1)

2 0 0 1 0
 
0 0 0 0
CIVL 7/8117 Chapter 16 - Structural Dynamics 59/85

Structural Dynamics
Beam Element – Example 2
Element 2: v2 2 v3 3 Element 3: v3 3 v4 4

1 0 0 0 1 0 0 0
 0 0 0  0 0 0
 AL 0   AL 0 
m  
(2)
m  
(2)

2 0 0 1 0 2 0 0 1 0
   
0 0 0 0 0 0 0 0
The assembled mass matrix is:
v2 2 v3 3 v2 2 v3 3

1  1 0 0 0 1 0 0 0
 0 0 0 0
 AL  0 0 0 0
m  2  0  M    AL 0 
0 1 0
0 1 1 0
   
 0 0 0 0 0 0 0 0

Structural Dynamics
Beam Element – Example 2
We can obtain the frequency equation as: K   2 M  0
 24 0 12 6L  1 0 0 0
 0 8L 6L 2L
2 2 0 0 0 0
EI     2  AL   0
L3  12 6L 24 0  0 0 1 0
 2 2  
 6L 2L 0 8L  0 0 0 0

24 EI
L3
  2  AL 0 12 EI
L3
6 EI
L2

0 8 EIL 6 EI
L2
2 EIL
0
12 EI
L3
6 EI
L2
24 EI
L3
  2  AL 0
6 EI
L2
2 EIL 0 8 EIL
CIVL 7/8117 Chapter 16 - Structural Dynamics 60/85

Structural Dynamics
Beam Element – Example 2
Simplifying and assuming:    AL
24 EI
L3
 2 0 12 EI
L3
6 EI
L2

0 8 EIL 6 EI
L2
2 EIL
0
12 EI
L3
6 EI
L2
24 EI
L3
  2 0
6 EI
L2
2 EIL 0 8 EIL

Evaluating the 4x4 determinate gives:

60 4  2E 2I 2 2,304 2  E 3I 3 11,664 E 4I 4
  0
L2 L5 L8
2.4495 EI 5.6921 EI
1   2  
L2 A L2 A

Structural Dynamics
Beam Element – Example 2
Ignoring the negative root as it is not physically possible gives:
2.4495 EI 5.6921 EI
1  2 
L2 A L2 A

To compare with the two-element solution, assume L = ⅔L:

2.4495 EI 5.5113 EI
1  
 23L  A
2
L2 A
CIVL 7/8117 Chapter 16 - Structural Dynamics 61/85

Structural Dynamics
Beam Element – Example 1 Revisited
Determine the first natural frequency for the beam shown in the
figure below. Assume the bar has a length 2L, modulus of
elasticity E, mass density , and cross-sectional area A.

In summary:
4.90 EI Exact solution:
Two elements: 
L2 A
5.59 EI

5.5114 EI L2 A
Three elements:  
L2 A

Structural Dynamics
Beam Element – Example 3
Determine the first natural frequency for the beam shown in the
figure below.

Assume the bar has a length L = 30 in, modulus of elasticity

E = 3 x 107 psi, mass density  = 0.00073 lb-s2/in4, and cross-
sectional area A = 1 in2, moment of inertia I = 0.0833 in4, and
Poisson’s ratio  = 0.3.
CIVL 7/8117 Chapter 16 - Structural Dynamics 62/85

Structural Dynamics
Beam Element – Example 3
Determine the first natural frequency for the beam shown in the
figure below.

Let’s discretize the beam into two elements each of length

L = 15 in. We will use the lumped-mass matrix.

We can obtain the natural frequencies by using the following

equation.
K   2 M  0

Structural Dynamics
Beam Element – Example 3
In this example, the elemental stiffness matrices are:
v1 1 v2 2

Element 1:  12 6L 12 6L 
 2 
EI  6L 4L 6L 2L 
2
k   3
(1)

L  12 6L 12 6L 

 2 
 6L 2L 6L 4L 
2

v2 2 v3 3

Element 2:  12 6L 12 6L 
 2 
EI 6L 4L 6L 2L 
2
k (2)   3 
L  12 6L 12 6L 
 2 
 6L 2L 6L 4L 
2
CIVL 7/8117 Chapter 16 - Structural Dynamics 63/85

Structural Dynamics
Beam Element – Example 3
In this example, the elemental mass matrices are:
v1 1 v2 2

Element 1: 1 0 0 0
 0 0 0
 AL 0 
m(1)  
2 0 0 1 0
 
0 0 0 0
v2 2 v3 3

Element 2: 1 0 0 0
 0
 AL 0 0 0

m(2)  
2 0 0 1 0
 
0 0 0 0

Structural Dynamics
Beam Element – Example 3
The boundary conditions are u1 = 1 = 0. Therefore the global
stiffness and lumped mass matrices is:
v2 2 v3 3 v2 2 v3 3

 24 0 12 6L  2 0 0 0
 0 8L2 6L 2L2   0 0 0
EI   AL 0
  L3 12 6L 12 6L 
K  M
2 0

0 1 0
 2   
 6L 2L 6L 4L 
2
0 0 0 0
CIVL 7/8117 Chapter 16 - Structural Dynamics 64/85

Structural Dynamics
Beam Element – Example 3
Substituting the global stiffness and mass matrices into the
global dynamic equations gives:
 24 0 12 6L  2 0 0 0
 6L 2 0 0 0 0
   2  AL 
EI  0 8L2 2L
 0
L3  12 6L 12 6L  2 0 0 1 0
 2  
 6L 2L 2
6L 4L  0 0 0 0

 24 0 12 6L  1 0 0 0
 0 8L2
6L 2L2 0 0 0 0
   2   0
 12 6L 12 6L  0 0 21 0 
 2   EI
 6L 2L2 6L 4L  0 0 0 0 
 AL4

Structural Dynamics
Beam Element – Example 3
Evaluating the determinant of the above equations gives:
1  0.7891 
14  4 4  240  3 2  144  4  0
2  4.0645 
The solution for the first two natural frequencies are:
0.7891 EI 4.0647 EI
1  2 
L2 A L2 A
CIVL 7/8117 Chapter 16 - Structural Dynamics 65/85

Structural Dynamics
Beam Element – Example 3
In this example, the length of each element is actual L/2,
therefore:
0.7891 EI 3.1564 EI
1  
L2  A A
2
L2

4.0645 EI 16.2580 EI
2  
L2 A A
2
L2

The exact solution for the first natural frequency is:

3.516 EI

L2 A

Structural Dynamics
Beam Element – Example 3
In the exact solution of the vibration of a clamped-free beam,
the higher natural frequencies to the first natural frequency
can be given as:
2 3
 6.2669  17.5475
1 1
CIVL 7/8117 Chapter 16 - Structural Dynamics 66/85

Structural Dynamics
Beam Element – Example 3
The figures below shows the first and second mode shapes
corresponding to the first two natural frequencies for a 2-
element FEA cantilever beam.

First mode shape Second mode shape

Structural Dynamics
Beam Element – Example 3
The figures below shows the first and second mode shapes
corresponding to the first two natural frequencies for a 8-
element FEA cantilever beam.

First mode shape Second mode shape

CIVL 7/8117 Chapter 16 - Structural Dynamics 67/85

Structural Dynamics
Beam Element – Example 3
The figures below shows the third and fourth mode shapes
corresponding to the first two natural frequencies for a 8-
element FEA cantilever beam.

Third mode shape Forth mode shape

Structural Dynamics
Beam Element – Example 3
The table below shows various finite element solutions
compared to the exact solution.
1 (rad/s) 2 (rad/s)
Exact Solution 228 1,434
FE Solution
Using 2 elements 205 1,286
Using 6 elements 226 1,372
Using 10 elements 227.5 1,410
Using 30 elements 228.5 1,430
Using 60 elements 228.5 1,432
CIVL 7/8117 Chapter 16 - Structural Dynamics 68/85

Structural Dynamics
Truss Analysis
The dynamics of trusses and plane frames are preformed by
extending the concepts of bar and beam element.

The truss element requires the same transformation of the mass

matrix from local to global coordinates as that used for the
stiffness matrix given as:
m  T  mT 
T

Structural Dynamics
Truss Analysis
Considering two-dimensional motion, the axial and the
transverse displacement are given as:
 u1   u1 
 
u  N1 0 N2 0  v1  1 L  x  0 x  0  v1 
        
v   0 N1 0 N2  u2  L  0 L  x  0 x  u2 
v 2  v 2 

The shape functions for the matrix are:

1 L  x  0 x 0 
N   L  L  x  0 x 
 0
CIVL 7/8117 Chapter 16 - Structural Dynamics 69/85

Structural Dynamics
Truss Analysis
The consistent-mass matrix can be obtained by applying:
2 0 1 0 
 
[m]    [N ]T [N ]dV  AL 0 2 0 1
m   6 1 0 2 0

V
 
0 1 0 2 
The lumped-mass matrix for two-dimensional motion is obtained
by simply lumping mass at each node (mass is the same in
both the x and y directions): 1 0 0 0
 
 AL 0 1 0 0 
m  2 0 0 1 0
 
0 0 0 1

Structural Dynamics
Plane Frame Analysis
The plane frame element requires combining the bar and beam
elements to obtain the local mass matrix.

There are six degrees of freedom associated with a plane frame

element.
CIVL 7/8117 Chapter 16 - Structural Dynamics 70/85

Structural Dynamics
Plane Frame Analysis
The plane frame analysis requires first expanding and then
combining the bar and beam mass matrices to obtain the local
mass matrix.

The bar and beam mass matrices are expanded to a 6 x 6 and

superimposed

Structural Dynamics
Plane Frame Analysis
Combining the local axis consistent-mass matrices for the bar
and beam elements gives:
2 6 0 0 1
6 0 0 
 
0 0
156 22L 54 13 L
420 420 420 420 

0 22L 4L 2
0 13 L 3 L2
420 
m   AL  1 420 420 420

 6 0 0 2
6 0 0 
0 54
420
13 L
420 0 156
420
22L
420

 
0 3 L2 0
2
13 L 22L
420 
4L
420 420 420

- Beam element components

CIVL 7/8117 Chapter 16 - Structural Dynamics 71/85

Structural Dynamics
Plane Frame Analysis
The resulting lumped-mass matrix for a plane frame element is
give as:
1 0 0 0 0 0
0 1 0 0 0 0
 
 AL 0 0 0 0 0 0
m  2 0 
0 0 1 0 0

0 0 0 0 1 0
 
0 0 0 0 0 0

Structural Dynamics
Plane Frame Analysis
The global mass matrix for the plane frame element arbitrarily
oriented in x-y coordinates is transformed by:

m  T  mT 
T

where the transformation matrix is given as:

 C S 0 0 0 0
 S C 0 0 0 0 
 
 0 0 1 0 0 0
T    0 0 0 C S 0
 
 0 0 0 S C 0 
 
 0 0 0 0 0 1
CIVL 7/8117 Chapter 16 - Structural Dynamics 72/85

Structural Dynamics
Plane Stress/Strain Elements
The plane stress/strain constant-strain triangle consistent-mass
matrix is obtained using the shape functions given below as:

N 0 N2 0 N3 0
N    01 N1 0 N2 0 N3 


The linear triangular shape functions are illustrated below:

N1 N2 N3

1 y y y

1
x 3 x 3 x 3
1 1 1

2 2 2

Structural Dynamics
Plane Stress/Strain Elements

The linear triangular shape functions are illustrated below:

N1 N2 N3

1 y y y

1
x 3 x 3 x 3
1 1 1

2 2 2
CIVL 7/8117 Chapter 16 - Structural Dynamics 73/85

Structural Dynamics
Plane Stress/Strain Elements
The consistent-mass matrix can be obtained by applying:
[m]    [N ]T [N ]dV
V

where dV = tdA

The CST global consistent-mass matrix is:

2 0 1 0 1 0
0 2 0 1 0 1
 
 tA  1 0 2 0 1 0
[m ]   
12 0 1 0 2 0 1
1 0 1 0 2 0
 
0 1 0 1 0 2

Structural Dynamics
Frame Example Problem 1
Use SAP2000 to determine the motion of the frame structure
below.
CIVL 7/8117 Chapter 16 - Structural Dynamics 74/85

Structural Dynamics
Frame Example Problem 1
Assume the modulus of elasticity E = 3 x 107 psi. The nodal
lumped mass values are obtained by dividing the total weight
(dead loads included) of each floor or wall section by gravity.

Structural Dynamics
Frame Example Problem 1
Below is the SAP2000 model for this example:

WALLB

WALLT

FLOORT

FLOORB
CIVL 7/8117 Chapter 16 - Structural Dynamics 75/85

Structural Dynamics
Frame Example Problem 1
For example, compute the total mass of the uniform vertical
load on elements 4, 5, and 6:

W4  50 psf  30 ft 15 ft 
 58.28 lb  s
2
M4  
g 386.04 in s2 in

W5 104 psf  30 ft 15 ft 

 121.23 lb  s
2
M5  
g 386.04 in s2 in

W6 104 psf  30 ft 15 ft 

 121.23 lb  s
2
M6  
g 386.04 in s2 in

Structural Dynamics
Frame Example Problem 1
Next, lump the mass equally to each node of the beam element.

For this example calculation, a lumped mass of 29.14 lbs2/in

should be added to nodes 7 and 8 and a mass of 60.62
lbs2/in should be added to nodes 3, 4, 5, and 6, all in
the x direction.
CIVL 7/8117 Chapter 16 - Structural Dynamics 76/85

Structural Dynamics
Frame Example Problem 1
In an identical manner, masses for the dead loads for additional
wall sections should be added to their respective nodes.

In this example, additional wall loads should be converted to

mass added to the appropriate loads.

Structural Dynamics
Frame Example Problem 1
For example, the additional mass due to the wall load on
element 7 is:
W7  20 psf 10 ft 15 ft 
 7.77 lb  s
2
M7  
g 386.04 in s2 in

Therefore, an additional 3.88 lbs2/in should be added to nodes

6 and 8.
CIVL 7/8117 Chapter 16 - Structural Dynamics 77/85

Structural Dynamics
Frame Example Problem 1
For example, the additional mass due to the wall load on
element 8 is:
W8  20 psf 10 ft 15 ft 
 7.77 lb  s
2
M8  
g 386.04 in s2 in

Therefore, an additional 3.88 lbs2/in should be added to nodes

4 and 6.

Structural Dynamics
Frame Example Problem 1
For example, the additional mass due to the wall load on
element 9 is:
W9  20 psf 15 ft 15 ft 
 11.66 lb  s
2
M9  
g 386.04 in s2 in

Therefore, an additional 5.83 lbs2/in should be added to node 4.

CIVL 7/8117 Chapter 16 - Structural Dynamics 78/85

Structural Dynamics
Frame Example Problem 1
The final values for the nodal masses on this frame are shown
below.

Structural Dynamics
Frame Example Problem 1
The deflected shape from SAP2000 is shown below:
CIVL 7/8117 Chapter 16 - Structural Dynamics 79/85

Structural Dynamics
Frame Example Problem 1
A trace of the displacements of nodes 8, 6, and 4 as a function
of time can be generated using the SAP2000 Display Menu.

Structural Dynamics
Frame Example Problem 1
A plot of the displacements of nodes 8, 6, and 4 over the first 5
seconds of the analysis generate by SAP2000 is shown
below:
CIVL 7/8117 Chapter 16 - Structural Dynamics 80/85

Structural Dynamics
Frame Example Problem 1
The maximum displacement of node 8 is 3.16 in. and the period
of the vibration is approximately 1.2 seconds.

Structural Dynamics
Frame Example Problem 1
The maximum velocity of node 8 is 8.08 in/s
CIVL 7/8117 Chapter 16 - Structural Dynamics 81/85

Structural Dynamics
Frame Example Problem 1
The maximum acceleration of node 8 is -41.9 in/s2

Structural Dynamics
Frame Example Problem 1
A plot of the displacement, velocity, and acceleration of node 8
is show below:
CIVL 7/8117 Chapter 16 - Structural Dynamics 82/85

Structural Dynamics
Frame Example Problem 2
Use SAP2000 to determine the motion of the frame structure
below.

Structural Dynamics
Frame Example Problem 2
A trace of the displacements as a function of time of nodes 8, 6,
and 4 over the first 5 seconds of the analysis generate by
SAP2000 is shown below:
CIVL 7/8117 Chapter 16 - Structural Dynamics 83/85

Structural Dynamics
Frame Example Problem 2
The maximum displacement of node 8 is 0.877 in. and the
period of the vibration is approximately 1.3 seconds.

Structural Dynamics
Frame Example Problem 2
The maximum velocity of node 8 is 4.68 in/s
CIVL 7/8117 Chapter 16 - Structural Dynamics 84/85

Structural Dynamics
Frame Example Problem 2
The maximum acceleration of node 8 is -75.2 in/s2

Structural Dynamics
Frame Example Problem 2
A plot of the displacement, velocity, and acceleration of node 8
is show below:
CIVL 7/8117 Chapter 16 - Structural Dynamics 85/85

Development of the Plate Bending Element

Problems
23. Do problems 16.6 and 16.11c in your textbook.

24. Do problems 16.14 and 16.16 in your textbook using

SAP2000.

End of Chapter 16