2-14 STATIC AND DYNAMIC ANALYSIS

4. Many materials, such as soils and fluids, do not satisfy the compatibility
equations. Also, locked in construction stresses, creep and slippage within
 joints are real violations of displacement compatibility. Therefore,
approximate methods that satisfy statics may produce more realistic results
for the purpose of design.

5. In addition, engineering students are not normally required to take a course in

geometry; whereas, all students take a course in statics. Hence, there has not
been an emphasis on the application of the equations of geometry.

The relaxation of the displacement compatibility requirement has been justified

for hand calculation to minimize computational time. Also, if one must make a
choice between satisfying the equations of statics or the equations of geometry,
in general, we should satisfy the equations of statics for the reasons previously
stated.

However, because of the existence of inexpensive powerful computers and

efficient modern computer programs, it is not necessary to approximate the
compatibility requirements. For many structures, such approximations can
produce significant errors in the force distribution in the structure in addition to
incorrect displacements.

2.15 REFERENCES

1. Cook, R. D., D. S. Malkus and M. E. Plesha. 1989. Concepts and 

 Applications of Finite Element Analysis, Third Edition. John Wiley & Sons,
Inc. ISBN 0-471-84788-7.

2. Boresi, A. P. 1985.  Advanced Mechanics of Materials. John Wiley & Sons,

Inc. ISBN 0-471-88392-1.
 3.

ENERGY AND WORK

 All External Wor k Supplied to a Real Structural

System is Stored or Dissipated as Energy

3.1 INTRODUCTION

A large number of energy methods have been presented during the last 150 years
for the analysis of both determinate and statically indeterminate structures.
However, if all methods are formulated in matrix notation, it can be shown that
only two fundamental methods exist. They are generally defined as the force and
displacement methods. One can use minimum energy principles or methods of 
virtual-work to derive the general equations for linear structural analysis.  Energy
is defined as the ability to do work. Both have the units of force-distance.

For many types of structural elements, however, there can be many advantages in
using both force and displacement methods in approximating the stiffness
properties of the element. For example, the classical non-prismatic beam element
uses a force approach to define the forces at a typical cross-section within the
beam; however, a displacement approximation, such as plane sections remain
plane, is used to define the strain distribution over the cross-section.

In recent years, assumed-stress hybrid formulations have been used to produce

element stiffness properties. In addition, assumed-stress distributions, virtual
work methods and the least-square error approach have been used to calculate
accurate stresses in displacement-based finite elements. Therefore, no one
method can be used to solve all problems in structural analysis. The only
3-2 STATIC AND DYNAMIC ANALYSIS

restriction on the computational techniques used is that the results must converge
to the exact values as the elements become smaller.

3.2 VIRTUAL AND REAL WORK

The principles of virtual work are very simple and are clear statements of 
conservation of energy. The principles apply to structures that are in equilibrium
in a real displaced position u when subjected to loading R . The corresponding
real internal deformations and internal forces are d and f respectively. All terms
are illustrated in Figures 3.1 and 3.2.

 R  f 

 R  f 

 R  f 
u d 

 A. External Virtual Work   R u  B. Internal Virtual Work   f d 

 Figure 3.1 Method of Virtual Forces

The principle of virtual forces states (in my words) if a set of infinitesimal

external forces , R , in equilibrium with a set of infinitesimal internal forces
f that exist before the application of the real loads and displacements, the
external virtual work is equal to the internal virtual work. Or, in terms of the
notation defined previously:

T
R u = f Td (3.1)

If only one joint displacement ui is to be calculated, only one external virtual

load exists, Ri = 1 . For this case, the equation is the same as the unit load
method. It is apparent for nonlinear analysis that the principle of virtual forces
ENERGY AND WORK 3-3

 f 
 R

 R  f 

u d 
u d 
 A. External Virtual Work  u  R  B. Internal Virtual Work  d  f 

 Figure 3.2 Method of Virtual Displacements

cannot be used, because the linear relationship between R and f may not hold
after the application of the real loads and displacements.

The principle of virtual displacements states (in my words) if a set of 

infinitesimal external displacements, u , consistent with a set of internal virtual
displacements, d , and boundary conditions are applied after the application of 
the real loads and displacements, the external virtual work is equal to the
internal virtual work. Or, in terms of matrix notation:

T
u R =d T
f (3.2)

It is important to note that the principle of virtual displacements does apply to the
solution of nonlinear systems because the virtual displacements are applied to
real forces in the deformed structure.

In the case of finite element analysis of continuous solids, the virtual work 
principles are applied at the level of stresses and strains; therefore, integration
over the volume of the element is required to calculate the virtual work terms.

For linear analysis, it is apparent that the real external work, or energy, is given
by:

1 T 1 T
W E = u R = R u (3.3)
2 2
3-4 STATIC AND DYNAMIC ANALYSIS

The real internal work, or strain energy, is given by:

1 T 1 T
W I  = d f = f d (3.4)
2 2

3.3 POTENTIAL ENERGY AND KINETIC ENERGY

One of the most fundamental forms of energy is the position of a mass within a
gravitational field near the earth's surface. The gravitational potential energy
V  g is defined as the constant weight w moved against a constant gravitational
field of distance h . Or:

V  g = mgh or V g = Wh (3.5)

A mass that is moving with velocity v has kinetic energy given by the following
equation:

1
V k  = mv 2 (3.6)
2

One of the most common examples that illustrates the physical significance of 
both the potential and kinetic energy is the behavior of a pendulum shown in
Figure 3.3.

If the mass of the pendulum has an initial position of  h max , the kinetic energy is
zero and the potential energy is h maxW  . When h equals zero, the potential
energy is zero; therefore, from conservation of energy, the kinetic energy is:

W  v 2
V k  = h max W  = (3.7)
2 g

Hence, the maximum horizontal velocity is:

v max = 2  g h max (3.8)

ENERGY AND WORK 3-5

Massless
Rigid Bar
θ 
v=0
 L
V k  = 0
V g = W  hmax V g =0
V k  = W  hmax
h
v vmax = 2 g hmax

 Figure 3.3 Oscillation of Pendulum

It is important to note that the total energy in the oscillating system is always
constant; therefore, the following energy equation, at any time t , must be
satisfied:

V  g (t) + V k (t) = W  hmax = constant (3.9)

The physical behavior of the oscillating pendulum can be considered to be an

energy pump, where there is an interchange between potential and kinetic
energy.

The tangential force accelerating the mass is W  sin θ . From Newton's Second
Law, the following nonlinear, differential equation of equilibrium can be written:

g
mLθ + W  sin θ = 0 or, θ + sin θ = 0 (3.10)
L

For very small angles, sin θ ≈ θ , the approximate linear differential equation is:

θ +  g θ = 0 (3.11)
L
3-6 STATIC AND DYNAMIC ANALYSIS

Hence, the small displacement  period of oscillation of a pendulum is:

L
T  = 2π (3.12)
 g

3.4 STRAIN ENERGY

The strain energy stored in an element " i" within a general structural system is
the area under the stress-strain diagram integrated over the volume of the
element. For linear systems, the stress-strain matrix E (i ) , including initial thermal
stresses ft (i ) , can be written in matrix form as:

(i )
f = E (i ) d (i ) + ft (i ) (3.13)

(i) i
The column matrices f and d ( ) are the stresses and strain respectively.
Therefore, the strain energy within one element is given by:

1 (i )

2 ∫  ∫ 
(i ) (i ) T (i) (i ) T (i)
W I  = d E d dV  + d ft dV  (3.14)

Within each element, an approximation can be made on the displacements. Or:

u x ( i ) = N( i ) u x , u y (i ) = N (i ) u y and u z(i ) = N(i ) u z (3.15)

Hence, after the application of the strain-displacement equations, the element

strains can be expressed in terms of nodal displacements. Or:

(i)
d = B( i ) u or d (i )T = u T B(i)T (3.16)

The column matrix u contains all of the node, or joint, displacements of the
complete structural system. In addition, it may contain displacement patterns
within the element. When equation (3.16) is written in this form, it is apparent
that the B( i ) matrix can be very large; however, it only has non-zero terms
associated with the displacements at the nodes connected to nodes adjacent to the
element. Therefore, the B(i ) matrix is always formed and used in compacted form
within a computer program, and an integer location array , L(i)a , is formed for
ENERGY AND WORK 3-7

(i)
each element that is used to relate the local node displacements u to the global 
node displacements u .

After integration over the volume of the element, the strain energy, in terms of 
the global node displacements, can be written as:

1
W I ( i ) = u
T
k
(i )
u + uT (i)
Ft (3.17)
2

Therefore, the element stiffness matrix is by definition:

k
(i)
= ∫  B
(i )T
E
(i )
B
( i)
dV  (3.18)

And the element thermal force matrix is:

F
(i )
= ∫ 
B
( i )T
ft
(i )
dV  (3.19)

The total internal strain energy is the sum of the element strain energies. Or:

1 T
W I  = u Ku + uT Ft (3.20)
2

( i)
The global stiffness matrix K is the sum of the element stiffness matrices k .
Or:

K = ∑ k
(i )
(3.21)

The summation of element stiffness matrices to form the global stiffness matrix
is termed the  direct stiffness method . The global thermal load vector Ft is the
sum of the element thermal load matrices:

Ft = ∑ Ft
(t)
(3.22)

3.5 EXTERNAL WORK

The external work  W c performed by a system of concentrated node, or joint,

loads Fc is: