ಓಂ ಶ್ರ ೀ ಗಣೀಶಯನಮಃ

One-Dimensional Problems
• The potential energy and the stress-stain and stain-displacement relationships are used
in developing the finite element method for a one-dimensional problem.

→ 3-Dimension

→ 1-Dimension

• The basic procedure is the same for two- and three-dimensional problems.

• For the one-dimensional problem, the stress, stain, displacement, and loading depend
only on the variable x.

• That is, the vectors u, 𝝈, 𝝐, T, and f are reduced to

• Furthermore, the stress-strain, stain-displacement and the differential volume dV are

• The loading consists of three types:

the body force f

the traction force T

the point load Pi

• A body force is a distributed force acting on every elemental volume of the body (force per
unit volume). Example; Self-weight due to gravity

• Traction force is defined as force per unit area.

• For the one-dimensional problem considered here,
the traction force is defined as force per unit length.

• This is done by taking the traction force to be the

product of the force per unit area with the perimeter
of the cross-section. (Frictional resistance, viscous
drag and surface shear are examples of traction forces
in one-dimensional)

• Pi is a force acting at a point i

One-dimensional bar loaded by
• ui is the x displacement at that point traction, body and point loads
Finite Element Modelling
Element Division: Consider the bar shown in Fig.

Finite element modeling of a bar

• The first step is to model the bar as a stepped shaft, consisting of a discrete
number of elements, each having a uniform cross section.

• The average cross-sectional area within each region is evaluated and then
used to define an element with uniform cross section.

• A four-element, five node finite element model is shown in Fig.

• In the finite element model, every element connects to two nodes.

• Cross-sectional area, traction, and body forces are considered as constant

for each element but can differ in magnitude from element to element.

• Better approximation are obtained by increasing the number of elements.

• It is convenient to define a node at each location where a point load is

applied.
Numbering Scheme

• In a one-dimensional problem, every node is permitted to displace only in the

± 𝒙 direction.

• Thus, each node has only one degree of freedom (dof).

• The displacement along each dof are denoted by Q1 , Q2 ,Q3 , Q4 , Q5 ,.

𝑇 𝑇
• The column vectors 𝐐 = Q1 , Q2 , Q3 , Q4 , Q5 and 𝐅 = F1 , F2 , F3 , F4 , F5 are called
global displacement vector and global load vector respectively.

• The sign convention used is that a displacement or load has a positive value if acting along
the +𝒙 direction.

• The node 1 is fixed, which implies Q1 = 0.

Q and F vectors
Element connectivity
• Each element has two nodes; the element
connectivity information can be represented as
shown Fig.

• In the connectivity table, the headings 1 and 2 refer to

local node numbers of an element

• The corresponding node numbers on the body are

called global numbers

• In this simple example, the connectivity can be easily

generated since local node 1 is the same as the
element number e, and local node 2 is e +1.

• Other ways of numbering nodes or more complex

geometries suggest the need for a connectivity table.
Coordinates and Shape Functions

• Consider a typical finite element e in Fig a.

• In the local number scheme, the first node will be numbered 1 and the second node 2.
• The notation x1 = x-coordinate of node 1, x2 = x-coordinate of node 2 is used.
• In FEA we define a natural or intrinsic coordinate system, denoted by ξ , as

Typical element in x- and ξ- coordinates

• From Fig. b, we see that ξ = -1 at node 1 and ξ = 1 at node 2.

• The length of an element is covered when ξ changes from -1 to 1.

• In FEA, this system of coordinates are used in defining shape functions, which are used in
interpolating the displacement field.
• The unknown displacement field within an element will be interpolated by a linear
distribution as shown in Fig.

Linear interpolation of the displacement within the element

• This approximation becomes increasingly accurate as more elements are considered in the
model.

• To implement this linear interpolation, linear shape functions will be introduced as

 N1 = 1 at ξ = -1, and N1 = 0 at ξ = +1 N2 = 0 at ξ = -1, and N2 = 1 at ξ = +1
N1 is a straight line between the two points. N2 is a straight line between the two points.

(a) Shape function N1 (b) Shape function N2

 (a) Shape function N1 (b) Shape function N2

At node 1, N1 = 1 and N2 = 0 → u1 = q1
At node 2, N2 = 1 and N1 = 0 → u2 = q2
Displacement field
u = N1 q1 + N2 q2
(c) Linear interpolation using N1 and N2
• The graph of the shape function N1 in Fig. (a) is obtained from

as N1 = 1 at ξ = -1, N1 = 0 at ξ = 1, and N1 is a straight line between the two points.

• Similarly, the graph of N2 in Fig. (b) is obtained from

• Once the shape functions are defined, the linear displacement field within the element
can be written in terms of the nodal displacements q1 and q2 as
Or in matrix form as
u = Nq

where

and

• In these equations, q is referred to as the element displacement vector.

• It is verified that u = q1 at node 1, u = q2 at node 2, and that u varies linearly (Fig.).

• It may be noted that the transformation from x to ξ can be written in terms of N1 and N2 as
Comparing Eqs. and

We see that both the displacement u and the coordinate x are interpolated within the
element using the same shape functions N1 and N2.

• This is referred to as the isoparametric formulation.

• Shape functions need to satisfy the following:

1. First derivatives must be finite within an element.

2. Displacements must be continuous across the element boundary.

Example:

Referring to Fig. do the following:

(a) Evaluate ξ , N1 and N2. at point P.

(b) If q1 = 0.003 in. and q2 = -0.005 in. determine the value of the displacement q at point P
Solution:

(a) Evaluate ξ , N1 and N2. at point P.

Using equation

The ξ coordinate of point P is given by

2
𝝃𝑷 = 24 − 20 − 1 = - 0.5
(36−20)
From Eqs.

𝝃𝑷 = - 0.5 and

N1 = 0.75 and N2 = 0.25

(b) Using Eq.

𝑢𝑝 = 0.75 0.003 + 0.25 −0.005

= 0.001 in.
The strain displacement relations for one-dimensional problems is given as

where

Upon using the chain rule of differentiation, we obtain

From relationship between x and ξ in equation

We have
From,

we have
and

Thus Eqn. ×

yields

This can be written as

 →

where (1 x 2) matrix B, called the element strain-displacement matrix, is given by

and is the (2 x 1) displacement matrix q

Use of linear shape functions results in a constant B matrix and, hence, is a constant strain
within the element.

The stress, from Hooke’s law, is

• The stress given by above equation is also constant within the element.

• For interpolation purposes, however, the above stress equation can be considered to be
the value at the centroid of the element.

• The expressions u = Nq, 𝜖 = Bq and 𝜎 = EBq relate the displacement, strain and stress,
respectively, in terms of nodal values.

• These expression will now be substituted into the potential-energy expression for the bar
to obtain the element stiffness and load matrices.
The Potential-energy approach
The general expression for the potential energy (for one-dimension) is given by

Since the continuum has been discretized into finite elements, the expression for П
becomes

The above equation can be written as

where
is the element strain energy

Element Stiffness Matrix

Consider the strain energy term

Substituting for 𝝈 = 𝑬𝑩𝒒 and 𝝐 = 𝑩𝒒 into above Eqn. yields

 or as

In the finite element model (one-dimensional bar),

The cross-sectional area of element e, denoted by Ae , is constant.

Also, B is a constant matrix.

Further, the transformation from x to ξ yields →

→
Now the element strain energy

can be written as

where Ee is Young’s modulus of element e.

Substituting and (where 𝑥2 − 𝑥1 = 𝑙𝑒 )

we get,
which results in,

This equation is of the form

where the element stiffness matrix ‘ ke ’ is given by

we can observe here the similarity of the stain energy expression in the above equation
with the strain energy in a simple spring, which is given as

Also, observe that ke is linearly proportional to the product AeEe and inversely proportional
to the length ℓe
Force Terms
The force terms in the potential energy are shown in shaded area

Body force Term

The element body force term is considered first

Substituting for u

we have
As Ae and 𝒇 are constant within the element, the above equation can be written as

where

Substituting and integrating we get

Alternatively,
න 𝑵𝟏 𝒅𝒙 is simply the area under the 𝑁1 curve as shown in Fig.

which equals
𝟏 𝒍𝒆
× 𝒍𝒆 × 𝟏 =
𝟐 𝟐

𝟏 𝒍𝒆
Similarly, න 𝑵𝟐 𝒅𝒙 = × 𝒍𝒆 × 𝟏 =
𝟐 𝟐
 where

Now this body force term is reduced to

which is of the form

The right side of this equation is of the form

Displacement × Force

Thus, the element body force vector, fe , is identified as

As Aeℓe is the volume of the element and 𝒇 is the body force per
unit volume, we see that Aeℓe 𝒇 gives the total body force acting
on the element.
1
The factor 2 tells that this total body force is equally distributed to
the two nodes of the element.

Traction force Term

Consider the traction force term shown in the potential energy equation, i.e.,
Substituting for

We get
Since the traction force T is constant within the element, we have

here

Thus the above equation can be written as

where the element traction force vector is given by

Here also this total traction force is equally distributed to the two nodes of the element.
At this stage, element matrices ke , fe , and Te have been obtained. After that we go for
element connectivity.

Then the total potential energy equation can be written as

Which can be written in the form

Where K is the global stiffness matrix, F is the global load vector, and Q is the global
displacement vector.
Example: In the finite element model shown in Fig., K is a (5 x 5) matrix, and Q and F are
each (5 x 1) vectors.

Using the element connectivity information, K is obtained as follows:

The elements of each ke are placed in the appropriate

locations in the larger K matrix.

And overlapping elements are then summed.

The F vector is similarly assembled.

Assembly of Global Stiffness Matrix and Load Vector

We noted earlier that the total potential energy can be written in the form

can be written as

This step involves assembling K and F from element stiffness and force matrices.
Elemental Stiffness Matrix will be

Global Stiffness Matrix will be

1 2 3 4 5
1 𝒆𝟏 −1 0 0 0 1
𝐸𝐴 −1 1 + 1 𝒆𝟐 −1 0 0 2
𝐊= 0 −1 1 + 1 𝒆𝟑 −1 0 3
𝑙
0 0 −1 1 + 1 −1 4
0 0 0 −1 𝒆𝟒 1 5
Assembly of the stiffness matrix K from element stiffness matrices ke

Referring to the finite element model shown in Fig.,

consider the strain energy in element 3.

We have

Substituting for k3

And for element 3, we have

Thus, we can write U3 as

From this equations, we see that elements of the matrix k3 occupy the third and fourth
rows and columns of the K matrix.

Consequently, when adding element-strain energies, the elements of ke are placed in the
appropriate locations of the global K matrix.
Based on the element connectivity; overlapping elements are simply added.

This assembly can be denoted symbolically as

Similarly, the global load vector F is assembled from element-force vectors and point load as

In actual computation, K is stored in banded or skyline form to take advantage of symmetry

and sparsity.
• Consider the bar as shown in Fig. For each element i, Ae
and ℓe are the cross-sectional area and length,
respectively.

• Each element i, is subjected to a traction force Ti per unit

length and a body force 𝒇 per unit volume.

• The Young’s modulus of the material is E.

• A concentrated load P2 is applied at node 2.

• Now assemble the structural stiffness matrix and nodal

vector.
The element stiffness matrix for each element i is obtained from Eq.

= 𝒌(𝒊)

The element connectivity table is

The element stiffness matrices can be “expanded” using the connectivity table and then
summed (or assembled) to obtain the structural stiffness matrix as follows:
The element stiffness matrices can be “expanded” using the connectivity table and then
summed (or assembled) to obtain the structural stiffness matrix as follows:

which gives
This “expansion” of element stiffness matrices as shown above is merely for illustration
purpose and is never explicitly carried out in the computer, since storing zeroes is inefficient.
Instead, K is assembled directly from ke using the connectivity table.

The global load vector is assemble as

where
Properties of K
1. The dimension of the global stiffness K is (N x N), where N is number of nodes.

2. K is symmetric.

3. K is a banded matrix. That is, all elements outside of the band are zero.

4. In the above example, K can be compactly represented in banded form as

• Note that Kbanded is of dimension [N x NBW] where NBW is the half-band width.

• In one-dimensional problems (as in the above example), the connectivity of element (i)
is i, i+1.

• In such cases, the banded matrix has only two columns (NBW = 2).

• In two and three dimensions, the direct formation of K in banded or skyline form from
the element matrices involves some bookkeeping.

• The following general formula for the half-bandwidth can be verified.

 For example, consider a four-element model of a bar that is numbered as shown in Fig. a.

Using the formula,

we have

The numbering scheme in Fig. a is bad since K is almost “filled up” and consequently requires
more computer storage and computation.

Fig. b shows the optimum numbering for minimum NBW.

Now the potential energy or Galerkin’s approach has to be applied, considering the
boundary conditions of the problem, to yield the finite element (equilibrium) equations.

Solution of these equations yields the global displacement vector Q.

Then the stresses and reaction forces can be calculated.

The Finite Element Equations; Treatment of Boundary Conditions

After using a discretization scheme to model the continuum, we have obtained an

expression for the total potential energy in the body as

Now we have arrived at the equations of equilibrium, from which we can determine nodal
displacements, element stresses, and support reactions.

The minimum potential-energy theorem is now considered.

This theorem states that “of all possible displacements that satisfy the boundary conditions
of a structural system, those corresponding to equilibrium configurations make the total
potential energy assume a minimum value”.

Consequently, the equations of equilibrium can be obtained by minimizing the potential

energy with respect to Q.

Boundary conditions are usually of the type

That is, the displacements along dofs 𝑝1 , 𝑝2 , ... 𝑝𝑟 , are specified to be equal to 𝑎1 , 𝑎2 , ... 𝑎𝑟 ,
respectively.

In other words, there are r number of supports in the structure, with each support node
given a specified displacement.
For example, consider the bar shown in Fig. There is
only one boundary condition in this problem, 𝑄1 = 0.

It is noted that this treatment of boundary conditions is

also applicable to two- and three-dimensional problems
as well. For this reason, the term dof is used here
instead of node, since a two-dimensional problem will
have two degrees of freedom per node.
It should be noted that improper specification of boundary conditions can lead to error in
results.

Boundary conditions eliminate the possibility of the structure moving as a rigid body.

Two approaches have been used for handling specified displacement boundary conditions;
the elimination approach and the penalty approach.

Elimination Approach

To illustrate the basic idea, consider the single boundary condition 𝑄1 = 𝑎1 . The equilibrium
equations are obtained by minimizing Π with respect to Q, subjected to the boundary
condition 𝑄1 = 𝑎1.
For an N dof structure, we have

The global stiffness matrix is of the form

Note that K is a symmetric matrix.

The potential energy can be written in expanded form as

𝑄1 𝐹1
1 𝑄2 − 𝐹2
Π = 2( . .
)
𝑄𝑁 𝐹𝑁
The potential energy can be written in expanded form as

𝑄1 𝐹1
1 𝑄2 𝐹2
Π= ( .
−
.
)
2
𝑄𝑁 𝐹𝑁

If the boundary condition 𝑄1 = 𝑎1 is substituted into this expression for Π, we obtain

Note that the displacement 𝐐𝟏 , has been eliminated in the potential-energy expression.

Consequently, the requirement that Π take on a minimum value implies that

Thus, after differentiation we obtain

These finite element equations can be expressed in matrix form as

Now we observe that the (N-1 x N-1) stiffness matrix is obtained simply by deleting or
eliminating the first row and column (in view of 𝑄1 = 𝑎1 ) from the original (N x N) stiffness
matrix.

This can be denoted as KQ = F

where K is a reduced stiffness matrix obtained by eliminating the row and column
corresponding to the specified of “support” dof.

The above equations can be solved for the displacement vector Q using Gaussian
elimination.
Note that the reduced K matrix is nonsingular, provided the boundary conditions have been
specified properly.

On the other hand, the original K matrix is a singular matrix

(i.e. det (original K) = 0)

Once Q has been determined, the element stress can be evaluated using 𝝈 = 𝑬𝑩𝒒, where q
for each element is extracted from Q using element connectivity information.

Assume that displacements and stresses have been determined. It is now necessary to
calculate the reaction force 𝑹𝟏 at the support.

This reaction force can be obtained from the finite element equation for node 1:
• Here, 𝑸𝟏 , 𝑸𝟐 , ... 𝑸𝑵 , are known.

• 𝑭𝟏 , which equals the load applied at the support, is also known.

• Consequently, the reaction force at the node that maintains equilibrium, is

• Note that the elements 𝐾11 , 𝐾12 , ... 𝐾1𝑁 , which form the first row of K, need to be stored
separately.

• This is because K is obtained by deleting this row and column from the original K.