Null 10

FINITE DIFFERENCE APPROACH TO STABILITY
AND VIBRATION ANALYSIS OF LINE

CONTINUUM
BY
UGOCHUKWU OBINNA OKPARA (B.ENG)

REG. NO: 20114775898
A THESIS SUBMITTED TO THE POSTGRADUATE SCHOOL,

FEDERAL UNIVERSITY OF TECHNOLOGY, OWERRI
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR

THE AWARD OF MASTER OF ENGINEERING (M.ENG)
DEGREE IN CIVIL ENGINEERING (STRUCTURES)
JANUARY, 2015.
i
© Federal University of Technology, Owerri
ii
CERTIFICATION
This is to certify that this work on “Finite Difference Approach to Stability and
Vibration Analysis of Line Continuum” was carried out by OKPARA, U. O. with
Reg. No. 20114775898 in partial fulfillment of the requirement for award of Master
of Engineering (M.Eng.) Degree in Civil Engineering (Structures) in School of
Engineering and Engineering Technology, Federal University of Technology,
Owerri.
____________________ __________________
Engr. Dr J. C. Ezeh Date
Supervisor
____________________ _________________
Engr. Dr. O. M. Ibearugbulem Date
Supervisor
____________________ _________________
Engr. Dr. Mrs. B. U. Dike Date
Head of Department
__________________ _________________
Engr. Prof E. E. Anyanwu Date
Dean of School of Engineering and
Engineering Technology
____________________ _________________
Engr. Prof. Mrs. K. B. Oyoh Date
Dean of Postgraduate School
____________________ _________________
Engr. Prof. O. E. Alutu Date
External Examiner
iii
DEDICATION
This project report is dedicated to God Almighty and to my family.
iv
ACKNOWLEDGEMENTS
I wish to appreciate my supervisors, Engr. Dr. J.C. Ezeh and Engr. Dr. O.M.
Ibearugbulem for their immense effort in initiating and supervising this research
project. Their guidance helped me during the period of this research and writing of
this thesis. I could not have imagined having better supervisors for this research
project, for they have been of great help and encouragement to me, may the
Almighty God bless and reward them in all their endeavors in life.
I will also like to thank my lecturers, Engr. Dr. D.O. Onwuka, Engr. Dr. Mrs.
Okere, Engr. M.S.W Mbajiogu, Engr. Dr. L.O. Ettu, Engr. U.C. Anya, Engr. C.F.
Njoku, Engr. Mrs. J. Arimanwa, Engr. Dr. H. U. Nwoke, Engr. Dr. Osugwu, Engr.
Dr. B. C. Okoro and Engr. Dr. Dike, B. U for taking out their time to make sure we
completed our coursework successfully. May God bless you all. I would like to
express my heartfelt gratitude to my parents Hon. and Mrs. B. I. Okpara for their
care, encouragement and financial support to me throughout the period of this work.
Also to everyone who has been of assistance by way of advice, finance, loaning of
literature materials and encouragement in this research, may God bless you all.
Above all, this research work would not have been a reality without the
Almighty God who provided life, wisdom and sound health for this work. May his
name be praised, adored, glorified and worshiped forever and ever, Amen.
v
TABLE OF CONTENTS
Title page
Certification i
Dedication ii
Acknowledgements iii
Table of Contents iv
List of Tables v
List of Nomenclature viii
Abstract ix
CHAPTER ONE: INTRODUCTION 1
1.1. Background of Study 1

1.2. Statement of Problem 2
1.3. Objective of Study 2
1.4. Justification of Study 3
1.5. Scope of Study 3
CHAPTER TWO: LITERATURE REVIEW 4
2.1. Historical Development. 4

2.2. Previous work on Stability Analysis of Line Continuum 5
2.3. Previous work on Vibration Analysis of Line Continuum 7
2.4. Approaches to Stability and Vibration Analysis 10
2.4.1. Approaches to Stability Analysis 10
2.4.1.1. Equilibrium Approach 10
2.4.1.2. Energy Approach 11
2.4.1.3. Principle of Conservation of Energy 12
2.4.1.4. Calculus of Variation 13
vi
2.4.1.5. Principle of Minimum Potential Energy 14
2.4.1.6. Rayleigh-Ritz technique 15
2.4.1.7. Galerkin’s technique 17
2.4.1.8. Numerical technique 18
2.4.1.9. Finite Difference method 18
2.4.1.10. Improved Finite Difference method 19
2.4.2. Approaches to Vibration Analysis 20
2.4.2.1. Theoretical Analysis of transverse vibration 20
2.4.2.2. Approximate method. Rayleigh’s method 22
2.4.2.3. Rayleigh-Kitz method 23
2.4.2.4 Numerical method. Improved Finite Difference Method 24
CHAPTER THREE: METHODOLOGY 26
3.1. Finite Difference Formulation. 26

3.2. Stability Analysis Approach. 33
3.2.1. Case 1 (n=3): For three flexible nodes 35
3.2.2. Case 2 (n=5): For five flexible nodes 40
3.2.3. Case 3 (n=7): For seven flexible nodes 47
3.3. Vibration Analysis Approach. 54
3.3.1. Case 1 (n=3): For three flexible nodes 54
3.3.2. Case 2 (n=5): For five flexible nodes 58
3.3.3. Case 3 (n=7): For seven flexible nodes 63
CHAPTER FOUR: RESULTS AND DISCUSSION 69
4.1. Results of Stability Analysis of Line Continuum 69

4.1.1. Case 1 (n = 3) 70
4.1.2. Case 2 (n = 5) 70
4.1.3. Case 3 (n = 7) 71
4.2. Results of Vibration Analysis of Line Continuum 72
4.2.1. Case 1 (n = 3) 73
vii
4.2.2. Case 2 (n = 5) 74
4.2.3. Case 3 (n = 7) 75
4.3. Discussion of Results 76
4.3.1. Stability Analysis of Line continuum 76
4.3.2. Vibration Analysis of Line Continuum 77
CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS 79
5.1. Conclusions 79
5.2. Contribution to knowledge 79
5.3. Recommendations 79
REFERENCES 80
APPENDICES 84
viii
LIST OF TABLES
Table 3.1: Schematic representation of Fourth – order Finite Difference
Expression 33
Table 3.2: Line continua and their boundary Conditions 34
Table 4.1: Result Data for P - R support and Exact Result 76
Table 4.2: Result Data for C - C support and Exact Result 76
Table 4.3: Result Data for C - R support and Exact Result 76
Table 4.4: Result Data for P - R support and Exact Result 77
Table 4.5: Result Data for C - C support and Exact Result 77
Table 4.6: Result Data for C - R support and Exact Result 78
ix
LIST OF NOMENCLATURE
A Cross-sectional area
EI Flexural rigidity
I second moment of Inertia
M Bending moment
P Axial load
𝑃𝑐𝑟 /N critical buckling load
Q Shear force
U Strain energy
y/w Lateral displacement/deflection
l length of line continuum
W Change in work
U Change in strain energy
 Total potential energy
 Density
 Natural frequency
𝑑𝑦
yI First derivative of y with respect to x
𝑑𝑥
𝑑2𝑦
yII Second derivative of y with respect to x
𝑑𝑥 2
𝑑𝑤
wI First order derivative of w with respect to x
𝑑𝑥
𝑑2𝑤
wII Second order derivative of w with respect to x
𝑑𝑥 2
𝑑3𝑤
wIII Third order derivative of w with respect to x
𝑑𝑥 3
𝑑4𝑤
wIV Fourth order derivative of w with respect to x
𝑑𝑥 4
x
ABSTRACT
This research work presents Stability and Vibration Analysis of Line Continuum
with various boundary conditions using Finite Difference Method developed from
Polynomial series expansion, truncated at the fourth term. The boundary conditions
considered were pin – roller support (P – R), clamped – clamped support (C – C)
and clamped – roller support (C – R). Considering the three boundary conditions
stated above, three cases of line continuum were studied using the various boundary
conditions which are Line continuum with three flexible nodes, Line continuum
with five flexible nodes and Line continuum with seven flexible nodes. The Finite
Difference pattern derived from polynomial expansion series was substituted into
each case of line continuum to obtain matrices as well as Eigen values for stability
and vibration analysis respectively and the natural frequency and buckling load
were then sought. The values from this study were compared to the exact result
from previous work. For Stability analysis, the results of critical buckling load
obtained for P – R support has a percentage difference ranging from -5.03% to -
1.31% as compared to exact result. For C – C support, it gave a percentage
difference of -18.94% to -5.03% while that of C – R support gave a percentage
difference of -11.19% to -2.88%, all when compared to exact result. For Vibration
analysis, the natural frequency obtained for P – R support gave a percentage
difference ranging from -2.75% to -0.65%. For C – C support , it gave a percentage
difference of -13.1% to -3.35% and C – R support revealed a percentage difference
ranging from -7.51% to -1.84%, all when compared to values from exact result. The
percentage difference obtained gave negative differences which depicts Finite
Difference method provides a lower bound solution. These results showed that this
Finite Difference Method, truncated at the fourth – order gave approximate
solutions close to the exact results from previous works and hence, can be applied
in structural engineering.
Keywords: Line Continuum, Finite Difference, Polynomial Series, Boundary

Conditions, Flexible Nodes, Buckling Load, Natural Frequency
xi
CHAPTER ONE
INTRODUCTION
1.1. BACKGROUND OF STUDY.

The primary objective of any structural member is to carry and transmit loads safely
and to perform well in a serviceable condition. The importance of the subject is
evident from the history of structural collapse caused by neglect or
misunderstanding of the stability and vibration aspects of design.
It is a well-known fact that most buildings collapse due to instability of the column
elements rather than foundation failures, of excessive deflections and resonances
due to vibrations at natural frequencies which induce inertia stresses and results in
high stresses in structural members (Ibearugbulem et al. 2013). Thus, it is important
to always verify the critical buckling load and natural frequency of the continuum.
These are many studies on the theory and analysis of line continuum, the oldest and
well-known is Euler or Equilibrium Approach, which is a classical analysis. It gives
exact solutions for buckling and vibration analysis of line continuum by applying
direct integration of the governing differential equation and thus satisfying the
boundary conditions of the system. Unfortunately, it is insufficient to solve
indeterminate problems. Hence, other approaches became very necessary and
imperative.
Energy and Numerical approaches are also other method of analysis of line
continuum. Example of energy methods are Galerkin, Rayleigh-Ritz etc. They aim
at minimizing the total potential energy functional in order to get algebraic
equations.
Numerical methods of analysis gives approximate solutions to problems of stability
and vibration analysis of line continuum.
One of the numerical approaches is the Finite Difference Method, which applies a
mathematical discretization of the continuum to evaluate derivative of functions,
yielding a set of algebraic equations (Szilard, 2004). Due to its slow convergence
and inaccuracy to solutions of higher-order derivatives and large number of mesh
1
points, the Improved Finite Difference Method was developed. Therefore,
Improved Finite Difference technique is applied when high accuracy in the finite
difference derivatives is required (Szilard, 2004). The general tool for derivation of
the improved derivatives is the Polynomial expansion series.
The work of Ibearugbulem et al. (2014) studied the pure bending analysis of
line continuum using a new Finite Difference Approach, determining the deflection
of the line continuum only without considering the critical bucking load which can
cause instability in the member or the natural frequency which can also make the
member to vibrate and collapse. Hence, the basis of this presents study.
1.2. STATEMENT OF PROBLEM.

Past researchers made use of simple higher – order differential techniques to obtain
solutions and results which were much higher than exact approximate solutions
gotten from other methods (Greenwood, 1961). Also, some analysis was done using
ordinary finite difference technique which is slow to converge and sometimes gives
inaccurate solutions. Hence, the need for a new finite difference method.
1.3. OBJECTIVE OF STUDY.

The main objective of this research is Finite Difference Approach to Stability and
Vibration Analysis of Line Continuum. The specific objectives of this research are
as follows:
i. To formulate a Finite difference pattern, using Polynomial series expansion.
ii. To apply the Finite difference pattern in stability and free-vibration analysis
of line continuum.
iii. To determine the fundamental critical bucking load and natural frequency of
a line continuum satisfying all necessary boundary conditions.
iv. To examine how close, the obtained solution are to the exact solution.
2
1.4. JUSTIFICATION OF STUDY.
This study would help increase the insufficient number of literature concerning the
use of Finite Difference technique. It would provide a new finite difference pattern
in handling stability and vibration analysis for line continuum problem. Also, it
would expose the finite difference technique as a veritable tool in solving line
continuum with different boundary conditions and serve as a reference research
material for future studies in the related field of study.
1.5. SCOPE OF STUDY.

This research studied the Stability and Vibration Analysis of Line Continuum of
different boundary conditions using Finite Difference Method. The method was
developed using polynomial expansion series and truncated at the fourth–order
derivative. The various boundary condition covered in this research include.
i. Pin – Roller support (P – R)
ii. Clamped – Clamped support (C – C)
iii. Clamped – Roller support (C – R)
Also, the line continuum was studied under 3 cases:
Case 1: Dividing the line continuum into four equal lengths using three nodes.
Case 2: Dividing the line continuum into six equal lengths using five nodes.
Case 3: Dividing the line continuum into eight equal lengths using seven nodes.
3
CHAPTER TWO
LITERATURE REVIEW
2.1. HISTORICAL DEVELOPMENT.
A line continuum, in this context, refers to a one – dimensional field structural

member whose configuration is straight either in the x-axis or y-axis. Based on this,
a line continuum can be referred to as either beam member or a column member.
Aristizabal –Ochoa (1993) proposed a method that evaluates the static, stability and
vibration response of non-prismatic beams and columns, by which the various
stiffness matrices and the consistent mass, damping and load matrices can be
obtained for any- type of non- prismatic straight beam without using approximate
shape function or breaking up the element into sub-elements. The simple and
innovative matrix method utilizes the classical conjugate beam method and
common integration technique such as Gaussian quadrature. The conjugate beam
method is first utilized to determine the basic flexural stiffness coefficient and then
by simple numerical integration, the remaining characteristic matrices of the beams
and columns can be determined including the mass, damping and geometric
matrices and load vectors.
An analysis and sample results of the lateral buckling and vibration of a

compressively loaded column whose cross-section is a piecewise constant along its
length was presented by Lake and Mikulas - Jr (1991). A parametric structural
efficiency analysis was performed using a non-dimensionalized set of governing
equations to determine the optimum ratio between the lengths of the centre section
and the outboard section based on both buckling load and vibration frequency
requirements. Both buckling and vibration characteristics of the column were
determined from a numerical solution of the exact Eigen value problems.
Ranjbaran et al. (2008) developed a mathematical formulation for the buckling and
vibration analysis of line continuum, assuming the member to be non-uniform and
4
cracked. Based on the calculus of variations, the problem is expressed as an
optimization problem and a technique of optimization was used for analyzing of the
buckling load. Considering the similarity between governing equation for buckling
and free vibration, the fundamental frequency and mode shape was computed with
the same method.
Ibearugbulem et al. (2013) presented a new stiffness matrix for stability and
vibration analysis of line continuum, the new 5x5 stiffness matrices for the analysis
was developed using Energy Variational Principle with five term Mclaurin’s
Polynomial Series as the shape function. The newly developed stiffness matrices
gave approximate solution for critical buckling load and natural frequency of line
continuum. Also, a new shape function for analysis of line continuum by Direct
Variational Calculus was developed by Ibearugbulem et al. (2013).The new
polynomial shape function was derived from the exact general solutions of the
governing differential equation, these peculiar polynomial shape function were
applied in analyzing the line continuum.
2.2 PREVIOUS WORKS ON STABILITY ANALYSIS OF LINE

CONTINUUM
Stability analysis began with Euler’s solution of buckling of an elastic column.

Most basic linear elastic problems of structural stability were solved by the end of
the 19th century. Although further solutions have been appearing as new structural
types were being introduced (Bažant, 2000). The twentieth century has witnessed a
great expansion of the stability theory into non-linear behavior, caused by large
deflections or by non-linearity of the constitutive law of the material (Bažant,
2000).
Iremonger (1980) determined buckling of non- uniform column by Finite

Difference Method using a matrix iteration solution technique, In calculating the
buckling loads, none of the column need be divided into more than 8 segments to
give an accuracy within 0.1% provided “h2” extrapolation of the solution is used.
5
To obtain a successful solution, the fourth-order equation would have to be
formulated using forward/ backward differences in the region of the step change of
the section.
Galambos (1998) gave a general introduction to the use of the finite element
method to conduct stability analysis. The governing differential equation for an
axially loaded prismatic column is provided as an example along with the statement
of total potential energy. He then goes on to discuss weighted- residual and
variational formulation of the problem. Coverage of higher- order analysis was
briefly provided and extensive recommendations for the finite element software
were given.
Saha and Banu (2007) presented a method for identifying the buckling load of a line
continuum using a technique named “Multi-Segment Integration technique”
considering the various boundary value problem for the linear continuum. The
governing differential equation and boundary condition are given in a matrix form
which are resolved to give approximate solutions.
Aristizabal – Ochoa (2008) derived and developed in a classical manner a new set
of slope – deflection equations that include the combined effects of shear and
bending deformations for Timoshenko beam – columns symmetrical cross- section.
The method also included the shear component of the applied axial forces (Haringx
Model) and was utilized in the stability and second- order analysis of the structural
member. The proposed method was based on the “Modified” stability functions for
beam- columns with semi-rigid connections (Aristizabal-Ochoa, 2004, 2008).
Pinarbasi (2011) analyzed buckling of beams using Homotopy Perburtion Method

(HPM). This analytical technique is applied to two fundamental stability
problems:(a) A simple supported rectangular beam under pure bending and (b) A
cantilever rectangular beam subjected to a concentrated load at their free ends. In
the analysis, two different types of stiffness variations, Linear, Exponential
variations were considered.
6
Šapalas et al. (2005), in their paper, dealt with a theoretical and numerical analysis
of tapered beam- column subjected to a bending moment and at axial force using a
standard Finite Element code for the numerical estimation of a critical load
multiplier. It was assumed that the critical force of an axially loaded tapered column
could be calculated in an analogous way as for uniform member just with an
additional correction factor.
In his article, Javidinejad (2012) examined the buckling behavior of an I- beam

under combined axial and horizontal side loading. He showed that the actual
application location of the axial loading governs the buckling behavior of the
member. Theoretical formulation was developed to determine the critical buckling
load for such combined loading configuration from the elastic static theory and
Finite Element Analysis was utilized to apply the axial load on the beam at various
configurations as it was shown that this application location determine the buckling
behavior and the critical load of the buckling of the beam.
Sapountzakis and Tsiatas (2005) developed a Boundary Element Method (BEM) for
the flexural buckling analysis of composite Euler- Bernoulli beams of arbitrary
variable cross- section, using variable coefficients of the governing differential
equations. The beam was subjected to a compressive centrally applied load together
with arbitrarily axial and transverse distributed loading, which its edges are
restrained by the most general linear boundary conditions, which were solved by
employing a boundary integral equation approach.
2.3. PREVIOUS WORKS ON VIBRATION ANALYSIS OF LINE

CONTINUUM
Greenwood (1961), in his thesis, made use of simple and higher- order differential
methods for the solution for the natural frequency of vibration of a uniform
continuous beam. Three different methods of boundary conditions representation
were given.
7
Chondros et al. (1998) developed a continuous cracked beam vibration theory for
the lateral vibration of cracked Euler- Bernoulli beams with single- edge or double-
edge open cracks; they used the Hu- Washizu- Bar variational formulation to
develop the differential equation and the boundary conditions of the cracked beam
as a one- dimensional continuum. The displacement field about the crack was used
to modify the stress and displacement field throughout the bar. They extended their
theory for a beam with breathing crack (Chondros et al. 2001).
Shifrin et al. (1999) proposed a technique for calculating natural frequencies of a

vibrating beam with arbitrary number of transverse open cracks. The main feature
of their method is related to decreasing the dimension of the matrix involved in the
calculation, so that reduced computation time is required for evaluating natural
frequencies compared to alternative methods
Mesut and Turgut (2006) analyzed free vibration of beams with different boundary
conditions using third-order shear deformation theory. The boundary condition was
satisfied using Lagrange multipliers. To apply the Lagrange equations, trial function
denoting the deflection and the rotations of the cross- sections of the beam are
expressed in polynomial forms.
Chopade and Barjibhe (2013), in their paper, focused on the theoretical analysis of
transverse vibration of fixed free beam and investigate their mode shape frequency.
All the theoretical values were analyzed with the numerical approach method by
using Finite Element package, ANSYS to also find out their mode shape and
frequencies with high accuracy, which concluded that both were in good agreement.
Demirdag and Murat (2009) presented a study on elastically supported Timoshenko

column with attached masses for free vibration analysis using fuzzy neural network.
In the study, neuro- fuzzy algorithm was used to obtain frequencies of model by
establishing neuro fuzzy frequency estimation (NFFE) models, which reduces
computational effort and time to obtain free vibration frequencies. Governing
equation of free vibration includes effects of bending and shear deformation with
8
rotary inertia of columns. Eigen frequencies are determined by transfer matrix
method (TMM) using separation of variable method.
Chati et al. (1997), in their article, addressed the problem of vibrations of a cracked
beam. It focused on the modal analysis of a cantilever beam with a transverse edge
cracks. In an attempt to define effective natural frequencies for the piecewise –
linear system, the idea of a “bilinear frequency” was utilized. They were obtained
by computing the associated frequencies of each of the linear pieces of the
piecewise – linear system. Finite element method was used to obtain natural
frequencies in each linear region.
Gupta and Sharma (2002) analyzed free vibrations of rectangular beams having two
linear variations in thickness along the length by classical theory of beams. The
solution of the equations of motion was obtained by Frobenius method. Numerical
results for natural frequencies and normalized mode shapes were computed for first
four normal modes for a clamped – clamped and cantilever beam.
Mao (2012) employed the Adomian Decomposition Method (ADM) to investigate

the free vibrations of a uniform cracked Euler – Bernoulli Beams. The proposed
method was used to analyze the vibration of beams consisting of an arbitrary
number of cracks through a recursive way. Based on the ADM and employing some
simple mathematical operations, the natural frequencies and the corresponding
closed – form series solution of the mode shapes can be obtained by solving a set of
algebraic equations with only three unknown parameters. It was shown that ADM
offers an accurate and effective method of free vibration analysis of beam with
arbitrary number of cracks under general boundary conditions.
Free vibrations characteristics of Timoshenko beams having different boundary

conditions was analyzed by Kocatürk and Şimşek (2005), using Lagrange
Equations. The constraints conditions of support are taken account of by using
Lagrange multipliers. According to their study, in applying the Lagrange equation,
trial functions denoting the deflection and the rotation of the cross – section of the
9
beam are expressed in the polynomial form. By using the Lagrange equations, the
problem was reduced to the solution of a system of algebraic equations. The first
eight Eigen values of Timoshenko beam were calculated and tabulated for different
thickness – to – length ratios and it was believed that the results proved useful to
designers and researchers.
2.4. APPROACHES TO STABILITY AND VIBRATION ANALYSIS.
2.4.1. APPROACHES TO STABILITY ANALYSIS.
The concept of a critical load of an elastic structure at which equilibrium bifurcates

was introduced by Euler, who also provided the solutions of critical loads of a line
continuum with various end restraints. Various methods can be used to determine
the elastic critical buckling load of line continuum and these methods are
summarized below.
2.4.1.1. EQUILIBRIUM APPROACH (EULER METHOD).
The concept of stability is frequently explained considering equilibrium. It deals

with the equilibrium configuration of a perfect system. The formulation of the line
continuum equation began with a much idealized model, the Euler Column
(Iyengar, 1988), which sums all the forces acting on the continuum to zero,
referring the summation as the governing equation.
P x P
y
L
Figure 2.1. An idealized Euler Column. (Iyengar, 1988)
The equilibrium equation was derived based on the following assumptions.
a. The line continuum is axially loaded.

10
b. The material is linearly elastic.
c. The deflections are small.
d. The ends are simply supported.
In view of these assumptions, the moment – curvature relation becomes linear and
can be obtained from the second – order differential equation as given in equation
2.1
d2 w
−EI = M(x) 2.1
dx2
Where M is the bending moment, I is the second moment of area and EI is the
flexural rigidity of the beam in the plane of bending that is the x – y plane assumed
to be the plane of symmetry. For a deformed shape to be in equilibrium, the internal
resistive moment and the external moment must be equal. (Iyengar, 1988). External
moment is given as
M = Pw 2.2
d2 w
Hence, EI + Pw = 0 2.3
dx2
𝑑2𝑤
Equation 2.3 can be re-written as + 𝑘2𝑤 = 0 2.4
𝑑𝑥 2
𝑃
Where 𝑘 2 = and P is the axial load.
𝐸𝐼
Equation 2.4 is a linear homogeneous differential equation with constant

coefficients whose general solution can be written in trigonometric function for ease
of analysis.
2.4.1.2 ENERGY APPROACH.
Energy methods are a convenient means of computing unknown force and

displacement in elastic structures (Yoo and Lee, 2011). They can be the basis for
deriving governing differential equations and the required boundary conditions of
the problem (Yoo and Lee, 2011).
11
When exact solution of a governing differential equation is not possible or is
difficult to obtain, the energy approach can be applied to get a good approximate
solution due to its inherent ability to converge to the exact solution if a large
number of term is taken to represent the deflection shape (Iyengar, 1988). This
approach is based on an extremum principle in mechanics which employ an energy
criterion, indicating the sums of all the work (strain energy and external work) on
the continuum must be equal to the total loads energy of the system at a specified
point along the continuum
2.4.1.3. PRINCIPLE OF CONSERVATION OF ENERGY.
Equilibrium is established by satisfying the law of conservation of energy which

states that “A conservative system is in equilibrium if the strain energy stored is
equal to the work performed by the external load” (Chajes, 1974). Thus, the energy
criterion for determining the critical load is therefore given as
ΔW = ΔU 2.5
Where ΔW is the change in work and ΔU is the changes in energy that occurs
during bending at constant axial load.
Change in external work, ΔW is given as
𝑃 𝐿
ΔW = ∫0 (𝑦 𝐼 )2 𝑑𝑥 2.6
2
𝑑𝑥
Where P is the axial load and 𝑦 𝐼 =
𝑑𝑦
Corresponding increase in strain energy due to bending is given as
𝐸𝐼 𝐿
ΔU =
2
∫0 (𝑦 𝐼𝐼 )2 𝑑𝑥 2.7
Trigonometric or polynomial functions are frequently used because of ease of

𝜋𝑥
analysis. Assuming a suitable function for deflection, y = A sin
𝑙
12
𝐴2 𝐸𝐼𝜋4 𝐿 𝜋𝑥
one obtains ΔU =
2𝑙 4
∫0 𝑠𝑖𝑛2 𝑙
𝑑𝑥 2.8
𝐴2 𝑃𝜋2 𝐿 𝜋𝑥
and ΔW =
2𝑙 2
∫0 𝑐𝑜𝑠 2 𝑙
𝑑𝑥 2.9
Solving the integrals, we obtain
𝐴2 𝐸𝐼𝜋4
ΔU = 2.10
4𝐿3
𝐴2 𝑃𝜋2
ΔW = 2.11
4𝐿
Using principle of conservation of energy requires that ΔU = ΔW from which
𝜋2 𝐸𝐼
Pcr = 2.12
𝐿2
where Pcr is the critical load.
2.4.1.4. CALCULUS OF VARIATIONS.
The rigorous application of the stationary energy theorem to continuous systems

requires the use of Calculus of Variations. Calculus of variation is a generalization
of all maximum and minimum problems of ordinary calculus. It seeks to determine
the function y=y(x) that maximizes / minimizes a definite integral called a
functional and whose integrand contains y and its derivatives and the independent
variable x (Yoo and Lee, 2011; Chajes, 1974). In structural mechanics, these
amounts to finding the deformed shape of a system that will cause the total potential
energy of the system to have a stationary value. The deformation that satisfies this
criterion corresponds to the equilibrium state of the system.
13
2.4.1.5. PRINCIPLE OF MINIMUM POTENTIAL ENERGY.
Applying the Calculus of Variations, the strain energy, U for a line continuum due
to flexural action is given as
𝐿 𝐸𝐼
U = ∫0 (𝑦 𝐼𝐼 )2 𝑑𝑥 2.13
2
𝑑2𝑦
Where 𝑦 𝐼𝐼 = 2.14
𝑑𝑥 2
Also, the potential energy, V of external loads consists of the negative product of
axial load and the vertical distance that the load moves as the member bends.
Hence,
𝐿𝑃
V= –∫0 (𝑦 𝐼 )2 𝑑𝑥 2.15
2
𝑑𝑦
where 𝑦𝐼 =
𝑑𝑥
Therefore, Total Potential Energy functional, П of the system is a combination of

the strain energy, U and potential energy of external loads, V. (Iyengar, 1988; Yoo
and Lee, 2011). That is
𝐿 𝐸𝐼 𝑃
Пp = U + V = ∫0 [ (𝑦 𝐼𝐼 )2 − (𝑦 𝐼 )2 ] 𝑑𝑥 2.16
2 2
Equating equation 2.16 to zero gives the mathematical statement of the principle of
minimum potential energy which states that “An elastic structure is in equilibrium if
no change occurs in the total potential energy (stationary value) of the system when
its displacement is changed by a small arbitrary amount (Yoo and Lee, 2011).
This shows that δПp = 0
That is
𝐿 𝐸𝐼 𝑃
δ(U+V) = ∫0 [ (𝑦 𝐼𝐼 )2 − (𝑦 𝐼 )2 ] 𝑑𝑥 = 0 2.17
2 2
14
The critical load can be obtained by assuming a suitable admissible function for y
and integrating it over the entire domain (Iyengar, 1988).
Equation 2.17 was also explained by Timoshenko and Gere (1963), known as the
Timoshenko method, indicating that a straight configuration (or line continuum) is
stable if Пp> 0 and unstable if Пp< 0. A critical condition is reached when Пp = 0,
which leads to
1 𝐿 𝑃𝑐𝑟 𝐿
∫ 𝐸𝐼(𝑦 𝐼𝐼 )2 𝑑𝑥
2 0
=
2
∫0 (𝑦 𝐼 )2 𝑑𝑥 2.18
where Pcr is the critical load.
Other energy methods include Principle of Virtual work, Castigliano’s theorem

(Part I and II), Principle of Minimum Complimentary Energy etc. They give a value
higher than the exact. By choosing a proper shape function when using this
approach, the difference is minimized.
2.4.1.6. RAYLEIGH- RITZ TECHNIQUE.
This approximate technique was first employed by Lord Rayleigh in 1945 in studies
of vibrations and by Timoshenko (1963) in buckling problems. The method was
refined and extended by Ritz and has hence been applied to mechanics, stability and
vibration of complex systems.
Rayleigh – Ritz method is very useful for approximate computation of the solution
in a situation where exact solutions are difficult and complicated or nearly
impossible to obtain (Yoo and Lee, 2011). The technique is based on the vibrational
formulation of the boundary –value and is considered as a derivative of the
principle of the stationary quasistatic loads (Simitses and Hodges, 2006).
In this method, an assumed displacement function satisfying the geometric

boundary conditions of the system is used in the expression for the total potential
energy function,Пp (Chen and Lui, 1987). By using such function, a structural
system with an infinite degree of freedom is now reduced to a system of finite
15
degrees of freedom. Due to this simplification, the total potential energy functional
is reduced to a function to directly obtain solutions by use of ordinary calculus
instead of calculus of variations (Chen and Lui, 1987).
Assuming the buckled shape of member has the approximate form
y = a 2.19
where y is the assumed lateral displacement of the line continuum; ϕ is the arbitrary
function satisfying the boundary condition of the member; a is the undetermined
coefficient for the static problem (Chen and Lui, 1987).
Using the principle of minimum total potential energy, the equilibrium

configuration of the member is identified if
δ (U+V) = 0 2.20
Since U and V are functions of a, equation 2.17 can be rewritten as
𝜕(𝑈 + 𝑉)
𝜕𝑎 = 0 2.21
𝜕𝑎
Because ∂a is arbitrary, we have,
𝜕(𝑈 + 𝑉) 𝜕П
=0 ⇒ =0 2.22
𝜕𝑎 𝜕𝑎
𝜕П
= 0; where i = 1, 2, 3, … … … … n 2.23
𝜕𝑎𝑖
Equation 2.20 represent a system of n simultaneous homogeneous equations with

the a’s and the load P as unknowns. For non– trivial solution of a’s, the determinant
of the coefficient matrix of the system of equations (which contains P as a variable)
must varnish. The lowest value of P that renders the determinant of the coefficient
matrix zero is the critical load of the member (Chen and Lui, 1987).
16
2.4.1.7. GALERKIN’S TECHNIQUE.
This technique, developed by Galerkin, uses approximate deflected shapes to obtain

approximate critical loads. This method deals directly with the differential equation
(Chen and Lui, 1987).
It has been shown that the total potential energy of a hinged - hinged column will
assume a stationary value if the following condition is satisfied
𝐿
(EI yII)(δyI)│x=l – (EI yII) (δyI) │x=0+∫0 (𝐸𝐼𝑦 𝐼𝑉 + 𝑃𝑦 𝐼𝐼 )𝑑𝑥 𝛿𝑦 = 0 2.2 4
Assuming the deflection function has the polynomial form
y = a11 + a22 + a33+……………. +ann = ∑𝑛𝑖=1 𝑎𝑖 𝜙𝑖 2.25
where i are independent continuous functions satisfying both boundary conditions

of the member; ai are undetermined coefficients (Chen and Lui, 1987).
Substituting the assumed deflection polynomial into equation 2.21, we obtain
𝐿
∫ (𝐸𝐼𝑦 𝐼𝑉 + 𝑃𝑦 𝐼𝐼 )𝑑𝑥 𝛿𝑦 = 0 2.26
0
Which can be further simplified if we denote an operator Q as
d4 d2
Q = EI 4
+ P 2.27
dx dx2
recognizing that δy = δ(∑𝑛𝑖=1 𝑎𝑖 𝜙𝑖 ) = ∑𝑛𝑖=1 𝜙𝑖 𝛿𝑎𝑖 2.28
Thus, using equation 2.23, 2.24 and 2.25, it can be written as
𝐿 𝑛
∫ 𝑄(𝑦) ∑ 𝜙𝑖 𝛿𝑎𝑖 𝑑𝑥 = 0 2.29

0 𝑖=1
Since δai is arbitrary, we must have
17
𝐿
∫ 𝑄(𝑦)𝜙𝑖 𝑑𝑥 = 0 2.30
0
Equation 2.27 is called Galerkin’s Integral.
2.4.1.8. NUMERICAL TECHNIQUES.
The differential equation approach is useful when it’s possible to get a closed –
form solution. In Rayleigh – Ritz and Galerkin’s method, where the approximate
function is quite involved in large number of terms, it is difficult to obtain a closed
– form solution because the resulting system of algebraic equations are so
complicated. Hence, numerical techniques find a wider application (Iyengar, 1988;
Chen and Liu, 1987; Yoo and Lee, 2011).
2.4.1.9. FINITE DIFFERENCE METHOD (FDM).
This is a numerical technique for obtaining approximate solutions to differential

equations (Salvadori, 1961). It refers to the process of replacing the partial
derivatives by finite difference quotient and then obtains solution of resulting
system of algebraic equations (Yoo and Lee, 2011). The basis of this technique is
that a derivative of a function at a point can be approximated by an algebraic
expression consisting of the value of the function at that point and at several nearby
points known as nodes. These nodes make the finite difference mesh.
Thus, finite difference method applies a mathematical discretization of the

continuum yielding a set of algebraic equations, from which the deflection at the
mesh points can be obtained (Szilard, 2004). Depending on the deformation of the
derivatives at a reference point in terms of the differences of the neighboring points,
a variety of practical approximation for the derivatives can be obtained. Notably
among them are: a). Backward differences b). Forward differences c.). Central
differences.
The first forward difference is given by the first order expansion in the form,
18
𝑓𝑖+1 − 𝑓𝑖
𝑓 𝐼 (𝑥) = 2.31
𝛥𝑥
𝑓𝑖 − 𝑓𝑖+1
and the first backward difference as, 𝑓 𝐼 (𝑥) = 2.32
𝛥𝑥
𝑓𝑖+1 − 𝑓𝑖−1
Hence, central difference is given as, 𝑓𝐼 = 2.33
2𝛥𝑥
2.4.1.10. IMPROVED FINITE DIFFERENCE METHOD. (IFDM)
While the use of an ordinary finite difference technique is very simple and is quite
general, it is characterized by slow convergence. More so, when higher – order
derivatives and a large number of mesh points are involved, the solution may
converge to a wrong number (Szilard, 2004). Consequently, improved finite
difference method is applied when a high accuracy in finite difference solution of
the problem is required.
The general tool for derivation of improved derivatives is the Polynomial expansion
series. To obtain improved finite difference expansions, additional mesh points are
considered in the vicinity of the pivotal point (Szilard, 2004).
Expanding a function f(x) at points (x+1) and (x-1) with xi as the pivotal point
yields
𝛥𝑥 𝐼 (𝛥𝑥)2 𝐼𝐼 (𝛥𝑥)3 𝐼𝐼𝐼 (𝛥𝑥)4 𝐼𝑉

𝑓(𝑥+1) = 𝑓(𝑥𝑖 ) + 𝑓 (𝑥𝑖 ) + 𝑓 (𝑥𝑖 ) + 𝑓 (𝑥𝑖 ) + 𝑓 (𝑥𝑖 )
1! 2! 3! 4!
(𝛥𝑥)𝑛 𝑛
+ ⋯+ 𝑓 (𝑥𝑖 ) 2.34
𝑛!
𝛥𝑥 𝐼 (𝛥𝑥)2 𝐼𝐼 (𝛥𝑥)3 𝐼𝐼𝐼 (𝛥𝑥)4 𝐼𝑉

𝑓(𝑥−1) = 𝑓(𝑥𝑖 ) − 𝑓 (𝑥𝑖 ) + 𝑓 (𝑥𝑖 ) − 𝑓 (𝑥𝑖 ) + 𝑓 (𝑥𝑖 )
1! 2! 3! 4!
(−𝛥𝑥)𝑛 𝑛
+ ⋯+ 𝑓 (𝑥𝑖 ) 2.35
𝑛!
Adding the above equations, we obtain a second – order derivative. Improved finite
quantities for the other derivatives can be derived in a similar manner.
19
Other numerical methods include Finite Element Method (FEM), Newmark’s
method, Boundary Element Method (BEM) etc.
2.4.2. APPROACHES TO VIBRATION (DYNAMIC) ANALYSIS.
Vibration analysis is one vital task in designing of structural system. The effect of
vibration is important information as it helps to design a system to control excessive
amplitude of the vibration (Chopade and Barjibhe, 2013).
For a dynamic approach, D’Alembert principle is invoked which is based on the

notion of a fictitious inertia force, a force equal to the product of mass times its
acceleration and acting in a direction opposite to the acceleration. It states that
“with inertia forces included, a system is in equilibrium at each time instant”
(Chopra, 1995).
2.4.2.1. THEORETICAL ANALYSIS OF TRANSVERSE VIBRATION.
Considering a line continuum with modulus of rigidity EI, density ρ, and cross
sectional area A. E is the modulus of elasticity of the material and I is the moment
of inertia about the axis of bending.
Let P be the external normal force per unit length and Q and M be the shear force
and bending moment a distance x from the origin, O.
From Euler – Bernoulli equation, the relationship between bending moment,

deflection and shear force can be expressed as
𝑑2𝑦 𝑑𝑀
𝑀 = −𝐸𝐼 and Q= 2.36
𝑑𝑥 2 𝑑𝑥
Using equation 2.36 and considering equation of motion, we can obtain
𝑑2𝑦 𝑑2𝑀
𝜌𝐴 2 𝑑𝑥 = 𝑑𝑥 + 𝑃𝑑𝑥 2.37
𝑑𝑡 𝑑𝑥 2
20
𝑑4𝑦 𝑑2𝑦
That is 𝐸𝐼 4
+ 𝜌𝐴 =𝑃 2.38
𝑑𝑥 𝑑𝑡 2
If there is no external load, then P is zero. Hence, the equation of motion for free
vibration of a line continuum can be written as:
𝑑4𝑦 2
𝑑2𝑦
+𝑐 =0 2.39
𝑑𝑥 4 𝑑𝑥 2
𝐸𝐼
Where c2 = ; ρ is the density of the material; A is the cross sectional area; E is
𝜌𝐴
the modulus of elasticity of material; I is the moment of inertia about the axis of
bending (Chakraverty, 2007).
The differential equation (Equation 2.39) is solved by the separation of variables as

one depends on position and another on time (Chakraverty, 2007) and this is given
as:
y = W(x) T (t) 2.40
Substituting equation 2.40 into equation 2.39 and simplifying it, we obtain
𝑐 2 𝑑4𝑊 1 𝑑 2 𝑇(𝑡)
= − = −𝜔2 2.41
𝑊(𝑥) 𝑑𝑥 4 𝑇(𝑡) 𝑑𝑡 2
Equation 2.41 can be written as two separate differential equation
𝑑4𝑤
− 𝛽 4 𝑊(𝑥) = 0 2.42𝑎
𝑑𝑥 4
𝑑2𝑇
2
+ 𝜔2 𝑇 = 0 2.42𝑏
𝑑𝑡
𝜔2 𝜌𝐴𝜔2
where β4 = = 2.43
𝑐2 𝐸𝐼
Equation 2.42a is the differential equation of free vibration of a line continuum. The
solution to differential equation 2.42a is obtained using a suitable trigonometric or
21
polynomial function, upon which solved, satisfying all boundary conditions of the
line continuum, gives a corresponding infinite number of natural frequencies as
given in equation 2.41
𝐸𝐼
𝜔𝑖 = 𝛽𝑖2 √ 2.44
𝜌𝐴𝐿4
where βi = value of roots, i = 1, 2, 3, ….., n obtained from solution of the

trigonometric/polynomial function (Chakraverty, 2007).
2.4.2.2. APPROXIMATE METHODS. RAYLEIGH’S METHOD
This method requires expressions for maximum kinetic and potential energies of a
system. The maximum kinetic and potential energies must be equal since no energy
is lost and no energy is fed into the system over one cycle of vibration. This gives
the Rayleigh quotient (Chakraverty, 2009).
If P and m denote the tensile force and mass per unit length of a line continuum,
then the potential energy for vibration is given as
𝑃 𝑙 𝑑𝑦 2
𝑉 = ∫ ( ) 𝑑𝑥 2.45
2 0 𝑑𝑥
and corresponding kinetic energy, T given as
𝑚 𝑙 𝑑𝑦 2
𝑇= ∫ ( ) 𝑑𝑥 2.46
2 0 𝑑𝑥
If the maximum value of y(x,t) is W(x), then the maximum value of the potential
energy is written as
𝑃 𝑙 𝑑𝑊(𝑥) 2
𝑉𝑚𝑎𝑥 = ∫ ( ) 𝑑𝑥 2.47
2 0 𝑑𝑥
22
The maximum kinetic energy can be obtained by assuming a harmonic function of
the form y(x, t) = W(x) cos ωt in equation 2.46 and we obtain
𝜔2 𝑚 𝑙
𝑇𝑚𝑎𝑥 = ∫ (𝑊(𝑥))2 𝑑𝑥 2.48
2 0
Equating maximum kinetic and potential energies of the system, we obtain the
Rayleigh quotient as
𝜔2
𝑃 𝑙 𝑑(𝑊(𝑥)) 2
𝑑𝑥
2 ∫0 ( 𝑑𝑥 )
= 𝑚 𝑙 2.49
(𝑊(𝑥))2 𝑑𝑥
2 ∫0
The deflection function can be identified using a polynomial function and hence the
natural frequency, ω of the system can be computed (Chakraverty, 2009).
2.4.2.3. RAYLEIGH – RITZ METHOD.
In Rayleigh’s method, approximate value of the lowest frequency is obtained using

a single function approximation. In Rayleigh – Ritz method, a linear combination of
assumed function is considered, satisfying some boundary with some boundary
condition to obtain a closer approximation to the exact values of the natural modes
of vibrations. However, approximate values of the frequencies depend on the
assumed functions (Chakraverty, 2009).
Assuming linear functions are selected for approximating a deflection function

W(x) for vibration, and then the deflection expression is written as
𝑊(𝑥) = ∑ 𝑐𝑖 𝑤𝑖 = 𝑐1 𝑤1 + 𝑐2 𝑤2 + 𝑐3 𝑤3 + ⋯ + 𝑐𝑛 𝑤𝑛 2.50
𝑖=1
where c1, c2, c3, …, cn are the constants and w1, w2, w3, …, wn are the functions of
the spatial coordinate that satisfy the boundary condition of the problem. Putting
23
equation 2.50 in equation 2.49 (Rayleigh’s quotient) and for stationary of the
natural frequencies, we equate the first partial derivatives with respect to each of the
constants ci to zero, that is
𝜕𝜔2
= 0, 𝑖 = 1, 2, … . , 𝑛 2.51
𝜕𝑐𝑖
Equations 2.48 represents a asset of n algebraic equations in n unknowns c1, c2,…,cn

which can be solved for n natural frequencies and mode shapes (Chakraverty,
2009).
2.4.2.4. NUMERICAL METHOD. FINITE DIFFERRNCE METHOD.
Assuming the line continuum oscillates about its equilibrium position, then the
lateral deflection can be expressed as
w(x,t) = W(x) sin ωt 2.52
where W(x) is the shape function that satisfies the boundary conditions and
represents the shape of the deflection of the member. The time dependent part of the
harmonic oscillation is represented by sin ωt, where ω is the natural frequency of
the line continuum. Rewriting equations 2.42a, we have
𝑑 4 𝑤 𝜌𝐴𝜔2
+ =0 2.53
𝑑𝑥 4 𝐸𝐼
Replacing the derivatives by the Finite Difference expressions as described in

Section 2.4.1.9 and Section 2.4.1.10 into equation 2.53 and treating the boundary
value problem in the same manner, an algebraic equation are obtained for each
finite difference mesh points. These algebraic equations can be written in more
general terms as
𝑎11 𝑥1 + 𝑎12 𝑥2 + 𝑎13 𝑥3 + 𝑎14 𝑥4 + ⋯ + 𝑎1𝑛 𝑥𝑛 = 𝜆𝑥1 𝑏1
𝑎21 𝑥1 + 𝑎22 𝑥2 + 𝑎23 𝑥3 + 𝑎24 𝑥4 + ⋯ + 𝑎2𝑛 𝑥𝑛 = 𝜆𝑥2 𝑏2 2.54
24
Or using matrix notation AX = λBx 2.55
Therefore, (A – λB) x = 0 2.56
Where x = {xi} is a column matrix whose element xi represent the amplitudes of the
free vibrations; A = [aij] is a square matrix obtained from the finite difference
expression; B = [bij] is a diagonal matrix representing the constant in the right hand
term and λ = ω2 is the natural frequency.
Pre-multiplying equation 2.55 by B-1, we obtain
(C – λI) x = 0 2.57
Where C = B-1A and I is an identity matrix.
Equation 2.54 is the general Eigen value equation for free vibration analysis. For
non – trivial solution of equation 2 .54, the determinant of the coefficients must be
zero. The expansion of this determinant yields a polynomial equation of nth – order
equation in λ which is called characteristic equation. The roots (λ 1, λ2…λn) are the
Eigen values of the vibrating system, from which the natural frequencies
𝜔𝑖 = √𝜆 fori = 1, 2, 3, … . , n 2.58
can be calculated (Szilard, 2004).
25
CHAPTER THREE
METHODOLOGY
3.1. FINITE DIFFERENCE FORMULATION.

For the formulation of the new derivatives, the polynomial expansion series is
applied once the grid parts are evenly spaced. Figure 3.1 schematically indicates the
grid parts for central differences where y3 is the pivotal point and h is the grid
spacing / segment length.
6
y5
y4
y3
y2
y0 y1
0 h 1 h 2 h 3 h 4 h 5 h 6
Figure 3.1: Computational grid for finite difference approximation.
Taking y3 as the pivotal point and expanding the function y3 using the series as
given in Equations 2.31 and 2.32, truncating at the fourth – order derivative yIV
gives;
ℎ2 ℎ3 ℎ4
𝑦4 = 𝑦3 + ℎ𝑦3 + ( ) 𝑦3 + ( ) 𝑦3 + ( ) 𝑦3𝐼𝑉
𝐼 𝐼𝐼 𝐼𝐼𝐼
3.1
2 6 24
ℎ2 ℎ3 ℎ4
𝑦2 = 𝑦3 − + ( ) 𝑦3 − ( ) 𝑦3 + ( ) 𝑦3𝐼𝑉
ℎ𝑦3𝐼 𝐼𝐼 𝐼𝐼𝐼
3.2
2 6 24
Subtracting Equation 3.2 from Equation 3.1 gives:
ℎ3
𝑦4 − 𝑦2 = 2ℎ𝑦3 + ( ) 𝑦3𝐼𝐼𝐼 .
𝐼
That is
3
26
𝑦4 𝑦2 ℎ2
𝑦3𝐼 = − − ( ) 𝑦3𝐼𝐼𝐼 3.3
2ℎ 2ℎ 6
Adding Equations 3.1 and 3.2 gives:
2 𝐼𝐼
ℎ4 𝐼𝑉
𝑦4 + 𝑦2 = 2𝑦3 + ℎ 𝑦3 + 𝑦 . That is
12 3
1 ℎ2
𝑦3 = 2 (𝑦2 − 2𝑦3 + 𝑦4 ) − ( ) 𝑦3𝐼𝑉
𝐼𝐼
3.4
ℎ 12
Taking differences between nodes y3 and y5 and y3 and y1, Equations 3.1 and 3.2
shall give:
4ℎ2 8ℎ3 16ℎ4
𝑦5 = 𝑦3 + 2ℎ𝑦3𝐼
+( 𝐼𝐼
) 𝑦3 + ( 𝐼𝐼𝐼
) 𝑦3 + ( ) 𝑦3𝐼𝑉 3.5
2 6 24
2 3
4ℎ 8ℎ 16ℎ4
𝐼
𝑦1 = 𝑦3 – 2ℎ𝑦3 + ( 𝐼𝐼
) 𝑦3 − ( 𝐼𝐼𝐼
) 𝑦3 + ( ) 𝑦3𝐼𝑉 3.6
2 6 24
𝐼
8ℎ3
𝑦5 − 𝑦1 = 4ℎ𝑦3 + ( ) 𝑦3𝐼𝐼𝐼 3.7
3
16ℎ4
2 𝐼𝐼
𝑦3 + 𝑦1 = 2𝑦3 + 4ℎ 𝑦3 + ( ) 𝑦3𝐼𝑉 3.8
12
From Equation 3.7, we have:
𝐼
1 2ℎ2
𝑦3 = (𝑦 )
5 + 𝑦1 − ( ) 𝑦3𝐼𝐼𝐼 3.9
4ℎ 3
From Equation 3.8, we have:
1 4ℎ2
𝐼𝐼
𝑦3 = 2
(𝑦5 + 𝑦1 − 2𝑦3 ) − ( ) 𝑦3𝐼𝑉 3.10
4ℎ 12
𝑦4 𝑦5 𝑦2 𝑦1 ℎ2 2ℎ2
0= − − + − ( ) 𝑦3𝐼𝐼𝐼 + ( ) 𝑦3𝐼𝐼𝐼 . That is;
2ℎ 4ℎ 2ℎ 4ℎ 6 3
𝑦4 𝑦5 𝑦2 𝑦1
0= 3 − 3 − 3 + 3 − ℎ2 𝑦3𝐼𝐼𝐼 + 4ℎ2 𝑦3𝐼𝐼𝐼 . This implies that
ℎ 2ℎ ℎ 2ℎ
−3𝑦1 3𝑦2 3𝑦4 3𝑦5
3ℎ2 𝑦3𝐼𝐼𝐼 = + − +
2ℎ ℎ ℎ 2ℎ
1
∴ 𝑦3𝐼𝐼𝐼 = (−𝑦1 + 2𝑦2 − 2𝑦4 + 𝑦5 ) 3.11
2ℎ3
27
𝑦2 2𝑦3 𝑦4 ℎ2 𝑦3 𝐼𝑉 𝑦5 2𝑦3 4ℎ2
0= − + − − + + ( ) 𝑦3𝐼𝑉 . That is;
ℎ2 ℎ2 ℎ 2 12 4ℎ 2 4ℎ 2 12
0 = 12y2 − 24y3 + 12y4 − h4 y3IV − 3y5 − 3y1 + 6y3 + 4h4 y3IV . This
implies;
3ℎ4 𝑦3𝐼𝑉 = 3𝑦1 − 12𝑦2 + 18𝑦3 − 12𝑦4 + 3𝑦5
1
∴ 𝑦3𝐼𝑉 = (𝑦 − 4𝑦2 + 6𝑦3 − 4𝑦4 + 𝑦5 ) 3.12
ℎ4 1
Substituting Equation 3.11 into Equation 3.3 gives:
𝑦4 𝑦2 ℎ2 1
𝑦3𝐼 = − − (2ℎ3 (−𝑦1 + 2𝑦2 − 2𝑦4 + 𝑦5 )). That is;
2ℎ 2ℎ 6
𝑦4 𝑦2 𝑦1 𝑦2 𝑦4 𝑦5
𝑦3𝐼 = − + − + −
2ℎ 2ℎ 12ℎ 6ℎ 6ℎ 12ℎ
𝟏
∴ 𝒚𝑰𝟑 = (𝒚 − 𝟖𝒚𝟐 + 𝟖𝒚𝟒 − 𝒚𝟓 ) 3.13
𝟏𝟐𝒉 𝟏
Equation 3.13 shows the first derivative of the new finite difference method.
Substituting Equation 3.12 into Equation 3.4 gives:
ℎ2
(1⁄ 4 ( 𝑦1 − 4𝑦2 + 6𝑦3 − 4𝑦4 + 𝑦5 ) ). That is;
𝑦2 2𝑦3 𝑦4
𝑦3𝐼𝐼 = 2
− + −
ℎ ℎ2 ℎ 2 12 ℎ
𝑦2 2𝑦3 𝑦4 𝑦1 𝑦2 𝑦3 𝑦4 𝑦5
𝑦3𝐼𝐼 = − + − + − + −
ℎ2 ℎ2 ℎ2 12ℎ2 3ℎ2 2ℎ2 3ℎ2 12ℎ2
𝟏
∴ 𝒚𝑰𝑰
𝟑 = (−𝒚𝟏 + 𝟏𝟔𝒚𝟐 − 𝟑𝟎𝒚𝟑 + 𝟏𝟔𝒚𝟒 − 𝒚𝟓 ) 3.14
𝟏𝟐𝒉𝟐
Equation 3.14 shows the second derivative of the new finite difference method.
From Equation 3.1, we can obtain:
ℎ2 𝐼𝐼𝐼 ℎ3 𝐼𝑉 ℎ4 𝑉
𝑦4𝐼= +𝑦3𝐼 + ℎ𝑦3𝐼𝐼
𝑦 + 𝑦 + 𝑦 3.15
2 3 6 3 24 3
ℎ2 𝐼𝐼𝐼 ℎ3 𝐼𝑉 ℎ4 𝐼𝑉
𝑦2𝐼
= 𝑦3𝐼
− ℎ𝑦3𝐼𝐼
+ 𝑦3 − 𝑦 + 𝑦 3.16
2 2 3 24 3
𝑦 𝐼𝐼 𝐼𝑉
𝑦41
− = 𝑦21 + 2ℎ𝑦3𝐼𝐼
𝑦
3 3
𝐼𝐼
1 𝐼 𝐼
ℎ2 𝐼𝑉
∴ 𝑦3 = (𝑦 − 𝑦2 ) − 𝑦 3.17
2ℎ 4 6 3
28
Adding Equation 3.15 and Equation 3.16 gives:
ℎ4 𝑉
𝑦4𝐼 + 𝑦2𝐼 = 2𝑦3𝐼 + ℎ2 𝑦3𝐼𝐼𝐼 + 𝑦
12 3
1 ℎ2 𝑉
∴ 𝑦3𝐼𝐼𝐼 𝐼 𝐼 𝐼
= 2 (𝑦2 − 2𝑦3 + 𝑦4 ) − 𝑦 3.18
ℎ 12 3
From Equation 3.5 and Equation 3.6, we have:
4ℎ2 𝐼𝐼𝐼 8ℎ3 𝐼𝑉 16ℎ4 𝑉
𝑦5𝐼 = 𝑦3𝐼 + 2ℎ𝑦3𝐼𝐼 + 𝑦 + 𝑦 + 𝑦 3.19
2 3 6 3 24 3
4ℎ2 𝐼𝐼𝐼 8ℎ3 𝐼𝑉 16ℎ4 𝑉
𝑦1𝐼 = −𝑦3𝐼 + 2ℎ𝑦3𝐼𝐼
𝑦 − 𝑦 + 𝑦 3.20
2 3 6 3 24 3
𝐼 𝐼 𝐼𝐼
8ℎ3 𝐼𝑉
𝑦5 − 𝑦1 = 4ℎ𝑦3 + 𝑦 3.21
3 3
16ℎ4 𝑉
𝑦5𝐼 + 𝑦1𝐼 = 2𝑦3𝐼 + 4ℎ𝑦 2 𝑦3𝐼𝐼𝐼 + 𝑦 3.22
12 3
1 𝐼
2ℎ2 𝐼𝑉
𝑦3𝐼𝐼
= 𝐼
(𝑦 − 𝑦1 ) − 𝑦 3.23
4ℎ 5 3 3
1 𝐼
4ℎ2 𝑉
𝑦3𝐼𝐼𝐼
= 𝐼 𝐼
(𝑦 − 2𝑦3 + 𝑦5 ) − 𝑦 3.24
4ℎ2 1 12 3
𝑦4 𝐼 𝑦2𝐼 ℎ2 𝑦5 𝐼 𝑦1𝐼 2ℎ2
0 = − − 𝑦3𝐼𝑉 − + + 𝑦3𝐼𝑉 . That is;
2ℎ 2ℎ 6 4ℎ 4ℎ 3
6𝑦4𝐼 6𝑦2𝐼 3𝑦5𝐼 3𝑦1𝐼

0= − − 2𝑦3𝐼𝑉 − + + 8𝑦3𝐼𝑉 . This implies that
ℎ3 ℎ3 ℎ3 ℎ3
1
6𝑦3𝐼𝑉 = (−3𝑦1𝐼 + 6𝑦2𝐼 − 6𝑦4𝐼 + 3𝑦5𝐼 )
ℎ3
1
∴ 𝑦3𝐼𝑉 = (−𝑦1𝐼 + 2𝑦2𝐼 − 2𝑦4𝐼 + 𝑦5𝐼 ) 3.25
2ℎ3
𝑦2𝐼 2𝑦3𝐼 𝑦4𝐼 ℎ4 𝑦1𝐼 2𝑦3𝐼 𝑦5𝐼 4ℎ2
0= 2
− + − 𝑦3𝑉 − + − + 𝑦3𝑉 . That is
ℎ ℎ2 ℎ 2 12 4ℎ 2 4ℎ 2 4ℎ 2 12
29
𝑦1𝐼 4𝑦2𝐼 6𝑦5𝐼 4𝑦4𝐼 𝑦5𝐼
ℎ2 𝑦3𝑉 = 2− 2 + 2 − 2 + 2
ℎ ℎ ℎ ℎ ℎ
1
∴ 𝑦3𝐼 = 4 (𝑦1𝐼 − 4𝑦2𝐼 + 6𝑦3𝐼 − 4𝑦4𝐼 + 𝑦5𝐼 ) 3.26
ℎ
Substituting Equation 3.25 into Equation 3.17, we have:
1 ℎ2 1
𝑦3𝐼𝐼 =
2ℎ
(𝑦4𝐼 − 𝑦2𝐼 ) −
6
( 2ℎ3 (− 𝑦1𝐼 + 2𝑦2𝐼 − 2𝑦4𝐼 + 𝑦5𝐼 )). This is
𝑦4𝐼 𝑦2𝐼 1 𝐼 1 𝐼 1 𝐼 1 𝐼
= − + 𝑦1 − 𝑦2 + 𝑦4 − 𝑦
2ℎ 2ℎ 12ℎ 6ℎ 6ℎ 12ℎ 5
1
∴ 𝑦3𝐼𝐼 = (𝑦 𝐼 − 8𝑦1𝐼 + 8𝑦4𝐼 − 𝑦5𝐼 ) 3.27
12ℎ 1
Substituting Equation 3.26 into Equation 3.18, we have:
1 ℎ2 1
𝑦3𝐼𝐼𝐼 =
ℎ 2
(𝑦2𝐼 − 2𝑦3𝐼 + 𝑦4𝐼 ) − ( (𝑦1𝐼 − 4𝑦2𝐼 + 6𝑦3𝐼 − 4𝑦4𝐼 + 𝑦5𝐼 )). This is
12 ℎ4
1 1
= (𝑦 𝐼
− 2𝑦 𝐼
+ 𝑦4
𝐼)
− ( 𝑦1𝐼 − 4𝑦2𝐼 + 6𝑦3𝐼 − 4𝑦4𝐼 + 𝑦5𝐼 )
ℎ2 2 3
12ℎ2
1
∴ 𝑦3𝐼𝐼𝐼 = 2
(𝑦1𝐼 − 16𝑦2𝐼 − 30𝑦3𝐼 + 16𝑦4𝐼 − 𝑦5𝐼 ) 3.28
12ℎ
Ignoring higher order terms in Equation 3.3 gives:
1
𝑦3𝐼 = (𝑦 − 𝑦2 ) 3.29
2ℎ 4
Similarly;
1
𝑦2𝐼 = (𝑦 − 𝑦1 ) 3.30
2ℎ 3
1
𝑦1𝐼 = (𝑦 − 𝑦2 ) 3.31
2ℎ 2
1
𝑦4𝐼 = (𝑦 − 𝑦3 ) 3.32
2ℎ 5
1
𝑦5𝐼 = (𝑦 − 𝑦4 ) 3.33
2ℎ 6
Substituting Equations 3.29 to 3.33 into Equation 3.28 gives:
1
𝑦3𝐼𝐼𝐼 = (– 𝑦2 + 𝑦0 + 16𝑦3 − 16𝑦1 − 30𝑦4 + 30𝑦2 + 16𝑦5 − 16𝑦3 −
24ℎ3
𝑦6 + 𝑦4 ) . This is;
1
= (𝑦0 − 16𝑦1 + 29𝑦2 − 29𝑦4 + 16𝑦5 − 𝑦6 )
24ℎ3
30
𝟏
∴ 𝒚𝑰𝑰𝑰
𝟑 = (𝒚 − 𝟏𝟔𝒚𝟏 + 𝟐𝟗𝒚𝟐 + 𝟎𝒚𝟑 − 𝟐𝟗𝒚𝟒 + 𝟏𝟔𝒚𝟓 − 𝒚𝟔 ) 3.34
𝟐𝟒𝒉𝟑 𝟎
Equation 3.34 shows the third derivative of the new finite difference method.
From Equations 3.15 and 3.16, we can obtain:
ℎ2 𝐼𝑉 ℎ3 𝑉 ℎ4 𝑉𝐼
𝑦4𝐼𝐼 = 𝑦3𝐼𝐼 + ℎ𝑦3𝐼𝐼𝐼 + 𝑦 + 𝑦 + 𝑦 3.35
2 3 6 3 24 3
ℎ2 𝐼𝑉 ℎ3 𝑉 ℎ4 𝑉𝐼
𝑦2𝐼𝐼 = 𝑦3𝐼𝐼 − ℎ𝑦3𝐼𝐼𝐼 + 𝑦 − 𝑦 + 𝑦 3.36
2 3 6 3 24 3
ℎ4
𝑦4𝐼𝐼 + 𝑦2𝐼𝐼 = 2𝑦3𝐼𝐼 + ℎ2 𝑦3𝐼𝑉 + 𝑦 𝑉𝐼 . This implies that
12 3
1 ℎ2 𝑉𝐼
𝑦3𝐼𝑉 (𝑦 𝐼𝐼 𝐼𝐼 𝐼𝐼 )
= 2 2 − 2𝑦3 + 𝑦4 − 𝑦 3.37
ℎ 12 3
From equations 3.19 and 3.20, we can obtain;
4ℎ2 𝐼𝑉 8ℎ3 𝑉 16ℎ2 𝑉𝐼

𝑦5𝐼𝐼 = 𝑦3𝐼𝐼 + 2ℎ𝑦3𝐼𝐼𝐼 + 𝑦 + 𝑦 + 𝑦 3.38
2 3 6 3 24 3
4ℎ2 𝐼𝑉 8ℎ3 𝑉 16ℎ4 𝑉𝐼
𝑦1𝐼𝐼 = 𝑦3𝐼𝐼 − 2ℎ𝑦3𝐼𝐼𝐼 + 𝑦 − 𝑦 + 𝑦 3.39
2 3 6 3 24 3
Adding equations 3.38 and 3.39 gives:
16ℎ4
𝑦5𝐼𝐼 + 𝑦1𝐼𝐼 = 2𝑦3𝐼𝐼 + 4ℎ2 𝑦3𝐼𝑉 + 𝑦3𝑉𝐼 . This implies that:
12
1 4ℎ2 𝑉𝐼
𝑦3𝐼𝑉 = 𝐼𝐼 𝐼𝐼 𝐼𝐼
(𝑦 − 2𝑦3 + 𝑦5 ) − 𝑦 3.40
4ℎ2 1 12 3
Subtracting equation 3.40 from equation 3.37 gives:
𝑦2𝐼𝐼 2𝑦3𝐼𝐼 𝑦4𝐼𝐼 ℎ2 𝑉𝐼 𝑦1𝐼𝐼 2𝑦3𝐼𝐼 𝑦5𝐼𝐼 4ℎ2 𝑉𝐼

− 2 + 2− 𝑦 − + − + 𝑦 =0
ℎ2 ℎ ℎ 12 3 4ℎ2 4ℎ2 4ℎ2 12 3
This implies that:
4𝑦2𝐼𝐼 6𝑦3𝐼𝐼 4𝑦4𝐼𝐼 𝑦1𝐼𝐼 𝑦5𝐼𝐼

𝑦3𝑉𝐼 =− 4 + 4 − 4 + 4 + 4
ℎ ℎ ℎ ℎ ℎ
1
∴ 𝑦3𝑉𝐼 = 4
(𝑦1𝐼𝐼 − 4𝑦2𝐼𝐼 + 6𝑦3𝐼𝐼 − 4𝑦4𝐼𝐼 + 𝑦5𝐼𝐼 ) 3.41
ℎ
31
Substituting equation 3.41 into equation 3.40 gives:
1 4ℎ2 1 𝑦1𝐼𝐼 − 4𝑦2𝐼𝐼 + 6𝑦3𝐼𝐼

𝑦3𝐼𝑉 = 𝐼𝐼 𝐼𝐼 𝐼𝐼
(𝑦 − 2𝑦3 + 𝑦5 ) − ( ( ))
4ℎ2 1 12 ℎ4 − 4𝑦4𝐼𝐼 + 𝑦5𝐼𝐼
This is
𝑦1𝐼𝐼 2𝑦3𝐼𝐼 𝑦5𝐼𝐼 𝑦1𝐼𝐼 4𝑦2𝐼𝐼 6𝑦3𝐼𝐼 4𝑦4𝐼𝐼 𝑦5𝐼𝐼
= − + − + − + −
4ℎ2 4ℎ2 4ℎ2 3ℎ2 3ℎ2 3ℎ2 3ℎ2 3ℎ2
This is
−𝑦1𝐼𝐼 4𝑦2𝐼𝐼 5𝑦3𝐼𝐼 4𝑦4𝐼𝐼 𝑦5𝐼𝐼
= + − + −
12ℎ2 3ℎ2 2ℎ2 3ℎ2 12ℎ2
1
∴ 𝑦3𝐼𝑉 = 2
(−𝑦1𝐼𝐼 + 16𝑦2𝐼𝐼 − 30𝑦3𝐼𝐼 + 16𝑦4𝐼𝐼 − 𝑦5𝐼𝐼 ) 3.42
12ℎ
Ignoring higher order terms in equation 3.4 gives:
1
𝑦3𝐼𝐼 = (𝑦 − 2𝑦3 + 𝑦4 ) 3.43
ℎ2 2
Similarly;
1
𝑦2𝐼𝐼 = (𝑦 − 2𝑦2 + 𝑦3 ) 3.44
ℎ2 1
1
𝑦1𝐼𝐼 = (𝑦 − 2𝑦1 + 𝑦2 ) 3.45
ℎ2 0
1
𝑦4𝐼𝐼 = (𝑦 − 2𝑦4 + 𝑦5 ) 3.46
ℎ2 3
1
𝑦5𝐼𝐼 = (𝑦 − 2𝑦5 + 𝑦6 ) 3.47
ℎ2 4
Substituting equations 3.43 to 3.47 into equation 3.42 gives:
1
𝑦3𝐼𝑉 = (– 𝑦0 + 2𝑦1 − 𝑦2 + 16𝑦1 − 32𝑦2 + 16𝑦3 − 30𝑦2 + 60𝑦3 − 30𝑦4
12ℎ4
+ 16𝑦3 − 32𝑦4 + 16𝑦5 − 𝑦4 + 2𝑦5 − 𝑦6 .
This is
1
= (−𝑦0 + 18𝑦1 − 63𝑦2 + 92𝑦3 − 63𝑦4 + 18𝑦5 − 𝑦6 )
12ℎ4
𝟏
∴ 𝒚𝑰𝑽𝟑 = (−𝒚𝟎 + 𝟏𝟖𝒚𝟏 − 𝟔𝟑𝒚𝟐 + 𝟗𝟐𝒚𝟑 − 𝟔𝟑𝒚𝟒 + 𝟏𝟖𝒚𝟓 − 𝒚𝟔 ) 3.48
𝟏𝟐𝒉𝟒
Equation 3.48 shows the fourth derivative of the new finite difference method.
32
The coefficients of the corresponding function values of the new finite difference
expressions of the first-order, second-order, third-order and fourth-order derivatives
as given in Equations 3.13, 3.14, 3.34 and 3.48 are used to develop the finite
difference patterns and are schematically shown on Table 3.1 below.
Table 3.1: Schematic representation of Fourth – order Finite Difference

Expression.
𝒚𝒏𝟑 COEFFICIENTS
𝟏𝟐𝒉𝟐 𝒚𝑰𝑰 -1 16 -30 16 -1

𝟑
𝟏𝟐𝒉𝒚𝑰𝟑 1 -8 0 8 -1
𝟐𝟒𝒉𝟑 𝒚𝑰𝑰𝑰 -1 16 29 0 -29 16 -1

𝟑
-1 18 -63 92 -63 18 -1
𝟏𝟐𝒉𝟒 𝒚𝑰𝑽
𝟑
Three line continua with different boundary conditions were studied in this work
namely: Pin – Roller supports (P – R), Clamped – Clamped supports (C – C) and
Clamped – Roller supports (C – R). Their boundary conditions are shown in table 2.
Also, the line continua were studied using 3 different cases namely:
Case 1: Dividing the linear continuum into 4 equal lengths using 3 nodes.
3.2. STABILITY ANALYSIS APPROACH.
Governing Differential Equation for buckling of a line continuum (stability) was

given by Iyengar (1988) as
𝑑4𝑤 𝑁 𝑑2𝑤
+ =0 3.49
𝑑𝑥 4 𝐸𝐼 𝑑𝑥 2
33
Table 3.2: Line continua and their boundary conditions.
Linear Continuum Boundary Conditions
𝑑 2 𝑤(0) 𝐼𝐼
𝑤(0) = 0, 2
= 𝑤(0) =0
𝑑𝑥
Pin – Roller Linear
Continuum ( P – R) 𝑑 2 𝑤(𝐿) 𝐼𝐼
𝑤(𝐿) = 0, 2
= 𝑤(𝐿) =0
𝑑𝑥
𝑑𝑤(0) 𝐼
𝑤(0) = 0, = 𝑤(0) = 0;
𝑑𝑥
Clamped – Clamped Linear 𝑑𝑤(𝐿) 𝐼
𝑤(𝐿) = 0, = 𝑤(𝐿) =0
𝑑𝑥
Continuum (C – C)
𝑑𝑤(0) 𝐼
𝑤(0) = 0, = 𝑤(0) = 0;
𝑑𝑥
Clamped – Roller Linear 𝑑 2 𝑤(𝐿) 𝐼𝐼
𝑤(𝐿) = 0, 2
= 𝑤(𝐿) = 0.
𝑑𝑥
Continuum (C – R)
This means that
𝑑4𝑤 2
𝑑2𝑤
+𝑘 =0 3.50
N
where k 2 =
EI
Equation (3.2) can be re-written as
𝑤 𝐼𝑉 + 𝑘 2 𝑤 𝐼𝐼 = 0 3.51
IV
d4 w II
d2 w
Where w = 4 and w =
dx dx 2
The Finite Difference pattern formulated for fourth – order and second – order
derivatives of the Governing Differential Equation respectively is given thus:
34
wIV= 1
12h4 -1 18 -63 92 -63 18 -1
wII = 1 -1 16 -30 16 -1
12h2
Studying the various cases, we have:
3.2.1. CASE 1(n = 3): FOR THREE FLEXIBLE NODES.
A. BOUNDARY CONDITION FOR PIN – ROLLER (P –R) SUPPORT:
4 5 8 1 2 3 9 6 7
Nodes 4, 5, 6 and 7 are fictitious nodes; Nodes 1, 2, 3 are the flexible nodes while
Nodes 8 and 9 are nodes at support
The various boundary conditions apply for a pin – roller support:
𝐼𝐼
𝑤(0) = 0; 𝑤(0) =0 3.52𝑎
𝐼𝐼
𝑤(𝐿) = 0; 𝑤(𝐿) =0 3.52𝑏
This implies that
𝑤8 = 𝑤9 = 0; 𝑤4 = 𝑤7 = −𝑤2 ; 𝑤5 = −𝑤1 ; 𝑤6 = −𝑤3 3.53
Applying the new finite difference pattern and boundary conditions to the various
nodes, we have:
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 )
12ℎ4
1
+ 2
(−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤5 + 16𝑤8 )𝑘 2 = 0 3.54
12ℎ
Applying boundary conditions, we have:
35
1
(92𝑤1 − 63𝑤2 + 18𝑤3 + 𝑤2 − 18𝑤1 − 0 − 0)
12ℎ4
1
+ (−30𝑤1 + 16𝑤2 − 𝑤3 + 𝑤1 + 0)𝑘 2 = 0
12ℎ2
This gives
1 1
4
(74𝑤1 − 62𝑤2 + 18𝑤3 ) + 2
(−29𝑤1 + 16𝑤2 − 𝑤3 )𝑘 2 = 0 3.55
12ℎ 12ℎ
= (74𝑤1 − 62𝑤2 + 18𝑤3 ) + ℎ2 𝑘 2 (−29𝑤1 + 16𝑤2 − 𝑤3 ) = 0
let  = ℎ2 𝑘 2 3.56
(74𝑤1 − 62𝑤2 + 18𝑤3 ) + (−29𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.57
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 )
12ℎ4
1
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤8 − 𝑤9 )𝑘 2 = 0 3.58
12ℎ2
Applying boundary conditions, this implies that
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 𝑤1 + 𝑤3 + 0) + ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 0)

=0
(−62𝑤1 + 92𝑤2 − 62𝑤3 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 ) = 0 3.59
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 − 𝑤7 + 𝑤8 + 63𝑤9 )
12ℎ4
1
+ 2
(−𝑤1 + 16𝑤2 − 30𝑤3 − 𝑤6 + 16𝑤9 )𝑘 2 = 0 3.60
12ℎ
Applying boundary conditions, this implies that
(18𝑤1 − 63𝑤2 + 92𝑤3 − 18𝑤3 + 𝑤2 − 0 − 0)

+ ℎ2 𝑘 2 (−𝑤1 + 16𝑤2 − 30𝑤3 + 𝑤3 + 0) = 0
36
(18𝑤1 − 62𝑤2 + 74𝑤3 ) + (−𝑤1 + 16𝑤2 − 29𝑤3 ) = 0 3.61
Re-arranging equations 3.57, 3.59 and 3.61 into matrix form gives:
74 − 62 18 −29 16 − 1 𝑤1
[−62 92 − 62] +  [ 16 − 30 16 ] {𝑤2 } = 0
18 − 62 74 −1 16 − 29 𝑤3
B. BOUNDARY CONDITION FOR CLAMPED-CLAMPED (C - C)

SUPPORTS.
4 5 8 1 2 3 9 6 7
The various boundary conditions apply:
𝐼
𝑤(0) = 0; 𝑤(0) = 0; this implies that 𝑤8 = 𝑤9 = 0
𝐼
𝑤(𝑙) = 0; 𝑤(𝑙) = 0; this implies that 𝑤8𝐼 = 𝑤9𝐼 = 0
Applying the pattern for 𝑤 𝐼 at node 8 and 9, we have
wI = 1 1 -8 0 8 -1
12h
This implies that
𝑤4 − 8𝑤5 + 0 + 8𝑤1 − 𝑤2 = 0 3.62
For this to be possible, then 𝑤4 = 𝑤2 ; 𝑤5 = 𝑤1
Similarly, 𝑤6 = 𝑤3 ; 𝑤7 = 𝑤2
Applying these boundary conditions into the various nodes, we have
37
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 )
12ℎ4
1
+ 2
(−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤5 + 16𝑤8 )𝑘 2 = 0 3.63
12ℎ
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤2 + 18𝑤1 − 0 − 0)
+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤1 + 0) = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 ) + (−31𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.64
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 )
12ℎ4
1
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤8 − 𝑤9 )𝑘 2 = 0 3.65
12ℎ2
(−63𝑤1 + 92𝑤2 − 63𝑤3 − 𝑤1 − 𝑤3 + 0 + 0) + ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 )
=0
(−64𝑤1 + 92𝑤2 − 64𝑤3 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 ) = 0 3.66
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 − 𝑤7 − 𝑤8 − 63𝑤9 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 − 𝑤6 + 16𝑤9 ) = 0 3.67
12ℎ2
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤3 − 𝑤2 − 0 − 0)
+ ℎ2 𝑘 2 (−𝑤1 + 16𝑤2 − 30𝑤3 − 𝑤3 + 0) = 0
(18𝑤1 − 64𝑤2 + 110𝑤3 ) + (−𝑤1 + 16𝑤2 − 31𝑤3 ) = 0 3.68
Re-arranging equations 3.64, 3.66 and 3.68 into matrix form:
110 − 64 18 −31 16 − 1 𝑤1
[−64 92 − 64] +  [ 16 − 30 16 ] {𝑤2 } = 0 3.69
18 − 64 110 −1 16 − 31 𝑤3
38
C. BOUNDARY CONDITION FOR CLAMPED-ROLLER (C – R)
SUPPORTS.
4 5 8 1 2 3 9 6 7
𝐼
𝑤(0) = 0, 𝑤(0) = 0; this implies that 𝑤4 = 𝑤2 ; 𝑤5 = 𝑤1 ; 𝑤8 = 0
𝐼𝐼
𝑤(𝐿) = 0, 𝑤(𝐿) = 0; this implies that 𝑤6 = −𝑤3; 𝑤7 = −𝑤2 ; 𝑤9 = 0
Applying boundary conditions into various nodes, we have
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 )
12ℎ4
𝑘2
+ 2
(−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤5 + 16𝑤8 )𝑘 2 = 0 3.70
12ℎ
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤2 + 18𝑤1 − 0 − 0)
+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤1 + 0) = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 ) + (−31𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.71
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 )
12ℎ4
1
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤8 − 𝑤9 )𝑘 2 = 0 3.72
12ℎ2
(−63𝑤1 + 92𝑤2 − 63𝑤3 − 𝑤1 + 𝑤3 + 0 + 0) + ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 )
=0
(−64𝑤1 + 92𝑤2 − 62𝑤3 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 ) = 0 3.73
39
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 − 𝑤7 + 𝑤8 + 63𝑤9 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 − 𝑤6 + 16𝑤9 ) = 0 3.74
12ℎ2
(18𝑤1 − 63𝑤2 + 92𝑤3 − 18𝑤3 + 𝑤2 − 0 − 0)
+ ℎ2 𝑘 2 (−𝑤1 + 16𝑤2 − 30𝑤3 + 𝑤3 + 0) = 0
(18𝑤1 − 62𝑤2 + 74𝑤3 ) + (−𝑤1 + 16𝑤2 − 29𝑤3 ) = 0 3.75
Re-arranging equations 3.71, 3.73 and 3.75 into matrix form:
110 − 64 18 −31 16 − 1 𝑤1
[−64 92 − 62] +  [ 16 − 30 16 ] {𝑤2 } = 0
18 − 62 74 −1 16 − 29 𝑤3
3.2.2. CASE 2(n = 5):- FOR 5 FLEXIBLE NODES.
Applying the pattern for 𝑤 𝐼𝑉 and 𝑤 𝐼𝐼 as developed to the various nodes, we obtain:
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤6 − 63𝑤7 − 𝑤10 )
12ℎ4
𝑘2
+ (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤7 + 16𝑤10 ) = 0 3.76
12ℎ2
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 )
12ℎ4
𝑘2
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 𝑤10 ) = 0 3.77
12ℎ2
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.78
12ℎ2
40
Node 4:
1
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 18𝑤8 + 𝑤11 )
12ℎ4
𝑘2
+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤11 ) = 0 3.79
12ℎ2
Node 5:
1
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 )
12ℎ4
𝑘2
+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤11 − 𝑤8 ) = 0 3.80
12ℎ2
Applying the various boundary conditions to equations 3.76 – 3.80, we obtain:
A. BOUNDARY CONDITION FOR PIN – ROLLER (P – R) SUPPORTS.
6 7 10 1 2 3 4 5 11 8 9
Nodes 6, 7, 8 and 9 are fictitious nodes; Nodes 1, 2, 3, 4 and 5 are the flexible nodes
while Nodes 10 and 11 are nodes at support
𝐼𝐼
𝑤(0) = 0; 𝑤(0) = 0; This implies that 𝑤6 = −𝑤2 ; 𝑤7 = −𝑤1 ; 𝑤10 = 𝑤11 = 0
𝐼𝐼
𝑤(𝐿) = 0; 𝑤(𝐿) = 0; This implies that 𝑤8 = −𝑤5 ; 𝑤9 = −𝑤4
Substituting the various boundary conditions into equations 3.76 to 3.80, we obtain:
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 )
12ℎ4
𝑘2
+ (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤7 + 16𝑤10 ) = 0 3.81
12ℎ2
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 𝑤2 − 18𝑤1 − 0)
+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 + 𝑤1 + 0) = 0
41
(74𝑤1 − 62𝑤2 + 18𝑤3 − 𝑤4 ) + (−29𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.82
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 )
12ℎ4
𝑘2
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 𝑤10 ) = 0 3.83
12ℎ2
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 + 𝑤1 + 0)
+ ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 ) = 0
(−62𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 )

=0 3.84
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.85
12ℎ2
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 ) + (−𝑤1 + 16𝑤2 − 30𝑤3 + 𝑤4 − 𝑤5 )
=0 3.86
Node 4:
1
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 + 18𝑤11 )
12ℎ4
𝑘2
+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤11 ) = 0 3.87
12ℎ2
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 𝑤5 + 0)
+ ℎ2 𝑘 2 (−𝑤2 + 16𝑤3 − 30𝑤4 +16𝑤5 + 0) = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 62𝑤5 ) + (−𝑤2 + 16𝑤3 − 30𝑤4 +16𝑤5 )

=0 3.88
42
Node 5:
1
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 )
12ℎ4
𝑘2
+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤11 − 𝑤8 ) = 0 3.89
12ℎ2
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 18𝑤5 + 𝑤4 − 0)
+ ℎ2 𝑘 2 (−𝑤3 + 16𝑤4 − 30𝑤5 +𝑤5 ) = 0
(−𝑤2 + 18𝑤3 − 62𝑤4 + 74𝑤5 ) + (−𝑤3 + 16𝑤4 − 29𝑤5 ) = 0 3.90
Re-arranging equations 3.82, 3.84, 3.86, 3.88 and 3.90 into matrix form, we have:
74 −62 18 −1 0 −29 16 −1 0 0 𝑤1
−62 92 −63 18 −1 16 −30 16 −1 0 𝑤2
18 −63 92 −63 18 +  −1 16 −30 16 −1 𝑤3 = 0 3.91
−1 18 −63 92 −62 0 −1 16−30 16 𝑤4
[ 0 −1 18 −62 74 ] [ 0 0 −1 16 −29] {𝑤5 }
B. BOUNDARY CONDITION FOR C – C SUPPORTS.
6 7 10 1 2 3 4 5 11 8 9
𝐼
𝑤(0) = 0, 𝑤(0) = 0; This implies that 𝑤6 = 𝑤2 ; 𝑤7 = 𝑤1 ; 𝑤10 = 𝑤11
=0
𝐼
𝑤(𝐿) = 0; 𝑤(𝐿) = 0; This implies that 𝑤8 = 𝑤5 ; 𝑤9 = 𝑤4
Substituting the various boundary conditions into equations 3.76 to 3.80
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 )
12ℎ4
𝑘2
+ (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤7 + 16𝑤10 ) = 0 3.92
12ℎ2
43
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0)
+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤1 + 0) = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 ) + (−31𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.93
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 )
12ℎ4
𝑘2
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 𝑤10 ) = 0 3.94
12ℎ2
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0)
+ ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 ) = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 )

=0 3.95
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.96
12ℎ2
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 )
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.97
Node 4:
1
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 + 18𝑤11 )
12ℎ4
𝑘2
+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤11 ) = 0 3.98
12ℎ2
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤5 + 0)
+ ℎ2 𝑘 2 (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 ) = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 64𝑤5 ) + (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 )

=0 3.99
44
Nodes 5:
1
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 )
12ℎ4
𝑘2
+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤11 − 𝑤8 ) = 0 3.100
12ℎ2
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤5 − 𝑤4 − 0)
+ ℎ2 𝑘 2 (−𝑤3 + 16𝑤4 − 30𝑤5 + 0 − 𝑤5 ) = 0
(−𝑤2 + 18𝑤3 − 64𝑤4 + 110𝑤5 ) + (−𝑤3 + 16𝑤4 − 31𝑤5 ) = 0 3.101
Re-arranging equations 3.93, 3.95, 3.97, 3.99 and 3.101 into matrix form:
110 −64 18 −1 0 −31 16 −1 0 0 𝑤1

−64 92 −63 18 −1 16 −30 16 −1 0 𝑤2
18 −63 92 −63 18 +  −1 16 −30 16 −1 𝑤3 = 0 3.102
−1 18 −63 92 −64 0 −1 16 −30 16 𝑤4
[ 0 −1 18 −64 110 ] [ 0 0 −1 16−31] {𝑤5 }
C. BOUNDARY CONDITION FOR C – R SUPPORT.
6 7 10 1 2 3 4 5 11 8 9
The various boundary conditions applies
𝐼
𝑤(0) = 0, 𝑤(0) = 0; This implies that 𝑤6 = 𝑤2 ; 𝑤7 = 𝑤1 ; 𝑤10 = 𝑤11 = 0
𝐼𝐼
𝑤(𝐿) = 0, 𝑤(𝐿) = 0; This implies that 𝑤8 = −𝑤5 ; 𝑤9 = −𝑤4
Substituting the various boundary conditions into equations 3.76 to 3.80
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 )
12ℎ4
𝑘2
+ (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤7 + 16𝑤10 ) = 0 3.103
12ℎ2
45
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0)
+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 + 𝑤1 + 0) = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 ) + (−31𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.104
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 )
12ℎ4
𝑘2
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 𝑤10 ) = 0 3.105
12ℎ2
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0)
+ ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 ) = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 )

=0 3.106
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.107
12ℎ2
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 )
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.108
Node 4:
1
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 + 18𝑤11 )
12ℎ4
𝑘2
+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤11 ) = 0 3.109
12ℎ2
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 𝑤5 + 0)
+ ℎ2 𝑘 2 (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 ) = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 62𝑤5 ) + (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 )

=0 3.110
46
Nodes 5:
1
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤11 + 18𝑤8 − 𝑤9 )
12ℎ4
𝑘2
+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤11 − 𝑤8 ) = 0 3.111
12ℎ2
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 0 − 18𝑤5 + 𝑤4 )
+ ℎ2 𝑘 2 (−𝑤3 + 16𝑤4 − 30𝑤5 + 0 + 𝑤5 ) = 0
(−𝑤2 + 18𝑤3 − 62𝑤4 + 74𝑤5 ) + (−𝑤3 + 16𝑤4 − 29𝑤5 ) = 0 3.112
Re-arranging equations 3.104, 3.106, 3.108, 3.110 and 3.112 into matrix form, we
have:
110 −64 18 −1 0 −31 16 −1 0 0 𝑤1

−64 92 −63 18 −1 16 −30 16 −1 0 𝑤2
18 −63 92 −63 18 +  −1 16−30 16 −1 𝑤3 = 0 3.113
−1 18 −63 92 −62 0 −1 16 −30 16 𝑤4
[ 0 −1 18 −62 74 ] [ 0 0 −1 16−29] {𝑤5 }
3.2.3. CASE 3 (n = 7):- FOR 7 FLEXABLE NODES.
Applying the pattern for 𝑤 𝐼𝑉 and 𝑤 𝐼𝐼 to the various nodes, we obtain:
A. BOUNDARY CONDITION FOR PIN – ROLLER (P – R) SUPPORT.

8 9 10 1 2 3 4 5 6 7 11 12 13
Nodes 8, 9, 12 and 13 are fictitious nodes; Nodes 1, 2, 3, 4, 5, 6 and 7 are the

flexible nodes while Nodes 10 and 11 are nodes at support
𝐼𝐼
𝑤(0) = 0, 𝑤(0) = 0; This implies that 𝑤8 = −𝑤2 ; 𝑤9 = −𝑤2 ; 𝑤10 = 𝑤11 = 0
𝐼𝐼
𝑤(𝐿) = 0; 𝑤(𝐿) = 0; This implies that 𝑤12 = −𝑤7 ; 𝑤13 = −𝑤6
Substituting the various boundary conditions into the various nodes, we obtain:
47
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤8 + 18𝑤9 − 63𝑤10 )
12ℎ4
𝑘2
+ (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤9 + 16𝑤10 ) = 0 3.114
12ℎ2
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 𝑤2 − 18𝑤1 − 0)
+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 + 𝑤1 + 0) = 0
(74𝑤1 − 62𝑤2 + 18𝑤3 − 𝑤4 ) + (−29𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.115
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤9 + 18𝑤10 )
12ℎ4
𝑘2
+ (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 𝑤10 ) = 0 3.116
12ℎ2
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 + 𝑤1 + 0)
+ ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 0) = 0
(−62𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 )

=0 3.117
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 − 𝑤10 )
12ℎ4
𝑘2
+ (−𝑤1 + 16𝑤2 − 30𝑤3 − 16𝑤4 − 𝑤5 ) = 0 3.118
12ℎ2
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 − 𝑤6 )
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.119
Node 4:
1
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 )
12ℎ4
𝑘2
+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤6 ) = 0 3.120
12ℎ2
48
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 )
+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤6 ) = 0 3.121
Node 5:
1
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 − 𝑤11 )
12ℎ4
𝑘2
+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤6 − 𝑤7 ) = 0 3.122
12ℎ2
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 )
+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤6 − 𝑤7 ) = 0 3.123
Node 6:
1
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 18𝑤11 − 𝑤12 )
12ℎ4
𝑘2
+ (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 − 𝑤11 ) = 0 3.124
12ℎ2
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 0 + 𝑤7 )
+ ℎ2 𝑘 2 (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 ) = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 62𝑤7 ) + (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 )

=0 3.125
Node 7:
1
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 18𝑤11 + 18𝑤12 − 𝑤13 )
12ℎ4
𝑘2
+ (−𝑤5 + 16𝑤6 − 30𝑤7 + 16𝑤11 − 𝑤12 ) = 0 3.126
12ℎ2
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 0 − 18𝑤7 + 𝑤6 )
+ ℎ2 𝑘 2 (−𝑤5 + 16𝑤6 − 30𝑤7 + 0 + 𝑤7 ) = 0
(−𝑤4 + 18𝑤5 − 62𝑤6 + 74𝑤7 ) + (−𝑤5 + 16𝑤6 − 29𝑤7 ) = 0 3.127
Re-arranging equations 3.115, 3.117, 3.119, 3.121, 3.123, 3.125 and 3.127 into
matrix form, we obtain:
49
74 −62 18 −1 0 0 0 −29 16 −1 0 0 0 0 𝑤1
−62 92−63 18 −1 0 0 16 −30 16 −1 0 0 0 𝑤2
18 −63 92 −63 18 −1 0 −1 16−30 16 −1 0 0 𝑤3
−1 18−63 92−63 18 −1 +  0 −1 16 −30 16 −1 0 𝑤4
0 −1 18 −63 92 −63 18 0 0 −1 16 −30 16 −1 𝑤5
0 0 −1 18 −63 92−62 0 0 0 −1 16−30 16 𝑤6
[ 0 0 0 −1 18 −62 74 ] [ 0 0 0 0 −1 16 −29] {𝑤7 }
=0 3.128
B. BOUNDARY CONDITION FOR CLAMPED – CLAMPED SUPPORT.

8 9 10 1 2 3 4 5 6 7 11 12 13
𝐼
𝑤(0) = 0, 𝑤(0) = 0; This imples that 𝑤8 = 𝑤2 ; 𝑤9 = 𝑤2 ; 𝑤10 = 𝑤11 = 0
𝐼
𝑤(𝐿) = 0, 𝑤(𝐿) = 0; This implies that 𝑤12 = 𝑤7 ; 𝑤13 = 𝑤6
Substituting the various boundary conditions into the various nodes, we obtain:
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 − 18𝑤1 − 0)

+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 − 𝑤1 + 0) = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 ) + (−31𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.129
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0)

+ ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 − 0) = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 )

=0 3.130
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 − 𝑤6 − 0)

+ ℎ2 𝑘 2 (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0
50
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 − 𝑤6 )
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.131
Node 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 )

+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤6 ) = 0 3.132
Node 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 − 0)

+ ℎ2 𝑘 2 (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤6 − 𝑤7 ) = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 )

+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤6 − 𝑤7 ) = 0 3.133
Node 6:
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 0 − 𝑤7 )

+ ℎ2 𝑘 2 (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 − 0) = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 64𝑤7 ) + (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 )

=0 3.134
Node 7:
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 0 + 18𝑤7 − 𝑤6 )

+ ℎ2 𝑘 2 (−𝑤5 + 16𝑤6 − 30𝑤7 + 0 − 𝑤7 ) = 0
(−𝑤4 + 18𝑤5 − 64𝑤6 + 110𝑤7 ) + (−𝑤5 + 16𝑤6 − 31𝑤7 ) = 0 3.135
Re-arranging equations 3.129, 3.130, 3.131, 3.132, 3.133, 3.134 and 3.135 into
matrix form:
51
110−64 18 −1 0 0 0 −31 16 −1 0 0 0 0 𝑤1
−64 92 −63 18 −1 0 0 16−30 16 −1 0 0 0 𝑤2
18 −63 92 −63 18 −1 0 −1 16 −30 16 −1 0 0 𝑤3
−1 18 −63 92−63 18 −1 +  0 −1 16 −30 16 −1 0 𝑤4
0 −1 18 −63 92 −63 18 0 0 −1 16−30 16 −1 𝑤5
0 0 −1 18 −63 92 −64 0 0 0 −1 16 −30 16 𝑤6
[ 0 0 0 −1 18 −64 110] [ 0 0 0 0 −1 16 −31] {𝑤7 }
=0 3.136
C. BOUNDARY CONDITION FOR CLAMPED – ROLLER (C – R)

SUPPORT.
8 9 10 1 2 3 4 5 6 7 11 12 13
𝐼
𝑤(0) = 0, 𝑤(0) = 0; This implies that 𝑤8 = 𝑤2 ; 𝑤9 = 𝑤2 ; 𝑤10 = 𝑤11 = 0
𝐼𝐼
𝑤(𝐿) = 0, 𝑤(𝐿) = 0; This implies that 𝑤12 = −𝑤7 ; 𝑤13 = −𝑤6
Substituting the various boundary conditions into the various nodes:
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0)

+ ℎ2 𝑘 2 (−30𝑤1 + 16𝑤2 − 𝑤3 + 𝑤1 + 0) = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 ) + (−31𝑤1 + 16𝑤2 − 𝑤3 ) = 0 3.137
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0)

+ ℎ2 𝑘 2 (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 + 0) = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) + (16𝑤1 − 30𝑤2 + 16𝑤3 − 𝑤4 )

=0 3.138
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 − 𝑤6 − 0)

+ ℎ2 𝑘 2 (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0
52
(18𝑤1 − 63𝑤2 + 92𝑤3 − 634 + 18𝑤5 − 𝑤6 )
+ (−𝑤1 + 16𝑤2 − 30𝑤3 + 16𝑤4 − 𝑤5 ) = 0 3.139
Node 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤)

+ (−𝑤2 + 16𝑤3 − 30𝑤4 + 16𝑤5 − 𝑤6 ) = 0 3.140
Node 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 − 0)

+ ℎ2 𝑘 2 (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤6 − 𝑤7 ) = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 )

+ (−𝑤3 + 16𝑤4 − 30𝑤5 + 16𝑤6 − 𝑤7 ) = 0 3.141
Node 6:
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 0 + 𝑤7 )

+ ℎ2 𝑘 2 (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 − 0) = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 62𝑤7 ) + (−𝑤4 + 16𝑤5 − 30𝑤6 + 16𝑤7 )

=0 3.142
Node 7:
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 0 − 18𝑤7 + 𝑤6 )

+ ℎ2 𝑘 2 (−𝑤5 + 16𝑤6 − 30𝑤7 + 0 + 𝑤7 ) = 0
(−𝑤4 + 18𝑤5 − 62𝑤6 + 74𝑤7 ) + (−𝑤5 + 16𝑤6 − 29𝑤7 ) = 0 3.143
Re-arranging equations 3.90, 3.91, 3.92, 3.93, 3.94, 3.95 and 3.96 into matrix form:
110−64 18 −1 0 0 0 −31 16 −1 0 0 0 0 𝑤1
−64 92 −63 18 −1 0 0 16 −30 16 −1 0 0 0 𝑤2
18 −63 92 −63 18 −1 0 −1 16 −30 16 −1 0 0 𝑤3
−1 18 −63 92−63 18 −1 +  0 −1 16−30 16 −1 0 𝑤4
0 −1 18 −63 92 −63 18 0 0 −1 16 −30 16 −1 𝑤5
0 0 −1 18 −63 92 −62 0 0 0 −1 16 −30 16 𝑤6
[ 0 0 0 −1 18 −62 74 ] [ 0 0 0 0 −1 16−29] {𝑤7 }
=0 3.144
53
3.3. VIBRATION ANYALYSIS APPROACH.
Governing Differential Equation for vibration analysis of line continuum was given
by Chakraverty (2009) as:
𝑑 4 𝑤 𝜌𝐴2
− 𝑤(𝑥) = 0 3.145
𝑑𝑥 4 𝐸𝐼
𝜌𝐴2
Let  =
𝐸𝐼
𝑑4𝑤
− 𝑤(𝑥) = 0; which is 𝑤 𝐼𝑉 − 𝜆𝑤(𝑥) 3.146
𝑑𝑥 4
The new finite Difference pattern developed for wiv is given as
wIV= 1
12h4 -1 18 -63 92 -63 18 -1
3.3.1. CASE 1 (n = 3); -FOR 3 FLEXABLE NODES.
Applying the pattern to the various nodes:
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 ) − 𝑤1 = 0
12ℎ4
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 ) − 𝛽𝑤1 = 0 3.147
Where β = 12h4𝜆
Node 2:
1
(−63𝑤1 − 92𝑤2 − 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 ) − 𝑤2 = 0
12ℎ4
(−63𝑤1 − 92𝑤2 − 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 ) − β𝑤2 = 0 3.148
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 + 𝑤7 − 𝑤8 − 63𝑤9 ) − 𝑤3 = 0
12ℎ4
54
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 + 𝑤7 − 𝑤8 − 63𝑤9 ) − 𝑤3 = 0 3.149
Applying the various boundary conditions, we obtain:

4 5 8 1 2 3 9 6 7
𝐼𝐼
𝑤(0) = 0; 𝑤(0) =0
𝐼𝐼
𝑤(𝐿) = 0; 𝑤(𝐿) =0; This implies that
𝑤8 = 𝑤9 = 0; 𝑤4 = 𝑤7 = −𝑤2 ; 𝑤5 = −𝑤1 ; 𝑤6 = −𝑤3
Applying these boundary conditions into equations 3.147 – 3.149, we obtain:
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 + 𝑤2 − 18𝑤1 − 0 − 0) − 𝛽𝑤1 = 0
(74𝑤1 − 62𝑤2 + 18𝑤3 ) − 𝛽𝑤1 = 0 3.150
Node 2:
(−63𝑤1 + 92𝑤2 + 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 𝑤1 + 𝑤3 + 0 + 0) − 𝛽𝑤2 = 0
 (−62𝑤1 + 92𝑤2 − 62𝑤3 ) − 𝛽𝑤2 = 0 3.151
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 − 𝑤7 − 𝑤8 − 63𝑤9 ) − 𝛽𝑤3 = 0
(18𝑤1 − 63𝑤2 + 92𝑤3 − 18𝑤3 + 𝑤2 − 0 − 0) − 𝛽𝑤3 = 0
(18𝑤1 − 62𝑤2 + 74𝑤3 ) − 𝛽𝑤3 = 0 3.152
Arranging Equations 3.150, 3.151 and 3.152 in matrix form, we obtain:
55
74 − 62 18 1 0 0 𝑤1
[ −62 92 − 62 ] − 𝛽 [0 1 0] [𝑤2 ] = 0 3.153
18 − 62 74 0 0 1 𝑤2
4 5 8 1 2 3 9 6 7
𝐼 𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤4 = 𝑤7 = 𝑤2 ; 𝑤6 = 𝑤3 ; 𝑤5 = 𝑤1 ; 𝑤8 = 𝑤9 = 0
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤2 + 18𝑤1 − 0 − 0) − 𝛽𝑤1 = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 ) − 𝛽𝑤1 = 0 3.154
Node 2:
(−63𝑤1 − 92𝑤2 − 63𝑤3 − 𝑤5 + 𝑤6 − 18𝑤8 − 18𝑤9 ) − 𝛽𝑤2 = 0
(−63𝑤1 − 92𝑤2 − 63𝑤3 − 𝑤1 − 𝑤3 + 0 + 0) − 𝛽𝑤2 = 0
 (−64𝑤1 − 92𝑤2 + 64𝑤3 ) − 𝛽𝑤2 = 0 3.155
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤6 − 𝑤7 − 𝑤8 − 63𝑤9 ) − 𝛽𝑤3 = 0
(18𝑤1 − 63𝑤2 + 92𝑤3 + 18𝑤3 − 𝑤2 − 0 − 0) − 𝛽𝑤3 = 0
(18𝑤1 − 64𝑤2 + 110𝑤3 ) − 𝛽𝑤3 = 0 3.156
Arranging Equations 3.107 – 3.109 into matrix form gives:
56
110 − 64 18 1 0 0 𝑤1
[−64 92 − 64] − 𝛽 [0 1 0] {𝑤2 } = 0 3.157
18 − 64 110 0 0 1 𝑤3
C. BOUNDARY CONDITION FOR CLAMPED – ROLLER(C – R)

SUPPORT.
4 5 8 1 2 3 9 6 7
𝐼 𝐼𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤4 = 𝑤2 ; 𝑤6 = −𝑤3 ; 𝑤8 = 𝑤9 = 0; 𝑤5 = 𝑤1 ; 𝑤7 = −𝑤2
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 18𝑤5 − 63𝑤8 − 𝑤9 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤2 + 18𝑤1 − 0 − 0) − 𝛽𝑤1 = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 ) − 𝛽𝑤1 = 0 3.158
Node 2:
(−63𝑤1 − 92𝑤2 + 63𝑤3 − 𝑤5 − 𝑤6 + 18𝑤8 + 18𝑤9 ) − 𝛽𝑤2 = 0
(−63𝑤1 − 92𝑤2 − 63𝑤3 − 𝑤1 + 𝑤3 + 0 + 0) − 𝛽𝑤2 = 0
 (−64𝑤1 − 92𝑤2 − 62𝑤3 ) − 𝛽𝑤2 = 0 3.159
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 18𝑤6 + 𝑤7 − 𝑤8 − 63𝑤9 ) − 𝛽𝑤3 = 0
(18𝑤1 − 63𝑤2 + 92𝑤3 − 18𝑤3 + 𝑤2 − 0 − 0) − 𝛽𝑤3 = 0
(18𝑤1 − 62𝑤2 + 74𝑤3 ) − 𝛽𝑤3 = 0 3.160
Arranging Equations 3.158 – 3.160 in matrix form gives:
57
110 − 64 18 1 0 0 𝑤1
[−64 92 − 62] − 𝛽 [0 1 0] {𝑤2 } = 0 3.161
18 − 62 74 0 0 1 𝑤3
3.3.2. CASE 2 (n = 5):- FOR 5 FLEXABLE NODES.
Applying the new finite difference pattern to the various nodes, we obtain:
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 ) − 𝜆𝑤1 = 0
12ℎ4
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 ) − 𝛽𝑤1 = 0 3.162
Where β = 12h4𝜆
Node 2:
1
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 ) − 𝜆𝑤2 = 0
12ℎ4
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 ) − β𝑤2 = 0 3.163
Node 3:
1
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 ) − 𝑤3 = 0
12ℎ4
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 ) − β𝑤3 = 0 3.164
Nodes 4:
1
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 − 18𝑤11 ) − 𝑤4 = 0
12ℎ4
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 − 18𝑤11 ) − β𝑤4 = 0 3.165
Nodes 5:
1
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 ) − 𝑤5 = 0
12ℎ4
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 ) − β𝑤5 = 0 3.166
58
Applying the various boundary conditions, we obtain:
6 7 10 1 2 3 4 5 11 8 9
𝐼𝐼 𝐼𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤10 = 𝑤11 = 0; 𝑤8 = −𝑤5 ; 𝑤7 = −𝑤1 ; 𝑤9 = −𝑤4 ; 𝑤6

= −𝑤2
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 𝑤2 − 18𝑤1 ) − 𝛽𝑤1 = 0
(74𝑤1 − 62𝑤2 + 18𝑤3 − 𝑤4 ) 𝛽𝑤1 = 0 3.167
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 + 𝑤1 + 0) − 𝛽𝑤2 = 0
(−62𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) − 𝛽𝑤2 = 0 3.168
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 ) − 𝛽𝑤3 = 0
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 ) − 𝛽𝑤3 = 0 3.169
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 + 18𝑤11 ) − 𝛽𝑤4 = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 62𝑤5 ) − 𝛽𝑤4 = 0 3.170
59
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 ) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 62𝑤4 + 74𝑤5 ) − 𝛽𝑤5 = 0 3.171
74 −62 18 −1 0 1 0 0 0 0 𝑤1
−62 92 −63 18 −1 0 1 0 0 0 𝑤2
18 −63 92 −63 18 − 𝛽 0 0 1 0 0 𝑤3 = 0 3.172
−1 18 −63 92 −62 0 0 0 1 0 𝑤4
[ 0 −1 18 −62 74 ] [0 0 0 0 1] {𝑤5 }
6 7 10 1 2 3 4 5 11 8 9
𝐼 𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤6 = 𝑤2 ; 𝑤8 = 𝑤5 ; 𝑤10 = 𝑤11 = 0; 𝑤7 = 𝑤1 ; 𝑤9 = 𝑤4
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0) − 𝛽𝑤1 = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 )𝛽𝑤1 = 0 3.173
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0) − 𝛽𝑤2 = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) − 𝛽𝑤2 = 0 3.174
60
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 ) − 𝛽𝑤3 = 0
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 0) − 𝛽𝑤3 = 0
 (18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 ) − 𝛽𝑤3 = 0 3.175
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 + 18𝑤11 ) − 𝛽𝑤4 = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤5 + 0) − 𝛽𝑤4 = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 64𝑤5 ) − 𝛽𝑤4 = 0 3.176
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 ) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤5 − 𝑤4 − 0) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 64𝑤4 + 110𝑤5 ) − 𝛽𝑤5 = 0 3.177
110 −64 18 −1 0 1 0 0 0 0 𝑤1
−64 92 −63 18 −1 0 1 0 0 0 𝑤2
18 −63 92 −63 18 −  0 0 1 0 0 𝑤3 = 0 3.178
−1 18 −63 92 −64 0 0 0 1 0 𝑤4
[ 0 −1 18 −64 110 ] [0 0 0 0 1] {𝑤5 }
6 7 10 1 2 3 4 5 11 8 9
𝐼 𝐼𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤6 = 𝑤2 ; 𝑤8 = −𝑤5 ; 𝑤10 = 𝑤11 = 0 ; 𝑤7 = 𝑤1 ; 𝑤9

= −𝑤4

61
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤6 + 18𝑤7 − 63𝑤10 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0) − 𝛽𝑤1 = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 )𝛽𝑤1 = 0 3.179
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤7 + 18𝑤10 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0) − 𝛽𝑤2 = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) − 𝛽𝑤2 = 0 3.180
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤10 − 𝑤11 ) − 𝛽𝑤3 = 0
 (18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 ) − 𝛽𝑤3 = 0 3.181
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 − 𝑤8 + 18𝑤11 ) − 𝛽𝑤4 = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 𝑤5 + 0) − 𝛽𝑤4 = 0
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 62𝑤5 ) − 𝛽𝑤4 = 0 3.182
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 + 18𝑤8 − 𝑤9 − 63𝑤11 ) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 18𝑤5 + 𝑤4 − 0) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 62𝑤4 + 74𝑤5 ) − 𝛽𝑤5 = 0 3.183
110 −64 18 −1 0 1 0 0 0 0 𝑤1
−64 92 −63 18 −1 0 1 0 0 0 𝑤2
18 −63 92 −63 18 − 𝛽 0 0 1 0 0 𝑤3 = 0 3.184
−1 18 −63 92 −62 0 0 0 1 0 𝑤4
[ 0 −1 18 −62 74 ] [0 0 0 0 1] {𝑤5 }
62
3.3.3. CASE 3 (n = 7):- FOR 7 FLEXIBLE NODES
Applying the new finite difference pattern to the various nodes, we obtain:
Node 1:
1
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤8 + 18𝑤9 − 63𝑤10 ) − 𝑤1 = 0
12ℎ4
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤8 + 18𝑤9 − 63𝑤10 ) − 𝛽𝑤1 = 0 3.185
Similarly, at other nodes we have:
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤9 + 18𝑤10 ) − 𝛽𝑤2 = 0 3.186
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 − 𝑤10 ) − 𝛽𝑤3 = 0 3.187
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 ) − 𝛽𝑤4 = 0 3.188
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 + 18𝑤7 − 𝑤11 ) − 𝛽𝑤5 = 0 3.189
Nodes 6:
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 18𝑤11 − 𝑤12 ) − 𝛽𝑤6 = 0 3.190
Node 7:
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 63𝑤11 + 18𝑤12 − 𝑤13 ) − 𝛽𝑤7 = 0 3.191
Applying the various boundary conditions to the nodes, we obtain:
A. BOUNDARY CONDITION FOR P – R SUPPORT.
The various boundary conditions apply for pin-roller support:
𝐼𝐼 𝐼𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
63
This implies that 𝑤8 = −𝑤2 ; 𝑤12 = −𝑤7 ; 𝑤9 = −𝑤1 ; 𝑤13 = −𝑤6 ; 𝑤10 = 𝑤11
=0
Applying these boundary conditions to equations 3.185 – 3.191, we obtain:
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤8 + 18𝑤9 − 63𝑤10 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 + 𝑤2 − 18𝑤1 − 0) − 𝛽𝑤1 = 0
(74𝑤1 − 62𝑤2 + 18𝑤3 − 𝑤4 ) 𝛽𝑤1 = 0 3.192
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤9 + 18𝑤10 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 + 𝑤1 + 0) − 𝛽𝑤2 = 0
(−62𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) − 𝛽𝑤2 = 0 3.193
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 − 𝑤10 ) − 𝛽𝑤3 = 0
 (18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 ) − 𝛽𝑤3 = 0 3.194
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 ) − 𝛽𝑤4 = 0 3.195
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 − 18𝑤7 − 𝑤11 ) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 − 92𝑤5 + 63𝑤6 − 18𝑤7 ) − 𝑤5 = 0 3.196
Nodes 6:
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 18𝑤11 − 𝑤12 ) − 𝛽𝑤6 = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 0𝑤 + 𝑤7 ) − 𝛽𝑤6 = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 62𝑤7 ) − 𝛽𝑤6 = 0 3.197
64
Node 7:
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 63𝑤11 + 18𝑤12 − 𝑤13 ) − 𝛽𝑤7 = 0
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 0 − 18𝑤7 + 𝑤6 ) − 𝛽𝑤7 = 0
(−𝑤4 + 18𝑤5 − 62𝑤6 + 74𝑤7 ) − 𝛽𝑤7 = 0 3.198
74−62 18 −1 0 0 0 1 0 0 0 0 0 0 𝑤1
−62 92 −63 18 −1 0 0 0 1 0 0 0 0 0 𝑤2
18−63 92 −63 18 −1 0 0 0 1 0 0 0 0 𝑤3
−1 18 −63 92−63 18 −1 − 𝛽 0 0 0 1 0 0 0 𝑤4 = 0 3.199
0 −1 18 −63 92−63 18 0 0 0 0 1 0 0 𝑤5
0 0 −1 18 −63 92 −62 0 0 0 0 0 1 0 𝑤6
[ 0 0 0 −1 18−62 74 ] [0 0 0 0 0 0 1] {𝑤7 }
B. BOUNDARY CONDITION FOR C – C SUPPORT.
The boundary conditions for C – C are given as:
𝐼 𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤8 = 𝑤2 ; 𝑤9 = 𝑤1 ; 𝑤10 = 𝑤11 = 0; 𝑤12 = 𝑤7 ; 𝑤13 = 𝑤6
Applying these boundary conditions to equations 3.185 – 3.191, we obtain:
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤8 + 18𝑤9 − 63𝑤10 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0) − 𝛽𝑤1 = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 ) 𝛽𝑤1 = 0 3.200
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤9 + 18𝑤10 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0) − 𝛽𝑤2 = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) − 𝛽𝑤2 = 0 3.201

65
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 − 𝑤10 ) − 𝛽𝑤3 = 0
 (18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 ) − 𝛽𝑤3 = 0 3.202
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 ) − 𝛽𝑤4 = 0 3.203
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 − 18𝑤7 − 𝑤11 ) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 − 92𝑤5 + 63𝑤6 − 18𝑤7 ) − 𝛽𝑤5 = 0 3.204
Nodes 6:
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 18𝑤11 − 𝑤12 ) − 𝛽𝑤6 = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 0 − 𝑤7 ) − 𝛽𝑤6 = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 64𝑤7 ) − 𝛽𝑤6 = 0 3.205
Node 7:
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 63𝑤11 + 18𝑤12 − 𝑤13 ) − 𝛽𝑤7 = 0
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 0 + 18𝑤7 − 𝑤6 ) − 𝛽𝑤7 = 0
(−𝑤4 + 18𝑤5 − 64𝑤6 + 110𝑤7 ) − 𝛽𝑤7 = 0 3.206
Rearranging Equations 3.200 – 3.206 in matrix form gives:
110−64 18 −1 0 0 0 1000 0 0 0 𝑤1
−64 92−63 18 −1 0 0 0100 0 0 0 𝑤2
18 −63 92 −63 18 −1 0 00100 0 0 𝑤3
−1 18−63 92−63 18 −1 −  0 0 0 1 0 0 0 𝑤4 = 0 3.207
0 −1 18 −63 92 −63 18 0000 1 0 0 𝑤5
0 0 −1 18 −63 92 −64 0000 0 1 0 𝑤6
[ 0 0 0 −1 18 −64 110 ] [0000 0 0 1 ] {𝑤7 }
66
The boundary conditions are given as:

𝐼 𝐼𝐼
𝑤(0) = 0; 𝑤(0) = 0 and 𝑤(𝐿) = 0; 𝑤(𝐿) =0
This implies that 𝑤8 = 𝑤2 ; 𝑤9 = 𝑤1 ; 𝑤10 = 𝑤11 = 0 ; 𝑤12 = −𝑤7 ; 𝑤13

= −𝑤6
Substituting these boundary conditions into equations 3.185 – 3.191, we obtain:
Node 1:
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤8 + 18𝑤9 − 63𝑤10 ) − 𝛽𝑤1 = 0
(92𝑤1 − 63𝑤2 + 18𝑤3 − 𝑤4 − 𝑤2 + 18𝑤1 − 0) − 𝛽𝑤1 = 0
(110𝑤1 − 64𝑤2 + 18𝑤3 − 𝑤4 ) 𝛽𝑤1 = 0 3.208
Node 2:
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤9 + 18𝑤10 ) − 𝛽𝑤2 = 0
(−63𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 − 𝑤1 + 0) − β𝑤2 = 0
(−64𝑤1 + 92𝑤2 − 63𝑤3 + 18𝑤4 − 𝑤5 ) − 𝛽𝑤2 = 0 3.209
Node 3:
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 − 𝑤10 ) − 𝛽𝑤3 = 0
(18𝑤1 − 63𝑤2 + 92𝑤3 − 63𝑤4 + 18𝑤5 − 𝑤6 ) − 𝛽𝑤3 = 0 3.210
Nodes 4:
(−𝑤1 + 18𝑤2 − 63𝑤3 + 92𝑤4 − 63𝑤5 + 18𝑤6 − 𝑤7 ) − 𝛽𝑤4 = 0 3.211
Nodes 5:
(−𝑤2 + 18𝑤3 − 63𝑤4 + 92𝑤5 − 63𝑤6 − 18𝑤7 − 𝑤11 ) − 𝛽𝑤5 = 0
(−𝑤2 + 18𝑤3 − 63𝑤4 − 92𝑤5 − 63𝑤6 + 18𝑤7 ) − 𝛽𝑤5 = 0 3.212
67
Nodes 6:
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 18𝑤11 − 𝑤12 ) − 𝛽𝑤6 = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 63𝑤7 + 0 + 𝑤7 ) − 𝛽𝑤6 = 0
(−𝑤3 + 18𝑤4 − 63𝑤5 + 92𝑤6 − 62𝑤7 ) − 𝛽𝑤6 = 0 3.213
Node 7:
(−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 63𝑤11 + 18𝑤12 − 𝑤13 ) − 𝛽𝑤7 = 0
⟹ (−𝑤4 + 18𝑤5 − 63𝑤6 + 92𝑤7 − 0 − 18𝑤7 + 𝑤6 ) − 𝛽𝑤7 = 0
(−𝑤4 + 18𝑤5 − 62𝑤6 + 74𝑤7 ) − 𝛽𝑤7 = 0 3.214
Rearranging Equations 3.208 – 3.214 in matrix form gives:
110−64 18 −1 0 0 0 1000 0 0 0 𝑤1
−64 92 −63 18 −1 0 0 0100 0 0 0 𝑤2
18 −63 92 −63 18 −1 0 00100 0 0 𝑤3
−1 18 −63 92−63 18 −1 − 𝛽 0 0 0 1 0 0 0 𝑤4 = 0 3.215
0 −1 18 −63 92−63 18 0000 1 0 0 𝑤5
0 0 −1 18 −63 92 −62 0000 0 1 0 𝑤6
[ 0 0 0 −1 18−62 74 ] [0000 0 0 1 ] {𝑤7 }
68
CHAPTER FOUR
RESULTS AND DISCUSSION
4.1. RESULTS OF STABILITY ANALYSIS OF LINE CONTINUUM.
The governing differential equation for buckling of a line continuum (stability) is

given by Iyengar (1988) as
𝑑4𝑤 𝑁 𝑑2𝑤
+ =0 4.1
𝑑𝑥 4 𝐸𝐼 𝑑𝑥 2
𝑑4𝑤 2
𝑑2𝑤
+𝑘 =0 4.2
N
where k 2 =
EI
The algebraic equations are written in a matrix notation of the form
[A +λB] x= 0 4.3
Where x = xi is a column matrix whose element xi represent the deflections of
line continuum
A = [aij] is a square matrix obtained from the finite difference method expression,
B = [bij] is also a square matrix obtained from the finite difference method
expression,
Pre-multiplying equation 4.3 with B-1, we can obtain
(C + λI) x = 0 4.4
Where C = (-A) B-1 and I is identity matrix
Solving the matrices obtained from the new finite difference method for stability
analysis, we obtain the Eigen values as shown in Appendix A:
69
4.1.1. CASE 1 (n = 3)
𝑙
For case 1 (n = 3), h = . The lowest values of the Eigen values are taken to
4
calculate the required critical loads; that is:
I. PIN – ROLLER (P – R) SUPPORT :-

 = 0.5858 4.5
N
but = h2 k 2 and k 2 = 4.6
EI
N 𝑙 𝑙2
⟹ h2 = 0.5858; but h = 2
which means h = 4.7
EI 4 16
Substituting theses values, we obtain:
𝑁 𝑙2 0.5858 × 16 × 𝐸𝐼
. = 0.5858 ⟹ 𝑁 = 4.8
𝐸𝐼 16 𝑙2
9.3728𝐸𝐼
𝑁 = 4.9
𝑙2
II. CLAMPED – CLAMPED (C – C) SUPPORT:-
 = 2; 4.10
𝑁 2 𝑁 𝑙2
ℎ = 2; which gives . =2 ⟹ 𝑁
𝐸𝐼 𝐸𝐼 16
16 × 2 × 𝐸𝐼
= 4.11
𝑙2
32𝐸𝐼
𝑁 = 4.12
𝑙2
III. CLAMPED – ROLLER (C – R) SUPPORT:-
 = 1.1207; 4.13
𝑁 2 𝑁 𝑙2
ℎ = 1.1207; which gives . = 1.1207 4.14
16 × 1.1207 × 𝐸𝐼 17.9312𝐸𝐼
𝑁 = = 4.15
𝑙2 𝑙2
4.1.2. CASE 2 (n = 5).
70
𝑙
For case 2 (n = 5), ℎ = and picking the lowest values of the Eigen values, the
6
required critical loads are calculated; thus

 = 0.2679 4.16
𝑁 2 𝑙 2
𝑙2
that is ℎ = 0.2679; but h = ; which means ℎ = 4.17
𝐸𝐼 6 36
Substituting these values, we have
𝑁 𝑙2
. = 0.2679; which gives
𝐸𝐼 36
0.2679 × 36 × 𝐸𝐼 9.6444𝐸𝐼
𝑁 = = 4.18
𝑙2 𝑙2
 = 1.0251 4.19
𝑁 2 𝑁 𝑙2
That is . ℎ = 1.0251; which gives . = 1.0251 4.20
36 × 1.025 × 𝐸𝐼 36.9036𝐸𝐼
 𝑁= = 4.21
𝑙2 𝑙2
 = 0.5324
𝑁 2 𝑁 𝑙2
⟹ ℎ = 0.5324; which gives . = 0.5324 4.22
36 × 0.5324 × 𝐸𝐼 19.1664𝐸𝐼
𝑁 = = 4.23
𝑙2 𝑙2
4.1.3. CASE 3 (n = 7).

𝑙
For case 3 (n = 7), ℎ = and picking the lowest values of the Eigen values, the
8
required critical loads are calculated; thus

 = 0.1522 4.24
71
𝑁 2 𝑙 2
𝑙2
that is ℎ = 0.1522; but h = ; which means ℎ = 4.25
𝐸𝐼 8 64
Substituting these values, we have
𝑁 𝑙2
. = 0.1522; which gives
𝐸𝐼 64
0.1522 × 64 × 𝐸𝐼 9.7408𝐸𝐼
𝑁 = = 4.26
𝑙2 𝑙2

 = 0.5858 4.27
𝑁 2 𝑁 𝑙2
That is . ℎ = 0.5858, which gives . = 0.5858 4.28
64 × 0.5858 × 𝐸𝐼 37.4912𝐸𝐼
 𝑁= = 4.29
𝑙2 𝑙2
 = 0.3064
𝑁 2 𝑁 𝑙2
ℎ = 0.3064 ; which gives . = 0.3064 4.30
64 × 0.3064 × 𝐸𝐼 19.6096𝐸𝐼
𝑁 = = 4.31
𝑙2 𝑙2
4.2. RESULTS OF VIBRATION ANALYSIS OF LINE CONTINUUM.
Governing Differential Equation for vibration analysis of line continuum is given

by Chakraverty (2009) as:
𝑑 4 𝑤 𝜌𝐴2
− 𝑤(𝑥) = 0 4.32
𝑑𝑥 4 𝐸𝐼
𝑑4𝑤 ρAω2
which is − 𝜆𝑤(𝑥), where λ = 4.33
𝑑𝑥 4 EI
𝑑4𝑤 1
4
= 𝑤 𝐼𝑉
= 4
4 ; where 4
𝑑𝑥 12ℎ
= fourth − order new finite difference pattern
72
1
This means that 4
4 − 𝑤(𝑥) = 0 4.34
12ℎ
Hence, 4 − 12ℎ4 𝑤(𝑥) = 0 4.35
let 12h4  = 
4 − 𝑤(𝑥) = 0 4.36
The algebraic equations are written in matrix notation of the form
AX = λBx 4.37
Therefore, (A – λB) x = 0 4.38
Where x = {xi} is a column matrix whose element xi represent the amplitudes of the
free vibrations; A = [aij] is a square matrix obtained from the new finite difference
expression; B = [bij] is a diagonal matrix representing the constant in the right hand
term and λ = ω2 is the natural frequency.
Pre-multiplying equation 4.27 by B-1, we obtain
(C – λI) x = 0 4.39
Where C = B-1A and I is an identity matrix.
Solving the matrices obtained the new finite difference method for free vibration
analysis; we obtain the Eigen values as shown in Appendix B:
4.2.1. CASE 1 (n = 3)
𝑙
For case 1 (n = 3), h = and the lowest values of the Eigen values are obtained to
4
calculate the required natural frequency; that is
I. PIN – ROLLER SUPPORT (P – R) :  = 4.3188

𝐴2 𝑙4 𝑙4
but  = 12ℎ  ; 4
 = ; and ℎ = 4 = 4
4.40
𝐸𝐼 4 256
𝑙 4
𝐴 2
This gives 4.3188 = 12. . 4.41
256 𝐸𝐼
73
4.3188 × 256 × 𝐸𝐼
That is 2 = 4.42
12 × 𝑙 4 × 𝐴
9.5987 𝐸𝐼
 = √ 4.43
𝑙2 𝐴
II. CLAMPED – CLAMPED SUPPORT(C – C):  = 17.7178

𝑙 4 𝐴2
This gives 17.7178 = 12 . . 4.44
256 𝐸𝐼
17.7178 × 256 × 𝐸𝐼
That is 2 = 4.45
12 × 𝑙 4 × 𝐴
19.4417 𝐸𝐼
 = √ 4.46
𝑙2 𝐴
III. CLAMPED– ROLLER SUPPORT:  = 9.533

𝑙 4 𝐴2
This gives 9.533 = 12 . . 4.47
256 𝐸𝐼
9.533 × 256 × 𝐸𝐼
That is 2 = 4.48
12 × 𝑙 4 × 𝐴
14.2609 𝐸𝐼
 = √ 4.49
𝑙2 𝐴
4.2.2. CASE 2 (n = 5).

𝑙
6
I. PIN – ROLLER SUPPORT (P – R):  = 0.8808

𝑙 4 𝐴2
This gives 0.8808 = 12. 4 . 4.50
6 𝐸𝐼
0.8808 × 1296 × 𝐸𝐼
That is 2 = 4.51
12 × 𝑙 4 × 𝐴
9.7533 𝐸𝐼
 = √ 4.52
𝑙2 𝐴
74
II. CLAMPED – CLAMPED SUPPORT:  = 4.0992
This gives 4.0992
𝑙 4 𝐴2
= 12 . . 4.53
1296 𝐸𝐼
4.0992 × 1296 × 𝐸𝐼
That is 2 = 4.54
12 × 𝑙 4 × 𝐴
21.0408 𝐸𝐼
 = √ 4.55
𝑙2 𝐴
III. CLAMPED – ROLLER SUPPORT:  = 2.0582

This gives 2.0582
𝑙 4 𝐴2
= 12 . . 4.56
1296 𝐸𝐼
2.0582 × 1296 × 𝐸𝐼
That is 2 = 4.57
12 × 𝑙 4 × 𝐴
14.29092 𝐸𝐼
 = √ 4.58
𝑙2 𝐴
4.2.3. CASE 3 (n = 7)
𝑙
8
I. PIN – ROLLER SUPPORT:  = 0.2817

𝑙 4 𝐴2
This gives 0.2817 = 12. 4 . 4.59
8 𝐸𝐼
0.2817 × 4096 × 𝐸𝐼
That is 2 = 4.60
12 × 𝑙 4 × 𝐴
9.8058 𝐸𝐼
 = √ 4.61
𝑙2 𝐴
II. CLAMPED – CLAMPED SUPPORT:  = 1.3699
75
𝑙4 𝐴2
This gives 1.3699 = 12 . . 4.62
4096 𝐸𝐼
1.3699 × 4096 × 𝐸𝐼
That is 2 = 4.63
12 × 𝑙 4 × 𝐴
21.6239 𝐸𝐼
 = √ 4.64
𝑙2 𝐴
III. CLAMPED – ROLLER SUPPORT:  = 0.6711

𝑙4 𝐴2
This gives 0.6711 = 12 . . 4.65
4096 𝐸𝐼
0.6711 × 4096 × 𝐸𝐼
That is 2 = 4.66
12 × 𝑙 4 × 𝐴
15.135 𝐸𝐼
 = √ 4.67
𝑙2 𝐴
4.3. DISCUSSION OF RESULTS.

4.3.1. STABILITY ANALYSIS OF LINE CONTINNUM.
The Tables below summarize the results of the critical buckling load obtained from
Stability analysis of line continuum.
Table 4.1. Result Data for P – R Support and Exact Result.

Exact Result Case 1 (n = 3) Case 2 (n = 5) Case 3 (n = 7)
Values 𝐸𝐼 𝐸𝐼 𝐸𝐼 𝐸𝐼
9. 8696 9.3728 9.6444 9.7408
𝑙2 𝑙2 𝑙2 𝑙2
% Difference -5.03% -2.29% -1.31%
Table 4.2. Result Data for C – C Support and Exact Result.

Exact Result Case 1 (n =3) Case 2 (n = 5) Case 3 (n = 7)
39. 4784 32 36.9036 37.4912
𝑙2 𝑙2 𝑙2 𝑙2
% Difference -18.94% -6.52% -5.03%
76
Table 4.3. Result Data for C – R Support and Exact Result.
20.191 17.9312 19.1664 19.6096
𝑙2 𝑙2 𝑙2 𝑙2
% Difference -11.19% -5.07% -2.88%
The result of the stability analysis of the line continua of the three boundary
conditions using the new finite difference method are presented in Tables 4.1, 4.2
and 4.3. For verification of purpose, the accuracy of results obtained by the present
research study was compared with the exact result obtained from previous analysis.
From Table 4.1, it can be seen from the analysis of P – R boundary conditions that
the percentage difference ranges from -5.03% to -1.31%. Result emerging from C –
C boundary condition as shown in Table 4.2 shows a percentage difference ranging
from -18.94% to -5.03% while that of C – R boundary condition ranges from -
11.19% to -2.88% as shown in Table 4.3. All these differences are small and within
the range of acceptance in statistics.
4.3.2. VIBRATION ANALYSIS OF LINE CONTINUUM.
The Tables below summarize the results of the natural frequencies obtained from
the free – vibration analysis of line continuum.
Table 4.4. Result Data for P – R Support and Exact Result.
PREVIOUS STUDY PRESENT STUDY
Exact Result (ω) Case 1 (n = 3) Case 2 (n = 5) Case 3 (n = 7)
Value
9.8967 𝐸𝐼 9.5987 𝐸𝐼 9.7533 𝐸𝐼 9.8058 𝐸𝐼
√ √ √ √
𝑙2 𝜌𝐴 𝑙2 𝜌𝐴 𝑙2 𝜌𝐴 𝑙2 𝜌𝐴
% -2.75% -1.18% -0.65%

Difference
Table 4.5. Result Data for C – C Support and Exact Result.

77
Exact Result (ω) Case 1 (n = 3) Case 2 (n = 5) Case 3 (n = 7)
Value
22.3733 𝐸𝐼 19.4417 𝐸𝐼 21.0408 𝐸𝐼 21.6239 𝐸𝐼
√ √ √ √
% -13.10% -5.96% -3.35%

Difference
Table 4.6. Result Data for C – R Support and Exact Result.


(ω)
Value
15.4182 𝐸𝐼 14.2609 𝐸𝐼 14.29092 𝐸𝐼 15.135 𝐸𝐼
√ √ √ √
% -7.51% -7.31% -1.84%

Difference
The result of the free-vibration analysis of line continua of the three boundary
conditions using the new finite difference method are presented in Tables 4.4, 4.5,
and 4.6. Also, values of exact result from previous analysis were presented for
comparison. It can be seen in Table 4.4 for the analysis of P – R boundary
conditions that the percentage difference ranges from -2.75% to -1.18%. Also
results emerging from analysis of C – C boundary shows a percentage difference
ranging from -13.10% to -3.35% as shown in Table 4.5 while result gotten from the
analysis of C – R boundary condition reveals percentage difference from -7.51% to
-1.84% as shown in Table 4.6.
The percentage differences obtained from this present study gave negative
differences which depicts that Finite Difference Approach provides lower bound
solution and hence, they are within the range of acceptance in statistics. Also, the
results obtained from this present study indicates that the higher the number of
nodes, the greater the accuracy of result as they tend to converge. Thus, the present
78
research study gives a close approximation to the exact solution gotten from
previous analysis.
CHAPTER FIVE
CONCLUSIONS AND RECOMMENDATION
5.1 CONCLUSIONS.
Based on the results of this study, the following conclusions were drawn.
I. The Finite Difference Approach used in this research study is simple and
straight forward. Thus, this approach gives straight forward equations for
the stability and free vibration analysis of line continuum.
II. The Finite Difference pattern developed by the polynomial series,
truncated at fourth term is satisfactory for the various boundary conditions
used.
III. Critical stability load values obtained by this approach gave satisfactory
approximate values that are close to the exact values.
IV. Natural frequency values obtained from this new pattern are close to the
exact value and hence, are satisfactory for vibration analysis of line
continuum.
5.2. CONTRIBUTION TO KNOWLEDGE.
I. This research study provides a new finite difference pattern and approach
that is less rigorous in handling stability and free vibration analysis of line
continuum problems.
II. This research has exposed this finite difference technique as a veritable
tool in treating line continuum with different boundary conditions.
III. This research work will serve as a reference material for future studies in
the related field of study.
5.3. RECOMMENDATIONS.
79
The following recommendations were made:
I. Future studies should be encouraged to extend their research work in
curvilinear structural members such as shells using various boundary
conditions,
II. Future studies should be encouraged to develop pattern of this approach,
truncating the polynomial series at the sixth terms and eight terms.
REFERENCES
Aristizabal – Ochoa, J.D. (1993). Statics, Stability and Vibration of Non-Prismatic

Beams and Columns. Journal of Sound and Vibration, Vol. 162, No. 3, pp 441 –
455.
Aristizabal- Ochoa, J.D. (2004). Column Stability and Minimum Lateral Bracing:
Effects of Shear Deformations. Journal of Engineering Mechanics; 130 (10):1223-
32.
Aristizabal- Ochoa, J.D. (2008). Slope – Deflection Equations for Stability and
Second- Order Analysis of Timoshenko Beam – Column Structures with Semi-
Rigid Connections. Journal of Engineering Structures; 30 (2008): 2517-2527.
Bažant, Z.P. (2000). Structural Stability. International Journal of Solids and

Structures. 37, pp 55 – 67
Chajes, A. (1974). Principles of Structural Stability Theory. New Jersey: Prentice-

Hall. Inc. Eagle wood Cliffs.
Chakraverty, S. (2009). Vibration of Plates. USA: CRC Press, Taylor and Francis
Group
80
Chan, S.L. (2004). Stability of Structure with Application to Practical Design.
M.Sc. Lecture Note, Department of Civil and Structural Engineering, Hong Kong
Polytechnic
Chen, W. F. and Lui, E. M. (1987). Structural Stability: Theory and

Implementation. New York, El Sevier Science Publishing Co. Inc.
Chondoras, T.G., et al. (1998). A Continuous Cracked Beam Vibrations Theory.

Journal of Sound and Vibration, Vol.215, No1, pp 17-34.
Chopade, J. P. and Barjibhe, R. B. (2013). Free Vibration Analysis of Fixed Free

Beam Theoretical and Numerical Approach Method. International Journal of
Innovation in Engineering and Technology (IJIET).Vol. 2, Issue 1, pp 352 – 356.
Chopra, A. K. (1995). Dynamics of Structures. New Jersey: Prentice-Hall, Inc.
Demirdag, O. and Murat, Y.S. (2009). Free Vibration Analysis of Elastically

Supported Timoshenko Columns with Attached Masses using Fuzzy Neural
Network. Journal of Scientific and Industrial Research. Vol. 68, 285-291
Ezeh, J. C., Ibearugbulem, O. M. and Onyechere, C. I. (2013). Free Vibration

Analysis of Thin Rectangular Flat Plates using Ordinary Finite Different Method.
SAVAP International Journal of Natural and Applied Sciences.Vol. 4, No. 2. pp
187 – 192.
Galambos, T. V. and Surovek, A. E. (2008). Structural Stability of Steel: Concepts

& Applications for Structural Engineers. New Jersey. John Wiley and Sons Inc.
Greenwood, D.T. (1961). The Use of Higher- Order Difference Methods in Beams
Vibration Analysis. Washington: National Aeronautic and Space Administration
(NASA). Technical Note. D- 964
81
Gupta, A.P. and Sharma, N. (2003). Free Vibration of a Rectangular Beam with
Two Linear Vibrations in Thickness. Indian Journal of Pure Applied Math. 34(1):
3-15
Hamoush, S. A., Terro, M.J. and McGinley, W.M. (2002). Elastic and Inelastic
Analysis of Non- Prismatic Members using Finite Difference. Kuwait Journal of
science and Engineering; 29(2).
Ibearugbulem, O. M., Ettu, L.O., Ezeh, J. C. and Anya, U.C. (2013). A New Shape
Function for Analysis of Line Continuum by Direct Variational Calculus.
International Journal of Engineering and Science (IJES). Vol. 2, Issue 3 pp 66 – 71.
Ibearugbulem, O. M., Ezeh, J. C. and Okpara, U.O. (2014). Fourth-Order Improved

Finite Difference Approach to Pure Bending Analysis of Line Continuum.
International Journal of Research in Engineering and Technology (IJRET).
Volume: 03 Issue 01; ISSN: 2319-1163
Iremonger, M. J. (1980). Finite Difference Buckling Analysis of Non-uniform

Columns. Computers and Structures.Vol. 12, pp 741 – 748
Iyengar, N. G. R (1988). Structural Stability of Columns and Plates. UK: Ellis

Horwood Limited, England
Javidinejad, A. (2012). Buckling of Beams and Columns under Combined Axial

and Horizontal Loading with Various Axial Loading Application Locations. Journal
of Theoretical and Applied Mechanics; Vol. 42, No 4, pp 19-30.
Kocatürk, T. and Şimşek, M. (2005). Free Vibration Analysis of Timoshenko

Beams under Various Boundary Conditions. Journal of Engineering and Natural
Science.Sigma 2005/1, pp 30-44
Lake, M. S. and Mikulas- Jr., M. M. (1991). Buckling and Vibration Analysis of

Simply Supported Column with A Piecewise Constant Cross section. NASA
Technical paper 3090 (1991). Springfield, Virginia.
82
Majkut, L. (2009). Free and Forced Vibrations of Timoshenko Beams Described by
Single Different Equation. Journal of Theoretical and Applied Mechanics (47), 1,
pp 193 – 210
Mao, Q. (2012). Free Vibration Analysis of Uniform Beams with Arbitrary number
of Cracks by using Adomian Decomposition Method. World Applied Science
Journal 19(12): 1721-1730.
Mohammadi, S. and Hassanirad, A. (2012). Applied and Theoretical Cantilever

Beam Free Vibration Analysis. World Academy Science, Engineering and
Technology.
Pinarbasi, S. (2012). Stability Analysis of Non-uniform Rectangular Beams using

Homotopy Perturbation Method. Mathematical Problems in Engineering, Hindawi
publishing corporation, Article ID 197483
Ranjbaran, A., Hashemi, S. and Ghaffarian, A.R (2008). A New Approach for
Buckling and Analysis of Cracked Column. IJE Transactions A: Basics, Vol.21,
No. 3, pp 225-230.
Reddy, J. N., Wang, C. M. and Wang, C. Y. (2005). Exact Solutions for Buckling
Structural members. USA: CRC press,
Saha, G. and Banu, S. (2007). Buckling Load of a Beam-Column for Different End
Conditions using Multi-Segment Integration Technique. ARPN Journal of
Engineering and Applied Sciences, Vol. 2, No. 1, pp 27 – 32
Šapalas, V., Samofalov, M. and Šaraškinas, V. (2005). FEM Stability Analysis of

Tapered Beam-Columns. Journal of Civil Engineering and Management, Vol. XI,
No. 3, 211 – 216
Şimşek, M. and Kocatürk, T. (2006). Free Vibration Analysis of Beams by using

Third-Order Shear Deformation Theory. Sadhana Journal, Vol. 32, Part 3, pp 167 –
179
Simitses, G. J. and Hodges, D. H. (2006). Fundamentals of Structural Stability.

USA: Elsevier Inc.
83
Szilard, R. (2004). Theories and Applications of Plate Analysis. Classical,
Numerical and Engineering Methods. New Jersey: John Wiley & Sons Inc.
Timoshenko, S. P. and Gere, J. M. (1963). Theory of Elastic Stability. Singapore:

McGraw Hill Book Company
Thorby, D. (2008). Structural Dynamics and Vibration in Practice: An Engineering

Handbook. U. K: Butterworth-Heinemann Publication.
Yoo, C. H. and Lee, S. C (2011). Stability of Structures: Principles and Application.

Butterworth- Heinemann.
APPENDIX A. SOLUTIONS TO MATRICES FOR STABILITY ANALYSIS.
4.1.1. CASE 1 ( n = 3)
B= -29 16 -1 B-1 = 0.0565 0.041 -0.021

16 -30 16 0.0412 0.077 -0.041
-1 16 -29 0.0208 0.041 -0.057
I. P - R SUPPORT.
A= 74 -62 18 (-A) = -74 62 -18
-62 92 -62 62 -92 62
18 -62 74 -18 62 -74
C = (-A)B-1 = Eigen values =
2 -1 0 0.5858
-1 2 -1 2
2.2E-16 -1 2 3.4142
II. C - C SUPPORT.
A= 110 -64 18 (-A) = -110 64 -18
-64 92 -64 64 -92 64
18 -64 110 -18 64 -110
84
B= -31 16 -1 B-1 = -0.05 0.0357 0.017
16 -30 16 -0.04 0.0714 0.036
-1 16 -31 -0.02 0.0357 -0.05
C = (-A)B-1
3.53333 -1 0.47 Eigen values = 4
0 2 0 3.0667
0.46667 -1 3.53 2
III. C - R SUPPORT.
A= 110 -64 18 (-A) = -110 64 -18
-64 92 -62 64 -92 62
18 -62 74 -18 62 -74
B= -31 16 -1 B-1 = -0.05 -0.037 0.019

16 -30 16 -0.04 0.0743 -0.04
-1 16 -29 -0.02 0.0397 0.056
C = (-A)B-1 Eigen values =

3.55078 -1 0 1.1207
0.03705 2 -1 3.7651
0.51869 -1 2 2.665
85
4.1.2. CASE 2 (n = 5)
I. P - R SUPPORT.
B= -29 16 -1 0 0
16 -30 16 -1 0
-1 16 -30 16 -1
0 -1 16 -30 16
0 0 -1 16 -29
B-1 = 0.0635 0.0551 0.0416 0.0278 0.0139
0.0551 0.1051 0.0829 0.0555 0.0278
0.0416 0.0829 -0.119 0.0829 0.0416
0.0278 0.0555 0.0829 0.1051 0.0551
0.0139 0.0278 0.0416 0.0551 0.0635
A= 74 -62 18 -1 0
-62 92 -63 18 -1
18 -63 92 -63 18
-1 18 -63 92 -62
0 -1 18 -62 74
(-A) = -74 62 -18 1 0
62 -92 63 -18 1
-18 63 -92 63 -18
1 -18 63 -92 62
0 1 -18 62 -74
86
C = (-A)B-1= Eigen values =
2 -1 -6E-16 -3E-16 -1E-16 3.7321
-1 2 -1 3E-16 2E-16 0.2679
6.1E-16 -1 2 -1 -2E-16 1
-4E-16 -4E-16 -1 2 -1 2
2.2E-16 0 -1E-15 -1 2 3
II. C - C SUPPORT.
A= 110 -64 18 -1 0
-64 92 -63 18 -1
18 -63 92 -63 18
-1 18 -63 92 -64
0 -1 18 -64 110
(-A) = -110 64 -18 1 0
64 -92 63 -18 1
-18 63 -92 63 -18
1 -18 63 -92 64
0 1 -18 64 -110
B= -29 16 -1 0 0
16 -30 16 -1 0
-1 16 -30 16 -1
0 -1 16 -30 16
0 0 -1 16 -29
B-1 -0.0635 0.0551 0.0416 0.0278 0.0139
-0.0551 0.1051 0.0829 0.0555 0.0278
-0.0416 0.0829 -0.119 0.0829 0.0416
-0.0278 0.0555 0.0829 0.1051 0.0551
-0.0139 0.0278 0.0416 0.0551 0.0635
C=
4.17436 1.1269 -2E-16 0.0278 0.4444 Eigen values = 1.0251
0.77434 1.8897 -1 0.0556 0.8889 4.2392
1.33309 1.0833 2 1.0833 1.3331 3.7979
0.88887 0.0556 -1 1.8897 0.7743 3.1887
0.44444 0.0278 2E-16 1.1269 4.1744 1.8773
87
III. C - R SUPPORT.
A= 110 -64 18 -1 0
-64 92 -63 18 -1
18 -63 92 -63 18
-1 18 -63 92 -62
0 -1 18 -62 74
(-A) = -110 64 -18 1 0
64 -92 63 -18 1
-18 63 -92 63 -18
1 -18 63 -92 62
0 1 -18 62 -74
B= -31 16 -1 0 0
16 -30 16 -1 0
-1 16 -30 16 -1
0 -1 16 -30 16
0 0 -1 16 -29
B-1 = -0.0563 0.0489 0.0369 0.0246 0.0123
-0.0489 0.0997 0.0788 0.0528 0.0264
-0.0369 0.0788 0.1159 0.0808 0.0406
-0.0246 0.0528 0.0808 0.1037 0.0544
-0.0123 0.0264 0.0406 0.0544 0.0631
C= Eigen values =
3.70421 -1 -1E-15 2E-16 3E-16 0.5324
0.36594 2 -1 -4E-16 4E-16 3.8955
1.02463 -1 2 -1 1E-15 3.3717
0.6831 9E-16 -1 2 -1 2.4718
0.34155 3E-16 -1E-15 -1 2 1.4328
88
4.1.3. CASE 3 (n = 7)
I. P - R SUPPORT.
A= 74 -62 18 -1 0 0 0
-62 92 -63 18 -1 0 0
18 -63 92 -63 18 -1 0
-1 18 -63 92 -63 18 -1
0 -1 18 -63 92 -63 18
0 0 -1 18 -63 92 -62
0 0 0 -1 18 -62 74
(-A) = -74 62 -18 1 0 0 0
62 -92 63 -18 1 0 0
-18 63 -92 63 -18 1 0
1 -18 63 -92 63 -18 1
0 1 -18 63 -92 63 -18
0 0 1 -18 63 -92 62
0 0 0 1 -18 62 -74
B= -29 16 -1 0 0 0 0
16 -3 16 -1 0 0 0
-1 16 -30 16 -1 0 0
0 -1 16 -30 16 -1 0
0 0 -1 16 -30 16 -1
89
0 0 0 -1 16 -30 16
0 0 0 0 -1 16 -29
B-1 = -0.0199 0.0281 0.0265 0.0214 0.0161 0.011 0.00536
0.02805 0.0538 0.0469 0.0376 0.0282 0.019 0.00942
0.02652 0.0469 0.0189 0.0191 0.0146 -0.01 -0.0049
0.02143 0.0376 0.0191 -0.076 -0.061 -0.04 -0.0205
0.01609 0.0282 0.0146 -0.061 0.1026 -0.07 -0.0362
0.01073 0.0188 0.0098 0.0409 -0.072 -0.1 -0.0515
0.00536 0.0094 0.0049 0.0205 0.0362 -0.05 -0.0616
C = (-A)B-1 =
2.75743 2.5149 0.7574 9E-17 3E-16 -0 2.8E-16
0.45195 0.9039 0.452 3E-16 6E-16 -0 0
1.26585 3.5317 3.2658 -1 -6E-16 3E-16 -2E-16
1.01651 -2.033 0.0165 2 -1 2E-15 -7E-16
0.76264 1.5253 0.7626 -1 2 -1 -1E-15
0.50844 1.0169 0.5084 2E-16 -1 2 -1
0.25422 0.5084 0.2542 -2E-16 2E-16 -1 2
Eigen values =
3.8478
3.4142
2.7654
2
0.1522
0.5858
1.2346
II. C - C SUPPORT.
A= 110 -64 18 -1 0 0 0
-64 92 -63 18 -1 0 0
18 -63 92 -63 18 -1 0
-1 18 -63 92 -63 18 -1
0 -1 18 -63 92 -63 18
0 0 -1 18 -63 92 -64
0 0 0 -1 18 -64 110
(-A) = -110 64 -18 1 0 0 0
64 -92 63 -18 1 0 0
-18 63 -92 63 -18 1 0
1 -18 63 -92 63 -18 1
0 1 -18 63 -92 63 -18
90
0 0 1 -18 63 -92 64
0 0 0 1 -18 64 -110
B= -31 16 -1 0 0 0 0
16 -30 16 -1 0 0 0
-1 16 -30 16 -1 0 0
0 -1 16 -30 16 -1 0
0 0 -1 16 -30 16 -1
0 0 0 -1 16 -30 16
0 0 0 0 -1 16 -31
B-1 = -0.0589 -0.0544 -0.0454 -0.0361 -0.0267 -0.02 -0.0081
-0.0544 0.1115 -0.097 0.0773 0.0573 -0.04 -0.0174
-0.0454 -0.097 0.1438 0.1186 0.0881 -0.06 -0.0267
-0.0361 0.0773 0.1186 0.1546 0.1186 -0.08 -0.0361
-0.0267 0.0573 0.0881 0.1186 0.1438 -0.1 -0.0454
-0.0174 0.0372 0.0573 0.0773 -0.097 -0.11 -0.0544
-0.0081 0.0174 0.0267 0.0361 0.0454 -0.05 -0.0589
C = (-A)B-1 =
3.77539 -1 7E-16 -1E-15 5E-16 -0 0.22461
0.51852 2 -1 -1E-15 2E-16 -0 0.48148
1.25935 -1 2 -1 1E-15 -0 0.74065
1 -9E-16 -1 2 -1 -0 1
0.74065 6E-16 -1E-15 -1 2 -1 1.25935
0.48148 8E-16 -2E-15 2E-15 -1 2 0.51852
0.22461 0 -3E-16 7E-17 2E-16 -1 3.77539
Eigen values = 0.5858
1.1207
2.665
4
3.7651
3.4142
2
III. C - R SUPPORT.
A= 110 -64 18 -1 0 0 0
-64 92 -63 18 -1 0 0
18 -63 92 -63 18 -1 0
-1 18 -63 92 -63 18 -1
0 -1 18 -63 92 -63 18
0 0 -1 18 -63 92 -62
91
0 0 0 -1 18 -62 74
(-A) = -110 64 -18 1 0 0 0
64 -92 63 -18 1 0 0
-18 63 -92 63 -18 1 0
1 -18 63 -92 63 -18 1
0 1 -18 63 -92 63 -18
0 0 1 -18 63 -92 62
0 0 0 1 -18 62 -74
B= -31 16 -1 0 0 0 0
16 -30 16 -1 0 0 0
-1 16 -30 16 -1 0 0
0 -1 16 -30 16 -1 0
0 0 -1 16 -30 16 -1
0 0 0 -1 16 -30 16
0 0 0 0 -1 16 29
B-1 = -0.0573 -0.0511 -0.0403 -0.0292 -0.018 -0.01 0.0032
-0.0511 0.1044 -0.086 0.0625 0.0386 -0.01 0.00686
-0.0403 -0.086 0.1269 0.0957 0.0593 -0.02 0.01055
-0.0292 0.0625 0.0957 0.1238 0.0797 -0.03 0.01425
-0.018 0.0386 0.0593 0.0797 -0.095 -0.04 0.01793
-0.0069 0.0148 0.0228 0.0308 0.0384 -0.04 0.02149
0.0032 0.0069 0.0106 0.0142 0.0179 0.021 0.02325
C = (-A)B-1 =
3.73227 0.4261 1.1171 0.808 0.499 0.192 -0.0887
-1 2 -1 2E-16 4E-17 -0 6.9E-18
-2E-16 -1 2 -1 2E-16 3E-16 -8E-17
1.3E-16 -3E-16 -1 2 -1 1E-16 2.7E-16
-4E-16 -1E-17 -2E-16 -1 2 -1 -2E-16
0.18559 0.3978 0.612 0.8262 0.0399 3.246 0.3483
-0.3712 0.7957 -1.224 1.6525 2.0799 -3.49 -0.6966
Eigen values =
-0.2192
3.9244
3.6085
0.62
1.2299
2.9194
2.1988
92
APPENDIX B. SOLUTIONS TO MATRICES FOR FREE VIBRATION
ANALYSIS
4.2.1. CASE 1 (n = 3)
B= 1 0 0
0 1 0
0 0 1
B-1 = 1 0 0
0 1 0
0 0 1
I. P - R SUPPORT.
A AB-1
74 -62 18 74 -62 18
-62 92 -62 -62 92 -62
18 -62 74 18 -62 74
Eigen values =
4.3188
56
179.68
II. C - C SUPPORT.
A C = AB-1
93
110 -64 18 110 -64 18
-64 92 -64 -64 92 -64
18 -64 110 18 -64 110
Eigen values =
17.718
92
202.28
III. C - R SUPPORT.
A C = AB-1
110 -64 18 110 -64 18
-64 92 -62 -64 92 -62
18 -62 74 18 -62 74
Eigen values =
9.533
74.179
192.29
4.2.2. CASE 2 (n = 5)
B-1 = 1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
0 0 0 1 0
0 0 0 0 1
I. P - R SUPPORT.
A
74 -62 18 -1 0
-62 92 -63 18 -1
18 -63 92 -63 18
-1 18 -63 92 -62
0 -1 18 -62 74
C = AB-1 Eigenvalues =
74 -62 18 -1 0
-62 92 -63 18 -1 0.8808
18 -63 92 -63 18 13
-1 18 -63 92 -62 56
0 -1 18 -62 74 135
94
II. C - C SUPPORT. 219.12
A
110 -64 18 -1 0
-64 92 -63 18 -1
18 -63 92 -63 18
-1 18 -63 92 -64
0 -1 18 -64 110
110 -64 18 -1 0
-64 92 -63 18 -1 4.0992
18 -63 92 -63 18 26.479
-1 18 -63 92 -64 80.236
0 -1 18 -64 110 157.52
III. C - R SUPPORT. 227.67
A
110 -64 18 -1 0
-64 92 -63 18 -1
18 -63 92 -63 18
-1 18 -63 92 -62
0 -1 18 -62 74
110 -64 18 -1 0
-64 92 -63 18 -1 2.0582
18 -63 92 -63 18 19.149
-1 18 -63 92 -62 68.096
0 -1 18 -62 74 146.9
223.79
95
4.2.3. CASE 3 (n = 7)
B-1 = 1 0 0 0 0 0 0
0 1 0 0 0 0 0
0 0 1 0 0 0 0
0 0 0 1 0 0 0
0 0 0 0 1 0 0
0 0 0 0 0 1 0
0 0 0 0 0 0 1
I. P - R SUPPORT.
A
74 -62 18 -1 0 0 0
-62 92 -63 18 -1 0 0
18 -63 92 -63 18 -1 0
-1 18 -63 92 -63 18 -1
0 -1 18 -63 92 -63 18
0 0 -1 18 -63 92 -62
0 0 0 -1 18 -62 74
C = AB-1 Eigen values =

74 -62 18 -1 0 0 0
-62 92 -63 18 -1 0 0 0.2817
18 -63 92 -63 18 -1 0 4.3188
96
-1 18 -63 92 -63 18 -1 20.174
0 -1 18 -63 92 -63 18 56
0 0 -1 18 -63 92 -62 112.91
0 0 0 -1 18 -62 74 179.68
234.63
II. C - C SUPPORT. A
110 -64 18 -1 0 0 0
-64 92 -63 18 -1 0 0
18 -63 92 -63 18 -1 0
-1 18 -63 92 -63 18 -1
0 -1 18 -63 92 -63 18
0 0 -1 18 -63 92 -64
0 0 0 -1 18 -64 110
C =AB-1 Eigen values =

110 -64 18 -1 0 0 0
-64 92 -63 18 -1 0 0 1.3699
18 -63 92 -63 18 -1 0 9.533
-1 18 -63 92 -63 18 -1 32.301
0 -1 18 -63 92 -63 18 74.179
0 0 -1 18 -63 92 -64 131.69
0 0 0 -1 18 -64 110 192.29
238.64
III. C - R SUPPORT. C = AB-1

110 -64 18 -1 0 0 0
-64 92 -63 18 -1 0 0
18 -63 92 -63 18 -1 0
-1 18 -63 92 -63 18 -1
0 -1 18 -63 92 -63 18
0 0 -1 18 -63 92 -62
0 0 0 -1 18 -62 74
Eigen values =
0.6711
6.6066
25.922
97
65.054
122.61
186.36
236.79
98

Null 10

Uploaded by

Copyright:

Available Formats

Null 10

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Null 10

Uploaded by

Copyright:

Available Formats

FINITE DIFFERENCE APPROACH TO STABILITY

AND VIBRATION ANALYSIS OF LINE

UGOCHUKWU OBINNA OKPARA (B.ENG)

A THESIS SUBMITTED TO THE POSTGRADUATE SCHOOL,

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR

List of Nomenclature viii

CHAPTER ONE: INTRODUCTION 1

1.1. Background of Study 1

CHAPTER TWO: LITERATURE REVIEW 4

2.1. Historical Development. 4

CHAPTER THREE: METHODOLOGY 26

3.1. Finite Difference Formulation. 26

CHAPTER FOUR: RESULTS AND DISCUSSION 69

4.1. Results of Stability Analysis of Line Continuum 69

CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS 79

Table 3.1: Schematic representation of Fourth – order Finite Difference

Table 3.2: Line continua and their boundary Conditions 34

Table 4.1: Result Data for P - R support and Exact Result 76

Table 4.2: Result Data for C - C support and Exact Result 76

Table 4.3: Result Data for C - R support and Exact Result 76

Table 4.4: Result Data for P - R support and Exact Result 77

Table 4.5: Result Data for C - C support and Exact Result 77

Table 4.6: Result Data for C - R support and Exact Result 78

Keywords: Line Continuum, Finite Difference, Polynomial Series, Boundary

1.1. BACKGROUND OF STUDY.

1.2. STATEMENT OF PROBLEM.

1.3. OBJECTIVE OF STUDY.

1.5. SCOPE OF STUDY.

Also, the line continuum was studied under 3 cases:

2.1. HISTORICAL DEVELOPMENT.

A line continuum, in this context, refers to a one – dimensional field structural

An analysis and sample results of the lateral buckling and vibration of a

2.2 PREVIOUS WORKS ON STABILITY ANALYSIS OF LINE

Stability analysis began with Euler’s solution of buckling of an elastic column.

Iremonger (1980) determined buckling of non- uniform column by Finite

Pinarbasi (2011) analyzed buckling of beams using Homotopy Perburtion Method

In his article, Javidinejad (2012) examined the buckling behavior of an I- beam

2.3. PREVIOUS WORKS ON VIBRATION ANALYSIS OF LINE

Shifrin et al. (1999) proposed a technique for calculating natural frequencies of a

Demirdag and Murat (2009) presented a study on elastically supported Timoshenko

Mao (2012) employed the Adomian Decomposition Method (ADM) to investigate

Free vibrations characteristics of Timoshenko beams having different boundary

2.4. APPROACHES TO STABILITY AND VIBRATION ANALYSIS.

2.4.1. APPROACHES TO STABILITY ANALYSIS.

The concept of a critical load of an elastic structure at which equilibrium bifurcates

2.4.1.1. EQUILIBRIUM APPROACH (EULER METHOD).

The concept of stability is frequently explained considering equilibrium. It deals

Figure 2.1. An idealized Euler Column. (Iyengar, 1988)

The equilibrium equation was derived based on the following assumptions.

a. The line continuum is axially loaded.

Equation 2.4 is a linear homogeneous differential equation with constant

2.4.1.2 ENERGY APPROACH.

Energy methods are a convenient means of computing unknown force and

2.4.1.3. PRINCIPLE OF CONSERVATION OF ENERGY.

Equilibrium is established by satisfying the law of conservation of energy which

Change in external work, ΔW is given as

Corresponding increase in strain energy due to bending is given as

Trigonometric or polynomial functions are frequently used because of ease of

Solving the integrals, we obtain