Applied Mathematics, 2011, 2, 1327-1338

doi:10.4236/am.2011.211186 Published Online November 2011 (http://www.SciRP.org/journal/am)

An Introduction to Numerical Methods for the Solutions of

Partial Differential Equations
Manoj Kumar, Garima Mishra
Department of Mathematics, Motilal Nehru National Institute of Technology, Allahabad, India
E-mail: manoj@mnnit.ac.in, garima.iitg08@gmail.com
Received August 8, 2011; revised September 27, 2011; accepted October 5, 2011

Abstract
Partial differential equations arise in formulations of problems involving functions of several variables such
as the propagation of sound or heat, electrostatics, electrodynamics, fluid flow, and elasticity, etc. The present paper deals with a general introduction and classification of partial differential equations and the numerical methods available in the literature for the solution of partial differential equations.
Keywords: Partial Differential Equations, Eigenvalue, Finite Difference Method, Finite Volume Method,
Finite Element Method

1. Introduction
An equation involving derivatives or differentials of one
or more dependent variables with respect to one or more
independent variables is called a differential equation.
The study of differential equations is a wide field in pure
and applied mathematics, physics, meteorology, and engineering, etc. All of these disciplines are concerned with
the properties of differential equations of various types.
Pure mathematics focuses on the existence and uniqueness of solutions, while applied mathematics emphasizes the rigorous justification of the methods for approximating solutions. Differential equations play an important role in modeling virtually every physical, technical, or biological process, from celestial motion, to bridge design, and interactions between neurons. Differential equations which are used to solve real-life problems may not necessarily be directly solvable, that is,
do not have closed form solutions. Instead, solutions can
be approximated using numerical methods. Mathematicians also study weak solutions (relying on weak derivatives), which are types of solutions that do not have
to be differentiable everywhere. This extension is often
necessary for solutions to exist, and it also results in
more physically reasonable properties of solutions, such
as possible presence of shocks for equations of hyperbolic type.
The theory of differential equations is quite developed
and the methods used to study them vary significantly
with the type of the equation.
Copyright 2011 SciRes.

A differential equation involving derivatives with respect to single independent variable is called an ordinary
differential equation. In the simplest form, the dependent
variable is a real or complex valued function, but more
generally, it may be vector-valued or matrix-valued: this
corresponds to considering a system of ordinary differential equations for a single variable. Ordinary differential
equations are classified according to the order of the
highest derivative of the dependent variable with respect
to the independent variable appearing in the equation.
The most important cases for applications are first-order
and second-order differential equations. In the classical
literature, the distinction is also made between differential equations explicitly solved with respect to the
highest derivative and differential equations in an implicit form.
A differential equation involving partial derivatives
with respect to two or more independent variables is
called partial differential equation. The partial differential equations can also be classified on basis of highest
order derivative.
Some topics in differential geometry as minimal surfaces and imbedding problems, which give rise to the
Monge-Ampere equations, have stimulated the analysis
of partial differential equations, especially nonlinear equations. Moreover, the theory of systems of first order
partial differential equations has a significant interaction
with Lie theory and with the work of E. Cartan.
The development of partial differential equations in
the 18th and 19th century is given in Klines book [1].
AM

1328

M. KUMAR

Until the 1870, the study of partial differential equation

was mainly concerned with heuristic methods for finding
solutions of boundary value problems as well as explicit
solutions for particular problems (for an example, the
solution of Dirichlet boundary value problem u = 0,
u in 2 , introduced by Riemann).
Poincar [2] gave the first complete proof of the existence and uniqueness of a solution of the Laplace equation for any continuous Dirichlet boundary condition in
1890. In a fundamental paper of, Poincare [3] established
the existence of an infinite sequence of eigenvalues and
corresponding eigen-functions for the Laplace operator
under the Dirichlet boundary condition. Picard applied
the method of successive approximation to obtain solutions of nonlinear problems which were mild perturbations of uniquely solvable linear problems. The construction of elementary solutions and Greens functions for
general higher order linear elliptic operators was carried
through to the analytic case by E. E. Levi [4]. Up to
about 1920s solutions of partial differential equations
were generally understood to be classical solutions, that
is, C k for a differential operator of order k .
Keeping in view the requirement of the new researchers, the present paper describes the basic fundamentals
of partial differential equations which has been collected
from a large number of research articles published in reputed journals and literature available in the books with
the intension to provide all important relevant material in
a condense form related to partial differential equations
and numerical methods for their solutions. Also, since
analytical and computational solution of partial differential equations is the major concern from the early
years, this paper gives a small step towards the development of computational analysis of partial differential
equations, which have lot of utilization in the field of
science and engineering.

2. Classification of Partial Differential

Equations
Both ordinary and partial differential equations are broadly classified as linear and nonlinear. A linear partial differential equation is one in which all of the partial derivatives appears in linear form and none of the coefficients depends on the dependent variables. The coefficient may be function of the independent variables. A nonlinear partial differential equation can be described as a
partial differential equation involving nonlinear terms.

2.1. Types of Non-Linear Partial Differential

Equations
The non-linear partial differential equations describe many
Copyright 2011 SciRes.

ET AL.

different physical systems, ranging from gravitation to

fluid dynamics and have been used in mathematics to
solve problems such as Poincare conjecture and Calabi
conjecture.
A non-linear partial differential equation is known as
semi-linear if it is linear in highest order derivatives and
the coefficients of highest order derivatives depend only
on independent variables.

a D u a0 D k 1u, , Du, u, x = 0 .

(1)

| |= k

A nonlinear partial differential equation is known as

quasi-linear if it is linear in highest order derivatives and
the coefficients of highest order derivatives depend on
independent variables as well on lesser order derivatives.

a D k 1u, , Du, u, x D u

| |= k

(2)

a0 D k 1u , , Du, u, x = 0.

A nonlinear partial differential equation is known as

fully non-linear if the coefficients depends on dependent
variable or the derivatives appear in nonlinear form.
Example 2.1
f xx f yy f x f y = f xy is linear equation (3)
af xx bf yy f x2 f y = c is semilinear

(4)

f x f xx f y f yy f x2 f = 0 is quasilinear

(5)

ff xx f yy af x bf y = 0 is nonlinear

(6)

where a, b are functions of x, y and c is function of x, y

and f.
But further classification into elliptic, hyperbolic, and
parabolic equations, especially for second-order linear
equations, is of utmost importance. For more study on
linear and quasi linear elliptic equations, see [5,6].

2.2. Classification Based on Discriminant

The general quasi linear second-order non-homogeneous
partial differential equation in two independent variable
is
(7)
Af xx Bf xy Cf yy Df x Ef y Ff = G
The classification of partial differential equations depends on the sign of discriminant B 2 4 AC as fallow:
1) If B 2 4 AC > 0, the partial differential equation
is hyperbolic.
2) If B 2 4 AC = 0, the partial differential equation
is parabolic.
3) If B 2 4 AC < 0, the partial differential equation
is elliptic.
AM

M. KUMAR

where M is Mac number.

1) First we characterize by discriminant

2.3. Eigenvalue Based Classification of Partial

Differential Equations
Since method for classification of partial differential equations fails if it is partial differential equation in three or
more independent variable then we are not able to find
discriminant.
Let us consider a general second order partial equation
in n independent variables
a11u x1x1 a12 u x1x2 a13u x1x3

a1n u x1xn

a21u x2 x1 a22 u x2 x2 a23u x2 x3

a2 n u x2 xn

an1u xn x1 an 2 u xn x2 an3u xn x3

ann u xn xn

b1u x1

b2 u x2

bn u xn cu = 0.

b3u x3

B = 0, A = 1 M 2 , C = 1,
then B 2 4 AC = 4(1 M 2 ) = 4( M 2 1).

If flow is subsonic i.e. M < 1, then

B 2 4 AC < 0 .
The equation of flow is elliptic.
If flow is sonic i.e. M = 1, then
B 2 4 AC = 0 .
The equation of flow is parabolic.
If flow is subsonic i.e. M > 1, then
B 2 4 AC > 0 .
The equation of flow is hyperbolic.
2) Now by eigenvalues

(8)

1 M 2 0
A=

1
0

which also can be written in compact form as:

n

2u

aij x x
i =1 j =1

bi
j

i =1

u
cu = 0.
xi

(9)

Then coefficient matrix of highest order derivatives is

A = [aij ] .
Let be an eigenvalue of A corresponding to eigenvector X
AX = X
AX X = 0

(A I)X = 0
A I = 0

Since X 0

then
a11
a21

an1

a1n

a22

a2 n

a12

an 2

ann

= 0.

The characterisic equation will have n roots as n

eigenvalues.
Characterization of differential equation is based on
following:
If any eigenvalue is zero, then partial differential equation is parabolic.
If all eigenvalues are non-zero and one eigenvalue
has opposite sign, then partial differential equation is
hyperbolic.
If all eigenvalues are nonzero having same sign, then
partial differential equation is elliptic.
Example 2.2 Consider the following differential equation of flow:
2 2

1 M 2
= 0,
x 2 y 2

Copyright 2011 SciRes.

1329

ET AL.

A I =

1 M

= 1, 1 M 2 .

If M < 1, then all eigenvalues are nonzero and of

same sign thus equation of flow is elliptic.
If M = 1, then all eigenvalues are zero thus equation of flow is parabolic.
If M > 1, then all eigenvalues are nonzero and of
one eigenvalue of opposite sign thus equation of flow
is hyperbolic.

2.4. Significance of Classification

The classification of a partial differential equations is
intimately related to the characteristics of the partial
differential equations. The characteristics are n 1 dimensional hyper-surfaces in n-dimensional hyperspace that have some very special features. In two-dimensional space, which is the case considered generally,
characteristics are paths in the solution domain along
which information propagates. In other words we can say
information propagates throughout the solution domain
along characteristics.

2.5. Classification by Physical Problems

Physical problems falls into one of the following general
classification:
1) Equilibrium Problem;
2) Propagation Problem;
3) Eigenvalue Problem.
Equilibrium Problem: Equilibrium problems are
steady state problems in closed domain D( x, y ) in
which the solution f ( x, y ) is governed by an elliptic
AM

M. KUMAR

1330

partial differential equations subject to boundary conditions specified at each point on the boundary B of
the domain.
Such type of problems have no real characteristic and
thus the solution at every point in the solution domain is
influenced by the solution at all other points and the
solution at each points influence the solution at all other
points.
Equilibrium problems are solved by method of relaxation numerically.
Propagation Problem: Propagation problems are initial value problems in open domains. Here by open domain means that open with respect to one of the independent variables.
Example 2.3
ft = f xx with initial condition f ( x, t0 ) = f ( x).

Eigenvalue Problem: Eigenvalue problem are special

problems in which the solution exits only for special
values (i.e. eigenvalues) of a parameter of the problem.
The eigenvalues are to be determined in addition to the
corresponding configuration of the system.

Initial Value Problem: An initial value problem is one

in which the dependent variable and possibly its derivatives are specified initially (i.e. at time t = 0 ) or at the
same value of independent variable in the equation.
Initial value problems are generally time-dependent problems.
u
2u
= 2 , u ( x, t0 ) = f ( x).
Example 2.4
t
x
Boundary Value Problems: A boundary value problem is one in which the dependent variable and possibly
its derivatives are specified at the extreme of the independent variable. For steady state equilibrium problems, the auxiliary conditions consists of boundary conditions on the entire boundary of the closed solution
domain. There are three types of boundary condition.
Example 2.5 Let be a bounded domain in n
with a smooth boundary , and let f :
be a locally Hlder continuous function. The BVP

u = 0, on

(10)
(11)

is called nonlinear elliptic boundary value problem. This

type of BVP arises in several domains, for example in
physical problems involving the steady-state temperature distribution see [7-9].
1) Dirichlet boundary condition: The value of the
function f is specified on the boundary. The dependent variable of the partial differential equation are preCopyright 2011 SciRes.

scribed in domain at different points. For example if an

iron rod had one end held at absolute zero then the value
of the problem would be known at that point in space. A
Dirichlet boundary condition imposed on an ordinary or
a partial differential equation specifies the values of a
solution is to take on the boundary of the domain. The
question of finding solutions to such equations is known
as the Dirichlet problem as in (10).
2) Neumann boundary condition: The value of def
rivative normal to the boundary is specified (
is
n
specified on the boundary). For example if one iron rod
had heater at one end then energy would be added at a
constant rate but the actual temperature would not be
known. A Neumann boundary condition imposed on an
ordinary or a partial differential equation specifies the
derivative values of a solution is to take on the boundary
of the domain. For example
u = f ( x, u ),in
u
= 0, on .
x

3) Mixed boundary condition: The linear combination of Dirichlet and Neumann boundary conditions:

2.6. Types of Conditions

u = f ( x, u ),in

ET AL.

af b

f
n

is specified on the boundary.

Mixed boundary conditions are also known as Cauchy
boundary condition. A Cauchy boundary condition imposed on an ordinary or a partial differential equation
specifies both the values a solution of a differential equation is to take on the boundary of the domain and the
normal derivative at the boundary.
u = f ( x, u ),in
c1

u
c2 u = 0, on .
x

3. Various Methods for Solving Partial

Differential Equation
In literature various method exits for solution of partial
differential equations. Here we will discuss some of them
briefly as following:
1) Finite Difference Method: The finite difference
method is a numerical procedure which solves a partial
differential equation by discretizing the continuous physical domain into a discrete finite difference grid, approximating the individual exact partial derivatives in the
partial differential equations by algebraic finite difference approximations (i.e. FDA), substituting the FDAs
into the partial differential equations to obtain an algeAM

M. KUMAR

braic finite difference equation(FDE), and solving the

resulting algebraic finite difference equations for the
dependent variable.
For detailed study on finite difference method, we encourage to readers to consult the references [10-19].
2) Finite Volume Method: The finite volume method is a method for representing and evaluating partial
differential equations in the form of algebraic equations.
Similar to the finite difference method, values are calculated at discrete places on a meshed geometry. Finite
volume refers to the small volume surrounding each
node point on a mesh. In the finite volume method, volume integrals in a partial differential equation that contain
a divergence term are converted to surface integrals, using the divergence theorem. These terms are then evaluated as fluxes at the surfaces of each finite volume.
Because the flux entering a given volume is identical to
that leaving the adjacent volume, these methods are conservative. The method is used in many computational
fluid dynamics packages
One advantage of the finite volume method over finite
difference methods is that it does not require a structured
mesh (although a structured mesh can also be used). Furthermore, the finite volume method is preferable to other
methods as a result of the fact that boundary conditions
can be applied non-invasively. This is true because the
values of the conserved variables are located within the
volume element, and not at nodes or surfaces. Finite volume methods are especially powerful on coarse nonuniform grids and in calculations where the mesh moves
to track interfaces or shocks.
For more details on finite volume method, we refer to
[20-22].
3) Finite Element Method: The finite element method, where functions are represented in terms of basis
functions and the partial differential equations is solved
in its integral (weak) form. In the finite element method
(FEM) the domain is partitioned in a finite set of
elements i , so that i j = for i j , and

i = . Usually one takes for i triangles or

quadrangles. Then the function is approximated by
uh = aii ( x), where i are functions that are polynomials on each i (i.e. piecewise polynomials). Usually the functions i are continuous polynomials of a
low degree. Further they are constructed so that their support extends only over a small number of elements.
Now we will give detail discussion of Finite element
method.

4. Finite Element Method

The key idea: The finite element is a numerical method
Copyright 2011 SciRes.

1331

ET AL.

like finite difference method but it is more general and

powerful in its application to real-world problems that
involve complicated physical geometry and boundary
conditions.
In FEM, a given domain is viewed as a collection of
sub-domains, and over each sub-domain the governing
equation is approximated by any of the traditional variational methods.
The main reason behind taking approximate solution
on a collection of sub-domains is the fact that it is easier
to represent a complicated function as a collection of
simple polynomials.
The method is characterized by three features:
1) The domain of the problem is represented by a collection of simple sub-domains called finite elements.
The collection of finite elements is called the finite element mesh.
2) Over each finite element, the physical process is
approximated by functions of the desired type and algebraic equations relating physical quantities at selective
points, called nodes of the element are developed.
3) The element equation are assembled using continuity and/or balance of physical quantities.
In FEM, we seek an approximation un of u in the form
n

j =1

j =1

u un = u j j c j j

(12)

where u j are the values of un at the element nodes

j are the interpolation function, c j are coefficients
that are not associated with nodes, and j are the
associated approximation functions. Direct substitution
of the such approximation into the governing differential
equation does not always result, for an arbitrary choice
of the data of the problem, in a necessary and sufficient
no. of equations for the under-determined coefficients
u j and c j Therefore a procedure whereby a necessary
and sufficient number of equations can be obtained is
needed.
There is only one method of finite element model of
the same problem. There can be more than one finite
element model of the same problem. The type of model
depends on the differential equations, method used to
derive the algebraic equations for the undetermined coefficients over an element, and nature of the approximations function used.

4.1. Variational Principles and Methods

The idea of using a variational formulation of a boundary
value problem for its numerical solution goes back to
Lord Rayleigh (1894,1896) and Ritz (1908), see, e.g.,
Kantorovich and Krylov [13]. In Ritzs approach the
approximation solution was sought as a finite linear comAM

1332

M. KUMAR

bination of functions such as, for instance, polynomial or

trigonometric polynomials. The use in this context of
continuous piecewise linear approximating function
based on triangulation adapted to the geometry of the
domain proposed by Courant (1943) in the paper based
on an address delivered to the American Mathematical
Society in 1941. Even though this idea had appeared
earlier, also in work by Courant himself (see Babuska
[23]), this is often thought as the starting point of the
finite element method, but the further development and
analysis of the method would occur much later.
Classical sense of the variational principle is to find
the extremum or the variables of the problem. The functional includes all the intrinsic features of the problem
such as the governing equations, boundary conditions,
and constraints condition, if any.
In solid and structural mechanics problems, the functional represents the total energy of the system and in
other problems it is simply an integral representation of
the governing equations.
First many problems of mechanics are posed in terms
of finding the extremum and thus by their nature, can be
formulated in terms of variational statement. Second,
there are problems that can be formulated by other means,
but these can also be formulated by means of variational
principles. Third, variational formulations form a powerful basis for obtaining approximate solutions to practical
problems, many of which are intractable otherwise. The
principle of minimum total potential energy, for example,
can be regarded as a substitute to the equations of equilibrium of an elastic body as well as a basis for the development of displacement finite element models that
can be used to determine approximate displacement and
stress fields in the body.

4.2. Variational Formulations

The classical sense of the phrase Variational Formulation refers to the construction of a functional or variational principle that is equivalent to the governing equations of the problem. The modern use of the phrase
refers to the formulation in which the governing equations are translated into equivalent weighted integral
statements that are not necessarily equivalent to a variational principle.
The importance of variational formulation of physical
laws, in the modern or general sense of the phrase, goes
for beyond its use as simply an alternate to other formulation. In fact, variational forms of the laws of continuum
physics may be only natural and rigorously correct way
to think of them. While all sufficiently smooth fields
leads to meaningful variational forms the converse is not
true. There exits physical phenomena which can be adeCopyright 2011 SciRes.

ET AL.

quately modeled mathematically only in a variational

setting they are nonsensical when viewed locally.
The starting point for the discussion of the finite
element method is differential equations governing the
physical phenomena under study. As such, we shall first
discuss why integral statement of differential equations
are needed.
4.2.1. Need for Weighted-Integral Statements
The weighted-integral statement are required in order to
generate the necessary and sufficient number of algebraic
equations to solve for the parameters c j in Equation
(13) of approximate solution.
N

u ( x) u N = c j j .

(13)

j =1

The use of integral statement is equivalent to the

governing differential equation is necessitated by the fact
that substitution of Equation (13) into the governing
differential does not always results in the required
number of linearly independent algebraic equations for
the unknown coefficients c j . One way to insure that
there are exactly the same number n of equations as
there are unknowns is to require weighted integrals of the
error in the equation to be zero. We can require the
approximate solution u to satisfy the given differential
equation in the weighted integral sense,

w Rdx = 0,

(14)

where R is called residual.

4.2.2. Linear and Bilinear Functional
Consider the integral expression
b

I (u ) = a F ( x, u , u )dx, u = u ( x), u =

du
.
dx

(15)

For a given real function u = u ( x), I (u ) is a real

number. Therefore, I can be viewed as an operator that
transforms functions u ( x) into real numbers, and such
operators called functionals.
Example 4.1
b
du

I (u ) = a P ( x ) q ( x)u 2 Pu (a ).
d
x

(16)

du dv

q( x, y )v dxdy Q u ds.
I (u , v) = P( x, y )
d
d
x
x

(17)
A functional l (u ) is said to be linear in u if and only
if it satisfies the relation
l ( u v) = l (u ) l (v).

(18)

for any real numbers and and dependent variable u and v.

AM

M. KUMAR

Example 4.2

4.2.4. Natural and Essential Boundary Conditions

Essential boundary conditions which require v and possible its derivatives to vanish at the boundary. Thus, we
have

1) I (u ) = a f ( x)udx q( x)u (b).

2) I (u , v) =

f ( x, y)u q( x, y)v dxdy.

A function B(u , v) is said to be bilinear if it is linear

in each of its arguments u and v :
B u1 u2 , v = B(u1 , v) B (u2 , v)
linear in first argument.
B u, v1 v2 = B(u, v1 ) B (u, v2 )
linear in second argument.
A bilinear form B(u, v) is said to be symmetric in its
arguments u and v if
B(u , v) = B(v, u ), for all u , v.

(19)

4.2.3. Variational Operator and First Variation

Consider the function F ( x, u , u ). For any fixed value
of the independent variable x, F depends on u and
u . The change v in u where is a constant quantity and v is a function of x is called the variation of
u and is defined by u

u = v.

(20)

is called variational operator.

The variation u of a function u represents an admissible change in the function u ( x) at a fixed value of
the independent variable x. If u is specified value
cannot be varied. Thus, the variation of a function u is
zero there because the specified on the boundary, the
variation of u is zero there because the specified value
cannot be varied. Thus, the variation of a function u
should satisfy the homogeneous form of the boundary
conditions for u
F = F ( x, u v, u v) .

(21)

Fundamental lemma of variational calculus: The

fundamental lemma of variations can be stated as follows
for any integrable function G ( x), if the statement
b

a G( x) ( x)dx = 0,

(22)

holds for any arbitrary function ( x), x (a, b), then

it follows that G ( x) = 0 in (a, b).
General statement of fundamental lemma is as following:
If ( x) is arbitrary in a < x < b and (a) is arbitrary then the statement
b

a G ( x)dx B(a) (a) = 0

G = 0 in a < x < b and B (a ) = 0
because ( x) is independent of (a ).
Copyright 2011 SciRes.

1333

ET AL.

(23)

Specify v = 0, or u = u on the boundary.

(24)

These conditions are also known as Dirichlet or geometric boundary conditions.

Natural boundary condition, which require specification of the coefficient of v (and possibly its derivatives). Thus we have
Specify

F
= Q, on the boundary.
u

(25)

Natural boundary conditions are also known as Neumann boundary conditions.

Example 4.3 Finding a function u = u ( x) such that
u (a) = ua , u (b) = ub and
b

I (u ) = a F ( x, u ( x), u ( x))dx is a minimum

(26)

F d F

= 0 in a < x < b.
u dx u

(27)

F ( x, u , u ) =

The necessary condition for I to attain a minimum

yields
b

d F

a v u dx u dx u v

Qa v(a ) Qb v(b) = 0.
a

(28)
F
and v are selected such that
Now suppose that
u
F

Qa v = 0 for x = a, Qb v = 0 for x = b.
u
u

(29)
Then using the fundamental lemma of the calculus of
variations, we obtain the same Euler equation.
Equations in (29) are satisfied identically for any of
the following combination:
1) v(a ) = 0, v(b) = 0.

2) v(a ) = 0,

F
u

b
a

Qb = 0.

F
a Qa = 0, v (b) = 0.
u
F
F
4)
a Qa = 0,
b Qb = 0.
u
u
Consider the problem of finding (u, v), defined on a
two dimensional region such that the following functional is to be minimized:
3)

I (u , v) = F ( x, y, u, v, u x , vx , u y , v y )dxdy

(30)

with condition
AM

M. KUMAR

1334

u = u so that u = 0 on
v = v so that v = 0 on
F
F
x
y = 0 on
u x
u y
F
F
x
y = 0 on .
vx
v y

(31)
(32)
(33)

Since B (, ) is linear in u , we have

N

B(i , j )c j = l (i ) B(i , j )
j =1

(34)

Equations (31) and (32) represent the essential boundary condition and Equations (33) and (34) represent
the natural boundary conditions. The pair of elements
(u,v) are called the primary variables and
Qx =

ET AL.

F
F
F
F
x y and Qy =
x
y (35)
x
y
vx
v y

Kij c j = Fi , (i = 1: N )

(39)

K ij = B (i , j ), F = l (i ) B (i , j ) .

(40)

j =1

The algebraic equations in Equation (39) can be expressed in matrix form as:
[ K ]{c} = [ F ] or Kc = F .

4.2.7. Approximation Functions

Let approximate solution is sought in the form

are secondary variables.

U N ( x) = c j j ( x)

4.2.5. Weak Form

Weak form is defined to be a weighted integral statement
of a differential equation in which the differentiation is
transformed the dependent variable to the weight function such that all natural boundary conditions of the problem are also introduced.
4.2.6. Ritz Method
In Ritz method, the coefficients c j of the approximations are determined using the weak form of the
problem, and hence choice of weight functions is restricted to the approximation function w = j .
Consider the variational problem resulting from the
weak form: find the solution u such that
B ( w, u ) = l ( w)
(36)
for all sufficiently differentiable functions w that satisfy the homogeneous form of the specified essential boundary conditions on u. In general, B (, ) can be unsymmetric in w and u and l is linear, the problem
in (36) is equivalent to minimization of the quadratic
functional
1
I (u ) = B (u, u ) l (u ).
(37)
2
In Ritz method, we seek an approximation solution to
(36) in the form of a finite series
N

U N ( x) = c j j ( x) 0 ( x),

(38)

j =1

where the constants c j , called the Ritz coefficients are

determined such that (36) holds for each
w = i (i = 1: N ).
The function j and 0 , called approximation functions are chosen such that U N satisfies the specified
essential boundary conditions.
N

B i , c j j 0 = l (i ), (i = 1 : N ).
j =1

Copyright 2011 SciRes.

(41)

j =1

and suppose that the specified essential boundary conditions is u ( x0 ) = u0 , then U N must satisfy the condition U N ( x0 ) = u0 at a boundary point x = x0 :
N

c j j ( x0 ) = u0 .
j =1

Since c j are unknown parameters to be determined,

it is not easy to choose j ( x) such that the above relation holds. If u0 , then we cans select all j such that
j ( x0 ) and satisfy the condition U N ( x0 ) = 0.
N

U N ( x0 ) = c j j ( x0 ) 0 ( x0 )

(42)

j =1

j =1

j =1

u0 = c j j ( x0 ) u0 c j j ( x0 ) = 0

(43)

which is satisfied, for arbitrary c j , by choosing

j ( x0 ) = 0. If all specified essential boundary conditions are homogeneous, then 0 is taken to be zero and
j must still satisfy the same conditions, j ( x0 ) = 0,
j = 1: N . Note that the requirement that w be zero at
the boundary conditions are specified is satisfied by the
choice w = j ( x).
4.2.8. The Method of Weighted Residuals
The method of weighted residual can be described in its
generality by considering the operator equation
A(u ) = f in .
(44)

where A is an operator (linear or nonlinear), often a

differential operator, acting on the dependent variables
and f is a known function of the independent variables.
In the weighted residual method, the solution u is
approximated, in much the same way as in Ritz method
by the expression
N

U N ( x0 ) = c j j ( x) 0 ( x)

(45)

j =1

AM

M. KUMAR

except that the requirements on 0 and j for the

weighted residual method are more stringent than those
for the Ritz method
N

R = A(U N ) f = A c j j 0 f 0 .
j =1

(46)

As the name suggests, the parameter c j . are determined by requiring the residual R to vanish in the
weighted integral sense:

i ( x) R( x, c j )dxdy.

(47)

where i ( x) are weight functions. There are following

type of well known method based on weight functions.
1) The Patrov-Galerkin method: The weighted residual method is referred as the Petrov-Galerkin method
when i i . When operator A is linear Equation
(47) is simplified to the form

i A( j )dx c j = i f A(0 ) dx
N

(48)

j =1

Aij c j = Fi

or

( Ac = F )

(49)

j =1

where Aij = i A( j )dx.

2) The Galerkin method: If the weight function i

is chosen to be equal to the approximation function i
the weighted residual method is better known as Galerkin method. The algebraic equations of the Galerkin
approximation are
Ac = F
(50)
where

1335

ET AL.

(i = 1: N ) in the domain
R ( x i , c j ) = 0 (i = 1: N ) .

(52)

The selection of points x i is crucial in obtaining a

well-conditioned system of equations and ultimately in
obtaining an accurate solution. The collocation method
can be shown to be a specified case of Equation (47)
with i = x x i , where ( x ) is the Dirac Delta
fun- ctions which is defined by

f ( x ) x dx =

f ( )

(53)

with this choice of the weight functions, the weighted

residual statement becomes
i
x x R( x, c j )dx = 0

R ( x i , c j ) = 0.

The steps involved in finite element method of a problem:

1) Discretization of the given domain into a collection
of prescribed finite element.
(a) Construct the finite element mesh of the prescribed
elements.
(b) Number of nodes and elements.
(c) Generate the geometry properties needed for problem.
2) Derivation of element equations for all typical element in the mesh:
(a) Construct the variational formulation of the given
differential equation over the typical element.
(b) Assume that a typical dependent variable u is of
the form
N

u = ui i

Aij = i A( j )dx, Fi = i f A(0 ) dx,

i =1

the approximation functions are used in Galerkin method

required to be higher order than those in the Ritz method.
For further readings we suggest to see [24-28].
3) The least-square method: In the least-square method, we determine the parameters c j by minimizing
the integral of the square of the residual.

ci

R
R

( x, c j )dx = 0

Rdx = 0.

(51)

R
.
ci
4) The collocation method: In the collocation method, we seek an approximate solution U N to Equation
(44) in the form of U N by requiring the residual to
vanish identically at N selected points x i = ( x i , y i , z i )

Here weight function i =

Copyright 2011 SciRes.

and substitute it into step 2(a) to obtain element equation

in the form
[ K e ][u e ] = [ F e ].

(c) Select, if already available in the literature, or

derive element interpolation functions i and compute
the element matrices.
3) Assembly of element equations to obtain the equation of whole problem.
4) Imposition of the boundary conditions of the problem.
5) Solution of the assembled equation.
6) Postprocessing of the results.
Authors can view the references [29-47] for more study on finite element method.

5. Error Estimate
In this section, we derive error estimates for the finite
AM

1336

M. KUMAR

element method. The discrete version of the Generalized

Lax-Milgram theorem, gives the uniqueness of the
solution to the discrete equation and gives a first estimate
of the error. This theorem is in fact only applicable when
we use finite dimensional subspaces of our original Hilbert spaces. We have a more general case when we dont
have such subspaces or when the operators in the variational equation are replaced by approximations (for
instance by quadrature). We give also error estimates for
this case. The error estimates depend on how good we
can interpolate elements of Banach spaces subspaces of
these Banach spaces, so we have to discuss the interpolation theory in Banach spaces, preceded by a necessary discussion of the formalism of the finite element
method. This will give us estimates in the Sobolev norms
k : km; q;. We will also give an estimate in the L2 norm, but for this we need additional requirements on the
problem we consider.
1) The error estimate should give an accurate measure
of the discretization error for a wide range of mesh spacings and polynomial degrees.
2) The procedure should be inexpensive relative to the
cost of obtaining the finite element solution. This usually
means that error estimates should be calculated using
only local computations, which typically require an effort comparable to the cost of generating the stiffness matrix.
3) A technique that provides estimates of point-wise
errors which can subsequently be used to calculate error
measures in several norms is preferable to one that only
works in a specific norm. Point-wise error estimates and
error estimates in local (elemental) norms may also provide an indications as to where solution accuracy is insufficient and where refinement is needed.
A posteriori error estimates can roughly be divided
into four categories:
1) Residual error estimates: Local finite element problems are created on either an element or a sub-domain
and solved for the error estimate. The data depends on
the residual of the finite element solution.
2) Flux-projection error estimates: A new flux is calculated by post processing the finite element solution.
This flux is smoother than the original finite element flux
and an error estimate is obtained from the difference of
the two fluxes.
3) Extrapolation error estimates: Two finite element
solutions having different orders or different meshes are
compared and their differences used to provide an error
estimate.
4) Interpolation error estimates: Interpolation error
bounds are used with estimates of the unknown constants.
Copyright 2011 SciRes.

ET AL.

6. Conclusions
The present paper gives a comprehensive overview of
the fundamentals of partial differential equations and related tools for their numerical solution available in the
literature. Many fundamental ideas and techniques in finite difference and finite element methods have resemblance, and in some simple cases they coincide. Nevertheless, with its more systematic use of the variational
approach, its greater flexibility, and the way it more
easily lends itself to error analysis, the finite element
method has become the dominating approach for tackling
the partial differential equations together with their applications in science and engineering.

7. Acknowledgements
This research work is financially supported by the grant,
No. SR/FTP/MS-14/2007, sponsored by Department of
Science and Technology, New Delhi, Government of
India. The authors would like to thank the anonymous
reviewers for their valuable comments and suggestions to
improve the manuscript.

8. References
[1]

M. Kline, Mathematical Thought from Ancient to Modern Times, Oxford University Press, London, 1972.

[2]

H. Poincare, Sur les Equations aux Derivees Partielles de

la Physique Mathematique, American Journal of Mathematics, Vol. 12, No. 3, 1890, pp. 211-294.
doi:10.2307/2369620

[3]

H. Poincare, Sur les Equations de la Physique MatheMatique, Rendiconti del Circolo Mathematico di Palermo, Vol. 8, 1894, pp. 57-155.

[4]

E. E. Levi, Sulle Equazioni Lineare Totalmente Ellitiche, Rendiconti del Circolo Mathematico di Palermo,
Vol. 24, No. 1, 1907, pp. 275-317.
doi:10.1007/BF03015067

[5]

O. A. Ladyzhenskaya and N. N. Uraltseva, Linear and

Quasi-Linear Elliptic Equations, Academic Press, New
York, 1968.

[6]

A. M. Micheletti and A. Pistoia, On the Existence of

Nodal Solutions for a Nonlinear Elliptic Problem on
Symmetric Riemannian Manifolds, International Journal of Differential Equations, Vol. 2010, 2010, pp. 1-11.
doi:10.1155/2010/432759

[7]

I. M. Gelfand, Some Problems in Theory of Quasilinear

Equations, Transactions of the American Mathematical
Society, Vol. 29, 1963, pp. 295-381.

[8]

D. D. Joseph and E. M. Sparrow, Nonlinear Diffusion

Induced by Nonlinear Sources, Quarterly of Applied
Mathematics, Vol. 28, 1970, pp. 327-342.

[9]

H. B. Keller and D.S. cohen, Some Positive Problems

Suggested by Nonlinear Heat Generation, Journal of

AM

M. KUMAR
Mathematics and Mechanics, Vol. 16, No. 12, 1967, pp.
1361-1376.
[10] G. E. Forsthye and W. R. Wasow, Finite Difference
Methods for Partial Differential Equations, Wiley, New
York, 1960.
[11] J. D. Hoffman, Numerical Methods for Engineers and
Scientists, 2nd Edition, McGraw-Hill, Inc., New York,
1992.
[12] M. K. Jain, R. K. Jain, R. K. Mohanty, Fourth Order
Difference Methods for the System of 2-D Non-Linear Elliptic Partial Differential Equations, Numerical Methods
for Partial Differential Equations, Vol. 7, No. 3, 1991, pp.
227-244. doi:10.1002/num.1690070303
[13] L. V. Kantorovich and V. I. Krylov, Approximate Methods in Higher Analysis, 3rd Edition, Interscience, New
York, 1958.
[14] M. Kumar, P. Singh and P. Kumar, A Survey on Various
Computational Techniques for Nonlinear Elliptic Boundary Value Problems, Advances in Engineering Software,
Vol. 39, No. 9, 2008, pp. 725-736.
doi:10.1016/j.advengsoft.2007.11.001
[15] M. Kumar and P. Kumar, Computational Method for
Finding Various Solutions for a Quasilinear Elliptic Equation of Kirchhoff Type, Advances in Engineering Software, Vol. 40, No. 11, 2009, pp. 1104-1111.
doi:10.1016/j.advengsoft.2009.06.003
[16] R. K. Mohanty and S. Dey, A New Finite Difference
Discretization of Order Four for un for Two-Dimensional Quasi-Linear Elliptic Boundary Value Problems,
International Journal of Computer Mathematics, Vol. 76,
No. 4, 2001, pp. 505-516.
doi:10.1080/00207160108805043
[17] L. A. Ogenesjan and L. A. Ruchovec, Study of the Rate
of Convergence of Variational Difference Schemes for
Second-Order Elliptic Equations in a Two-Dimensional
Field with a Smooth Boundary, USSR Computational
Mathematics and Mathematical Physics, Vol. 9, No. 5,
1969, pp. 158-183. doi:10.1016/0041-5553(69)90159-1

ET AL.

1337

[24] A. Pedasa and E. Tamme, Discrete Galerkin Method for

Fredholm Integro-Differential Equations with Weakly
Singular Kernel, Journal of Computational and Applied
Mathematics, Vol. 213, No. 1, 2008, pp. 111-126.
doi:10.1016/j.cam.2006.12.024
[25] R. P. Kulkarni and N. Gnaneshwar, Iterated Discrete
Polynomially Based Galerkin Methods, Applied Mathematics and Computation, Vol. 146, No. 1, 2003, pp. 153165. doi:10.1016/S0096-3003(02)00533-7
[26] M. H. Schultz, Rayleigh-Ritz-Galerkin Methods for MultiDimensional Problems, SIAM Journal on Numerical
Analysis, Vol. 6, No. 4, 1969, pp. 523-538.
doi:10.1137/0706047
[27] M. H. Schultz, L2 Error Bounds for the Rayleigh-RitzGalerkin Method, SIAM Journal on Numerical Analysis,
Vol. 8, No. 4, 1971, pp. 737-748. doi:10.1137/0708067
[28] Y. Jianga and Y. Xu, Fast Fourier Galerkin Methods for
Solving Singular Boundary Integral Equations: Numerical Integration and Precondition, Journal of Computational and Applied Mathematics, Vol. 234, No. 9, 2010,
pp. 2792-2807. doi:10.1016/j.cam.2010.01.022
[29] I. Babuska and A. K. Aziz, Survey Lectures on the
Mathematical Foundation of the Finite Element Method,
In: A. K. Aziz, Ed., The Mathematical Foundations of the
Finite Element Method with Applications to Partial Differential Equations, Academic Press, New York, 1972,
pp. 5-359.
[30] K. Bhmer, Numerical Methods for Nonlinear Elliptic
Differential Equations, Oxford University Press, New
York, 2010.
[31] R. Courant, Variational Methods for the Solution of Problems of Equilibrium and Vibration, Bulletin of American, Mathematical Society, Vol. 49, 1943, pp. 1-23.
doi:10.1090/S0002-9904-1943-07818-4
[32] M. Ghimenti and A. M. Micheletti, On the Number of
Nodal Solutions for a Nonlinear Elliptic Problems on
Symmetric Riemannian Manifolds, Electronic Journal
of Differential Equations, Vol. 18, 2010, pp. 15-22.

[18] R. D. Richtmyer and K. W. Morton, Difference Methods

for Initial Value Problems, 2nd Edition, Wiley-Interscience, New York, 1967.

[33] N. Hirano, Multiple Existence of Solutions for a Nonlinear Elliptic Problem on a Riemannian Manifold, Nonlinear Analysis: Theory, Methods and Applications, Vol.
70, No. 2, 2009, pp. 671-692.

[19] A. A. Samarskii, Theory of Difference Schemes, Marcel Dekker Inc., New York, 2001.

[34] D. V. Hutton, Fundamentals of Finite Element Analysis, Tata McGraw-Hill, New York, 2005.

[20] R. Eymard and T. R. Gallout and R. Herbin, The Finite

Volume Method Handbook of Numerical Analysis, Vol.
7, 2000, pp. 713-1020.
doi:10.1016/S1570-8659(00)07005-8

[35] M. Kumar and P. Kumar, A Finite Element Approach

for Finding Positive Solutions of Semilinear Elliptic
Dirichlet Problems, Numerical Methods for Partial Differential Equations, Vol. 25, No. 5, 2009, pp. 1119-1128.
doi:10.1002/num.20390

[21] R. J. Leveque, Finite Volume Methods for Hyperbolic

Problems, Cambridge University Press, Cambridge, 2002.
[22] P. Wesseling, Principles of Computational Fluid Dynamics, Springer-Verlag, Berlin, 2001.
doi:10.1007/978-3-642-05146-3
[23] I. Babuska, Courant Element: Before and After, In: M.
Krizek, P. Neittanmaki and R. Stenberg, Eds., Finite Element Methods: Fifty Years of the Courant Element, Marcel Dekker, New York, 1994, pp. 37-57.
Copyright 2011 SciRes.

[36] M. Kumar and P. Kumar, Simulation of a Nonlinear

Steklov Eigen Value Problem Using Finite Element Approximation, Computational Mathematics and Modelling, Vol. 21, No.1, 2010, pp. 109-116.
doi:10.1007/s10598-010-9058-6
[37] R. Molle, Semilinear Elliptic Problems in Unbounded
Domains with Unbounded Boundary, Asymptotic Analysis, Vol. 38, No. 3-4, 2004, pp. 293-307.

AM

1338

M. KUMAR

[38] J. T. Oden and D. Somogyi, Finite Element Applications

in Fluid Dynamics, Journal of the Engineering Mechanics Division, ASCE, Vol. 95, No. 4, 1968, pp. 821-826.
[39] J. T. Oden, A General Theory of Finite Elements, II:
Applications, International Journal for Numerical Methods in Engineering, Vol. 1, No. 3, 1969, pp. 247-259.
doi:10.1002/nme.1620010304
[40] J. T. Oden, A Finite Element Analogue of the NavierStokes Equations, Journal of the Engineering Mechanics
Division, ASCE, Vol. 96, No. 4, 1970, pp. 529-534.
[41] E. R. De Arantes and E. Oliveira, Theoretical Foundation of the Finite Element Method, International Journal
of Solids and Structures Vol. 4, No. 10, 1968, pp. 926952.
[42] S. S. Rao, The Finite Element Method in Engineering,
4th Edition, Elsevier, Butterworth Heinemann, 2005.

Copyright 2011 SciRes.

ET AL.
[43] J. N. Reddy, An Introduction to the Finite Element
Method, 3rd Edition, McGraw-Hill, New York, 2005.
[44] M. Ramos and H. Tavares, Solutions with Multiple
Spike Patterns for an Elliptic System, Calculus of Variations and Partial Differential Equations, Vol. 31, No. 1,
2008, pp. 1-25. doi:10.1007/s00526-007-0103-z
[45] F. Williamson, A Historical Note on the Finite Element
Method, International Journal for Numerical Methods
in Engineering, Vol. 15, No. 6, 1980, pp. 930-934.
doi:10.1002/nme.1620150611
[46] O. C. Zienkiewicz, The Finite Element Method in Engineering Science, 3rd Edition, Mc-Graw-Hill, London,
1977.
[47] M. Zlamal, On the Finite Element Method, Numerische
Mathematik, Vol. 12, No. 5, 1968, pp. 394-409.
doi:10.1007/BF02161362

AM