Numerical Solutions of ODE and PDE

Lesson 1

Introduction

Ordinary Differential Equation

An ordinary differential equation (ODE) is an equation stating a relationship be-

tween a function of a single independent variable and the derivatives of this func-
tions with respect to the independent variable. For example:

ψ(t, y, y 0 . . . , y m )

The order of an ODE is the order of the highest order derivative in the differential
equation.
If no product of the dependent variable y(t) with itself or any of its derivatives
occurs, then the equation is called linear, otherwise it is non-linear.
Examples are

y 00 + y = 0 linear
y0 + y2 = 0 nonlinear
y 0 + t2 y = 0 linear
y 00 + sin(y) = 0 nonlinear

The general first order ODE is of the form

dy
= f (t, y)
dt
A general solution of an ODE of order m contains m arbitrary constants that can
be determined by prescribing m conditions. There are two different classes of of
ODE, depending on the type of auxiliary conditions specified.

1-1
 Introduction

Initial and Boundary Value Problem (IVP & BVP)

If all the auxiliary conditions are specified at the same value of the independent
variable and the solution is to be marched forward from that initial point, the dif-
ferential equation together with the specified condition is called an IVP.

If the auxiliary conditions are specified at two different values of the independent
variable, the end point or at the boundaries of the domain of interest, the differential
equation is called boundary value problem. For example:

y 00 + P (t, y)y 0 + Q(t, y)y = F (t) y(t0 ) = c1 and y 0 (t0 ) = c2 (IV P )

y 00 + P (t, y)y 0 + Q(t, y)y = F (t) y(t1 ) = d1 and y(t2 ) = d1 (BV P )

Reduction of higher order equations to the system of first order

differential equations

Suppose that an n-th order equation can be solved for the n-th derivative, i.e., it can
be written in the form:

y (n) = f (t, y 0 , y 00 , . . . , y n−1 )

This equation can now be written in a system of first order differential equations by
a standard change of variables:

y1 = y
y2 = y 0
y3 = y 00
..
.
yn = y n−1 .

1-2
 Introduction

Then, the resulting first-order system is the following:

y10 = y 0 = y2
y20 = y 00 = y3
y30 = y 000 = y4
..
.
yn0 = y n = f (t, y2 , y3 , . . . , yn ).

In vector form this can simply be written as

y0 = f (t, y)

where y = [y1 , y2 , . . . , yn ]T and f = [y2 , y3 , . . . yn , f ]T .

Let us assume that the initial values for the nth order problem are given as

y(t0 ) = y0 , y 0 (t0 ) = y1 , . . . , y n−1 (t0 ) = yn−1

Clearly, it follows
y(t0 ) = [y0 , y1 , . . . , yn−1 ]T .

Example 1.1 Convert the second order IVP into a system of first order IVP

2y 00 − 5y 0 + y = 0

y(0) = 6; y 0 (0) = −1;

Sol: Let
y1 = y y2 = y 0 .
It follows then

y10 = y2
1 5
y20 = − y1 + y2
2 2
and

y1 (0) = 6; y2 (0) = −1

1-3
 Introduction

Remark 1.2 The methods of solution of first order initial value problem may be
used to solve the system of first order initial value problems and the nth order
initial value problem.

Suggested Readings

A. Quarteroni, R. Sacco, F. Saleri (2007). Numerical Mathematics. Second Edition.

Springer Berlin.
M.K. Jain, S.R.K. Iyengar, R.K. Jain (2009). Numerical Mathematics. Fifth Edi-
tion. New age international publishers, New Delhi.

1-4
 Numerical Solutions of ODE and PDE

Lesson 2

Numerical Solutions of IVP

Let us consider the following IVP and discuss its existence and uniqueness of solu-
tion:
dy
= f (t, y), y(t0 ) = y0 t ∈ [t0 , b] (2.1)
dt

Existence and Uniqueness of the Solution of IVP

The IVP (2.1) admits a unique solution y(t) if f (t, y) is uniformly Lipschitz con-
tinuous, that is,

|f (t, y1 ) − f (t, y2 )| ≤ L|y1 − y2 | for any t ∈ [t0 , b] and any y1 and y2 .

Here L is called Lipschitz constant.

Finding exact solution of a practical problem is hardly possible, therefore numerical

solutions are required.

Numerical Solutions of IVP

The numerical methods produce approximate values of y(t) at certain points along
the t coordinate called grids or mesh points. In case of equally spaced grid points
we have

ti+1 = ti + h, i = 0, 1, . . . , N − 1; tN = b

where h is called the step size.

2-1
 Numerical Solutions of IVP

Figure 2.1: Grid points

We shall use the following notation for the the approximation of solution of IVPs

un ≈ y(tn ) =: yn .

Single or Multi Step Methods: If the method advances the solution from
one grid point to the next grid point using only data at the single grid point, that is,
un+1 depends only on un , it is called one-step or single step method otherwise it is
called multistep method.

Explicit and Implicit Methods: A method is called explicit method if

un+1 can be computed directly in terms of the previous values uk , k ≤ n, implicit
if un+1 depends implicitly on itself through f .

Single Step Method

A single step method can be written as

un+1 = un + hφ(tn , un , fn , h)

2-2
 Numerical Solutions of IVP

where φ is called an increment function.

Figure 2.2: Increment function

2-3
are