MA 1201- MATHEMATICS-II

Ordinary Differential Equations

Dr. Pratibhamoy Das

Department of Mathematics
Indian Institute of Technology Patna
India 801106

January 8, 2025

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 1 / 36

Happy New Year 2025

Note: 442 = 1936 452 = 2025, 462 = 2116.

& Hence 2025 will be the only Square Year (452 ) in Our Lifetimes.
In addition (1 + 2 + 3 + · · · + 9)2 = 2025, (10 + 20 + 30 + · · · + 20250 ) = 2025

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 2 / 36

Lets Start the Class

Dr. Pratibhamoy Das (IITP) Figure:

MA 1201- MATHEMATICS-II January 8, 2025 3 / 36
Marks Distribution, Rules and Books

Instructors:
Dr. Pratibhamoy Das (ODE- 20 Lectures)
Dr. Sushmitha P. (Probability- 12 Lectures)
Dr. Yogesh Mani Tripathi (Statistics- 10 Lectures)

Books-
G. F. Simmons- ODE
S. L. Ross - ODE

My ODE Notes: http://172.16.1.252/~pratibhamoy/

75% Attendance Compulsory (in Core Classes)
Medical leaves fall under the rest of 25%. (Hence do not send any medical
sickness prescription/request)
Absolute Grading
Reserve Timings for Quizzes/Classes- "Saturday & Free Slots”
Any mistakes (typo/logical/mathematical etc.) in the slide- should be
reported with Slide Number to the Instructor. Otherwise you must
follow and read the correct ones from the books.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 4 / 36

Types of Differential Equations

An equation containing the derivatives or differentials of functions is called a

differential equation.
A differential equation can have ONE or MORE dependent and independent
variables:
The differential equations that have only one independent variable are called
ordinary differential equations (ODE).

dy d2 y dy dy dx
+ 5y = ex , − + 6y = 0, + = 2x + y
dx dx2 dx dt dt
The differential equations that have TWO or MORE independent variables are
called Partial differential equations (PDE).
∂2u ∂2u ∂u ∂v ∂v ∂2θ ∂θ ∂ϕ ∂2ϕ
2
= −2 , + = v, λ1 + λ2 + λ3 2 = 0
∂x ∂t2 ∂t ∂s ∂t ∂x 2 ∂y ∂x ∂y

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 5 / 36

Order and Degree of Differential equations

The order of a differential equation is the order of the highest derivative present
in the differential equation.
In the equation above, the differential equations in cyan are of first order, while
the others are of second order.
The degree of a differential equation is the highest degree of the highest order
derivatives when the equation is rationalized; that is when the equation is made
free from fractional powers of the derivatives.
All of the above equations have degree one.
d2 y dy 2
The order and the degree of the ODE + y}1/3 = 0 are 2 and 3,


dx2
+ 5{ dx
respectively.
IVP: Initial Value Problem : Solution will be given at a point. We have to find
the solution by solving the given ODE in a neighborhood of that point. E.g:
y ′ = 2y, x ∈ (0, 1], y(0) = 0. Solution is defined on [0, 1].
BVP: Boundary Value Problem : Solution (or information) will be given at more
than one point. We have to find the solution by solving the given ODE between
these points. For example, y ′′ = 3y, x ∈ (0, 1), y(0) = 0, y(1) = 5. The solution
is defined in [0, 1].

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 6 / 36

Application of ODEs(Undamped Simple Harmonic Motion)

An application of Differential Equation appears - By considering a mass suspended

from a spring attached to a rigid support, and there is no friction. This is called a
spring-mass system.

Figure: 1. A spring in its natural position (a), at equilibrium with a mass m attached (b),
and in oscillatory motion (c).

1 L denote the length of the spring.

2 m is the mass and k is the spring constant.
Suppose the mass is pulled downward. Then x(t) represent the displacement of the
mass from its equilibrium position at time t.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 7 / 36

Application
In the system (Figure 1(c)), two forces act on the system.
1 Gravitational force (mg) and

2 Restoring force.

According to Hooke’s Law, Restoring(f ) ∝ displacement, i.e., f = −k(s + x),

1 Why negative sign? (f is pulling it upward).

From Newton’s second law, we can write:

d2 x
F =m , (1)
dt2
where F is the total force acting on the system.
The total force acting on the system is −k(s + x) + mg. Substituting this into
equation (1), we obtain:
d2 x
m 2 = −k(s + x) + mg. (2)
dt
At the equilibrium position (Figure 1(b)), mg = ks. Then the equation simplifies to:
d2 x
m = −kx.
dt2

1 Equation (2) is an ordinary differential equation (ODE), where x is the

dependent variable and t is the independent variable.
2 The order of the ODE is 2, and its degree is 1.
Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 8 / 36
Space of Functions

C(I, R): the VS of all real cont. functions on I . Example: f (x) = |x| in [−1, 1].
Definition: C n (I, R) of all real functions on I having continuous derivatives up to
order n.
Definition: C 1 (I, R): Set of all real functions on I having continuous derivatives
up to order 1. Example: f (x) = x|x| in [−1, 1].
Definition: C 2 (I, R): Set of all real functions on I having continuous derivatives
up to order 2. Example: f (x) = x2 |x| in [−1, 1].
Definition: C ∞ (I, R): Set of all real functions on I which are infinite times
differentiable. Example: f (x) = exp(−1/x2 ), when x ̸= 0, and f (0) = 0. But its
power series will not converge to the function.
Q: Give an example of a continuous function that is nowhere differentiable.
Show it by deduction.
Q: Give an example of a function that is two times differentiable at the point
x = 0 only, but nowhere else. Show it by deduction.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 9 / 36

Explicit and Implicit Solutions

d2 y dn y
Notations: y ′ = dy
dx
, y ′′ = dx2
, y (n) = dxn
1st order ODEs may also be written using differentials. E.g., dy = (ex − 5y)dx
Now onward, we will consider only ODE and with one dependent variable.
An nth order ODE can be written as:

F (x, y, y ′ , . . . , y (n) ) = 0, (3)

where F : I × D → R for some interval I and D ⊆ Rn+1 .

Let I be an interval. A function f ∈ C n (I, R) is an explicit solution of the ODE
(1) on I if F (x, f (x), f ′ (x), . . . , f (n) (x)) = 0, ∀x ∈ I.
A relation g(x, y) = 0 is an implicit solution of (1) if there is an explicit solution
f of (1) on an interval I such that g(x, f (x)) = 0 ∀x ∈ I.
Both explicit and implicit solutions are usually called solutions.
Example: An explicit solution of y ′ − 3y = 0 is f (x) = 2e3x , x ∈ R.
An implicit solution of yy ′ + x = 0 is x2 + y 2 = 4, x ∈ (−2, 2).

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 10 / 36

General vs Particular Solutions

Consider y ′ + y = 0. For any c ∈ R, f (x) = ce−x is a solution on R. This is a

one-parameter family of solutions. Are these all solutions? Yes. We will see very
soon how.
f (x) = a cos(2x) + b sin(2x) is a solution of y ′′ + 4y = 0 on R for any a, b ∈ R.
This is a two-parameter family of solutions. Are these all solutions? Yes!
If a family y = fλ (x), λ ∈ Λ gives all the solutions of a DE, then it is called the
general solution of the DE. An individual solution of the DE is called a
particular solution.
Assuming the above answers as yes, y(x) = ce−x , c ∈ R, is the general solution
and y(x) = 2e−x is a particular solution of y ′ + y = 0 in R.
y(x) = a cos(2x) + b sin(2x), a, b ∈ R, is the general solution and
y(x) = 3 sin(2x) is a particular solution of y ′′ + 4y = 0 on R.
The particular solution and the particular integral are different. Particular
Integral comes as a solution of non-homogeneous ODEs.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 11 / 36

Initial Value Problem(IVP)

Initial Value Problem (IVP): Finding a solution of

F (x, y, y ′ , y ′′ , . · · · , y (n) ) = 0

on an interval I, which satisfies the supplementary conditions

y(x0 ) = a0 , y ′ (x0 ) = a1 , · · · , y (n−1) (x0 ) = an−1 (initial conditions) for some
x0 ∈ I.
The IVP y ′ + y = 0, y(0) = 1, has a solution on R: y = e−x .
Q. Does an IVP necessarily have a solution? No.
The IVP xy ′ = 4y, y(0) = 1 has no solution. [The general solution is
y = cx4 , c ∈ R.]
Q. Can an IVPphave more than one solution? Yes.
Consider y ′ = |y|, ( y(0) = 0.
x2
for x ≥ 0,
Both y = 0 and y = 4 are solutions.
0 for x < 0
In fact it has infinitely many solutions. (Find it Geometrically !!)

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 12 / 36

Wellposed Differential Equation

Existence of the Solution

Uniqueness of the Solution
Stability of the Solution (continuous dependence of the initial & given data)

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 13 / 36

Boundary Value Problems(BVP)

Boundary Value Problem (BVP): Finding a solution of

F (x, y, y ′ , y ′′ , . · · · , y (n) ) = 0

on an open interval I, together with a set of supplementary conditions about the

values of y, y ′ , y ′′ , ..., y (n) at two (or more) different values of x in I.
The solution of the BVP

y ′′ + y = 0,I = (−π/2, π/2), y(0) = 0, y(π/2) = 3, is y = 3 sin x.

Q. Does a BVP necessarily have a solution? No.

The BVP,

y ′′ + y = 0, I = (−π, π), y(−π) = 1, y(π) = 2, has no solution.

Q. Can an BVP have more than one solution? Yes.

y ′′ + y = 0, I = (0, π), y(0) = 0, y(π) = 0, has Infinitely Many Solutions

y = c sin x, where c is an arbitrary constant.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 14 / 36

First Order ODE

A first-order ODE can be written as F (x, y, dy

dx
) =0

If the ODE is of degree 1, we can write it as dxdy

= f (x, y), where f is defined on
some region R in R2 . A solution y = ϕ(x) is then a function ϕ : I → R such that
(x, ϕ(x)) ∈ R and ϕ′ (x) = f (x, ϕ(x)) ∀x ∈ I.
Thus, it gives a curve {(x, ϕ(x)) : x ∈ I} in the region R; the slope at a point of
the curve is the value of f at that point.

A region R ⊆ R2 at each point of which a direction (slope for a line) f (x, y) is

specified is a direction field. The set of points {(x, y) : f (x, y) = k}, where k is
any constant, is called an isocline in the direction field. A curve in R at each
point of which the tangent is in the direction specified by the direction field is an
integral curve. These curves correspond to solutions of the ODE dx dy
= f (x, y).

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 15 / 36

Separation of Variable

Consider y ′ = f (x, y), and f (x, y) = h(y)

g(x)
in a rectangle R (or on R2 ).
The equation can then be written as
dy
h(y) = g(x) or h(y)dy = g(x)dx. (4)
dx
Assume
Rx g and h are continuous.
Rx If ϕ : I → R is a solution of (1), then ∀x ∈ I,
x0
h(ϕ(x))ϕ ′
(x)dx = x0
g(x)dx, and therefore
Z x Z y Z x
h(ϕ(x))ϕ′ (x)dx = h(y)dy = g(x)dx,
x0 y0 x0

where y = ϕ(x) is differentiable on I, y0 = ϕ(x0 ).RThus, ϕ is given (implicitly)

by H(y) = G(x) for some (anti-derivatives) H = h(y)dy and G = g(x)dx.
R

AssumeRthat ϕ is differentiable and defined implicitly on some I by the

relation h(y)dy = g(x)dx + c (for some c ∈ R). Then, y = ϕ(x) satisfies
R
dy
h(y) dx = g(x).
The (implicit) general solution of (1):
Z Z
h(y)dy = g(x)dx + c, c ∈ R.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 16 / 36

Homogeneous First Order ODE

If f : R2 → R is homogeneous of degree k, if f (λx, λy) = λk f (x, y), ∀λ ∈ R. In

that case, f (x, y) = xk f (1, y/x).
Suppose f (x, y) is homogeneous of degree 0. Then y ′ = f (x, y) is said to be
homogeneous.
If y ′ = f (x, y) is homogeneous, then the isoclines are the lines passing through
the origin, since f (λx, λy) = f (x, y), ∀λ ∈ R.
Suppose y = f (x, y) is homogeneous. Put u = y/x, that is, xu = y. Then
y ′ = xu′ + u. The equation reduces to y ′ = f (x, y) = f (1, y/x) i.e.,

xu′ + u = f (1, u),

i.e.,
f (1, u) − u
u′ = ,
x
an equation with variables separated.
x+y
Example of homogeneous form of f (x, y): y ′ = .
x−y
Solve this equation for the solutions of y ′ = f (x, y). Do not forget to check
whether some solutions may have been added or omitted.
Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 17 / 36
Exact First Order ODE

Consider y ′ = f (x, y), where f (x, y) = −M (x,y)

N (x,y)
in a rectangle D.
Thus, the equation is M (x, y) dx + N (x, y) dy = 0. The equation is exact if
∃ F : D → R with continuous first order partial derivatives such that
∂F ∂F
= M, = N.
∂x ∂y

In that case, the equation is dF (x, y) = ∂F

∂x
(x, y) dx + ∂F
∂y
(x, y) dy = 0, and the
general solution is F (x, y) = c, c ∈ R.
Example: ydx + xdy = 0. This implies d(xy) = 0, and hence xy = c, where c is
an arbitrary constant
Suppose M and N have continuous partial derivatives on D. Then
M (x, y) dx + N (x, y) dy = 0 is exact iff ∂M
∂y
= ∂N
∂x
.
In case M (x, y) dx + N (x, y) dy = 0 holds, the continuity of the partial
derivatives of M and N implies

∂M ∂2F ∂2F ∂N
= = = .
∂y ∂y∂x ∂x∂y ∂x

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 18 / 36

Exact First Order ODE

Proof: Fix (x0 , y0 ) ∈ D. For (x, y) ∈ D define

Z x Z y
F (x, y) = M (u, y) du + N (x0 , v) dv, and
x0 y0
Z x Z y
G(x, y) = M (u, y0 ) du + N (x, v) dv.
x0 y0

Then F (x0 , y0 ) = G(x0 , y0 ) = 0 and

∂F ∂G
= M, = N.
∂x ∂y
Now, F (x, y) − G(x, y)
Z x Z y
= [M (u, y) − M (u, y0 )] du − [N (x, v) − N (x0 , v)] dv
x y0
Z x0 Z y Z yZ x
∂M ∂N
= (u, v) dv du − (u, v) du dv
x0 y0 ∂y y0 x0 ∂x
Z xZ y 
∂M ∂N
= (u, v) − (u, v) dv du
x0 y0 ∂y ∂x
= 0.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 19 / 36

Example of Exact ODE

Solve: e2y − y cos(xy) dx + 2xe2y − x cos(xy) + 2y dy = 0.

Solution: Put M (x, y) = e2y − y cos(xy),

N (x, y) = 2xe2y − x cos(xy) + 2y
and get ∂M ∂y
= 2e2y − cos(xy) + xy sin(xy) = ∂N ∂x
. So exact.
∂F
holds if for some function h,
R
∂x
= M F (x, y) = M (x, y) dx + h(y),
that is, if
Z
e2y − y cos(xy) dx + h(y) = xe2y − sin(xy) + h(y).


F (x, y) =

But then
∂F
=N demands h′ (y) = 2y.
∂y
Substituting h(y) = y 2 , we get the general solution

F (x, y) = xe2y − sin(xy) + y 2 = c, c ∈ R.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 20 / 36

Integrating Factors

Is the equation ydx + 2xdy = 0 exact? No.

Multiply both sides by y to get y 2 dx + 2xydy = 0, which is exact.
The general solution of the new equation is xy 2 = c, c ∈ R.
Q. Is this the general solution of the given equation?

y 2 dx + 2xydy = 0, y = 0, or ydx + 2xydy = 0.

So, probably an extra solution y ≡ 0 has been cropped in. Note: y ≡ 0 is a
solution of the first equation also. So, the answer is YES.
The term y is called an integrating factor for the above equation.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 21 / 36

Integrating Factors

If M (x, y)dx + N (x, y)dy = 0 is not exact, but

µ(x, y)M (x, y)dx + µ(x, y)N (x, y)dy = 0 is exact, then µ(x, y) is called an
integrating factor for the former equation.
y(2y 2 − 3x2 )dx + 2x3 dy = 0 is not exact. An integrating factor: y −3 .
Resulting exact equation:

3x2 2x3
 
2 − 2 dx + 3 dy = 0
y y
3
has general solution 2x − xy2 = c, c ∈ R.
The solution y ≡ 0 of the original equation is lost. How did it happen? We have
taken the factor y 3 out!
Each of 1/xy, 1/y 2 and 1/(x2 + y 2 ) is an integrating factor for the equation
ydx − xdy = 0. Find the solutions of the resulting exact equations and compare
them with those of the given equation.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 22 / 36

Integrating Factors

Consider the differential equation

M (x, y)dx + N (x, y)dy = 0. (5)
Let ψ(x, y) := ∂
∂y
M (x, y) − ∂x∂
N (x, y).
Suppose Φ(x, y) := N (x,y) is independent
ψ(x,y)
of y. Then, set
and consider
R
µ(x) := exp Φ(x, y)dx
µ(x)M (x, y)dx + µ(x)N (x, y)dy = 0. (6)
We have,
∂ ∂
(µ(x)N (x, y)) = µ(x)Φ(x, y)N (x, y) + µ(x) N (x, y)
∂x  ∂x

∂
= µ(x) µ(x, y) − N (x, y)
∂x
∂
= µ(x) M (x, y)
∂y
∂
= (µ(x)M (x, y)).
∂y
Thus, µ(x) is an integrating factor for (5).
Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 23 / 36
Integrating Factor

We can consider, Φ(x, y) := − M

ψ(x,y)
(x,y)
is independent of x. Then,
Z 
µ(y) := exp Φ(x, y)dx

is an integrating factor for (5).

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 24 / 36

First Order Linear Equations

Consider the differential equation

y ′ + P (x)y = Q(x), (7)
where P and Q are continuous on some I (possibly R). It is a linear equation of
first-degree. Write (7) as [P (x)y − Q(x)]dx + dy = 0.
P (x)dx is an integrating factor, because
R
µ(x) = exp
 
1 ∂ ∂
M (x, y) − N (x, y) = P (x), independent of y.
N (x, y) ∂y ∂x
The resulting exact equation is
dF (x, y) = µ(x)[P (x)y − Q(x)]dx + µ(x)dy = 0,
where F (x, y) = µ(x)y − µ(x)Q(x)dx.
R

Since µ(x) ̸= 0, the general solution of (7) is

Z
µ(x)y − µ(x)Q(x)dx = c, c ∈ R,

i.e.,
 Z  Z Z  
y = exp − P (x)dx exp P (x)dx Q(x)dx + c , c ∈ R.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 25 / 36

Bernoulli’s Equation
Bernoulli’s differential equation:
y ′ + P (x)y = Q(x)y n . (8)
Here n is any real number. It is linear, if n = 0 or 1. Suppose n ̸= 0, 1.
Put v = y 1−n . Then, v ′ = (1 − n)y −n y ′ , and (8) becomes
v ′ + (1 − n)P (x)v = (1 − n)Q(x),
which is linear in v.
Q. Solve xy ′ + y = x2 y 2 on (0, ∞).
Solution: First, divide by x to get y ′ + x1 y = xy 2 , a Bernoulli’s equation with
n = 2. The substitution v = y −1 transforms the equation to v ′ − x1 v = −x. This
linear equation has the solution
 Z  Z Z  
1 1
v = exp − − dx exp − dx (−x)dx + c [c ∈ R]
x x
Z  Z
= exp{ln x} exp{− ln x}(−x)dx + c = x[ −dx + c]

= −x2 + cx
Thus, the general solution of the Bernoulli’s equation is
1
y= , c ∈ R.
cx − x2
Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 26 / 36
Ricati’s Equation

Riccati’s differential equation:

y ′ = A(x)y 2 + B(x)y + C(x), (9)

where A, B, C ∈ C(I, R). If A = 0, then (9) is a linear equation, whereas if

C = 0, then it is a Bernoulli’s equation.

Let y0 be a particular solution of (9). Put y = y0 + z. Then, y = y0 + z is a

solution of (9) iff Ay 2 + By + C = y ′ = y0′ + z ′ = Ay02 + By0 + C + z ′ , i.e., iff
2Ay0 z + Az 2 + Bz = z ′ , i.e., iff z is a solution of the Bernoulli’s equation
z ′ − (B + 2Ay0 )z = Az 2 .

If y0 is a particular solution of (9), then show that the transformation

y = f + (1/v) reduces the equation to a linear equation in v.

Solve the differential equation y ′ = 2 − 2xy + y 2 given that y0 = 2x is a

particular solution of the equation on R.

Find the general solution of the equation y ′ = y

x
+ x3 y 2 − x5 , which has
y0 (x) = x as a particular solution on (0, ∞)

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 27 / 36

Clairut’s Equation

Clairaut’s differential equation:

y = xy ′ + f (y ′ ), (10)

where f ∈ C 1 (I, R). Put p := y ′ . Then, (10) can be written as

y = px + f (p) (11)

Differentiate (11) with respect to x and get p = xp′ + p + f ′ (p)p′ , i.e.,

xp′ + f ′ (p)p′ = 0, i.e., (x + f ′ (p))p′ = 0. (12)

(12) is a first-order differential equation in x and p. Assuming that

x + f ′ (p) ̸= 0 on I, we get p′ = 0. Solving one gets p = c, c ∈ R. Put this in (11)
and get a one-parameter family of solutions of (10) as

y = cx + f (c), c ∈ R.

Solve y = xy ′ + 12 (y ′ )2 . Here, f (t) = 21 t2 . Therefore, y = cx + 12 c2 , c ∈ R is a

one-parameter family of solutions. Another solution is obtained from
x + f ′ (p) = 0, i.e., x + p = 0, i.e., y ′ = −x. The solution y = − 21 x2 obtained
from this is called a singular solution.
Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 28 / 36
Orthogonal Trajectories

Consider C : x2 + y 2 = a2 , a > 0, a one-parameter family of curves in xy-plane

(circles with center at the origin). Consider

L = {ℓ : y = mx | m ∈ R} ∪ {ℓ0 : x = 0}.

Any ℓ ∈ L intersects any C ∈ C orthogonally. We say ‘ℓ’ is an orthogonal

trajectory of C. Here, L is the family of the orthogonal trajectories of C, and
vice versa.
C is the solution set of yy ′ + x = 0. For y ̸= 0, this is dy
dx
= −x
y
= f (x, y).

L is the solution set of xy = y. For x ̸= 0, this is

′ dy
dx
= y
x
1
= − f (x,y) .
Let F = {F (x, y, c) = 0, c ∈ K ⊂ R} be a one parameter family of curves.
Suppose each curve satisfies y ′ = f (x, y) on a region R ⊆ R2 . If f (x, y) ̸= 0,
∀(x, y) ∈ R, the orthogonal trajectories of F are the solutions of the equation
y ′ = −1/f (x, y) on R.
Ex. Find the orthogonal trajectories of the family of hyperbolas x2 − y 2 = a,
where a is a parameter.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 29 / 36

First Order IVP

Suppose F (x, y, y ′ ) = 0 be of order 1 and degree d. Theoretically, solving for y ′

we get d equations of the type y ′ = f (x, y).
Therefore, we consider the IVP

y ′ = f (x, y), y(x0 ) = x0 ,

where f is a real valued continuous function on

R = {(x, y) : |x − x0 |≤ a, |y − y0 |≤ b}, a, b > 0.

we aim to show with some more assumptions on f that the IVP has a unique
solution.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 30 / 36

First Order IVP

The IVP :
y ′ = f (x, y), y(x0 ) = x0 , (13)
where f is continuous on

R = {(x, y) : |x − x0 |≤ a, |y − y0 |≤ b}, a, b > 0.

Let ϕ be a solution of (13). That means ϕ is a differentiable function defined on

some interval I such that

∀x ∈ I, (x, ϕ(x)) ∈ R, ϕ′ (x) = f (x, ϕ(x)) and ϕ(x0 ) = y0 .

Integrate to get, Z x
ϕ(x) = y0 + f (t, ϕ(t))dt.
x0

Thus ϕ satisfies the equation,

Z x
y = y0 + f (t, ϕ(t))dt. (14)
x0

on the interval I.
Equations such as (14) are called Integral Equations (IE).
Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 31 / 36
Integral Equation and IVP

Suppose ϕ is a solution of the IE, that is,

Z x
ϕ(X) = y0 + f (t, ϕ(t))dt.
x0

Then, ϕ is a solution of the IVP (13). But how??

Since we will then have ϕ′ (x) = f (x, ϕ(x))∀x ∈ I. Why? Use Fundamental
Theorem of Calculus. Moreover, ϕ(x0 ) = y0 .

THEOREM: A function ϕ is a solution of the IVP on an interval I if and only if

ϕ is a solution of the IE on I.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 32 / 36

Picard’s Successive Approximation

Rx
Consider the integral equation (IE) y = y0 + x0
f (t, ϕ(t))dt,
where f is continuous on R.

Think of a sequence of functions ϕk (k = 0, 1, . . . ) defined on some interval I

containing x0 successively as follows:
Z x
ϕ0 (x) = y0 and ϕk+1 (x) = y0 + f (t, ϕk (t))dt, k ≥ 0.
x0

Since f is continuous on R, ϕk+1 will be defined (and cont.) on I if

(x, ϕk (x)) ∈ R.

Suppose ϕk Rare defined and ϕRk → ϕ on some I. Then ϕ is a solution of the

x x
IE on I if x0 f (t, ϕk (t))dt → x0 f (t, ϕ(t))dt, ∀x ∈ I.

We will see that with some extra conditions on f , lim ϕk exists and is a
k→∞
solution of the IE on some I. The functions ϕk will be called successive
approximations for the solution ϕ.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 33 / 36

Example of Picard’s Successive Approximation

Consider the IVP: y ′ = y, y(0) = 1 on R. The corresponding IE is

Z x
y =1+ y dt, x ∈ R.
0
Rx
Put y0 = 1 (= y(0)), and yk+1 = 1 + 0
yk dt, x ∈ R, k ≥ 0.
Then

y1 = 1 + x,
x2
y2 = 1 + x + ,...
2
x2 xn
In general, yk = 1 + x + + ··· + . [Use Induction.]
2 n!
Note that yk → ex pointwise on R, and y = ex is a solution of the IVP.
Verify that ex is a solution of the IE.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 34 / 36

Existence and Uniqueness Theorem

A real valued function f on S ∈ R2 is said to satisfy a Lipschitz Condition if

∃K > 0 ( K is Lipschitz Constant) such that,

|f (x, y1 ) − f (x, y2 )|≤ K|y1 − y2 |.

Picard’s Existence and Uniqueness Theorem: If f is continuous and satisfies the

Lipschitz condition
|f (x, y1 ) − f (x, y2 )|≤ K|y1 − y2 |
on R = {(x, y) : |x − x0 |≤ a, |y − y0 |≤ b}, then the IVP

y ′ = f (x, y), y(x0 ) = y0 ,

has a unique solution on the interval

I = {x : |x − x0 |≤ α},

where α = min{a, b/M }, M = supR |f (x, y)|.

The (pointwise) limit of the successive approximations is the solution in that
interval.

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 35 / 36

 THANK YOU

Dr. Pratibhamoy Das (IITP) MA 1201- MATHEMATICS-II January 8, 2025 36 / 36