Differential Equations

Second-order ODEs
Intro-Definitions-Reduction of order

Dr. Ahmed Sayed AbdelSamea

Giza, Egypt, Spring 2021

[email protected]
1
Higher order ODEs
Definition 1
The non-homogeneous linear 𝒏-th order ODE is given by:
𝒂𝒏 𝒙 𝒚 𝒏 + 𝒂𝒏−𝟏 𝒙 𝒚 𝒏−𝟏 + ⋯ + 𝒂𝟏 𝒙 𝒚′ + 𝒂𝟎 𝒙 𝒚 = 𝒈 𝒙 ,
where 𝒂𝒏 𝒙 , 𝒂𝒏−𝟏 𝒙 , … , 𝒂𝟎 𝒙 and 𝒈 𝒙 ≠ 𝟎 are continuous
functions of 𝒙 on an interval 𝑰.
When 𝒈 𝒙 = 𝟎, it is said to be homogeneous.
Definition 2
When the 𝒏-th order ODE given by:
𝒂𝒏 𝒙 𝒚 𝒏 + 𝒂𝒏−𝟏 𝒙 𝒚 𝒏−𝟏 + ⋯ + 𝒂𝟏 𝒙 𝒚′ + 𝒂𝟎 𝒙 𝒚 = 𝒈 𝒙 ,
satisfies the following initial conditions:
𝒚 𝒙𝟎 = 𝒚𝟎 , 𝒚′ 𝒙𝟎 = 𝒚𝟏 , … , 𝐚𝐧𝐝 𝒚 𝒏−𝟏 𝒙𝟎 = 𝒚𝒏−𝟏 , it is called
an initial value problem (IVP).

2
Higher order ODEs
Theorem 1 (Existence and Uniqueness)
Let 𝒂𝒏 𝒙 , 𝒂𝒏−𝟏 𝒙 , … , 𝒂𝟎 𝒙 and 𝒈 𝒙 be continuous functions
on an interval 𝑰 , and let 𝒂𝒏 𝒙 ≠ 𝟎 for every 𝒙 in 𝑰. For any
point 𝒙𝟎 in this interval 𝑰, there is one and only one function
𝒚 𝒙 defined on 𝑰 which is a solution to the 𝒏-th order IVP .

Example The following IVP

𝒚′′ − 𝟒𝒚 = 𝟏𝟐𝒙, 𝒚 𝟎 = 𝟒, 𝒚′ 𝟎 = 𝟏
has a unique solution 𝒚 = 𝟑𝒆𝟐𝒙 + 𝒆−𝟐𝒙 − 𝟑𝒙

Example The following IVP

𝒙𝟐 𝒚′′ − 𝟐𝒙𝒚′ + 𝟐𝒚 = 𝟔, 𝒚 𝟎 = 𝟑, 𝒚′ 𝟎 = 𝟏
has a solution which is not unique
𝒚 = 𝒄𝒙𝟐 + 𝒙 + 𝟑
3
Higher order ODEs
Theorem 2 (Superposition Principle)
Suppose that 𝒚𝟏 , 𝒚𝟐 … , 𝒚𝒌 are solutions to the homogeneous
linear ODE on an interval 𝑰 :
𝒂𝒏 𝒙 𝒚 𝒏 + 𝒂𝒏−𝟏 𝒙 𝒚 𝒏−𝟏 + ⋯ + 𝒂𝟏 𝒙 𝒚′ + 𝒂𝟎 𝒙 𝒚 = 𝟎 ,
then
𝒌
𝒚 = 𝒄𝟏 𝒚𝟏 + 𝒄𝟐 𝒚𝟐 + ⋯ + 𝒄𝒌 𝒚𝒌 = ෍ 𝒄𝒊 𝒚𝒊
𝒊=𝟏
is also a solution on 𝑰 where 𝒄𝒊 ; 𝒊 = 𝟏, . . , 𝒌 are arbitrary
constants.

Notes
• A constant multiple of a solution is also a solution.
• A homogeneous linear ODE always possess the trivial solution
𝒚 = 𝟎. 4
Higher order ODEs
Definition 3
A set of functions 𝒇𝟏 (𝒙), 𝒇𝟐 (𝒙), … , 𝒇𝒏 (𝒙) is said to be a linear
dependent set (at least one function can be expressed as a linear
combination of the remaining functions) on an interval 𝑰, if
there exist constants 𝒄𝟏 , 𝒄𝟐 , … , 𝒄𝒏 , not all zeros, such that:
𝒄𝟏 𝒇𝟏 𝒙 + 𝒄𝟐 𝒇𝟐 𝒙 + ⋯ + 𝒄𝒏 𝒇𝒏 𝒙 = 𝟎
for every 𝒙 in the interval 𝑰 . Otherwise, it is a linearly
independent set.

In other words, if, 𝒄𝟏 = 𝒄𝟐 = ⋯ = 𝒄𝒏 = 𝟎, the set of 𝒏-functions

is called a linearly independent set (we can not represent any
function as a linear combinations of the others).
5
Higher order ODEs

Example 𝒇𝟏 𝒙 = cos𝟐 𝒙 , 𝒇𝟐 𝒙 = sin𝟐 𝒙 , 𝒇𝟑 𝒙 = tan𝟐 𝒙 ,

𝝅 𝝅
𝒇𝟒 𝒙 = sec 𝟐 𝒙 is a linear dependent set on the interval − , .
𝟐 𝟐

Example 𝒇𝟏 𝒙 = cos 𝒙 , 𝒇𝟐 𝒙 = sin 𝒙 , is a linear independent

𝝅 𝝅
set on the interval − , . What about −∞, ∞ ?
𝟐 𝟐

Example 𝒇𝟏 𝒙 = 𝒙 + 𝟓, 𝒇𝟐 𝒙 = 𝒙 + 𝟓𝒙, 𝒇𝟑 𝒙 = 𝒙 − 𝟏,
𝒇𝟒 𝒙 = 𝒙𝟐 is a linear dependent set on the interval 0, ∞ .

6
Higher order ODEs
Definition 4
The Wronskian of the functions 𝒇𝟏 (𝒙), 𝒇𝟐 (𝒙), … , 𝒇𝒏 (𝒙), each
possesses at least 𝒏 − 𝟏 derivatives, is defined by
𝒇𝟏 𝒇𝟐 … 𝒇𝒏
𝒇′𝟏 𝒇′𝟐 … 𝒇′𝒏
𝑾 𝒇𝟏 , 𝒇𝟐 , … , 𝒇𝒏 = ⋮ ⋮ … ⋮
(𝒏−𝟏) (𝒏−𝟏) (𝒏−𝟏)
𝒇𝟏 𝒇𝟐 … 𝒇𝒏
Theorem 3
The set of functions 𝒇𝟏 (𝒙), 𝒇𝟐 (𝒙), … , 𝒇𝒏 (𝒙) is linearly
independent set if and only if 𝑾 𝒇𝟏 , 𝒇𝟐 , … , 𝒇𝒏 ≠ 𝟎.

The Wronskian was named after the Polish mathematician H.

Wronski (1778–1863). 7
Higher order ODEs
Theorem 4
Let 𝒚𝟏 , 𝒚𝟐 , … , 𝒚𝒏 be 𝒏 solutions of the homogeneous linear 𝒏-th
order ODE on the interval 𝑰. Then, the set of solutions is
linearly independent on 𝑰 if and only if 𝑾 𝒚𝟏 , 𝒚𝟐 , … , 𝒚𝒏 ≠ 𝟎 for
every 𝒙 on 𝑰 and is called a fundamental set of solutions.

Theorem 5
Let 𝒚𝟏 , 𝒚𝟐 , … , 𝒚𝒏 be a fundamental set of solutions of the linear
homogeneous 𝒏-th order ODE on an interval 𝑰. Then, general
solution of ODE on interval 𝑰 is:
𝒚 = 𝒄𝟏 𝒚𝟏 + 𝒄𝟐 𝒚𝟐 + ⋯ + 𝒄𝒏 𝒚𝒏
where 𝒄𝒊 , 𝒊 = 𝟏, 𝟐, … , 𝒏 are arbitrary constants.

8
Higher order ODEs
Example
Show that 𝒚𝟏 = 𝒆𝒙 , 𝒚𝟐 = 𝒆𝟐𝒙 , 𝒚𝟑 = 𝒆𝟑𝒙 , form a fundamental set
of solution on −∞, ∞ for the 3rd order ODE:
𝒚′′′ − 𝟔𝒚′′ + 𝟏𝟏𝒚′ − 𝟔𝒚 = 𝟎.

Solution
𝑾 𝒚𝟏 , 𝒚𝟐 , 𝒚𝟑 = 𝟐𝒆𝟔𝒙 ≠ 𝟎

9
Second order ODEs
Definition 1
The general solution of homogeneous linear 2-nd order ODE
𝒂𝟐 𝒙 𝒚′′ + 𝒂𝟏 𝒙 𝒚′ + 𝒂𝟎 𝒙 𝒚 = 𝟎,
or, equivalently, 𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝟎.
using the fundamental set of solutions 𝒚𝟏 , 𝒚𝟐 is given by:
𝒚 = 𝒄𝟏 𝒚𝟏 + 𝒄𝟐 𝒚𝟐
where 𝒚𝟏 , 𝒚𝟐 are two linearly independent solutions and 𝒄𝟏 , 𝒄𝟐
𝒚
are arbitrary constants 𝟏 ≠ 𝐜𝐨𝐧𝐬𝐭𝐚𝐧𝐭 .
𝒚𝟐

Note
𝒚𝟏 𝒚𝟐 ′ ′
Wronskian: 𝑾 𝒚𝟏 , 𝒚𝟐 = 𝒚′ ′ = 𝒚𝟏 𝒚𝟐 − 𝒚𝟐 𝒚𝟏 ≠ 𝟎.
𝒚𝟐
𝟏
10
Second order ODEs
Proposition 1
Suppose that 𝒚𝟏 and 𝒚𝟐 are solutions to the homogenous linear
2-nd order ODE: 𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝟎 on the interval 𝑰.
Then the Wronskian: 𝑾 𝒚𝟏 , 𝒚𝟐 is either identically zero on 𝑰
or it is never equal to zero.

Proposition 2
Suppose that 𝒚𝟏 and 𝒚𝟐 are solutions to the homogenous linear
2-nd order ODE: 𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝟎 on the interval 𝑰.
Then 𝒚𝟏 and 𝒚𝟐 are linearly dependent if and only if their
Wronskian: 𝑾 𝒚𝟏 , 𝒚𝟐 is identically zero on 𝑰.

11
Second order ODEs
Proposition 3
Suppose that 𝒚𝟏 and 𝒚𝟐 are solutions to the homogenous linear
2-nd order ODE: 𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝟎 on the interval 𝑰.
Then 𝒚𝟏 and 𝒚𝟐 are linearly independent if and only if their
Wronskian: 𝑾 𝒚𝟏 , 𝒚𝟐 ≠ 𝟎.

Theorem
Suppose that 𝒚𝟏 and 𝒚𝟐 are independent solutions to the
homogenous linear 2-nd order ODE:
𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝟎
on the interval 𝑰. Then the general solution is given by:
𝒚 = 𝒄𝟏 𝒚𝟏 + 𝒄𝟐 𝒚𝟐
where 𝒄𝟏 , 𝒄𝟐 are arbitrary constants.
12
Second order ODEs
Exercise
Show that 𝒚𝟏 = 𝒆𝟑𝒙 , 𝒚𝟐 = 𝒆−𝟑𝒙 , form a fundamental set of
solutions for the 2nd order ODE: 𝒚′′ − 𝟗𝒚 = 𝟎, hence find the
general solution and the interval of existence.
Is the function 𝒚 = 𝟒 sinh 𝟑𝒙 − 𝟓𝒆𝟑𝒙 a solution to the above
ODE? Why?

13
Second order ODEs

The 2nd order DE is given by (normal form): 𝒚′′ = 𝒇(𝒙, 𝒚, 𝒚′ )

Example: object motion (Newton’s 2nd law)
𝒅 𝒅𝟐 𝒚
(𝒎𝒗) = 𝑭 ֜ 𝒎 𝟐 = 𝑭(𝒕, 𝒚, 𝒚′ )
𝒅𝒕 𝒅𝒕
֜ 𝒎𝒚′′ = 𝑭(𝒕, 𝒚, 𝒚′ )
where,
• 𝒎 : is the object mass,
• 𝒚 : is the object location (displacement),
• 𝒚′ : is the object velocity,
• 𝒚′′ : is the object acceleration.
14
Second order ODEs
Notes
• The nonhomogeneous linear second order ODE is given by
𝒂𝟐 𝒙 𝒚′′ + 𝒂𝟏 𝒙 𝒚′ + 𝒂𝟎 𝒙 𝒚 = 𝒈(𝒙)

• In the standard form, it is: 𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝒇(𝒙)

It has no powers for 𝒚, 𝒚′ , 𝐚𝐧𝐝 𝒚′′ more than 1, no products
and no non-linear functions. 𝒑, 𝒒 are continuous functions
of the independent variable (𝒙) on some interval I. 𝒇 𝒙 is
called the forcing (driving force) term.
• The homogenous linear 2nd order DE is given by:
𝒚′′ + 𝒑 𝒙 𝒚′ + 𝒒 𝒙 𝒚 = 𝟎 (no forcing term). 15
Second order ODEs
Boundary Value Problem (BVP)
The nonhomogeneous linear second order ODE given by
𝒂𝟐 𝒙 𝒚′′ + 𝒂𝟏 𝒙 𝒚′ + 𝒂𝟎 𝒙 𝒚 = 𝒈(𝒙)
defined on an interval 𝑰 ≡ (𝒂, 𝒃) with the following boundary
conditions:
𝒚 𝒂 = 𝒚 𝟎 , 𝒚 𝒃 = 𝒚𝟏
is called a boundary value problem (BVP).
However, there are other boundary conditions forms as
𝒚′ 𝒂 = 𝒚 𝟎 , 𝒚 𝒃 = 𝒚𝟏
𝒚 𝒂 = 𝒚𝟎 , 𝒚′ 𝒃 = 𝒚𝟏
𝒚′ 𝒂 = 𝒚𝟎 , 𝒚′ 𝒃 = 𝒚𝟏

16
Second order ODEs
Differential Operators
𝒅
The differential operator symbol 𝑫 can be used instead of to
𝒅𝒙
simplify notation. Therefore, the 2nd order ODE can be given by
𝒂𝟐 𝒙 𝑫𝟐 𝒚 + 𝒂𝟏 𝒙 𝑫𝒚 + 𝒂𝟎 𝒙 𝒚 = 𝒈(𝒙)
where, in general,
𝒏
𝒏 𝒏
𝒅 𝒚
𝑫 𝒚=𝒚 = 𝒏
.
𝒅𝒙
Then, we can define the 2nd order linear operator 𝑳 as
𝑳 ≡ 𝒂𝟐 𝒙 𝑫𝟐 + 𝒂𝟏 𝒙 𝑫 + 𝒂𝟎 𝒙 .
Therefore the 2nd order ODE can by written in a short hand
notation as
𝑳 𝒚 =𝒈 𝒙 .

17
Second order ODEs
Example: the vibrating spring
𝒎𝒚′′ + 𝝁𝒚′ + 𝒌𝒚 = 𝑭(𝒕)
where,
• 𝒎 is the spring mass.
• 𝒚, 𝒚′ , 𝐚𝐧𝐝 𝒚′′ are the displacement, velocity and acceleration.
• 𝝁 is the damping constant.
• 𝒌 is the spring constant.
• 𝑭(𝒕) is the driving force.

18
Second order ODEs
Example: Consider the 2nd order ODE: 𝒚′′ − 𝒚 = 𝟎.
Given that 𝒚𝟏 𝒕 = 𝒆𝒕 is a solution in the interval −∞, ∞ ,
find 𝒚𝟐 𝒕 , hence give the general solution.

Solution
Let 𝒚𝟐 𝒕 = 𝒖 𝒕 𝒚𝟏 𝒕 ֜ 𝒆𝒕 𝒖′′ + 𝟐𝒖′ = 𝟎 ֜ 𝒖′′ + 𝟐𝒖′ = 𝟎.
Using the reduction method: let 𝒖′ = 𝒘, then the 2nd-order DE
becomes 𝒘′ + 𝟐𝒘 = 𝟎 (1st-order separable) ֜ 𝒘 𝒕 = 𝒌𝟏 𝒆−𝟐𝒕
֜ 𝒖 𝒕 = 𝒌𝟏 𝒆−𝟐𝒕 + 𝒌𝟐
𝒚𝟐 𝒕 = 𝒌𝟏 𝒆−𝒕 + 𝒌𝟐 𝒆𝒕 (choose 𝒌𝟏 = 𝟏, 𝒌𝟐 = 𝟎) ֜ 𝒚𝟐 𝒕 = 𝒆−𝒕
The general solution: 𝒚 = 𝒄𝟏 𝒚𝟏 + 𝒄𝟐 𝒚𝟐 = 𝒄𝟏 𝒆𝒕 + 𝒄𝟐 𝒆−𝒕
19
Second order ODEs
Consider the 2nd order ODE in the standard form:
𝒚′′ + 𝒑 𝒙 𝒚′ + 𝒒 𝒙 𝒚 = 𝟎
If 𝒚𝟏 (𝒙) is one solution of the above 2nd order DE, one can get
another independent solution, by reduction of order as follows:
Let 𝒚𝟐 𝒙 = 𝒖(𝒙)𝒚𝟏 (𝒙), then get 𝒚′𝟐 , 𝒚′′
𝟐 and substitute back in

the original DE and simplify to get:

𝒆− ‫𝒙 𝒑 ׬‬ 𝒅𝒙
𝒚𝟐 𝒙 = 𝒚𝟏 𝒙 න 𝟐
𝒅𝒙
𝒚𝟏 𝒙
Exercise
What about the nonhomogeneous case?
𝒚′′ + 𝒑(𝒙)𝒚′ + 𝒒 𝒙 𝒚 = 𝒇(𝒙) 20
Second order ODEs
Example: Find the general solution of the ODE:
𝒙𝟐 𝒚′′ − 𝟑𝒙𝒚′ + 𝟒𝒚 = 𝟎.
Given that 𝒚𝟏 𝒙 = 𝒙𝟐 is a solution in the interval 𝟎, ∞ .

Solution
Write the ODE in the standard form:
′′
𝟑 ′ 𝟒
𝒚 − 𝒚 + 𝟐𝒚= 𝟎
𝒙 𝒙
𝟏
− ‫𝒙𝒅 𝒙 𝒑 ׬‬ 𝟑 ‫𝒙𝒅𝒙׬‬
𝒆 𝒆
Then 𝒚𝟐 𝒙 = 𝒚𝟏 𝒙 ‫׬‬ 𝒅𝒙 ֜ 𝒚𝟐 𝒙 = 𝒙𝟐 ‫׬‬ 𝒅𝒙
𝒚𝟏 𝒙 𝟐 𝒙𝟒
𝒚𝟐 𝒙 = 𝒙𝟐 ln 𝒙
The general solution: 𝒚 = 𝒄𝟏 𝒚𝟏 + 𝒄𝟐 𝒚𝟐 = 𝒄𝟏 𝒙𝟐 + 𝒄𝟐 𝒙𝟐 ln 𝒙
21
 Second order ODEs
Example: Find the general solution of the ODE:
𝒚′′ − 𝟒𝒚 = 𝟐.
where 𝒚𝟏 𝒙 = 𝒆−𝟐𝒙 is a solution of the associated
homogeneous equation.

Solution
Let 𝒚 𝒙 = 𝒖 𝒙 𝒚𝟏 𝒙 ֜ 𝒆−𝟐𝒙 𝒖′′ − 𝟒𝒖′ = 𝟐 ֜ 𝒖′′ − 𝟒𝒖′ = 𝟐𝒆𝟐𝒙
Using the reduction method: let 𝒖′ = 𝒘, then the 2nd-order DE
becomes 𝒘′ − 𝟒𝒘 = 𝟐𝒆𝟐𝒙 (1st-order linear) ֜ 𝝁 𝒙 = 𝒆−𝟒𝒙
′ 𝟒𝒙 𝟐𝒙 𝟒𝒙
𝟏 𝟐𝒙
֜ 𝒘 = 𝒖 = 𝒄𝒆 − 𝒆 ֜ 𝒖 = 𝒌𝟐 𝒆 − 𝒆 + 𝒌𝟏
𝟐
−𝟐𝒙 𝟐𝒙 𝟏
The general solution: 𝒚 𝒙 = 𝒖 𝒙 𝒚𝟏 𝒙 = 𝒌𝟏 𝒆 + 𝒌𝟐 𝒆 −
𝟐
𝟏
where 𝒚𝟐 𝒕 = 𝒆𝟐𝒙 and 𝒚𝒑 = − . 22
𝟐
Second order ODEs

Exercise Find the general solution of the ODE:

𝒚′′ + 𝒚′ = 𝟏.
where 𝒚𝟏 𝒙 = 𝟏 is a solution of the associated homogeneous
equation.

Exercise Find the general solution of the ODE:

𝒚′′ − 𝟒𝒚′ + 𝟑𝒚 = 𝒙.
where 𝒚𝟏 𝒙 = 𝒆𝒙 is a solution of the associated homogeneous
equation.

23