Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Introduction

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
Introduction to Ordinary
Differential Equations (ODE)
Recall basic definitions of ODE,
◦ order
◦ linearity
◦ initial conditions
◦ solution

Classify ODE based on( order, linearity, conditions)

Classify the solution methods
 Differential equations
➢Differential equations involve derivatives of unknown solution function

➢Ordinary differential equation (ODE): all derivatives are for a single independent variable, often
representing time

➢The solution of a differential equation is a function in an infinite-dimensional space of functions

➢The numerical solution of differential equations is based on finite-dimensional approximation

➢A differential equation is replaced by an algebraic equation whose solution approximates that of a given
differential equation
Basic Concepts: Modelling
Basic Concepts: Modelling
Basic Concepts: Modelling
To build a model can be a long and arduous process; it may take many years of
research. Once they are formulated, models may be virtually impossible to solve
analytically. Then the researcher has two options:

• Simplify, or “tweak”, the model so that it can be dealt with in a more

manageable way. This is a valid approach, provided the simplification does not
overly compromise the “real-world” connection, and therefore, its usefulness.

• Retain the model as is and use other techniques, such as numerical or graphical
methods. This represents a qualitative approach. While we do not possess an
exact, analytical solution, we do obtain some information which can shed some
light on the model and its application. Technological tools can be extremely
helpful with this approach
Basic Concepts: Modelling
Basic Concepts: Modelling
Basic Concepts: Modelling
Derivatives
Derivatives

Ordinary Derivatives Partial Derivatives

dy u
dx y
u is a function of
y is a function of one more than one
independent variable independent variable
Differential Equations
Differential
Equations

Ordinary Differential Equations Partial Differential Equations

d y 2
u u
2 2
+ 6 xy = 1 − 2 =0
dx 2
y 2
x
- involve one or more - involve one or more - involve one or more - involve one or more
ordinary derivatives of dependent variables partial derivatives of dependent variables of
unknown functions for a single independent unknown functions two or more independent
variable variables
Ordinary Differential Equations
Ordinary Differential Equations (ODE) involve one or more ordinary derivatives
of unknown functions with respect to one independent variable

Examples :
dy
− y = ex y(x): unknown function
dx
d2y dy
2
− 5 + 2 y = cos( x)
dx dx
x: independent variable
Notation
Throughout this lesson ordinary derivatives will be written by using either the Leibniz notation
𝒅𝒚 𝒅𝟐 𝒚 𝒅𝟑 𝒚
, , 𝒅𝒙𝟑 , …
𝒅𝒙 𝒅𝒙𝟐
or the prime notation
𝒚′ , 𝒚′′ , 𝒚′′′ , …
Actually, the prime notation is used to denote only the first three derivatives; the fourth derivatives is
(𝟒) ′′′′ 𝒅𝒏 𝒚 (𝒏)
written 𝒚 instead of 𝒚 . In general, the nth derivative of y is written 𝒅𝒙𝒏 or 𝒚 .
You should also be aware that Newton’s dot notation is sometimes used to denote derivatives with respect
𝑑2 𝑠
to time (t) in physical sciences and engineering. For example, the differential equation 𝑑𝑡 2 = −32 becomes
𝑠ሷ = −32.
Partial derivatives (PDE) are often denoted by a subscript notation indicating the independent variables.
For example, in subscript notation, a PDE equation is written as:
𝒖𝒙𝒙 = 𝒖𝒕𝒕 − 𝟐𝒖𝒕
Classification of ODE
ODE can be classified in different ways
Order Derivatives
◦ First order ODE Ordinary DE
◦ Second order ODE Partial DE
◦ Nth order ODE (Higher order)
Linearity coefficients
◦ Linear ODE constant coefficients ODE
◦ Nonlinear ODE variable coefficients ODE
Auxiliary conditions homogeneity
◦ Initial value problems homogeneous
◦ Boundary value problems non-homogeneous
Order of ODE
➢Order of ODE is determined by highest-order derivative of solution function
appearing in ODE

➢ODE with higher-order derivatives can be transformed into equivalent first-

order system

➢We will discuss numerical solution methods only for first-order ODEs

➢Most ODE software is designed to solve only first-order equations

Order of a differential equation
The order of an ordinary differential equations is the order of the highest order
derivative

Examples :
dy
− y = ex First order ODE
dx
d2y dy
2
− 5 + 2 y = cos( x) Second order ODE
dx dx
3
 d y
2
 dy
 2
 − + 2 y 4 = 1 Second order ODE
 dx  dx
Higher-order ODEs
Example:
Newton’s second law

u1 = solution y of the original

equation of 2nd order
u2 = velocity y’
Can solve this by methods for 1st order equations
Solution of a differential equation
A solution to a differential equation is a function that
satisfies the equation.
Example :
Solution x(t ) = e −t
dx(t )
+ x(t ) = 0 Proof :
dt dx(t ) −t
= −e
dt
dx(t )
+ x(t ) = −e −t + e −t = 0
dt
 Linear ODE
An ODE is linear if the unknown function and its derivatives appear to power one.
No product of the unknown function and/or its derivatives

an ( x) y n ( x) + an −1 ( x) y n −1 ( x) +  + a1 ( x) y ' ( x) + a0 ( x) y ( x) = g ( x)
Examples :
dy 𝑑𝑦
− y = ex Linear ODE 1−𝑦 + 2𝑦 = 𝑒 𝑥 Non-linear ODE
𝑑𝑥
dx
(Non-linear term)
d2y dy
− 5 + 2 x 2
y = cos( x) Linear ODE 𝑑2𝑦
+ 𝑠𝑖𝑛𝑦 = 0
dx 2
dx 𝑑𝑥 2 Non-linear ODE
3 (Non-linear function)
 d 2 y  dy
 2  − + y = 1 Non-linear ODE
 dx  dx (Non-linear function)
(Non-linear degree)
Boundary-Value and Initial value Problems
Initial-Value Problems Boundary-Value Problems

The auxiliary conditions are at

The auxiliary conditions are not at one
one point of the independent point of the independent variable
variable
More difficult to solve than initial value
problem

y ' '+2 y '+ y = e −2 x y ' '+2 y '+ y = e −2 x

y (0) = 1, y ' (0) = 2.5 y (0) = 1, y (2) = 1.5
same different
Example:
Example:

y-dependent, x-independent variable, first order, power:1,linear, non-homogeneous

y-dependent, x-independent variable, second order, power:1,linear, non-homogeneous

y-dependent, x-independent variable, first order, power:3,non-linear, non-homogeneous

y-dependent, x-independent variable, first order, power:2,non-linear, non-homogeneous

y-dependent, x-independent variable, fourth order, power:1,linear, homogeneous

Example:

Fourier Dif. Eq., used for heat transfer, first order, power:1,

T dependent, x and y independent variables, Partial Dif. Eq.

s and r dependent, t independent variables, Ordinary Dif. Eq.

u and v dependent, x and y independent variables, Partial Dif. Eq.

Solutions
Analytical Solutions to ODE are available for linear ODE and special classes of
nonlinear differential equations.

Numerical method are used to obtain a graph or a table of the unknown

function

We focus on solving first order linear ODE and second order linear ODE and
Euler equation
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Problem and Differential Solution

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
Differential Solutions
Introduction

Engineering Problem
Definition of Problem
Solution of Problem

Physical Quantities
Experimental Study
Theoretical Study and Differential Solution
Introduction

 Engineering is a science of computation that researches to solve a problem.

 A calculation process is possible by using different functions called relation-equation-equality, which are
developed due to different methods.

 In other words, it cannot be mentioned without an equation for the solution of a problem.
Introduction
 If the relations or equations in the solution of the problem are found experimentally, they are called AMPIRIC, if
they are found by numerical methods, NUMERIC and if they are obtained by theoretical techniques,
ANALYTICAL relations.

 This course aims is to give the necessary mathematical and basic physical techniques to form the basic structure
in which THEORETICAL and NUMERICAL solutions can be applied from the mentioned equations.

 In addition, the question of how to solve equations that are the product of the basic structure, which can also
be named as solution model, will be answered.
Introduction
 In other words, engineer means "the person who has the physical and mathematical knowledge required by the
related problem to reflect the effect of an error, which is considered necessary to find a mathematically and
physically correct solution, on the result after the solution".

 From another point of view, the engineer is the person who can solve the problem in the easiest, most accurate
way possible and determine the reality of the results found.

 Such a capability can only be possible if the mathematical and physical information is fully known and their
blending.
Example:
The material point with an initial velocity of 1 m / s continues its motion with acceleration a depending
on time. Find the equation that gives the speed of the motion over time. The unit of t in the equation is a
second.

Solution: velocity from integration of acceleration over time can be found.

İntegral result: time-dependent velocity equation of motion

The solution of this integral will give an equation V = V (t) valid for any time t between 0 and ∞.

However, the given integral cannot be solved. Therefore, a relation in the form of V = V (t) will not be produced.

At this stage, an approximate solution can be considered rather than a complete solution. The first of this approximate
solution is the numerical solution of the integral that cannot be solved analytically. The second is to try to remove the
functional structure that causes the insolubility of the integral with a certain error.
Example:
The material point with an initial velocity of 1 m / s continues its motion with acceleration a depending
on time. Find the equation that gives the speed of the motion over time. The unit of t in the equation is a
second.

Instead of sin (0.25t) in the integral, about 0.25t can be written for
small 0.25t values. Thus, the integral becomes solvable. The accuracy
of this approach can be observed in the figure on the right. There is a
great agreement between the values of 0 and 3 seconds of time-t
between sin (0.25t) and 0.25t.

Since the calculation of the area under the integral curve will be in
the meantime, the area under both curves will be almost the same
between 0-3 seconds. However, the difference becomes larger for
time values greater than 3 seconds. Thus, an equation describing the
speed of the motion can be produced to be used between 0-3.
Example:
The material point with an initial velocity of 1 m / s continues its motion with acceleration a depending
on time. Find the equation that gives the speed of the motion over time. The unit of t in the equation is a
second.

The integral that cannot be solved can be solved very precisely by Numerical and analytical velocity
numerical methods. When the velocity values calculated for values (m/s) for different times
different times are compared with the numerical solution and the Time t(s) Num. V(t)
equation V (t), it will be seen that the velocity values after 3 s are
very different. The table on the right shows the speed values
calculated by numerical integration with MATHCAD and V (t)
equation. As stated, a difference occurs between the speed values
after 3 s. If the velocity is calculated from the equation V (t) for 10
seconds, a 12% error is made.
 Example:
Find the velocity and path equations of the motions that occur when a parachute and a steel ball of the same
mass are released from a height H.

Solution: The forces acting on the steel ball and the parachute are weight (W) and air resistance (FR ) forces.

Air resistance is closely related to the object's surface area (A).

Large resistance forces will act on large surfaces, small resistance forces will affect small surfaces.

In general, fluid resistive forces are proportional to the fluid or body velocity in n order (V n ).

In this case, if the Second Newton Law (F = ma) is applied for the parachute and steel ball, the acceleration (a) is found as
follows.
 Example:
Find the velocity and path equations of the motions that occur when a parachute and a steel ball of the same
mass are released from a height H.

According to the last expression, acceleration is a function of velocity. As the velocity will change during the
movement, the acceleration will change. One can pass from acceleration to velocity and from velocity to the road by
integration.

The final integration depends on the numerical value of n. The number "n" can take a value between 0.5 and 2 depending
on the body geometry, fluid type and flow type. In this case, since integration is not possible in general terms, it can be
said "there is no solution". It can be considered to neglect the resistive force, which causes the solution to be unsolved. FR
= 0 can be taken for steel balls with very small surface area. In this case, the acceleration is equal to the acceleration of
gravity and will be constant. Time-dependent speed and path equations in the new situation.
Engineering Problem

It is all of the Design, Production and Control calculations that concern the Energy, Thermodynamics, Manufacturing-
Construction, Machine Dynamics and Mechanics divisions, which are the areas of interest in mechanical engineering.

All processes such as defect detection, adjustment operations, post-production control are within the scope of a
problem.

A problem may cover more than one area, which may be related to only one of these departments. Both numerical,
experimental or both theoretical and experimental solution of a problem can be done.

Since the experiment will represent the real environment, its results will be healthier. However, experimental work is
an expensive and delicate process. For this reason, experimental study is not preferred after the accuracy of the
theory is confirmed by experimental study in the literature.

Results are sought with analytical and numerical solutions, which are more economical and easier.
Physical Quantities

All quantities with or without units (4 basic units: m; meter, °C;

degrees centigrade, s; seconds and kg; kilograms) that affect
the relevant problem. In other words, they are Constant and
Variables that will appear in the solution equation and solution
phase.

It is possible to determine these sizes by mastering the physics

of the related problem. Determining the physical magnitudes
affecting a problem will directly affect the health of the
solution. An incorrect determination will result in a difference
between the resulting and actually occurring values.
Experimental Study

In some highly complex and analytical problems, some equations used for practical calculations are found as a result
of experimental studies. These are called empirical correlations because they are found experimentally.

In order to find such a correlation, the experimenter must be an expert in the usage capacity and error-making size of
the measuring devices.

The experiment set should be a small mock-up of the real problem for the result to be as accurate as possible.

At the same time, the data found as a result of experimental measurements should be evaluated with a good
statistical information.
 Theoretical Study and Differential Solution

Theoretical work is preferred because it is cheaper than experimental work.

Generally, the outcome of a theory is tried to be verified by an experimental study.

After this verification process is performed for several areas of a problem, it is concluded that the theory put forward
gives correct results.

After determining the physical magnitudes that affect the course of the problem in the theoretical study, the role of
these variables during the validity of the problem should also be determined.

The purpose of these works is to divide the physical sizes into two groups as fixed and variable.
 Theoretical Study and Differential Solution

The basic logic of thinking of larger problems and structures as smaller problems or structures is that the solution condition
can be reached.

As a result, it is essential to divide large systems into a certain number of subsystems and divide one big problem into N
numbers of small problems.

If the solution of an unsolvable problem is sought with infinite fragmentation, each piece obtained is called a DIFFERENTIAL
ELEMENT.

In other words, it is possible to call the smallest physical structure a differential element.

However, although it is the smallest, it never has a zero value. An interpretation as the closest and smallest to zero is more
accurate.
Theoretical Study and Differential Solution

Differential element, mathematically, for time (t) (differential element or differential of time);

It can be described with an example. Fragmentation is derivation with another definition.

The smallest part of a physical quantity is its mathematically derivative (differential).

This differential element can also be called Differential Range, Differential Surface or Differential Volume depending
on the situation. In particular, it is also called Control Volume in thermodynamics and fluid mechanics.
Example:
A tank with two different geometries is considered to store a 1000 m3 hot liquid at a temperature of 300 °C at an
ambient temperature of 25 °C. The first geometry is a cylindrical structure with radius R and height H, and the second
geometry is a prismatic structure in K×L×M dimensions. What should be the dimensions of the geometries considered
in order to minimize the cooling of the hot liquid. Which geometry is suitable for the least heat loss.

Heat flows from high temperature environment to low temperature environment. Heat flow directions are indicated by
arrows in the figure.

Q=C.A.(Tin-Tout)=C.A.(300-25)
The heat transfer that occurs can be calculated simply with the above equation.
Example:
A tank with two different geometries is considered to store a 1000 m3 hot liquid at a temperature of 300 °C at an
ambient temperature of 25 °C. The first geometry is a cylindrical structure with radius R and height H, and the second
geometry is a prismatic structure in KxLxM dimensions. What should be the dimensions of the geometries considered
in order to minimize the cooling of the hot liquid. Which geometry is suitable for the least heat loss.

Here C is a constant value of the storage material and medium. A is the surface to which heat is transferred. In this
case, only surface A can be played with for the least amount of heat loss. Since Q and A will be directly
proportional, A must be the smallest for the least heat transfer. The smallest A surface that gives a volume of 1000
m3 should be searched.
Example:
A tank with two different geometries is considered to store a 1000 m3 hot liquid at a temperature of 300 °C at an
ambient temperature of 25 °C. The first geometry is a cylindrical structure with radius R and height H, and the second
geometry is a prismatic structure in KxLxM dimensions. What should be the dimensions of the geometries considered
in order to minimize the cooling of the hot liquid. Which geometry is suitable for the least heat loss.

Solution for cylindrical structure:

VOLUME
The derivative must be zero to find the smallest value of Amin . The change of Amin with respect to R is given in the
figure. According to this figure, R has a minimum value of surface area A around 5 m.
Example:
A tank with two different geometries is considered to store a 1000 m3 hot liquid at a temperature of 300 °C at an ambient
temperature of 25 °C. The first geometry is a cylindrical structure with radius R and height H, and the second geometry is
a prismatic structure in KxLxM dimensions. What should be the dimensions of the geometries considered in order to
minimize the cooling of the hot liquid. Which geometry is suitable for the least heat loss.

Solution for prismatic structure:

VOLUME

Unlike the previous structure, A could be obtained here as a function of both L and M. It was changing only as
a function of R for the cylinder. If A = A (M, L) its differential is as follows.

To find the smallest value of A, its derivative (dA) must be

zero. In this, the partial derivatives of A with respect to L and
M must be zero.
The volume of 1000 m3 is provided with a lower
surface area with a cylindrical structure. The surface
area of the cylinder (553.58 m2 ) is smaller than the
= 1000,
surface area of the prism (600 m2 ). By choosing the
cylindrical structure, another 7.7% of heat leakage can
m m be saved.
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Differential Equations Solutions

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
Differential Equation Solutions
1- The Method of integration

2-Checksum Principle

3-Undetermined Coefficients Method

4- Decreasing the order of the equation

5- Variation of Parameters
1- The Method of integration:

The basic method of solving a differential equation is integration.

At least two integrals should be taken for the solution, one on the dependent and one on the independent variable.

In principle, dependent and independent variables cannot be integrated over each other.

In order for the differential equation to be integrated, the variables must be separable.

Differential equations that can be separated are called differential equations that can be separated into variables.
Example: 𝑑2 𝑢
= 100 2. order, linear, non-homogenious UH + UP = U
𝑑𝑥 2
It can be separated into its variables and integrated directly.
UH UP
İt is multiplied side by side with ‫𝑥𝑑 ׬‬.

𝑑2 𝑢 𝑑 𝑑𝑢 𝑑𝑢
඲ 2 𝑑𝑥 = න ⋅ 𝑑𝑥 = න𝑑 = න100 𝑑𝑥
𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑑𝑥

𝑑𝑢 100𝑥 2
= 100𝑥 + 𝑐1 ⇒ 𝑢 = + 𝑐1 𝑥 + 𝑐2
𝑑𝑥 2

UP UH
Example: 𝑑4 𝑤
= 16si n 2 𝑡 − 𝑡 4. order, linear, non-homogenious, It can be separated into its variables
dt 4

wH wP
𝑑3 𝑤 𝑡2
= −8co s 2 𝑡 − + 𝑐1
𝑤 = 𝑐1 𝑓1 (𝑡) + 𝑐2 𝑓2 (𝑡) + 𝑐3 𝑓3 (𝑡) + 𝑐4 ⋅ 𝑓4 t + 𝑓5 t 𝑑𝑡 3 2
Taking 4 consecutive 𝑑2 𝑤 𝑡3
wH wP integrals: = −4si n 2 𝑡 − + 𝑐1 𝑡 + 𝑐2
𝑑𝑡 2 6
𝑑𝜔 𝑡 4 𝑐1 𝑡 2
= 2co s 2 𝑡 − + + 𝑐2 𝑡 + 𝑐3
𝑑𝑡 24 2
Indeterminate constants (𝑐1 , 𝑐2, 𝑐3, 𝑐4) can be used in
𝑡5 𝑐1 𝑡 3 𝑐2 𝑡 2
place of certain constants. 𝑤 = 𝜔(𝑡) = si n 2 𝑡 − + + + 𝑐3 𝑡 + 𝑐4
120 6 2
𝑐1 𝑐2
≡ 𝑐1 , = 𝑐2 ….
6 2

𝑡5 𝑐1 𝑡 3 𝑐2 𝑡 2
𝑤 = 𝜔(𝑡) = si n 2 𝑡 − + + + 𝑐3 𝑡 + 𝑐4
120 6 2

wP wH
Example: 𝑑𝑦 y:dependent, x: independent variable
= 1− 𝑥2 1. order, non-homogenious, it can be separated into its variables
𝑑𝑥

𝑑𝑥 1
= x:dependent, y: independent variable
𝑑𝑦 1 − 𝑥2

න𝑑𝑦 = න 1 − 𝑥 2 𝑑𝑥 𝑥 = sin 𝜃 ⇒ dx = cos𝜃𝑑𝜃 1 − 𝑥 2 = 1 − sin2 𝜃 = cos2 𝜃

1
න𝑑𝑦 = න( 1 + cos2𝜃)𝑑𝜃
නd y = න cos2 𝜃 co s 𝜃 𝑑𝜃 = නcos2 𝜃𝑑𝜃 2

1 𝑠𝑖𝑛2𝜃 sin 2𝜃 = 2sin𝜃 ⋅ cos𝜃

𝑦= 𝜃+ +𝐶
2 2 sin 𝜃 = 𝑥

cos 𝜃 = 1 − 𝑥2

1
𝑦= arcsi𝑛𝑥 + 𝑥 1 − 𝑥 2 + 𝑐
2
Example: 2
𝑑2 𝑢 𝑤0 𝑑𝑢 2. order,non-linear, non-homogenious, it can be separated into its variables
= 1 +
𝑑𝑥 2 𝑇0 𝑑𝑥
𝑑𝑢 𝑑𝑢
= tan𝜃 → 𝑑 = sec 2 𝜃𝑑𝜃
𝑑𝑢 𝑑𝑥 𝑑𝑥
2 𝑑
𝑑 𝑑𝑢 𝜔0 𝑑𝑢 𝑤0 𝑑𝑥
= 1+ න 𝑑𝑥 = ඳ
𝑑𝑥 𝑑𝑥 𝑇0 𝑑𝑥 𝑇0 1 + 𝑑 u Τ𝑑 𝑥 2
𝑑𝑢
2
1+ = 1 + tan2 𝜃 = sec 2 𝜃
𝑑𝑥
𝑤0 sec 2 𝜃 𝑑𝜃 d𝜃
x=඲ = නs ec𝜃𝑑𝜃 = න
𝑇0 sec 𝜃2 cos 𝜃

𝑤0 1 + si𝑛𝜃
𝑥 = ln + 𝐶1
𝑇0 1 − si𝑛𝜃
 𝑑2 𝑦
Example: − 5𝑦 = 100 2. order, linear, non-homogenious, it can not be separated into its
𝑑𝑥 2
variables
yH yP

𝑑2 𝑦
඲ 2 𝑑𝑥 = 5 නy dx = 100 න𝑑 𝑥
𝑑𝑥
100x
𝑑𝑦
𝑦′ =
𝑑x
2-Checksum Principle:

The calculated solution itself and its derivatives are written in the equation and after the necessary
simplifications to obtain the 0 = 0 result, the solution is called the checksum differential equation or the
principle of checksum.

This principle is used in practice for two purposes.

In order to determine the correctness of the solution found in the first, the solution itself and its derivatives
are written in place.

The second is used to calculate solution functions in equations that are not separated into variables.

The functions selected by trial and error method and those that give a result of 0 = 0 by writing instead of
themselves and their derivatives are evaluated as solutions.
Example: 𝑑2 𝑦 𝑑𝑦
+ 4 + 4𝑦 = 0 Which of the following is the solution of the equation?
𝑑𝑥 2 𝑑𝑥
𝒊) 𝑦 = 𝑐1 sin 2𝑥 + 𝑐2 cos 2𝑥,
𝑑𝑦
𝑦 = 𝑐1 sin 2𝑥 + 𝑐2 cos 2𝑥 = 𝑦 ′ = 2𝑐1 cos 2𝑥 − 2𝑐2 sin 2𝑥
𝑑𝑥
𝑦 = 𝑐1 + 𝑐2 𝑥 ⋅ 𝑒 −2𝑥
𝑑2 𝑦
2 = 𝑦 ′′ = −4𝑐1 sin 2𝑥 − 4𝑐2 cos 2𝑥
𝑑𝑥
−41 𝑐1 sin2𝑥 − 4𝑐2 cos2𝑥 + 4 2𝑐1 cos 2𝑥 − 2𝑐2 sin 2𝑥 + 4 𝑐1 sin 2𝑥 + 𝑐2 cos 2𝑥 = 0 ⇒ 8𝑐1 cos2𝑥 − 8𝑐2 sin2𝑥 ≠ 0
(The 0 = 0 condition did not occur.)
𝑦 = 𝑐1 sin2𝑥 + 𝑐2 cos2𝑥, (it is not solution equation or the mistake can be made.)

𝒊𝒊) 𝑦 = 𝑐1 + 𝑐2 𝑥 ⋅ 𝑒 −2𝑥 𝑦 ′′ + 4𝑦 ′ + 4𝑦 = 0
−4𝑐2 𝑒 −2𝑥 + 4 𝑐1 + 𝑐2 𝑥 𝑒 −2𝑥 + 4 𝑐2 𝑒 −2𝑥 − 2 𝑐1 + 𝑐2𝑥 𝑒 −2𝑥 +
𝑦′ = 𝑐2 𝑒 −2𝑥 − 2 𝑐1 + 𝑐2 𝑥 𝑒 −2𝑥
+4 𝑐1 + 𝑐2 𝑥 𝑒 −2𝑥 = 0 ⇒ 0 = 0
𝑦 ′′ = −2𝑐2 𝑒 −2𝑥 +4𝑐1 𝑒 −2𝑥 − 2𝑐2 𝑒 −2𝑥 +4𝑐2 𝑥𝑒 −2𝑥 ⇒ 0 = 0, (it is a solution equation.)
𝑦 ′′ = −4𝑐2 𝑒 −2𝑥 + 4 𝑐1 + 𝑐2 𝑥 𝑒 −2𝑥
 𝑑2 𝑦
Example: + 9𝑦 = 0 Which of the following are the solution functions of the equation?
𝑑𝑥 2

a) 𝑦 = 𝑥 2 b) 𝑦 = 𝑒 𝑥 c) 𝑦 = cos 𝑥 d) 𝑦 = cos 3𝑥

Solution: 𝑦 = 𝑐1 𝑓1 (𝑥) + 𝑐2 𝑓2 (𝑥) ⇒ This pattern is not present in nonlinear equations.

a) 𝑦 = 𝑥 2 −−−⇒ 𝑦 ′ = 2𝑥 −−−⇒ 𝑦 ′′ = 2 𝑦 ′′ + 9𝑦 = 0, 2 + 9𝑥 2 ≠ 0 ⇒ 0 = 0 not occur, no solution for 𝑦 = 𝑥 2

b) 𝑦 = 𝑒 𝑥 ⇒ 𝑦 ′′ = 𝑦 ′ = 𝑦 = 𝑒 𝑥 ⇒ 𝑦 ′′ + 9𝑦 = 0 ⇒ 𝑒 𝑥 + 9𝑒 𝑥 = 0 ⇒ 10𝑒 𝑥 ≠ 0 0 = 0 not ocur, no solution for 𝑦 = 𝑒 𝑥

c) 𝑦 = cos 𝑥 => 𝑦 ′ = − sin 𝑥 ⇒ 𝑦 ′′ = − cos 𝑥 ⇒ 𝑦 ′′ + 9𝑦 = 0 ⇒ − cos 𝑥 + 9 cos 𝑥 = 0 ; 8 cos 𝑥 ≠ 0 ⇒

0 = 0 not ocur, no solution for 𝑦 = cos 𝑥

d) 𝑦 = cos 3𝑥 ⇒ 𝑦 ′ = −3 sin 3𝑥 ⇒ 𝑦 ′′ = − 9cos 3𝑥 ⇒ 𝑦 ′′ + 9𝑦 = 0 ⇒ −9 cos 3𝑥 + 9 cos 3𝑥 = 0 ⇒ 0 = 0

The condition of 0 = 0 is provided, cos 3𝑥 is a solution.
𝑓1 𝑥 = cos 3𝑥 ⇒ 𝑓2 (𝑥) = sin 3𝑥

General solution is: 𝑦 = 𝑐1 cos 3𝑥 + 𝑐2 sin 3𝑥

 𝑑2 𝑥 𝑑𝑥 𝑥 = 𝑥𝐻 + 𝑥𝑝 = 𝑐1 𝑓1 (𝑡) + 𝑐2 𝑓2 (𝑡) + 𝑓3 (𝑡 ൯
Example: − 10 + 3𝑥 = 𝑒 2𝑡 𝑓3 (𝑡) ≈ 𝑒 2𝑡
𝑑𝑡 2 𝑑𝑡
xP
xH xP e itself, its derivatives and
𝑒𝑡 𝑒 −𝑡 integrals are equal to each
Is it 𝑥𝑝 = −3𝑒 2𝑡 ? other.

𝑑𝑥𝑃 𝑑2 𝑥𝑃 𝑑2 𝑥𝑝 𝑑𝑥𝑝
= −6𝑒 2+ , = −12𝑒 2𝑡 If x ---> 𝑥𝑃 − 10 + 3𝑥𝑝 = 𝑒 2𝑡
𝑑𝑡 𝑑𝑡 2 𝑑𝑡 2 𝑑𝑡

−12𝑒 2𝑡 − 10 −6𝑒 2𝑡 + 3 −3𝑒 2𝑡 = 𝑒 2𝑡

−12𝑒 2𝑡 + 60𝑒 2𝑡 − 9𝑒 2𝑡 = 𝑒 2𝑡 39𝑒 2𝑡 = 𝑒 2𝑡 , 39 = 1

No, 0 = 0

It is not 𝑥𝑝 = −3𝑒 2𝑡
3-Undetermined Coefficients Method:

It is a proposed method to calculate non-homogeneous solutions in linear equations.

For this method to be valid, the function that causes non-homogeneity must be a function that simplifies as it is derived.

The functions that satisfy this condition are polynomials, exponential functions, sinus and cosine functions.

The non-homogeneous solution for functions such as division functions and logarithms tan sec cannot be calculated by
this method.

In this method, the non-homogeneous solution is found by analogy with itself and its derivatives.

A solution format is proposed by multiplying the inhomogeneous solution itself and its derivatives by unknown
coefficients.
According to the checksum principle, uncertain coefficients are determined as numerical.
Example: 𝑑2 𝑦 𝑑𝑦 𝑓(𝑥) = 𝑒 2𝑥
2 − 10 + 3𝑦 = 26𝑒 2𝑥
𝑑𝑥 𝑑𝑥 𝑓 ′ (𝑥) = 2𝑒 2𝑥 𝑦𝑝 = 𝐴𝑒 2𝑥
𝑓 ′′ (𝑥) = 4𝑒 2𝑥
yH yP ,Non- 𝑓 ′′′ (𝑥) = 8𝑒 2𝑥
homogeneous Undetermined
function cooeficient

𝑦𝑝 = A𝑒 2𝑥
𝑑2 𝑦 𝑑𝑦
𝑦𝑝 ′ = 2𝐴𝑒 2𝑥 − 10 + 3𝑦 = 26𝑒 2𝑥 4A𝑒 2𝑥 − 10(2𝐴𝑒 2𝑥 ) + 3𝐴𝑒 2𝑥 = 26𝑒 2𝑥
𝑑𝑥 2 𝑑𝑥
𝑦𝑝 ′′ = 4A𝑒 2𝑥
−13A𝑒 2𝑥 = 26𝑒 2𝑥 A = −2

𝑦𝑝 = −2𝑒 2𝑥
Example: 𝑑2 𝑦 𝑑𝑦 𝑓(𝑥) = 10𝑥 2
2 − 10 + 3𝑦 = 6𝑥 2
𝑑𝑥 𝑑𝑥 𝑓 ′ (𝑥) = 20𝑥 A,B and C, undetermined
𝑦𝑝 = 𝐴𝑥 2 + 𝐵𝑥 + 𝐶
𝑓 ′′ (𝑥) = 20 coeficients
yH 𝑓 ′′′ (𝑥) = 0
yP

𝑦𝑝 = 𝐴𝑥 2 + 𝐵𝑥 + 𝐶 𝑦𝑝 ′ = 2𝐴𝑥 + 𝐵 𝑦𝑝 ′′ = 2𝐴

𝑑2 𝑦 𝑑𝑦
− 10 + 3𝑦 = 6𝑥 2 2𝐴 − 10 2𝐴𝑥 + 𝐵 + 3 𝐴𝑥 2 + 𝐵𝑥 + 𝐶 = 6𝑥 2
𝑑𝑥 2 𝑑𝑥
3𝐴𝑥 2 + (3𝐵 − 20𝐴)𝑥 + 2𝐴 − 10𝐵 + 3𝐶 = 6𝑥 2 ⟹ 3𝐴𝑥 2 = 6𝑥 2 ⟹ 2𝐴 = 6
𝐴 =2
=0 =0

10𝐵 − 2𝐴 388 20𝐴 40

2𝐴 − 10𝐵 + 3𝐶 = 0 ⟹ 𝐶 = ⟹𝐶= 3𝐵 − 20𝐴 = 0 ⟹ 𝐵 = =
3 9 3 3

𝟒𝟎 𝟑𝟖𝟖
𝒚𝒑 = 𝟐𝒙𝟐 + 𝒙+
𝟑 𝟗
Example: 𝑦 ′′ − 10𝑦 ′ + 3𝑦 = l n 𝑥

𝑓(𝑥) = 𝑙 𝑛 𝑥
1
𝑓 ′ (𝑥) =
𝑥
1
𝑓 ′′ (𝑥) = − 2 There are functions that go to infinity and the non-homogeneous solution cannot be
𝑥 calculated by this method.
2
𝑓 ′′ (𝑥) = 3
𝑥
−6
𝑓 ′ (𝑥) = − 4
𝑥
4- Decreasing the order of the equation :

One of the solution methods of differential equations is to decrease the order of the equation.

It is primarily used as a solution method of higher order non-linear equations.

The condition of decreasing orders should not appear in the equation as a function of either dependent or
independent variables.

Decreasing the order of the equation is in a way a transformation of variables.

A chain derivative rule can be used to reduce the number of variables during operations.
Example: 𝑥𝑦 ′′ = 𝑦 ′ + 𝑥 𝑦 ′ 2

2 2. order, non-linear, homogeneous dif. Equ.

𝑑2 𝑦 𝑑𝑦 𝑑𝑦 y: dependent, x: independent variable
𝑥 = +𝑥
𝑑𝑥 2 𝑑𝑥 𝑑𝑥

𝑑2𝑦 𝑑𝑦 𝑑𝑦 2
𝑑𝑚 𝑑𝑥
If the equ. is such as 𝑦 ⋅ = +𝑥 න =න
𝑑𝑥 2 𝑑𝑥 𝑑𝑥 𝑚 + 𝑥𝑚2 𝑥
The equation would be unsolvable because the order would
not decrease. ‘𝑥’ cannot be integrated over 𝑚.
or
There is no 𝑦 in the equation.
There is 𝑦’s derivatives, order can decrease. 𝑑𝑚 1 + 𝑥𝑚
න =න 𝑑𝑥
𝑚 𝑥
𝑑𝑦 𝑚 : new dependent variable
=𝑚 𝑦 : old dependent variable ‘𝑚’ cannot be integrated over 𝑥.
𝑑𝑥
𝑦 = 𝑦 𝑥 ⇒ 𝑚 = 𝑚(𝑥) This equation cannot be divided into variables.
It can be solved by a special transformation
𝑑𝑦 𝑑2 𝑦 𝑑𝑚 known as the bernoulli equation.
=𝑚 ⇒ =
𝑑𝑥 𝑑𝑥 2 𝑑𝑥

𝑑𝑚
𝑥⋅ = 𝑚 + 𝑥𝑚2 1. order,non-linear,
𝑑𝑥 homogeneous
Example: 𝑥𝑦 ′′ = 𝑦 ′ + 𝑦 ′ 2

2
2. order,non-linear, homogeneous dif. Equ.
𝑑2 𝑦
𝑑𝑦 𝑑𝑦 y: dependent, x: independent variable
𝑥 2= +
𝑑𝑥 𝑑𝑥 𝑑𝑥

There is no 𝑦 in the equation. ⟹ 1 = 𝐴(1 + 𝑚) + 𝐵𝑚 𝐴 = 1 and 𝐵 = −1

There is 𝑦’s derivatives, order can decrease.
1 1 𝑑𝑥
𝑦 = 𝑦 𝑥 ⇒ 𝑚 = 𝑚(𝑥) ඲ − 𝑑𝑚 = න
𝑚 1+𝑚 𝑥

𝑑𝑦 𝑑2 𝑦 𝑑𝑚 ln 𝑚 − ln 1 + 𝑚 = ln(x) + ln(𝑐)
=𝑚 ⇒ =
𝑑𝑥 𝑑𝑥 2 𝑑𝑥
For functional accordance, ln (c), sin (c), c 2 or ec
𝑑𝑚 can be written instead of c.
𝑥⋅ = 𝑚 + 𝑚2 1. order,non-linear, 𝑚 𝑚 𝑐𝑥
𝑑𝑥 homogeneous ln = ln(cx) ⟹ = 𝑐𝑥 ⟹ 𝑚 =
1+𝑚 1+𝑚 1−𝑐𝑥
𝑑𝑦 𝑐𝑥
𝑑𝑚 𝑑𝑥 1 𝐴 𝐵 𝑚= ⟹ ‫ = 𝑦𝑑 ׬‬න( )𝑑𝑥 ⟹ 𝑢 = 1 − 𝑐𝑥 ⟹
න
𝑚 + 𝑚2
=න
𝑥
⟹ = +
𝑚 + 𝑚2 𝑚 1 + 𝑚
⟹ 𝑑𝑥 1−𝑐𝑥
1 1−𝑢
𝑑𝑢 = −𝑐𝑑𝑥 ⟹ ‫ = 𝑦𝑑 ׬‬− න( )𝑑𝑢 ⟹
𝑐 𝑢
The equation is separated into its variables. 𝟏
𝒚 = − 𝐥𝐧 𝟏 − 𝒄𝒙 − 𝟏 + 𝒄𝒙 + 𝒄𝟏
𝒄
Example: 𝑦 ′′ + 2𝑦 𝑦 ′ 3 =0

3
2. order,non-linear, homogeneous dif. Equ.
𝑑2 𝑦 𝑑𝑦 y: dependent, x: independent variable
+ 2𝑦 =0
𝑑𝑥 2 𝑑𝑥

1 1 𝑑𝑦
⟹ 𝑚 = y 2 + 𝑐1 ⟹ 𝑚 = y2+𝑐1 = 𝑑𝑥 ⟹
There is no 𝑥 in the equation.
There is 𝑦’s derivatives, order can decrease. 𝑑𝑥 = (y 2 +𝑐1 )𝑑𝑦 ⟹
𝑑𝑦 𝑑2 𝑦 𝑑𝑚 y3
=𝑚 ⇒ = 𝑥= + 𝑐1 𝑦 + 𝑐2
𝑑𝑥 𝑑𝑥 2 𝑑𝑥 3
𝑑𝑚 𝑑𝑚 𝑑𝑦 𝑑𝑚 There is 3 variables (x, y, m), Since the equation is non-linear, the
= = 𝑚
𝑑𝑥 𝑑𝑦 𝑑𝑥 𝑑𝑦 the equation can’t be solved. result is found as 𝑥 = 𝑥 (𝑦).
𝑑𝑚 One of the variables is
𝑚 + 2y𝑚3 = 0 removed with Chain rule.
𝑑𝑦
𝑑𝑚 1
න 2 = න−2𝑦𝑑𝑦
𝑚
⟹ − = −y 2 − 𝑐1
𝑚
⟹

The equation is separated into its variables.

 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Dif. Equ. Solutions and App. Dif. Equ.

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
Differential Equation Solutions
1- The Method of integration

2-Checksum Principle

3-Undetermined Coefficients Method

4-Decreasing the order of the equation

5-Variation of Parameters
5- Variation of parameters :

This method is used to solve differe intial equations generating a new solution over a specified solution.

It is primarily used as a solution method of higher order non-linear equations.

The condition of decreasing orders should not appear in the equation as a function of either dependent or
independent variables.

Decreasing the order of the equation is in a way a transformation of variables.

A chain derivative rule can be used to reduce the number of variables during operations.
Example:
𝑦 ′′ + 4𝑦 ′ + 4y = 16𝑒 −2x
𝑦𝐻 𝑦𝑃 2. order,non-linear, nonhomogeneous dif. Equ.
y: dependent, x: independent variable

Homogeneous solution (yH): 𝑐(𝑥) is calculated with Checksum principle and Variation of parameters ;
With trial and error method
𝑑𝑦
and checksum principle; y=𝑐(𝑥)𝑒 −2x ⟹ = 𝑦 ′ =𝑐 ′ 𝑒 −2x -2𝑐𝑒 −2x ⟹ 𝑦 ′ =(𝑐 ′ −2c)𝑒 −2x ⟹
𝑑𝑥
y = 𝑒 −2x 𝑦 ′ = −2𝑒 −2x 𝑦 ′′ = 4𝑒 −2x
𝑑𝑦 𝑑𝑐
𝑦′ = = ( −2c)𝑒 −2x
4𝑒 −2x + 4 −2𝑒 −2x + 4𝑒 −2x = 0 𝑑𝑥 𝑑𝑥
𝑑 2 𝑦 𝑑 2𝑐 𝑑𝑐
8𝑒 −2x − 8𝑒 −2x = 0 ⟹ 0 = 0 𝑦 ′′ = 2 = ( 2 −4 +4c)𝑒 −2x
𝑑𝑥 𝑑𝑥 𝑑𝑥
𝑦𝐻 = 𝑐1 𝑓1 𝑥 + 𝑐2 𝑓2 (𝑥) If these derivatives are substituted in the 𝑦 ′′ + 4𝑦 ′ + 4y =0equation,
𝑦1 𝑦2
𝑑2𝑐 𝑑𝑐 𝑑𝑐
𝑓1 𝑥 =𝑒 −2x 𝑓2 𝑥 =? ( 2 −4 +4𝑐)𝑒 −2x+( −2𝑐)𝑒 −2x+4𝑐𝑒 −2x=0
𝑑𝑥 𝑑𝑥 𝑑𝑥

𝑦1 =𝑐𝑒 −2x ⟹ 𝑦2 =𝑐(𝑥)𝑒 −2x 𝑑 2 𝑐 −2x

𝑒 =0⟹ 𝑒 −2x ≠ 0 ve
𝑑2 𝑐
=0⟹
𝑑 𝑑𝑐
( )=0 ⟹
𝑑𝑐
=1⟹𝑐=𝑥
𝑑𝑥 2 𝑑𝑥 2 𝑑𝑥 𝑑𝑥 𝑑𝑥
Constant Dependent on independent variable
 ⟹ 𝑐 𝑥 = 𝑥 ⟹ 𝑦2 =𝑥𝑒 −2x 𝑦𝑃 =𝑐 𝑥 𝑒 −2x and 𝑦𝑃 =𝑐 𝑥 𝑥𝑒 −2x

𝒚𝑯 = 𝒄𝟏 𝒆−𝟐𝒙 + 𝒄𝟐 𝒙𝒆−𝟐𝒙 = (𝒄𝟏 +𝒄𝟐 𝒙)𝒆−𝟐𝒙 will be preferred

𝑦𝑃 ′ = (𝑐 ′ −2c)𝑒 −2x ⟹ 𝑦𝑃 ′′ = (𝑐 ′′ −4𝑐 ′ + 4𝑐)𝑒 −2x
Nonhomogeneous solution (yP):
(𝑐 ′′ −4𝑐 ′ + 4𝑐)𝑒 −2x + 4(𝑐 ′ −2c) 𝑒 −2x+4c𝑒 −2x =16𝑒 −2x
If 𝑦𝑃 is substituted in the 𝑦 ′′ + 4𝑦 ′ + 4y = 16𝑒 −2x equation,
c terms must absolutely cancel out each other in the
′′ ′
𝑦𝑃 + 4𝑦𝑃 + 4𝑦𝑃 = 16𝑒 −2x Variation of parameters method. Otherwise there is
an error.
𝑓(𝑥) = 𝑒 −2x There is only 𝑒 −2x . By the Undetermined
Coefficients Method(UCM) 𝑑2𝑐 𝑑𝑐
𝑓 ′ (𝑥) = −2𝑒 −2x 𝑐 ′′ 𝑒 −2x=16𝑒 −2x ⟹ ′′
𝑐 = 2=16⟹ = 16x⟹ 𝑐 = 8𝑥 2
𝑑𝑥 𝑑𝑥
𝑓 ′′ (𝑥) = 4𝑒 −2x 𝑦𝑃 = 𝐴𝑒 −2x ⟹ 𝑦𝑃 ′ = −2𝐴𝑒 −2x Since it is not required, the integral constant may not
𝑦𝑃 ′′ = 4𝐴𝑒 −2x be taken.

𝒚𝑷=𝟖𝒙𝟐 𝒆−𝟐𝒙
4𝐴𝑒 −2x − 8𝐴𝑒 −2x + 4𝐴𝑒 −2x = 16𝑒 −2x
0 = 16𝑒 −2x 𝒚=(𝒄𝟏 +𝒄𝟐 𝒙)𝒆−𝟐𝒙 + 𝟖𝒙𝟐 𝒆−𝟐𝒙
It cannot be solved with UCM. By Variation of parameters There are 2 different constants (𝑐1 ,𝑐2) and 3 different
method over 𝑦𝐻 ; solution functions (𝑒 −2𝑥 ,x𝑒 −2𝑥 ,𝑥 2 𝑒 −2𝑥 ). The result is
𝑦𝐻 = 𝑐1 𝑒 −2x + 𝑐2 𝑥𝑒 −2x correct.
2. Applications of Differential Equations
2.1. Separable differential equations and Reduction to Separable Variable Form
Equations that can be solved by direct integral are called variables separable differential equations.

Although it cannot be separated into variables, some differential equations can be separated into variables with
variable transformations.
Generally, transformations that reduce two variables to one variable are used.
In practice, many ordinary differential equations can be of the form

𝒈 𝒚 𝒚′ = 𝒇(𝒙)

If we integrate the left and right sides of this expression with respect to x, we get
𝑑𝑦
න𝑔 𝑦 𝑑𝑥 = න 𝑓(𝑥) 𝑑𝑥 + 𝑐 ⟹ න 𝒈 𝒚 𝒅𝒚 = න 𝒇(𝒙) 𝒅𝒙 + 𝒄
𝑑𝑥
 General Examples
Ex:1
𝑑𝑦
9𝑦 + 4𝑥 = 0 solve the given differantial equation.
𝑑𝑥

9𝑦 2 𝐶 𝑥 2 𝑦2
න9 𝑦𝑑𝑦 = − න4 𝑥𝑑𝑥 ⟹ 2
= −2𝑥 2 + 𝐶 ⟹ 𝐶1 =
18 ⟹ 9
+
4
= 𝐶1

Checksum Prencible:

𝑥 2 𝑦2 2𝑥 2𝑦𝑦 ′
If the solution is equal to + = 𝐶1 the first derivative of the solution is equal to + =0
9 4 9 4
4𝑥 + 9𝑦𝑦 ′
36
=0 ⟹ 4𝑥 + 9𝑦𝑦 ′ = 0
Ex:1

𝑑𝑦
9𝑦 + 4𝑥 = 0 Solve the given differantial equation.
𝑑𝑥

Checksum Prencible:
8𝑥
𝑥2 𝑦2 36𝐶1 − 4𝑥 2 − −4𝑥
→ 𝑦′ = 9
9
+
4
= 𝐶1 ⟹ 𝑦= 9 36𝐶1 − 4𝑥 2
=
36𝐶1 − 4𝑥 2
2 9
9 9

𝑑𝑦 36𝐶1 − 4𝑥 2 −4𝑥
9𝑦 + 4𝑥 = 0 ⟹ 9 + 4𝑥 = 0 → −4𝑥 + 4𝑥 = 0 → 0 = 0
𝑑𝑥 9 36𝐶1 − 4𝑥 2
9
9
Ex:2

d3 y
3 = 5e−2x Solve the given differantial equation.
dx

𝑑3 𝑦 𝑑 𝑑2 𝑦 −2𝑥
𝑑2 𝑦 −2𝑥
𝑑2 𝑦 −5 −2𝑥
= = 5𝑒 ⇒ න𝑑 = න5 𝑒 𝑑𝑥 ⇒ 2 = 𝑒 + 𝐶1
𝑑𝑥 3 𝑑𝑥 𝑑𝑥 2 𝑑𝑥 2 𝑑𝑥 2

d dy −5 −2x dy −5 −2x dy 5 −2x

= e + C1 ⇒ නd =඲ e + C1 dx ⇒ = e + C1 x + C2
dx dx 2 dx 2 dx 4

5 −2x
නd𝑦 = ඲ e + C1 x + C2 dx
4

𝟓 −𝟐𝒙 𝒙𝟐
𝒚=− 𝒆 + 𝑪𝟏 + 𝑪𝟐 𝒙 + 𝑪𝟑
𝟖 𝟐
Ex:3
𝑑𝑦
+ 5𝑥 4 𝑦 2 = 0 Solve the given equation with the condition of y(0) = 1
𝑑𝑥

𝑑𝑦 1
න 2 = −5 න𝑥 4 𝑑𝑥
𝑦
⟹ − = −𝑥 5 + 𝐶
𝑦 ⟹ x = 0 𝑦=1 𝑦(0) = 1  C = −1

1
𝑦= 5
𝑥 +1
Ex:4
𝑑𝑦 𝑥
= Solve the given equation with the condition of y(1) = 3
𝑑𝑥 𝑦

𝑦2 𝑥 2 33 12
න𝑦 𝑑𝑦 = න𝑥 𝑑𝑥 ⟹ 2
=
2
+𝐶 ⟹ 𝑥=1  𝑦=3 [𝑦(1) = 3ሿ ⟹ 2
=
2
+𝐶  C=4

𝑦2 = 𝑥 2 + 8
Ex:5
𝑑𝑦
= 𝑘𝑦 Solve the given equation with the condition of y(0) = 2
𝑑𝑥

𝑑𝑦
න
𝑦
= න𝑘 𝑑𝑥 ⟹ l n 𝑦 = kx + lnC ⟹ lny = kx + l n 𝐶 ⟹ y = 𝐶ekx

x=0 y=2 [y(0) = 2ሿ ⟹ 𝐶1 = 2 ⟹ 𝑦 = 2𝑒 𝑘𝑥

Ex:6 Ex:7

Ex:8 Ex:9
Ex:10 Ex:11

Ex:12 Ex:13
Ex:14 Ex:15
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: App. Dif. Equ. and Examples

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
 General Examples - Continue
Ex:16
𝑥𝑦𝑑𝑥 − 𝑥 + 2 𝑑𝑦 = 0 Solve the given differantial equation.

𝑑𝑦 𝑥𝑦 𝑑y 𝑥
𝑥y𝑑𝑥 = 𝑥 + 2 𝑑𝑦 ⟹ = ⟹ = 𝑑𝑥 It can be separated into variables.
𝑑𝑥 𝑥 + 2 𝑦 𝑥+2

𝑥 + 2 = 𝑚 ⟹ 𝑑𝑥 = 𝑑𝑚

𝑑𝑦 m−2 2
න =඲ 𝑑𝑚 = ඲ 1 − 𝑑𝑚 ⟹ 𝑙𝑛𝑦 = 𝑚 − 2𝑙𝑛𝑚 + 𝑙𝑛𝑐 ⟹ 𝑚 = 𝑙𝑛𝑦 + 𝑙𝑛𝑚2 − 𝑙𝑛𝑐
y m m

y𝑚2 𝑚
y𝑚2 𝑐𝑒 𝑚 𝑐𝑒 x+2 2
𝑒x
𝑚 = 𝑙𝑛 ⟹𝑒 = ⟹𝑦= 2 ⟹𝑦= ⟹ 𝑦 = 𝑐𝑒
c c 𝑚 (x + 2)2 (x + 2)2

𝒆𝒙
𝒚 = 𝒄𝟏
(𝐱 + 𝟐)𝟐
Ex:17

𝑦 2 𝑒 2𝑥 𝑑𝑥 = 4 + 𝑒 2𝑥 𝑑𝑦 Solve the given differantial equation.

d𝑦 𝑒 2𝑥
= 𝑑𝑥 It can be separated into variables.
𝑦 2 4 + 𝑒 2𝑥

d𝑦 𝑒 2𝑥 2𝑥 2𝑥
𝑑𝑢
න 2=න 2𝑥
𝑑𝑥 ⟹ 4 + 𝑒 = 𝑢 ⟹ 𝑒 𝑑𝑥 = ⟹
𝑦 4+𝑒 2

𝑑𝑢
d𝑦 2 1 1 1 2𝑥
1
න 2=න ⟹ 𝑙𝑛 𝑢 = − + 𝑐 ⟹ 𝑙𝑛 4 + 𝑒 = − +𝑐 ⟹
𝑦 𝑢 2 𝑦 2 𝑦
𝟏
𝒚=
𝟏
− 𝟐 𝒍𝒏 𝟒 + 𝒆𝟐𝒙 + 𝒄
Ex:18

𝑦 ′ = 𝑐𝑜𝑠 2 𝑥. 𝑐𝑜𝑠𝑦 Solve the given differantial equation.

d𝑦 2
d𝑦
= 𝑐𝑜𝑠 𝑥. 𝑐𝑜𝑠𝑦 ⟹ = 𝑐𝑜𝑠 2 𝑥. 𝑑𝑥 It can be separated into variables.
𝑑𝑥 𝑐𝑜𝑠𝑦

d𝑦 𝑐𝑜𝑠𝑦 2
𝑐𝑜𝑠𝑦 𝑐𝑜𝑠𝑦 1 + 𝑐𝑜𝑠2𝑥
න . = න 𝑐𝑜𝑠 𝑥. 𝑑𝑥 ⟹ න 2
d𝑦 = න 2
d𝑦 = න 𝑑𝑥
𝑐𝑜𝑠𝑦 𝑐𝑜𝑠𝑦 𝑐𝑜𝑠 𝑦 1 − 𝑠𝑖𝑛 𝑦 2

𝑑𝑢 1 + 𝑐𝑜𝑠2𝑥 1/2 1/2 1 + 𝑐𝑜𝑠2𝑥

𝑠𝑖𝑛𝑦 = 𝑢 ⟹ 𝑐𝑜𝑠𝑦𝑑𝑦 = 𝑑𝑢 ⟹ න 2
=න 𝑑𝑥 ⟹ න + 𝑑𝑢 = න 𝑑𝑥
1−𝑢 2 1−𝑢 1+𝑢 2

1 1 1 sin2x 1 1+𝑢 1 sin2x

− 2 𝑙𝑛 1−𝑢 + 𝑙𝑛 1+𝑢 = (x+ 2 )+c⟹ 2 𝑙𝑛 1−𝑢 = (x+ 2 )+c
2 2 2

𝟏 𝟏 + 𝒔𝒊𝒏𝒚 𝟏 sin2x
𝒍𝒏 = (x+ )+c
𝟐 𝟏 − 𝒔𝒊𝒏𝒚 𝟐 𝟐
Ex:19

𝑑𝑥 = 𝑡 1 + 𝑡 2 𝑠𝑒𝑐 2 𝑥𝑑𝑡 Solve the given differantial equation.

d𝑥 2
2
= 𝑡 1 + 𝑡 𝑑𝑡 It can be separated into variables.
𝑠𝑒𝑐 𝑥

d𝑥 3 2 3
1 + 𝑐𝑜𝑠2𝑥
න = 𝑡 + 𝑡 𝑑𝑡 ⟹ න 𝑐𝑜𝑠 𝑥𝑑𝑥 = න 𝑡 + 𝑡 𝑑𝑡 ⟹ න 𝑑𝑥 = න 𝑡 + 𝑡 3 𝑑𝑡
1 2
𝑐𝑜𝑠 2 𝑥

𝟏 𝟐 𝟏 𝟒 𝟏 sin2x
𝒕 + 𝒕 = (x+ )+c
𝟐 𝟒 𝟐 𝟐
Ex:20
2𝑥 + 𝑥𝑦 2
𝑦′ = Solve the given differantial equation.
3𝑦 + 𝑦𝑥 2

𝑑𝑦 𝑥(2 + 𝑦 2 ) 𝑦𝑑𝑦 𝑥𝑑𝑥

= ⟹ = It can be separated into variables.
𝑑𝑥 𝑦(3 + 𝑥 ) 2 2+𝑦 2 3+𝑥 2

2
𝑑𝑚
2 + 𝑦 = 𝑚 ⟹ 𝑦𝑑𝑦 = 1 𝑑𝑚 1 𝑑𝑢 1 1 1
2 න =න ⟹ 𝑙𝑛 𝑚 = 𝑙𝑛 𝑢 + 𝑙𝑛 𝑐
2 𝑚 2 𝑢 2 2 2
2
𝑑𝑢
3 + 𝑥 = 𝑢 ⟹ 𝑥𝑑𝑥 =
2

𝑚 = 𝑢. 𝑐 ⟹ 2 + 𝑦 2 = 𝑐(3 + 𝑥 2 )

𝒚= 𝒄 𝟑 + 𝒙𝟐 − 𝟐
Ex:21
𝑑𝑟 = 𝑏 𝑐𝑜𝑠𝜃𝑑𝑟 + 𝑟𝑠𝑖𝑛𝜃𝑑𝜃 Solve the given differantial equation.

𝑑𝑟 − 𝑏𝑐𝑜𝑠𝜃𝑑𝑟 = 𝑏𝑟𝑠𝑖𝑛𝜃𝑑𝜃 ⟹ 1 − 𝑏𝑐𝑜𝑠𝜃 𝑑𝑟 = 𝑏𝑟𝑠𝑖𝑛𝜃𝑑𝜃 It can 𝑟beand 𝜃 are variables.

separated into variables.
𝑑𝑟 𝑏𝑠𝑖𝑛𝜃
න =න 𝑑𝜃 ⟹ 1 − 𝑏𝑐𝑜𝑠𝜃 = 𝑚 ⟹ 𝑏𝑠𝑖𝑛𝜃𝑑𝜃 = 𝑑𝑚
𝑟 1 − 𝑏𝑐𝑜𝑠𝜃

𝑑𝑟 𝑑𝑚
න =න ⟹ 𝑙𝑛 𝑟 = 𝑙𝑛 𝑚 + 𝑙𝑛 𝑐
𝑟 𝑚

𝒓 = 𝒎. 𝒄 ⟹ 𝒓 = 𝒄(𝟏 − 𝒃𝒄𝒐𝒔𝜽)
Ex:22
𝑑4 𝑦
= 100 Solve the given differantial equation.
𝑑𝑥 4
𝑑4 𝑦
= 100 It can be separated into variables.
𝑑𝑥 4
𝑑4 𝑦 𝑑3 𝑦 𝑑3 𝑦 𝑑2 𝑦 2+𝑐 𝑥+𝑐
න 𝑑𝑥 = න 100 𝑑𝑥 ⟹ = 100𝑥 + 𝑐1 ⟹ න 𝑑𝑥 = න 100𝑥 + 𝑐1 𝑑𝑥 ⟹ = 50𝑥 1 2
𝑑𝑥 4 𝑑𝑥 3 𝑑𝑥 3 𝑑𝑥 2

𝑑2 𝑦 2 + 𝑐 𝑥 + 𝑐 ) 𝑑𝑥 ⟹
𝑑𝑦 50 3 𝑐1 2 50 3 𝑐1 2
න 𝑑𝑥 = න(50𝑥 1 2 = 𝑥 + 𝑥 + 𝑐2 𝑥 + 𝑐3 ⟹ න 𝑑𝑦 = න 𝑥 + 𝑥 + 𝑐2 𝑥 + 𝑐3 𝑑𝑥
𝑑𝑥 2 𝑑𝑥 3 2 3 2

𝟐𝟓 𝟒 𝒄𝟏 𝟑 𝒄𝟐 𝟐
𝒚= 𝒙 + 𝒙 + 𝒙 + 𝒄𝟑 𝒙 + 𝒄 𝟒
𝟔 𝟔 𝟐
2.2. Reduction to Separable Variable Form
If some equations of the first order that cannot be separated into variables can be transformed from the
form 𝑦 ′ = 𝑓(𝑥, 𝑦) to the forms 𝑦 ′ = 𝑓(𝑦/𝑥), 𝑦 ′ = 𝑓(𝑥/𝑦), 𝑦 ′ = 𝑓(𝑥 + 𝑦), they can be made separable
into variables by variable transformation.

For this purpose, variable transformations should be made in the form of 𝒖 = 𝒚/𝒙, 𝒖 = 𝒙/𝒚 or 𝒖 = 𝒙 +
𝒚
𝒚. In the equation of 𝒚′ = 𝒇 , a solution is sought by making the 𝒚 = 𝒖. 𝒙 transformation. 𝒖 is the new
𝒙
dependent variable, which depends on the old independent variable 𝒙.

𝒚 ′
𝒅𝒖
𝒖(𝒙) = ⟹ 𝒚 = 𝒖 𝒙 . 𝒙 ⟹ 𝒚 = 𝒖 + 𝒙
𝒙 𝒅𝒙

After this variable transformation, the differential equation structure will be

𝒅𝒖
𝒖+𝒙 = 𝒇(𝒖)
𝒅𝒙
Ex:23

𝑥𝑦 2 𝑦′ = 𝑦 3 −𝑥 3 Solve the given differantial equation.

3 3
𝑦 −𝑥
න 𝑦 2 𝑑𝑦 = න 𝑑𝑥 It can not be separated into variables.
d𝑦 𝑥
𝑥𝑦 2 = 𝑦 3 −𝑥 3 ⟹ The equation is nonlinear due to 𝑥 3 and 𝑦 3 .
𝑑𝑥 𝑦2 𝑑𝑥 A transformation is needed for solution.
න 3 𝑑𝑦 = න
𝑦 −𝑥 3 𝑥
3 3
′
𝑦 −𝑥 ′
𝑦 1 𝑦
𝑦 = = 𝑓 𝑥, 𝑦 ⟹ 𝑦 = − The equationis a function of both 𝑥 and 𝑦, and .
𝑥𝑦 2 𝑥 𝑦 2 𝑥

𝑦 𝑥
= 𝑚 ⟹ 𝑦 ′ = 𝑓(𝑚) In this case, it can be separated into variables.
𝑥
𝑦 𝑑𝑦 𝑑𝑚 𝑑𝑚 1 𝑑𝑚 1 1
= 𝑚 ⟹ 𝑦 = 𝑚𝑥 ⟹ =𝑚+𝑥 ⟹𝑚+𝑥 =𝑚− 2 ⟹𝑥 = − 2 ⟹ න 𝑚2 𝑑𝑚 = − න 𝑑𝑥
𝑥 𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑚 𝑑𝑥 𝑚 𝑥
𝑚3 1/3
𝑦 1/3 𝟏/𝟑
= −𝑙𝑛𝑥 + 𝑐 ⟹ 𝑚 = −3𝑙𝑛𝑥 + 𝑐1 ⟹ = −3𝑙𝑛𝑥 + 𝑐1 ⟹ 𝒚 = 𝒙 −𝟑𝒍𝒏𝒙 + 𝒄𝟏
3 𝑥
Ex:24

𝑦′ − 𝑦 = 𝑥 Solve the given differantial equation.

d𝑦 It can not be separated into variables.

= 𝑥 + 𝑦 ⟹ න 𝑑𝑦 = න(𝑥 + 𝑦) 𝑑𝑥 A transformation is needed for solution.
𝑑𝑥
𝑑𝑦 𝑑𝑚 𝑑𝑦 𝑑𝑚
𝑥+𝑦 =𝑚 ⟹ +1 = ⟹ = −1 In this case, it can be separated into variables.
𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑑𝑥
𝑑𝑚 𝑑𝑚 𝑑𝑚
−1= 𝑚 ⟹ = 𝑚+1⟹ ‫𝑚 ׬‬+1 = ‫⟹ 𝑥𝑑 ׬‬
𝑑𝑥 𝑑𝑥

𝑙𝑛 𝑚+1 = 𝑥 + 𝑐 ⟹ 𝑚 + 1 = 𝑒 𝑥+𝑐 = 𝑒 𝑐 . 𝑒 𝑥 ⟹ 𝑚 + 1 = 𝑐1 . 𝑒 𝑥 ⟹ 𝑚 = 𝑐1 . 𝑒 𝑥 −1

𝑥 + 𝑦 = 𝑐1 . 𝑒 𝑥 −1 ⟹ 𝒚 = 𝒄𝟏 . 𝒆𝒙 −𝟏 − 𝒙
Ex:25
𝑦 𝑥
2 − 3 𝑑𝑥 + 3 − 2 𝑑𝑦 = 0 Solve the given differantial equation.
𝑥 𝑦
𝑦 𝑥 It can not be separated into variables.
න 2 − 3 𝑑𝑥 + න 3 − 2 𝑑𝑦 = 0 A transformation is needed for solution.
𝑥 𝑦
𝑦
𝑑𝑦 2 𝑥 −3 𝑦 𝑦
=− 𝑥 ⟹ 𝑓(𝑥 ) ⟹ 𝑥 = 𝑢 In this case, it can be separated into variables.
𝑑𝑥 3 − 2𝑦
𝑦 𝑑𝑦 𝑑𝑢 𝑑𝑢 2𝑢−3 𝑑𝑢 2𝑢2 −3𝑢
𝑥
= 𝑢 ⟹ 𝑦 = 𝑢𝑥 ⟹ 𝑑𝑥 = 𝑢 + 𝑥 𝑑𝑥 ⟹ 𝑢 + 𝑥 𝑑𝑥 =− 1 ⟹ 𝑥
𝑑𝑥
= − 3𝑢−2 − 𝑢
3−2
𝑢
𝑑𝑢 2𝑢2 − 3𝑢 𝑑𝑢 −5𝑢2 + 5𝑢 3𝑢 − 2 𝑑𝑥
𝑥 =− −𝑢 ⟹𝑥 = ⟹න 2
𝑑𝑢 = න
𝑑𝑥 3𝑢 − 2 𝑑𝑥 3𝑢 − 2 −5𝑢 + 5𝑢 𝑥
3𝑢 − 2 𝑑𝑥 1 2 𝑑𝑥
න 𝑑𝑢 = න ⟹න − 𝑑𝑢 = න 5 ⟹ −𝑙𝑛 1 − 𝑢 − 2𝑙𝑛𝑢 = 5𝑙𝑛𝑥 + 𝑙𝑛𝑐
5𝑢(1 − 𝑢) 𝑥 1−𝑢 𝑢 𝑥
1 𝟏
− 𝑙𝑛 𝑢2 1−𝑢 = 5𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 2 = 𝑐𝑥 5 ⟹ 𝟐
= 𝒄𝒙𝟓
𝑢 1−𝑢 𝒚 𝒚
𝟏−
𝒙 𝒙
Ex:26
𝑦𝑑𝑥 + 𝑦 + tan(𝑥 + 𝑦) 𝑑𝑦 = 0 Solve the given differantial equation.

It can not be separated into variables.

න 𝑦 + tan(𝑥 + 𝑦) 𝑑𝑦 = − න 𝑦𝑑𝑥 A transformation is needed for solution.
𝑑𝑦 𝑑𝑚 𝑑𝑦 𝑑𝑚
𝑥+𝑦 = 𝑚 ⟹ 1+ = ⟹ 𝑑𝑥 1 + = ⟹ 𝑑𝑥 = 𝑑𝑚 − 𝑑𝑦
𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑑𝑥
𝑦 + tan(𝑚) 𝑑𝑦 = −𝑦 𝑑𝑚 − 𝑑𝑦 ⟹ 𝑦𝑑𝑦 + tan 𝑚 𝑑𝑦 = −𝑦𝑑𝑚 + 𝑦𝑑𝑦 ⟹
𝑑𝑚 𝑑𝑦
tan 𝑚 𝑑𝑦 = −𝑦𝑑𝑚 ⟹ න = න− Now, it can be separated into variables.
tan 𝑚 𝑦
cos(𝑚) 𝑑𝑦 𝑑𝑢
න 𝑑𝑚 = න − ⟹ sin 𝑚 = 𝑢 ⟹ cos 𝑚 𝑑𝑚 = 𝑑𝑢 ⟹ න = −𝑙𝑛𝑦 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑢 = −𝑙𝑛𝑦 + 𝑙𝑛𝑐
sin 𝑚 𝑦 𝑢

𝑐 𝑐 𝒄
𝑙𝑛 sin 𝑚 = 𝑙𝑛 ⟹ sin 𝑚 = ⟹ 𝒔𝒊𝒏 𝒙 + 𝒚 =
𝑦 𝑦 𝒚
Ex:27
𝑦′ = 𝑥 − 𝑦 2
−2 𝑥−𝑦 −2 Solve the given differantial equation.

2 It can not be separated into variables.

න 𝑑𝑦 = − න 𝑥 𝑦 − 2 𝑥 − 𝑦 − 2 𝑑𝑥 A transformation is needed for solution.
𝑑𝑦 𝑑𝑢
𝑥−𝑦 = 𝑢 ⟹ 𝑦 = 𝑥−𝑢 ⟹ =1− ⟹
𝑑𝑥 𝑑𝑥
𝑑𝑢
1− = 𝑢2 − 2𝑢 − 2 Now, it can be separated into variables.
𝑑𝑥
𝑑𝑢 2
𝑑𝑢 1 1 1
− = 𝑢 − 2𝑢 − 3 ⟹ 2 = −𝑑𝑥 ⟹ න − 𝑑𝑢 = − න 𝑑𝑥
𝑑𝑥 𝑢 − 2𝑢 − 3 4 𝑢−3 𝑢+1
1 𝑢−3 𝑥−𝑦−3
𝑙𝑛 𝑢 − 3 − 𝑙𝑛 𝑢 + 1 = −𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛 = −4𝑥 + 𝑐1 ⟹ = 𝑒 −4𝑥+𝑐1
4 𝑢+1 𝑥−𝑦+1
𝑥−𝑦−3 𝑥 − 𝑦 − 3 𝒄 𝒆 −𝟒𝒙 + 𝟑
𝟐
= 𝑒 −4𝑥 𝑒 𝑐1 ⟹ = 𝑐2 𝑒 −4𝑥 ⟹ 𝒚 = 𝒙 +
𝑥−𝑦+1 𝑥−𝑦+1 𝒄𝟐 𝒆−𝟒𝒙 − 𝟏
Ex:28
𝑥 3 + 𝑦 3 𝑑𝑥 − 3𝑥𝑦 2 𝑑𝑦 = 0 Solve the given differantial equation.

𝑦 3 𝑦 2
It can not be separated into variables.
න 1+ 𝑑𝑥 − 3 න 𝑑𝑦 = 0 A transformation is needed for solution.
𝑥 𝑥
𝑦 3
𝑑𝑦 1 + 𝑥 𝑦 𝑦
= 2 ⟹ 𝑓( ) ⟹ =𝑢 In this case, it can be separated into variables.
𝑑𝑥 𝑦 𝑥 𝑥
3 𝑥
𝑦 𝑑𝑦 𝑑𝑢 𝑑𝑢 1+𝑢3 𝑑𝑢 1+𝑢3
𝑥
= 𝑢 ⟹ 𝑦 = 𝑢𝑥 ⟹ 𝑑𝑥
=𝑢+ 𝑥 𝑑𝑥 ⟹𝑢 + 𝑥 𝑑𝑥 = 3𝑢2 ⟹ 𝑥 𝑑𝑥 = 3𝑢2 −𝑢
𝑑𝑢 1 + 𝑢3 − 3𝑢3 𝑑𝑢 1 − 2𝑢3 3𝑢2 𝑑𝑥 3 2
1
𝑥 = ⟹𝑥 = ⟹ −න 𝑑𝑢 = න ⟹ 𝑣 = 1 − 2𝑢 ⟹ 3𝑢 𝑑𝑢 = − 𝑑𝑣
𝑑𝑥 3𝑢2 𝑑𝑥 3𝑢2 1 − 2𝑢3 𝑥 2
1 𝑙𝑛𝑐
1
− 2 𝑑𝑣 𝑑𝑥 1 𝑐1
3 3
න =න ⟹ − 𝑙𝑛 𝑣 = 𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛 1 − 2𝑢 = −2𝑙𝑛𝑥 − 2𝑙𝑛𝑐 ⟹ 1 − 2𝑢 = 2
𝑣 𝑥 2 𝑥
𝑦 3 𝑐1
1−2 = 2 ⟹ 𝒙𝟑 − 𝟐𝒚𝟑 = 𝒄𝟏 𝒙
𝑥 𝑥
Ex:29
𝑥 2 + 3𝑦 2 𝑑𝑥 + 2𝑥𝑦𝑑𝑦 = 0 Solve the given differantial equation.

𝑦 2 𝑦 It can not be separated into variables.

න 1+3 𝑑𝑥 + 2 න 𝑑𝑦 = 0 A transformation is needed for solution.
𝑥 𝑥
𝑦 2
𝑑𝑦 1 + 3 𝑦 𝑦
= 𝑥 ⟹ 𝑓( ) ⟹ = 𝑢
𝑦 In this case, it can be separated into variables.
𝑑𝑥 2 𝑥 𝑥 𝑥
2 2
𝑦 𝑑𝑦 𝑑𝑢 𝑑𝑢 1+3𝑢 𝑑𝑢 1+3𝑢
= 𝑢 ⟹ 𝑦 = 𝑢𝑥 ⟹ =𝑢+ 𝑥 𝑑𝑥 ⟹𝑢 + 𝑥 𝑑𝑥 = ⟹𝑥 = −𝑢
𝑥 𝑑𝑥 2𝑢 𝑑𝑥 2𝑢
𝑑𝑢 1 + 𝑢2 2𝑢 𝑑𝑥 2 ⟹ 2𝑢𝑑𝑢 = 𝑑𝑣
𝑥 = ⟹ −න 𝑑𝑢 = න ⟹ 𝑣 = 1 + 𝑢
𝑑𝑥 2𝑢 1 + 𝑢2 𝑥

𝑑𝑣 𝑑𝑥
න =න ⟹ 𝑙𝑛 𝑣 = 𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛 1 + 𝑢2 = 𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 1 + 𝑢2 = 𝑐. 𝑥
𝑣 𝑥
𝑦 2
1+ = 𝑐. 𝑥 ⟹ 𝒙𝟐 + 𝒚𝟐 = 𝒄𝒙𝟑
𝑥
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: First-order linear differential equations and
linearization

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
2.2. First-Order Linear Differential Equations and Linearization
A first-order linear differential equation has the form
𝑑𝑦
+ 𝑝(𝑥)𝑦 = 𝑞 𝑥 1. order, linear, non-homogeneous,
𝑑𝑥 y: dependent, x: independent variable
𝑦𝐻 𝑦𝑃

𝑦 = 𝑦𝐻 + 𝑦𝑃 (General Solution)

Or, in the other forms:

d𝑥 d𝑥 d𝑟
+ 𝑃 𝑦 𝑥 = 𝑄(𝑦) 1. order, linear, + 𝑃 𝑡 𝑥 = 𝑄(𝑡) 1. order, linear, + 𝑃 𝜃 𝑟 = 𝑄(𝜃) 1. order, linear,
𝑑𝑦 non-homogeneous,
𝑑𝑡 non-homogeneous, 𝑑𝜃 non-homogeneous,
x: dependent, x: dependent, r: dependent,
𝑥𝐻 𝑥𝑃 𝑥𝐻 𝑥𝑃 t: independent variable 𝑟𝐻 𝑟𝑃
y: independent variable 𝜃 : independent variable
𝑥 = 𝑥𝐻 + 𝑥𝑃 𝑥 = 𝑥𝐻 + 𝑥𝑃 𝑟 = 𝑟𝐻 + 𝑟𝑃

First, the homogeneous solution and then, the non-homogeneous solution is calculated by the variation of
parameters method (VPM) over the homogeneous solution. The general solution is the sum of the homogeneous
and the non-homogeneous (particular) solution.
2.2.1 Variation of Parameters Method (VPM)
𝑑𝑦
+ 𝑝(𝑥)𝑦 = 𝑞 𝑥
𝑑𝑥
If the right side of the equation above is zero, it is said that the equation is homogeneous. Therefore, in this type of
equations, the general solution is obtained by adding the homogeneous and non-homogeneous solutions found in two
ways. The first solution is found only from the homogeneous part, while the second solution is found from the non-
homogeneous part.

Homogeneous solution:
If 𝑞 𝑥 = 0, the equation becomes homogeneous. In this case, the equation can be separated into variables. 𝑞 𝑥 =
0 is a temporary assumption made to reach the solution.
𝑑𝑦
+ 𝑝(𝑥)𝑦 = 0
𝑑𝑥
By solving this homogeneous equation separated into variables;
𝑑𝑦
= −𝑝 𝑥 𝑑𝑥 ⟹ 𝑙𝑛𝑦 = − න 𝑝 𝑥 𝑑𝑥 + 𝑙𝑛𝑐 ⟹ 𝒚𝑯 = 𝒄𝒆− ‫( 𝒙𝒅𝒑 ׬‬Here, 𝒄 is an arbitrary constan
𝑦
2.2.1 Variation of Parameters Method (VPM)
According to VPM, the previously made assumption of 𝒒(𝒙) = 𝟎 shows that a structure that is only a function of 𝑥 is neglected
during the homogeneous solution. Therefore, the neglect made can be compensated by accepting the constant 𝒄 in the
homogeneous solution as a function of 𝑥. In other words, the effect of 𝒒(𝒙) on the general solution is tried to be found by
accepting 𝒄 = 𝒄(𝒙). The 𝒄(𝒙) function is determined to satisfy the main differential equation. Here, a general solution (𝒚) proposal
is created starting from the homogeneous solution (𝒚𝑯 ). The fact that this proposed general solution satisfies the main differential
equation will be sufficient for the solution to be completed. For this purpose, the proposed solution itself and its derivative are
transferred to the main differential equation and the 𝒄(𝒙) function that satisfies the equation is determined.
Non-homogeneous and
General solution:
𝑑𝑦𝑝
− ‫𝑥𝑑𝑝 ׬‬ 𝑑𝑐 𝑑 − ‫𝑥𝑑𝑝 ׬‬
𝑦𝑝 = 𝒄 𝒙 𝑒 ⟹ = 𝑒 − ‫𝑥𝑑𝑝 ׬‬ +𝑐 𝑒 After finding
𝑑𝑥 𝑑𝑥 𝑑𝑥 c(x), the
solution is
𝑑𝑦𝑝 𝑑𝑐 𝑑 − ‫𝑥𝑑𝑝 ׬‬
+ 𝑝 𝑥 𝑦𝑝 = 𝑞 𝑥 ⟹ 𝑒 − ‫𝑥𝑑𝑝 ׬‬ +𝑐 𝑒 + 𝑝 𝑥 𝑐𝑒 − ‫)𝑥(𝑞 = 𝑥𝑑𝑝 ׬‬ completed
𝑑𝑥 𝑑𝑥 𝑑𝑥 by writing it
in its place.
𝑑 − ‫𝑥𝑑𝑝 ׬‬ 𝑑
𝑒 = − න 𝑝𝑑𝑥 𝑒 − ‫ = 𝑥𝑑𝑝 ׬‬−𝑝(𝑥)𝑒 − ‫𝑥𝑑𝑝 ׬‬
𝑑𝑥 𝑑𝑥

𝑑𝑐 𝑑𝑐
𝑒 − ‫𝑥𝑑𝑝 ׬‬ + 𝑐𝑝 𝑥 𝑒 − ‫ 𝑥𝑑𝑝 ׬‬− 𝑐𝑝 𝑥 𝑒 − ‫⟹ 𝑥 𝑞 = 𝑥𝑑𝑝 ׬‬ = 𝑞 𝑥 𝑒 ‫ ⟹ 𝑥𝑑𝑝 ׬‬න 𝒅𝒄 = න 𝒒 𝒙 𝒆‫⟹ 𝒙𝒅 𝒙𝒅𝒑 ׬‬
𝑑𝑥 𝑑𝑥
𝒚 = 𝒚𝑯 + 𝒚𝑷 (General Solution)
or the solution can be formulated as follows.

⟹ න 𝑑𝑐 = 𝑐 𝑥 ⟹ 𝒄 𝒙 = න 𝒆‫𝒙𝒅 𝒙 𝒒 𝒙𝒅𝒑 ׬‬

𝑦 = 𝑐 𝑥 𝑒 − ‫𝒆 = 𝐲 ⟹ 𝑥𝑑𝑝 ׬‬− ‫ 𝒙𝒅𝒑 ׬‬න 𝒆‫ 𝒙𝒅 𝒙 𝒒 𝒙𝒅𝒑 ׬‬+ 𝒄

Here, ℎ = ‫𝒆 = 𝐲 ⟹ 𝑥𝑑𝑝 ׬‬−𝒉 න 𝒆𝒉 𝒒 𝒙 𝒅𝒙 + 𝒄 , ℎ = ℎ(𝑥)

𝐲 = 𝒆−𝒉 න 𝒆𝒉 𝒒 𝒙 𝒅𝒙 + 𝒄𝒆−𝒉

𝑦𝑃 𝑦𝐻
Ex-1:
𝑦′ − 𝑦 = 𝑒 2𝑥 Solve the given differantial equation.
Homogeneous solution (yH):
1. order, linear, nonhomogeneous dif. Equ. 𝑑𝑦 𝑑𝑦
𝑝 𝑥 = −1 and 𝑞 𝑥 = 𝑒 2𝑥 y: dependent, x: independent variable 𝑦′ − 𝑦 = 0 ⟹ =𝑦⟹න = න𝑑𝑥 ⟹
𝑑𝑥 𝑦
𝑑𝑦 It can not be 𝑦
= 𝑒 2𝑥 + y ⟹ ‫ 𝑒(׬ = 𝑦𝑑 ׬‬2𝑥 +y)𝑑𝑥 𝑙𝑛𝑦 = 𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑦 − 𝑙𝑛𝑐 = 𝑥 ⟹ 𝑙𝑛 = 𝑥
𝑑𝑥 𝑐
separated into variables.
𝒚𝑯 = 𝒄. 𝒆𝒙
𝑒 2𝑥 + y = 𝑚 ⟹ y = 𝑚 − 𝑒 2𝑥 ⟹ Nonhomogeneous solution (𝑦𝑃 ):
𝑑𝑦 𝑑𝑚 𝑦𝐻 = 𝑐. 𝑒 𝑥 ⟹ 𝑦𝑃 = 𝑐(𝑥). 𝑒 𝑥 Variation of Parameters
= − 2𝑒 2𝑥 Method(VPM)
𝑑𝑥 𝑑𝑥 𝑑𝑦𝑃 𝑑𝑐 𝑥 writing 𝑦𝑃′ and 𝑦𝑃 into
𝑑𝑚 = 𝑒 + 𝑐𝑒 𝑥 By
− 2𝑒 2𝑥 − 𝑚 − 𝑒 2𝑥 = 𝑒 2𝑥 ⟹ 𝑑𝑥 𝑑𝑥 the nonhomogeneous equation:
𝑑𝑥
𝑦𝑃′ 𝑦𝑃
𝑦′ 𝑦
𝑑𝑐 𝑥 𝑑𝑐 𝑥
𝑒 + 𝑐𝑒 𝑥 − 𝑐𝑒 𝑥 = 𝑒 2𝑥 ⟹ 𝑒 = 𝑒 2𝑥 ⟹
⟹ න𝑑𝑚 = න(2𝑒 2𝑥 +𝑚)𝑑𝑥 (It still can not be 𝑑𝑥 𝑑𝑥
separated into variables)
න𝑑𝑐 = න𝑒 𝑥 𝑑𝑥 ⟹ 𝑐 𝑥 = 𝑒 𝑥
𝑦′ − 𝑦 = 𝑒 2𝑥
𝑦𝑃 = 𝑒 𝑥 . 𝑒 𝑥 ⟹ 𝒚𝑷 = 𝒆𝟐𝒙
𝑦𝐻 𝑦𝑃
𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝒚 = 𝒄. 𝒆𝒙 + 𝒆𝟐𝒙
2. Method: Formulation solution

𝑦′ − 𝑦 = 𝑒 2𝑥 ⟹ 𝑝 𝑥 = −1 and 𝑞 𝑥 = 𝑒 2𝑥

y = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬−1 𝑑𝑥 ⟹ 𝒉 = −𝒙

y = 𝑒 −(−𝑥) න 𝑒 −𝑥 𝑒 2𝑥 𝑑𝑥 + 𝑐 = 𝑒 𝑥 න 𝑒 𝑥 𝑑𝑥 + 𝑐 ⟹ 𝑦 = 𝑒 𝑥 𝑒 𝑥 + 𝑐

𝒚 = 𝒄𝒆𝒙 + 𝒆𝟐𝒙
Ex-2:
𝑥 4 𝑦 ′ + 2𝑥 3 𝑦 = 1 Solve the given differantial equation.
It can not be Homogeneous solution (yH):
න𝑥 4 𝑑𝑦 = න(1 − 2𝑥 3 𝑦)𝑑𝑥 2 𝑑𝑦 −2 𝑑𝑦 −2
separated into variables. 𝑦′ + 𝑦 = 0 ⟹ = 𝑦⟹න =න 𝑑𝑥
𝑥 𝑑𝑥 𝑥 𝑦 𝑥
1−𝑚 𝑑𝑦 −2𝑥 3 𝑑𝑚 − 6𝑥 2 (1 − 𝑚)𝑑𝑥 𝑐 𝒄
1− 2𝑥 3 𝑦 =𝑚⟹y= ⟹ = 𝑙𝑛𝑦 = −2𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑦 = 𝑙𝑛 2 ⟹ 𝒚𝑯 = 𝟐
2𝑥 3 𝑑𝑥 4𝑥 6 𝑥 𝒙
3 𝑑𝑚 − 6𝑥 2 (1 − 𝑚)𝑑𝑥 Nonhomogeneous solution (𝑦𝑃 ):
2𝑥
𝑥4 = 1 − 2𝑥 3 𝑦 = 𝑚 ⟹ 𝑐 𝑐(𝑥) Variation of Parameters
4𝑥 6 𝑦𝐻 = 2 ⟹ 𝑦𝑃 = 2 Method(VPM)
𝑚 𝑥 𝑥
𝑥 3 𝑑𝑦𝑃 1 𝑑𝑐 2𝑐 By writing 𝑦 ′ and 𝑦 into
𝑑𝑚 − (1 − 𝑚)𝑑𝑥 = 𝑚 ⟹ 𝑥𝑑𝑚 − 3 1 − 𝑚 𝑑𝑥 = 2𝑚 ⟹ = 2 − 3 𝑃 𝑃
2 2 𝑑𝑥 𝑥 𝑑𝑥 𝑥 the nonhomogeneous equation:
′ 𝑦𝑃
(It still can not be 𝑦𝑃
න 𝑥𝑑𝑚 = 2𝑚 + න 3 1 − 𝑚 𝑑𝑥 separated into variables.) 1 𝑑𝑐 2𝑐 2 𝑐 1 𝑑𝑐 1
− + = ⟹ = ⟹
Dif. the equation is linear. Therefore; 𝑥 2 𝑑𝑥 𝑥 3 𝑥 𝑥 2 𝑥 4 𝑑𝑥 𝑥 2
1 1
1 4 𝑦 ′ + 2𝑥 3 𝑦 = 1) ⟹ 2
′+ 𝑦=
1 න𝑑𝑐 = න 2 𝑑𝑥 ⟹ 𝑐 𝑥 = −
(𝑥 𝑦 𝑥 𝑥
𝑥4 𝑥 𝑥4
1 1 𝟏
𝑦𝐻 𝑦𝑃 𝑦𝑃 = − . 2 ⟹ 𝒚𝑷 = − 𝟑
𝑥 𝑥 𝒙
2 1 1. order, linear, nonhomogeneous dif. Equ. 𝒄 𝟏
𝑝 𝑥 = and 𝑞 𝑥 = 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝒚 = 𝟐 − 𝟑
𝑥 𝑥4 y: dependent, x: independent variable
𝒙 𝒙
2. Method: Formulation solution

1 4 ′ 3 ′
2 1 2 1
𝑥 𝑦 + 2𝑥 𝑦 = 1 ⟹ 𝑦 + 𝑦 = 4 ⟹ 𝑝 𝑥 = and 𝑞 𝑥 =
𝑥4
𝑥4 𝑥 𝑥 𝑥

y = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

2
ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫׬‬ 𝑑𝑥 ⟹ ℎ = 2 ln 𝑥 ⟹ 𝒉 = 𝒍𝒏 𝒙𝟐
𝑥

2) 2 1 1 1 1 1
y = 𝑒 −(ln 𝑥 න 𝑒 ln 𝑥 𝑑𝑥 + 𝑐 = න 𝑑𝑥 + 𝑐 = − +𝑐
𝑥4 𝑥2 𝑥2 𝑥2 𝑥
1
𝑥2.
𝑥4
𝒄 𝟏
𝒚= 𝟐− 𝟑
𝒙 𝒙
Ex-3:
𝑑𝑥 𝑥 1 Solve the given differantial equation.
+ 2= 2
𝑑𝑡 𝑡 𝑡
1 1
1. order, linear, nonhomogeneous dif. Equ. 𝑥𝑃′ 𝑥𝑃
x: dependent, t: independent variable
𝑥′ + 2 𝑥 = 2
𝑡 𝑡 𝑑𝑐 1 𝑐 1 1 1 1 𝑑𝑐 1 1
1 1 𝑒 − 2 𝑒 + 2 𝑐𝑒 = 2 ⟹
𝑡 𝑡 𝑡 𝑒 = 2⟹
𝑡
𝑝 𝑡 = and 𝑞 𝑡 = 𝑑𝑡 𝑡 𝑡 𝑡 𝑑𝑡 𝑡
𝑥𝐻 𝑥𝑃 𝑡2 𝑡2

Homogeneous solution (𝑥𝐻 ): −

1
𝑒 𝑡 1 𝑑𝑡
1 𝑑𝑥 1 𝑑𝑥 1 න𝑑𝑐 = ඳ 2 𝑑𝑡 ⟹ 𝑚 = − ⟹ 𝑑𝑚 = 2 ⟹
𝑥′ + 2 𝑥 = 0 ⟹ = − 2𝑥 ⟹ න = න − 2 𝑑𝑡 𝑡 𝑡 𝑡
𝑡 𝑑𝑡 𝑡 𝑥 𝑡
1 1 𝟏
𝑙𝑛𝑥 = + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑥 − 𝑙𝑛𝑐 = ⟹ 𝒙𝑯 = 𝒄𝒆 𝒕 1
𝑡 𝑥 𝑡 න𝑑𝑐 = න𝑒 𝑚 𝑑𝑚 ⟹𝑐= 𝑒𝑚 ⟹ 𝑐(𝑡) = −
𝑒 𝑡
𝑙𝑛
𝑐
Nonhomogeneous solution (𝑥𝑃 ): −
1 1
1 1
𝑥𝑃 = 𝑒 𝑡. 𝑒𝑡 ⟹ 𝒙𝑷 = 𝟏
𝑥𝐻 = 𝑐𝑒 𝑡 ⟹ 𝑥𝑃 = 𝑐(𝑡)𝑒 𝑡 Variation of Parameters
Method(VPM) 𝟏
𝑥 = 𝑥𝐻 + 𝑥𝑃 ⟹ 𝒙 = 𝒄𝒆 𝒕 +𝟏
𝑑𝑥𝑃 𝑑𝑐 1 𝑐 1 By writing 𝑥𝑃′ and 𝑥𝑃 into
= 𝑒𝑡 − 2 𝑒𝑡 the nonhomogeneous equation:
𝑑𝑡 𝑑𝑡 𝑡 Alternative solution method: By separating the
𝑥𝑃 1 diff. equation into variables, it can be solved by
𝑥𝑃′ + 2 = 2 ⟹ integration.
𝑡 𝑡
2. Method: Formulation solution
𝑑𝑥 𝑥 1 1 1 1 1
+ 2 = 2 ⟹ 𝑥′ + 2 𝑥 = 2 ⟹ 𝑝 𝑡 =
𝑡2
and 𝑞 𝑡 =
𝑡2
𝑑𝑡 𝑡 𝑡 𝑡 𝑡

y = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑡 𝑑𝑡 + 𝑐 ⟹

1 𝟏
ℎ = ‫ ⟹ 𝑡𝑑𝑝 ׬‬ℎ = ‫׬‬ 𝑑𝑡 ⟹ 𝒉 −
𝑡2 𝒕
𝟏
−(− ) −
𝟏 1 𝟏
𝑚
1
𝑚
1
−
1
y= 𝑒 𝒕 න𝑒 𝒕 𝑑𝑡 + 𝑐 = 𝑒 𝒕 න 𝑒 𝑑𝑚 + 𝑐 = 𝑒 𝑡 𝑒 + 𝑐 = 𝑒 𝑡 𝑒 𝑡 + 𝑐
𝑡2
1
− = 𝑚 ⟹ 𝑡 −2 𝑑𝑡 = 𝑑𝑚
𝑡
𝟏
𝒚= 𝟏+ 𝒄𝒆 𝒕
Ex-4:
𝑑𝑟
+ 𝑟𝑡𝑎𝑛𝜃 = 𝑐𝑜𝑠𝜃 Solve the given differantial equation.
𝑑𝜃
1. order, linear, nonhomogeneous dif. Equ. 𝑟𝑃′ 𝑟𝑃
𝑟′ + 𝑟𝑡𝑎𝑛𝜃 = 𝑐𝑜𝑠𝜃 r: dependent, 𝜃: independent variable
𝑑𝑐
𝑟𝐻 𝑟𝑃 𝑝 𝜃 = 𝑡𝑎𝑛𝜃 and 𝑞 𝜃 = 𝑐𝑜𝑠𝜃 𝑐𝑜𝑠𝜃 − 𝑐. 𝑠𝑖𝑛𝜃 + 𝑐. 𝑐𝑜𝑠𝜃. 𝑡𝑎𝑛𝜃 = 𝑐𝑜𝑠𝜃 ⟹
𝑑𝜃
𝑑𝑟
+ 𝑟𝑡𝑎𝑛𝜃 = 𝑐𝑜𝑠𝜃 ⟹ 𝑑𝑟 = 𝑐𝑜𝑠𝜃 − 𝑟𝑡𝑎𝑛𝜃 𝑑𝜃 𝑑𝑐 𝑠𝑖𝑛𝜃
𝑑𝜃 𝑐𝑜𝑠𝜃 − 𝑐. 𝑠𝑖𝑛𝜃 + 𝑐. 𝑐𝑜𝑠𝜃. = 𝑐𝑜𝑠𝜃 ⟹
𝑑𝜃 𝑐𝑜𝑠𝜃
It can not be separated into variables. May not be
solved by transformation.
න𝑑𝑐 = න𝑑𝜃 ⟹ 𝑐 = 𝜃 ⟹ 𝑐(𝜃) = 𝜃
Homogeneous solution (𝑟𝐻 ):
𝑑𝑟 𝑑𝑟 𝑠𝑖𝑛𝜃 𝑟𝑃 = 𝜃. 𝑐𝑜𝑠𝜃 ⟹ 𝒓𝑷 = 𝜽. 𝒄𝒐𝒔𝜽
𝑟 ′ + 𝑟𝑡𝑎𝑛𝜃 = 0 ⟹ = −𝑟𝑡𝑎𝑛𝜃 ⟹ න = න− 𝑑𝜃
𝑑𝜃 𝑟 𝑐𝑜𝑠𝜃
𝑙𝑛𝑟 = 𝑙𝑛 𝑐𝑜𝑠𝜃 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑟 = 𝑙𝑛 𝑐. 𝑐𝑜𝑠𝜃 ⟹ 𝒓𝑯 = 𝒄. 𝒄𝒐𝒔𝜽 𝑟 = 𝑟𝐻 + 𝑟𝑃 ⟹ 𝒓 = 𝒄. 𝒄𝒐𝒔𝜽 + 𝜽. 𝒄𝒐𝒔𝜽
Nonhomogeneous solution (𝑟𝑃 ): 𝒓 = 𝒄 + 𝜽 𝒄𝒐𝒔𝜽
Variation of Parameters
𝑟𝐻 = 𝑐. 𝑐𝑜𝑠𝜃 ⟹ 𝑟𝑃 = 𝑐(𝜃). 𝑐𝑜𝑠𝜃 Method(VPM)
𝑑𝑟𝑃 𝑑𝑐 By writing 𝑟𝑃′ and 𝑟𝑃 into
= 𝑐𝑜𝑠𝜃 − 𝑐. 𝑠𝑖𝑛𝜃 the nonhomogeneous equation:
𝑑𝜃 𝑑𝜃
𝑟𝑃′ + 𝑟𝑃 𝑡𝑎𝑛𝜃 = 𝑐𝑜𝑠𝜃 ⟹
2. Method: Formulation solution
𝑑𝑟
+ 𝑟 tan 𝜃 = cos 𝜃 ⟹ 𝑟 ′ + 𝑟 tan 𝜃 = cos 𝜃 ⟹ 𝑝 𝜃 = tan 𝜃 and 𝑞 𝜃 = cos 𝜃
𝑑𝜃

y = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝜃 𝑑𝜃 + 𝑐 ⟹

sin 𝜃 𝑑𝑢 𝟏
ℎ = ‫ ⟹ 𝜃𝑑𝑝 ׬‬ℎ = ‫ ׬‬tan 𝜃 𝑑𝜃 = ‫׬‬ 𝑑𝜃 = ‫׬‬− ⟹ ℎ = −ln 𝑢 ⟹ 𝒉 = 𝐥𝐧
cos 𝜃 𝑢 𝒄𝒐𝒔 𝜽

cos 𝜃 = 𝑢 ⟹ −sin 𝜃 𝑑𝜃 = 𝑑𝑢

1 1
−(ln ) ln
y=𝑒 𝑐𝑜𝑠 𝜃 න𝑒 𝑐𝑜𝑠 𝜃 𝑐𝑜𝑠 𝜃 𝑑𝜃 + 𝑐 = 𝑐𝑜𝑠 𝜃 න(1) 𝑑𝜃 + 𝑐 = 𝑐𝑜𝑠 𝜃 (𝜃 + 𝑐) ⟹

𝒓 = 𝒄 + 𝜽 𝒄𝒐𝒔𝜽
Ex-5:
𝑦 2 𝑒 2𝑥 𝑑𝑥 = 4 + 𝑒 2𝑥 𝑑𝑦 Solve the given differantial equation. Nonhomogeneous solution (𝑚𝑃 ):
(VPM)
1. order, nonlinear, nonhomogeneous dif. Equ. 2 2
𝑦 2 𝑒 2𝑥 𝑑𝑥 = 4 + 𝑒 2𝑥 𝑑𝑦 y: dependent, x: independent variable 𝑚𝐻 = 𝑐. 𝑒
−
𝑦 ⟹ 𝑚𝑃 = 𝑐(𝑦). 𝑒
−
𝑦

It can be separated into variables. 𝑑𝑚𝑃 𝑑𝑐 −2 2 −2

It has nonlinearity due to 𝑦 2 and 𝑒 2𝑥 . It can be made linear. = 𝑒 + 𝑐. 2 𝑒 𝑦
𝑦
𝑑𝑦 𝑑𝑦 𝑦
𝑑𝑚
𝑒 2𝑥 = 𝑚 ⟹ 2𝑒 2𝑥 𝑑𝑥 = 𝑑𝑚 ⟹ 𝑒 2𝑥 𝑑𝑥 = 2 8 By writing 𝑚𝑃′ and 𝑚𝑃
2 ′
𝑚𝑃 − 2 𝑚 𝑃 = 2 ⟹ into the nonhom.
𝑑𝑚 𝑑𝑚 8 + 2𝑚 𝑦 𝑦 equation:
𝑦 2 = 4 + 𝑚 𝑑𝑦 ⟹ = ⟹ ′
2 𝑑𝑦 𝑦2 𝑚𝑃 𝑚𝑃
𝑑𝑚 2 8 𝑑𝑐 −𝑦
2
2 −
2
2
2
− 8
− 2 𝑚 = 2 ⟹ Now, it is linear, m: dependent, 𝑒 + 𝑐. 2 𝑒 𝑦 − 𝑐. 𝑒 𝑦 = ⟹
𝑑𝑦 𝑦 𝑦 y: independent variable 𝑑𝑦 𝑦 𝑦 2 𝑦2

𝑚𝐻 𝑚𝑃 𝑑𝑐 −𝑦
2
8 1
2 𝟐
𝑒 = ⟹ ‫= 𝑐𝑑 ׬‬ 8 න 2 𝑒 𝑦 𝑑𝑦 ⟹ 𝒄 = −𝟒𝒆 𝒚
Homogeneous solution (𝑚𝐻 ): 𝑑𝑦 𝑦2 𝑦
𝟐 2
2 𝑑𝑚 2 𝑑𝑚 2 𝒚 −
′ 𝑚𝑃 = −𝟒𝒆 . 𝑒 𝑦 ⟹ 𝑚𝑃 = −4
𝑚 − 2𝑚=0⟹ = 𝑚⟹න = න 2 𝑑𝑦
𝑦 𝑑𝑦 𝑦 2 𝑚 𝑦 −
𝟐

2 𝟐 𝑚 = 𝑚𝐻 + 𝑚𝑃 ⟹ 𝒎 = 𝒄. 𝒆 𝒚 −𝟒
−
𝑙𝑛𝑚 = − + 𝑙𝑛𝑐 ⟹ 𝒎𝑯 = 𝒄. 𝒆 𝒚 𝟐
𝑦 −
2 𝟏 −
𝑒 2𝑥 = 𝑐. 𝑒 𝑦 −4⟹ 𝒙 = 𝐥𝐧 𝒄. 𝒆 𝒚 − 𝟒
𝟐
2. Method: Formulation solution

𝑑𝑚 2 8 2 8 2 8
′
− 𝑚 = 2 ⟹ 𝑚 − 𝑦2 𝑚 = 𝑦2 ⟹ 𝑝 𝑦 = − and 𝑞 𝑦 =
𝑑𝑦 𝑦 2 𝑦 𝑦2 𝑦2

𝑚 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑦 𝑑𝑦 + 𝑐 ⟹

2 𝟐
ℎ = ‫ ⟹ 𝑦𝑑𝑝 ׬‬ℎ = ‫ ׬‬− 𝑑𝑦 ⟹ 𝒉 =
𝑦2 𝒚

2 2 8 2 𝑑𝑣 2 2 2 2
− − 𝑣 − 𝑣 − −
𝑚= 𝑒 𝑦 න𝑒 𝑦 𝑑𝑦 + 𝑐 = 𝑒 𝑦 8න𝑒 − +𝑐 =𝑒 𝑦 −4𝑒 + 𝑐 = 𝑒 𝑦 𝑦 𝑦
−4𝑒 + 𝑐 = 𝑐𝑒 − 4
𝑦2 2
2 2 1 𝑑𝑣
= 𝑣 ⟹ − 2 𝑑𝑦 = 𝑑𝑣 ⟹ 2 𝑑𝑦 = −
𝑦 𝑦 𝑦 2
2
−
𝑚= 𝑒 2𝑥 = 𝑐𝑒 𝑦 −4⟹
𝟏 𝟐
−
𝒙 = 𝐥𝐧 𝒄. 𝒆 𝒚 − 𝟒
𝟐
Ex-6:
𝑥𝑦 2 𝑦 ′ = 𝑦 3 − 𝑥 3 Solve the given differantial equation. Nonhomogeneous solution (𝑢𝑃 ):
1. order, nonlinear, nonhomogeneous dif. Equ. (VPM)
𝑥𝑦 2 𝑦 ′ = 𝑦 3 − 𝑥 3 y: dependent, x: independent variable
𝑢𝐻 = 𝑐. 𝑥 3 ⟹ 𝑢𝑃 = 𝑐(𝑥). 𝑥 3
It can be separated into variables.
𝑑𝑢𝑃 𝑑𝑐 3
It has nonlinearity due to 𝑦 3 and 𝑥 3 . It should be made linear. = 𝑥 + 3𝑐𝑥 2
𝑑𝑥 𝑑𝑥
1 𝑑𝑦 𝑑𝑢
𝑦3 = 𝑢 ⟹ (3𝑦 2 𝑑𝑦 = 𝑑𝑢) ⟹ 3𝑦 2 = ⟹ 3 By writing 𝑚𝑃′ and 𝑚𝑃
′
𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑢𝑃 − 𝑢𝑃 = −3𝑥 2 ⟹ into the non-hom.
1 𝑑𝑢 1 𝑑𝑢 𝑥 equation:
2 ′ = 𝑢 − 𝑥3
⟹𝑦 𝑦 = ⟹𝑥 𝑢𝑃′ 𝑢𝑃
3 𝑑𝑥 3 𝑑𝑥
𝑑𝑢 3 𝑑𝑐 3 3
− 𝑢 = −3𝑥 2 ⟹ Now, it is linear, u: dependent, 𝑥 + 3𝑐𝑥 − 𝑐. 𝑥 3 = −3𝑥 2 ⟹
2
𝑑𝑥 𝑥 x: independent variable 𝑑𝑥 𝑥
𝑢𝐻 𝑢𝑃 𝑑𝑐 3 𝑑𝑥
2
𝑥 = −3𝑥 ⟹ න𝑑𝑐 = −3 න ⟹ 𝒄 = −𝟑𝒍𝒏𝒙
Homogeneous solution (𝑢𝐻 ): 𝑑𝑥 𝑥
𝑑𝑢 3 𝑑𝑢 3 𝑑𝑢 3 𝑢𝑃 = −3𝑙𝑛𝑥. 𝑥 3 ⟹ 𝑢𝑃 = −3𝑙𝑛𝑥. 𝑥 3
− 𝑢=0⟹ = 𝑢⟹න = න 𝑑𝑥
𝑑𝑥 𝑥 𝑑𝑥 𝑥 𝑢 𝑥
𝑢 = 𝑢𝐻 + 𝑢𝑃 ⟹ 𝒖 = 𝒄. 𝒙𝟑 − 𝟑𝒍𝒏𝒙. 𝒙𝟑
𝑙𝑛𝑢 = 3𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑢 = 𝑙𝑛 𝑐. 𝑥 3
1 𝟑 𝟏
𝒖𝑯 = 𝒄. 𝒙𝟑 𝑦 = 𝑐. 𝑥 + 𝑙𝑛 3 . 𝑥 ⟹ 𝒚 = 𝒙. 𝒄 + 𝒍𝒏 𝒙𝟑
3 3 3
𝑥
2. Method: Formulation solution

𝑑𝑢 3 3 3
2 ′ 2
− 𝑢 = −3𝑥 ⟹ 𝑢 − 𝑢 = −3𝑥 ⟹ 𝑝 𝑥 = − and 𝑞 𝑥 = −3𝑥 2
𝑑𝑥 𝑥 𝑥 𝑥

𝑢 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

3 𝟏
ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬− 𝑑𝑥 = −3 ln 𝑥 ⟹ 𝒉 = 𝐥𝐧
𝑥 𝒙𝟑

1 1 1 1
− ln 3 ln 3
𝑢= 𝑒 𝑥 න𝑒 𝑥 −3𝑥 2 𝑑𝑥 + 𝑐 = 𝑥 3 න −3𝑥 2 + 𝑐 = 𝑥 3 −3 න + 𝑐
𝑥3 𝑥

1 1
𝑢 = −3𝑥 3 ln 𝑥 + 𝑐𝑥 3 ⟹ 𝑢 = 𝑥 3 𝑙𝑛 3 + 𝑐𝑥 3 ⟹ 𝑦 3 = 𝑥 3 𝑙𝑛
3 + 𝑐𝑥 3
𝑥 𝑥

𝑢 = 𝑦3

𝟑 𝟏
𝒚 = 𝒙. 𝒄 + 𝒍𝒏 𝟑
𝒙
Ex-7:
𝑦 ′ − 1 = 𝑒 −𝑦 . 𝑠𝑖𝑛𝑥 Solve the given differantial equation.

1. order, nonlinear, nonhomogeneous dif. Equ. Nonhomogeneous solution (𝑚𝑃 ):

𝑦′ −1= 𝑒 −𝑦 . 𝑠𝑖𝑛𝑥 y: dependent, x: independent variable (VPM)

It can be separated into variables. 𝑚𝐻 = 𝑐. 𝑒 𝑥 ⟹ 𝑚𝑃 = 𝑐(𝑥). 𝑒 𝑥

It has nonlinearity due to 𝑒 −𝑦 and 𝑠𝑖𝑛𝑥. It should be made linear. 𝑑𝑚𝑃 𝑑𝑐 𝑥
= 𝑒 + 𝑐𝑒 𝑥
𝑒 𝑦 = 𝑚 ⟹ 𝑚′ = 𝑦 ′ 𝑒 𝑦 and 𝑒 𝑦 (𝑦 ′ − 1 = 𝑒 −𝑦 . 𝑠𝑖𝑛𝑥) ⟹ 𝑑𝑥 𝑑𝑥
′ By writing 𝑚𝑃′ and 𝑚𝑃
𝑚′ − 𝑚 = 𝑠𝑖𝑛𝑥 ⟹ Now, it is linear, m: dependent,
𝑚𝑃 − 𝑚𝑃 = 𝑠𝑖𝑛𝑥 ⟹
into the nonhom.
x: independent variable
𝑑𝑚 𝑚𝑃′ 𝑚𝑃 equation:
− 𝑚 = 𝑠𝑖𝑛𝑥 ⟹
𝑑𝑥 𝑑𝑐 𝑥
𝑒 + 𝑐𝑒 𝑥 − 𝑐𝑒 𝑥 = 𝑠𝑖𝑛𝑥 ⟹
𝑚𝐻 𝑚𝑃 𝑑𝑥
Homogeneous solution (𝑚𝐻 ): 𝑑𝑐
= 𝑒 −𝑥 𝑠𝑖𝑛𝑥 ⟹ න𝑑𝑐 = න𝑒 −𝑥 𝑠𝑖𝑛𝑥𝑑𝑥 ⟹
𝑑𝑥
𝑑𝑚 𝑑𝑚 𝑑𝑚
−𝑚 = 0⟹ =𝑚⟹න = න𝑑𝑥
𝑑𝑥 𝑑𝑥 𝑚
𝐼 = න𝑒 −𝑥 𝑠𝑖𝑛𝑥𝑑𝑥 ⟹ 𝑢 = 𝑒 −𝑥 ⟹ 𝑑𝑢 = −𝑒 −𝑥 𝑑𝑥
𝑙𝑛𝑚 = 𝑥 + 𝑙𝑛𝑐 ⟹ 𝑚 = 𝑐. 𝑒 𝑥 𝑑𝑣 = 𝑠𝑖𝑛𝑥𝑑𝑥 ⟹ 𝑣 = −𝑐𝑜𝑠𝑥
𝒎𝑯 = 𝒄. 𝒆𝒙
𝐼 = න𝑒 −𝑥 𝑠𝑖𝑛𝑥𝑑𝑥 = න𝑢𝑑𝑣 = 𝑢. 𝑣 − න 𝑣𝑑𝑢 ⟹
 𝟏
𝐼 = 𝑢. 𝑣 − න 𝑣𝑑𝑢 ⟹ 𝐼 = −𝑒 −𝑥 . 𝑐𝑜𝑠𝑥 − න 𝑐𝑜𝑠𝑥. 𝑒 −𝑥 𝑑𝑥 𝑚 = 𝑚𝐻 + 𝑚𝑃 ⟹ 𝒎 = 𝒄. 𝒆𝒙 − (𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙)
𝟐
𝐼1 1
𝑚 = 𝑒 𝑦 = 𝑐. 𝑒 𝑥 − (𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥)
𝑢 = 𝑒 −𝑥 ⟹ 𝑑𝑢 = −𝑒 −𝑥 𝑑𝑥 2
𝑑𝑣 = 𝑐𝑜𝑠𝑥𝑑𝑥 ⟹ 𝑣 = 𝑠𝑖𝑛𝑥 𝟏
𝒚 = 𝒍𝒏 𝒄. 𝒆𝒙 − (𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙)
𝟐
𝐼1 = 𝑒 −𝑥 𝑠𝑖𝑛𝑥 + න 𝑠𝑖𝑛𝑥. 𝑒 −𝑥 𝑑𝑥

𝐼1

𝐼 = න𝑒 −𝑥 𝑠𝑖𝑛𝑥𝑑𝑥 = −𝑒 −𝑥 . 𝑐𝑜𝑠𝑥 − 𝑒 −𝑥 𝑠𝑖𝑛𝑥 + න 𝑠𝑖𝑛𝑥. 𝑒 −𝑥 𝑑𝑥 ⟹

𝐼
𝐼 = −𝑒 −𝑥 . 𝑐𝑜𝑠𝑥 − 𝑒 −𝑥 𝑠𝑖𝑛𝑥 − 𝐼 ⟹ 2𝐼 = −𝑒 −𝑥 (𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥)
−𝑒 −𝑥
𝐼= (𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥)
2
−𝒆−𝒙 −𝑒 −𝑥
𝒄= (𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙) ⟹ 𝑚𝑃 = 𝑐(𝑥). 𝑒 𝑥 ⟹ 𝑚𝑃 = (𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥). 𝑒 𝑥
𝟐 2
𝟏
𝒎𝑷 = − (𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙)
𝟐
2. Method: Formulation solution
𝑑𝑚
− 𝑚 = sin 𝑥 ⟹ 𝑚 − 𝑚 = sin 𝑥 ⟹ 𝑝 𝑥 = −1 and 𝑞 𝑥 = sin 𝑥
′
𝑑𝑥

𝑚 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬−1 𝑑𝑥 ⟹ 𝒉 = −𝒙

−𝑒 −𝑥
𝑚 = 𝑒 −(−𝑥) න 𝑒 −𝑥 sin 𝑥 𝑑𝑥 + 𝑐 = 𝑒 𝑥 න 𝑒 −𝑥 sin 𝑥 𝑑𝑥 + 𝑐 ⟹ 𝑚 = 𝑒 𝑥 (𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥) + 𝑐 ⟹
2
−𝑒 −𝑥
(𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥)
2
𝟏 1
𝒎 = − (𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙) + 𝒄𝒆 ⟹ 𝑚 = 𝑒 = − (𝑐𝑜𝑠𝑥 + 𝑠𝑖𝑛𝑥) + 𝑐𝑒 𝑥
𝒙 𝑦
𝟐 2
𝟏
𝒚 = 𝒍𝒏 𝒄. 𝒆𝒙 − (𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙)
𝟐
Ex-8:

𝑦′ + 𝑦 = 3 Solve the given differantial equation.

𝑦𝐻 𝑦𝑃
It can be separated into variables and it is linear. 𝒚𝑷 = 𝟑 ⟹ 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝒚 = 𝟑 + 𝒄. 𝒆−𝒙
Homogeneous solution (yH):
𝑑𝑦 𝑑𝑦
𝑦′ + 𝑦 = 0 ⟹ = −𝑦 ⟹ න = − න𝑑𝑥 ⟹ 2. Method: Formulation solution
𝑑𝑥 𝑦
𝑦 𝑦 ′ + 𝑦 = 3 ⟹ 𝑝 𝑥 = 1 and 𝑞 𝑥 = 3
𝑙𝑛𝑦 = −𝑥 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑦 − 𝑙𝑛𝑐 = −𝑥 ⟹ 𝑙𝑛 = −𝑥
𝑐
𝒚𝑯 = 𝒄. 𝒆−𝒙 y = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹
Nonhomogeneous solution (𝑦𝑃 ):
𝑦𝐻 = 𝑐. 𝑒 −𝑥 ⟹ 𝑦𝑃 = 𝑐(𝑥). 𝑒 −𝑥 Variation of Parameters ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬1 𝑑𝑥 ⟹ 𝒉 = 𝒙
Method(VPM)
𝑑𝑦𝑃 𝑑𝑐 −𝑥 writing 𝑦𝑃′ and 𝑦𝑃 into
= 𝑒 − 𝑐𝑒 −𝑥 By
𝑑𝑥 𝑑𝑥 the nonhomogeneous equation:

𝑦𝑃′ 𝑦𝑃 y = 𝑒 −(𝑥) න 𝑒 𝑥 3 𝑑𝑥 + 𝑐 = 𝑒 −𝑥 3 න 𝑒 𝑥 𝑑𝑥 + 𝑐
𝑑𝑐 −𝑥 −𝑥 −𝑥
𝑑𝑐 −𝑥
𝑒 − 𝑐𝑒 + 𝑐𝑒 = 3 ⟹ 𝑒 =3⟹
𝑑𝑥 𝑑𝑥 y = 𝑒 −𝑥 3𝑒 𝑥 + 𝑐
න𝑑𝑐 = න3𝑒 𝑥 𝑑𝑥 ⟹ 𝑐 𝑥 = 3𝑒 𝑥 ⟹ 𝑦𝑃 = 3𝑒 𝑥 . 𝑒 −𝑥 ⟹ 𝒚 = 𝟑 + 𝒄𝒆−𝒙
Ex-9:

𝑥 2 𝑙𝑛𝑥. 𝑦 ′ + 𝑥𝑦 = 1 Solve the given differantial equation.

1 2 𝑙𝑛𝑥. 𝑦 ′ + 𝑥𝑦 = 1 ⟹ Nonhomogeneous solution (𝑦𝑃 ):

𝑥
𝑥 2 𝑙𝑛𝑥
𝑦 1 𝑐 𝑐(𝑥) Variation of Parameters
𝑦′ + = 𝑦𝐻 = ⟹ 𝑦𝑃 = Method(VPM)
𝑥𝑙𝑛𝑥 𝑥 2 𝑙𝑛𝑥 𝑙𝑛𝑥 𝑙𝑛𝑥
𝑑𝑦𝑃 𝑑𝑐 1 1/𝑥 By writing 𝑦𝑃′ and 𝑦𝑃 into
𝑦𝐻 𝑦𝑃 = −𝑐
𝑑𝑥 𝑑𝑥 𝑙𝑛𝑥 (𝑙𝑛𝑥)2 the nonhomogeneous equation:
Homogeneous solution (yH): 𝑦𝑃 1
𝑦 𝑑𝑦 𝑦 𝑑𝑦 𝑑𝑥 𝑦𝑃 ′ + = 2
′
𝑦 + =0⟹ =− ⟹න = −න ⟹ 𝑥𝑙𝑛𝑥 𝑥 𝑙𝑛𝑥
𝑥𝑙𝑛𝑥 𝑑𝑥 𝑥𝑙𝑛𝑥 𝑦 𝑥𝑙𝑛𝑥 𝑑𝑐 1 1 1 1
1 −𝑐 +𝑐 = ⟹
𝑙𝑛𝑥 = 𝑢 ⟹ 𝑑𝑥 = 𝑑𝑢 𝑑𝑥 𝑙𝑛𝑥 𝑥. (𝑙𝑛𝑥)2 𝑥. (𝑙𝑛𝑥)2 𝑥 2 𝑙𝑛𝑥
𝑥 𝑑𝑐 1 𝑑𝑥 1
𝑑𝑦 𝑑𝑢 = 2 ⟹ න𝑑𝑐 = න 2 ⟹ 𝑐 𝑥 = − ⟹
න = −න ⟹ 𝑙𝑛𝑦 = −𝑙𝑛𝑢 + 𝑙𝑛𝑐 ⟹ 𝑑𝑥 𝑥 𝑥 𝑥
𝑦 𝑢
1 1 𝟏
𝑐 𝑐 𝑐 𝑦𝑃 = − . ⟹ 𝒚𝑷 = −
𝑙𝑛𝑦 = 𝑙𝑛 ⟹ 𝑙𝑛𝑦 = 𝑙𝑛 ⟹𝑦= 𝑥 𝑙𝑛𝑥 𝒙𝒍𝒏𝒙
𝑢 𝑙𝑛𝑥 𝑙𝑛𝑥
𝑐 1 𝒄 − 𝟏/𝒙
𝒄 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝑦 = − ⟹ 𝒚=
𝒚𝑯 = 𝑙𝑛𝑥 𝑥𝑙𝑛𝑥 𝒍𝒏𝒙
𝒍𝒏𝒙
 2. Method: Formulation solution
1 2 ′ ′
𝑦 1 1 1
𝑥 𝑙𝑛𝑥. 𝑦 + 𝑥𝑦 = 1 ⟹ 𝑦 + = ⟹ 𝑝 𝑥 = and 𝑞 𝑥 =
𝑥 2 𝑙𝑛𝑥 𝑥𝑙𝑛𝑥 𝑥 2 𝑙𝑛𝑥 𝑥𝑙𝑛𝑥 𝑥 2 𝑙𝑛𝑥

𝑦 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

1 1 1 1
ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫׬‬ 𝑑𝑥 = ‫׬‬ 𝑑𝑥 = ‫ = 𝑢𝑑 ׬‬ln 𝑢 ⟹ 𝒉 = 𝐥𝐧 𝐥𝐧 𝒙
𝑥𝑙𝑛𝑥 ln 𝑥 𝑥 𝑢
1
ln 𝑥 = 𝑢 ⟹ 𝑑𝑥 = 𝑑𝑢
𝑥

1 1 1 1 1 1 1
𝑦 = 𝑒 − ln ln 𝑥 න 𝑒 ln ln 𝑥 𝑑𝑥 + 𝑐 = න ln 𝑥 𝑑𝑥 + 𝑐 = න 𝑑𝑥 + 𝑐 = − +𝑐 ⟹
𝑥 2 𝑙𝑛𝑥 ln 𝑥 𝑥 2 𝑙𝑛𝑥 ln 𝑥 𝑥2 ln 𝑥 𝑥

𝒄 − 𝟏/𝒙
𝒚=
𝒍𝒏𝒙
Ex-10:

𝑥𝑦 ′ = 𝑦 + 2𝑥 3 Solve the given differantial equation.

1 𝑦 න𝑑𝑐 = න2𝑥𝑑𝑥 ⟹ 𝑐 𝑥 = 𝑥 2 ⟹ 𝑦𝑃 = 𝑥 2 . 𝑥 ⟹ 𝒚𝑷 = 𝒙𝟑
𝑥𝑦 − 𝑦 = 2𝑥 ⟹ 𝑦 − = 2𝑥 2
′ 3 ′
𝑥 𝑥
𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝑦 = 𝑐𝑥 + 𝑥 3
𝑦𝐻 𝑦𝑃
Homogeneous solution (yH): 𝒚 = 𝒙(𝒙𝟐 + 𝒄)
′
𝑦 𝑑𝑦 𝑦 𝑑𝑦 𝑑𝑥
𝑦 − =0⟹ = ⟹න =න ⟹
𝑥 𝑑𝑥 𝑥 𝑦 𝑥
𝑙𝑛 𝑦 = 𝑙𝑛 𝑥 + 𝑙𝑛 𝑐 ⟹ 𝑙𝑛 𝑦 = 𝑙𝑛 𝑐𝑥 ⟹ 𝑦 = 𝑐𝑥
𝒚𝑯 = 𝒄𝒙
Nonhomogeneous solution (𝑦𝑃 ):
(VPM)
𝑦𝐻 = 𝑐𝑥 ⟹ 𝑦𝑃 = 𝑐 𝑥 . x
𝑑𝑦𝑃 𝑑𝑐 By writing 𝑦𝑃′ and 𝑦𝑃 into
= 𝑥+𝑐 the nonhomogeneous equation:
𝑑𝑥 𝑑𝑥
𝑦𝑃′ 𝑦𝑃
𝑑𝑐 1 𝑑𝑐
𝑥 + 𝑐 − 𝑐𝑥 = 2𝑥 2 ⟹ = 2𝑥 ⟹
𝑑𝑥 𝑥 𝑑𝑥
2. Method: Formulation solution
1 1 1
𝑥𝑦 ′ = 𝑦 + 2𝑥 3 ⟹ 𝑦 ′ − 𝑦 = 2𝑥 2 ⟹ 𝑝 𝑥 = − 𝑥 and 𝑞 𝑥 = 2𝑥 2
𝑥 𝑥

𝑦 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

1 𝟏
ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬− 𝑑𝑥 = −ln 𝑥 ⟹ 𝒉 = 𝐥𝐧
𝑥 𝒙

− ln
1
ln
1 1 𝑥2
𝑦=𝑒 𝑥 න𝑒 𝑥 2𝑥 2 2
𝑑𝑥 + 𝑐 = 𝑥 න 2𝑥 𝑑𝑥 + 𝑐 = 𝑥 2 න 𝑥 𝑑𝑥 + 𝑐 = 𝑥 2 + 𝑐 ⟹
𝑥 2

𝒚 = 𝒙 𝒙𝟐 + 𝒄
Ex-11:

𝑥 2 𝑦 ′ + 𝑥𝑦 + 2 = 0 Solve the given differantial equation.

𝑦𝑃′ 𝑦𝑃
1 2 ′ ′
𝑦 2
2
𝑥 𝑦 + 𝑥𝑦 + 2 = 0 ⟹ 𝑦 + = − 2 𝑑𝑐 1 1 𝑐1 2 𝑑𝑐 2
𝑥 𝑥 𝑥
−𝑐 2+ =− 2⟹ =− ⟹
𝑦𝐻 𝑦𝑃 𝑑𝑥 𝑥 𝑥 𝑥𝑥 𝑥 𝑑𝑥 𝑥
Homogeneous solution (yH): 2 1
′
𝑦 𝑑𝑦 𝑦 𝑑𝑦 𝑑𝑥 න𝑑𝑐 = න − 𝑑𝑥 ⟹ 𝑐 𝑥 = −2𝑙𝑛𝑥 ⟹ 𝑦𝑃 = −2𝑙𝑛𝑥. ⟹
𝑦 + =0⟹ =− ⟹න = −න ⟹ 𝑥 𝑥
𝑥 𝑑𝑥 𝑥 𝑦 𝑥
𝑐 𝑐 𝟐𝒍𝒏𝒙
𝑙𝑛 𝑦 = −𝑙𝑛 𝑥 + 𝑙𝑛 𝑐 ⟹ 𝑙𝑛 𝑦 = 𝑙𝑛 ⟹𝑦= 𝒚𝑷 = −
𝒄 𝑥 𝑥 𝒙
𝒄 𝟐𝒍𝒏𝒙
𝒚𝑯 = 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝒚= −
𝒙 𝒙 𝒙
Nonhomogeneous solution (𝑦𝑃 ):
(VPM)
𝑐 𝑐(𝑥)
𝑦𝐻 = ⟹ 𝑦𝑃 =
𝑥 𝑥
𝑑𝑦𝑃 𝑑𝑐 1 1 By writing 𝑦𝑃′ and 𝑦𝑃 into
= − 𝑐 2 the nonhomogeneous equation:
𝑑𝑥 𝑑𝑥 𝑥 𝑥
2. Method: Formulation solution
1 2 ′ 1
′+ 𝑦=−
2 1 2
𝑥 𝑦 + 𝑥𝑦 + 2 = 0 ⟹ 𝑦 ⟹ 𝑝 𝑥 = and 𝑞 𝑥 = −
𝑥2 𝑥 𝑥2 𝑥 𝑥2

𝑦 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

1
ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫׬‬ 𝑑𝑥 = ln 𝑥 ⟹ 𝒉 = 𝐥𝐧 𝒙
𝑥

2 1 2 1 1 1
𝑦 = 𝑒 − ln 𝑥 න 𝑒 ln 𝑥 − 𝑑𝑥 + 𝑐 = න 𝑥 − 𝑑𝑥 + 𝑐 = −2 න 𝑑𝑥 + 𝑐 = −2ln 𝑥 + 𝑐 ⟹
𝑥2 𝑥 𝑥2 𝑥 𝑥 𝑥

𝒄 𝟐𝒍𝒏𝒙
𝒚= −
𝒙 𝒙
Ex-12:

′
𝑦 2
𝑦 − = 𝑥+1 Solve the given differantial equation.
𝑥+1 𝑦𝑃′ 𝑦𝑃
𝑦
𝑦′ − = 𝑥+1 2
𝑑𝑐 1
𝑥+1 𝑥+1 −𝑐− 𝑐(𝑥 + 1) = 𝑥 + 1 2 ⟹
𝑑𝑥 𝑥+1
𝑦𝐻 𝑦𝑃
Homogeneous solution (yH): 1 2
𝑦 𝑑𝑦 𝑦 𝑑𝑦 𝑑𝑥 න𝑑𝑐 = න(𝑥 + 1)𝑑𝑥 ⟹ 𝑐 𝑥 = 𝑥 + 𝑥 ⟹
𝑦′ − =0⟹ = ⟹න =න ⟹ 2
𝑥+1 𝑑𝑥 𝑥 + 1 𝑦 𝑥+1 𝟏
𝒚𝑷 = ( 𝒙𝟐 + 𝒙)(𝒙 + 𝟏)
𝑙𝑛 𝑦 = 𝑙𝑛 𝑥 + 1 + 𝑙𝑛 𝑐 ⟹ 𝑙𝑛 𝑦 = 𝑙𝑛 𝑐 𝑥 + 1 ⟹ 𝟐
1 2
𝑦 = 𝑐(𝑥 + 1) ⟹ 𝒚𝑯 = 𝒄(𝒙 + 𝟏) 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝑦 = 𝑐(𝑥 + 1) + ( 𝑥 + 𝑥)(𝑥 + 1)
2

Nonhomogeneous solution (𝑦𝑃 ): 𝟏 𝟐

𝒚 = (𝒙 + 𝟏) 𝒙 +𝒙+𝒄
(VPM) 𝟐
𝑦𝐻 = 𝑐(𝑥 + 1) ⟹ 𝑦𝑃 = 𝑐(𝑥)(𝑥 + 1)

𝑑𝑦𝑃 𝑑𝑐 By writing 𝑦𝑃′ and 𝑦𝑃 into

= (𝑥 + 1) − 𝑐
𝑑𝑥 𝑑𝑥 the nonhomogeneous equation:
2. Method: Formulation solution
𝑦 1
𝑦′ − = 𝑥+1 2 ⟹𝑝 𝑥 =− and 𝑞 𝑥 = 𝑥+1 2
𝑥+1 𝑥+1

𝑦 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

1 𝟏
ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬− 𝑑𝑥 = − ln 𝑥 + 1 ⟹ 𝒉 = 𝐥𝐧
𝑥+1 𝒙+𝟏

−ln
1
ln
1
2 𝑑𝑥
1 2 𝑑𝑥
𝑦=𝑒 𝑥+1 න𝑒 𝑥+1 𝑥+1 + 𝑐 = (𝑥 + 1) න 𝑥+1 + 𝑐 = (𝑥 + 1) න(𝑥 + 1) 𝑑𝑥 + 𝑐 ⟹
𝑥+1

𝒙𝟐
𝒚 = (𝒙 + 𝟏) +𝒙+𝒄
𝟐
Ex-13:
3
𝑦 ′ + 3𝑥 2 𝑦 = 𝑥𝑒 −𝑥 Solve the given differantial equation for y 0 = 1.

𝑦𝐻 𝑑𝑐 1 2
𝑦𝑃 = 𝑥 ⟹ න𝑑𝑐 = න𝑥𝑑𝑥 ⟹ 𝑐 𝑥 = 𝑥 ⟹
𝑑𝑥 2
Homogeneous solution (yH):
𝑑𝑦 𝑑𝑦 𝟏 𝟐 −𝒙𝟑
𝑦 ′ + 3𝑥 2 𝑦 = 0 ⟹ = −3𝑥 2 𝑦 ⟹ න = −3 න𝑥 2 𝑑𝑥 ⟹ 𝒚𝑷 = 𝒙 𝒆
𝑑𝑥 𝑦 𝟐
3
𝑦 1
−𝑥 3 + 𝑥 2 𝑒 −𝑥 3
𝑙𝑛 𝑦 = −𝑥 + 𝑙𝑛 𝑐 ⟹ 𝑙𝑛 = −𝑥 3 ⟹ 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝑦 = 𝑐𝑒
𝑐 2
3 𝟑
𝑦 = 𝑐𝑒 −𝑥 ⟹ 𝒚𝑯 = 𝒄𝒆−𝒙 𝟑 𝟏
𝒚 = 𝒆−𝒙 (𝒄 + 𝒙𝟐 )
Nonhomogeneous solution (𝑦𝑃 ): 𝟐
(VPM)
𝑦𝐻 = 𝑐𝑒 −𝑥 3 ⟹ 𝑦𝑃 = 𝑐(𝑥)𝑒 −𝑥
3

𝑑𝑦𝑃 𝑑𝑐 −𝑥 3 3 By writing 𝑦𝑃′ and 𝑦𝑃 into

= 𝑒 − 3𝑥 2 𝑐𝑒 −𝑥 the nonhomogeneous equation:
𝑑𝑥 𝑑𝑥
𝑦𝑃′ 𝑦𝑃
𝑑𝑐 −𝑥 3 3 3 3
𝑒 − 3𝑥 2 𝑐𝑒 −𝑥 + 3𝑥 2 𝑐𝑒 −𝑥 = 𝑥𝑒 −𝑥 ⟹
𝑑𝑥
2. Method: Formulation solution

3 3
𝑦 ′ + 3𝑥 2 𝑦 = 𝑥𝑒 −𝑥 ⟹ 𝑝 𝑥 = 3𝑥 2 and 𝑞 𝑥 = 𝑥𝑒 −𝑥

𝑦 = 𝑒 −ℎ න 𝑒 ℎ 𝑞 𝑥 𝑑𝑥 + 𝑐 ⟹

ℎ = ‫ ⟹ 𝑥𝑑𝑝 ׬‬ℎ = ‫ ׬‬3𝑥 2 𝑑𝑥 = 𝑥 3 ⟹ 𝒉 = 𝒙𝟑

3 3 3 3 3 𝑥2
𝑦= 𝑒 −𝑥 න 𝑒𝑥 𝑥𝑒 −𝑥 𝑑𝑥 + 𝑐 = 𝑒 −𝑥 න 𝑥 𝑑𝑥 + 𝑐 = 𝑒 −𝑥 +𝑐
2

𝟑 𝒙𝟐
𝒚= 𝒆−𝒙 +𝒄
𝟐
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Nonlinear differential equations: Riccati and
Bernoulli Equations

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
2.4. Nonlinear differential equations: Riccati and Bernoulli Equations
Some nonlinear differential equations can be solved by reducing them to linear form with special transformations. These
equations are named after the names of the scientists who put forward the transformations. Two typical examples of these
are Bernoulli and Riccati. Since they are similar in form to 1st order linear differential equations, transformations on
reduction to linear differential equations are based on the solution. Common formats are as follows:

𝑑𝑦
+ 𝑝(𝑥)𝑦 = 𝑄 𝑥 1. order, linear, non-homogeneous
𝑑𝑥
𝑦𝐻 𝑦𝑃 𝑦 = 𝑦𝐻 + 𝑦𝑃

𝑑𝑦 𝑛: Constant, nonlinear, Bernoulli equation

+ 𝑝 𝑥 𝑦 = 𝑄 𝑥 . 𝑦𝑛
𝑑𝑥
𝑑𝑦
+ 𝑝 𝑥 𝑦 = 𝑄 𝑥 . 𝑦2 2. degree, nonlinear, Bernoulli equation
𝑑𝑥

where 𝑝(𝑥) and 𝑄(𝑥) are continuous functions on the interval we’re working on and 𝑛 is a real number. Differential
equations in this form are called Bernoulli Equations.
Bernoulli Diff. Equ. Solution Method:
𝑑𝑦
+ 𝑝 𝑥 𝑦 = 𝑄 𝑥 . 𝑦 𝑛 ⟹ where 𝑛 is constant and it is a non-linear homogenous Bernoulli Equations.
𝑑𝑥
In order to solve this equation, we’ll first divide the differential equation by 𝒚𝒏 to get,

𝑦 −𝑛 𝑦′ + 𝑝 𝑥 𝑦1−𝑛 = 𝑄 𝑥

Now, we are going to use the transformation 𝒎 = 𝒚𝟏−𝒏 to express the differential equation in terms of dependent variable
𝒎. In the final case, 𝒎 is the new variable and 𝒚 is the old variable. With this substitution, a linear differential equation can
be obtained that we can easily solve.
𝑚 = 𝑦1−𝑛 ⟹ 𝑚′ = 1 − 𝑛 𝑦 1−𝑛 −1 𝑦′ ⟹
⟹ 𝑚′ = 1 − 𝑛 𝑦 −𝑛 𝑦′
From the original differential equation;
𝑦 −𝑛 𝑦 ′ + 𝑝 𝑥 𝑦1−𝑛 = 𝑄 𝑥 ⟹ 𝑦 −𝑛 𝑦 ′ = 𝑄(𝑥) − 𝑝 𝑥 𝑦1−𝑛 ⟹
⟹ 𝒚′ = 𝑄(𝑥) − 𝑝 𝑥 𝑦
𝑚′ = 1 − 𝑛 𝑦 −𝑛 𝒚′ = 1 − 𝑛 𝑦 −𝑛 𝑸(𝒙) − 𝒑 𝒙 𝒚 ⟹
⟹ 𝑚′ = 1 − 𝑛 𝑄(𝑥) − 𝑝 𝑥 𝑦1−𝑛 ⟹
Now the diff. equation is linear and can be solved by the
⟹ 𝒎′ + 𝟏 − 𝒏 𝒑 𝒙 𝒎 = 𝟏 − 𝒏 𝑸(𝒙) method of solving first order linear diff. equations!
 Riccati Diff. Equ. Solution Method:
𝑑𝑦
+ 𝑝 𝑥 . 𝑦 = 𝑄 𝑥 . 𝑦 2 + 𝑅(𝑥) ⟹ 𝑛: Constant, nonlinear, Riccati equation
𝑑𝑥
𝑦 = 𝑦𝐻 + 𝑦𝑃
𝑦𝐻 𝑦𝑃
In the Riccati equation, 𝑦𝑝 can be calculated using the functions arising from the function, derivative or integral of 𝑅(𝑥) and
the undetermined coefficients method. The non-homogenous (particular) solution also can be predict trial and error method.
After the calulation of 𝑦𝑝 the homogenous solution of 𝑦𝐻 can be calculated. To calculate the homogenous section the given
equation below can be used;

In the solution of Riccati equation, 𝑦𝑃 is found firstly.

𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ it is passed from Riccati to Bernoulli equation.

1
𝑦= + 𝑦𝑃 ⟹ it is passed from Riccati to linear equation.
𝑚

Here, the particular solution (𝑦𝑃 ) is any function that satisfies the main differential equation. It got this name because the
person who solved the problem previously determined its functional structure. According to the structure of the differential
equation, it is estimated that it provides the main differential equation. The reason for looking for the particular solution is
to eliminate the function 𝑅 𝑥 in the equation. The particular solution can be found from the function 𝑅 𝑥 . If the function
𝑅 𝑥 is eliminated, the equation becomes a 2nd order Bernoulli differential equation.
 Example: 𝑦 ′ + 𝑦 = 𝑥𝑦 3 Solve the given
differantial equation. 𝑑𝑚𝑃 𝑑𝑐 2𝑥
= 𝑒 + 2𝑐. 𝑒 2𝑥
𝑑𝑥 𝑑𝑥
𝑛 = 3, Bernoulli equation
1 𝑑𝑚𝑃 1 𝑑𝑐 2𝑥
1 𝑦′ 1 − + 𝑚𝑃 = 𝑥 ⟹ − 𝑒 + 2𝑐. 𝑒 2𝑥 + 𝑐. 𝑒 2𝑥 = 𝑥 ⟹
. 𝑦 ′ + 𝑦 = 𝑥𝑦 3 ⟹ + =𝑥⟹ 2 𝑑𝑥 2 𝑑𝑥
𝑦3 𝑦3 𝑦2
1 𝑑𝑐 2𝑥 𝑢 = 𝑥 ⟹ 𝑑𝑢 = 𝑑𝑥
1 2𝑦 ′
𝑑𝑚 𝑦′
1 𝑑𝑚 − 𝑒 = 𝑥 ⟹ න𝜕𝑐 = −2 න 𝑥. 𝑒 −2𝑥 𝜕𝑥 𝑑𝑣 = 𝑒 −2𝑥 𝜕𝑥
= 𝑚 ⟹ − = ⟹ = − (Linearization)
2 𝑑𝑥
𝑦2 𝑦3 𝑑𝑥 𝑦3 2 𝑑𝑥 1
𝑢 𝑑𝑣 𝑣 = − 𝑒 −2𝑥
1 𝑑𝑚 ‫ 𝑣𝑢 = 𝑣𝜕𝑢 ׬‬− ‫( 𝑢𝑑𝑣 ׬‬K. İnt.) 2
− +𝑚 = 𝑥 Now the equation is linear. 𝟏 𝟏
2 𝑑𝑥 1 −2𝑥 1
න𝜕𝑐 = 𝑥. − 𝑒 + න 𝑒 −2𝑥 𝜕𝑥 ⟹𝒄 = − 𝒙. 𝒆−𝟐𝒙 − 𝒆−𝟐𝒙
2 2 𝟐 𝟒
𝑚𝐻 𝑚𝑃
𝟏 𝟏
Homogeneous solution (𝑚𝐻 ): 𝑚𝑃 = 𝑐 𝑥 . 𝑒 2𝑥 ⟹ 𝒎𝑷 = − 𝒙+
𝟐 𝟐
1 𝑑𝑚 𝑑𝑚 𝑑𝑚
− +𝑚 =0⟹ = 2𝑚 ⟹ න = 2 න𝑑𝑥 𝟐𝒙
𝟏 𝟏
2 𝑑𝑥 𝑑𝑥 𝑚 𝑚 = 𝑚𝐻 + 𝑚𝑃 ⟹ 𝒎 = 𝒄. 𝒆 − − 𝒙+
𝟐 𝟐
𝑙𝑛𝑚 = 2𝑥 + 𝑙𝑛𝑐 ⟹ 𝒎𝑯 = 𝒄. 𝒆𝟐𝒙
1 2𝑥 −
1 1 𝟏
Nonhomogeneous solution (𝑚𝑃 ): = 𝑐. 𝑒 𝑥 + ⟹ 𝒚 =
𝑦2 2 2 1 1
𝟏/𝟐

(VPM) 𝒄. 𝒆𝟐𝒙 − 𝑥 +
2 2
𝑚𝐻 = 𝑐. 𝑒 2𝑥 ⟹ 𝑚𝑃 = 𝑐(𝑥). 𝑒 2𝑥
Example: 3𝑦′ + 𝑦 = 1 − 2𝑥 𝑦 4 Solve the given
differantial equation. 𝑑𝑚𝑃 𝑑𝑐 𝑥
= 𝑒 + 𝑐. 𝑒 𝑥
𝑑𝑥 𝑑𝑥
𝑛 = 4, Bernoulli equation
𝑑𝑚𝑃 𝑑𝑐
1 3𝑦′ 1 − + 𝑚𝑃 = 1 − 2𝑥 ⟹ − 𝑒 𝑥 − 𝑐. 𝑒 𝑥 + 𝑐. 𝑒 𝑥 = 1 − 2𝑥 ⟹
. 3𝑦′ + 𝑦 = 1 − 2𝑥 𝑦 4 ⟹ + = 1 − 2𝑥 ⟹ 𝑑𝑥 𝑑𝑥
𝑦4 𝑦4 𝑦3
1 3𝑦 ′ 𝑑𝑚 න𝜕𝑐 = − න 1 − 2𝑥 𝑒 −𝑥 𝜕𝑥 ⟹
=𝑚⟹ 4 =− (Linearization)
𝑦3 𝑦 𝑑𝑥 𝑢 = 𝑥 ⟹ 𝑑𝑢 = 𝑑𝑥
𝑑𝑚 න𝜕𝑐 = 2 න 𝑥𝑒 −𝑥 𝜕𝑥 − න 𝑒 −𝑥 𝜕𝑥 ⟹ 𝑑𝑣 = 𝑒 −𝑥 𝜕𝑥
− + 𝑚 = 1 − 2𝑥 The equation is linear. 𝑣 = −𝑒 −𝑥
𝑑𝑥 𝑢 𝑑𝑣
𝑚𝐻 𝑚𝑃
න𝜕𝑐 = 2 −𝑥𝑒 −𝑥 + න 𝑒 −𝑥 𝜕𝑥 − න 𝑒 −𝑥 𝜕𝑥 ⟹
Homogeneous solution (𝑚𝐻 ):
𝑑𝑚 𝑑𝑚
-𝑚′ + 𝑚 = 0 ⟹ − +𝑚 =0⟹න = ‫𝑥𝑑 ׬‬ 𝑐 = −2𝑥𝑒 −𝑥 + න 𝑒 −𝑥 𝜕𝑥 ⟹ 𝒄 = −𝟐𝒙𝒆−𝒙 − 𝒆−𝒙
𝑑𝑥 𝑚

𝑙𝑛𝑚 = 𝑥 + 𝑙𝑛𝑐 ⟹ 𝒎𝑯 = 𝒄. 𝒆𝒙 𝑚𝑃 = −2𝑥𝑒 −𝑥 − 𝑒 −𝑥 . 𝑒 𝑥 ⟹ 𝒎𝑷 = −𝟐𝒙 − 𝟏

Nonhomogeneous solution (𝑚𝑃 ): 𝑚 = 𝑚𝐻 + 𝑚𝑃 ⟹ 𝒎 = 𝒄. 𝒆𝒙 − 𝟐𝒙 − 𝟏
(VPM) 1 𝑥 − 2𝑥 − 1 ⟹
𝟏
= 𝑐. 𝑒 𝒚=
𝑚𝐻 = 𝑐. 𝑒 𝑥 ⟹ 𝑚𝑃 = 𝑐(𝑥). 𝑒 𝑥 𝑦3 (𝒄. 𝒆𝒙 − 𝟐𝒙 − 𝟏)𝟏/𝟑
Example: 2𝑥 3 𝑦′ = 𝑦 2 + 3𝑥 2 𝑦 Solve the given
differantial equation. Nonhomogeneous solution (𝑚𝑃 ):
2𝑥 3 𝑦 ′ − 3𝑥 2 𝑦 = 𝑦 3 𝑛 =3, Bernoulli equation (VPM)

1 𝑦 ′ 𝑥 2 𝑐 𝑐 𝑥 𝑑𝑚𝑃 1 𝑑𝑐 3𝑐
. 2𝑥 3 𝑦 ′ − 3𝑥 2 𝑦 = 𝑦 3 ⟹ 2𝑥 3 3 − 3 2 = 1 ⟹ 𝑚𝐻 = 3 ⟹ 𝑚 𝑃 = 3 ⟹ = 3 −
𝑦 3 𝑦 𝑦 𝑥 𝑥 𝑑𝑥 𝑥 𝑑𝑥 𝑥 4
1 2𝑦 ′ 𝑑𝑚 3 1 1 𝑑𝑐 3𝑐 3 𝑐 1
=𝑚⟹− 3 = 𝑚 𝑃 ′ + 𝑚𝑃 = − 3 ⟹ − + =− 3⟹
𝑦2 𝑦 𝑑𝑥 𝑥 𝑥 𝑥 3 𝑑𝑥 𝑥 4 𝑥 𝑥 3 𝑥
1 𝑑𝑚 1 𝑑𝑐 3𝑐 3𝑐 1
− 3 . −𝑥 3 − 3𝑥 2 𝑚 = 1 − + = − ⟹ න𝜕𝑐 = − න𝜕𝑥 ⟹
𝑥 𝑑𝑥 𝑥 3 𝑑𝑥 𝑥 4 𝑥 4 𝑥3
𝟏 𝒄 𝟏
3 1 𝒄 = −𝒙 and 𝒎𝑷 = − 𝟐 and 𝒎 = 𝟑 − 𝟐
𝑚′ + 𝑚 = − 3 𝒙 𝒙 𝒙
𝑥 𝑥
1 𝑐 1 1 𝑐−𝑥 𝑥 3
𝑚𝐻 𝑚𝑃 = 3 − 2 ⟹ 2 = 3 ⟹ 𝑦2 = ⟹
𝑦 2 𝑥 𝑥 𝑦 𝑥 𝑐−𝑥
The equation is linear.
𝒙𝟑
Homogeneous solution (𝑚𝐻 ): 𝒚=
3 𝑑𝑚 3 𝑑𝑚 3 𝒄−𝒙
′
𝑚 + 𝑚=0⟹ + 𝑚=0⟹න = − න 𝑑𝑥
𝑥 𝑑𝑥 𝑥 𝑚 𝑥
𝒄
𝑙𝑛𝑚 = −3𝑙𝑛𝑥 + 𝑙𝑛𝑐 ⟹ 𝒎𝑯 = 𝟑
𝒙
Example: 2𝑥𝑦 ′ = 10𝑥 3 𝑦 5 + 𝑦 Solve the given
differantial equation. Nonhomogeneous solution (𝑚𝑃 ):
2𝑥𝑦 ′ −𝑦= 10𝑥 3 𝑦 5 𝑛 =5, Bernoulli equation
(VPM)
1 ′ 3 5
𝑦′ 1 3 ⟹
𝑐 𝑐(𝑥) 𝑑𝑚𝑃 1 𝑑𝑐 2𝑐
. 2𝑥𝑦 − 𝑦 = 10𝑥 𝑦 ⟹ 2𝑥 − = 10𝑥 𝑚𝐻 = 2 ⟹ 𝑚𝑃 = 2 ⟹ = 2 −
𝑦5 𝑦5 𝑦4 𝑥 𝑥 𝑑𝑥 𝑥 𝑑𝑥 𝑥 3
2 1 𝑑𝑐 2𝑐 2𝑐
1 4𝑦 ′ 𝑑𝑚 2𝑦 ′ 1 𝑑𝑚 𝑚𝑃 ′ + 𝑚𝑃 = −20𝑥 2 ⟹ 2 − 3 + 3 = −20𝑥 2 ⟹
= 𝑚 ⟹ − = ⟹ = − ⟹ 𝑥 𝑥 𝑑𝑥 𝑥 𝑥
𝑦4 𝑦5 𝑑𝑥 𝑦5 2 𝑑𝑥 1 𝑑𝑐 𝑑𝑐
= −20𝑥 2 ⟹ = −20𝑥 4 ⟹
1 𝑑𝑚 2
𝑥 𝑑𝑥 𝑑𝑥
𝑥 − − 𝑚 = 10𝑥 3 ⟹
2 𝑑𝑥
න𝑑𝑐 = −20 න𝑥 4 𝑑𝑥 ⟹ 𝒄 = −𝟒𝒙𝟓
2 𝑥 2
− . − 𝑚′ − 𝑚 = 10𝑥 3 ⟹ 𝑚′ + 𝑚 = −20𝑥 2
𝑥 2 𝑥 𝑐(𝑥) −4𝑥 5 𝟑
𝑚𝑃 = 2 ⟹ 𝑚 𝑃 = ⟹ 𝒎 𝑷 = −𝟒𝒙
𝑚𝐻 𝑚𝑃 𝑥 𝑥2
𝒄
The equation is linear. 𝑚 = 𝑚𝐻 + 𝑚𝑝 ⟹ 𝒎 = 𝟐 − 𝟒𝒙𝟑
𝒙
Homogeneous solution (𝑚𝐻 ): 1 1 𝑐
2 𝑑𝑚 2 𝑑𝑚 𝑑𝑥 4 = 𝑚 ⟹ 4 = 2 − 4𝑥 3 ⟹
𝑚′ + 𝑚 = 0 ⟹ + 𝑚=0⟹න = −2 න 𝑦 𝑦 𝑥
𝑥 𝑑𝑥 𝑥 𝑚 𝑥 𝟏/𝟒
𝑐 𝒄 𝒙𝟐
𝑙𝑛 𝑚 = −2𝑙𝑛 𝑥 + l n 𝑐 ⟹ 𝑙𝑛 𝑚 = 𝑙𝑛 2 ⟹ 𝒎𝑯 = 𝟐 𝒚=
𝑥 𝒙 𝒄 − 𝟒𝒙𝟓
 2
Example: Solve the given differantial equation. −𝑚′ 1 1
𝑦 ′ = 𝑦 2 + 1 − 2𝑥 𝑦 + 𝑥 2 − 𝑥 + 1 + 1 = +𝑥 + 1 − 2𝑥 + 𝑥 + 𝑥 2 − 𝑥 + 1⟹
𝑚2 𝑚 𝑚
′
𝑦 ′ = 𝑦 2 + 1 − 2𝑥 𝑦 + 𝑥 2 − 𝑥 + 1 ⟹ −𝑚 1 1
𝑚2 . 2 = + 2 ⟹ − 𝑚′ = 𝑚 + 1⟹𝑚′ + 𝑚 = −1
𝑦 ′ + 𝑃(𝑥)𝑦 = 𝑄(𝑥)𝑦 2 + 𝑅(𝑥) 𝑚 𝑚 𝑚
𝑃 𝑥 = − 1 − 2𝑥 , 𝑄 𝑥 = 1, 𝑅 𝑥 = 𝑥 2 − 𝑥 + 1 𝑚′ + 𝑚 = −1
It is a Riccati diff. equation. The last equation is a 1st order linear differential
1 equation and it can be separated into variables.
𝑦 = + 𝑦𝑃 ⟹ Here, 𝑚 is new dependent variable.
𝑚 𝑑𝑚 𝑑𝑚
= − 𝑚 + 1 ⟹න = − න 𝑑𝑥 ⟹
𝑑𝑥 𝑚+1
Calculation of 𝑦𝑃 :
𝑦𝑃 must satisfy the main differential equation. The 𝑙𝑛 𝑚 + 1 = −𝑥 + 𝑙𝑛𝑐⟹m=c𝑒 −𝑥 − 1⟹
most suitable function for this is 𝑦𝑃 = 𝑥. 1
𝑦 = +𝑥⟹
𝑦𝑃 = 𝑥 ⟹ 𝑦𝑃 ′ = 1 ⟹ 𝑦 ′ = 𝑦 2 + 1 − 2𝑥 𝑦 + 𝑥 2 − 𝑥 + 1 𝑚
𝟏
1 = 𝑥 2 + 1 − 2𝑥 𝑥 + 𝑥 2 − 𝑥 + 1 ⟹ 0 = 0 𝒚 = −𝒙 +𝒙
c𝒆 − 𝟏
1 −𝑚 ′
𝑦 = + 𝑥 ⟹ 𝑦′ = 2 + 1
𝑚 𝑚
𝑦 ′ = 𝑦 2 + 1 − 2𝑥 𝑦 + 𝑥 2 − 𝑥 + 1 ⟹
 2
Example: Solve the given differantial equation. −𝑚′ 𝑥
1 1
2 +𝑒 − + 𝑒𝑥 +2 + 𝑒 𝑥 𝑒 𝑥 − 𝑒 2𝑥 − 𝑒 𝑥 =0⟹
𝑦 ′ − 𝑦 2 + 2𝑦𝑒 𝑥 − 𝑒 2𝑥 −𝑒 𝑥 = 0 𝑚 𝑚 𝑚

−𝑚 ′ 1
𝑦′ +𝑃 𝑥 𝑦 = 𝑄 𝑥 𝑦2 +𝑅 𝑥 (It is a Riccati diff. equation) 𝑚2 . 2 = 2 ⟹ 𝑚′ = −1
𝑚 𝑚
𝑦 ′ + 2𝑦𝑒 𝑥 = 𝑦 2 + 𝑒 2𝑥 +𝑒 𝑥 ⟹ ⟹ 𝑚′ + 1 = 0
The last equation is a 1st order differential equation
𝑃 𝑥 = 2𝑒 𝑥 , 𝑄 𝑥 = 1, 𝑅 𝑥 = 𝑒 2𝑥 +𝑒 𝑥 and it can be separated into variables.
1 𝑑𝑚
𝑦= + 𝑦𝑃 ⟹ Here, 𝑚 is new dependent variable. = −1⟹ න 𝑑𝑚 = − න 𝑑𝑥 ⟹
𝑚 𝑑𝑥
Calculation of 𝑦𝑃 :
𝑚 = −𝑥 + 𝑐 ⟹
𝑦𝑃 must satisfy the main differential equation. The
most suitable function for this is 𝑦𝑃 = 𝑒 𝑥 . 1
𝑦 = + 𝑒 𝑥⟹
𝑚
𝑦𝑃 = 𝑒 𝑥 ⟹ 𝑦𝑃 ′ = 𝑒 𝑥 ⟹ 𝑦 ′ − 𝑦 2 + 2𝑦𝑒 𝑥 − 𝑒 2𝑥 −𝑒 𝑥 = 0 𝟏
𝒚= + 𝒆𝒙
𝑒 𝑥 − 𝑒 2𝑥 + 2𝑒 𝑥 𝑒 𝑥 − 𝑒 2𝑥 − 𝑒 𝑥 = 0 ⟹ 0 = 0 c−𝒙
1 −𝑚 ′
𝑦 = + 𝑒 𝑥 ⟹ 𝑦′ = 2 + 𝑒 𝑥
𝑚 𝑚
𝑦 ′ − 𝑦 2 + 2𝑦𝑒 𝑥 − 𝑒 2𝑥 −𝑒 𝑥 = 0 ⟹
 𝑦 4 Solve the given
Example: 𝑦′ + = 𝑦2 − 2
𝑥 𝑥 differantial equation. 𝑦𝑃 4
𝑦𝑃′ + = 𝑦𝑃 2 − 2 ⟹
𝑦 4 𝑥 𝑥
𝑦′ + = 𝑦 2 − 2 ⟹ 𝑦 ′ + 𝑃(𝑥)𝑦 = 𝑄(𝑥)𝑦 2 + 𝑅(𝑥)
𝑥 𝑥 𝐴
1 4 2
Riccati −2𝐴 𝑥 2 𝐴 4 𝐴 𝐴2 4
𝑃 𝑥 = , 𝑄 𝑥 = 1, 𝑅 𝑥 = 2 diff. equ. + = − 2⟹− 3= 4− 2
𝑥 𝑥 𝑥3 𝑥 𝑥2 𝑥 𝑥 𝑥 𝑥
1 1
𝑦 = + 𝑦𝑃 ⟹ Here, 𝑚 is new dependent variable. There is no on the left side, so 𝐴 must be zero. Also,
𝑚 𝑥4
𝐴
Calculation of 𝑦𝑃 : 𝑦𝑃 = 0. Therefore, 𝑦𝑃 does not equal .
𝑥2
4 𝐴 𝐴
𝑅 𝑥 = 2 ⟹ 𝑦𝑃 = 2 We will try for 𝑦𝑃 =
𝑥 𝑥 𝑥
𝑑𝑅 8 𝐴 𝐴 ′
𝐴
= − 3 ⟹ 𝑦𝑃 = 3 it can be tried one by 𝑦𝑃 = ⟹ 𝑦𝑃 = − 2
𝑑𝑥 𝑥 𝑥 one 𝑥 𝑥
4 𝐴 𝑦𝑃 4
න 𝑅 𝑥 𝑑𝑥 = − ⟹ 𝑦𝑃 = 𝑦𝑃′ + = 𝑦𝑃 2 − 2 ⟹
𝑥 𝑥 𝑥 𝑥
From Undetermined Coefficients Method; 𝐴 2
𝐴 𝑥 𝐴 4 𝐴2 4
𝐴 − 2+ = − 2⟹0= 2− 2⟹𝐴=2
𝑦𝑃 = is a proposal equation. 𝑥 𝑥 𝑥 𝑥 𝑥 𝑥
𝑥2
𝟐
𝐴 ′
−2𝐴 𝒚𝑷 =
𝑦𝑃 = ⟹ 𝑦𝑃 = 𝒙
𝑥2 𝑥3
 1 2 Nonhomogeneous solution (𝑚𝑃 ):
𝑦= + ⟹ Here, 𝑚 is new dependent variable.
𝑚 𝑥 (VPM)
−𝑚′ 2 𝑐 𝑐 𝑥 𝑑𝑚𝑃 1 𝑑𝑐 3𝑐
𝑦′ = 2 − 2 𝑚𝐻 = ⟹ 𝑚 𝑃 = ⟹ = −
𝑚 𝑥 𝑥3 𝑥3 𝑑𝑥 𝑥 3 𝑑𝑥 𝑥 4
1 2 2
𝑦 4 −𝑚 ′ 2 + 1 2 4
′ 2
𝑦 + =𝑦 − 2⟹ 2 − 2+ 𝑚 𝑥 = + − 2 3𝑚𝑃 1 𝑑𝑐 3𝑐 3𝑐
𝑥 𝑥 𝑚 𝑥 𝑥 𝑚 𝑥 𝑥 𝑚𝑃′ + = −1 ⟹ 3 − + = −1
𝑥 𝑥 𝑑𝑥 𝑥 4 𝑥 4
−𝑚′ 2 1 2 1 4 4 4 1 𝑑𝑐 3
𝒙𝟒
− + + = + + − = −1 ⟹ න 𝑑𝑐 = − න 𝑥 𝑑𝑥 ⟹ 𝒄 = −
𝑚2 𝑥 2 𝑚𝑥 𝑥 2 𝑚2 𝑚𝑥 𝑥 2 𝑥 2 𝑥 3 𝑑𝑥 𝟒
−𝑚′ 1 1 4 𝑚 4𝑚 𝑥4
2
𝑚 . + = + ′
⟹ −𝑚 + = 1 + ⟹ − 𝒙
𝑚2 𝑚𝑥 𝑚2 𝑚𝑥 𝑥 𝑥 𝑚𝑃 = 34 ⟹ 𝒎𝑷 = −
𝑥 𝟒
3𝑚 𝒄 𝒙
𝑚′ + = −1 𝑚 = 𝑚𝐻 + 𝑚𝑃 ⟹ 𝒎 = 𝟑 −
𝑥 𝒙 𝟒
𝑚𝐻 𝑚𝑃 1 1 2
𝑦 = + 𝑦𝑃 ⟹ 𝑦 = 𝑐 𝑥+𝑥 ⟹
Homogeneous solution (𝑚𝐻 ): 𝑚 −
3𝑚 𝑑𝑚 3𝑚 𝑥3 4
′
𝑚 + =0⟹ =− ⟹
𝑥 𝑑𝑥 𝑥 𝟒𝒙𝟑 𝟐
𝑑𝑚 1 𝒄 𝒚= +
න 𝒎 =
= −3 න 𝑑𝑥 ⟹ 𝑙𝑛 𝑚 = −3𝑙𝑛 𝑥 + 𝑙𝑛𝑐 ⟹ 𝑯 𝒙𝟑 𝟒𝒄 − 𝒙𝟒 𝒙
𝑚 𝑥
Example: Solve the given differantial equation.
𝑦 1
′ 3
𝑦 = 𝑥 𝑦−𝑥 + 2 𝑦𝑃 = 𝑥 ⟹ 𝑦𝑃 ′ = 1 ⟹ 𝑦′ + 2𝑥 4 − 𝑦 = 𝑥 3𝑦2 + 𝑥 5 ⟹
𝑥 𝑥
When the parenthesis in the equation is expanded, it 1
will be seen that it is a Riccati equation. 1 + 2𝑥 4 − 𝑥 = 𝑥 3 𝑥 2 + 𝑥 5 ⟹ 1 + 2𝑥 5 − 1 = 𝑥 5 + 𝑥 5 ⟹
𝑥
𝑦
𝑦 ′ = 𝑥 3 𝑦 2 − 2𝑥𝑦 + 𝑥 2 + ⟹ 𝑥5 = 𝑥5 ⟹ 0 = 0
𝑥 ′
𝑦 1 −𝑚
𝑦 ′ = 𝑥 3 𝑦 2 − 2𝑥 4 𝑦 + 𝑥 5 + ⟹ 𝑦 = + 𝑥 ⟹ 𝑦′ = 2 + 1
𝑥 𝑚 𝑚
1 1
𝑦 ′ + 2𝑥 4 − 𝑦 = 𝑥 3𝑦2 + 𝑥 5 ⟹ 𝑦 ′ + 2𝑥 4 − 𝑦 = 𝑥 3𝑦2 + 𝑥 5 ⟹
𝑥 𝑥
2
𝑦′ + 𝑃(𝑥)𝑦 = 𝑄(𝑥)𝑦 2 + 𝑅(𝑥) −𝑚′ 4−
1 1 1
+ 1 + 2𝑥 + 𝑥 = 𝑥3 +𝑥 + 𝑥5⟹
1 𝑚2 𝑥 𝑚 𝑚
𝑃 𝑥 = 2𝑥 4 − , 𝑄 𝑥 = 𝑥 3, 𝑅 𝑥 = 𝑥5 ′ 3
𝑥 −𝑚 𝑥 1 𝑚
1 𝑚2 . = + ⟹ − 𝑚 ′ = 𝑥3 + ⟹
𝑦 = + 𝑦𝑃 ⟹ Here, 𝑚 is new dependent variable. 𝑚2 𝑚2 𝑚𝑥 𝑥
𝑚
𝑚
Calculation of 𝑦𝑃 : −𝑚′ − = 𝑥3
𝑥
If 𝑦𝑃 = 𝑥 is tested, it will be seen that it satisfies the
main differential equation. A differential equation The last equation is a 1st order linear differential
can have more than one particular solution (𝑦𝑃 ). It is equation.
sufficient to use only one of these solutions.
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Exact differential equations and Integrating
factors

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
2.3. Exact Differential Equations
Example: 2𝑥𝑦 + 𝑒 𝑦 𝑑𝑥 + 𝑥 2 + 𝑥𝑒 𝑦 𝑑𝑦 = 0 Let explain the exact dif. eq. over an example:
 This differential equation example cannot be separated into variables. Since there are two variables (𝑥 and 𝑦)
inside both parentheses, they can be solved by transforming variables. But, after transformation, the new eq.
cannot be seperated into variables.

 This differential equation is a four-term equation and is not a linear differential equation due to functions 𝑒 𝑦 and
𝑥 2 . It can not be linearized by transformations.

 This equation is a non-linear differential equation, and it can be solved by the exact differential equation
approach. In order to be solved with the exact differential equation approach, the functions in the parentheses
must be similar as in the example.

 If this equation could be solved, a solution like y = f(𝑥, 𝑐) would be obtained. If the constant 𝑐 is taken from this
solution, 𝑐 = 𝑐 𝑥, 𝑦 = 𝑓(𝑥, 𝑦) would be obtained. In this new equation, if the constant 𝑐 is accepted as a
variable, it is transformed from ordinary differential equation with the two variables to a partial differential
structure with three variables.

 Since the number of independent variables is more than 1 in partial structures, it is preferred to take a differential
instead of a derivative.
The first and second differentials of 𝑐 = 𝑐 𝑥, 𝑦 are as follows:

𝜕𝑐 𝜕𝑐
𝑑𝑐 = 𝑑𝑥 + 𝑑𝑦 𝟏. 𝐝𝐢𝐟𝐟. (A part of 𝑐 depends on 𝑥 and a part on 𝑦)
𝜕𝑥 𝜕𝑦
𝜕 𝜕𝑐 𝜕𝑐 𝜕 𝜕𝑐 𝜕𝑐
𝑑𝑐 2 = 𝑑𝑥 + 𝑑𝑦 𝑑𝑥 + 𝑑𝑥 + 𝑑𝑦 𝑑𝑦 ⟹ 𝟏
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑦

𝜕 2𝑐 𝜕 2𝑐 𝜕 2𝑐
𝑑𝑐 2 = 2 + 2 + 𝟐. 𝐝𝐢𝐟𝐟.
𝜕𝑥 𝜕𝑥𝜕𝑦 𝜕𝑦 2
Here, if c = constant and dc = 0, it is called the exact differential state.
𝜕𝑐 𝜕𝑐
𝑑𝑥 + 𝑑𝑦 = 0 𝟐
𝜕𝑥 𝜕𝑦

If the example equation is generalized,

𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝟑

It is not certain that the equation [3] is an exact differential equation, while the equation [2] is an exact differential. If
equation [2] is equaled to equation [3];
𝜕𝑐 𝜕𝑐
= 𝑀(𝑥, 𝑦) and = 𝑁(𝑥, 𝑦) 𝟒
𝜕𝑥 𝜕𝑦
If [4] equations are derived;
𝜕 2 𝑐 𝜕𝑀 𝜕2𝑐 𝜕𝑀
= =
𝜕𝑥 2 𝜕𝑥 𝜕𝑥𝜕𝑦 𝜕𝑦

𝜕 2 𝑐 𝜕𝑁 𝜕2𝑐 𝜕𝑁
= =
𝜕𝑦 2 𝜕𝑦 𝜕𝑥𝜕𝑦 𝜕𝑥

The red colored derivatives are equal to each other.

𝜕2𝑐 𝜕𝑁 𝜕𝑀 𝜕𝑀 𝜕𝑁
= = ⟹ = 𝟓
𝜕𝑥𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑥

If Equation 𝟑 is an exact differential, it must satisfy the criterion 𝟓 . The equation satisfying the condition [5] can be solved
by the exact differential approach.
2𝑥𝑦 + 𝑒 𝑦 𝑑𝑥 + 𝑥 2 + 𝑥𝑒 𝑦 𝑑𝑦 = 0

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝜕𝑀
𝑀 𝑥, 𝑦 = 2𝑥𝑦 + 𝑒 𝑦 ⟹ = 2𝑥 + 𝑒 𝑦
𝜕𝑦
2 𝑦
𝜕𝑁
𝑁 𝑥, 𝑦 = 𝑥 + 𝑥𝑒 ⟹ = 2𝑥 + 𝑒 𝑦
𝜕𝑥
According to the assumption 𝑐 = 𝑐(𝑥, 𝑦), 𝑥 and 𝑦 are independent, and 𝑐 is dependent variable. In derivative and integral,
𝜕𝑀 𝜕𝑁
independent variables are taken as constant with respect to each other. Briefly, and 𝜕𝑥 are shown as
𝜕𝑦
𝜕𝑀 𝜕𝑁
= 𝑀𝑦 and = 𝑁𝑥
𝜕𝑦 𝜕𝑥
𝑀𝑦 = 𝑁𝑥 = 2𝑥 + 𝑒 𝑦 (Example is an Exact Diff. Equ.)

From equation [4],

𝜕𝑐
= 𝑀 𝑥, 𝑦 = 2𝑥𝑦 + 𝑒 𝑦
𝜕𝑥 𝑐 must be found by integrating over both 𝑥 and 𝑦.
𝜕𝑐
= 𝑁 𝑥, 𝑦 = 𝑥 2 + 𝑥𝑒 𝑦
𝜕𝑦
𝑐 can be found from these two equation as can be seen:
𝜕𝑐
= 2𝑥𝑦 + 𝑒 𝑦 ⟹ න 𝜕𝑐 = න 2𝑥𝑦 + 𝑒 𝑦 𝜕𝑥 ⟹ න 𝜕𝑐 = 2𝑦 න 𝑥𝜕𝑥 + 𝑒 𝑦 න 𝜕𝑥 ⟹
𝜕𝑥
𝑐 = 𝑦𝑥 2 + 𝑥𝑒 𝑦 + 𝑐1 (𝑦)
The integral was taken on 𝒙, so the integral constant may depend on 𝒚. 𝑐1 (𝑦) can be calculated from 𝑁(𝑥, 𝑦).
𝜕𝑐 𝜕𝑐1
𝑐= 𝑦𝑥 2 + 𝑥𝑒 𝑦 + 𝑐1 𝑦 ⟹ 2 𝑦
= 𝑥 + 𝑥𝑒 + = 𝑁 𝑥, 𝑦 = 𝑥 2 + 𝑥𝑒 𝑦 ⟹
𝜕𝑦 𝜕𝑦
𝜕𝑐1
= 0 ⟹ 𝑐1 = a constant number
𝜕𝑦
Since there is 𝑐 constant in the equation, 𝑐1 may not be written in the equation. As a result, the solution is as below:
𝒄 = 𝒚𝒙𝟐 + 𝒙𝒆𝒚

Or, the solution can be obtained as below:

𝜕𝑐
= 𝑥 2 + 𝑥𝑒 𝑦 ⟹ න 𝜕𝑐 = න 𝑥 2 + 𝑥𝑒 𝑦 𝜕𝑦 ⟹ න 𝜕𝑐 = 𝑥 2 න 𝜕𝑦 + 𝑥 න 𝑒 𝑦 𝜕𝑦 ⟹
𝜕𝑦
𝑐 = 𝑦𝑥 2 + 𝑥𝑒 𝑦 + 𝑐1 (𝑥)
The integral was taken on 𝒚, so the integral constant may depend on 𝒙. 𝑐1 (𝑥) can be calculated from 𝑀(𝑥, 𝑦).
𝜕𝑐 𝜕𝑐1
𝑐 = 𝑦𝑥 2 + 𝑥𝑒 𝑦 + 𝑐1 𝑥 ⟹ = 2𝑥𝑦 + 𝑒 𝑦 + = 𝑀 𝑥, 𝑦 = 2𝑥𝑦 + 𝑒 𝑦 ⟹
𝜕𝑥 𝜕𝑥
𝜕𝑐1
= 0 ⟹ 𝑐1 = a constant number
𝜕𝑥
Since there is 𝑐 constant in the equation, 𝑐1 may not be written in the equation. As a result, the solution is as below:
𝒄 = 𝒚𝒙𝟐 + 𝒙𝒆𝒚
The solution found must satisfy the differential equation.

𝑑𝑐 = 𝑑(𝑦𝑥 2 + 𝑥𝑒 𝑦 ) ⟹ Since 𝑐 is constant 0 = 𝑑(𝑦𝑥 2 ) + 𝑑(𝑥𝑒 𝑦 ) ⟹

0 = 𝑥 2 𝑑𝑦 + 2𝑥𝑦𝑑𝑥 + 𝑒 𝑦 𝑑𝑥 + 𝑥𝑒 𝑦 𝑑𝑦 ⟹

0 = (2𝑥𝑦 + 𝑒 𝑦 )𝑑𝑥 + (𝑥 2 + 𝑥𝑒 𝑦 )𝑑𝑦 ⟹ The solution is true.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
Example: 𝑥 − 𝑦𝐶𝑜𝑠 𝑥 𝑑𝑥 − 𝑆𝑖𝑛 𝑥 𝑑𝑦 = 0 Solve the given 𝑑𝑐1
= 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝑁(𝑥, 𝑦) differantial equation. 𝑑𝑦
𝑀(𝑥, 𝑦)
𝒙𝟐
𝑀(𝑥, 𝑦)𝑑𝑥 − 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝒄= − 𝒚𝑺𝒊𝒏 𝒙
𝟐
𝜕𝑐 𝜕𝑀 2. Method:
𝑀 𝑥, 𝑦 = = 𝑥 − 𝑦𝐶𝑜𝑠 𝑥 ⟹ = −𝐶𝑜𝑠 𝑥
𝜕𝑥 𝜕𝑦
If 𝑦 is dependent, the equation is linear.
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = −𝑆𝑖𝑛 𝑥 ⟹ = −𝐶𝑜𝑠 𝑥
𝜕𝑦 𝜕𝑥 1
𝑥 − 𝑦𝐶𝑜𝑠 𝑥 𝑑𝑥 − 𝑆𝑖𝑛 𝑥 𝑑𝑦 = 0
𝜕𝑀 𝜕𝑁 It is an exact differential equation.
𝑑𝑥
=
𝜕𝑦 𝜕𝑥 𝑑𝑦
𝜕𝑐 𝜕𝑐 𝑥 − 𝑦𝐶𝑜𝑠 𝑥 − 𝑆𝑖𝑛 𝑥 =0
𝑀 𝑥, 𝑦 = = 𝑥 − 𝑦𝐶𝑜𝑠 𝑥 ⟹ = 𝑥 − 𝑦𝐶𝑜𝑠 𝑥 ⟹ 𝑑𝑥
𝜕𝑥 𝜕𝑥
𝑑𝑦
𝑆𝑖𝑛 𝑥 + 𝑦𝐶𝑜𝑠 𝑥 = 𝑥
𝒙𝟐 𝑑𝑥
න𝜕𝑐 = න 𝑥 − 𝑦𝐶𝑜𝑠 𝑥 𝜕𝑥 ⟹ 𝒄 = − 𝒚𝑺𝒊𝒏 𝒙 + 𝒄𝟏 (𝒚)
𝟐
𝑦𝐻 𝑦𝑃
𝜕𝑐 𝑑𝑐1 𝑑𝑐1
= −𝑆𝑖𝑛 𝑥 + = 𝑁 𝑥, 𝑦 ⟹ −𝑆𝑖𝑛 𝑥 + = −𝑆𝑖𝑛 𝑥 ⟹
𝜕𝑦 𝑑𝑦 𝑑𝑦
Homogeneous solution (yH):
𝑑𝑦 𝑑𝑦 𝐶𝑜𝑠 𝑥 𝑑𝑐 𝑐. 𝐶𝑜𝑠(𝑥) 𝑐. 𝐶𝑜𝑠 𝑥
𝑆𝑖𝑛 𝑥 + 𝑦𝐶𝑜𝑠 𝑥 = 0 ⟹ න = න− 𝑑𝑥 − + =𝑥⟹
𝑑𝑥 𝑦 𝑆𝑖𝑛 𝑥 𝑑𝑥 𝑆𝑖𝑛 𝑥 𝑆𝑖𝑛 𝑥
𝑑𝑦 𝑑𝑢
𝑆𝑖𝑛 𝑥 = 𝑢 ⟹ 𝐶𝑜𝑠 𝑥 𝑑𝑥 = 𝑑𝑢 ⟹ න = න− 𝑑𝑐 𝑥2
𝑦 𝑢 = 𝑥 ⟹ න 𝑑𝑐 = න 𝑥𝑑𝑥 ⟹ 𝑐 𝑥 =
1 𝑐 𝑑𝑥 2
𝑙𝑛𝑦 = −𝑙𝑛𝑢 + 𝑙𝑛𝑐 ⟹ 𝑙𝑛𝑦 = 𝑙𝑛 + 𝑙𝑛𝑐 ⟹ 𝑦 =
𝑢 𝑢 𝑥2 1 𝑥2
𝑐 𝑦𝑃 = . ⟹ 𝒚𝑷 =
𝑦𝐻 = 2 𝑆𝑖𝑛 𝑥 2𝑆𝑖𝑛 𝑥
𝑆𝑖𝑛 𝑥
Nonhomogeneous solution (𝑦𝑃 ): 𝒄 𝒙𝟐
(VPM)
𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹ 𝒚 = +
𝑺𝒊𝒏 𝒙 𝟐𝑺𝒊𝒏 𝒙
𝑐 𝑐(𝑥)
𝑦𝐻 = ⟹ 𝑦𝑃 =
𝑆𝑖𝑛 𝑥 𝑆𝑖𝑛 𝑥
𝑑𝑦𝑃 1 𝑑𝑐 𝑐. 𝐶𝑜𝑠(𝑥)
= −
𝑑𝑥 𝑆𝑖𝑛 𝑥 𝑑𝑥 𝑆𝑖𝑛 𝑥 2
𝑑𝑦𝑃
𝑆𝑖𝑛 𝑥 + 𝑦𝑃 𝐶𝑜𝑠 𝑥 = 𝑥
𝑑𝑥
1 𝑑𝑐 𝑐. 𝐶𝑜𝑠(𝑥) 𝑐
𝑆𝑖𝑛 𝑥 − + 𝐶𝑜𝑠 𝑥 = 𝑥 ⟹
𝑆𝑖𝑛 𝑥 𝑑𝑥 𝑆𝑖𝑛 𝑥 2 𝑆𝑖𝑛 𝑥
Example: Solve the given differantial equation.
𝑥𝐶𝑜𝑠 𝑥 + 𝑦 + 𝑆𝑖𝑛 𝑥 + 𝑦 𝑑𝑥 + 𝑥𝐶𝑜𝑠 𝑥 + 𝑦 𝑑𝑦 = 0 ⟹ 𝒄 = 𝒙𝑺𝒊𝒏 𝒙 + 𝒚 + 𝒄𝟏 (𝒙)

𝜕𝑐 𝑑𝑐1
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) = 𝑆𝑖𝑛 𝑥 + 𝑦 + 𝐶𝑜𝑠 𝑥 + 𝑦 + = 𝑀 𝑥, 𝑦 ⟹
𝜕𝑥 𝑑𝑥
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝑑𝑐1
𝑆𝑖𝑛 𝑥 + 𝑦 + 𝑥𝐶𝑜𝑠 𝑥 + 𝑦 + = 𝑥𝐶𝑜𝑠 𝑥 + 𝑦 + 𝑆𝑖𝑛 𝑥 + 𝑦
𝑑𝑥
𝜕𝑐
𝑀 𝑥, 𝑦 = = 𝑥𝐶𝑜𝑠 𝑥 + 𝑦 + 𝑆𝑖𝑛 𝑥 + 𝑦 ⟹ 𝑑𝑐1
𝜕𝑥 = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝜕𝑀 𝑑𝑥
= −𝑥𝑆𝑖𝑛 𝑥 + 𝑦 + 𝐶𝑜𝑠(𝑥 + 𝑦) 𝒄 = 𝒙𝑺𝒊𝒏 𝒙 + 𝒚
𝜕𝑦
𝜕𝑐
𝑁 𝑥, 𝑦 = = 𝑥𝐶𝑜𝑠 𝑥 + 𝑦 ⟹
𝜕𝑦
𝜕𝑁
= −𝑥𝑆𝑖𝑛 𝑥 + 𝑦 + 𝐶𝑜𝑠 𝑥 + 𝑦
𝜕𝑥
𝜕𝑀 𝜕𝑁 It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝜕𝑐
𝑁 𝑥, 𝑦 = = 𝑥𝐶𝑜𝑠 𝑥 + 𝑦 ⟹ න𝜕𝑐 = 𝑥 න𝐶𝑜𝑠 𝑥 + 𝑦 𝜕𝑦 ⟹
𝜕𝑦
2. Method:

𝑥 + 𝑦 = 𝑚 ⟹ 𝑦 = 𝑚 − 𝑥 ⟹ 𝑑𝑦 = 𝑑𝑚 − 𝑑𝑥

𝑥𝐶𝑜𝑠 𝑚 + 𝑆𝑖𝑛 𝑚 𝑑𝑥 + 𝑥𝐶𝑜𝑠 𝑚 (𝑑𝑚 − 𝑑𝑥) = 0

𝑥𝐶𝑜𝑠 𝑚 𝑑𝑥 + 𝑆𝑖𝑛 𝑚 𝑑𝑥 + 𝑥𝐶𝑜𝑠 𝑚 𝑑𝑚 − 𝑥𝐶𝑜𝑠 𝑚 𝑑𝑥 = 0

𝑆𝑖𝑛 𝑚 𝑑𝑥 + 𝑥𝐶𝑜𝑠 𝑚 𝑑𝑚 = 0
It can be separated into variables.

𝑑𝑥 𝐶𝑜𝑠 𝑚
න = −න 𝑑𝑚 ⟹ 𝑢 = 𝑆𝑖𝑛 𝑚 ⟹ 𝑑𝑢 = 𝐶𝑜𝑠 𝑚 𝑑𝑚
𝑥 𝑆𝑖𝑛 𝑚

𝑑𝑥 𝑑𝑢 𝑐 𝑐
න = −න ⟹ 𝑙𝑛 𝑥 = −𝑙𝑛 𝑢 + 𝑙𝑛𝑐 ⟹ 𝑥 = = ⟹
𝑥 𝑢 𝑢 𝑆𝑖𝑛(𝑚)
𝒄
𝒙=
𝑺𝒊𝒏(𝒙 + 𝒚)
Example: 3−𝑦 𝑦 2 − 2𝑥 Solve the given
𝑑𝑥 − 𝑑𝑦 = 0
𝑥2 𝑥𝑦 2 differantial equation.
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) 𝑑𝑐1 3 3
= 2 ⟹ න 𝑑𝑐1 = න 2 𝑑𝑥
𝑑𝑥 𝑥 𝑥
𝑀(𝑥, 𝑦)𝑑𝑥 − 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝟑
𝜕𝑐 3 𝑦 𝜕𝑀 1 𝒄𝟏 = −
𝒙
𝑀 𝑥, 𝑦 = = 2− 2⟹ =− 2
𝜕𝑥 𝑥 𝑥 𝜕𝑦 𝑥 𝒚 𝟐 𝟑
𝒄= + −
𝜕𝑐 1 2 𝜕𝑁 1 𝒙 𝒚 𝒙
𝑁 𝑥, 𝑦 = = − 2⟹ =− 2
𝜕𝑦 𝑥 𝑦 𝜕𝑥 𝑥
𝜕𝑀 𝜕𝑁 It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝜕𝑐 1 2 1 2
𝑁 𝑥, 𝑦 = = − 2 ⟹ න 𝜕𝑐 = න − 𝜕𝑦 ⟹
𝜕𝑦 𝑥 𝑦 𝑥 𝑦2
𝒚 𝟐
𝒄 = + + 𝒄𝟏 (𝒙)
𝒙 𝒚
𝜕𝑐 𝑦 𝑑𝑐1 𝑦 𝑑𝑐1 3 𝑦
=− 2+ = 𝑀 𝑥, 𝑦 ⟹ − 2 + = 2− 2⟹
𝜕𝑥 𝑥 𝑑𝑥 𝑥 𝑑𝑥 𝑥 𝑥
Example: Solve the given differantial equation.
𝑦𝑒 𝑥𝑦 + 4𝑦 3 𝑑𝑥 − 𝑥𝑒 𝑥𝑦 + 12𝑥𝑦 2 − 2𝑦 𝑑𝑦 = 0
𝑑𝑐1
= −2𝑦 ⟹ 𝑐1 = −𝑦 2
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) 𝑑𝑦

𝑀(𝑥, 𝑦)𝑑𝑥 − 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝒄 = 𝒆𝒙𝒚 + 𝟒𝒙𝒚𝟑 − 𝒚𝟐

𝜕𝑐 𝜕𝑀
𝑀 𝑥, 𝑦 = = 𝑦𝑒 𝑥𝑦 + 4𝑦 3 ⟹ = 𝑒 𝑥𝑦 + 𝑥𝑦𝑒 𝑥𝑦 + 12𝑦 2
𝜕𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑥𝑒 𝑥𝑦 + 12𝑥𝑦 2 − 2𝑦 ⟹ = 𝑒 𝑥𝑦 + 𝑥𝑦𝑒 𝑥𝑦 + 12𝑦 2
𝜕𝑦 𝜕𝑥
𝜕𝑀 𝜕𝑁 It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝜕𝑐
𝑀 𝑥, 𝑦 = = 𝑦𝑒 𝑥𝑦 + 4𝑦 3 ⟹ න 𝜕𝑐 = න 𝑦𝑒 𝑥𝑦 + 4𝑦 3 𝜕𝑥 ⟹
𝜕𝑥
𝒄 = 𝒆𝒙𝒚 + 𝟒𝒙𝒚𝟑 + 𝒄𝟏 (𝒚)

𝜕𝑐 𝑥𝑦 2
𝑑𝑐1 𝑥𝑦 2
𝑑𝑐1
= 𝑥𝑒 + 12𝑥𝑦 + = 𝑁 𝑥, 𝑦 ⟹ 𝑥𝑒 + 12𝑥𝑦 + = 𝑥𝑒 𝑥𝑦 + 12𝑥𝑦 2 − 2𝑦 ⟹
𝜕𝑦 𝑑𝑦 𝑑𝑦
Example: Solve the given differantial equation.

The two partial derivatives are equal, so this equation is exact; therefore, we can begin solving it. First, we need to take the
integrals of M with respect to 𝑥 and N with respect to y:

Notice that we added g (𝑦) and f (𝑥); that is because the integral is only taken with respect to one variable, therefore there
may be some function of 𝑥 or 𝑦, respectively, that was left out. Think of it as adding a C when integrating.

Now, let's look at these two integrals. There is one common term among them (2xy), and each has an odd term out. The odd
term can be thought of as the g (y) or f (x) in the opposite equation:
Example: Solve the given differantial equation.

First, we need to verify that this equation is exact by comparing the partial derivatives of M and N:

The partial derivatives match up, so this is an exact equation. To solve, we need to take the integrals of M dx and N dy:

From this, we can see that 3𝑥𝑦 2 is common to the two integrals, and:
Example: Solve the given differantial equation.

First, we need to verify that this equation is exact. To do that, we need to compare the partial derivatives of M
(with respect to y) and N (with respect to x):

Since the partial derivatives come out to the same thing, this is an exact equation. To solve, we need to take the integrals
of M (with respect to x) and N (with respect to y):

From this, we can see that 2𝑥 2 𝑦 is the common term, and:

 𝑚 𝒄 𝒙𝟒
𝑚′ + = −𝑥 3 𝑚 = 𝑚 𝐻 + 𝑚𝑃 ⟹ 𝒎 = −
𝑥 𝒙 𝟓
𝑚𝐻 𝑚𝑃 1 1
𝑦 = + 𝑦𝑃 ⟹ 𝑦 = +𝑥 ⟹
Homogeneous solution (𝑚𝐻 ): 𝑚 𝒄 𝒙𝟒
𝑚 𝑑𝑚 𝑚 −
𝒙 𝟓
𝑚′ + = 0 ⟹ =− ⟹
𝑥 𝑑𝑥 𝑥 𝟓𝒙
𝑑𝑚 𝑑𝑥 𝒚= +𝒙
න = −න ⟹ 𝑙𝑛 𝑚 = −𝑙𝑛 𝑥 + 𝑙𝑛𝑐 ⟹ 𝟓𝒄 − 𝒙𝟓
𝑚 𝑥
𝒄
𝒎𝑯 =
𝒙
Nonhomogeneous solution (𝑚𝑃 ):
(VPM)
𝑐 𝑐 𝑥 𝑑𝑚𝑃 1 𝑑𝑐 𝑐
𝑚𝐻 = ⟹ 𝑚𝑃 = ⟹ = − 2
𝑥 𝑥 𝑑𝑥 𝑥 𝑑𝑥 𝑥
𝑚𝑃 1 𝑑𝑐 𝑐 𝑐
𝑚𝑃 ′ + = −𝑥 3 ⟹ − 2 + 2 = −𝑥 3
𝑥 𝑥 𝑑𝑥 𝑥 𝑥
1 𝑑𝑐 𝒙𝟓
= −𝑥 3 ⟹ න 𝑑𝑐 = − න 𝑥 4 𝑑𝑥 ⟹ 𝒄 = −
𝑥 𝑑𝑥 5
𝟓
𝑥
− 𝒙𝟒
𝑚𝑃 = 5 ⟹ 𝒎𝑷 = −
𝑥 𝟓
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Exact differential equations and Integrating
factors

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
2.3. Integrating factors
Differential equations in the form 𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0 may not satisfy the exact differential condition.
It can be multiplied by a specific function to transform a non-exact differential equation into an exact differential state. This
multiplier function that transforms the differential equation into an exact differential equation is called the integral factor.
This multiplier is denoted by the most general expression 𝜆(𝑥, 𝑦). The multiplier 𝜆(𝑥, 𝑦) can be calculated as follows with the
condition of exact differential.
𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 Assuming that the differential equation is not an exact differential equation,

𝜆 𝑥, 𝑦 . 𝑀 𝑥, 𝑦 𝑑𝑥 + 𝜆 𝑥, 𝑦 . 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 Now, it has became an exact differential equation.

According to the criteria for an exact differential equation, it must be as:

𝜕 𝜕
𝜆 𝑥, 𝑦 . 𝑀 𝑥, 𝑦 = 𝜆 𝑥, 𝑦 . 𝑁(𝑥, 𝑦)𝑑𝑦
𝜕𝑦 𝜕𝑥
If derivatives are taken:
𝜕𝜆 𝜕𝑀 𝜕𝜆 𝜕𝑁
𝑀+𝜆 = 𝑁+𝜆
𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑥

𝑀𝑦 𝑁𝑥
𝜕𝜆 𝜕𝜆
𝑀− 𝑁 = 𝜆 𝑁𝑥 − 𝑀𝑦
𝜕𝑦 𝜕𝑥
For the solution of this equation, some simplistic assumptions are made.

i) 𝜆 = 𝜆 𝑥 : If 𝜆 is only the function of 𝑥 :

𝜕𝜆 𝜕𝜆 𝑑𝜆
= 0 and = reduces to ordinary differential equation form
𝜕𝑦 𝜕𝑥 𝑑𝑥
𝑑𝜆 𝑑𝜆 𝑀𝑦 − 𝑁𝑥
0. 𝑀 − 𝑁 = 𝜆 𝑁𝑥 − 𝑀𝑦 ⟹ න =න 𝑑𝑥
𝑑𝑥 𝜆 𝑁

From this integration, 𝜆(𝑥) is calculated. To reach a solution of the assumption, it must be
𝑀𝑦 − 𝑁𝑥
= 𝑓(𝑥)
𝑁
ii) 𝜆 = 𝜆 𝑦 : If 𝜆 is only the function of 𝑦 :

𝜕𝜆 𝜕𝜆 𝑑𝜆
= 0 and = reduces to ordinary differential equation form
𝜕𝑥 𝜕𝑦 𝑑𝑦
𝜕𝜆 𝑑𝜆 𝑁𝑥 − 𝑀𝑦
𝑀 − 0. 𝑁 = 𝜆 𝑁𝑥 − 𝑀𝑦 ⟹ න =න 𝑑𝑦
𝜕𝑦 𝜆 𝑀
 From this integration, 𝜆(𝑦) is calculated. To reach a solution of the assumption, it must be
𝑁𝑥 − 𝑀𝑦
= 𝑓(𝑦)
𝑀
iii) Otherwise, 𝜆 = 𝑁(𝑥, 𝑦) and can be calculated by trial and error or by achieving a functional balance between 𝑀 and 𝑁.
Example: 1 + 𝑥𝑦 𝑑𝑥 + 𝑥 2 𝑑𝑦 = 0 Solve the given
differantial equation.
i) iii) 1 + 𝑥𝑦 𝑑𝑥 + 𝑥 2 𝑑𝑦 = 0
𝑚 𝑥𝑑𝑚 − 𝑚𝑑𝑥
𝑥𝑦 = 𝑚 ⟹ 𝑦 = ⟹ 𝑑𝑦 = 𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝑥 𝑥2
2
𝑥𝑑𝑚 − 𝑚𝑑𝑥 𝜕𝑐 𝜕𝑀
1 + 𝑚 𝑑𝑥 + 𝑥 =0 𝑀 𝑥, 𝑦 = = 1 + 𝑥𝑦 ⟹ = 𝑀𝑦 = 𝑥
𝑥2 𝜕𝑥 𝜕𝑦
𝑑𝑥 + 𝑚𝑑𝑥 + 𝑥𝑑𝑚 − 𝑚𝑑𝑥 = 0 ⟹ 𝑑𝑥 + 𝑥𝑑𝑚 = 0 𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑥2 ⟹ = 𝑁𝑥 = 2𝑥
It was separated into its variables. 𝜕𝑦 𝜕𝑥
𝜕𝑀 𝜕𝑁
≠
ii) 𝜕𝑦 𝜕𝑥
If 𝑦 is dependent, the equation is linear.
Now, it is not an exact differential equation but there
𝑑𝑦 is little difference between 𝑀𝑦 and 𝑁𝑥 . It can be
1 + 𝑥𝑦 + 𝑥 2
=0 made into an exact differential equation.
𝑑𝑥
𝑑𝑦 𝑦 1
+ =− 2 1st order linear diff. equation 𝑀𝑦 − 𝑁𝑥 = 𝑥 − 2𝑥 = −𝑥
𝑑𝑥 𝑥 𝑥
solution.
This result simplifies with 𝑁.
𝑦𝐻 𝑦𝑃
𝑀𝑦 − 𝑁𝑥 −𝑥 1
= 2 = − ⟹ it will be 𝜆 = 𝜆(𝑥)
𝑁 𝑥 𝑥
 𝑑𝜆 𝑀𝑦 − 𝑁𝑥 𝑑𝜆 1 𝜕𝑐 𝑑𝑐1 𝑑𝑐1
⟹න =න 𝑑𝑥 ⟹ න = න − 𝑑𝑥 ⟹ =𝑥+ = 𝑁 𝑥, 𝑦 ⟹ 𝑥 + =𝑥⟹
𝜆 𝑁 𝜆 𝑥 𝜕𝑦 𝑑𝑦 𝑑𝑦
𝟏 𝑑𝑐1
𝑙𝑛𝜆 = −𝑙𝑛𝑥 ⟹ 𝝀 = = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝒙 𝑑𝑦
The equation is multiplied by 𝜆.
𝒄 = 𝒍𝒏𝒙 + 𝒙𝒚
1 1
. 1 + 𝑥𝑦 𝑑𝑥 + 𝑥 2 𝑑𝑦 = 0 ⟹ + 𝑦 𝑑𝑥 + 𝑥𝑑𝑦 = 0 ⟹
𝑥 𝑥
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝜕𝑐 1 𝜕𝑀
𝑀 𝑥, 𝑦 = = +𝑦 ⟹ = 𝑀𝑦 = 1
𝜕𝑥 𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = =𝑥⟹ = 𝑁𝑥 = 1
𝜕𝑦 𝜕𝑥
𝜕𝑀 𝜕𝑁 Now, It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝜕𝑐 1 1
𝑀 𝑥, 𝑦 = = + 𝑦 ⟹ න𝜕𝑐 = ඲ + 𝑦 𝜕𝑥 ⟹
𝜕𝑥 𝑥 𝑥
𝒄 = 𝒍𝒏𝒙 + 𝒙𝒚 + 𝒄𝟏 (𝒚)
Example: 3𝑥𝑒 𝑦 + 2𝑦 𝑑𝑥 + 𝑥 2 𝑒 𝑦 + 𝑥 𝑑𝑦 = 0 Solve the given
differantial equation.
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
⟹ 𝑙𝑛𝜆 = 𝑙𝑛𝑥 ⟹ 𝝀 = 𝒙
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0
The equation is multiplied by 𝜆.
𝜕𝑐 𝜕𝑀 𝑥. 3𝑥𝑒 𝑦 + 2𝑦 𝑑𝑥 + 𝑥 2 𝑒 𝑦 + 𝑥 𝑑𝑦 = 0 ⟹
𝑀 𝑥, 𝑦 = = 3𝑥𝑒 𝑦 + 2𝑦 ⟹ = 𝑀𝑦 = 3𝑥𝑒 𝑦 + 2
𝜕𝑥 𝜕𝑦
3𝑥 2 𝑒 𝑦 + 2𝑥𝑦 𝑑𝑥 + 𝑥 3 𝑒 𝑦 + 𝑥 2 𝑑𝑦 = 0 ⟹
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑥 2𝑒𝑦 + 𝑥 ⟹ = 𝑁𝑥 = 2𝑥𝑒 𝑦 + 1
𝜕𝑦 𝜕𝑥 𝑁(𝑥, 𝑦)
𝑀(𝑥, 𝑦)
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not 𝜕𝑐 𝜕𝑀
≠ satisfy the exact differential condition. 𝑀 𝑥, 𝑦 = = 3𝑥 2 𝑦
𝑒 + 2𝑥𝑦 ⟹ = 𝑀 = 3𝑥 2 𝑦
𝑒 + 2𝑥
𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝑦

𝜕𝑐 3𝑒 𝑦 + 𝑥 2 ⟹
𝜕𝑁 2 𝑒 𝑦 + 2𝑥
𝑦
𝑀𝑦 − 𝑁𝑥 = 3𝑥𝑒 + 2 − 2𝑥𝑒 + 1 = 𝑥𝑒 + 1 𝑦 𝑦 𝑁 𝑥, 𝑦 = = 𝑥 = 𝑁𝑥 = 3𝑥
𝜕𝑦 𝜕𝑥
This result simplifies with 𝑁. 𝜕𝑀 𝜕𝑁
= Now, It is an exact differential equation.
𝑀𝑦 − 𝑁𝑥 𝑥𝑒 𝑦 + 1 1 𝜕𝑦 𝜕𝑥
= = ⟹ it will be 𝜆 = 𝜆(𝑥) 𝜕𝑐
𝑁 𝑥 𝑥𝑒 𝑦 + 1 𝑥 𝑁 𝑥, 𝑦 = = 𝑥 3 𝑒 𝑦 + 𝑥 2 ⟹ න𝜕𝑐 = න 𝑥 3 𝑒 𝑦 + 𝑥 2 𝜕𝑦 ⟹
𝜕𝑦
𝑑𝜆 𝑀𝑦 − 𝑁𝑥 𝑑𝜆 1 𝒄 = 𝒙𝟑 𝒆𝒚 + 𝒚𝒙𝟐 + 𝒄𝟏 (𝒙)
න =න 𝑑𝑥 ⟹ න = න 𝑑𝑥 ⟹
𝜆 𝑁 𝜆 𝑥
𝜕𝑐 2 𝑦
𝑑𝑐1 2 𝑦
𝑑𝑐1
= 3𝑥 𝑒 + 2𝑥𝑦 + = 𝑀 𝑥, 𝑦 ⟹ 3𝑥 𝑒 + 2𝑥𝑦 + = 3𝑥 2 𝑒 𝑦 + 2𝑥𝑦 ⟹
𝜕𝑥 𝑑𝑥 𝑑𝑥
𝑑𝑐1
= 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝑑𝑦

𝒄 = 𝒙𝟑 𝒆𝒚 + 𝒚𝒙𝟐
Example: 1 + 2𝑥 2 + 4𝑥𝑦 𝑑𝑥 + 2𝑑𝑦 = 0 Solve the given
differantial equation.
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝟐
⟹ 𝑙𝑛𝜆 = 𝑥 2 ⟹ 𝝀 = 𝒆𝒙
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0
The equation is multiplied by 𝜆.
𝜕𝑐 𝜕𝑀 𝟐
𝑀 𝑥, 𝑦 = = 1 + 2𝑥 2 + 4𝑥𝑦 ⟹ = 𝑀𝑦 = 4𝑥 𝒆𝒙 . 1 + 2𝑥 2 + 4𝑥𝑦 𝑑𝑥 + 2𝑑𝑦 = 0 ⟹
𝜕𝑥 𝜕𝑦 2 2
𝜕𝑐 𝜕𝑁 𝑒 𝑥 1 + 2𝑥 2 + 4𝑥𝑦 𝑑𝑥 + 2𝑒 𝑥 𝑑𝑦 = 0 ⟹
𝑁 𝑥, 𝑦 = =2⟹ = 𝑁𝑥 = 0
𝜕𝑦 𝜕𝑥 𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not 𝜕𝑐 2 𝜕𝑀 2
≠ satisfy the exact differential condition. 𝑀 𝑥, 𝑦 = 𝑥
= 𝑒 1 + 2𝑥 + 4𝑥𝑦 ⟹ 2
= 𝑀𝑦 = 4𝑥𝑒 𝑥
𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑦
𝜕𝑐 2 𝜕𝑁 2
𝑀𝑦 − 𝑁𝑥 = 4𝑥 − 0 = 4𝑥 𝑁 𝑥, 𝑦 = = 2𝑒 𝑥 ⟹ = 𝑁𝑥 = 4𝑥𝑒 𝑥
𝜕𝑦 𝜕𝑥
This result simplifies with 𝑁. 𝜕𝑀 𝜕𝑁 Now, It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝑀𝑦 − 𝑁𝑥 4𝑥
= = 2𝑥 ⟹ it will be 𝜆 = 𝜆(𝑥) 𝜕𝑐 2 2
𝑁 2 𝑁 𝑥, 𝑦 = = 2𝑒 𝑥 ⟹ න𝜕𝑐 = න2𝑒 𝑥 𝜕𝑦 ⟹
𝜕𝑦
𝑑𝜆 𝑀𝑦 − 𝑁𝑥 𝑑𝜆
න =න 𝑑𝑥 ⟹ න = න 2𝑥𝑑𝑥 ⟹ 𝟐
𝒄 = 𝟐𝒚𝒆𝒙 + 𝒄𝟏 (𝒙)
𝜆 𝑁 𝜆
𝜕𝑐 𝑥 2 𝑑𝑐1 𝑥 2 𝑑𝑐1 2
= 4𝑥𝑦𝑒 + = 𝑀 𝑥, 𝑦 ⟹ 4𝑥𝑦𝑒 + = 𝑒 𝑥 1 + 2𝑥 2 + 4𝑥𝑦 ⟹
𝜕𝑥 𝑑𝑥 𝑑𝑥
𝑑𝑐1 2 2 2 2
= 𝑒 𝑥 + 2𝑥 2 𝑒 𝑥 ⟹ න 𝑑𝑐1 = න 𝑒 𝑥 + 2𝑥 2 𝑒 𝑥 𝑑𝑥
𝑑𝑥
2 2
න 𝑑𝑐1 = න 𝑒 𝑥 𝑑𝑥 + 2 න 𝑥 2 𝑒 𝑥 𝑑𝑥 ⟹

𝑢 = 𝑥 ⟹ 𝑑𝑢 = 𝑑𝑥
2 𝑑𝑚
𝑑𝑣 = 𝑥𝑒 𝑥 𝑑𝑥 ⟹ 𝑥 2 = 𝑚 ⟹ 𝑥𝑑𝑥 =
2
න 𝑥. 𝑥𝑒 𝑥 𝑑𝑥 = 𝑢. 𝑣 − න 𝑣𝑑𝑢 2
𝑑𝑚 1 𝑚 1 𝑥2
න 𝑑𝑣 = න 𝑒𝑚 ⟹𝑣= 𝑒 = 𝑒
𝑢 𝑑𝑣 2 2 2

2 1 2 1 2
න 𝑥 2 𝑒 𝑥 𝑑𝑥 = 𝑥𝑒 𝑥 − න 𝑒 𝑥 𝑑𝑥
2 2

2 1 𝑥2 1 𝑥2 𝟐 𝟐 𝟐
න 𝑑𝑐1 = න 𝑒 𝑥 𝑑𝑥 + 2 𝑥𝑒 − න 𝑒 𝑑𝑥 ⟹ 𝒄𝟏 = 𝒙𝒆𝒙 ⟹ 𝒄 = 𝟐𝒚𝒆𝒙 + 𝒙𝒆𝒙
2 2
 Example: 2𝑥𝑑𝑥 = 3𝑦 2 + 𝑥 2 − 𝑦 3 tan(𝑦) 𝑑𝑦 Solve the given
differantial equation.
𝑑𝜆 𝑁𝑥 − 𝑀𝑦 𝑑𝜆
3𝑦 2 + 𝑥 2 − 𝑦 3 𝑡𝑎𝑛𝑦 𝑑𝑦 − 2𝑥𝑑𝑥 = 0 න =න 𝑑𝑦 ⟹ න = න −tan(𝑦)𝑑𝑦 ⟹
𝜆 𝑀 𝜆
𝑁(𝑥, 𝑦) 𝑀(𝑥, 𝑦) 𝑑𝜆 Sin(𝑦)
න = න− 𝑑𝑦 ⟹ 𝑙𝑛𝜆 = 𝑙𝑛 𝐶𝑜𝑠(𝑦) ⟹
𝜆 𝐶𝑜𝑠(𝑦)
𝑁 𝑥, 𝑦 𝑑𝑦 − 𝑀 𝑥, 𝑦 𝑑𝑥 = 0
𝝀 = 𝑪𝒐𝒔(𝒚)
𝜕𝑐 𝜕𝑀
𝑀 𝑥, 𝑦 = = −2𝑥 ⟹ = 𝑀𝑦 = 0 The equation is multiplied by 𝜆.
𝜕𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 3𝑦 2 + 𝑥 2 − 𝑦 3 𝑡𝑎𝑛𝑦 ⟹ = 𝑁𝑥 = 2𝑥𝑡𝑎𝑛(𝑦)
𝜕𝑦 𝜕𝑥 𝐶𝑜𝑠(𝑦). 3𝑦 2 + 𝑥 2 − 𝑦 3 𝑡𝑎𝑛𝑦 𝑑𝑦 − 2𝑥𝑑𝑥 = 0 ⟹
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not
≠ satisfy the exact differential condition.
𝐶𝑜𝑠 𝑦 3𝑦 2 + 𝑥 2 − 𝑦 3 𝑡𝑎𝑛𝑦 𝑑𝑦 − 2𝑥𝐶𝑜𝑠(𝑦)𝑑𝑥 = 0 ⟹
𝜕𝑦 𝜕𝑥

𝑀𝑦 − 𝑁𝑥 = 0 − 2𝑥𝑡𝑎𝑛(𝑦) = −2𝑥𝑡𝑎𝑛(𝑦) 𝑁(𝑥, 𝑦) 𝑀(𝑥, 𝑦)

This result simplifies with 𝑀. 𝜕𝑐 𝜕𝑀

𝑀 𝑥, 𝑦 = = −2𝑥𝐶𝑜𝑠(𝑦) ⟹ = 𝑀𝑦 = 2𝑥𝑆𝑖𝑛(𝑦) ⟹
𝜕𝑥 𝜕𝑦
𝑁𝑥 − 𝑀𝑦 2𝑥𝑡𝑎𝑛(𝑦)
= = −tan(𝑦) ⟹ it will be 𝜆 = 𝜆(𝑦)
𝑀 −2𝑥
 𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝐶𝑜𝑠 𝑦 3𝑦 2 + 𝑥 2 − 𝑦 3 𝑡𝑎𝑛𝑦 ⟹ = 𝑁𝑥 = 2𝑥𝑆𝑖𝑛(𝑦)
𝜕𝑦 𝜕𝑥
𝜕𝑀 𝜕𝑁 Now, It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝜕𝑐
𝑀 𝑥, 𝑦 = = −2𝑥𝐶𝑜𝑠 𝑦 ⟹ න𝜕𝑐 = −2 න𝑥𝐶𝑜𝑠 𝑦 𝜕𝑥 ⟹
𝜕𝑥
𝒄 = −𝒙𝟐 𝑪𝒐𝒔 𝒚 + 𝒄𝟏 (𝒚)

𝜕𝑐 𝑑𝑐1 𝑑𝑐1
= 𝑥 2 𝑆𝑖𝑛 𝑦 + = 𝑁 𝑥, 𝑦 ⟹ 𝑥 2 𝑆𝑖𝑛 𝑦 + = 3𝑦 2 𝐶𝑜𝑠 𝑦 + 𝑥 2 𝑆𝑖𝑛 𝑦 − 𝑦 3 𝑆𝑖𝑛 𝑦 ⟹
𝜕𝑦 𝑑𝑦 𝑑𝑦

න 𝑑𝑐1 = න 3𝑦 2 𝐶𝑜𝑠 𝑦 − 𝑦 3 𝑆𝑖𝑛 𝑦 𝑑𝑦 ⟹ 𝒄𝟏 = 𝒚𝟑 𝑪𝒐𝒔 𝒚 ⟹ 𝒄 = −𝒙𝟐 𝑪𝒐𝒔 𝒚 + 𝒚𝟑 𝑪𝒐𝒔 𝒚

Derivative of 𝑦 3 𝐶𝑜𝑠 𝑦
 Example: 𝑦𝑑𝑥 + 𝑦 + tan(𝑥 + 𝑦) 𝑑𝑦 = 0 Solve the given
differantial equation.
𝜆 ≠ 𝜆(𝑥)
𝑦𝑑𝑥 + 𝑦 + tan(𝑥 + 𝑦) 𝑑𝑦 = 0
𝑑𝜆 𝑁𝑥 − 𝑀𝑦 𝑑𝜆 𝑆𝑒𝑐 2 (𝑥 + 𝑦)−1
න =න 𝑑𝑦 ⟹ න =න 𝑑𝑦
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) 𝜆 𝑀 𝜆 𝑦
𝜆 ≠ 𝜆(𝑦) 𝑥 terms can not be eliminated
𝑀 𝑥, 𝑦 𝑑𝑦 + 𝑁 𝑥, 𝑦 𝑑𝑥 = 0
For the functional equilibrium, the differential equation is
𝜕𝑐 𝜕𝑀 multiplied with Cos(x+y).
𝑀 𝑥, 𝑦 = =𝑦⟹ = 𝑀𝑦 = 1
𝜕𝑥 𝜕𝑦 𝐶𝑜𝑠(𝑥 + 𝑦). 𝑦𝑑𝑥 + 𝑦 + tan(𝑥 + 𝑦) 𝑑𝑦 = 0 ⟹
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑦 + tan(𝑥 + 𝑦) ⟹ = 𝑁𝑥 = 𝑆𝑒𝑐 2 (𝑥 + 𝑦) 𝐶𝑜𝑠 𝑥 + 𝑦 𝑦𝑑𝑥 + 𝐶𝑜𝑠 𝑥 + 𝑦 𝑦 + tan(𝑥 + 𝑦) 𝑑𝑦 = 0 ⟹
𝜕𝑦 𝜕𝑥
𝜕𝑀 𝜕𝑁 𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
Currently, the differential equation does not
≠ satisfy the exact differential condition. 𝜕𝑐 𝜕𝑀
𝜕𝑦 𝜕𝑥 𝑀 𝑥, 𝑦 = = 𝐶𝑜𝑠 𝑥 + 𝑦 𝑦 ⟹ ⟹
𝜕𝑥 𝜕𝑦
𝑀𝑦 − 𝑁𝑥 = 1 − 𝑆𝑒𝑐 2 (𝑥 + 𝑦) 𝑀𝑦 = 𝐶𝑜𝑠 𝑥 + 𝑦 − ySin(x + y)
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝐶𝑜𝑠 𝑥 + 𝑦 𝑦 + tan(𝑥 + 𝑦) ⟹ ⟹
𝑑𝜆 𝑀𝑦 − 𝑁𝑥 𝑑𝜆 1 − 𝑆𝑒𝑐 2 (𝑥 + 𝑦) 𝜕𝑥 𝜕𝑥
න =න 𝑑𝑥 ⟹ න =න 𝑑𝑥 ⟹ 𝑁𝑥 = 𝐶𝑜𝑠 𝑥 + 𝑦 − ySin(x + y)
𝜆 𝑁 𝜆 𝑦 + tan(𝑥 + 𝑦)
𝜕𝑀 𝜕𝑁
𝑦 terms can not be eliminated = Now, It is an exact differential equation.
𝜕𝑦 𝜕𝑥
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Exact differential equations and Integrating
factors-Examples

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
 1 𝑦
Example: 𝑥 3 + 𝐶𝑜𝑠𝑦 + 𝑑𝑦 − 2 − 3𝑦𝑥 2 𝑑𝑥 = 0 Solve the given differantial equation.
𝑥 𝑥
𝑁(𝑥, 𝑦) 𝑀(𝑥, 𝑦) 𝑑𝑐1
⟹ = 𝐶𝑜𝑠𝑦 ⟹ න 𝑑𝑐1 = න 𝐶𝑜𝑠𝑦𝑑𝑦
−𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝑑𝑦
𝒄𝟏 = 𝑺𝒊𝒏 𝒚
𝜕𝑐 𝑦 𝜕𝑀 1
𝑀 𝑥, 𝑦 = = − 2 − 3𝑦𝑥 2 ⟹ = 𝑀𝑦 = − 2 − 3𝑥 2
𝜕𝑥 𝑥 𝜕𝑦 𝑥 𝒚
𝜕𝑐 1 𝜕𝑁 1 𝒄 = 𝒚𝒙𝟑 + + 𝑺𝒊𝒏 𝒚
3 2 𝒙
𝑁 𝑥, 𝑦 = = 𝑥 + 𝐶𝑜𝑠𝑦 + ⟹ = 𝑁𝑥 = 3𝑥 − 2
𝜕𝑦 𝑥 𝜕𝑥 𝑥
𝜕𝑀 𝜕𝑁
= It is an exact differential equation.
𝜕𝑦 𝜕𝑥
𝜕𝑐 𝑦
𝑀 𝑥, 𝑦 = = − 2 − 3𝑦𝑥 2 ⟹
𝜕𝑥 𝑥
𝑦
න𝜕𝑐 = − න 2 − 3𝑦𝑥 2 𝜕𝑥 ⟹
𝑥
𝒚
𝒄 = 𝒚𝒙𝟑 + + 𝒄𝟏 (𝒚)
𝒙
𝜕𝑐 1 𝑑𝑐1 1 𝑑𝑐1 1
= 𝑥3 + + = 𝑁 𝑥, 𝑦 ⟹ 𝑥 3 + + = 𝑥 3 + 𝐶𝑜𝑠𝑦 + ⟹
𝜕𝑦 𝑥 𝑑𝑦 𝑥 𝑑𝑦 𝑥
Example: 4𝑥 3 𝑦 3 − 2𝑥𝑦 𝑑𝑥 + 3𝑥 4 𝑦 2 − 𝑥 2 𝑑𝑦 = 0 Solve the given differantial equation.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝑑𝑐1
⟹ = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 𝑑𝑦
𝜕𝑐 𝜕𝑀 𝒄 = 𝒙𝟒 𝒚𝟑 − 𝒚𝒙𝟐
𝑀 𝑥, 𝑦 = = 4𝑥 3 𝑦 3 − 2𝑥𝑦 ⟹ = 𝑀𝑦 = 12𝑥 3 𝑦 2 − 2𝑥
𝜕𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 3𝑥 4 𝑦 2 − 𝑥 2 ⟹ = 𝑁𝑥 = 12𝑥 3 𝑦 2 − 2𝑥
𝜕𝑦 𝜕𝑥

𝜕𝑀 𝜕𝑁
= It is an exact differential equation.
𝜕𝑦 𝜕𝑥
𝜕𝑐
𝑁 𝑥, 𝑦 = = 3𝑥 4 𝑦 2 − 𝑥 2 ⟹
𝜕𝑦

න𝜕𝑐 = න 3𝑥 4 𝑦 2 − 𝑥 2 𝜕𝑦 ⟹

𝒄 = 𝒙𝟒 𝒚𝟑 − 𝒚𝒙𝟐 + 𝒄𝟏 (𝒙)

𝜕𝑐 𝑑𝑐1 𝑑𝑐1
= 4𝑥 3 𝑦 3 − 2𝑥𝑦 + = 𝑀 𝑥, 𝑦 ⟹ 4𝑥 3 𝑦 3 − 2𝑥𝑦 + = 4𝑥 3 𝑦 3 − 2𝑥𝑦 ⟹
𝜕𝑥 𝑑𝑥 𝑑𝑦
 Example: 𝑥 3 + 3𝑥𝑦 2 𝑑𝑥 + 𝑦 3 + 3𝑥 2 𝑦 𝑑𝑦 = 0 Solve the given differantial equation.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝑑𝑐1
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 ⟹ = 𝑦 3 ⟹ න 𝑑𝑐1 = න 𝑦 3 𝑑𝑦 ⟹
𝑑𝑦
𝜕𝑐 𝜕𝑀
𝑀 𝑥, 𝑦 = = 𝑥 3 + 3𝑥𝑦 2 ⟹ = 𝑀𝑦 = 6𝑥𝑦 𝒚𝟒
𝜕𝑥 𝜕𝑦 𝒄𝟏 =
𝟒
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑦 3 + 3𝑥 2 𝑦 ⟹ = 𝑁𝑥 = 6𝑥𝑦
𝜕𝑦 𝜕𝑥 𝒙 𝟒 𝟑 𝟐 𝟐 𝒚𝟒
𝒄= + 𝒙 𝒚 +
𝟒 𝟐 𝟒
𝜕𝑀 𝜕𝑁
= It is an exact differential equation.
𝜕𝑦 𝜕𝑥 if the result is multiplied by 4, and 𝒄𝟏 = 𝟒𝒄
𝜕𝑐
𝑀 𝑥, 𝑦 = = 𝑥 3 + 3𝑥𝑦 2 ⟹ 𝒄𝟏 = 𝒙𝟒 + 𝟔𝒙𝟐 𝒚𝟐 + 𝒚𝟒
𝜕𝑥

න𝜕𝑐 = න 𝑥 3 + 3𝑥𝑦 2 𝜕𝑥 ⟹

𝒙𝟒 𝟑 𝟐 𝟐
𝒄= + 𝒙 𝒚 + 𝒄𝟏 (𝒚)
𝟒 𝟐
𝜕𝑐 2
𝑑𝑐1 2
𝑑𝑐1
= 3𝑥 𝑦 + = 𝑁 𝑥, 𝑦 ⟹ 3𝑥 𝑦 + = 𝑦 3 + 3𝑥 2 𝑦 ⟹
𝜕𝑦 𝑑𝑦 𝑑𝑦
 Example: 𝑆𝑖𝑛 𝑥 𝐶𝑜𝑠ℎ(𝑦) 𝑑𝑥 − 𝐶𝑜𝑠 𝑥 𝑆𝑖𝑛ℎ(𝑦) 𝑑𝑦 = 0 Solve the given differantial equation.
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝑑𝑐1
𝑀 𝑥, 𝑦 𝑑𝑥 − 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 ⟹ 𝑆𝑖𝑛(𝑥)𝐶𝑜𝑠ℎ(𝑦) + = 𝑆𝑖𝑛(𝑥)𝐶𝑜𝑠ℎ(𝑦)
𝑑𝑥
𝜕𝑐 𝜕𝑀 𝑑𝑐1
𝑀 𝑥, 𝑦 = = 𝑆𝑖𝑛 𝑥 𝐶𝑜𝑠ℎ(𝑦) ⟹ = 𝑀𝑦 = 𝑆𝑖𝑛 𝑥 𝑆𝑖𝑛ℎ 𝑦 ⟹ = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝜕𝑥 𝜕𝑦 𝑑𝑦
𝜕𝑐 𝜕𝑁 𝒄 = −𝑪𝒐𝒔 𝒙 𝑪𝒐𝒔𝒉(𝒚)
𝑁 𝑥, 𝑦 = = − 𝐶𝑜𝑠 𝑥 𝑆𝑖𝑛ℎ(𝑦) ⟹ = 𝑁𝑥 = 𝑆𝑖𝑛 𝑥 𝑆𝑖𝑛ℎ 𝑦
𝜕𝑦 𝜕𝑥
𝜕𝑀 𝜕𝑁
= It is an exact differential equation.
𝜕𝑦 𝜕𝑥
𝜕𝑐
𝑁 𝑥, 𝑦 = = − 𝐶𝑜𝑠 𝑥 𝑆𝑖𝑛ℎ(𝑦) ⟹
𝜕𝑦
න𝜕𝑐 = − න 𝐶𝑜𝑠 𝑥 𝑆𝑖𝑛ℎ(𝑦) 𝜕𝑦 ⟹

𝒄 = −𝑪𝒐𝒔 𝒙 𝑪𝒐𝒔𝒉(𝒚) + 𝒄𝟏 (𝒙)

𝜕𝑐 𝑑𝑐1
= 𝑆𝑖𝑛(𝑥)𝐶𝑜𝑠ℎ(𝑦) + = 𝑀 𝑥, 𝑦 ⟹
𝜕𝑥 𝑑𝑥
Example: 2𝑥𝑦 4 𝑒 𝑦 + 2𝑥𝑦 3 + 𝑦 𝑑𝑥 + 𝑥 2 𝑦 4 𝑒 𝑦 − 𝑥 2 𝑦 2 − 3𝑥 𝑑𝑦 =0 Solve the given differantial equation.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) 𝑁𝑥 − 𝑀𝑦 4
= − ⟹ it will be 𝜆 = 𝜆(𝑦)
𝑀 𝑦
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0 𝑑𝜆 𝑁𝑥 − 𝑀𝑦 𝑑𝜆 4
න =න 𝑑𝑦 ⟹ න = න − 𝑑𝑦 ⟹
𝜕𝑐 𝜆 𝑀 𝜆 𝑦
𝑀 𝑥, 𝑦 = = 2𝑥𝑦 4 𝑒 𝑦 + 2𝑥𝑦 3 + 𝑦 ⟹
𝜕𝑥 𝑑𝜆 4 𝟏
𝜕𝑀 න = න − 𝑑𝑦 ⟹ 𝑙𝑛𝜆 = −4𝑙𝑛 𝑦 ⟹ 𝝀 = 𝟒
= 𝑀𝑦 = 8𝑥𝑦 3 𝑒 𝑦 + 2𝑥𝑦 4 𝑒 𝑦 + 6𝑥𝑦 2 + 1 𝜆 𝑦 𝒚
𝜕𝑦
𝜕𝑐 The equation is multiplied by 𝜆.
𝑁 𝑥, 𝑦 = = 𝑥 2 𝑦 4 𝑒 𝑦 − 𝑥 2 𝑦 2 − 3𝑥 ⟹
𝜕𝑦 𝟏 2𝑥𝑦 4 𝑒 𝑦 + 2𝑥𝑦 3 + 𝑦 𝑑𝑥 +
𝜕𝑁 . ⟹
= 𝑁𝑥 = 2𝑥𝑦 4 𝑒 𝑦 − 2𝑥𝑦 2 − 3 𝒚𝟒 𝑥 2 𝑦 4 𝑒 𝑦 − 𝑥 2 𝑦 2 − 3𝑥 𝑑𝑦 = 0
𝜕𝑥
𝑥 1 𝑥 2 3𝑥
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not ⟹ 2𝑥𝑒 𝑦 2 𝑦
+ 2 + 3 𝑑𝑥 + 𝑥 𝑒 − 2 − 4 𝑑𝑦 = 0
≠ satisfy the exact differential condition. 𝑦 𝑦 𝑦 𝑦
𝜕𝑦 𝜕𝑥
𝑀𝑦 − 𝑁𝑥 = (8𝑥𝑦 3 𝑒 𝑦 + 2𝑥𝑦 4 𝑒 𝑦 + 6𝑥𝑦 2 + 1) − (2𝑥𝑦 4 𝑒 𝑦 − 2𝑥𝑦 2 − 3)
𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝑀𝑦 − 𝑁𝑥 = 8𝑥𝑦 3 𝑒 𝑦 + 8𝑥𝑦 2 +4
𝜕𝑐 𝑦
𝑥 1
This result simplifies with 𝑀. 𝑀 𝑥, 𝑦 = = 2𝑥𝑒 + 2 + 3 ⟹
𝜕𝑥 𝑦 𝑦
𝑁𝑥 − 𝑀𝑦 8𝑥𝑦 3 𝑒 𝑦 + 8𝑥𝑦 2 + 4 4(2𝑥𝑦 3 𝑒 𝑦 + 2𝑥𝑦 2 + 1) 𝜕𝑀 𝑥 1
𝑀
=−
2𝑥𝑦 4 𝑒 𝑦 + 2𝑥𝑦 3 + 𝑦
=−
𝑦(2𝑥𝑦 3 𝑒 𝑦 + 2𝑥𝑦 2 + 1)
⟹ = 𝑀𝑦 = 2𝑥𝑒 𝑦 − 2 2 − 3 4 ⟹
𝜕𝑦 𝑦 𝑦
 𝜕𝑐 𝑦
𝑥 1
𝑀 𝑥, 𝑦 = = 2𝑥𝑒 + 2 + 3 ⟹
𝜕𝑥 𝑦 𝑦
𝜕𝑀 𝑥 1
= 𝑀𝑦 = 2𝑥𝑒 𝑦 − 2 2 − 3 4 ⟹
𝜕𝑦 𝑦 𝑦
𝜕𝑐 2 𝑦
𝑥2 𝑥
𝑁 𝑥, 𝑦 = =𝑥 𝑒 − 2−3 4 ⟹
𝜕𝑦 𝑦 𝑦
𝜕𝑁 𝑦
𝑥 1
= 𝑁𝑥 = 2𝑥𝑒 − 2 2 − 3 4
𝜕𝑥 𝑦 𝑦
𝜕𝑀 𝜕𝑁 Now, It is an exact differential equation.
=
𝜕𝑦 𝜕𝑥
𝜕𝑐 𝑥 1 𝑥 1
𝑀 𝑥, 𝑦 = = 2𝑥𝑒 𝑦 + 2 + 3 ⟹ න𝜕𝑐 = ඲ 2𝑥𝑒 𝑦 + 2 + 3 𝜕𝑥 ⟹
𝜕𝑥 𝑦 𝑦 𝑦 𝑦
𝟐
𝒙 𝒙
𝒄 = 𝒙𝟐 𝒆𝒚 + + 𝟑 + 𝒄𝟏 (𝒚)
𝒚 𝒚

𝜕𝑐 2 𝑦
𝑥2 𝑥 𝑑𝑐1 2 𝑦
𝑥2 𝑥 𝑑𝑐1 2 𝑦
𝑥2 𝑥
=𝑥 𝑒 − 2−3 4+ = 𝑁 𝑥, 𝑦 ⟹ 𝑥 𝑒 − 2 − 3 4 + =𝑥 𝑒 − 2−3 4 ⟹
𝜕𝑦 𝑦 𝑦 𝑑𝑦 𝑦 𝑦 𝑑𝑦 𝑦 𝑦
𝑑𝑐1 𝟐 𝒚
𝒙𝟐 𝒙
⟹ = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 𝒄=𝒙 𝒆 + + 𝟑
𝑑𝑦 𝒚 𝒚
Example: 2y𝑑𝑥 + 3𝑥𝑑𝑦 =0 Solve the given differantial equation.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) The equation is multiplied by 𝜆.

𝟏 1 2
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0 − −
𝒙 𝟑. 2y𝑑𝑥 + 3𝑥𝑑𝑦 = 0 ⟹ 2y𝑥 3 𝑑𝑥 + 3𝑥 3 𝑑𝑦 =0
𝜕𝑐 𝜕𝑀
𝑀 𝑥, 𝑦 = = 2𝑦 ⟹ = 𝑀𝑦 = 2 𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝜕𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁 𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0
𝑁 𝑥, 𝑦 = = 3𝑥 ⟹ = 𝑁𝑥 = 3
𝜕𝑦 𝜕𝑥 𝜕𝑐 1
−3 𝜕𝑀 1
−3
𝑀 𝑥, 𝑦 = = 2𝑦𝑥 ⟹ = 𝑀𝑦 = 2𝑥
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not 𝜕𝑥 𝜕𝑦
≠ satisfy the exact differential condition.
𝜕𝑦 𝜕𝑥 𝜕𝑐 2 𝜕𝑁 1
−3
𝑁 𝑥, 𝑦 = = 3𝑥 ⟹
3 = 𝑁𝑥 = 2𝑥
𝑀𝑦 − 𝑁𝑥 = 3 − 2 = −1 𝜕𝑦 𝜕𝑥
𝜕𝑐 1
−3
1
−3
This result simplifies with 𝑁. 𝑀 𝑥, 𝑦 = = 2y𝑥 ⟹ 𝜕𝑐 = න 2y𝑥 𝜕𝑥 ⟹
𝜕𝑥
𝑀𝑦 − 𝑁𝑥 1 𝟐
=− ⟹ it will be 𝜆 = 𝜆(𝑥)
𝑁 3𝑥 𝒄= 𝟑𝒚𝒙𝟑 + 𝒄𝟏 (𝒚)
𝑑𝜆 𝑀𝑦 − 𝑁𝑥 𝑑𝜆 1 𝜕𝑐 2 𝑑𝑐1 2 𝑑𝑐1 2
න =න 𝑑𝑥 ⟹ න = න − 𝑑𝑥 ⟹ 𝑁 𝑥, 𝑦 = = 3𝑥 3 + ⟹ 3𝑥 3 + = 3𝑥 3
𝜆 𝑁 𝜆 3𝑥 𝜕𝑦 𝑑𝑦 𝑑𝑦
1 −𝟑
𝟏 𝑑𝑐1 𝟐
𝑙𝑛𝜆 = − 𝑙𝑛 𝑥 ⟹ 𝝀 = 𝒙 ⟹ = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 ⟹ 𝒄 = 𝟑𝒚𝒙𝟑
3 𝑑𝑦
Example: 2𝑆𝑖𝑛(𝑦 2 )𝑑𝑥 + 𝑥𝑦𝐶𝑜𝑠(𝑦 2 )𝑑𝑦 =0 Solve the given differantial equation.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) The equation is multiplied by 𝜆.

𝒙𝟑 . 2𝑆𝑖𝑛(𝑦 2 )𝑑𝑥 + 𝑥𝑦𝐶𝑜𝑠(𝑦 2 )𝑑𝑦 = 0 ⟹
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0
2𝑥 3 𝑆𝑖𝑛(𝑦 2 )𝑑𝑥 + 𝑥 4 𝑦𝐶𝑜𝑠(𝑦 2 )𝑑𝑦 = 0 ⟹
𝜕𝑐 2
𝜕𝑀
𝑀 𝑥, 𝑦 = = 2𝑆𝑖𝑛(𝑦 ) ⟹ = 𝑀𝑦 = 4𝑦𝐶𝑜𝑠(𝑦 2 ) 𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝜕𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑥𝑦𝐶𝑜𝑠(𝑦 2 ) ⟹ = 𝑁𝑥 = 𝑦𝐶𝑜𝑠(𝑦 2 ) 𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0
𝜕𝑦 𝜕𝑥
𝜕𝑐 𝜕𝑀
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not 𝑀 𝑥, 𝑦 = = 2𝑥 3 𝑆𝑖𝑛(𝑦 2 ) ⟹ = 𝑀𝑦 = 4y𝑥 3 𝐶𝑜𝑠(𝑦 2 )
≠ 𝜕𝑥 𝜕𝑦
satisfy the exact differential condition.
𝜕𝑦 𝜕𝑥 𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 𝑥 4 𝑦𝐶𝑜𝑠(𝑦 2 ) ⟹ = 𝑁𝑥 = 4y𝑥 3 𝐶𝑜𝑠(𝑦 2 )
𝑀𝑦 − 𝑁𝑥 = 4𝑦𝐶𝑜𝑠(𝑦 2 ) − 𝑦𝐶𝑜𝑠(𝑦 2 ) = 3𝑦𝐶𝑜𝑠(𝑦 2 ) 𝜕𝑦 𝜕𝑥
𝜕𝑐
This result simplifies with 𝑁. 𝑀 𝑥, 𝑦 = = 2𝑥 3 𝑆𝑖𝑛(𝑦 2 ) ⟹ 𝜕𝑐 = න 2𝑥 3 𝑆𝑖𝑛(𝑦 2 ) 𝜕𝑥 ⟹
𝜕𝑥
𝑀𝑦 − 𝑁𝑥 3𝑦𝐶𝑜𝑠(𝑦 2 ) 3 𝒙𝟒
= = ⟹ it will be 𝜆 = 𝜆(𝑥) 𝒄 = 𝑺𝒊𝒏(𝒚𝟐 ) + 𝒄𝟏 (𝒚)
𝑁 𝑥𝑦𝐶𝑜𝑠(𝑦 2 ) 𝑥 𝟐
𝑑𝜆 𝑀𝑦 − 𝑁𝑥 𝑑𝜆 3 𝜕𝑐 𝑑𝑐1
න =න 𝑑𝑥 ⟹ න = න 𝑑𝑥 ⟹ 𝑁 𝑥, 𝑦 = = 𝑦𝑥 4 𝐶𝑜𝑠 𝑦 2 + = 𝑦𝑥 4 𝐶𝑜𝑠(𝑦 2 ) ⟹
𝜆 𝑁 𝜆 𝑥 𝜕𝑦 𝑑𝑦
𝑑𝑐1
𝑙𝑛𝜆 = 3𝑙𝑛 𝑥 ⟹ 𝝀 = 𝒙𝟑 ⟹ = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 ⟹ 𝒄𝟐 = 𝒙𝟒 𝑺𝒊𝒏(𝒚𝟐 )
𝑑𝑦
Example: 2𝑥𝑦𝑑𝑥 + 4𝑦 + 3𝑥 2 𝑑𝑦 =0 Solve the given differantial equation.

𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦) The equation is multiplied by 𝜆.

𝒚𝟐 . 2𝑥𝑦𝑑𝑥 + 4𝑦 + 3𝑥 2 𝑑𝑦 = 0 ⟹
𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0
2𝑥𝑦 3 𝑑𝑥 + 4𝑦 3 + 3𝑥 2 𝑦 2 𝑑𝑦 = 0 ⟹
𝜕𝑐 𝜕𝑀
𝑀 𝑥, 𝑦 = = 2𝑥𝑦 ⟹ = 𝑀𝑦 = 2𝑥 𝑀(𝑥, 𝑦) 𝑁(𝑥, 𝑦)
𝜕𝑥 𝜕𝑦
𝜕𝑐 𝜕𝑁 𝑀 𝑥, 𝑦 𝑑𝑥 + 𝑁 𝑥, 𝑦 𝑑𝑦 = 0
𝑁 𝑥, 𝑦 = = 4𝑦 + 3𝑥 2 ⟹ = 𝑁𝑥 = 6𝑥
𝜕𝑦 𝜕𝑥
𝜕𝑐 𝜕𝑀
𝜕𝑀 𝜕𝑁 Currently, the differential equation does not 𝑀 𝑥, 𝑦 = = 2𝑥𝑦 3 ⟹ = 𝑀𝑦 = 6x𝑦 2
≠ 𝜕𝑥 𝜕𝑦
satisfy the exact differential condition.
𝜕𝑦 𝜕𝑥 𝜕𝑐 𝜕𝑁
𝑁 𝑥, 𝑦 = = 4𝑦 3 + 3𝑥 2 𝑦 2 ⟹ = 𝑁𝑥 = 6x𝑦 2
𝑀𝑦 − 𝑁𝑥 = 2𝑥 − 6𝑥 = −4𝑥 𝜕𝑦 𝜕𝑥
𝜕𝑐
This result simplifies with 𝑀. 𝑁 𝑥, 𝑦 = = 4𝑦 3 + 3𝑥 2 𝑦 2 ⟹ 𝜕𝑐 = න 4𝑦 3 + 3𝑥 2 𝑦 2 𝜕𝑦 ⟹
𝜕𝑦
𝑁𝑥 − 𝑀𝑦 −(−4𝑥) 2
= = ⟹ it will be 𝜆 = 𝜆(𝑦) 𝒄 = 𝒚𝟒 + 𝒙𝟐 𝒚𝟑 + 𝒄𝟏 (𝒙)
𝑀 2𝑥𝑦 𝑦
𝑑𝜆 𝑁𝑥 − 𝑀𝑦 𝑑𝜆 2 𝜕𝑐 𝑑𝑐1
න =න 𝑑𝑦 ⟹ න = න 𝑑𝑦 ⟹ 𝑀 𝑥, 𝑦 = = 2𝑥𝑦 3 + = 2𝑥𝑦 3 ⟹
𝜆 𝑀 𝜆 𝑦 𝜕𝑥 𝑑𝑥
𝑑𝑐1
𝑙𝑛𝜆 = 2𝑙𝑛 𝑦 ⟹ 𝝀 = 𝒚𝟐 ⟹ = 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡 ⟹ 𝒄𝟐 = 𝒚𝟒 + 𝒙 𝟐 𝒚𝟑
𝑑𝑦
 Department of Mechanical Engineering

ME 201 Differential Equations in Engineering

Lecture Notes: Higher order linear differential equations with
constant coefficients

Prof. Dr. Recep EKİCİ

Doç. Dr. Orhan KEKLİKCİOĞLU

References: Advanced Engineering Mathematics-9th - Erwin Kreyszig

SCHAUM’S Outlines Differential Equations-4th - Richard Bronson,
A first course in differential equations-3th, Springer, 2015 - J. David Logan
3.1. Higher order linear differential equations with constant coefficients
A linear differential equation of order 𝑛 is defined to be an equation of the form
𝑑𝑛 𝑦 𝑑𝑛−1 𝑦 𝑑𝑦
𝑎𝑛 𝑥 + 𝑎 𝑛−1 𝑥 + ⋯ + 𝑎1 𝑥 + 𝑎0 𝑥 𝑦 = 𝑓 𝑥 n-th order linear diff. equation
𝑑𝑥 𝑛 𝑑𝑥 𝑛−1 𝑑𝑥

where 𝑎𝑛 , … , 𝑎1 , 𝑎0 are constants and 𝑎𝑛 ≠ 0. 𝑓 𝑥 is continuous real valued function. Higher order differential equations
are divided into two groups according to the structure of the dependent variable or its derivatives. If the power of
dependent variables and derivatives is 1, it is called a linear differential equation; if it is greater than one or in terms of a
special function, it is called a nonlinear differential equation. Analytical solutions of nonlinear equations are only possible
for first order differential equations and are not possible for higher order equations. The analytical solution for higher order
equations exists only for linear equations.
2
𝑑2 𝑦 𝑑𝑦
𝑎1 𝑥 + 𝑎2 𝑥 + 𝑎3 𝑥 𝑦 = 𝑓 𝑥 2. order nonlinear diff. equation
𝑑𝑥 2 𝑑𝑥

𝑑2 𝑦 𝑑𝑦
𝑎1 𝑥 + 𝑎 2 𝑥 + 𝑎3 𝑥 𝐶𝑜𝑠(𝑦) = 𝑓 𝑥 2. order nonlinear diff. equation
𝑑𝑥 2 𝑑𝑥
In practice, the highest order for differential equations is 4, and however, the most used differential equation types are the
2nd order. A 4th order differential equation is as follows:

𝑑4 𝑦 𝑑3 𝑦 𝑑2 𝑦 𝑑𝑦
𝑎4 𝑥 + 𝑎 3 𝑥 + 𝑎 2 𝑥 + 𝑎1 𝑥 + 𝑎0 𝑥 𝑦 = 𝑓 𝑥
𝑑𝑥 4 𝑑𝑥 3 𝑑𝑥 2 𝑑𝑥
𝑦𝐻 𝑦𝑃
The solution of this equation is carried out in two stages:

I. Homogeneous solution(𝒚𝑯 ):
𝑑4 𝑦 𝑑3 𝑦 𝑑2 𝑦 𝑑𝑦
𝑎4 𝑥 + 𝑎 3 𝑥 + 𝑎 2 𝑥 + 𝑎1 𝑥 + 𝑎0 𝑥 𝑦 = 0 ⟹
𝑑𝑥 4 𝑑𝑥 3 𝑑𝑥 2 𝑑𝑥
𝑎4 𝑦 𝚤𝑣 + 𝑎3 𝑦 ′′′ + 𝑎2 𝑦 ′′ + 𝑎1 𝑦 ′ + 𝑎0 𝑦 = 0

In a homogeneous solution, the right-hand member of the equation is zero. The solution of this differential equation can
be found by proposal functions. First, a solution function is proposed, and then, edits are made on the proposed functions
to ensure that the proposed equation provides the differential equation.

𝑦𝐻 = 𝑐1 𝑓1 𝑥 + 𝑐2 𝑓2 𝑥 + 𝑐3 𝑓3 𝑥 + 𝑐4 𝑓4 (𝑥)
The function that will provide the solution of this differential equation must have certain properties. The function itself and
all its derivatives must be functionally similar. However, similar functions cancel each other out and make the left side of the
equation zero. The only function that is equal to itself and its derivatives is 𝒆𝒌𝒙 , where 𝒌 is a real number or an imaginary
constant. This is because the functional structure of the exponential function does not change when it is derived 𝒏 times.

If 𝒚 = 𝒆𝒌𝒙:
𝑦 = 𝑒 𝑘𝑥 ⟹ 𝑦 ′ = 𝑘𝑒 𝑘𝑥 , 𝑦 ′′ = 𝑘 2 𝑒 𝑘𝑥 , 𝑦 ′′′ = 𝑘 3 𝑒 𝑘𝑥 , 𝑦 𝚤𝑣 = 𝑘 4 𝑒 𝑘𝑥
𝑎4 𝑘 4 𝑒 𝑘𝑥 + 𝑎3 𝑘 3 𝑒 𝑘𝑥 + 𝑎2 𝑘 2 𝑒 𝑘𝑥 + 𝑎1 𝑘𝑒 𝑘𝑥 + 𝑎0 𝑒 𝑘𝑥 = 0 ⟹

For the solution,

𝑒 𝑘𝑥 𝑎4 𝑘 4 + 𝑎3 𝑘 3 + 𝑎2 𝑘 2 + 𝑎1 𝑘 + 𝑎0 = 0 ⟹

𝑒 𝑘𝑥 ≠ 0 and 𝑎4 𝑘 4 + 𝑎3 𝑘 3 + 𝑎2 𝑘 2 + 𝑎1 𝑘 + 𝑎0 = 0 (Characteristic equation)

Roots of characteristic equation can have in three states:
i) Distinct Real roots state:
𝑘1 ≠ 𝑘2 ≠ 𝑘3 ≠ 𝑘4 ⟹ 𝑘’s are real numbers.
𝑦𝐻 = 𝑐1 𝑒 𝑘1 𝑥 + 𝑐2 𝑒 𝑘2𝑥 + 𝑐3 𝑒 𝑘3𝑥 + 𝑐4 𝑒 𝑘4𝑥
ii) Repeated roots state:
𝑘1 = 𝑘2 = 𝑘3 = 𝑘4 ⟹ 𝑘’s are real numbers.

One solution has been obtained. This state is called linear independence. In this case, the missing solutions are defined by
Linear Independent Solutions (LIS).
𝑒 𝑘𝑥 Known solution
𝑥. 𝑒 𝑘𝑥 1. LIS
4 unequal solutions have been found
𝑥 2 . 𝑒 𝑘𝑥 2. LIS
𝑥 3 . 𝑒 𝑘𝑥 3. LIS

𝑦𝐻 = 𝑐1 + 𝑐2 𝑥 + 𝑐3 𝑥 2 + 𝑐4 𝑥 3 𝑒 𝑘𝑥

Linear independent solution expresses the situation where the new function obtained by multiplying the first solution by a
linear equation such as (𝒂𝒙 + 𝒃) also satisfies the differential equation. That is, new solutions can be produced by
multiplying the solution found by the linear equation. This multiplication process is repeated consecutively for the desired
amount. If the differential equation requires 𝑛 solutions, it is enough to multiply by the linear equation 𝒏 − 𝟏 times. Since 𝒄
will be the integral constant in a homogeneous solution, Constants 𝑎 and 𝑏 in a linear equation may not be considered.
Therefore, linear independent solutions are produced by taking 𝒙 instead of (𝒂𝒙 + 𝒃).
iii) Complex roots state:
𝑘1 = 𝛼 + 𝛽𝑖, 𝑘2 = 𝛼 − 𝛽𝑖 ⟹ Roots are complex numbers.
𝛼 and 𝛽 are real and imaginary parts, respectively. Keeping in mind that 𝑖 2 = −1.

If these complex roots are wrote in (i) state for the solution of two roots:
𝑦𝐻 = 𝑐1 𝑒 (𝛼+𝛽𝑖)𝑥 + 𝑐2 𝑒 (𝛼−𝛽𝑖)𝑥 = 𝑐1 𝑒 𝛼𝑥 𝑒 𝛽𝑖𝑥 + 𝑐2 𝑒 𝛼𝑥 𝑒 −𝛽𝑖𝑥 ⟹

𝑦𝐻 = 𝑒 𝛼𝑥 𝑐1 𝑒 𝛽𝑖𝑥 + 𝑐2 𝑒 −𝛽𝑖𝑥 ⟹

According to Euler’s Formula:

𝑒 ∓𝛽𝑖𝑥 = cos 𝛽𝑥 ∓ 𝑖. 𝑠𝑖𝑛 𝛽𝑥 ⟹
𝑦𝐻 = 𝑒 𝛼𝑥 𝑐1 𝑒 𝛽𝑖𝑥 + 𝑐2 𝑒 −𝛽𝑖𝑥 = 𝑒 𝛼𝑥 𝑐1 cos 𝛽𝑥 + 𝑖. 𝑠𝑖𝑛 𝛽𝑥 + 𝑐2 cos 𝛽𝑥 − 𝑖. 𝑠𝑖𝑛 𝛽𝑥 ⟹

𝑦𝐻 = 𝑒 𝛼𝑥 𝑐1 + 𝑐2 cos 𝛽𝑥 + 𝑐1 − 𝑐2 𝑖. 𝑠𝑖𝑛 𝛽𝑥 ⟹

𝑐1 𝑐2
𝑦𝐻 = 𝑒 𝛼𝑥 𝑐1 cos 𝛽𝑥 + 𝑐2 𝑠𝑖𝑛 𝛽𝑥
II. Nonhomogeneous solution(𝒚𝑷 ):
𝑦𝑃 is calculated either by the Undetermined Coefficients Method (UCF) or by the Variation of Parameters Method (VPM),
depending on the function 𝑓(𝑥). Generally, if 𝑓(𝑥) is in a simple form, 𝑦𝑃 is calculated by the Undetermined Coefficients
Method, if it is in a difficult form, by the variation of parameters method over 𝑦𝐻 .
1
𝑓 𝑥 = 𝑥, 𝑥 2 , 𝑥 3 … 𝑓 𝑥 =
Undetermined 𝑥 Variation of
1
𝑓 𝑥 = 𝑆𝑖𝑛 𝑥 , 𝐶𝑜𝑠(𝑥) Coefficients 𝑓 𝑥 = Parameters
𝑆𝑖𝑛(𝑥) Method (VPM)
Method (UCF)
𝑓 𝑥 = 𝑒𝑥 𝑓 𝑥 = 𝑙𝑛𝑥

1-Undetermined Coefficients Method (UCF) –(Establishing Analogy with f(x) and its Derivatives):
To find the non-homogeneous solution of 𝒚𝑷 with this method, the function 𝒇 𝒙 itself and its derivatives must be finite in
number. For 𝑓 𝑥 structures that give a new function when differentiated, Variation of Parameters Method (VPM) should be
used. For example, n new functions are formed by deriving functions such as 𝑓(𝑥) = 1/𝑥, 𝑓(𝑥) = sin 𝑥 /𝑥, 𝑓(𝑥) = tan 𝑥,
𝑓 𝑥 = sin 𝑥 . ln 𝑥 n times. Therefore, the undetermined coefficients method cannot be used for 𝑓(𝑥). If there are functions
such as 𝑓(𝑥) = 𝑥 2 , 𝑓(𝑥) = 𝑒 𝑥 , 𝑓(𝑥) = sin 𝑥, several new functions are formed by deriving them n times. These new functions
are multiplied by undetermined constants such as A, B, C, and... and added to determine the particular solution form. The
method is called undetermined coefficients due to its multiplication with undetermined coefficients.
Examples: 𝟏. 𝑓 𝑥 = 𝑥 2 ⟹ 𝑓 ′ 𝑥 = 2𝑥, 𝑓 ′′ 𝑥 = 2, 𝑓 ′′′ 𝑥 = 0
There are 3 functions (𝑥 2 , 𝑥,constant) ⟹ 𝒚𝑷 = 𝑨𝒙𝟐 + 𝑩𝒙 + 𝑪 (This 𝑦𝑃 specific solution should be suggested.)

𝟐. 𝑓 𝑥 = 𝑒 2𝑥 ⟹ 𝑓 ′ 𝑥 = 2𝑒 2𝑥 , 𝑓 ′′ 𝑥 = 4𝑒 2𝑥 , 𝑓 ′′′ 𝑥 = 8𝑒 2𝑥
There is one function (𝑒 2𝑥 ) ⟹ 𝒚𝑷 = 𝑨𝒆𝟐𝒙 (This 𝑦𝑃 specific solution should be suggested.)

𝟑. 𝑓 𝑥 = sin 𝑥 ⟹ 𝑓 ′ 𝑥 = cos 𝑥 , 𝑓 ′′ 𝑥 = −sin 𝑥 , 𝑓 ′′′ 𝑥 = −cos 𝑥

There are 2 functions (sin 𝑥 and cos 𝑥) ⟹ 𝒚𝑷 = 𝑨 𝒔𝒊𝒏 𝒙 + 𝑩 𝒄𝒐𝒔 𝒙 (This 𝑦𝑃 specific solution should be suggested.)

𝟒. 𝑓 𝑥 = 𝑥 cos 𝑥 ⟹ 𝑓 ′ 𝑥 = cos 𝑥 − 𝑥 sin 𝑥 , 𝑓 ′′ 𝑥 = −2sin 𝑥 − 𝑥 cos 𝑥 , 𝑓 ′′′ 𝑥 = −3cos 𝑥 + 𝑥 sin 𝑥

There are 4 functions (sin 𝑥, cos 𝑥, 𝑥 sin 𝑥 and 𝑥 cos 𝑥) ⟹ 𝒚𝑷 = 𝑨 𝒔𝒊𝒏 𝒙 + 𝑩 𝒄𝒐𝒔 𝒙 + 𝑪𝒙 𝒔𝒊𝒏 𝒙 + 𝑫𝒙 𝐜𝐨𝐬 𝒙
(This 𝑦𝑃 specific solution should be suggested.)

In the examples, the given 𝒚𝑷 and the required number of derivatives are taken and written in their places in the non-
homogeneous differential equation. Undetermined coefficients such as A, B, C, D, ... are determined according to the
principle of providing a non-homogeneous differential equation.
• One thing to be careful about when using this method is that the proposed particular solution should never resemble the
homogeneous solution of the differential equation. This similarity is based on functions and has previously been called a
linear independent solution. If there is a similarity between homogeneous (𝒚𝑯 ) and non-homogeneous (𝒚𝑷) solutions, the
numerical value of all or several of the coefficients A, B, C, D in the 𝒚𝑷 proposal cannot be calculated, or these constants
are found as a function of the independent variable. The coefficients A, B, C, ... in the 𝒚𝑷 proposal must definitely have a
fixed and numerical value. If the special solution and homogeneous solutions are similar to each other, linear independent
solutions of the proposed 𝒚𝑷 solutions are produced for the correct solution. For this purpose, the solution at hand is
multiplied by the linear equation (𝒙). This multiplication process is continued sequentially as many times as necessary.
Therefore, more than one linear independent solution can be produced from a solution. Determine the 𝒚𝑷 special solution
forms according to the 𝒇(𝒙) and 𝒚𝑯 homogeneous solution forms given below.
Examples: 1. 𝑦𝐻 = 𝐶1 sin 𝑥 + 𝐶2 𝑒 −2𝑥 and 𝑓 𝑥 = sin 𝑥 ⟹ 𝑓 ′ 𝑥 = cos 𝑥 , 𝑓 ′′ 𝑥 = −sin 𝑥 , 𝑓 ′′′ 𝑥 = −cos 𝑥
There are 2 functions (sin 𝑥 and cos 𝑥) ⟹ 𝒚𝑷 = 𝑨 𝒔𝒊𝒏 𝒙 + 𝑩 𝒄𝒐𝒔 𝒙
Here, 𝑪𝟏 𝒔𝒊𝒏 𝒙 and 𝑨 𝒔𝒊𝒏 𝒙 solutions are the same. Therefore, the expression 𝑨 𝒔𝒊𝒏 𝒙 should be multiplied by 𝒙
when proposing 𝒚𝑷 until the similarity disappears. As a result;
𝒚𝑷 = 𝑨𝒙 𝒔𝒊𝒏 𝒙 + 𝑩 𝒄𝒐𝒔 𝒙
2. 𝑦𝐻 = 𝐶1 𝑒 2𝑥 + 𝐶2 𝑒 −2𝑥 and 𝑓 𝑥 = 𝑒 2𝑥 ⟹ 𝑓 ′ 𝑥 = 2𝑒 2𝑥 , 𝑓 ′′ 𝑥 = 4𝑒 2𝑥 , 𝑓 ′′′ 𝑥 = 8𝑒 2𝑥
There is 1 function (𝑒 2𝑥 ) ⟹ 𝒚𝑷 = 𝑨𝒆𝟐𝒙
Here, 𝑪𝟏 𝒆𝟐𝒙 and 𝑨𝒆𝟐𝒙 solutions are the same. Therefore, the expression𝑨𝒆𝟐𝒙 should be multiplied by 𝒙 when
proposing 𝒚𝑷 until the similarity disappears. As a result;
𝒚𝑷 = 𝑨𝒙𝒆𝟐𝒙
2-Variation of Parameters Method (VPM):
The particular solution is obtained by analogy with the homogeneous solution of the differential equation. For this purpose,
the integral constants in the homogeneous solution are accepted as functions depending on the independent variable. These
new functions are determined based on the principle that the particular solution must satisfy the main differential equation. It
is called the change of constants due to the transformation of the fixed integral constants into variable-functional form. This
method, where mathematical operations are many and difficult, should not be used if possible. This method should only be
used in cases where the function 𝑓(𝑥) is complex, 𝑓(𝑥) and its derivatives do not give a finite number of functions, and
therefore a specific solution proposal cannot be made according to f(x) and its derivatives. If the particular solution is written
by looking at the function 𝑓 𝑥 and its derivatives, and 𝑓 𝑥 has a finite number of derivatives, this method should not be
preferred. One or all of the integral constants in the homogeneous solution can be made functional.

Examples: 𝟏. 𝑦𝐻 = 𝐶1 𝑒 −2𝑥 + 𝐶2 𝑒 𝑥 ⟹ 3 particular solutions can be suggested.

i. 𝑦𝑃 = 𝐶1 𝑥 𝑒 −2𝑥 + 𝐶2 𝑥 𝑒 𝑥
ii. 𝑦𝑃 = 𝐶1 𝑥 𝑒 −2𝑥 𝑥
iii. 𝑦𝑃 = 𝐶2 𝑥 𝑒 𝑥
2. 𝑦𝐻 = 𝐶1 sin 3𝑥 + 𝐶2 cos 3𝑥 ⟹ 3 particular solutions can be suggested.
i. 𝑦𝑃 = 𝐶1 (𝑥) sin 3𝑥 + 𝐶2 (𝑥) cos 3𝑥
ii. 𝑦𝑃 = 𝐶1 (𝑥) sin 3𝑥
iii. 𝑦𝑃 = 𝐶2 (𝑥) cos 3𝑥
One of these suggested forms is preferred. These suggested solutions should provide the non-homogeneous differential equation.
For this purpose, these 𝒚𝑷 solutions themselves and the necessary number of their derivatives are transferred to the non-
homogeneous equation. 𝑪𝟏 (𝒙) and 𝑪𝟐 (𝒙) functions are determined in a way that will provide the non-homogeneous differential
equation. Since they require more processing, ii and iii cases should be preferred from the 3 suggested solutions.

III. General solution(y):

Finally, the general solution is obtained by adding the homogeneous solution (𝒚𝑯 ) obtained and the particular solution (𝒚𝑷 )
obtained by the above methods (UCF and VPM).

𝒚 = 𝒚𝑯 + 𝒚𝑷
Examples: Solve the given homogeneous differantial equations.

1) 𝑦 ′′ − 9𝑦 = 0 3) 𝑦 ′′ − 8𝑦 ′ + 16𝑦 = 0
𝑘2 − 9 = 0 (Characteristic equation) 𝑘 2 − 8𝑘 + 16 = 0 (Characteristic equation)
𝑘1 = +3, 𝑘2 = −3 ( i) Distinct Real roots state ) 𝑘−4 2 = 0 ⟹ 𝑘1 = 𝑘2 = +4

𝑦𝑃 = 0 and 𝑦 = 𝑦𝐻 = 𝑐1 𝑒 3𝑥 + 𝑐2 𝑒 −3𝑥 ( ii) Repeated roots state )

𝑦𝑃 = 0 and 𝑦 = 𝑦𝐻 = 𝑐1 + 𝑐2 𝑥 𝑒 4𝑥

2) 𝑦 ′′ + 9𝑦 = 0 4) 𝑦 𝚤𝑣 − 64𝑦 = 0
𝑘2 + 9 = 0 (Characteristic equation) 𝑘 4 − 64 = 0 (Characteristic equation)
𝑘1 = −3𝑖, 𝑘2 = +3𝑖 ( iii) Complex roots state ) 𝑘2 − 8 . 𝑘2 + 8 = 0
𝑦𝑃 = 0 and 𝛼 = 0, 𝛽 = 3 𝑘 2 − 8 = 0 ⟹ 𝑘1 = 8, 𝑘2 = − 8

𝑦 = 𝑦𝐻 = 𝑒 0 𝑐1 𝑆𝑖𝑛(3𝑥) + 𝑐2 𝐶𝑜𝑠(3𝑥) ⟹ 𝑘 2 + 8 = 0 ⟹ 𝑘3 = 8𝑖, 𝑘4 = − 8𝑖 ⟹ 𝛼 = 0, 𝛽 = 8

( i) Distinct Real and ii) Complex roots state )
𝑦 = 𝑦𝐻 = 𝑐1 𝑆𝑖𝑛(3𝑥) + 𝑐2 𝐶𝑜𝑠(3𝑥)
𝑦𝑃 = 0 and
𝑦 = 𝑦𝐻 = 𝑐1 𝑒 8𝑥 + 𝑐2 𝑒 − 8𝑥 + 𝑐3 𝑆𝑖𝑛( 8𝑥) + 𝑐4 𝐶𝑜𝑠( 8𝑥)
5) 𝑦 ′′′ − 6𝑦 ′′ + 12𝑦 ′ − 8𝑦 = 0 7) 𝑑2 𝑠
+ 𝑤2𝑠 = 0
𝑘 3 − 6𝑘 2 + 12𝑘 − 8 = 0 (Characteristic equation) 𝑑𝑡 2

3
𝑘−2 = 0 ⟹ 𝑘1 = 𝑘2 = 𝑘3 = +2 𝑘2 + 𝑤 2 = 0 (Characteristic equation)
( ii) Repeated roots state ) 𝑘 2 + 𝑤 2 = 0 ⟹ 𝑘1 = −𝑤𝑖, 𝑘2 = +𝑤𝑖 ⟹ 𝛼 = 0, 𝛽 = 𝑤
𝑦𝑃 = 0 and 𝑦 = 𝑦𝐻 = 𝑐1 + 𝑐2 𝑥 + 𝑐3 𝑥 2 𝑒 2𝑥 ( iii) Complex roots state )

𝑠𝑃 = 0 and 𝑠 = 𝑠𝐻 = 𝑐1 𝑆𝑖𝑛(𝑤𝑡) + 𝑐2 𝐶𝑜𝑠(𝑤𝑡)

6) 𝑦 𝚤𝑣 − 𝑦′′′ = 0
𝑑2 𝑅 𝑑𝑅
𝑘3 𝑘−1 =0 (Characteristic equation) 8) + 4 + 5𝑅 = 0
𝑑𝜃 2 𝑑𝜃
𝑘1 = 𝑘2 = 𝑘3 = 0, 𝑘4 =1
𝑘 2 + 4𝑘 + 5 = 0 (Characteristic equation)
( i) Distinct Real and ii) Repeated roots state )
∆= 𝑏 2 − 4𝑎𝑐 = 42 − 4.1.5 = −4
𝑦𝑃 = 0 and 𝑦 = 𝑦𝐻 = 𝑐1 + 𝑐2 𝑥 + 𝑐3 𝑥 2 + 𝑒 𝑥
−𝑏 ± ∆ −4 ± −4
𝑘1,2 = = ⟹ 𝑘1,2 = −2 ± 𝑖 ⟹
2𝑎 2.1
𝑅𝑃 = 0 and 𝛼 = −2, 𝛽 = 1 ( iii) Complex roots state )

𝑅 = 𝑅𝐻 = 𝑒 −2𝜃 𝑐1 𝑆𝑖𝑛(𝜃) + 𝑐2 𝐶𝑜𝑠(𝜃)

Examples: Solve the given nonhomogeneous differantial equations.

1) 𝑦 ′′ − 4𝑦 = 16𝑒 𝑥 𝑦𝑃 = 𝑦𝑃 ′ = 𝑦𝑃 ′′ = 𝐴𝑒 𝑥
𝑦𝐻 𝑦𝑃 𝑦 ′′ − 4𝑦 = 16𝑒 𝑥 ⟹ 𝑦𝑃 ′′ − 4𝑦𝑃 = 16𝑒 𝑥 ⟹
Homogeneous solution (𝑦𝐻 ):
16
𝑦 ′′ − 4𝑦 = 0 ⟹ 𝑘2 −4=0 (Characteristic equation) 𝐴𝑒 𝑥 − 4𝐴𝑒 𝑥 = 16𝑒 𝑥 ⟹ 𝐴 = −
3
𝑘1 = +2, 𝑘2 = −2 ( i) Distinct Real roots state ) 16 𝑥
𝑦𝑃 = − 𝑒
3
𝑦𝐻 = 𝑐1 𝑒 2𝑥 + 𝑐2 𝑒 −2𝑥
𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹
Nonhomogeneous solution (𝑦𝑃 ): 16 𝑥
𝑦 = 𝑐1 𝑒 2𝑥 + 𝑐2 𝑒 −2𝑥 − 𝑒
𝑓(𝑥) = 16𝑒 𝑥 It is a simple function. Undetermined 3
Coefficients Method (UCF) can be used
for the solution.

𝑓 𝑥 = 𝑓 ′ 𝑥 = 𝑓 ′′ 𝑥 = 𝑓 ′′′ 𝑥 = ⋯ = 𝑓 𝑛 𝑥 = 16𝑒 𝑥

There is no similarity between 𝑦𝑃 and 𝑦𝐻 . Particular solution

can be proposed as 𝑦𝑃 = 𝐴𝑒 𝑥
2) 𝑦 ′′ − 4𝑦 = 9𝑒 2𝑥 According to Linear Independent Solution (LIS) Method;
𝑦𝐻 𝑦𝑃
𝑦𝑃 = 𝐴. 𝑥. 𝑒 2𝑥
Homogeneous solution (𝑦𝐻 ):
𝑦𝑃 ′ = 𝐴𝑒 2𝑥 + 2𝐴𝑥𝑒 2𝑥
𝑦 ′′ − 4𝑦 = 0 ⟹ 𝑘 2 − 4 = 0 (Characteristic equation)
𝑦𝑃 ′′ = 4𝐴𝑒 2𝑥 + 4𝐴𝑥𝑒 2𝑥
𝑘1 = +2, 𝑘2 = −2 ( i) Distinct Real roots state )
𝑦 ′′ − 4𝑦 = 9𝑒 2𝑥 ⟹ 𝑦𝑃 ′′ − 4𝑦𝑃 = 9𝑒 2𝑥 ⟹
𝑦𝐻 = 𝑐1 𝑒 2𝑥 + 𝑐2 𝑒 −2𝑥 9
4𝐴𝑒 2𝑥 + 4𝐴𝑥𝑒 2𝑥 − 4𝐴𝑥𝑒 2𝑥 = 9𝑒 2𝑥 ⟹ 𝐴 =
4
Nonhomogeneous solution (𝑦𝑃 ): 9 𝑥
𝑦𝑃 = 𝑥𝑒
𝑓(𝑥) = 9𝑒 2𝑥 It is a simple function. Undetermined 4
Coefficients Method (UCF) can be used 𝑦 = 𝑦𝐻 + 𝑦𝑃 ⟹
for the solution.
9 𝑥
𝑦 = 𝑐1 𝑒 2𝑥 + 𝑐2 𝑒 −2𝑥 + 𝑥𝑒
𝑓 𝑥 = 𝑒 2𝑥 𝑓 ′ 𝑥 = 2𝑒 2𝑥 𝑓 ′′ 𝑥 = 4𝑒 2𝑥 4

Particular solution can be proposed as 𝑦𝑃 = 𝐴𝑒 2𝑥 . But,

𝑦𝑃 and 𝑦𝐻 are partially similar. Therefore,

𝑐1 𝑒 2𝑥 ≈ 𝐴𝑒 2𝑥 ⟹
 𝑒 2𝑥
Any of these can be
3) 𝑦 ′′ − 3𝑦 ′ + 2𝑦 = 𝑥 i) 𝑦𝑃 = 𝑐1 (𝑥)𝑒 𝑥 +𝑐2 (𝑥)𝑒 2𝑥 ,
𝑒 +1 preferred. In terms of ease
𝑦𝐻 𝑦𝑃 ii) 𝑦𝑃 = 𝑐1 (𝑥)𝑒 𝑥 , of operation, it is logical to
prefer single-function (ii) or
Homogeneous solution (𝑦𝐻 ): iii) 𝑦𝑃 = 𝑐2 (𝑥)𝑒 2𝑥
(Characteristic (iii) proposals. With (ii)
𝑦 ′′ − 3𝑦 ′ + 2𝑦 = 0 ⟹ 𝑘 2 − 3𝑘 + 2 = 0 preference, the
equation)
development of the
𝑘1 = +1, 𝑘2 = +2 ( i) Distinct Real roots state )
solution will be as follows:
𝑦𝐻 = 𝑐1 𝑒 𝑥 + 𝑐2 𝑒 2𝑥
𝑦𝑃 = 𝑐1 (𝑥)𝑒 𝑥 ,
Nonhomogeneous solution (𝑦𝑃 ):
𝑦𝑃′ = 𝑐1′ 𝑥 𝑒 𝑥 + 𝑐1 𝑥 𝑒 𝑥 ,
For the solution with the UCF, the derivative of the function
𝑓 𝑥 must be considered n → ∞ times. To determine the 𝑦𝑃′′ = 𝑐1′′ 𝑥 𝑒 𝑥 + 2𝑐1′ 𝑥 𝑒 𝑥 + 𝑐1 𝑥 𝑒 𝑥
coefficients of so many functions, it is not possible to take their 𝑒 2𝑥
derivatives and transfer them to a non-homogeneous 𝑦𝑃′′ − 3𝑦𝑃′ + 2𝑦𝑃 = 𝑥 ⟹
𝑒 +1
differential equation. Therefore, the particular solution of this
differential equation cannot be found with UCF. ⟹ 𝑐1′′ 𝑥 𝑒 𝑥 + 2𝑐1′ 𝑥 𝑒 𝑥 + 𝑐1 𝑥 𝑒 𝑥 − 3 𝑐1′ 𝑥 𝑒 𝑥 + 𝑐1 𝑥 𝑒 𝑥
𝑒 2𝑥 𝑥
𝑒 2𝑥
𝑓(𝑥) = 𝑥 The VPM method can be used for the + 𝑐1 𝑥 𝑒 = 𝑥
𝑒 +1 𝒚𝑷 solution. Three different particular 𝑒 +1
solutions can be proposed in the VPM 𝑒 2𝑥 𝑒𝑥
method: 𝑐1′′ 𝑥 𝑒𝑥 − 𝑐1′ 𝑥 𝑒𝑥 = 𝑥 ′′ ′
⟹ 𝑐1 𝑥 − 𝑐1 𝑥 = 𝑥 ⟹
𝑒 +1 𝑒 +1
 𝑒 𝑥 𝑒 𝑥
The order of the final differential equation obtained is the
same as the equation that is being tried to be solved. While 𝑢′ − 𝑢 = 𝑥 ⟹ 𝑐′ 𝑥 𝑒 𝑥 + 𝑐 𝑥 𝑒 𝑥 − 𝑐 𝑥 𝑒 𝑥 = 𝑥 ⟹
𝑒 +1 𝑒 +1
𝒚 was the dependent variable in the first equation, 𝒄𝟏
became the dependent variable in the last equation. The 1 1 𝑒 −𝑥 𝑒 −𝑥 𝑑𝑥
⟹ 𝑐′ 𝑥 = 𝑥 ⟹ න 𝑑𝑐 = න 𝑥 𝑑𝑥 = න −𝑥
difference is that the order of the last equation can be 𝑒 +1 𝑒 + 1 𝑒 −𝑥 𝑒 +1
reduced. In order to reduce the order, either the
dependent or the independent variable should not be in ⟹ 𝑐 𝑥 = − ln 𝑒 −𝑥 + 1 ⟹ 𝑢 = 𝑐 𝑥 𝑒 𝑥
the equation. Therefore, in the last stage of the solution, 𝒄 ⟹ 𝒖 = −𝒆𝒙 𝒍𝒏(𝒆−𝒙 + 𝟏)
should not appear in order to be able to talk about a
progress. Only its derivatives should be in the equation. 𝑐1′ 𝑥 = 𝑢 𝑥 ⟹ 𝑐1′ 𝑥 = −𝑒 𝑥 𝑙𝑛(𝑒 −𝑥 + 1) ⟹
This is a must-have condition in the method of changing
constants. The solution will continue with decreasing the ⟹ න 𝑑𝑐1 = − න 𝑙𝑛 𝑒 −𝑥 + 1 𝑒 𝑥 𝑑𝑥 ⟹ 𝑢 = − 𝑙𝑛 𝑒 −𝑥 + 1
order.
𝑢 𝑑𝑣 𝑒 −𝑥 𝑑𝑥
𝑒𝑥 ⟹ 𝑑𝑢 = −𝑥
𝑐1′ 𝑥 = 𝑢 𝑥 ⟹ 𝑐1′′ 𝑥 = 𝑢′ 𝑥 ⟹ 𝑢′ 𝑥 − 𝑢 𝑥 = 𝑒 +1
𝑒𝑥 + 1
This equation can be solved by the first order linear differential 𝑑𝑣 = 𝑒 𝑥 𝑑𝑥 ⟹ 𝑣 = 𝑒 𝑥
equation approach. 𝑒 −𝑥 𝑑𝑥
𝑒𝑥 𝑑𝑢 𝑐1 = −𝑒 𝑥 𝑙𝑛 𝑒 −𝑥 +1 − න 𝑒𝑥 ⟹
′
𝑢 −𝑢 = 𝑥 ′
⟹𝑢 −𝑢 =0⟹න = න 𝑑𝑥 ⟹ 𝑒 −𝑥 + 1
𝑒 +1 𝑢
𝑥
ln 𝑢 = 𝑥 + ln 𝑐 ⟹ 𝒖𝑯 = 𝒄𝒆𝒙 𝑑𝑥 𝑒
𝑐1 = −𝑒 𝑥 𝑙𝑛 𝑒 −𝑥 + 1 − න −𝑥 ⟹
𝑢𝐻 = 𝑐𝑒 𝑥 ⟹ 𝑢 = 𝑐 𝑥 𝑒 𝑥 ⟹ 𝑢′ = 𝑐 ′ 𝑥 𝑒 𝑥 + 𝑐(𝑥)𝑒 𝑥 ⟹ 𝑒 + 1 𝑒𝑥
⟹ 𝒄𝟏 = −𝒆𝒙 𝒍𝒏 𝒆−𝒙 + 𝟏 − 𝒍𝒏 𝒆𝒙 + 𝟏

𝒚𝑷 = 𝒄𝟏 𝒙 𝒆𝒙 = −𝒆𝒙 𝒆𝒙 𝒍𝒏 𝒆−𝒙 + 𝟏 + 𝒍𝒏 𝒆𝒙 + 𝟏

𝒚𝑯 = 𝒄𝟏 𝒆𝒙 + 𝒄𝟐 𝒆𝟐𝒙

𝒚 = 𝒚𝑯 + 𝒚𝑷 = 𝒄𝟏 𝒆𝒙 + 𝒄𝟐 𝒆𝟐𝒙 − 𝒆𝒙 𝒆𝒙 𝒍𝒏 𝒆−𝒙 + 𝟏 + 𝒍𝒏 𝒆𝒙 + 𝟏
 𝜕𝑐
𝑁 𝑥, 𝑦 = = 𝐶𝑜𝑠 𝑥 + 𝑦 𝑦 + tan(𝑥 + 𝑦) ⟹
𝜕𝑦
න𝜕𝑐 = න 𝐶𝑜𝑠 𝑥 + 𝑦 𝑦 + 𝑆𝑖𝑛(𝑥 + 𝑦) 𝜕𝑦 ⟹

න𝜕𝑐 = න𝑦𝐶𝑜𝑠 𝑥 + 𝑦 𝑑𝑦 + න 𝑆𝑖𝑛(𝑥 + 𝑦)𝜕𝑦 ⟹

𝑢 𝑑𝑣

න𝜕𝑐 = 𝑦𝑆𝑖𝑛 𝑥 + 𝑦 − න 𝑆𝑖𝑛 𝑥 + 𝑦 𝜕𝑦 + න 𝑆𝑖𝑛(𝑥 + 𝑦)𝜕𝑦 ⟹

𝒄 = 𝒚𝑺𝒊𝒏 𝒙 + 𝒚 + 𝒄𝟏 (𝒙)

𝜕𝑐 𝑑𝑐1 𝑑𝑐1
= 𝑦𝐶𝑜𝑠 𝑥 + 𝑦 + = 𝑀 𝑥, 𝑦 ⟹ 𝑦𝐶𝑜𝑠 𝑥 + 𝑦 + = 𝑦𝐶𝑜𝑠 𝑥 + 𝑦 ⟹
𝜕𝑥 𝑑𝑥 𝑑𝑥
𝑑𝑐1
= 0 ⟹ 𝑐1 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝑑𝑥
𝒄 = 𝒚𝑺𝒊𝒏 𝒙 + 𝒚