__________________________________________________________________________________15MA21-Partial Differential Equations

PARTIAL DIFFERENTIAL EQUATIONS

MODULE 3
SUBJECT CODE: 15MA21

Module 3 syllabus:

1. Formulation of a Partial Differential Equation by elimination of arbitrary

constants/functions
2. Solution of nonhomogeneous Partial Differential Equations by direct
integration
3. Solution of homogeneous Partial Differential Equations involving
derivative with respect to one independent variable only
4. Derivation of one dimensional Heat and Wave equations
5. Solution of one dimensional Heat and Wave equations by variable
separable method

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

PARTIAL DIFFERENTIAL EQUATIONS

An equation involving one or more partial derivatives of a function of two or more

variables is called a Partial Differential Equation (PDE). The order of a PDE is the order of
the highest derivative present in it the degree of a PDE is the degree of the highest order
derivative present in it after clearing the equation of fractional powers. If the dependent variable
and its partial derivatives in a PDE are of first degree then the PDE is said to be linear. A PDE
is homogeneous if each term in it contains either the dependent variable or one of its partial
derivatives, otherwise, the PDE is nonhomogeneous.
Examples:
𝛛𝐳 𝛛𝐳
1. 𝟐 𝛛𝐱 + 𝟑 𝛛𝐲 = 𝟎 is a homogeneous PDE of order 1 and degree 1

𝛛𝟐 𝐳 𝛛𝟐 𝐳
2. 𝛛𝐱𝟐 + 𝛛𝐲𝟐 = 𝟎 is a homogeneous PDE of order 2 and degree 1

𝛛𝟐 𝐳
3. 𝛛𝐱𝟐 = 𝐱 + 𝐲 is a nonhomogeneous PDE of order 2 and degree 1
𝛛𝟐 𝐳
4. 𝛛𝐱𝛛𝐲 − 𝟒𝐱𝐲 = 𝟎 is a nonhomogeneous PDE of order 2 and degree 1
𝛛𝐳 𝛛𝐳
5. 𝐱 𝛛𝐱 + 𝐲 𝛛𝐲 = 𝐳 is a linear PDE of order 1 and degree 1, nonhomogeneous

𝛛𝟐 𝐳
6. (𝛛𝐱𝟐 )𝟐 − 𝟑𝐱 = 𝐲 is a nonhomogeneous PDE of order 2 and degree 2.

Formation of a PDE
PDE’s can be formed by two methods:
1. Elimination of arbitrary constants from a relation
2. Elimination of arbitrary functions from a relation
In the case of arbitrary constants, if the number of arbitrary constants is equal to the
number of independent variables, we obtain a PDE of order one. If the number of constants is
more than the number of independent variables, we obtain a PDE of order 2 or higher.
In the case of arbitrary functions, the order of the PDE is equal to the number of
functions eliminated. When eliminating arbitrary functions, constants (if present) need not be
eliminated.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Standard Notations
𝐖𝐞 𝐜𝐨𝐧𝐬𝐢𝐝𝐞𝐫 𝐳 𝐚𝐬 𝐚 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧 𝐨𝐟 𝐭𝐰𝐨 𝐢𝐧𝐝𝐞𝐩𝐞𝐧𝐝𝐞𝐧𝐭 𝐯𝐚𝐫𝐢𝐚𝐛𝐥𝐞𝐬 𝐱 𝐚𝐧𝐝 𝐲 𝐢. 𝐞. 𝐳 = 𝐳(𝐱, 𝐲)
𝛛𝐳 𝛛𝐳 𝛛𝟐 𝐳
𝐀𝐥𝐬𝐨, = 𝐳𝐱 = 𝐩, = 𝐳𝐲 = 𝐪, = 𝐳𝐱𝐱 = 𝐫,
𝛛𝐱 𝛛𝐲 𝛛𝐱 𝟐
𝛛𝟐 𝐳 𝛛𝟐 𝐳 𝛛𝟐 𝐳
= 𝐳𝐱𝐲 (= = 𝐳𝐲𝐱 ) = 𝐬, 𝟐 = 𝐳𝐲𝐲 = 𝐭.
𝛛𝐱𝛛𝐲 𝛛𝐲𝛛𝐱 𝛛𝐲

Examples:
Formation of a PDE by the elimination of arbitrary constants
Note:
The elimination of arbitrary constants is done by differentiating the given
expression partially. If the expression contains two arbitrary constants, then p and q are
sufficient. For more than two constants we require higher order partial derivatives.
_________________________________________
From the PDE by eliminating the arbitrary constants from the following relations:
1. 𝐳 = (𝐱 + 𝐚)(𝐲 + 𝐛)
Solution:
On differentiating partially w.r.t x and y, we obtain:
∂z
= p= y+b
∂x
∂z
and = q = x + a.
∂y
This gives pq = (y + b)(x + a) = z.
Hence 𝐳 = 𝐩𝐪 is the required PDE.

𝐱𝟐 𝐲𝟐
2. 𝐳 𝟐 = 𝐚𝟐 + 𝐛𝟐

Solution:
On differentiating partially w.r.t x, we get:
∂z 2x x
2z = 2 or zp = 2 … … . . (1)
∂x a a
On differentiating partially w.r.t y, we get:
∂z 2y y
2z = 2 or zq = 2 … … . . (2)
∂y b b
Multiplying (1) by x and (2) by y and adding, we get:

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

x2 y2
xzp + yzq = 2 + 2
a b
or xzp + yzq = z 2
This gives 𝐱𝐩 + 𝐲𝐪 = 𝐳 as the required PDE.

3. 𝐳 = (𝐱 𝟐 + 𝐚)(𝐲 𝟐 + 𝐛)
Solution:
Differentiating the given relation partially w.r.t x and y, we get:
∂z
= p = 2x(y 2 + b) … … . . (1)
∂x
and
∂z
= q = 2y(x 2 + a) … … . . (2)
∂y
Multiplying (1) and (2),
pq = 4xy(x 2 + a)(y 2 + b) = 4xyz.
Hence 𝒑𝒒 = 𝟒𝒙𝒚𝒛 is the required PDE.

4. 𝒛 = 𝒙𝒚 + 𝒚√𝒙𝟐 − 𝒂𝟐 + 𝒃)
Solution:
Differentiating the given relation partially w.r.t x and y, we get:
𝜕𝑧 1
=𝑝=𝑦+𝑦 2𝑥 𝑜𝑟
𝜕𝑥 2√𝑥 2 − 𝑎2
𝑥𝑦
𝑝=𝑦+ … … . . (1)
√𝑥 2 − 𝑎2
𝜕𝑧
= 𝑞 = 𝑥 + √𝑥 2 − 𝑎2 … . . (2)
𝜕𝑦
𝑥𝑦
From (1), (𝑝 − 𝑦) = √𝑥 2 … … . . (3)
−𝑎2

and from (2), (𝑞 − 𝑥) = √𝑥 2 − 𝑎2 ….(4)

Multiplying (3) and (4):
(𝑝 − 𝑦)(𝑞 − 𝑥) = 𝑥𝑦
or 𝑝𝑞 − 𝑥𝑝 − 𝑦𝑞 + 𝑥𝑦 = 𝑥𝑦.
This gives 𝒙𝒑 + 𝒚𝒒 = 𝒑𝒒, which is the required PDE.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

5. 𝒛 = (𝒙 − 𝐚)𝟐 + (𝒚 − 𝒃)𝟐
Solution:
Differentiating partially w.r.t x and y respectively,
𝜕𝑧
= 𝑝 = 2(𝑥 − 𝑎) ……..(1)
𝜕𝑥
𝜕𝑧
= 𝑞 = 2(𝑦 − 𝑎) ……..(2)
𝜕𝑦
𝑝
From equation (1), (𝑥 − 𝑎) = 2
𝑞
and from (2), (𝑦 − 𝑏) = 2

Substituting in the given relation,

𝟒𝒛 = 𝒑𝟐 + 𝒒𝟐 is the required PDE.

6. 𝒛 = 𝒂𝒙 + 𝒃𝒚 + 𝒄𝒙𝒚
Solution:
 This relation contains 3 arbitrary constants a, b and c which have to be eliminated
 The order of the PDE will be greater than one
 Hence, we need to obtain the second order partial derivatives also

Differentiating the given relation partially w.r.t x and y, we get:

𝜕𝑧
= 𝑝 = 𝑎 + 𝑐𝑦 ……..(1)
𝜕𝑥
𝜕𝑧
= 𝑞 = 𝑏 + 𝑐𝑥 ……..(2)
𝜕𝑦

Now differentiating (1) partially w.r.t x and (2) partially w.r.t y, we get:
𝜕2 𝑧
= 𝑟 = 0 ……..(3)
𝜕𝑥 2
𝜕2 𝑧
= 𝑡 = 0 ……..(4)
𝜕𝑦 2

Finally differentiating (1) partially w.r.t y (or equivalently differentiating (2) partially w.r.t
x), we get:
𝜕2 𝑧
= 𝑠 = 𝑐 ……..(5)
𝜕𝑥𝜕𝑦

Note:
When the PDE is of order 2 (or more), not all the partial derivatives will be
necessary during the elimination of arbitrary constants. In this case, the partial
derivatives r and t obtained in equations (3) and (4) are not necessary. A Trial and Error
method will ensure the correct PDE.
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Multiplying equation (1) by x and equation (2) by y and adding the resulting
expressions, we get:

𝑝𝑥 + 𝑞𝑦 = 𝑎𝑥 + 𝑐𝑥𝑦 + 𝑏𝑦 + 𝑐𝑥𝑦
or 𝑝𝑥 + 𝑞𝑦 = (𝑎𝑥 + 𝑏𝑦 + 𝑐𝑥𝑦) + (𝑐𝑥𝑦)
Using the given expression and (5), we immediately get:
𝒑𝒙 + 𝒒𝒚 = 𝒛 + 𝒔𝒙𝒚.
This is the required PDE.

𝒙𝟐 𝒚𝟐 𝒛𝟐
7. + 𝒃𝟐 + 𝒄𝟐 = 𝟏
𝒂𝟐

Solution:
 In this example there are three arbitrary constants a, b and c. Elimination of these
constants will result in a PDE of order 2
 This example can be solved using two methods:
i) By differentiating the given expression two times partially w.r.t x
OR
ii) By differentiating the given expression two times partially w.r.t y
Both methods are equivalent.

We illustrate the first method.

Differentiating the given expression partially w.r.t x, we get:
𝜕𝑧
2𝑥 2𝑧
𝜕𝑥
+ =0
𝑎2 𝑐2
𝜕𝑧
𝑥 𝑧
𝜕𝑥
or + = 0 ………(1)
𝑎2 𝑐2
Differentiating (1) partially w.r.t x again, we get:
𝜕2 𝑧 𝜕𝑧 𝜕𝑧
1 (𝑧 + )
𝜕𝑥2 𝜕𝑥𝜕𝑥
+ = 0 ……..(2)
𝑎2 𝑐2
From (1),
𝑥 𝑧𝑝
2
+ =0
𝑎 𝑐2
𝑥 𝑧𝑝
or =−
𝑎2 𝑐2
𝑐2 𝑧𝑝
or =− ……..(3)
𝑎2 𝑥
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

From (2),
1 𝑧𝑟+𝑝2
2
+ =0
𝑎 𝑐2
1 𝑧𝑟+𝑝2
or =−
𝑎2 𝑐2
𝑐2
or = −(𝑧𝑟 + 𝑝2 ) ……..(4)
𝑎2

From (3) and (4),

𝑧p
(𝑧𝑟 + 𝑝2 ) =
𝑥

or 𝒛𝒑 = 𝒙𝒛𝒓 + 𝒑𝟐 is the PDE.

Note:
The second method gives the PDE as 𝒛𝒒 = 𝒚𝒛𝒕 + 𝒒𝟐 .

8. Find the PDE of all spheres of radius k having their centres in the xy-plane (or the
plane z=0).
Solution:
The equation of the sphere having centre at (𝑎, 𝑏, 𝑐) 𝑎𝑛𝑑 𝑟𝑎𝑑𝑖𝑢𝑠 𝑘 𝑖𝑠:
(𝑥 − 𝑎)2 + (𝑦 − 𝑏)2 + (𝑧 − 𝑐)2 = 𝑘 2
Since the centre on the xy-plane, we have 𝑐 = 0. Hence the equation of the sphere is:
(𝑥 − 𝑎)2 + (𝑦 − 𝑏)2 + 𝑧 2 = 𝑘 2 … … . . (1)

Note:
We have to eliminate the arbitrary constants a and b since for different values of
a and b, the centre of the sphere changes. The constant k need not be eliminated because
it the radius of the sphere (fixed constant).

Differentiating (1) partially w.r.t x and y, we get respectively:

𝜕𝑧
2(𝑥 − 𝑎) + 2𝑧 =0
𝜕𝑥
𝑜𝑟 (𝑥 − 𝑎) + 𝑧𝑝 = 0
𝑜𝑟 (𝑥 − 𝑎) = −𝑧𝑝
and
𝜕𝑧
2(𝑦 − 𝑏) + 2𝑧 =0
𝜕𝑦

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑜𝑟 (𝑦 − 𝑏) + 𝑧𝑞 = 0
𝑜𝑟 (𝑦 − 𝑏) = −𝑧𝑞.
Substituting the values of (𝑥 − 𝑎) and (𝑦 − 𝑏) obtained above in (1) we obtain:
𝑧 2 𝑝2 + 𝑧 2 𝑞 2 + 𝑧 2 = 𝑘 2 .
This gives 𝒛𝟐 (𝒑𝟐 + 𝒒𝟐 + 𝟏) = 𝒌𝟐 as the required PDE.

9. Find the PDE representing all planes that are at a perpendicular distance P from a
fixed origin.
Solution:
The equation of the plane is
𝑙𝑥 + 𝑚𝑦 + 𝑛𝑧 = 𝑃 … … . . (1)
𝑤ℎ𝑒𝑟𝑒 𝑙, 𝑚, 𝑛 are the direction cosines of the normal and 𝑃 is the length of the perpendicular
drawn from the plane to the origin. The direction cosines 𝑙, 𝑚, 𝑛 satisfy the relation
𝑙 2 + 𝑚2 + 𝑛2 = 1 ……..(2).
Here, 𝑙 𝑎𝑛𝑑 𝑚 are the arbitrary constants since 𝑛 can be expressed in terms of 𝑙 𝑎𝑛𝑑 𝑚
from (2).
Differentiating (1) partially w.r.t x and y, we get:
𝜕𝑧
𝑙 + 𝑛 𝜕𝑥 =0

or 𝑙 + 𝑛𝑝 = 0……..(3)
and
𝜕𝑧
𝑚+𝑛 =0
𝜕𝑦
or 𝑚 + 𝑛𝑞 = 0 ……..(4)
Substituting (3) and (4) in (2), we get:
𝑛2 𝑝 2 + 𝑛2 𝑞 2 + 𝑛2 = 1
𝑜𝑟 𝑛2 (𝑝2 + 𝑞 2 + 1) = 1
1
This gives 𝑛 = .
√(𝑝2 +𝑞 2 +1)
𝑝
Consequently (3) gives: 𝑙 = −𝑛𝑝 = −
√(𝑝2 +q2 +1)
𝑞
and (4) gives: 𝑚 = −𝑛𝑞 = −
√(𝑝2 +𝑞 2 +1)

Substituting in (1) we get:

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑝𝑥 𝑞𝑦 𝑧
− − + =𝑃
√(𝑝2 +𝑞 2 +1) √(𝑝2 +𝑞 2 +1) √(𝑝2 +𝑞 2 +1)

or
𝑝𝑥 + 𝑞𝑦 − 𝑧 = 𝑃√(𝑝2 + 𝑞 2 + 1)
or
(𝒑𝒙 + 𝒒𝒚 − 𝒛)𝟐 = 𝑷𝟐 (𝒑𝟐 + 𝒒𝟐 + 𝟏)
This is the required PDE.

10. Form the PDE of the family of planes having equal x and y intercepts.
Solution:
The equation of the plane with x, y and z intercepts a, b and c respectively is
𝑥 𝑦 𝑧
+ + = 1.
𝑎 𝑏 𝑐
But given, x and y intercepts are equal, i.e. a=b. Hence the equation of
the plane is
𝑥 𝑦 𝑧
+𝑎+𝑐 =1
𝑎
𝑥+𝑦 𝑧
or + 𝑐 = 1 ……..(1)
𝑎

Differentiating (1) partially w.r.t x and y respectively:

𝜕𝑧
1 𝜕𝑥
+ =0
𝑎 𝑐
1 𝑝
or =−
𝑎 𝑐
𝑐
or − = 𝑝 ……..(2)
𝑎
and
𝜕𝑧
1 𝜕𝑦
+ =0
𝑎 𝑐
1 𝑞
or =−
𝑎 𝑐
𝑐
or − = 𝑞 ……..(3)
𝑎
From (2) and (3), the required PDE is
𝒑 = 𝒒 𝒐𝒓 𝒑 − 𝒒 = 𝟎.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Exercises:
1. 𝐳 = 𝐚𝐱 + 𝐛𝐲 + 𝐚𝐛
2. 𝐳 = 𝐚(𝐱 + 𝐲) + 𝐛
3. 𝐳 = 𝐚𝐱𝐲 + 𝐛
4. 𝐳 = 𝐚𝐱 + 𝐚𝟐 𝐲 𝟐 + 𝐛
𝟏
5. 𝐳 = 𝐚𝐱𝐞𝐲 + 𝟐 𝐚𝟐 𝐞𝟐𝐲 + 𝐛
𝐱𝟐 𝐲𝟐
6. 𝐧𝐳 = 𝐚𝟐 + 𝐛𝟐

7. Form the PDE of the family of planes the sum of whose x, y and z intercepts is 1.
8. Form the PDE of all spheres of radius r whose centres lie on the Z axis.
9. Form the PDE of all spheres whose radii are the same.
10. Form the PDE of all surfaces of revolution having the Z axis as the axis of revolution.

Examples:
Formation of a PDE by the elimination of arbitrary functions
Note:
The elimination of arbitrary functions is done by differentiating the given
expression partially. The order of the PDE obtained will be equal to the number of
arbitrary functions eliminated.
Constants (if present) need not be eliminated here.
_____________________________

Form the PDE by eliminating the arbitrary functions in the following:

𝟏
1. 𝒛 = 𝒚𝟐 + 𝟐𝒇(𝒙 + 𝒍𝒐𝒈𝒚)

Solution:
Differentiating the given relation partially w.r.t x and y respectively, we get:
𝜕𝑧 1 1
= 𝑝 = 2𝑓 ′ (𝑥 + 𝑙𝑜𝑔𝑦) (− 𝑥 2 ) ……..(1) and
𝜕𝑥
𝜕𝑧 1 1
= 𝑞 = 2𝑦 + 2𝑓 ′ (𝑥 + 𝑙𝑜𝑔𝑦) (𝑦) ……..(2)
𝜕𝑦

From (1), we have:

1
−𝑥 2 𝑝 = 2𝑓 ′ (𝑥 + 𝑙𝑜𝑔𝑦) ........(3)

and from (2), we have:

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

1
𝑦(𝑞 − 2𝑦) = 2𝑓 ′ (𝑥 + 𝑙𝑜𝑔𝑦) ……..(4).

Now, from (3) and (4), since the RHS values are same, we can equate the LHS.
Hence,
−𝑥 2 𝑝 = 𝑦(𝑞 − 2𝑦)
or −𝑥 2 𝑝 = 𝑦𝑞 − 2𝑦 2
This gives,
𝒙𝟐 𝒑 + 𝒚𝒒 = 𝟐𝒚𝟐 as the required PDE.

2. 𝒛 = 𝒆𝒂𝒙+𝒃𝒚 𝒇(𝒂𝒙 − 𝒃𝒚)

Solution:
Differentiating partially w.r.t x and y respectively,
𝜕𝑧
𝜕𝑥
= p = 𝑒 𝑎𝑥+𝑏𝑦 𝑓 ′ (𝑎𝑥 − 𝑏𝑦)𝑎 + 𝑓(𝑎𝑥 − 𝑏𝑦)𝑒 𝑎𝑥+𝑏𝑦 𝑎 (product rule)

or 𝑝 = 𝑎(𝑒 𝑎𝑥+𝑏𝑦 𝑓 ′ (𝑎𝑥 − 𝑏𝑦) + 𝑧) ……..(1) and

𝜕𝑧
= 𝑞 = 𝑒 𝑎𝑥+𝑏𝑦 𝑓 ′ (𝑎𝑥 − 𝑏𝑦)(−𝑏) + 𝑓(𝑎𝑥 − 𝑏𝑦)𝑒 𝑎𝑥+𝑏𝑦 𝑏 (product rule)
𝜕𝑦

or 𝑞 = 𝑏(−𝑒 𝑎𝑥+𝑏𝑦 𝑓 ′ (𝑎𝑥 − 𝑏𝑦) + 𝑧) ........(2)

Multiplying (1) by b and (2) by a, we get:
𝑏𝑝 = 𝑎𝑏(𝑒 𝑎𝑥+𝑏𝑦 𝑓 ′ (𝑎𝑥 − 𝑏𝑦) + 𝑧) ……..(3) and
𝑎𝑞 = 𝑎𝑏(−𝑒 𝑎𝑥+𝑏𝑦 f ′ (𝑎𝑥 − 𝑏𝑦) + 𝑧) ……..(4)
Adding (3) and (4),
we get
𝒃𝒑 + 𝒂𝒒 = 𝟐𝒂𝒃𝒛 as the PDE.
3. 𝒛 = 𝒚 𝒇(𝒙) + 𝒙 ∅(𝒚)
Solution:
 In this example, two arbitrary functions f and ∅ have to be eliminated
 The order of the PDE obtained will be 2.
Differentiating the given expression partially w.r.t x and y respectively,
𝜕𝑧
= 𝑝 = 𝑦𝑓 ′ (𝑥) + ∅(𝑦)
𝜕𝑥
𝜕𝑧
= 𝑞 = 𝑓(𝑥) + 𝑥∅′(𝑦)
𝜕𝑦
The second order partial derivatives are:
𝜕 2𝑧
= 𝑟 = 𝑦𝑓′′(𝑥)
𝜕𝑥 2
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝜕2 𝑧
= 𝑠 = 𝑓 ′ (𝑥) + ∅′ (𝑦) and
𝜕𝑥𝜕𝑦

𝜕2 𝑧
= 𝑡 = 𝑥 ∅′′(𝑦)
𝜕𝑦 2

For the required PDE, we consider:

𝑥𝑝 + 𝑦𝑞 = 𝑥{𝑦 𝑓 ′ (𝑥) + ∅(𝑦)} + 𝑦{𝑓(𝑥) + 𝑥 ∅′ (𝑦)}

= 𝑥∅(𝑦) + 𝑦𝑓(𝑥) + 𝑥𝑦 𝑓 ′ (𝑥) + 𝑥𝑦 ∅′(𝑦)

= 𝑥∅(𝑦) + 𝑦𝑓(𝑥) + 𝑥𝑦 {𝑓 ′ (𝑥) + ∅′ (𝑦)}
= 𝑧 + 𝑥𝑦𝑠
Hence,
𝒙𝒑 + 𝒚𝒒 = 𝒛 + 𝒙𝒚𝒔 is the required PDE.

4. 𝒛 = 𝒇(𝒙) + 𝒆𝒚 𝒈(𝒙)
Solution:
Differentiating the given expression partially w.r.t x and y respectively,
𝜕𝑧
= 𝑝 = 𝑓 ′ (𝑥) + 𝑒 𝑦 𝑔′(𝑥)
𝜕𝑥
𝜕𝑧
= 𝑞 = 𝑒 𝑦 𝑔(𝑥)
𝜕𝑦
Also,
𝜕 2𝑧
= 𝑡 = 𝑒 𝑦 𝑔(𝑥)
𝜕𝑦 2
Clearly the PDE is
𝑞=𝑡 or 𝒒 − 𝒕 = 𝟎.
Note:
In this example the partial derivatives r and s are not necessary.

5. 𝒛 = 𝒇(𝒚 + 𝒙) + 𝒈(𝒚 − 𝟐𝒙)

Solution:
From the given relation,
𝜕𝑧
= 𝑝 = 𝑓 ′ (𝑦 + 𝑥) + 𝑔′(𝑦 − 2𝑥)(−2)
𝜕𝑥
= 𝑓 ′ (𝑦 + 𝑥) − 2𝑔′(𝑦 − 2𝑥)
𝜕𝑧
= 𝑞 = 𝑓 ′ (𝑦 + 𝑥) + 𝑔′(𝑦 − 2𝑥)
𝜕𝑦

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝜕 2𝑧
= 𝑟 = 𝑓 ′′ (𝑦 + 𝑥) − 2𝑔′′ (𝑦 − 2𝑥)(−2)
𝜕𝑥 2
= 𝑓 ′′ (𝑦 + 𝑥) + 4𝑔′′(𝑦 − 2𝑥)
𝜕 2𝑧
= 𝑠 = 𝑓 ′′ (𝑦 + 𝑥) + 𝑔′′ (𝑦 − 2𝑥)(−2)
𝜕𝑥𝜕𝑦
= 𝑓 ′′ (𝑦 + 𝑥) − 2𝑔′′(𝑦 − 2𝑥)
𝜕 2𝑧
= 𝑡 = 𝑓 ′′ (𝑦 + 𝑥) + 𝑔′′ (𝑦 − 2𝑥)
𝜕𝑦 2
For the PDE, we consider:
𝑟 + 𝑠 = 2𝑓 ′′ (𝑦 + 𝑥) + 2𝑔′′ (𝑦 − 2𝑥)
= 2(𝑓 ′′ (𝑦 + 𝑥) + 𝑔′′ (𝑦 − 2𝑥)) = 2𝑡
Hence,
𝒓 + 𝒔 = 𝟐𝒕 is the required PDE.
_____________________________________
Similar Example:
𝒛 = 𝒇(𝒚 + 𝒙) + 𝒈(𝒚 + 𝟐𝒙)
Answer: 𝒓 + 𝟐𝒕 = 𝟑𝒔
______________________________________
6. 𝒛 = 𝒇𝟏 (𝒚 + 𝟐𝒙) + 𝒇𝟐 (𝒚 − 𝟑𝒙)
Solution:
From the given relation,
𝜕𝑧
= 𝑝 = 𝑓1 ′(𝑦 + 2𝑥)(2) + 𝑓2 ′(𝑦 − 3𝑥)(−3)
𝜕𝑥
= 2𝑓1 ′(𝑦 + 2𝑥)−3𝑓2 ′(𝑦 − 3𝑥)
𝜕𝑧
= 𝑞 = 𝑓1 ′(𝑦 + 2𝑥) + 𝑓2 ′(𝑦 − 3𝑥)
𝜕𝑦
𝜕 2𝑧
= 𝑟 = 2𝑓1′′ (𝑦 + 2𝑥)(2) − 3𝑓2 ′′(𝑦 − 3𝑥)(−3)
𝜕𝑥 2
= 4𝑓1′′ (𝑦 + 2𝑥) + 9𝑓2 ′′(𝑦 − 3𝑥)
𝜕 2𝑧
= 𝑠 = 𝑓1′′ (𝑦 + 2𝑥)(2) + 𝑓2 ′′(𝑦 − 3𝑥)(−3)
𝜕𝑥𝜕𝑦
= 2𝑓1′′ (𝑦 + 2𝑥) − 3𝑓2 ′′(𝑦 − 3𝑥)
𝜕 2𝑧
= 𝑡 = 𝑓1 ′′(𝑦 + 2𝑥) + 𝑓2 ′′(𝑦 − 3𝑥)
𝜕𝑦 2
For the PDE, we consider
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑟 + 𝑠 = 6𝑓1 ′′(𝑦 + 2𝑥) + 6𝑓2 ′′(𝑦 − 3𝑥)

= 6(𝑓1 ′′(𝑦 + 2𝑥) + 𝑓2 ′′(𝑦 − 3𝑥)) = 6𝑡

Hence,
𝒓 + 𝒔 = 𝟔𝒕 is the required PDE.
___________________________________
Similar Example:
𝒛 = 𝒇𝟏 (𝒚 − 𝟐𝒙) + 𝒇𝟐 (𝟐𝒚 − 𝒙)
Answer: 𝟐𝒓 + 𝟓𝒔 + 𝟐𝒕 = 𝟎
___________________________________
7. 𝒛 = 𝒇(𝒙 + 𝒄𝒕) + 𝒈(𝒙 − 𝒄𝒕)
Solution:
 In this example, two functions f and g have to be eliminated
 The independent variables are x and t

Differentiating the given relation partially w.r.t x and t, we get:

𝜕𝑧
= 𝑓 ′ (𝑥 + 𝑐𝑡) + 𝑔′ (𝑥 − 𝑐𝑡)
𝜕𝑥
𝜕𝑧
= 𝑓 ′ (𝑥 + 𝑐𝑡)𝑐 + 𝑔′ (𝑥 − 𝑐𝑡)(−𝑐)
𝜕𝑡
= 𝑐{𝑓 ′ (𝑥 + 𝑐𝑡) − 𝑔′ (𝑥 − 𝑐𝑡)}
The second order partial derivatives are:
𝜕2 𝑧
= 𝑓 ′′ (𝑥 + 𝑐𝑡) + 𝑔′′ ……..(1)
𝜕𝑥 2

𝜕 2𝑧
= 𝑐{𝑓 ′′ (𝑥 + 𝑐𝑡)𝑐 + 𝑔′′(𝑥 − 𝑐𝑡)(−𝑐)}
𝜕𝑡 2
= 𝑐 2 {𝑓 ′′ (𝑥 + 𝑐𝑡) + 𝑔′′(𝑥 − 𝑐𝑡)} ........(2)
From equations (1) and (2), we immediately obtain the PDE:
𝝏𝟐 𝒛 𝟐
𝝏𝟐 𝒛
= 𝒄
𝝏𝒕𝟐 𝝏𝒙𝟐
Note:
𝝏𝟐 𝒛
In the above example notice that the partial derivative 𝝏𝒙𝝏𝒕 is not necessary

8. 𝒛 = 𝒙𝒇(𝒙 + 𝒕) + 𝒈(𝒙 + 𝒕)
Solution:
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

 As in example 7, the independent variables are x and t

Differentiating partially w.r.t x and t,
𝜕𝑧
= 𝑥𝑓 ′ (𝑥 + 𝑡) + 𝑓(𝑥 + 𝑡) + 𝑔′(𝑥 + 𝑡)
𝜕𝑥
𝜕𝑧
= 𝑥𝑓 ′ (𝑥 + 𝑡) + 𝑔′(𝑥 + 𝑡)
𝜕𝑡
The second order partial derivatives are:
𝜕 2𝑧
2
= 𝑥𝑓 ′′ (𝑥 + 𝑡) + 𝑓 ′ (𝑥 + 𝑡) + 𝑔′′(𝑥 + 𝑡)
𝜕𝑥
𝜕 2𝑧
2
= 𝑥𝑓 ′′ (𝑥 + 𝑡) + 𝑓 ′ (𝑥 + 𝑡) + 𝑓 ′ (𝑥 + 𝑡) + 𝑔′′(𝑥 + 𝑡)
𝜕𝑥
= x𝑓 ′′ (𝑥 + 𝑡) + 2𝑓 ′ (𝑥 + 𝑡) + 𝑔′′(𝑥 + 𝑡)
𝜕 2𝑧
= 𝑥𝑓 ′′ (𝑥 + 𝑡) + 𝑓 ′ (𝑥 + 𝑡) + 𝑔′′(𝑥 + 𝑡)
𝜕𝑥𝜕𝑡
𝜕 2𝑧
= 𝑥𝑓 ′′ (𝑥 + 𝑡) + 𝑔′′(𝑥 + 𝑡)
𝜕𝑡 2
Consider
𝜕 2𝑧 𝜕 2𝑧
+ = 2𝑥𝑓 ′′ (𝑥 + 𝑡) + 2𝑓′(𝑥 + 𝑡) + 2𝑔′′ (𝑥 + 𝑡)
𝜕𝑥 2 𝜕𝑡 2
𝜕 2𝑧
=2
𝜕𝑥𝜕𝑡
Therefore PDE is
𝝏𝟐 𝒛 𝝏𝟐 𝒛 𝝏𝟐 𝒛
+ 𝝏𝒕𝟐 = 𝟐 𝝏𝒙𝝏𝒕.
𝝏𝒙𝟐

Expressions of the form 𝐮 = 𝐟(𝐯) where 𝐮 𝐚𝐧𝐝 𝐯 𝐚𝐫𝐞 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧𝐬 𝐨𝐟 𝐱, 𝐲 𝐚𝐧𝐝 𝐳

AIM:
To eliminate the function 𝒇  which will result in a PDE of order 1.

Examples:
1. 𝒍𝒙 + 𝒎𝒚 + 𝒏𝒛 = 𝒇(𝒙𝟐 + 𝒚𝟐 + 𝒛𝟐 )
Solution:
Differentiating the given expression partially w.r.t x and y,
𝜕𝑧 𝜕𝑧
𝑙 + 𝑛 𝜕𝑥 = 2𝑥 + 2𝑧 𝜕𝑥

or 𝑙 + 𝑛𝑝 = 2(𝑥 + 𝑧𝑝) ……..(1)

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝜕𝑧 𝜕𝑧
𝑚 + 𝑛 𝜕𝑦 = 2𝑥 + 2𝑧 𝜕𝑦

or 𝑚 + 𝑛𝑞 = 2(𝑦 + 𝑧𝑞) ……..(2)

Dividing equations (1) and (2), we get:
𝑙 + 𝑛𝑝 2(𝑥 + 𝑧𝑝)
=
𝑚 + 𝑛𝑞 2(𝑦 + 𝑧𝑞)
Simplifying we get:
(𝑙 + 𝑛𝑝)(𝑦 + 𝑧𝑞) = (𝑚 + 𝑛𝑞)(𝑥 + 𝑧𝑝) or
𝑙𝑦 + 𝑙𝑧𝑞 + 𝑛𝑦𝑝 + 𝑛𝑧𝑝𝑞 = m𝑥 + 𝑚𝑧𝑝 + 𝑛𝑥𝑞 + 𝑛𝑧𝑝𝑞 or
𝑛𝑦𝑝 − 𝑚𝑧𝑝 + 𝑙𝑧𝑞 − 𝑛𝑥𝑞 = 𝑚𝑥 − 𝑙𝑦 or
(𝒏𝒚 − 𝒎𝒛)𝒑 + (𝒍𝒛 − 𝒏𝒙)𝒒 = (𝒎𝒙 − 𝒍𝒚) is the required PDE.

𝒙𝒚
2. 𝒛 = ∅( 𝒛 )

Solution:
Differentiating partially w.r.t x and y,
𝜕𝑧
𝜕𝑧 𝑥𝑦 𝑧−𝑥
= 𝑝 = ∅′ ( 𝑧 ) 𝑦 [ 𝜕𝑥
] ........(1)
𝜕𝑥 𝑧2

𝜕𝑧
𝜕𝑧 𝑥𝑦 𝑧−𝑦
=𝑞= ∅′ ( 𝑧 ) 𝑥 [ 𝑧 2𝜕𝑦] ……..(2)
𝜕𝑦

Dividing equation (1) and (2), we get:

𝑝 𝑦 𝑧−𝑥𝑝
= [𝑥 ] [𝑧−𝑦𝑞]
𝑞

or 𝑝𝑥(𝑧 − 𝑦𝑞) = 𝑞𝑦(𝑧 − 𝑥𝑝)

or 𝑥𝑧𝑝 − 𝑥𝑦𝑝𝑞 = 𝑦𝑧𝑞 − x𝑦𝑝𝑞
Simplification gives
𝒙𝒑 = 𝒚𝒒, which is the required PDE.

3. 𝒙𝒚𝒛 = 𝒇(𝒙 + 𝒚 + 𝒛)
Solution:
Differentiating partially w.r.t x and y,
𝜕𝑧 𝜕𝑧
𝑦 (𝑥 𝜕𝑥 + 𝑧) = 𝑓′(𝑥 + 𝑦 + 𝑧)(1 + 𝜕𝑥) ……..(1)
𝜕𝑧 𝜕𝑧
𝑥 (𝑦 𝜕𝑦 + 𝑧) = 𝑓′(𝑥 + 𝑦 + 𝑧)(1 + 𝜕𝑦) ……..(2)

Dividing equations (1) and (2)

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑦(𝑥𝑝+𝑧) (1+𝑝)
= (1+𝑞) or
𝑥(𝑦𝑞+𝑧)

𝑦(𝑥𝑝 + 𝑧)(1 + 𝑞) = 𝑥(𝑦𝑞 + 𝑧)(1 + 𝑝) or

𝑥𝑦𝑝 + 𝑥𝑦𝑝𝑞 + 𝑦𝑧 + 𝑦𝑧𝑞 = 𝑥𝑦𝑞 + 𝑥𝑦𝑝𝑞 + 𝑥𝑧 + +𝑥𝑧𝑝
On Simplification, we obtain the PDE as:
𝒙(𝒚 − 𝒛)𝒑 + 𝒚(𝒛 − 𝒙)𝒒 = 𝒛(𝒙 − 𝒚).

Expressions of the form 𝐟(𝐮, 𝐯) = 𝟎 where 𝐮 𝐚𝐧𝐝 𝐯 𝐚𝐫𝐞 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧𝐬 𝐨𝐟 𝐱, 𝐲 𝐚𝐧𝐝 𝐳

AIM:
To eliminate the function 𝒇  which will result in a PDE of order 1.
________________________________________________
The procedure is as follows:
By data, 𝑓(𝑢, 𝑣) = 0. Differentiating partially w.r.t x and y,
𝜕𝑓 𝜕𝑢 𝜕𝑓 𝜕𝑣 𝜕𝑓 𝜕𝑢 𝜕𝑓 𝜕𝑣
+ =0 or =− ……..(1)
𝜕𝑢 𝜕𝑥 𝜕𝑣 𝜕𝑥 𝜕𝑢 𝜕𝑥 𝜕𝑣 𝜕𝑥
𝜕𝑓 𝜕𝑢 𝜕𝑓 𝜕𝑣 𝜕𝑓 𝜕𝑢 𝜕𝑓 𝜕𝑣
+ =0 or =− ……..(2)
𝜕𝑢 𝜕𝑦 𝜕𝑣 𝜕𝑦 𝜕𝑢 𝜕𝑦 𝜕𝑣 𝜕𝑦

Dividing equations (1) and (2),

𝜕𝑓 𝜕𝑢 𝜕𝑓 𝜕𝑣
𝜕𝑢 𝜕𝑥 𝜕𝑣𝜕𝑥
𝜕𝑓 𝜕𝑢 = 𝜕𝑓 𝜕𝑣
𝜕𝑢𝜕𝑦 𝜕𝑣𝜕𝑦

𝝏𝒖 𝝏𝒗 𝝏𝒖 𝝏𝒗
or = ……..(3)
𝝏𝒙 𝝏𝒚 𝝏𝒚 𝝏𝒙
𝜕𝑢 𝜕𝑢 𝜕𝑣 𝜕𝑣
Substituting the four partial derivatives , , 𝑎𝑛𝑑 , we obtain the required PDE.
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦

Examples:
1. 𝒇(𝒙 + 𝒚 + 𝒛, 𝒙𝟐 + 𝒚𝟐 − 𝒛𝟐 ) = 𝟎
Solution:
Given 𝑓(𝑢, 𝑣) = 0
where 𝑢 = 𝑥 + 𝑦 + 𝑧 ……..(1)
and 𝑣 = 𝑥 2 + 𝑦 2 − 𝑧 2 ……..(2)
Repeat the derivation up to equation (3) (see procedure explained above)
We consider the expression
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= ……..(3)
𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑥

Differentiating equation (1) partially w.r.t x and y, we get:

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝜕𝑢 𝜕𝑧
= 1+ = 1 + 𝑝,
𝜕𝑥 𝜕𝑥
𝜕𝑢 𝜕𝑧
= 1+ = 1+𝑞
∂𝑦 𝜕𝑦
Differentiating equation (2) partially w.r.t x and y, we get:
𝜕𝑣 𝜕𝑧
= 2𝑥 − 2𝑧 = 2(𝑥 − 𝑧𝑝),
𝜕𝑥 𝜕𝑥
𝜕𝑣 𝜕𝑧
= 2𝑦 − 2𝑧 = 2(𝑦 − 𝑧𝑞)
𝜕𝑦 𝜕𝑦
Substituting in equation (3),
(1 + 𝑝)2(𝑦 − 𝑧𝑞) = (1 + 𝑞)2(𝑥 − 𝑧𝑝)
On simplification, we obtain:
𝑦 − 𝑧𝑞 + 𝑝𝑦 − 𝑝𝑞𝑧 = 𝑥 − 𝑧𝑝 + 𝑥𝑞 − 𝑝𝑞𝑧 or
−𝑧𝑞 + 𝑝𝑦 + 𝑧𝑝 − 𝑥𝑞 = 𝑥 − 𝑦 or
(𝒚 + 𝒛)𝒑 − (𝒛 + 𝒙)𝒒 = (𝒙 − 𝒚)
This is the required PDE.
Note:
It is convenient to express the PDE obtained in the form
𝐏𝐩 + 𝐐𝐪 = 𝐑, 𝐰𝐡𝐞𝐫𝐞 𝐏, 𝐐 𝐚𝐧𝐝 𝐑 𝐚𝐫𝐞 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧𝐬 𝐨𝐟 𝐱, 𝐲 𝐚𝐧𝐝 𝐳. This equation is called the
Lagrange’s Linear Equation.

2. 𝒇(𝒙𝒚𝒛, 𝒙𝟐 + 𝒚𝟐 − 𝟐𝒛) = 𝟎
Solution:
Given 𝑓(𝑢, 𝑣) = 0
where 𝑢 = 𝑥𝑦𝑧 ……..(1)
and 𝑣 = x 2 + 𝑦 2 − 2𝑧……..(2)
Repeat the derivation up to equation (3)
We consider the expression
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= 𝜕𝑦 𝜕𝑥 ……..(3)
𝜕𝑥 𝜕𝑦

Differentiating equation (1) partially w.r.t x and y, we get:

𝜕𝑢 𝜕𝑧
= 𝑦 (𝑥 + 𝑧) = 𝑦(𝑥𝑝 + 𝑧),
𝜕𝑥 𝜕𝑥
𝜕𝑢 𝜕𝑧
= 𝑥(𝑦 + 𝑧) = 𝑥(𝑦𝑞 + 𝑧)
𝜕𝑦 𝜕𝑦
Differentiating equation (2) partially w.r.t x and y, we get:
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝜕𝑣 𝜕𝑧
= 2𝑥 − 2 = 2(𝑥 − 𝑝),
𝜕𝑥 𝜕𝑥
𝜕𝑣 𝜕𝑧
= 2𝑦 − 2 = 2(𝑦 − 𝑞)
𝜕𝑦 𝜕𝑦
Substituting in equation (3),
𝑦(𝑥𝑝 + 𝑧)2(𝑦 − 𝑞) = 𝑥(𝑦𝑞 + 𝑧)2(𝑥 − 𝑝) or
(𝑥𝑦𝑝 + 𝑦𝑧)(𝑦 − 𝑞) = (𝑥𝑦𝑞 + 𝑥𝑧)(𝑥 − 𝑝) or
𝑥𝑦 2 𝑝 − 𝑥𝑦𝑝𝑞 + 𝑦 2 𝑧 − 𝑦𝑧𝑞 = 𝑥 2 𝑦𝑞 − 𝑥𝑦𝑝𝑞 + 𝑥 2 𝑧 − 𝑥𝑧𝑝 or
𝑥𝑦 2 𝑝 − 𝑦𝑧𝑞 − 𝑥 2 𝑦𝑞 + 𝑥𝑧𝑝 = 𝑥 2 𝑧 − 𝑦 2 𝑧
Hence the required PDE is
𝒙(𝒚𝟐 + 𝒛)𝒑 − 𝒚(𝒙𝟐 + 𝒛)𝒒 = 𝒛(𝒙𝟐 − 𝒚𝟐 )

3. 𝒇(𝒙𝟐 + 𝟐𝒚𝒛, 𝒚𝟐 + 𝟐𝒙𝒛) = 𝟎

Solution:
Given 𝑓(𝑢, 𝑣) = 0
where 𝑢 = 𝑥 2 + 2𝑦𝑧 ……..(1)
and 𝑣 = 𝑦 2 + 2𝑥𝑧 ……..(2)
Repeat the derivation up to equation (3)
We consider the expression
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= 𝜕𝑦 𝜕𝑥……..(3)
𝜕𝑥 𝜕𝑦

Differentiating equation (1) partially w.r.t x and y, we get:

𝜕𝑢 𝜕z
= 2𝑥 + 2𝑦 = 2(𝑥 + 𝑦𝑝),
𝜕𝑥 𝜕𝑥
𝜕𝑢 𝜕𝑧
= 2(𝑦 + 𝑧) = 2(𝑦𝑞 + 𝑧)
𝜕𝑦 𝜕𝑦
Differentiating equation (2) partially w.r.t x and y, we get:
𝜕𝑣 𝜕𝑧
= 2 (𝑥 + 𝑧) = 2(𝑥𝑝 + 𝑧),
𝜕𝑥 𝜕𝑥
𝜕𝑣 𝜕𝑧
= 2𝑦 + 2𝑥 = 2(𝑦 + 𝑥𝑞)
𝜕𝑦 𝜕𝑦
Substituting in equation (3),
2(𝑥 + 𝑦𝑝)2(𝑦 + 𝑥𝑞) = 2(𝑦𝑞 + 𝑧)2(𝑥𝑝 + 𝑧) or
(𝑥 + 𝑦𝑝)(𝑦 + 𝑥𝑞) = (𝑦𝑞 + 𝑧)(𝑥𝑝 + 𝑧) or
𝑥𝑦 + 𝑥 2 𝑞 + 𝑦 2 𝑝 + 𝑥𝑦𝑝𝑞 = 𝑥𝑦𝑝𝑞 + 𝑦𝑧𝑞 + 𝑥𝑧𝑝 + 𝑧 2

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

This gives
𝑥 2 𝑞 + 𝑦 2 𝑝 − y𝑧𝑞 − 𝑥𝑧𝑝 = 𝑧 2 − 𝑥𝑦 or
(𝒚𝟐 − 𝒙𝒛)𝒑 + (𝒙𝟐 − 𝒚𝒛)𝒒 = (𝒛𝟐 − 𝒙𝒚)
This is the required PDE.

𝒙
4. 𝒇 (𝒙𝟐 − 𝒙𝒚, 𝒛 ) = 𝟎

Solution:
Given 𝑓(𝑢, 𝑣) = 0
where 𝑢 = 𝑥 2 − 𝑥𝑦 ……..(1)
𝑥
and 𝑣 = 𝑧 ……..(2)

Repeat the derivation up to equation (3)

We consider the expression
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= 𝜕𝑦 𝜕𝑥……..(3)
𝜕𝑥 𝜕𝑦

Differentiating equation (1) partially w.r.t x and y, we get:

𝜕𝑢
= 2𝑥 − 𝑦,
𝜕𝑥
𝜕𝑢
= −𝑥
𝜕𝑦
Differentiating equation (2) partially w.r.t x and y, we get:
𝜕𝑧
𝜕𝑣 (𝑧 − 𝑥 𝜕𝑥) (𝑧 − 𝑥𝑝)
= = ,
𝜕𝑥 𝑧2 𝑧2
𝜕𝑣 1 𝜕𝑧 𝑥𝑞
= 𝑥 (− 2 ) = − 2
𝜕𝑦 𝑧 𝜕𝑦 𝑧
Substituting in equation (3)
𝑥𝑞 𝑧−𝑥𝑝
(2𝑥 − y) (− ) = −𝑥 ( ) or
𝑧2 𝑧2

(2𝑥 − 𝑦)(𝑥𝑞) = 𝑥(𝑧 − 𝑥𝑝) or

2𝑥 2 𝑞 − 𝑥𝑦𝑞 = 𝑥𝑧 − 𝑥 2 𝑝 or
𝑥 2 𝑝 + 2𝑥 2 𝑞 − 𝑥𝑦𝑞 = 𝑥𝑧
The required PDE is therefore
𝒙𝟐 𝒑 + 𝒙(𝟐𝒙 − 𝒚)𝒒 = 𝒙𝒛

𝒙𝒚 𝒙−𝒚
5. ∅ ( 𝒛 , )=𝟎
𝒛

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Solution:
Given ∅(𝑢, 𝑣) = 0
𝑥𝑦
where 𝑢 = ……..(1)
𝑧
𝑥−𝑦
and 𝑣 = ……..(2)
𝑧

Repeat the derivation up to equation (3) (Replace ‘f’ by ′∅′)

We consider the expression
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= 𝜕𝑦 𝜕𝑥……..(3)
𝜕𝑥 𝜕𝑦

Differentiating equation (1) partially w.r.t x and y, we get:

𝜕𝑧
𝜕𝑢 (𝑧 − 𝑥 ) (𝑧 − 𝑥𝑝)
=𝑦 𝜕𝑥 =𝑦 ,
𝜕𝑥 𝑧 2 z2
𝜕𝑧
𝜕𝑢 (𝑧 − 𝑦 ) (𝑧 − 𝑦𝑞)
𝜕𝑦
=𝑥 2
=𝑥
𝜕𝑦 𝑧 𝑧2
Differentiating equation (2) partially w.r.t x and y, we get:
𝜕𝑧
𝜕𝑣 𝑧 − (𝑥 − 𝑦) (𝜕𝑥) 𝑧 − (𝑥 − 𝑦)𝑝
= = ,
𝜕𝑥 𝑧2 𝑧2
𝜕𝑧
𝜕𝑣 𝑧(−1) − (𝑥 − 𝑦)(𝜕𝑦) −𝑧 − (𝑥 − 𝑦)𝑞
= =
𝜕𝑦 𝑧2 𝑧2
Substituting in equation (3),
(𝑧−𝑥𝑝) −𝑧−(𝑥−𝑦)𝑞 (𝑧−𝑦𝑞) 𝑧−(𝑥−𝑦)𝑝
𝑦( )( ) = 𝑥( )( ) or
𝑧2 𝑧2 𝑧2 𝑧2

(𝑦𝑧 − 𝑥𝑦𝑝)(−𝑧 − 𝑥𝑞 + 𝑦𝑞) = (𝑥𝑧 − 𝑥𝑦𝑞)(𝑧 − 𝑥𝑝 + 𝑦𝑝) or

−𝑦𝑧 2 − 𝑥𝑦𝑧𝑞 + 𝑦 2 𝑧𝑞 + 𝑥𝑦𝑧𝑝 + 𝑥 2 𝑦𝑝𝑞 − 𝑥𝑦 2 𝑝𝑞
= 𝑥𝑧 2 − 𝑥 2 𝑧𝑝 + 𝑥𝑦𝑧𝑝 − 𝑥𝑦𝑧𝑞 + 𝑥 2 𝑦𝑝𝑞 − 𝑥𝑦 2 𝑝𝑞
This gives
−𝑦𝑧 2 + 𝑦 2 𝑧𝑞 = 𝑥𝑧 2 − 𝑥 2 𝑧𝑝 or
𝑥 2 𝑧𝑝 + 𝑦 2 𝑧𝑞 = 𝑥𝑧 2 + 𝑦𝑧 2
𝑜𝑟 𝑧(𝑥 2 𝑝 + 𝑦 2 𝑞) = 𝑧 2 (𝑥 + 𝑦)
Hence the PDE is:
𝒙𝟐 𝐩 + 𝒚𝟐 𝒒 = 𝒛 (𝒙 + 𝒚)

6. ∅(𝒙𝒚 + 𝒛𝟐 , 𝒙 + 𝒚 + 𝒛) = 𝟎
Solution:
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Given ∅(𝑢, 𝑣) = 0
where 𝑢 = 𝑥𝑦 + 𝑧 2 ……..(1)
and 𝑣 = 𝑥 + 𝑦 + 𝑧……..(2)
Repeat the derivation up to equation (3)
We consider the expression
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= 𝜕𝑦 𝜕𝑥……..(3)
𝜕𝑥 𝜕𝑦

Differentiating equation (1) partially w.r.t x and y, we get:

𝜕𝑢 𝜕𝑧
= 𝑦 + 2𝑧 = 𝑦 + 2𝑧𝑝,
𝜕𝑥 𝜕𝑥
𝜕𝑢 𝜕𝑧
= 𝑥 + 2𝑧 = 𝑥 + 2𝑧𝑞
𝜕𝑦 𝜕𝑦
Differentiating equation (2) partially w.r.t x and y, we get:
𝜕𝑣 𝜕𝑧
= 1+ =1+𝑝,
𝜕𝑥 𝜕𝑥
𝜕𝑣 𝜕𝑧
=1+ = 1+𝑞
𝜕𝑦 𝜕𝑦
Substituting in equation (3),
(𝑦 + 2𝑧𝑝)(1 + 𝑞) = (𝑥 + 2𝑧𝑞)(1 + 𝑝) or
𝑦 + 𝑦𝑞 + 2𝑧𝑝 + 2𝑧𝑝𝑞 = 𝑥 + 𝑥𝑝 + 2𝑧𝑞 + 2𝑧𝑝𝑞 or
𝑦𝑞 + 2𝑧𝑝 − 𝑥𝑝 − 2𝑧𝑞 = 𝑥 − 𝑦
Therefore, the required PDE is
(𝟐𝒛 − 𝒙)𝒑 + (𝒚 − 𝟐𝒛)𝒒 = (𝒙 − 𝒚).

Exercises:
1. 𝐟(𝐱𝐲 + 𝐳 𝟐 , 𝐱 + 𝐲 + 𝐳) = 𝟎
2. ∅(𝐱 𝟐 + 𝐲 𝟐 + 𝐳 𝟐 , 𝐱 + 𝐲 + 𝐳) = 𝟎
3. 𝐟(𝐱 𝟐 + 𝐲 𝟐 , 𝐱𝐲 − 𝐳) = 𝟎
4. 𝐟(𝐱 𝟐 + 𝐲 𝟐 + 𝐳 𝟐 , 𝐲 𝟐 − 𝐳 𝟐 − 𝟐𝐲𝐳) = 𝟎
5. 𝐠(𝐱 𝟐 − 𝐲𝐳, 𝐳 𝟐 − 𝐱𝐲) = 𝟎
6. 𝐱𝐲𝐳 = 𝐟(𝐱 𝟐 + 𝐲 𝟐 − 𝟐𝐳)
𝟏 𝟏 𝟏 𝟏
7. (𝐱 − 𝐲) = ∅(𝐱 − 𝐳 )

8. (𝐱 − 𝐲) = 𝐟(𝐥𝐨𝐠𝐳 − 𝐲)
9. 𝐳 = (𝐱 + 𝐲)∅(𝐱 𝟐 − 𝐲 𝟐 )

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

10. 𝐱 = 𝐟(𝐳) + 𝐠(𝐲)

11. 𝐳 = 𝐟(𝐱 + 𝐳) + 𝐠(𝐱 + 𝐲)
12. 𝐳 = 𝐟(𝐱𝐲) + 𝐠(𝐱 + 𝐲)
𝐱
13. 𝐱𝐲 − 𝐳 𝟐 = ∅ (𝐲)
𝐲
14. 𝐟(𝐳, 𝐱 𝟐 + 𝐲 𝟐 + 𝐳 𝟐 ) = 𝟎

15. 𝐟(𝐱 − 𝐲 − 𝟐𝐳, 𝐱𝐲 𝟐 𝐳 𝟐 ) = 𝟎

Solution of Non-Homogeneous PDE’s by direct integration

If each term in a PDE contains either the dependent variable or one of its partial
derivatives, the PDE is said to be homogeneous. Otherwise, the PDE is said to be
nonhomogeneous. A solution or integral of a PDE is a relation between the dependent and
independent variables that satisfies the PDE. In this section, we consider nonhomogeneous
PDE’s and obtain their solution by direct integration method. If the independent variables in
the PDE are x and y, then, during the integration procedure, the following points should be
noted:
i. When we integrate with respect to x, we must add a function of y as arbitrary
constant, like 𝒇(𝒚), 𝒈(𝒚), 𝒉(𝒚) … . 𝒆𝒕𝒄.
ii. When we integrate with respect to y, we must add a function of x as arbitrary
constant, like 𝒇(𝒙), 𝒈(𝒙), 𝒉(𝒙) … . 𝒆𝒕𝒄.

In some cases, the arbitrary functions in the solution of the PDE can be obtained using
the giving set of conditions (which will be specified along with the PDE). This solution is a
particular solution of the PDE.
Note:
During the integration process the following method can be applied: For example
if the partial derivative is 𝒛𝒙𝒚 , then, to obtain z as the solution, we first integrate w.r.t x
keeping y constant and then integrate w.r.t y. Similarly if we have 𝒛𝒙𝒙 , we first integrate
w.r.t x keeping y constant and then integrate w.r.t x to obtain the solution z.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Examples:
𝝏𝟐 𝒛
1. Solve 𝝏𝒙𝟐 = 𝒙 − 𝟑𝒚

Solution:
The given PDE is
𝑧𝑥𝑥 = 𝑥 − 3𝑦
Integrating w.r.t x treating y as constant,
𝑥2
𝑧𝑥 = − 3𝑥𝑦 + 𝑓(𝑦)
2
Integrating again w.r.t x,
𝒙𝟑 𝒙𝟐
𝒛= −𝟑 𝒚 + 𝒙𝒇(𝒚) + 𝒈(𝒚).
𝟔 𝟐

This is the required solution.

𝝏𝟐 𝒛
2. Solve 𝝏𝒙𝝏𝒚 = 𝒔𝒊𝒏(𝟐𝒙 − 𝟑𝒚) + 𝟏𝟖𝒙𝟐 𝒚

Solution:
The given PDE is
𝑧𝑥𝑦 = 𝑠𝑖𝑛(2𝑥 − 3𝑦) + 18𝑥 2 𝑦
Integrating w.r.t x treating y as constant,
−𝑐𝑜𝑠(2𝑥 − 3𝑦) 𝑥3
𝑧𝑦 = + 18 𝑦 + 𝑓(𝑦)
2 3
−𝑐𝑜𝑠(2𝑥 − 3𝑦)
→ 𝑧𝑦 = + 6𝑥 3 𝑦 + 𝑓(𝑦)
2
Integrating w.r.t y, the solution is:
−𝑠𝑖𝑛(2𝑥−3𝑦) 𝑦2
𝑧= + 6𝑥 3 + ∫ 𝑓(𝑦)𝑑𝑦 + 𝑔(𝑥) or
2(−3) 2
𝒔𝒊𝒏(𝟐𝒙−𝟑𝒚)
𝒛= + 𝟑𝒙𝟑 𝒚𝟐 + ∫ 𝒇(𝒚)𝒅𝒚 + 𝒈(𝒙).
𝟔

This is the solution to the given PDE.

𝝏𝟐 𝒛 𝝏𝒛
3. Solve = 𝒔𝒊𝒏𝒙𝒔𝒊𝒏𝒚 subject to the conditions = −𝟐𝒔𝒊𝒏𝒚 when 𝒙 = 𝟎, and 𝒛 = 𝟎
𝝏𝒙𝝏𝒚 𝛛𝒚
𝝅
when y is an odd multiple of 𝟐 .

Solution:
The given PDE is
𝑧𝑥𝑦 = 𝑠𝑖𝑛𝑥𝑠𝑖𝑛𝑦

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Integrating w.r.t x treating y as constant,

𝑧𝑦 = −𝑐𝑜𝑠𝑥𝑠𝑖𝑛𝑦 + 𝑓(𝑦) ……..(1)
Integrating w.r.t y:
𝑧 = 𝑐𝑜𝑠𝑥𝑐𝑜𝑠𝑦 + ∫ 𝑓(𝑦)𝑑𝑦 + 𝑔(𝑥) ……..(2)
Note:
 We now apply the given conditions to determine 𝒇(𝒚) and 𝒈(𝒙)
 The solution will be a particular solution of the given PDE
Consider equation (1):
𝑧𝑦 = −𝑐𝑜𝑠𝑥𝑠𝑖𝑛𝑦 + 𝑓(𝑦)
𝜕𝑧
Given that = −2𝑠𝑖𝑛𝑦 when 𝑥 = 0. 𝑖. 𝑒. , 𝑧𝑦 = −2𝑠𝑖𝑛𝑦 𝑤ℎ𝑒𝑛 𝑥 = 0.
𝜕𝑦

Hence equation (1) becomes:

−2𝑠𝑖𝑛𝑦 = −𝑠𝑖𝑛𝑦 + 𝑓(𝑦)
This gives:
𝑓(𝑦) = −𝑠𝑖𝑛𝑦
Substituting in (2), we get:
𝑧 = 𝑐𝑜𝑠𝑥𝑐𝑜𝑠𝑦 + ∫ −𝑠𝑖𝑛𝑦𝑑𝑦 + 𝑔(𝑥) or
𝑧 = 𝑐𝑜𝑠𝑥𝑐𝑜𝑠𝑦 + 𝑐𝑜𝑠𝑦 + 𝑔(𝑥) ……..(3)
Now, applying the condition:
𝜋 𝜋
𝑧 = 0 𝑤ℎ𝑒𝑛 𝑦 𝑖𝑠 𝑎𝑛 𝑜𝑑𝑑 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑒 𝑜𝑓 , 𝑖. 𝑒. 𝑦 = (2𝑛 + 1) 2 on (3), we get:
2

𝜋 𝜋
0 = 𝑐𝑜𝑠𝑥𝑐𝑜𝑠 ((2𝑛 + 1) ) + 𝑐𝑜𝑠 ((2𝑛 + 1) ) + 𝑔(𝑥)
2 2
𝜋
But since 𝑐𝑜𝑠 ((2𝑛 + 1) 2 ) = 0, we obtain

𝑔(𝑥) = 0
Substituting in (3), the solution to the PDE (particular solution) is:
𝒛 = 𝒄𝒐𝒔𝒙𝒄𝒐𝒔𝒚 + 𝒄𝒐𝒔𝒚
or
𝒛 = 𝒄𝒐𝒔𝒚(𝒄𝒐𝒔𝒙 + 𝟏).

𝝏𝟐 𝒛 𝝏𝒛
4. Solve 𝝏𝒙𝟐 = 𝒙𝒚, subject to the conditions 𝒛(𝟎, 𝒚) = 𝟎 and 𝝏𝒙 (𝟏, 𝒚) = 𝒍𝒐𝒈(𝟏 + 𝒚).

Solution:
The given PDE is

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑧𝑥𝑥 = 𝑥𝑦
Integrating w.r.t x treating y as constant,
𝑥2
𝑧𝑥 = 𝑦 + 𝑓(𝑦) ……..(1)
2

Integrating w.r.t x again:

𝑥3
𝑧= 𝑦 + 𝑥𝑓(𝑦) + 𝑔(𝑦) ……..(2)
6
𝜕𝑧 𝝏𝒛
By data, 𝜕𝑥 (1, 𝑦) = 𝑙𝑜𝑔(1 + 𝑦) which means 𝝏𝒙 = 𝒛𝒙 = 𝒍𝒐𝒈(𝟏 + 𝒚) 𝒘𝒉𝒆𝒏 𝒙 = 𝟏.

Hence equation (1) becomes:

𝑦
𝑙𝑜𝑔(1 + y) = + 𝑓(𝑦)
2
This gives:
𝑦
𝑓(𝑦) = 𝑙𝑜𝑔(1 + 𝑦) −
2
Substituting in equation (2) we get:
𝑥3 𝑦
𝑧= 𝑦 + 𝑥{𝑙𝑜𝑔(1 + 𝑦) − 2} + 𝑔(𝑦) ……..(3)
6

Again by data, 𝑧(0, 𝑦) = 0 which means 𝒛 = 𝟎 𝒘𝒉𝒆𝒏 𝒙 = 𝟎.

Hence equation (3) becomes:
0 = 0 + 0 + 𝑔(𝑦)
This gives
𝑔(𝑦) = 0
Substituting in (3), the required solution is:
𝒙𝟑 𝒚
𝒛= 𝒚 + 𝒙 {𝒍𝒐𝒈(𝟏 + 𝒚) − }.
𝟔 𝟐

𝝏𝟐 𝒛 𝒙 𝝏𝒛
5. Solve = 𝒚, subject to the conditions = 𝒍𝒐𝒈𝒆 𝒙 when 𝒚 = 𝟏 and 𝒛 = 𝟎 when
𝝏𝒙𝝏𝒚 𝝏𝒙

𝐱 = 𝟏.
Solution:
The given PDE is
𝒙
𝒛𝒙𝒚 = 𝒚 ……..(1)

Note:
 𝒛𝒙𝒚 on integration w.r.t x, keeping y as constant, will give 𝒛𝒚
 Later, integration w.r.t y will give the solution z
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

 BUT, NOTICE THAT the first integration w.r.t x, which gives 𝒛𝒚 IS NOT
𝝏𝒛
COMPATIBLE WITH THE INITIAL CONDITION 𝝏𝒙 = 𝒍𝒐𝒈𝒆 𝒙 when 𝒚 = 𝟏

Hence, to solve this example we need to use the fact that:

𝒛𝒙𝒚 =𝒛𝒚𝒙
_______________________________
Now, from equation (1),
𝑥
𝑧𝑦𝑥 =
𝑦
Integrating w.r.t y treating x constant, we get:
𝑧𝑥 = 𝑥𝑙𝑜𝑔𝑦 + 𝑓(𝑥) ……..(2)
Integrating (2) w.r.t x,
𝑥2
𝑧= 𝑙𝑜𝑔𝑦 + ∫ 𝑓(𝑥)𝑑𝑥 + 𝑔(𝑦) ……..(3)
2

Applying the condition

𝜕𝑧
= 𝑙𝑜𝑔𝑒 𝑥 when 𝑦 = 1 on equation (2),
𝜕𝑥

𝑙𝑜𝑔𝑥 = 𝑥𝑙𝑜𝑔1 + 𝑓(𝑥)

But since 𝑙𝑜𝑔1 = 0, we get
𝑓(𝑥) = 𝑙𝑜𝑔𝑥
Substituting in equation (3) we get:
𝑥2
𝑧= 𝑙𝑜𝑔𝑦 + ∫ 𝑙𝑜g𝑥𝑑𝑥 + 𝑔(𝑦) or
2
𝑥2
𝑧= 2
𝑙𝑜𝑔𝑦 + (𝑥𝑙𝑜𝑔𝑥 − 𝑥) + 𝑔(𝑦) ……..(4)

Now, applying the condition 𝑧 = 0 𝑤ℎ𝑒𝑛 𝑥 = 1, equation (4) becomes:

1
0 = 2 𝑙𝑜𝑔𝑦 − 1 + 𝑔(𝑦) or
1
𝑔(𝑦) = 1 − 𝑙𝑜𝑔𝑦
2
Substituting in equation (4), the solution to the PDE is:
𝑥2 1
𝑧= 𝑙𝑜𝑔𝑦 + (𝑥𝑙𝑜𝑔𝑥 − 𝑥) + 1 − 2 𝑙𝑜𝑔𝑦 or
2

(𝒙𝟐 − 𝟏)
𝒛= 𝒍𝒐𝒈𝒚 + 𝒙𝒍𝒐𝒈𝒙 − 𝒙 + 𝟏.
𝟐

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝝏𝟐 𝒛 𝝏𝒛
6. Solve = 𝒆−𝒕 𝒄𝒐𝒔𝒙 subject to the conditions 𝒛(𝒙, 𝟎) = 𝟎 and (𝟎, 𝒕) = 𝟎. Also show
𝝏𝒙𝝏𝒕 𝝏𝒕

that 𝒛 → 𝒔𝒊𝒏𝒙 𝐚𝒔 𝒕 → ∞.
Solution:
Note:
In this example the independent variables are x and t.
The given PDE is
𝑧𝑥𝑡 = 𝑒 −𝑡 𝑐𝑜𝑠𝑥
Integrating w.r.t x treating t as constant,
𝑧𝑡 = 𝑒 −𝑡 𝑠𝑖𝑛𝑥 + 𝑓(𝑡) ……..(1)
Integrating w.r.t t again:
𝑧 = −𝑒 −𝑡 𝑠𝑖𝑛𝑥 + ∫ 𝑓(𝑡)𝑑𝑡 + 𝑔(𝑥) ……..(2)
𝜕𝑧 𝜕𝑧
But (0, 𝑡) = 0 which means = 0 when 𝑥 = 0
𝜕𝑡 𝜕𝑡

Using this condition equation (1) becomes

0 = 0 + 𝑓(𝑡)
This gives
𝑓(𝑡) = 0
Substituting this in (2) we get:
𝑧 = −𝑒 −𝑡 𝑠𝑖𝑛𝑥 + 𝑔(𝑥) ……..(3)
Now applying the condition 𝑧(𝑥, 0) = 0 which means 𝑧 = 0 𝑤ℎ𝑒𝑛 𝑡 = 0,
equation (3) becomes:
0 = −𝑠𝑖𝑛𝑥 + 𝑔(𝑥) 𝑜𝑟
𝑔(𝑥) = 𝑠i𝑛𝑥
Substituting in (1), the solution is:
𝒛 = −𝒆−𝒕 𝒔𝒊𝒏𝒙 + 𝒔𝒊𝒏𝒙

Finally, as 𝒕 → ∞, 𝒔𝒊𝒏𝒄𝒆 𝒆−𝒕 → 𝟎, 𝒘𝒆 𝒊𝒎𝒎𝒆𝒅𝒊𝒂𝒕𝒆𝒍𝒚 𝒐𝒃𝒕𝒂𝒊𝒏 𝒛 → 𝒔𝒊𝒏𝒙.

𝝏𝟑 𝒛
7. Solve 𝝏𝒙𝟐 𝝏𝒚 = 𝒄𝒐𝒔(𝟑𝒙 + 𝟐𝒚)

Solution:
The given PDE is
𝑧𝑥𝑥𝑦 = 𝑐𝑜𝑠(3𝑥 + 2𝑦)

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Integrating w.r.t x treating y as constant,

𝑠𝑖𝑛(3𝑥 + 2𝑦)
𝑧𝑥𝑦 = + 𝑓(𝑦)
3
Integrating w.r.t x again treating y as constant,
−𝑐𝑜𝑠(3𝑥 + 2𝑦)
𝑧𝑦 = + 𝑥𝑓(𝑦) + 𝑔(𝑦)
9
Finally, integrating w.r.t y, the solution is:
−𝒔𝒊𝒏(𝟑𝒙 + 𝟐𝐲)
𝒛= + 𝒙 ∫ 𝒇(𝒚)𝒅𝒚 + ∫ 𝒈(𝒚)𝒅𝒚 + 𝒉(𝒙).
𝟏𝟖

Exercises:
𝛛𝟐 𝐳 𝛛𝐳
1. Solve 𝛛𝐱𝛛𝐲 = 𝐬𝐢𝐧𝐱𝐜𝐨𝐬𝐲 subject to the conditions (i) 𝛛𝐲 = −𝟐𝐜𝐨𝐬𝐲 when 𝐱 = 𝟎 and

(ii) 𝐳 = 𝟎 𝐰𝐡𝐞𝐧 𝐲 = 𝐧𝛑
𝛛𝟑 𝐳
2. Solve + 𝟒𝐱 𝟐 𝐲 𝟐 − 𝐬𝐢𝐧(𝟐𝐱 − 𝐲) = 𝟎
𝛛𝐱𝛛𝐲 𝟐

𝛛𝟐 𝐳
3. Solve 𝛛𝐲𝟐 = 𝟐𝐜𝐨𝐬(𝟑𝐱 − 𝐲) + 𝟑𝐬𝐢𝐧(𝐱 − 𝟐𝐲)

𝛛𝟐 𝐮
4. Solve = 𝐞−𝟐𝐭 𝐜𝐨𝐬𝟑𝐱 subject to the conditions (i) 𝐮(𝐱, 𝟎) = 𝟎 and
𝛛𝐱𝛛𝐭

(ii) 𝐮𝐭 (𝟎, 𝐭) = 𝟎
𝛛𝟑 𝐮
5. Solve 𝛛𝐱𝟐 𝛛𝐲 = 𝐞−𝟑𝐱 𝐬𝐢𝐧𝟑𝐲

𝛛𝟐 𝐮
6. Solve 𝛛𝐲𝛛𝐱 = 𝐬𝐢𝐧(𝐱 + 𝟐𝐲) − 𝐜𝐨𝐬(𝟐𝐱 + 𝐲)

𝛛𝟑 𝐮
7. Solve 𝛛𝐱𝟐 𝛛𝐭 = 𝐞−𝟑𝐱 𝐒𝐢𝐧(𝟐𝐭 − 𝟑)

8. Solve 𝐳𝐲𝐱 = 𝐜𝐨𝐬𝟐𝐱𝐬𝐢𝐧𝟑𝐲 subject to the conditions 𝐳𝐲 = 𝟑𝐜𝐨𝐬𝐲 when 𝐱 = 𝟎 and

𝛑
𝐳 = 𝟎 when 𝐲 = 𝟐.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Solution of Homogeneous PDE’s involving derivatives with respect to one

independent variable only

We recall that if each term in a PDE contains either the dependent variable or one of its
partial derivatives, the PDE is said to be homogeneous. Otherwise, the PDE is said to be
nonhomogeneous. Also, a solution or integral of a PDE is a relation between the dependent and
independent variables that satisfies the PDE. In this section, we consider homogeneous PDE’s
involving the derivatives with respect to one independent variable only.
If the dependent variable has been differentiated partially with respect to one
independent variable only (say z is differentiated with respect to x only), then the PDE can be
treated as an ordinary differential equation (ODE).
 We are familiar with the solution of a linear differential equation with constant
coefficients (ODE). These equations as we know can be either homogeneous or
nonhomogeneous.
 For homogeneous equations we know that the complimentary function (C.F) is
sufficient and the general solution is
Dependent variable = C.F
 For nonhomogeneous equations the solution has two components, the C.F and
Particular Integral (P.I). For such equations the general solution is
Dependent variable = C.F + P.I

The construction of the C.F in a linear differential equation with constant coefficients
requires the Auxillary Equation (A.E) and this A.E can be written by replacing the derivatives
in the equation by m.
For example, if the homogeneous ODE is
𝑦 ′′ − 3𝑦 ′ + 2𝑦 = 0,
then, the A.E is
𝑚2 − 3𝑚 + 2 = 0.
Now, consider the homogeneous PDE
𝜕2 𝑧
+ 𝑧 = 0.
𝜕𝑥 2

This PDE can be viewed as

𝑧 ′′ + 𝑧 = 0.
The A.E can be therefore written as
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑚2 + 1 = 0
Similarly, the homogeneous PDE
𝜕3 𝑧 𝜕𝑧
+4 =0
𝜕𝑦 3 𝜕𝑦

which can be viewed as

𝑧 ′′′ + 4𝑧 ′ = 0
has it’s A.E. as
𝑚3 + 4𝑚 = 0.

Method of Solution:
1. Identify the dependent variable and independent variable
2. Write the A.E by writing the given PDE in its equivalent form
3. Find the roots of the A.E (These roots may be real and distinct, real and
equal(coincident) or imaginary)
4. Based on the roots of the A.E, construct the C.F
Note:
WHEN WRITING THE C.F, THE FOLLOWING POINTS MUST BE
REMEMBERED:
1. THE VARIABLE IN THE C.F MUST BE THE INDEPENDENT VARIABLE
2. THE CONSTANTS IN THE C.F WILL BE REPLACED BY FUNCTIONS OF
THE OTHER INDEPENDENT VARIABLE (RECALL THAT WE WRITE
CONSTANTS IN THE C.F. OF A LINEAR DIFFERENTIAL EQUATION
WITH CONSTANT COEFFICIENTS)

For example, consider the homogeneous PDE

𝜕 2𝑧
+ 9𝑧 = 0
𝜕𝑥 2
This PDE can be viewed as
𝑧 ′′ + 9𝑧 = 0
The A.E is
𝑚2 + 9 = 0
The roots of this A.E are
𝑚 = ±3𝑖

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

For an ODE, since the independent variable is x, for such roots (complex roots),
the C.F is
𝐶. 𝐹 = 𝐶1 𝑐𝑜𝑠3𝑥 + 𝐶2 𝑠𝑖𝑛3𝑥
But, since the constants should be functions of y the C.F. is
𝐶. 𝐹 = 𝑓(𝑦)𝑐𝑜𝑠3𝑥 + 𝑔(𝑦)𝑠𝑖𝑛3𝑥
where 𝑓(𝑦) and 𝑔(𝑦) are functions of y.

𝜕2 𝑧 𝜕𝑧
Similarly, consider the PDE: 𝜕𝑦 2 + 3 𝜕𝑦 + 2𝑧 = 0

This PDE can be viewed as

𝑧 ′′ + 3𝑧 ′ + 2𝑧 = 0
The A.E is
𝑚2 + 3𝑚 + 2 = 0
The roots of this A.E are
𝑚 = −1, −2
For an ODE, since the independent variable is y, the C.F for such roots (real and
distinct) is
𝐶. 𝐹 = 𝐶1 𝑒 −𝑦 + 𝐶2 𝑒 −2𝑦
But, since the constants should be functions of x the C.F. is
𝐶. 𝐹 = 𝑓(𝑥)𝑒 −𝑦 + 𝑔(𝑥)𝑒 −2𝑦
where 𝑓(𝑥) and 𝑔(𝑥) are functions of x.

Another example is the homogeneous PDE

𝜕 3𝑧 𝜕 2 𝑧 𝜕𝑧
+ 2 + =0
𝜕𝑥 3 𝜕𝑥 2 𝜕𝑥
In this PDE, the independent variable is x. Hence constants in C.F should be functions of y.
This PDE is equivalent to
𝑧 ′′′ + 2𝑧 ′′ + 𝑧 ′ = 0
The A.E is
𝑚3 + 2𝑚2 + 𝑚 = 0
𝑜𝑟 𝑚(𝑚2 + 2𝑚 + 1) = 0
The roots of A.E are
𝑚 = 0, −1, −1
The C.F is therefore

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑪. 𝑭 = 𝒇(𝒚)𝒆𝟎𝒙 + [𝒈(𝒚) + 𝒉(𝒚)𝒙]𝒆−𝒙

or
𝑪. 𝑭 = 𝒇(𝒚) + [𝒈(𝒚) + 𝒉(𝒚)𝒙]𝒆−𝒙
Note that for the roots 𝑚 = 0, −1, −1, the C.F for a linear differential equation with constant
coefficients is
𝐶. 𝐹 = 𝐶1 𝑒 0𝑥 + (𝐶2 + 𝐶3 𝑥)𝑒 −𝑥
or
𝐶. 𝐹 = 𝐶1 + (𝐶2 + 𝐶3 𝑥)𝑒 −𝑥

Therefore, if z is the dependent variable and the independent variables are x and y,
1. If the PDE has independent variable x, then WRITE C.F IN THE VARIABLE x,
WRITE THE CONSTANTS AS FUNCTIONS OF y and the solution is
DEPENDENT VARIABLE = C.F
2. If the PDE has independent variable y, then WRITE C.F IN THE VARIABLE y,
WRITE THE CONSTANTS AS FUNCTIONS OF x and the solution is
DEPENDENT VARIABLE = C.F

Examples:
𝝏𝟐 𝒛 𝝏𝒛
1. Solve 𝝏𝒙𝟐 + 𝒛 = 𝟎 given that when 𝒙 = 𝟎, 𝒛 = 𝒆𝒚 𝒂𝒏𝒅 = 𝟏.
𝝏𝒙

Solution:
 In this example, the dependent variable is z
 The independent variable is x
 Hence, the constants in the C.F will be functions of y

The given PDE is equivalent to

𝑧 ′′ + 𝑧 = 0
A.E is
𝑚2 + 1 = 0 𝑜𝑟
𝑚2 = −1 𝑜𝑟
𝑚 = ±𝑖
Hence,
𝐶. 𝐹 = 𝑓(𝑦)𝑐𝑜𝑠𝑥 + 𝑔(𝑦)𝑠𝑖𝑛𝑥

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Therefore, the solution to the PDE is:

𝑧 = 𝑓(𝑦)𝑐𝑜𝑠𝑥 + 𝑔(𝑦)𝑠𝑖𝑛𝑥 ……..(1)
But, given 𝑧 = 𝑒 𝑦 𝑤ℎ𝑒𝑛 𝑥 = 0. Therefore equation (1) becomes:
𝑒 𝑦 = 𝑓(𝑦)
𝜕𝑧
Also from the given data, 𝜕𝑥 = 1 𝑤ℎ𝑒𝑛 𝑥 = 0 ……..(2)

From equation (1):

𝜕𝑧
= −𝑓(𝑦)𝑠𝑖𝑛𝑥 + 𝑔(𝑦)𝑐𝑜𝑠𝑥
𝜕𝑥
Applying the condition in (2), we get:
1 = 𝑔(𝑦)
Substituting the values of 𝑓(𝑦) and 𝑔(𝑦) in (1), the solution to the PDE (particular solution)
is 𝒛 = 𝒆𝒚 𝒄𝒐𝒔𝒙 + 𝒔𝒊𝒏𝒙.

𝝏𝟐 𝒛 𝝏𝒛 𝝏𝒛
2. Solve 𝝏𝒚𝟐 + 𝟑 𝝏𝒚 − 𝟒𝒛 = 𝟎 given that when 𝒚 = 𝟎, 𝒛 = 𝟏 𝒂𝒏𝒅 = 𝒙.
𝝏𝒚

Solution:
 In this example, the dependent variable is z
 The independent variable is y
 Hence, the constants in the C.F will be functions of x

The given PDE is equivalent to

𝑧 ′′ + 3𝑧 ′ − 4𝑧 = 0
A.E is
𝑚2 + 3𝑚 − 4 = 0 𝑜𝑟
(𝑚 + 4)(𝑚 − 1) = 0 𝑜𝑟
𝑚 = −4, 1
Hence,
𝐶. 𝐹 = 𝑓(𝑥)𝑒 −4𝑦 + 𝑔(𝑥)𝑒 𝑦
Therefore, the solution to the PDE is:
𝑧 = 𝑓(𝑥)𝑒 −4𝑦 + 𝑔(𝑥)𝑒 𝑦 ……..(1)
But, given 𝑧 = 1 𝑤ℎ𝑒𝑛 𝑦 = 0. Therefore equation (1) becomes:
1 = 𝑓(𝑥) + 𝑔(𝑥) ……..(2)
𝜕𝑧
Also from the given data, 𝜕𝑦 = 𝑥 𝑤ℎ𝑒𝑛 𝑦 = 0 ……..(3)

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

From equation (1):

𝜕𝑧
= −4𝑓(𝑥)𝑒 −4𝑦 + 𝑔(𝑥)𝑒 𝑦
𝜕𝑦
Applying the condition in (3), we get:
𝑥 = −4𝑓(𝑥) + 𝑔(𝑥) ……..(4)
Solving equations (2) and (4) we get:
1−𝑥
𝑓(𝑥) =
5
and
𝑥+4
𝑔(𝑥) =
5
Substituting in (1), the solution to the PDE is:
𝟏 − 𝒙 −𝟒𝒚 𝒙+𝟒 𝒚
𝒛=[ ]𝒆 +[ ]𝒆 .
𝟓 𝟓

𝝏𝟐 𝒛 𝝏𝒛 𝝏𝒛
3. Solve 𝝏𝒙𝟐 − 𝟐 𝝏𝒙 + 𝟐𝒛 = 𝟎 given that 𝒛 = 𝒆𝒚 𝒂𝒏𝒅 = 𝟏 𝒘𝒉𝒆𝒏 𝒙 = 𝟎.
𝝏𝒙

Solution:
 In this example, the dependent variable is z
 The independent variable is x
 Hence, the constants in the C.F will be functions of y
The given PDE is equivalent to
𝑧 ′′ − 2𝑧 ′ + 2𝑧 = 0
A.E is
𝑚2 − 2𝑚 + 2 = 0 𝑜𝑟
2 ± √4 − 8 2 ± 2𝑖
𝑚= = =1±𝑖
2 2
Hence,
𝐶. 𝐹 = 𝑒 𝑥 (𝑓(𝑦)𝑐𝑜𝑠𝑥 + 𝑔(𝑦)𝑠𝑖𝑛𝑥)
The solution to the PDE is
𝑧 = 𝑒 𝑥 (𝑓(𝑦)𝑐𝑜𝑠𝑥 + 𝑔(𝑦)𝑠𝑖𝑛𝑥) ……..(1)
But, from data given, 𝑧 = 𝑒 𝑦 𝑤ℎ𝑒𝑛 𝑥 = 0
Hence, equation (1) becomes:
𝑒 𝑦 = 𝑓(𝑦)
Again by data,

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝜕𝑧
= 1 𝑤ℎ𝑒𝑛 𝑥 = 0
𝜕𝑥
But from equation (1),
𝜕𝑧
= 𝑒 𝑥 (−𝑓(𝑦)𝑠𝑖𝑛𝑥 + 𝑔(𝑦)𝑐𝑜𝑠𝑥)+ (𝑓(𝑦)𝑐𝑜𝑠𝑥 + 𝑔(𝑦)𝑠𝑖𝑛𝑥) 𝑒 𝑥
𝜕𝑥

Applying the above condition, we get:

1 = 𝑔(𝑦) + 𝑓(𝑦)
But since
𝑓(𝑦) = 𝑒 𝑦
the equation 1 = 𝑔(𝑦) + 𝑓(𝑦)
will give 𝑔(𝑦) = 1 − 𝑓(𝑦) = 1 − e𝑦
Substituting in (1), the solution to the PDE is:
𝒛 = 𝒆𝒙 (𝒆𝒚 𝒄𝒐𝒔𝒙 + (𝟏 − 𝒆𝒚 ) 𝒔𝒊𝒏𝒙).

𝝏𝟐 𝒖
4. Solve 𝝏𝒙𝟐 = 𝒂𝟐 𝒖 satisfying the conditions 𝒖 = 𝟎 𝒂𝒏𝒅 𝒖𝒙 = 𝒂𝒔𝒊𝒏𝒚 𝒘𝒉𝒆𝒏 𝒙 = 𝟎.

Solution:
 In this example, the dependent variable is u
 The independent variable is x
 Hence, the constants in the C.F will be functions of y
 Note that the terms containing u (independent variable) must be in the LHS only.
The given PDE is
𝜕 2𝑢
= 𝑎2 𝑢
𝜕𝑥 2
This equation can be written as:
𝜕 2𝑢
− 𝑎2 𝑢 = 0
𝜕𝑥 2
which is equivalent to:
𝑢′′ − 𝑎2 𝑢 = 0
The A.E. is
𝑚2 − 𝑎2 = 0
𝒐𝒓 𝑚2 = 𝑎2 .
This will give
𝑚 = 𝑎, − 𝑎.
Hence C.F. is

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝐶. 𝐹 = 𝑓(𝑦)𝑒 𝑎𝑥 + 𝑔(𝑦)𝑒 −𝑎𝑥

The solution of the PDE is
𝑢 = 𝑓(𝑦)𝑒 𝑎𝑥 + 𝑔(𝑦)𝑒 −𝑎𝑥 ……..(1)
Applying the condition 𝑢 = 0 𝑤ℎ𝑒𝑛 𝑥 = 0, equation (1) changes to
0 = 𝑓(𝑦) + 𝑔(𝑦) ……..(2)
Also from data, 𝑢𝑥 = 𝑎𝑠𝑖𝑛𝑦 𝑤ℎ𝑒𝑛 𝑥 = 0
To apply this condition, we first find that from equation (1),
𝑢𝑥 = 𝑎𝑓(𝑦)𝑒 𝑎𝑥 − 𝑎𝑔(𝑦)𝑒 −𝑎𝑥 ……..(3)
Now, using the condition
𝑢𝑥 = 𝑎𝑠𝑖𝑛𝑦 𝑤ℎ𝑒𝑛 𝑥 = 0
equation (3) becomes:
𝑎𝑠𝑖𝑛𝑦 = 𝑎𝑓(𝑦) − 𝑎𝑔(𝑦) 𝑜𝑟
𝑠𝑖𝑛𝑦 = 𝑓(𝑦) − 𝑔(𝑦) ……..(4)
But from (2),
0 = 𝑓(𝑦) + 𝑔(𝑦)
Solving equations (2) and (4) we get:
𝑠𝑖𝑛𝑦
𝑓(𝑦) =
2
𝑎𝑛𝑑
𝑠𝑖𝑛𝑦
𝑔(𝑦) = −
2
Substituting in (1), solution is:
𝑠𝑖𝑛𝑦 𝑠𝑖𝑛𝑦
𝑢= 𝑒 𝑎𝑥 − 𝑒 −𝑎𝑥 or
2 2
𝑠𝑖𝑛𝑦
𝑢= (𝑒 𝑎𝑥 − 𝑒 −𝑎𝑥 )
2
𝑠𝑖𝑛𝑦
= 2𝑠𝑖𝑛ℎ𝑎𝑥
2
Hence the solution to the PDE is
𝒖 = 𝒔𝒊𝒏𝒚𝒔𝒊𝒏𝒉𝒂𝒙.

5. Solve 𝒛𝒚𝒚 + 𝟗𝒛 = 𝟎 satisfying the conditions 𝒛(𝒙, 𝟎) = 𝒆𝟐 and 𝒛𝒚 (𝒙, 𝟎) = −𝟏.

Solution:
 In this example, the dependent variable is z

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

 The independent variable is y

 Hence, the constants in the C.F will be functions of x
The given PDE is equivalent to
𝑧 ′′ + 9𝑧 = 0
The A.E is
𝑚2 + 9 = 0 or
𝑚2 = −9 or
𝑚 = ±3𝑖
Hence
𝐶. 𝐹 = 𝑓(𝑥)𝑐𝑜𝑠3𝑦 + 𝑔(𝑥)𝑠𝑖𝑛3𝑦
The solution of the PDE is
𝑧 = 𝑓(𝑥)𝑐𝑜𝑠3𝑦 + 𝑔(𝑥)𝑠𝑖𝑛3𝑦 ……..(1)
But 𝑧(𝑥, 0) = 𝑒 2 which means → when 𝒚 = 𝟎, 𝒛 = 𝒆𝟐
Applying this condition, (1) becomes:
𝑒 2 = 𝑓(𝑥)
Also by data, 𝑧𝑦 (𝑥, 0) = −1 which means → when 𝒚 = 𝟎, 𝒛𝒚 = −𝟏
To apply this condition, we find that from (1):
𝑧𝑦 = −3𝑓(𝑥)𝑠𝑖𝑛3𝑦 + 3𝑔(𝑥)𝑐𝑜𝑠3𝑦
so that
−1 = 3𝑔(𝑥) or
1
𝑔(𝑥) = −
3
Substituting the values of 𝑓(𝑥) 𝑎𝑛𝑑 𝑔(𝑥) in (1), solution to the PDE is:
𝟏
𝒛 = 𝒆𝟐 𝒄𝒐𝒔𝟑𝒚 − 𝒔𝒊𝒏𝟑𝒚.
𝟑

𝝏𝟑 𝒛 𝝏𝒛
6. Solve 𝝏𝒙𝟑 + 𝟒 𝝏𝒙 = 𝟎, given that when 𝒙 = 𝟎, 𝒛 = 𝟎 = 𝒛𝒙 and 𝒛𝒙𝒙 = −𝟐.

Solution:
 In this example, the dependent variable is z
 The independent variable is x
 Hence, the constants in the C.F will be functions of y
The given PDE is equivalent to
𝑧 ′′′ + 4𝑧′ = 0

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

The A.E is
𝑚3 + 4𝑚 = 0 or
𝑚(𝑚2 + 4) = 0 or
𝑚 = 0, 𝑚 = ±2𝑖
Hence
𝐶. 𝐹 = 𝑓(𝑦) + 𝑔(𝑦)𝑐𝑜𝑠2𝑥 + ℎ(𝑦)𝑠𝑖𝑛2𝑥
The solution of the PDE is
𝑧 = 𝑓(𝑦) + 𝑔(𝑦)𝑐𝑜𝑠2𝑥 + ℎ(𝑦)𝑠𝑖𝑛2𝑥……..(1)
By data, 𝑧 = 0 𝑤ℎ𝑒𝑛 𝑥 = 0. Hence
0 = 𝑓(𝑦) + 𝑔(𝑦) ……..(2)
Also by data, 𝑧𝑥 = 0 𝑤ℎ𝑒𝑛 𝑥 = 0. From (1):
𝑧𝑥 = −2𝑔(𝑦)𝑠𝑖𝑛2𝑥 + 2ℎ(𝑦)𝑐𝑜𝑠2𝑥 ……..(3)
Therefore
0 = 2ℎ(𝑦) or ℎ(𝑦) = 0
Again by data, 𝑧𝑥𝑥 = −2 𝑤ℎ𝑒𝑛 𝑥 = 0.
From (3),
𝑧𝑥𝑥 = −4𝑔(𝑦)𝑐𝑜𝑠2𝑥 − 4ℎ(𝑦)𝑠𝑖𝑛2𝑥
This gives
1
−2 = −4𝑔(𝑦) 𝑜𝑟 𝑔(𝑦) =
2
1
Now, from (2), 𝑓(𝑦) + 𝑔(𝑦) = 0. Also, ℎ(𝑦) = 0 𝑎𝑛𝑑 𝑔(𝑦) = 2.
1
This immediately gives 𝑓(𝑦) = − 2.

Substituting in equation (1) solution to the PDE is:

𝟏 𝟏 𝟏 𝟏
𝒛 = − 𝟐 + 𝟐 𝒄𝒐𝒔𝟐𝒙 = − 𝟐 (𝟏 − 𝒄𝒐𝒔𝟐𝒙) = − 𝟐 (𝟐𝒔𝒊𝒏𝟐 𝒙) = −𝒔𝒊𝒏𝟐 𝒙.

Exercises:
𝛛𝟐 𝐳 𝛛𝐳 𝛛𝐳
1. Solve 𝛛𝐱𝟐 + 𝟐 𝛛𝐱 + 𝟓𝐳 = 𝟎 given that 𝐳 = −𝟏 , 𝛛𝐱 = 𝐲 𝐰𝐡𝐞𝐧 𝐱 = 𝟎.
𝛛𝟐 𝐮
2. Solve 𝛛𝐲𝟐 + 𝟐𝟓𝐮 = 𝟎 𝐠𝐢𝐯𝐞𝐧 𝐭𝐡𝐚𝐭 𝐮(𝐱, 𝟎) = 𝟏 and 𝐮𝐲 (𝐱, 𝟎) = 𝐱 𝟐 .

3. Solve 𝐳𝐲𝐲 + 𝟓𝐳𝐲 + 𝟔𝐳 = 𝟎 𝐠𝐢𝐯𝐞𝐧 𝐭𝐡𝐚𝐭 𝐳(𝐱, 𝟎) = 𝟎 𝐚𝐧𝐝 𝐳𝐲 (𝐱, 𝟎) = 𝟐𝐱.

𝛛𝟐 𝐯 𝛛𝐯
4. Solve 𝛛𝐱𝟐 + 𝟔 𝛛𝐱 + 𝟗𝐯 = 𝟎 𝐠𝐢𝐯𝐞𝐧 𝐭𝐡𝐚𝐭 𝐯 = 𝟏, 𝐯𝐱 = 𝐞−𝐲 𝐰𝐡𝐞𝐧 𝐱 = 𝟎.
𝛛𝟑 𝐳 𝛛𝟐 𝐳 𝛛𝐳
5. Solve 𝛛𝐱𝟑 + 𝟔 𝛛𝐱𝟐 + 𝟏𝟏 𝛛𝐱 + 𝟔𝐳 = 𝟎 𝐠𝐢𝐯𝐞𝐧 𝐭𝐡𝐚𝐭 𝐳 = 𝟏, 𝐳𝐱 = −𝟐, 𝐳𝐱𝐱 = 𝟑 𝐰𝐡𝐞𝐧 𝐱 = 𝟎.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Derivation of one dimensional heat and wave equations

Partial differential equations is one of the most fundamental areas in applied analysis
and it is hard to imagine any area of science and engineering where the impact of this subject
is not felt. A number of problems in science and engineering like the vibration of strings, heat
conduction and electrostatics will lead us to partial derivatives and equations involving them,
namely, partial differential equations. In physical problems, we seek a solution of the
differential equation which satisfies some specified conditions called BOUNDARY
CONDITIONS. The differential equations together with these boundary conditions constitute
a Boundary Value Problem (BVP).
In this section, we derive two important partial differential equations called the Wave
equation and the Heat equation.

Derivation 1:
VIBRATIONS OF A STRETCHED STRING - WAVE EQUATION
Consider a tightly stretched elastic string of length 𝑙 having fixed ends 𝐴 and 𝐵 and
subjected to a constant tension T. The tension T will be considered large as compared to the
weight of the string so that the effects of gravity are negligible.
 Let the string be released from rest and allowed to vibrate
 we study the subsequent motion of the string with no external forces acting on
it. We assume that each point of the string makes small vibrations at right angles
to the equilibrium position 𝐴𝐵 of the string entirely in one plane
 we take the end 𝐴 as the origin, 𝐴𝐵 as the X-axis and 𝐴𝑌 as the Y-axis. The
motion of the string takes place in the XY-plane.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Figure shows the string in the position APB at time t. Consider the motion of the
element PQ of the string between the points 𝑃(𝑥, 𝑦) and 𝑄(𝑥 + 𝛿𝑥, 𝑦 + 𝛿𝑦). The tangents at P
and Q make angles of 𝜑 and 𝜑 + 𝛿𝜑 with the X-axis respectively.
𝜕2 𝑦
Since the element is moving upwards with acceleration , the vertical component of
𝜕𝑡 2
the force acting on this element is
= 𝑇𝑠𝑖𝑛(𝜑 + 𝜕𝜑) − 𝑇𝑠𝑖𝑛 𝜑
= 𝑇{𝑠𝑖𝑛(𝜑 + 𝜕𝜑) − 𝑠𝑖n 𝜑}
= 𝑇{𝑡𝑎𝑛(𝜑 + 𝜕𝜑) − 𝑡𝑎𝑛𝜑}, 𝑠𝑖𝑛𝑐𝑒 𝜑 𝑖𝑠 𝑠𝑚𝑎𝑙𝑙
𝜕𝑦 𝜕𝑦
= 𝑇 [( ) −( ) ]
𝜕𝑥 𝑥+𝛿𝑥 𝜕𝑥 𝑥
If 𝑚 is the mass per unit length of the string, by Newton’s second law of motion,
𝑚𝑎𝑠𝑠 × 𝑎𝑐𝑐𝑙𝑒𝑟𝑎𝑡𝑖𝑜𝑛 = 𝐹𝑜𝑟𝑐𝑒 i.e.
𝜕2 𝑦 𝜕𝑦 𝜕𝑦
(𝑚𝛿𝑥) ( ) = 𝑇 [(𝜕𝑥 ) − (𝜕𝑥 ) ] or
𝜕𝑡 2 𝑥+𝛿𝑥 𝑥

𝜕𝑦 𝜕𝑦
[( ) −( ) ]
𝜕 2𝑦 𝑇 𝜕𝑥 𝑥+𝛿𝑥 𝜕𝑥 𝑥
=
𝜕𝑡 2 𝑚 𝛿𝑥
Taking limits as 𝑄 → 𝑃 𝑖. 𝑒. , 𝛿𝑥 → 0, we obtain
𝜕 2𝑦 𝑇 𝜕 2𝑦
= , 0 < 𝑥 < 𝑙, 𝑡 > 0
𝜕𝑡 2 𝑚 𝜕𝑥 2
𝑇
Denoting 𝑐 2 = 𝑚 , we have:

𝝏𝟐 𝒚 𝟐
𝝏𝟐 𝒚
= 𝒄
𝝏𝒕𝟐 𝝏𝒙𝟐
This is the partial differential equation giving the transverse vibrations of the string. It
𝑇
is also called the one-dimensional Wave Equation. The physical constant is denoted by 𝑐 2
𝑚

(instead of 𝑐) to indicate that this constant is positive, a fact that will be essential to the form
of the solutions. “One-dimensional" means that the equation involves only one space
variable 𝑥.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Derivation 2:
ONE DIMENSIONAL HEAT FLOW - HEAT EQUATION
Consider a homogeneous bar of uniform cross-section 𝛼 𝑐𝑚2 . Suppose that the sides
are covered with a material impervious to heat so that the stream lines of heat flow are all
parallel and perpendicular to the area 𝛼. One end of the bar is taken as the origin and the
direction of heat flow is taken as the positive X-axis.

Let 𝜌 be the density, 𝑠 be the specific heat and 𝑘 be the thermal conductivity. Let 𝑢(𝑥, 𝑡)
be the temperature at a distance 𝑥 from O. If 𝛿𝑢 be the change in temperature in a slab of
thickness 𝛿𝑥 of the bar, then, from the fundamental principles of heat conduction we know that
“heat flows from a higher temperature to a lower temperature” and “the quantity of heat in a
body is proportional to its mass and temperature”.
But, the quantity of heat in the slab is = 𝑠𝜌𝛼 𝛿𝑥 𝛿𝑢
𝜕𝑢
Hence, the rate of increase of heat in this slab which is = 𝑠 𝑝 𝛼 𝛿𝑥 can be written as:
𝜕𝑡
𝜕𝑢
𝑠𝜌𝛼 𝛿𝑥 = 𝑅1 − 𝑅2 ……..(1)
𝜕𝑡

where 𝑅1 and 𝑅2 are respectively the rate of inflow and outflow of heat.
Again, from the fundamental principles of heat conduction we have the property:
“the rate of heat-flow across an area is proportional to the area and to the rate of change of
temperature with respect to its distance normal to the area”.
Hence,
𝜕𝑢
𝑅1 = −𝑘𝛼 [ ] and
𝜕𝑥 𝑥

𝜕𝑢
𝑅2 = −𝑘𝛼 [ ]
𝜕𝑥 𝑥+𝛿𝑥

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

the negative sign appears due to the fact that heat flows from a higher temperature to a lower
temperature.
Now, equation (1) takes the form:
𝜕𝑢 𝜕𝑢 𝜕𝑢
𝑠𝜌𝛼 𝛿𝑥 = −𝑘𝛼 [ ] + 𝑘𝛼 [ ] or
𝜕𝑡 𝜕𝑥 𝑥 𝜕𝑥 𝑥+𝛿𝑥

𝜕𝑢 𝜕𝑢 𝜕𝑢
𝑠𝜌𝛼 𝛿𝑥 = 𝑘𝛼 [[ ] − [ ] ] or
∂𝑡 𝜕𝑥 𝑥+𝛿𝑥 𝜕𝑥 𝑥

𝜕𝑢 𝜕𝑢
] [[ −[ ] ]
𝜕𝑢 𝑘 𝜕𝑥 𝑥+𝛿𝑥 𝜕𝑥 𝑥
=
𝜕𝑡 𝑠𝜌 𝛿𝑥
Taking the limit as 𝛿𝑥 → 0, we get:
𝜕𝑢 𝑘 𝜕 2𝑢
=
𝜕𝑡 𝑠𝜌 𝜕𝑥 2
𝑘
writing = 𝑐 2 , called the diffusivity of the substance we get
𝑠𝜌

𝝏𝒖 𝝏𝟐 𝒖
= 𝒄𝟐 𝟐
𝝏𝒕 𝝏𝒙
This is the one dimensional heat-flow equation.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

SOLUTION OF THE WAVE AND HEAT EQUATIONS BY THE

METHOD OF SEPARATION OF VARIABLES

This method is applicable for solving a linear homogeneous PDE involving the
derivatives with respect to two independent variables. The solution is obtained by solving two
ordinary differential equations.
Since both the wave equation and the heat equation have independent variables x and t,
we explain the method with these independent variables.
Method:
 Let 𝒖 be the dependent variable
 Assume the solution in the form 𝒖 = 𝑿𝑻 where 𝑿 is a function of 𝒙 and 𝑻 is a
function of 𝒕
𝝏𝒖 𝝏𝒖 𝝏𝟐 𝒖 𝝏𝟐 𝒖
 Since 𝒖 = 𝑿𝑻, 𝒘𝒆 𝒈𝒆𝒕 = 𝑿′ 𝑻, − 𝑿𝑻′ , = 𝑿′′ 𝑻 and = 𝑿𝑻′′
𝝏𝒙 𝝏𝒕 𝝏𝒙𝟐 𝝏𝒙𝟐
𝒅𝑿 𝒅𝑻 𝒅𝟐 𝑿 𝒅𝟐 𝑻
where 𝑿′ = , 𝑻′ = , 𝑿′′ = 𝟐
, 𝑻′′ =
𝒅𝒙 𝒅𝒕 𝒅𝒙 𝒅𝒕𝟐

 Substitute these values in the Wave/Heat equation

 Separate the variables (𝑿, 𝒙 terms from and 𝑻, 𝒕 terms)
 Equate both sides to a common constant say 𝒌
 Solve the resulting ODE’s and determine 𝑿 𝒂𝒏𝒅 𝑻
 Finally substitute in 𝒖 = 𝑿𝑻 for the required solution.

We solve the wave equation and the heat equation by equating the constant k to
i. Zero
ii. A positive constant (say 𝑘 = 𝑝2 )
iii. A negative constant (say 𝑘 = −𝑝2 )
Thus, we obtain three possible solutions of the Wave equation and the Heat equation.

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

1. Solution of the Wave equation by the method of separation of variables

The one dimensional wave equation is
𝜕2 𝑦 𝜕2 𝑦
= 𝑐2 ……..(1)
𝜕𝑡 2 𝜕𝑥 2
We seek a solution of the form 𝑦 = 𝑦(𝑥, 𝑡) using the method of separation of variables.
Since ‘y’ is the dependent variable and ‘x, t’ are the independent variables, we assume the
solution in the form
𝑦 = 𝑋𝑇 ……..(2)
where X is a function of x and T is a function of t.
𝜕2 𝑦 𝜕2 𝑦 𝑑2 𝑋 𝑑2 𝑇
Since 𝑦 = 𝑋𝑇, 𝑤𝑒 ℎ𝑎𝑣𝑒 = 𝑋𝑇 ′′ 𝑎𝑛𝑑 𝜕𝑥 2 = 𝑋 ′′ 𝑇 where 𝑋 ′′ = , 𝑇 ′′ =
𝜕𝑡 2 𝑑𝑥 2 𝑑𝑡 2

Substituting in equation (1), we get:

𝑋𝑇 ′′ = 𝑐 2 𝑋′′𝑇
Separating the variables,
𝑋′′ 𝑇′′
= 2
𝑋 𝑐 𝑇
Equating both sides to a common constant say k, we get:
𝑋′′ 𝑇′′
=𝑘 𝑎𝑛𝑑 =𝑘
𝑋 𝑐2𝑇
or
𝑋 ′′ = 𝑘𝑋 𝑎𝑛𝑑 𝑇 ′′ = 𝑘𝑐 2 𝑇
We obtain the two ODE’s:
𝑋 ′′ − 𝑘𝑋 = 0 … … . . (𝟑)
𝑎𝑛𝑑 𝑇 ′′ − 𝑘𝑐 2 𝑇 = 0 ……..(4)
For the various possible solutions of the wave equation, we consider the following
cases.
Case (i):
Let 𝑘 = 0.
Equation (3) takes the form
𝑋 ′′ = 0
Integrating, 𝑋 ′ = 𝐶1
Integrating again, 𝑿 = 𝑪𝟏 𝒙 + 𝑪𝟐 , which is the solution for X

Equation (4) takes the form

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑇 ′′ = 0
Integrating, 𝑇 ′ = 𝐶1 ′
Integrating again, 𝑻 = 𝑪′𝟏 𝒕 + 𝑪𝟐 ′, which is the solution for T
Substituting in equation (2), the solution for 𝑘 = 0 is:
𝒚 = (𝑪𝟏 𝒙 + 𝑪𝟐 )(𝑪′𝟏 𝒕 + 𝑪𝟐 ′) ……..(5)
Case (ii):
Let 𝑘 be positive, say 𝑘 = 𝑝2
Equation (3) takes the form
𝑋 ′′ − 𝑝2 𝑋 = 0
This is a linear homogeneous ODE.
A.E is 𝑚2 − 𝑝2 = 0 𝑜𝑟
𝑚2 = 𝑝2 𝑜𝑟
𝑚 = 𝑝, − 𝑝
𝐶. 𝐹 = 𝐶3 𝑒 𝑝𝑥 + 𝐶4 𝑒 −𝑝𝑥
Hence, the solution is 𝑋 = 𝐶. 𝐹
or 𝑿 = 𝑪𝟑 𝒆𝒑𝒙 + 𝑪𝟒 𝒆−𝒑𝒙

Equation (4) takes the form

𝑇 ′′ − 𝑝2 c 2 𝑇 = 0
This is again a linear homogeneous ODE.
A.E is 𝑚2 − 𝑝2 𝑐 2 = 0 𝑜𝑟
𝑚2 = 𝑝2 𝑐 2 𝑜𝑟
𝑚 = 𝑝𝑐, −𝑝𝑐
𝐶. 𝐹 = 𝐶3 ′𝑒 𝑝𝑐𝑡 + 𝐶4 ′𝑒 −𝑝𝑐𝑡
Hence, the solution is 𝑇 = 𝐶. 𝐹
or 𝑻 = 𝑪𝟑 ′𝒆𝒑𝒄𝒕 + 𝑪𝟒 ′𝒆−𝒑𝒄𝒕
Substituting in equation (2), the solution for 𝑘 = 𝑝2is:
𝒚 = (𝑪𝟑 𝒆𝒑𝒙 + 𝑪𝟒 𝒆−𝒑𝒙 )(𝑪𝟑 ′𝒆𝒑𝒄𝒕 + 𝑪𝟒 ′𝒆−𝒑𝒄𝒕 ) ……..(6)
Case (iii):
Let 𝑘 be negative, say 𝑘 = −𝑝2
Equation (3) takes the form
𝑋 ′′ + 𝑝2 𝑋 = 0
This is a linear homogeneous ODE.
__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

A.E is 𝑚2 + 𝑝2 = 0 𝑜𝑟
𝑚2 = −𝑝2 𝑜𝑟
𝑚 = ±𝑖𝑝
𝐶. 𝐹 = 𝐶5 𝑐𝑜𝑠𝑝𝑥 + 𝐶6 𝑠𝑖𝑛𝑝𝑥
Hence, the solution is 𝑋 = 𝐶. 𝐹
or 𝑿 = 𝑪𝟓 𝒄𝒐𝒔𝒑𝒙 + 𝑪𝟔 𝒔𝒊𝒏𝒑𝒙

Equation (4) takes the form

𝑇 ′′ + 𝑝2 𝑐 2 𝑇 = 0
This is again a linear homogeneous ODE.
A.E is 𝑚2 + 𝑝2 𝑐 2 = 0 𝑜𝑟
𝑚2 = −𝑝2 𝑐 2 𝑜𝑟
𝑚 = ±𝑖𝑝𝑐
𝐶. 𝐹 = 𝐶5′ 𝑐𝑜𝑠𝑝𝑐𝑡 + 𝐶6′ 𝑠𝑖𝑛𝑝𝑐𝑡
Hence, the solution is 𝑇 = 𝐶. 𝐹
or 𝑻 = 𝑪′𝟓 𝒄𝒐𝒔𝒑𝒄𝒕 + 𝑪′𝟔 𝒔𝒊𝒏𝒑𝒄𝒕
Substituting in equation (2), the solution for 𝑘 = −𝑝2is:
𝒚 = (𝑪𝟓 𝒄𝒐𝒔𝒑𝒙 + 𝑪𝟔 𝒔𝒊𝒏𝒑𝒙)(𝑪′𝟓 𝒄𝒐𝒔𝒑𝒄𝒕 + 𝑪′𝟔 𝒔𝒊𝒏𝒑𝒄𝒕) ……..(7)
Hence, the solution to the wave equation are the solutions obtained in equations
(5), (6) and (7).

2. Solution of the Heat equation by the method of separation of variables

The one dimensional heat equation is

𝜕𝑢 𝜕2 𝑢
= 𝑐2 ……..(1)
𝜕𝑡 𝜕𝑥 2
We seek a solution of the form 𝑢 = 𝑢(𝑥, 𝑡) using the method of separation of variables.
Since ‘u’ is the dependent variable and ‘x, t’ are the independent variables, we assume the
solution in the form
𝑢 = 𝑋𝑇 ……..(2)
where X is a function of x and T is a function of t.
𝜕𝑢 𝜕2 𝑢 𝑑2 𝑋 𝑑𝑇
Since 𝑢 = 𝑋𝑇, 𝑤𝑒 ℎ𝑎𝑣𝑒 = 𝑋𝑇 ′ 𝑎𝑛𝑑 𝜕𝑥 2 = 𝑋 ′′ 𝑇 where 𝑋 ′′ = 𝑑𝑥 2 , 𝑇 ′ =
𝜕𝑡 𝑑𝑡

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Substituting in equation (1), we get:

𝑋𝑇 ′ = 𝑐 2 𝑋′′𝑇
Separating the variables,
𝑋′′ 𝑇′
= 2
𝑋 𝑐 𝑇
Equating both sides to a common constant say k, we get:
𝑋′′ 𝑇′
=𝑘 a𝑛𝑑 =𝑘
𝑋 𝑐2𝑇
or
𝑋 ′′ = 𝑘𝑋 𝑎𝑛𝑑 𝑇 ′ = 𝑘𝑐 2 𝑇
We obtain the ODE’s:
𝑋 ′′ − 𝑘𝑋 = 0 … … . . (𝟑) 𝑎𝑛𝑑
𝑇 ′ = 𝑘𝑐 2 𝑇 ……..(4)
For the various possible solutions of the wave equation, we consider the following
cases.
Case (i):
Let 𝑘 = 0.
Equation (3) takes the form
𝑋 ′′ = 0
Integrating, 𝑋 ′ = 𝐶1
Integrating again, 𝑿 = 𝑪𝟏 𝒙 + 𝑪𝟐 , which is the solution for X.

Equation (4) takes the form

𝑇′ = 0
Integrating, 𝑻 = 𝑪𝟏 ′, which is the solution for T.

Substituting in equation (2), the solution for 𝑘 = 0 is:

𝒖 = (𝑪𝟏 𝒙 + 𝑪𝟐 )𝑪′𝟏
or
𝒖 = 𝑪𝟏 𝑪𝟏 ′𝒙 + 𝑪𝟐 𝑪′𝟏
or
𝒖 = 𝑲𝟏 𝒙 + 𝑲𝟐 ……..(5)
where 𝑲𝟏 = 𝑪𝟏 𝐂𝟏 ′ and 𝑲𝟏 = 𝑪𝟐 𝑪𝟏 ′

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

Case (ii):
Let 𝑘 be positive, say 𝑘 = 𝑝2
Equation (3) takes the form
𝑋 ′′ − 𝑝2 𝑋 = 0
This is a linear homogeneous ODE.
A.E is 𝑚2 − 𝑝2 = 0 𝑜𝑟
𝑚2 = 𝑝2 𝑜𝑟
𝑚 = 𝑝, − 𝑝
𝐶. 𝐹 = 𝐶3 𝑒 𝑝𝑥 + 𝐶4 𝑒 −𝑝𝑥
Hence, the solution is 𝑋 = 𝐶. 𝐹
or 𝑿 = 𝑪𝟑 𝒆𝒑𝒙 + 𝑪𝟒 𝒆−𝒑𝒙

Equation (4) takes the form

𝑇 ′ = 𝑝2 𝑐 2 𝑇
This gives
𝑑𝑇 𝑑𝑇
= 𝑝2 𝑐 2 𝑇 𝑜𝑟 = 𝑝2 𝑐 2 𝑑𝑡
𝑑𝑡 𝑇
On integration we get:
𝑙𝑜𝑔𝑇 = 𝑝2 𝑐 2 𝑡 + 𝐶3 ′
Hence, the solution is
2 𝑐 2 𝑡+𝐶 ′ 2𝑐 2𝑡
𝑇 = 𝑒𝑝 3 = 𝑒p 𝑒 𝐶3′
Substituting in equation (2), the solution for 𝑘 = 𝑝2 is:
𝟐 𝒄𝟐 𝒕
𝒖 = (𝑪𝟑 𝒆𝒑𝒙 + 𝑪𝟒 𝒆−𝒑𝒙 ) 𝒆𝒑 𝒆𝑪𝟑 ′ or
𝟐 𝒄𝟐 𝒕
𝒖 = (𝑪𝟑 𝒆𝑪𝟑 ′ 𝒆𝒑𝒙 + 𝑪𝟒 𝒆𝑪𝟑 ′ 𝒆−𝒑𝒙 ) 𝒆𝒑 or
𝟐 𝒄𝟐 𝒕
𝒖 = (𝑲𝟑 𝒆𝒑𝒙 + 𝑲𝟒 𝒆−𝒑𝒙 ) 𝒆𝒑 ……..(6)
where 𝑲𝟑 = 𝑪𝟑 𝒆𝑪𝟑 ′ 𝒂𝒏𝒅 𝑲𝟒 = 𝑪𝟒 𝒆𝑪𝟑 ′
Case (iii):
Let 𝑘 be negative, say 𝑘 = −𝑝2
Equation (3) takes the form
𝑋 ′′ + 𝑝2 𝑋 = 0
This is a linear homogeneous ODE.
A.E is 𝑚2 + 𝑝2 = 0 𝑜𝑟

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56
__________________________________________________________________________________15MA21-Partial Differential Equations

𝑚2 = −𝑝2 𝑜r
𝑚 = ±𝑖𝑝
𝐶. 𝐹 = 𝐶5 𝑐𝑜𝑠𝑝𝑥 + 𝐶6 𝑠𝑖𝑛𝑝𝑥
Hence, the solution is 𝑋 = 𝐶. 𝐹
or 𝑿 = 𝑪𝟓 𝒄𝒐𝒔𝒑𝒙 + 𝑪𝟔 𝒔𝒊𝒏𝒑𝒙
Equation (4) takes the form
𝑇 ′ = −𝑝2 𝑐 2 𝑇
This gives
𝑑𝑇
= −𝑝2 𝑐 2 𝑇 𝑜𝑟
𝑑𝑡
𝑑𝑇
= −𝑝2 𝑐 2 𝑑𝑡
𝑇
On integration we get:
𝑙𝑜𝑔𝑇 = −𝑝2 𝑐 2 𝑡 + 𝐶5 ′
Hence, the solution is
2 𝑐 2 𝑡+𝐶 ′
𝑇 = 𝑒 −𝑝 5

2 𝑐 2𝑡
= 𝑒 −𝑝 𝑒 𝐶5′
Substituting in equation (2), the solution for 𝑘 = −𝑝2 is:
𝟐 𝒄𝟐 𝒕
𝒖 = (𝑪𝟓 𝒄𝒐𝒔𝒑𝒙 + 𝑪𝟔 𝒔𝒊𝒏𝒑𝒙) 𝒆−𝒑 𝒆𝑪𝟓 ′ or
𝟐 𝒄𝟐 𝒕
𝒖 = (𝑪𝟓 𝒆𝑪𝟓 ′ 𝒄𝒐𝒔𝒑𝒙 + 𝑪𝟔 𝒆𝑪𝟓 ′ 𝒔𝒊𝒏𝒑𝒙) 𝒆−𝒑 or
𝟐 𝒄𝟐 𝒕
𝒖 = (𝑲𝟓 𝒄𝒐𝒔𝒑𝒙 + 𝑲𝟔 𝒔𝒊𝒏𝒑𝒙) 𝒆−𝒑 ……..(6)
where 𝑲𝟓 = 𝑪𝟓 𝒆𝑪𝟓 ′ 𝒂𝒏𝒅 𝑲𝟔 = 𝑪𝟔 𝒆𝑪𝟓 ′

Hence, the solution to the heat equation are the solutions obtained in equations
(5), (6) and (7).

********

__________________________________________________________________________________
Dr.R.Murali, Professor, Department of Mathematics,
Dr. Ambedkar Institute of Technology, Bengaluru-56