Module 4: Partial Differential Equations

Introduction: Many of the foundational theories of physics and engineering are

expressed by means of systems of partial differential equations. Fluid mechanics is often
formulated by the Euler equations of motion or the so-called Navier-Stokes equations,
electricity and magnetism by Maxwell’s equations, general relativity by Einstein’s field
equations. It is therefore important to develop techniques that can be used to solve a
wide variety of partial differential equations.
Definition: A partial differential equation (PDE) is a differential equation containing
partial derivatives (one or more) of the dependent variable with more than one
independent variable.
The order and degree of partial differential equations are defined same as for ordinary
differential equations.
Examples:
𝜕𝑧 𝜕 2 𝑧 order-2, degree-1
= ( )
𝜕𝑡 𝜕𝑥 2 𝑧-dependent & 𝑥, 𝑡-independent
order-2, degree-1
𝑧𝑥𝑥 + 𝑧𝑦 = cos 𝑥 ( )
𝑧-dependent & 𝑥, 𝑦-independent
4
𝜕 2 𝑢 𝜕𝑢 𝜕3𝑢 order-3, degree-4
+ + 𝑟 ( 3 ) = 2𝑢 ( )
𝜕𝜃 2 𝜕𝑡 𝜕𝑟 𝑢-dependent & 𝜃, 𝑡, 𝑟-independent
Linear & Non-linear PDE:
A PDE is said to be linear if the dependent variable and all its partial derivatives occur
only in the first degree and not multiplied together. A PDE which is not linear is called
non-linear PDE.
Homogeneous & Non-homogeneous PDE:
If all the terms of PDE contains either the dependent variable or its partial derivatives
then such PDE is called homogeneous PDE otherwise it is called non-homogeneous
PDE.
Examples:
𝜕2𝑧 𝜕2𝑧
−𝑐 2 =𝑧 (linear and homogeneous)
𝜕𝑡 2 𝜕𝑥
𝑧𝑥𝑦 + 𝑧𝑦 = 2𝑥 (linear and non-homogeneous)
𝜕2𝑧 𝜕𝑧 2 𝜕3𝑧
+ ( ) + 𝑟 3 = 2𝑧 (non-linear and homogeneous)
𝜕𝜃 2 𝜕𝑡 𝜕𝑟
2
𝑢𝑥𝑥 + (𝑢𝑦𝑦 ) = 2𝑦 (non-linear and non-homogeneous)
𝜕 2 𝑧 𝜕𝑧 𝜕3𝑧 𝜕𝑧
− + 𝑟 − 𝑧 =0 (non-linear and homogeneous)
𝜕𝜃 2 𝜕𝑡 𝜕𝑟 3 𝜕𝑟
2 2)
𝜕2𝑢 𝜕2𝑢
(𝑥 + 𝑦 − = 2 sin 𝑥 (linear and non-homogeneous)
𝜕𝑥 2 𝜕𝑥𝜕𝑦
Notations: When 𝑧 = 𝑓(𝑥, 𝑦), we use the following notations
𝜕𝑧
= 𝑧𝑥 = 𝑝
𝜕𝑥
𝜕𝑧
= 𝑧𝑦 = 𝑞
𝜕𝑦
𝜕2𝑧
= 𝑧𝑥𝑥 = 𝑟
𝜕𝑥 2
𝜕2𝑧
= 𝑧𝑥𝑦 = 𝑠
𝜕𝑥𝜕𝑦
𝜕2𝑧
= 𝑧𝑦𝑦 = 𝑡
𝜕𝑦 2
𝜕2𝑧 𝜕2𝑧
NOTE: In PDE = or 𝑧𝑥𝑦 = 𝑧𝑦𝑥
𝜕𝑥𝜕𝑦 𝜕𝑦𝜕𝑥
 4.1: Formation of Partial Differential Equations
4.1.1: Method of Elimination of Arbitrary Constants
Let 𝑓(𝑥, 𝑦, 𝑧, 𝑎, 𝑏) = 0 be an equation where 𝑧 is a function of 𝑥, 𝑦 and 𝑎, 𝑏 are arbitrary
constants. We differentiate the given equation with respect to 𝑥 & 𝑦 partially and
eliminate the arbitrary constants 𝑎, 𝑏 to form PDE.
If the number of arbitrary constants is more than the number of independent variables
then PDE of second or higher order arise.
PROBLEMS
1) Form the PDE by eliminating the arbitrary constants 𝑎 and 𝑏 from the following
𝑧 = 𝑎𝑥 + 𝑏𝑦 + 𝑎2 + 𝑏 2
𝜕𝑧 𝜕𝑧 𝜕𝑧 2 𝜕𝑧 2
Answer: 𝑧 = 𝑝𝑥 + 𝑞𝑦 + 𝑝2 + 𝑞2 Or 𝑧 = 𝑥 +𝑦 +( ) +( )
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦

2) Form the PDE by eliminating the arbitrary constants 𝑎, 𝑏 and 𝑐 from

𝑧 = 𝑎𝑥 + 𝑏𝑦 + 𝑐𝑥𝑦
Answer: 𝑧 = (𝑝 − 𝑠𝑦 )𝑥 + (𝑞 − 𝑠𝑥 )𝑦 + 𝑠𝑥𝑦 Or 𝑧 = 𝑥𝑝 + 𝑦𝑞 − 𝑥𝑦𝑠
𝜕𝑧 𝜕𝑧 𝜕2 𝑧
Or 𝑧 = 𝑥 +𝑦 − 𝑥𝑦
𝜕𝑥 𝜕𝑦 𝜕𝑥𝜕𝑦

4.1.2: Method of Elimination of Arbitrary Functions

Suppose 𝑧 is a function of two arbitrary functions, we need to find partial derivatives up
to second order to form the PDE by eliminating the arbitrary functions.
PROBLEMS
Form the PDE by eliminating the arbitrary functions from the following
1) 𝑧 = 𝑓(𝑥 2 − 𝑦 2 )
𝜕𝑧 𝜕𝑧
Answer: 𝑦𝑝 + 𝑥𝑞 = 0 Or 𝑦 +𝑥 =0
𝜕𝑥 𝜕𝑦

2) 𝑧 = 𝑒 𝑚𝑦 𝜙(𝑥 − 𝑦)
𝜕𝑧 𝜕𝑧
Answer: 𝑝 + 𝑞 = 𝑚𝑧 Or + = 𝑚𝑧
𝜕𝑥 𝜕𝑦

3) 𝑧 = 𝑓 (𝑥 ) + 𝑒 𝑦 𝑔(𝑥)
𝜕𝑧 𝜕2 𝑧
Answer: 𝑞 − 𝑡 = 0 Or − =0
𝜕𝑦 𝜕𝑦 2

4) 𝑧 = 𝑦𝑓 (𝑥 ) + 𝑥𝑔(𝑦)
𝜕2 𝑧 𝜕𝑧 𝜕𝑧
Answer: 𝑥𝑦𝑠 + 𝑧 = 𝑥𝑝 + 𝑦𝑞 Or 𝑥𝑦 +𝑧 =𝑥 +𝑦
𝜕𝑥𝜕𝑦 𝜕𝑥 𝜕𝑦
 4.2: Solutions of Partial Differential Equations
A first order PDE has a solution in the form 𝑓(𝑥, 𝑦, 𝑧, 𝑎, 𝑏) = 0 − − − −(1) which
contains two arbitrary constants is called a Complete Integral.
A solution obtained from the complete integral by assigning particular values to the
arbitrary constants is called a Particular Integral.
Replacing 𝑏 by 𝜙(𝑎) in (1) and find the envelope of the family of surfaces
𝑓(𝑥, 𝑦, 𝑧, 𝜙(𝑎)) = 0 then we get a solution containing an arbitrary function 𝜙 which is
called the General Integral.
The envelope of the family of surfaces (1) with parameters 𝑎 and 𝑏, if it exists, is called
a Singular Integral. The singular integral differs from the particular integral in that it
is not obtained from the complete integral by giving particular values to the constants.

4.2.1: Solution of Homogeneous PDE involving derivatives with

respect to one independent variable only
PROBLEMS
𝜕2 𝑧 𝜕𝑧
1) Solve + 𝑧 = 0, given that when 𝑥 = 0, 𝑧 = 𝑒 𝑦 and = 1.
𝜕𝑥 2 𝜕𝑥

Answer: 𝑧 = 𝑒 𝑦 cos 𝑥 + sin 𝑥

𝜕2 𝑧 𝜕𝑧
2) Solve = 𝑧, given that when 𝑦 = 0, 𝑧 = 𝑒 𝑥 and = 𝑒 −𝑥 .
𝜕𝑦 2 𝜕𝑦

Answer: 𝑧 = 𝑒 𝑦 cosh 𝑥 + 𝑒 −𝑦 sinh 𝑥

𝜕2 𝑧 𝜕𝑧
3) Solve + 4𝑧 = 0, given that when 𝑥 = 0, 𝑧 = 𝑒 2𝑦 and = 2.
𝜕𝑥 2 𝜕𝑥

Answer: 𝑧 = 𝑒 2𝑦 cos 2𝑥 + sin 2𝑥

𝜕2 𝑧 𝜕𝑧
4) Solve + 𝑧 = 0, given that when 𝑦 = 0, 𝑧 = cos 𝑥 and = sin 𝑥.
𝜕𝑦 2 𝜕𝑦

Answer: 𝑧 = cos(𝑥 − 𝑦)
 4.2.2: Solution of Lagrange’s Linear PDE of the type Pp+Qq=R
A linear PDE of the first order, commonly known as Lagrange’s linear equation is of the
𝜕𝑧 𝜕𝑧
form 𝑃𝑝 + 𝑄𝑞 = 𝑅, where 𝑝 = ,𝑞 = and 𝑃, 𝑄, 𝑅 are the functions of 𝑥, 𝑦, 𝑧. This
𝜕𝑥 𝜕𝑦

equation is called a quasi-linear equation. When 𝑃, 𝑄 & 𝑅 are independent of 𝑧 it is

known as linear equation.
Procedure to solve the equation 𝑷𝒑 + 𝑸𝒒 = 𝑹:
𝑑𝑥 𝑑𝑦 𝑑𝑧
i) Construct auxiliary or subsidiary equations = =
𝑃 𝑄 𝑅

ii) Solve the above subsidiary equations. Let the two independent solutions be
𝑢(𝑥, 𝑦, 𝑧) = 𝑐1 & 𝑣(𝑥, 𝑦, 𝑧) = 𝑐2
iii) Then 𝜙(𝑢, 𝑣 ) = 0 or 𝑢 = 𝑓 (𝑣 ) is the required solution.

PROBLEMS
1) Solve 𝑥𝑝 + 𝑦𝑞 = 3𝑧.
𝑥 𝑦3 𝑥 𝑦3
Answer: 𝜙 ( , )=0 Or = 𝑓( )
𝑦 𝑧 𝑦 𝑧

𝑦2 𝑧
2) Solve 𝑝 + 𝑥𝑧𝑞 = 𝑦 2 .
𝑥

Answer: 𝜙(𝑥 3 − 𝑦 3 , 𝑥 2 − 𝑧 2 ) = 0 Or 𝑥 3 − 𝑦 3 = 𝑓 (𝑥 2 − 𝑧 2 )
3) Solve 𝑥𝑝 − 𝑦𝑞 = 𝑦 2 − 𝑥 2 .
Answer: 𝜙(𝑥𝑦, 𝑥 2 + 𝑦 2 + 2𝑧) = 0 Or 𝑥𝑦 = 𝑓 (𝑥 2 + 𝑦 2 + 2𝑧)
4) Solve (𝑧 − 𝑦)𝑝 + (𝑥 − 𝑧)𝑞 = 𝑦 − 𝑥.
Answer: 𝜙(𝑥 + 𝑦 + 𝑧, 𝑥 2 + 𝑦 2 + 𝑧 2 ) = 0 Or 𝑥 2 + 𝑦 2 + 𝑧 2 = 𝑓 (𝑥 + 𝑦 + 𝑧 )
 4.2.3: Method of Separation of Variables (Product Method)
This method is a powerful technique to solve PDE with boundary conditions, when PDE
is linear and boundary conditions are homogeneous.
Working Procedure:
i) We assume the trial solution of the PDE (𝑢 is dependent & 𝑥, 𝑦 are independent) in
the form of a product. i.e., 𝑢(𝑥, 𝑦) = 𝑋(𝑥 )𝑌(𝑦), where 𝑋 is a function of 𝑥 alone
and 𝑌 is a function of 𝑦 alone.
ii) Substitute 𝑢(𝑥, 𝑦) = 𝑋𝑌 in the given PDE wherein the partial derivatives present in
the equation converts into ordinary derivatives.
iii) The resulting equation involving ordinary derivatives is rearranged in such a way
that LHS is a function of 𝑥 and RHS is a function of 𝑦 (or vice versa).
iv) We equate each side comprising ODE’s to a common constant 𝑘.
v) We solve the ODE’s to obtain 𝑋 = 𝑋(𝑥 ) & 𝑌 = 𝑌(𝑦).
vi) Substitution of 𝑋(𝑥 ) & 𝑌(𝑦) into 𝑢(𝑥, 𝑦) = 𝑋(𝑥 )𝑌(𝑦) results in the required
solution of the PDE.

PROBLEMS
𝜕𝑢 𝜕𝑢
1) Solve 𝑥 2 + 𝑦2 = 0 by the method of separation of variables.
𝜕𝑥 𝜕𝑦
1 1
𝑘( − )
Answer: 𝑢(𝑥, 𝑦) = 𝑐𝑒 𝑦 𝑥

2) Solve 𝑝𝑦 3 + 𝑞𝑥 2 = 0 by the method of separation of variables.

𝑘 3 −3𝑦 2 )
Answer: 𝑧(𝑥, 𝑦) = 𝑐𝑒 12(4𝑥
𝜕𝑢 𝜕𝑢
3) Solve 3 +2 = 0 where 𝑢(𝑥, 0) = 4𝑒 −𝑥 by the method of separation of
𝜕𝑥 𝜕𝑦

variables.
𝑥 𝑦 3
Answer: 𝑢(𝑥, 𝑦) = 𝑐1 𝑐2 𝑒 𝑘(3−2 ) and 𝑢(𝑥, 𝑦) = 4𝑒 −𝑥+(2)𝑦
𝜕𝑢 𝜕𝑢
4) Solve =4 where 𝑢(0, 𝑦) = 8𝑒 −3𝑦 by the method of separation of variables.
𝜕𝑥 𝜕𝑦
𝑦 𝑦
Answer: 𝑢(𝑥, 𝑦) = 𝑐1 𝑐2 𝑒 𝑘(𝑥+4 ) and 𝑢(𝑥, 𝑦) = 8𝑒 −12(𝑥+4 ) = 8𝑒 −(12𝑥+3𝑦)
 ADDITIONAL PROBLEMS
I. Form the PDE by eliminating the arbitrary constants from the followings
1. 𝑧 = 𝑎𝑥 + 𝑏𝑦 + √𝑎2 + 𝑏2
𝑥2 𝑦2
2. 2𝑧 = +
𝑎2 𝑏2

3. 𝑧 = 𝑎𝑥 3 + 𝑏𝑦 3
4. (𝑥 − 𝑎)2 + (𝑦 − 𝑏)2 = 𝑧 2 𝑐𝑜𝑡 2 𝛼
𝑥2 𝑦2 𝑧2
5.
𝑎 2 + 𝑏 2 + 𝑐2
=1

II. Form the PDE by eliminating the arbitrary functions from the following
1. 𝑧 = 𝑓 (𝑥 2 + 𝑦 2 ) + 𝑥 + 𝑦
1
2. 𝑧 = 𝑦 2 + 2𝑓 ( + log 𝑦)
𝑥

3. 𝑥𝑦𝑧 = 𝜙(𝑥 + 𝑦 + 𝑧)
4. 𝑧 = (𝑥 + 𝑦)𝜙(𝑥 2 − 𝑦 2 )
5. 𝑧 = 𝑥𝑓1 (𝑥 + 𝑡) + 𝑓2 (𝑥 + 𝑡)
6. 𝑧 = 𝑓1 (𝑦 + 2𝑥 ) + 𝑓2 (𝑦 − 3𝑥)
7. 𝑧 = 𝑓1 (𝑥 )𝑓2 (𝑦)
8. 𝑧 = 𝑥 2 𝑓(𝑦) + 𝑦 2 𝑔(𝑥 )
9. 𝑧 = 𝑓 (𝑥 + 𝑎𝑡) + 𝑔(𝑥 − 𝑎𝑡)

III. Solve the following homogeneous PDE’s

1
𝜕2 𝑢 𝜕𝑢
1. ( )
2 + 𝑢 = 0, given that 𝑢 0, 𝑦 = 𝑒
2 and (0, 𝑦) = 1.
𝜕𝑥 𝜕𝑥
𝜕2 𝑧 𝜕𝑧
2. = 𝑎2 𝑧, given that 𝑥 = 0, 𝑧 = 0 and = 𝑎 sin 𝑦.
𝜕𝑥 2 𝜕𝑥
𝜕2 𝑧 𝜕𝑧
3. = 𝑧, given that 𝑦 = 0, 𝑧 = 0 and = sin 𝑥.
𝜕𝑦 2 𝜕𝑦

𝜕2 𝑧 𝜕𝑧 𝜕𝑧
4. +3 − 4𝑧 = 0, given that 𝑧 = 1 and = 𝑦 when 𝑥 = 0.
𝜕𝑥 2 𝜕𝑥 𝜕𝑥
IV. Solve the following Lagrange’s Linear PDE’s
1. 𝑥𝑝 + 𝑦𝑞 = 𝑧
2. 𝑦 2 𝑧𝑝 − 𝑥 2 𝑧𝑞 = 𝑥 2 𝑦
3. (𝑦 + 𝑧)𝑝 − (𝑧 + 𝑥 )𝑞 = 𝑥 − 𝑦
4. 𝑥(𝑦 − 𝑧)𝑝 + 𝑦(𝑧 − 𝑥 )𝑞 = 𝑧(𝑥 − 𝑦)
5. 𝑥 2 (𝑦 − 𝑧)𝑝 + 𝑦 2 (𝑧 − 𝑥 )𝑞 = 𝑧 2 (𝑥 − 𝑦)
6. 𝑥(𝑦 2 − 𝑧 2 )𝑝 + 𝑦(𝑧 2 − 𝑥 2 )𝑞 = 𝑧(𝑥 2 − 𝑦 2 )

V. Solve the following PDE’s by the method of separation of variables

𝜕𝑧 𝜕𝑧
1. 𝑥 −𝑦 =0
𝜕𝑥 𝜕𝑦
𝜕𝑢 𝜕𝑢
2. + = 2(𝑥 + 𝑦)𝑢
𝜕𝑥 𝜕𝑦
𝜕𝑢 𝜕𝑢
3. 4 + = 3𝑢 with 𝑢(0, 𝑦) = 2𝑒 5𝑦
𝜕𝑥 𝜕𝑦
𝜕𝑢 𝜕𝑢
4. =2 + 𝑢 with 𝑢(𝑥, 0) = 6𝑒 −3𝑥
𝜕𝑥 𝜕𝑦