Engineering Analysis II:

Partial Differential Equations (PDEs)

and Applications to Engineering

By: Dr. Mahfouz Rostamzadeh

Soran University
Faculty of Engineering
2|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

1. Introduction

A PDE is an equation that contains one or more partial derivatives of an unknown function that
depends on at least two variables. Usually one of these deals with time t and the remaining with
space (spatial variable(s) like x, y and z). The most important PDEs are the wave equations that can
model the vibrating string and the vibrating membrane, the heat equation for temperature in a bar
or wire, and the Laplace equation for electrostatic potentials.
PDEs are very important in dynamics, elasticity, heat transfer, electromagnetic theory, and
quantum mechanics. They have a much wider range of applications than ODEs, which can model
only the simplest physical systems. Thus PDEs are subjects of many ongoing research and
development projects.
Realizing that modeling with PDEs is more involved than modeling with ODEs, we take a gradual,
well-planned approach to modeling with PDEs. To do this we carefully derive the PDE that models
the phenomena, such as the one-dimensional wave equation for a vibrating elastic string (say a
violin string), and then solve the PDE. In a similar vein, we derive the heat equation and then solve
and generalize it. We derive these PDEs from physics and consider methods for solving initial and
boundary value problems, that is, methods of obtaining solutions which satisfy the conditions
required by the physical situations. We show how PDEs can also be solved by Fourier and Laplace
transform methods.

2. Basic Concepts of PDEs

We recall that for a two variable function z = f(x, y) with independent variables x and y and
dependent variable z, the first partial differentials are
∂z ∂f
or (x, y) or fx = derivative of f 𝑤𝑖𝑡ℎ 𝑟𝑒𝑠𝑝𝑒𝑐𝑡 𝑡𝑜 𝑥 𝑤ℎ𝑒𝑛 𝑦 𝑖𝑠 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝜕𝑥 𝜕𝑥
∂z ∂f
or (x, y) or fy = derivative of f 𝑤𝑖𝑡ℎ 𝑟𝑒𝑠𝑝𝑒𝑐𝑡 𝑡𝑜 𝑦 𝑤ℎ𝑒𝑛 𝑥 𝑖𝑠 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝜕𝑦 𝜕𝑦
For other multivariable functions similarly the first partial differentials can be defined.

Example: For f(x, y) = x 2 𝑦 + cos(𝑥 2 − 𝑦) + √𝑥 + 1 the first partial differentials are:

∂f 1 ∂f
= fx = 2𝑥𝑦 − 2𝑥 sin(𝑥 2 − 𝑦) + ; = fy = 𝑥 2 + sin(𝑥 2 − 𝑦)
𝜕𝑥 2√𝑥 𝜕𝑦
3|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Second partial differentials can be defined as follows:

∂2 f 𝜕 𝜕𝑓 ∂2 f 𝜕 𝜕𝑓 ∂2 f 𝜕 𝜕𝑓
(x, y) or fxx = ( )  (x, y) or fxy = ( )  2 (x, y) or fyy = ( )
𝜕𝑥 2 𝜕𝑥 𝜕𝑥 𝜕𝑥𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑦

Note that for a differentiable function 𝑓, 𝑓𝑥𝑦 = 𝑓𝑦𝑥 .

4|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

• A partial differential equation (PDE) is an equation involving one or more partial

derivatives of an (unknown) function, call it u, that depends on two or more variables, often
time t and one or several variables in space.
 5|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

E X A M P L E 1 Important Second-Order PDEs

(1) One-dimensional wave equation

(2) One-dimensional heat equation

(3) Two-dimensional Laplace equation

(4) Two-dimensional Poisson equation

(5) Two-dimensional wave equation

(6) Three-dimensional Laplace equation

Here c is a positive constant, t is time, x, y, z are Cartesian coordinates and dimension is the
number of these coordinates in the equation.

• The order of the highest derivative is called the order of the PDE. Just as was the case for ODEs,
second-order PDEs will be the most important ones in applications.

• Just as for ordinary differential equations (ODEs) we say that a PDE is linear if it is of the first
degree in the unknown function u and its partial derivatives. Otherwise we call it nonlinear.
Thus, all the equations in Example 1 are linear and 𝑢𝑥 𝑢𝑦 = 𝑢𝑥𝑦 is nonlinear of order two and
(𝑢𝑥𝑥𝑥 )2 + sin(𝑢𝑦 + 𝑥) = 𝑢 is nonlinear of order 3.
• We call a linear PDE homogeneous if each of its terms contains either u or one of its partial
derivatives. Otherwise we call the equation nonhomogeneous. Thus, (4) in Example 1 (with f not
identically zero) is nonhomogeneous, whereas the other equations are homogeneous.

• A solution of a PDE in some region R of the space of the independent variables is a function
that has all the partial derivatives appearing in the PDE in some domain D containing R, and
satisfies the PDE everywhere in R. Often one merely requires that the function is continuous on
the boundary of R, has those derivatives in the interior of R, and satisfies the PDE in the interior
of R. Letting R lie in D simplifies the situation regarding derivatives on the boundary of R,
which is then the same on the boundary as it is in the interior of R.
 6|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

• In general, the totality of solutions of a PDE is very large. For example, the functions

which are entirely different from each other, are solutions of (3), as you may verify. For example
we check 𝑢 = sin 𝑥 cosh 𝑦:
𝑢𝑥 = cos 𝑥 cosh 𝑦 , 𝑢𝑥𝑥 = − sin 𝑥 cosh 𝑦
𝑢𝑦 = sin 𝑥 sinh 𝑦 , 𝑢𝑦𝑦 = sin 𝑥 cosh 𝑦
and so 𝑢𝑥𝑥 + 𝑢𝑦𝑦 = − sin 𝑥 cosh 𝑦 + sin 𝑥 cosh 𝑦 = 0. The others easily can be checked.

The unique solution of a PDE corresponding to a given physical problem will be obtained by the
use of additional conditions arising from the problem. The additional conditions may be imposed
on spatial boundaries belonging to a region D where the solution is required, and when this is done
the conditions are called boundary conditions. A typical boundary condition for a second order
PDE defined in a rectangle could be that the solution is required to assume specified values on the
sides of the rectangle. If time is involved, it is necessary to specify how the solution starts, and a
condition of this type is called an initial condition. Problems requiring initial and boundary
conditions are called initial boundary value problems (IBVPs).

A general first order PDE for the function u(x, y) is of the form

where F is an arbitrary function of its arguments. For example 𝑥𝑢𝑥 + 𝑢𝑦 = 𝑢 + 1 can be written
as 𝐹(𝑥, 𝑦, 𝑢, 𝑢𝑥 , 𝑢𝑦 ) = 𝑥𝑢𝑥 + 𝑢𝑦 − 𝑢 − 1 = 0.
7|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations
8|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Although in principle a general solution of a linear first order PDE can be found, unlike the general
solution of a linear first order ordinary differential equation (ODE) that contains an arbitrary
constant, the general solution of a linear first order PDE contains an arbitrary function. This
situation is illustrated by the first order PDE
9|Page Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations
10 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

3. Solving PDEs
We begin from the simplest partial differential equations.

1. 𝑢𝑥 = 0
Solution: we integrate the both sides with respect to x:
∫ 𝑢𝑥 (𝑥, 𝑦)𝑑𝑥 = ∫ 0𝑑𝑥 ⇒ 𝑢(𝑥, 𝑦) = 𝑐
But as the function is two variables so the constant c must be a function of y, hence the general
solution is 𝑢(𝑥, 𝑦) = 𝑓(𝑦) for any arbitrary function f of y. So we have an infinite number of
solutions. For example f(y) = y + 1, f(y) = cosy + lny and … are a solution of this PDE.

2. 𝑢𝑥 = 0 , 𝑢(𝑥, 0) = 𝑒 𝑦 + 𝑦 + 1

Solution: This one is a boundary value problem which its general solution as we already
obtained is 𝑢(𝑥, 𝑦) = 𝑓(𝑦). Now we need to verify the condition:
𝑢(0, 𝑦) = 𝑒 𝑦 + 𝑦 + 1 = 𝑓(𝑦) ⇒ 𝑢(𝑥, 𝑦) = 𝑒 𝑦 + 𝑦 + 1 is the unique solution of this
PDE.

3. 𝑢𝑦𝑦 = 0 , 𝑢𝑦 (𝑥, 0) = 𝑥 + 1 , 𝑢(𝑥, 0) = cos 𝑥

Solution:
We integrate from the both sides with respect to y:
∫ 𝑢𝑦𝑦 (𝑥, 𝑦)𝑑𝑦 = ∫ 0𝑑𝑦 ⇒ 𝑢𝑦 (𝑥, 𝑦) = 𝑓(𝑥) (i)
As we mentioned before, 𝑓(𝑥) is the constant parameter of the integration with respect to y.
Again we take another integration from the both sides of (i) and we have:
∫ 𝑢𝑦 (𝑥, 𝑦)𝑑𝑦 = ∫ 𝑓(𝑥)𝑑𝑦 ⇒ 𝑢(𝑥, 𝑦) = 𝑦𝑓(𝑥) + 𝑔(𝑥) (ii)
So the general solution is 𝑢(𝑥, 𝑦) = 𝑦𝑓(𝑥) + 𝑔(𝑥). Now we check the initial conditions:
(𝑖)
⇒ 𝑥 + 1 = 𝑢𝑦 (𝑥, 0) ⇒ 𝑓(𝑥) = 𝑥 + 1 (iii)
(𝑖𝑖),(𝑖𝑖𝑖)
⇒ 𝑢(𝑥, 𝑦) = 𝑦(𝑥 + 1) + 𝑔(𝑥)
(𝑖𝑖𝑖)
⇒ 𝑢(𝑥, 0) = 0 ⋅ (𝑥 + 1) + 𝑔(𝑥) = cos 𝑥 ⇒ 𝑔(𝑥) = 𝑐𝑜𝑠𝑥
Therefore 𝑢(𝑥, 𝑦) = 𝑦(𝑥 + 1) + cos 𝑥 is the particular and unique solution of this PDE.

4. 𝑢𝑥𝑦 = 𝑥 2
Solution: by integrating from the both sides
2
𝑥3
∫ 𝑢𝑥𝑦 (𝑥, 𝑦)𝑑𝑥 = ∫ 𝑥 𝑑𝑥 ⇒ 𝑢𝑦 (𝑥, 𝑦) = + 𝑓(𝑦)
3
Again we integrate

𝑥3 𝑥3
∫ 𝑢𝑦 (𝑥, 𝑦)𝑑𝑦 = ∫ [ + 𝑓(𝑦)] 𝑑𝑦 ⇒ 𝑢(𝑥, 𝑦) = 𝑦 + ∫ 𝑓(𝑦)𝑑𝑦 + 𝑔(𝑥)
3 3
11 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Note that we can rewrite again the general solution by putting ℎ(𝑦) ≔ ∫ 𝑓(𝑦)𝑑𝑦 as follows
𝑥3
𝑢(𝑥, 𝑦) = 𝑦 + ℎ(𝑦) + 𝑔(𝑥)
3

Remark (PDEs SOLVABLE AS ODEs):

This happens if a PDE involves derivatives with respect to one variable only (or can be
transformed to such a form), so that the other variable(s) can be treated as parameter(s). So
sometimes we can directly integrate like previous examples or use a suitable substitution to solve it
by ODE solution methods.

5. 𝑢𝑥𝑥 − 𝑢 = 0

Solution:
Since no y-derivatives occur, we can solve this PDE like 𝑢′′ − 𝑢 = 0 as an ordinary differential
equation. From ODEs we know 𝑟 2 − 1 = 0 is its characteristic equation and 𝑟 = ±1. So 𝑢1 = 𝑒 𝑥
and 𝑢1 = 𝑒 −𝑥 are its solutions thus 𝑢(𝑥, 𝑦) = 𝑐1 𝑒 𝑥 + 𝑐2 𝑒 −𝑥 . But u is a two variables function and
each of 𝑐1 and 𝑐2 is a function of y, so the general solution is of the following form:
𝑢(𝑥, 𝑦) = 𝑓(𝑦)𝑒 𝑥 + 𝑔(𝑦)𝑒 −𝑥

6. Solve 𝑢𝑥𝑡 + 𝑢𝑥 = 0 for 𝑢(𝑥, 𝑡).

Solution:
𝑦′
Set 𝑢𝑥 ≔ 𝑦. So 𝑢𝑥𝑡 = 𝑦 ′ ⇒ 𝑦 ′ + 𝑦 = 0 ⇒ ∫ 𝑑𝑡 = − ∫ 1𝑑𝑡
𝑦
⇒ ln|𝑦| = −𝑡 + 𝑐(𝑥) ⇒ 𝑦 = 𝑒 𝑐(𝑥) 𝑒 −𝑡 = ℎ(𝑥)𝑒 −𝑡 (note that we can write 𝑒 𝑐(𝑥) = ℎ(𝑥) )
⇒ 𝑢𝑥 = 𝑓(𝑥)𝑒 −𝑡 ⇒ ∫ 𝑢𝑥 𝑑𝑥 = ∫ ℎ(𝑥)𝑒 −𝑡 𝑑𝑥 = 𝑒 −𝑡 ∫ ℎ(𝑥)𝑑𝑥
⇒ 𝑢(𝑥, 𝑡) = 𝑓(𝑥)𝑒 −𝑡 + 𝑔(𝑡)
is the general solution. Note that f(x) and g(t) are arbitrary.

7. 𝑢𝑥𝑡 + 𝑢𝑥 = 𝑥𝑒 −𝑡
Setting 𝑢𝑥 ≔ 𝑦, we have 𝑢𝑥𝑡 = 𝑦 ′ ⇒ 𝑦 ′ + 𝑦 = 𝑥𝑒 −𝑡 . Now we can solve this first order linear ODE
with respect to t:
1
𝑦(𝑡) = (∫ 𝜇(𝑡)𝑥𝑒 −𝑡 𝑑𝑡 + 𝑐) ; 𝜇 = 𝑒 ∫ 1𝑑𝑡 = 𝑒 𝑡
𝜇(𝑡)
𝑦(𝑡) = 𝑒 −𝑡 (∫ 𝑒 𝑡 𝑥𝑒 −𝑡 𝑑𝑡 + 𝑐) = 𝑒 −𝑡 (𝑡𝑥 + 𝑓(𝑥)) (note that we can put c=f(x)).

⇒ 𝑦(𝑡) = 𝑢𝑥 (𝑥, 𝑡) = 𝑒 −𝑡 (𝑡𝑥 + 𝑓(𝑥))

𝑥2
⇒ ∫ 𝑢𝑥 (𝑥, 𝑡)𝑑𝑥 = ∫(𝑡𝑒 −𝑡 𝑥 + 𝑒 𝑡 𝑓(𝑥)) 𝑑𝑥 = 𝑡𝑒 −𝑡 ⋅ + 𝑒 −𝑡 ∫ 𝑓(𝑥)𝑑𝑥 + 𝑐
2
2
𝑐≔ℎ(𝑡) , ∫ 𝑓(𝑥)𝑑𝑥≔𝑔(𝑥) 𝑥
⇒ 𝑢(𝑥, 𝑡) = 𝑡𝑒 −𝑡 ⋅ + 𝑒 −𝑡 𝑔(𝑥) + ℎ(𝑡)
2
12 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

8. To see that not every first order PDE has a solution, it is only necessary to consider the
nonlinear equation
𝑢𝑥2 + 𝑢𝑦2 = −1
The expression on the left is nonnegative, so clearly this equation cannot be satisfied by any
real function u(x, y).
13 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations
14 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

4. The Method of Characteristics

The method of solution of a quasi linear first order PDE involving the unknown function u(x, y)
contains within it as special cases the solution of linear and semi linear first order PDEs. For a first-
order PDE (partial differential equation), the method of characteristics discovers curves (called
characteristic curves or just characteristics) along which the PDE becomes an ordinary differential
equation (ODE). Once the ODE is found, it can be solved along the characteristic curves and
transformed into a solution for the original PDE.
For the sake of simplicity, we confine our attention to the case of a function of two independent
variables x and y for the moment. Consider a quasi linear PDE of the form
𝑃(𝑥, 𝑦, 𝑢)𝑢𝑥 + 𝑄(𝑥, 𝑦, 𝑢)𝑢𝑦 = 𝑅(𝑥, 𝑦, 𝑢) (4.1)

where P, Q and R are assumed to be continuous functions of their arguments. To solve the
equation (4.1) once we consider the rate of changes in y to be zero, hence
𝜕𝑢
= 𝑢𝑦 = 0. Then we have 𝑃(𝑥, 𝑦, 𝑢)𝑢𝑥 = 𝑅(𝑥, 𝑦, 𝑢) and so
𝜕𝑦

𝑑𝑥 𝑑𝑢
= 𝑅(𝑥,𝑦,𝑢) (4.2)
𝑃(𝑥,𝑦,𝑢)

𝜕𝑢
Similarly we consider the rate of changes in x to be zero, hence = 𝑢𝑥 = 0. Then we have
𝜕𝑥
𝑄(𝑥, 𝑦, 𝑢)𝑢𝑦 = 𝑅(𝑥, 𝑦, 𝑢) and so

𝑑𝑦 𝑑𝑢
= 𝑅(𝑥,𝑦,𝑢) (4.3)
𝑄(𝑥,𝑦,𝑢)

From (4.2) and (4.3) we can deduce the method of characteristics to solve quasi-linear first order
PDEs as follows:

4.1 The Method of Characteristics:

To solve 𝑃(𝑥, 𝑦, 𝑢)𝑢𝑥 + 𝑄(𝑥, 𝑦, 𝑢)𝑢𝑦 = 𝑅(𝑥, 𝑦, 𝑢) we consider the following steps:

1. Set
𝑑𝑥 𝑑𝑦 𝑑𝑢
= =
𝑃(𝑥, 𝑦, 𝑢) 𝑄(𝑥, 𝑦, 𝑢) 𝑅(𝑥, 𝑦, 𝑢)
and solve the following two ordinary differential equations
𝑑𝑥 𝑑𝑦
= ⟹ 𝑢1 (𝑥, 𝑦, 𝑢) = 𝑐1
𝑃(𝑥, 𝑦, 𝑢) 𝑄(𝑥, 𝑦, 𝑢)

𝑑𝑥 𝑑𝑢
= ⟹ 𝑢2 (𝑥, 𝑦, 𝑢) = 𝑐2
𝑃(𝑥, 𝑦, 𝑢) 𝑅(𝑥, 𝑦, 𝑢)
2. The intersection of 𝑢1 and 𝑢2 gives the solution, hence u(x, y). For this purpose we
substitute 𝑐1 and 𝑐2 in 𝑐2 = 𝑓(𝑐1 ).
15 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Example 4.1: Find the general solution of 𝑢𝑥 + 2𝑢𝑦 = 𝑥 .

Solution:
First of all 𝑃(𝑥, 𝑦, 𝑢) = 1, 𝑄(𝑥, 𝑦, 𝑢) = 2 and 𝑅(𝑥, 𝑦, 𝑢) = 𝑥. So
1.
𝑑𝑥 𝑑𝑦 𝑑𝑢
= =
1 2 𝑥
𝑑𝑥 𝑑𝑦
= ⟹ 𝑑𝑦 = 2𝑑𝑥 ⟺ ∫ 𝑑𝑦 = ∫ 2𝑑𝑥 ⟺ 𝑦 = 2𝑥 + 𝑐1 ⟺ 𝑐1 = 𝑦 − 2𝑥
1 2
𝑑𝑥 𝑑𝑢 𝑥2 𝑥2
= ⟹ 𝑑𝑢 = 𝑥𝑑𝑥 ⟺ ∫ 𝑑𝑢 = ∫ 𝑥𝑑𝑥 ⟺ 𝑢 = + 𝑐2 ⟺ 𝑐2 = 𝑢 −
1 𝑥 2 2
𝑥2 𝑥2
2. 𝑐2 = 𝑓(𝑐1 ) ⟺ 𝑢 − = 𝑓(𝑦 − 2𝑥) ⟺ 𝑢(𝑥, 𝑦) = 𝑓(𝑦 − 2𝑥) +
2 2

Example 4.2: Solve the initial valued problem of 𝑢𝑥 + 3𝑢𝑦 = 2𝑢, 𝑢(𝑥, 0) = 𝑒 𝑥 .
Solution:
First of all 𝑃(𝑥, 𝑦, 𝑢) = 1, 𝑄(𝑥, 𝑦, 𝑢) = 3 and 𝑅(𝑥, 𝑦, 𝑢) = 2𝑢. So
1.
𝑑𝑥 𝑑𝑦 𝑑𝑢
= =
1 3 2𝑢
𝑑𝑥 𝑑𝑦
= ⟹ 𝑑𝑦 = 3𝑑𝑥 ⟺ ∫ 𝑑𝑦 = ∫ 3𝑑𝑥 ⟺ 𝑦 = 3𝑥 + 𝑐1 ⟺ 𝑐1 = 𝑦 − 3𝑥
1 3
𝑑𝑥 𝑑𝑢 𝑑𝑢 𝑒 2𝑥
= ⟺∫ = ∫ 𝑑𝑥 ⟺ ln(𝑐2 𝑢) = 2𝑥 ⟺ 𝑐2 𝑢 = 𝑒 2𝑥 ⟺ 𝑐2 =
1 2𝑢 𝑢 𝑢

𝑒 2𝑥 𝑒 2𝑥
2. 𝑐2 = 𝑓(𝑐1 ) ⟺ = 𝑓(𝑦 − 3𝑥) ⟺ 𝑢 = 𝑓(𝑦−3𝑥) = 𝑔(𝑦 − 3𝑥)𝑒 2𝑥
𝑢
1
Note that without losing of generality we can write 𝑓(𝑦−3𝑥) = 𝑔(𝑦 − 3𝑥) for a function g.
Thus 𝑢(𝑥, 𝑦) = 𝑔(𝑦 − 3𝑥)𝑒 2𝑥 is the general solution. Now we check the initial condition:
𝑢(𝑥, 0) = 𝑒 2𝑥 𝑔(0 − 3𝑥) = 𝑒 𝑥 ⟺ 𝑔(−3𝑥) = 𝑒 −𝑥
𝑡
t
Set −3x = t ⟹ x = − 3 ⟹ g(t) = e−(−3) = 𝑒 𝑡/3 ⟹ 𝑔(𝑥) = 𝑒 𝑥/3
So
𝑦−3𝑥 𝑦
𝑢(𝑥, 𝑦) = 𝑔(𝑦 − 3𝑥)𝑒 2𝑥 = 𝑒 3 𝑒 2𝑥 = 𝑒 3+𝑥
is the particular solution.

Example 4.3: Solve the problem of 𝑢𝑥 + 𝑢𝑢𝑦 + 𝑢 = 0, 𝑢(0, 𝑦) = 1 + 𝑦.

Solution:
First of all 𝑢𝑥 + 𝑢𝑢𝑦 = −𝑢 and 𝑃(𝑥, 𝑦, 𝑢) = 1, 𝑄(𝑥, 𝑦, 𝑢) = 𝑢 and 𝑅(𝑥, 𝑦, 𝑢) = −𝑢. So
1.
𝑑𝑥 𝑑𝑦 𝑑𝑢
= =
1 𝑢 −𝑢
𝑑𝑥 𝑑𝑦
= ⟹ 𝑑𝑦 = 𝑢𝑑𝑥 ⟺ ∫ 𝑑𝑦 = ∫ 𝑢𝑑𝑥 ⟺ 𝑦 = 𝑢𝑥 + 𝑐1 ⟺ 𝑐1 = 𝑦 − 𝑢𝑥
1 𝑢
𝑑𝑥 𝑑𝑢 𝑑𝑢 𝑒 −𝑥
= ⟺∫ = − ∫ 𝑑𝑥 ⟺ ln(𝑐2 𝑢) = −𝑥 ⟺ 𝑐2 𝑢 = 𝑒 −𝑥 ⟺ 𝑐2 =
1 −𝑢 𝑢 𝑢
16 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

𝑒 −𝑥
2. 𝑐2 = 𝑓(𝑐1 ) ⟺ = 𝑓(𝑦 − 𝑢𝑥)
𝑢

𝑒 −0 1
= 𝑓(𝑦 − 𝑢 ⋅ 0) ⟺ = 𝑓(𝑦)
𝑢(0, 𝑦) 1+𝑦
𝑒 −𝑥 1
⟺ = 𝑓(𝑦 − 𝑢𝑥) = ⟺ 𝑢𝑒 𝑥 = 1 + 𝑦 − 𝑢𝑥
𝑢 1 + 𝑦 − 𝑢𝑥
⟺ 𝑢(𝑒 𝑥 + 𝑥) = 1 + 𝑦
1+𝑦
⟺ 𝑢(𝑥, 𝑦) = 𝑥
𝑒 +𝑥
__________________________________________________________
17 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

5. An Application of First Order PDEs: derivation of the 1st order PDE

involving a transient heat balance
To illustrate one of the ways in which first order PDEs arise from physical situations, we will
derive the equation governing the transient heat balance between a pipe transporting a hot fluid and
the air surrounding the pipe at a constant temperature 𝑇0 . Let the length of the pipe be L, the
constant speed of the fluid through the pipe be u and the temperature of the fluid be 𝑇(𝑥, 𝑡), where
𝑥 is the distance along the pipe and t is the time measured from the moment a particle of fluid
enters the pipe. The physical situation is represented in Fig. 1, and in order to arrive at the transient
heat balance equation we will consider the situation in an element of the pipe of length Δ𝑥.
The instantaneous energy balance that is to be modeled in the element of pipe of length Δ𝑥 can be
represented as follows:

{Energy entering with fluid} − {energy leaving with fluid} − {heat transferred to air} ={energy
stored in fluid} (5.1)

Figure 5.1 Transient heat distributions in an element of the pipe of length Δ𝑥.

If Δ𝑡 is the time taken for a particle of fluid to travel through an element of the pipe of length Δ𝑥,
Δ𝑥
then the fluid speed 𝑢 ≈ ∆𝑡 . If we denote the mass of fluid present in this element by M and the
𝑚Δ𝑥
mass flow rate by m, the quantities M and m are related by 𝑀 = 𝑢 .
If the fluid enters the element at the temperature 𝑇(𝑥, 𝑡), its temperature when leaving it can be
𝜕𝑇
approximated by 𝑇 + Δ𝑥(𝜕𝑥 ). If we assume that the transfer of heat from the surface of the pipe to
the air is proportional to the temperature difference 𝑇(𝑥, 𝑡) − 𝑇0, and denote the heat transfer
Δ𝑥
coefficient by 𝛼, the heat transferred from the surface of the pipe to the air will be 𝛼𝑆( 𝐿 )(𝑇 − 𝑇0 ),
where S is the surface area of the pipe. The heat energy entering the element due to the fluid is
𝑚𝑐𝑇, where c is the specific heat of the fluid, and the heat energy leaving with the fluid is
18 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

𝜕𝑇 𝜕𝑇
𝑚𝑐𝑇 (𝑇 + Δ𝑥 (𝜕𝑥 )), whereas the stored energy in the fluid occupying the element is 𝑀𝑐( 𝜕𝑡 ).
Substituting these quantities into the energy balance equation gives

Δx
Cancelling terms, and dividing the above equation by Mc = cm , this balance equation becomes
u
the PDE for transient heat transfer:

𝜕𝑇 𝜕𝑇 𝛼𝑢𝑆
+𝑢 =− (𝑇 − 𝑇0 ).
𝜕𝑡 𝜕𝑥 𝑀𝑐𝐿

6. Classification and Reduction to Standard Form of a Second Order

Constant Coefficient PDE for u(x, y)

The purpose of the present section is to explain the basis of a classification where, for simplicity, in
the main the discussion will be limited to linear second order partial differential equations whose
coefficients are either constants or functions of the independent variables involved.
We have already seen that in the case of two dimensions, examples of these equations involving a
function u are as follows:
1. One dimensional wave equation: 𝑢𝑡𝑡 = 𝑐 2 𝑢𝑥𝑥
2. One dimensional heat equation: 𝑢𝑡 = 𝜅 2 𝑢𝑥𝑥
3. Two dimensional Laplace equation: 𝑢𝑥𝑥 + 𝑢𝑦𝑦 = 0

These three equations are all special cases of the general linear PDE for an
unknown twice differentiable classical solution u(x, y) of the two independent variables x and y, or
sometimes t and x, which is defined in some region D and can be written
𝐴𝑢𝑥𝑥 + 2𝐵𝑢𝑥𝑦 + 𝐶𝑢𝑦𝑦 + 𝑃𝑢𝑥 + 𝑄𝑢𝑦 + 𝑅𝑢 = 𝐹(𝑥, 𝑦) (∗)
where 𝐴, 𝐵, 𝐶, 𝑃, 𝑄, and 𝑅 are functions of 𝑥 and 𝑦.
The factor 2 multiplying B has been introduced for convenience as it simplifies the calculations
that are to follow. The functions 𝐴, 𝐵, … , 𝑅 multiplying u and its derivatives are called the
coefficients of the PDE, and 𝐹(𝑥, 𝑦) is called the nonhomogeneous term. Equation (∗) is called
homogeneous if 𝐹(𝑥, 𝑦) is identically zero.
To classify 2nd order linear PDEs we set 𝑑 = 𝐵 2 − 4𝐴𝐶 and we call it the discriminant of the
PDE. Then:
• If 𝑑 > 0 then the PDE is of hyperbolic type
• If 𝑑 = 0 then the PDE is of parabolic type
• If 𝑑 > 0 then the PDE is of elliptic type

Example 6.1:
The wave equation 𝑐 2 𝑢𝑥𝑥 − 𝑢𝑡𝑡 = 0 is of hyperbolic type, because 𝑑 = 0 − (𝑐 2 )(−1) > 0.
19 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Example 6.2:
The heat equation 𝜅 2 𝑢𝑥𝑥 − 𝑢𝑡 = 0 is of parabolic type, because 𝑑 = 0 − (𝜅 2 )(0) = 0.

Example 6.3:

The Laplce equation 𝑢𝑥𝑥 + 𝑢𝑡𝑡 = 0 is of elliptic type, because 𝑑 = 0 − (𝑐 2 )(1) < 0.

Example 6.4:
For the equation 𝑦𝑢𝑥𝑥 − 𝑢𝑥𝑦 + 𝑥𝑢𝑦𝑦 + 5𝑥𝑢𝑥 + 3 cos(𝑥 + 𝑦) 𝑢𝑦 = 3𝑢 we consider the
1 2 1
discriminant 𝑑 = (− 2) − (𝑥)(𝑦) = 4 − 𝑥𝑦. To
classify it we should discuss on 𝑑 in three cases,
because d is a function.
1
i. If 𝑑 = 4 − 𝑥𝑦 > 0 then the PDE is of
hyperbolic type for the all points between
1
two branches of the curve 𝑦 = 4𝑥 as
shown in figure with blue color.
1 1
ii. If 𝑑 = 4 − 𝑥𝑦 = 0 then 𝑥𝑦 = 4 the PDE is
of parabolic type for all the points exactly
1
on the curve 𝑦 = 4𝑥 as shown in figure in black color.
1
iii. If 𝑑 = 4 − 𝑥𝑦 < 0 then the PDE is of elliptic type outside of two branches of the curve 𝑦 =
1
as shown in figure by read color.
4𝑥

Example 6.5:
Classify the PDE 𝑢𝑥𝑥 − 2𝑥𝑢𝑥𝑦 + 𝑢𝑦𝑦 + ln(𝑥) 𝑢𝑦 = 𝑥𝑢 + 4.
Solution: We set 𝑑 = (−𝑥)2 − (1)(1) = 𝑥 2 − 1. We
verify the three cases:
i. If 𝑑 = 𝑥 2 − 1 > 0 then the PDE is of hyperbolic
type for all the points 𝑥 > 1 𝑜𝑟 𝑥 < −1 as
shown in figure with blue color.
ii. If 𝑑 = 𝑥 2 − 1 = 0 then the PDE is of parabolic
type for all the points exactly on the lines 𝑥 =
1 𝑎𝑛𝑑 𝑥 = −1 as shown in figure in black color.
iii. If 𝑑 = 𝑥 2 − 1 < 0 then the PDE is of elliptic
type for all the points on the region −1 < 𝑥 < 1
as shown in figure by read color.

This apparently arbitrary classification of the PDEs in (*) is of fundamental importance for the
following reasons:
(a) The classification of a PDE is independent of the choice of coordinate system used when
formulating the equation. Expressed differently, the classification is such that it does not
depend on the choice of independent variables. So, for example, if a PDE is of elliptic type
when expressed in terms of the Cartesian coordinates x and y, it will still be of elliptic type
20 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

when expressed in terms of any other coordinate system like the cylindrical polar coordinates r
, θ, and z.
(b) The nature of an appropriate domain D and the associated auxiliary conditions (initial
and/or boundary conditions) that must be imposed on the PDE in order to ensure a unique
solution throughout D differ according to the classification.

We will only justify statement (a), as the significance of (b) will become apparent when boundary
and initial conditions are considered.

6.1 Reduction of 2nd Order Linear PDEs to Standard Forms

Let us make a transformation of the independent coordinate variables x and y to ξ and η in such a
way that one point in the domain D in the (x, y)-plane corresponds to one point in the
corresponding domain in the (𝜉, 𝜂) −plane, and conversely (the transformation is one-one between
the two domains), by setting
𝜉 = 𝜉 (𝑥, 𝑦)𝑎𝑛𝑑 𝜂 = 𝜂(𝑥, 𝑦)
where the functions 𝜉 and 𝜂 are assumed to be twice continuously differentiable. Using the rules
from the calculus for a change of variables to express the partial derivatives of 𝑢 with respect to 𝑥
and 𝑦 in terms of those with respect to 𝜉 and 𝜂, one can derive the standard forms of a 2nd order
linear PDE as follows:
• Hyperbolic PDEs Standard Form:
𝑢𝜉𝜂 = 𝐹1 (𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 ) or 𝑢𝜉𝜉 = 𝑢𝜂𝜂 + 𝐹2 (𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 )
• Parabolic PDEs Standard Form: 𝑢𝜂𝜂 = 𝐺(𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 )
• Elliptic PDEs Standard Form: 𝑢𝜉𝜉 + 𝑢𝜂𝜂 = 𝐻(𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 )
where 𝐹1 , 𝐹2 ,G, and H are linear combinations of 𝑢, 𝑢𝜉 , 𝑎𝑛𝑑 𝑢𝜂 .

In the following, we reduce the PDE

𝐴𝑢𝑥𝑥 + 2𝐵𝑢𝑥𝑦 + 𝐶𝑢𝑦𝑦 + 𝑃𝑢𝑥 + 𝑄𝑢𝑦 + 𝑅𝑢 = 𝐹(𝑥, 𝑦)
to the standard form of 𝜉 and 𝜂:
𝐴∗ 𝑢𝜉𝜉 + 2𝐵 ∗ 𝑢𝜉𝜂 + 𝐶 ∗ 𝑢𝜂 𝜂 + 𝑃∗ 𝑢𝜉 + 𝑄 ∗ 𝑢𝜂 + 𝑅𝑢(𝜉, 𝜂) = 𝐹(𝑥(𝜉, 𝜂), 𝑦(𝜉, 𝜂))
By obtaining the values of 𝑢𝑥𝑥 , 𝑢𝑦𝑦 , 𝑢𝑥 , 𝑢𝑦 using the chain rule from calculus

and substituting and simplify the equation we get the values of 𝐴∗ , … , 𝐹 ∗ .

21 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Reduction to Standard Form:

We follow these steps:

Step 1: Determine the type of PDE by 𝑑 = 𝐵 2 − 𝐴𝐶

Step 2: Determine characteristic equations and characteristics curves as follows:

dy 𝐵+√𝐵2 −𝐴𝐶
= (1)
Case hyperbolic (d>0): we solve {dx 𝐴
which we call them
𝑑𝑦 𝐵−√𝐵2 −𝐴𝐶
= (2)
𝑑𝑥 𝐴
𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛
(1) ⇒ 𝜙1 (𝑥, 𝑦) = 𝑐1
characteristics equations of the PDE and call their solution { 𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛
(2) ⇒ 𝜙2 (𝑥, 𝑦) = 𝑐2
characteristics curves. Then we set 𝜙1 (𝑥, 𝑦) = 𝜉(𝑥, 𝑦) and 𝜙2 (𝑥, 𝑦) = 𝜂(𝑥, 𝑦).

dy 𝐵
Case Parabolic (d=0): dx = 𝐴 is the characteristic equation and
We set (𝑥,𝑦)=𝑥 and 𝜉(𝑥,𝑦)=𝜙(𝑥,𝑦) (𝜉=𝑦−𝐵/𝐴 𝑥+c when A and B are constant)
is the characteristic curve.

dy 𝐵±√𝐵2 −𝐴𝐶
Case Elliptic (d<0): dx = are the characteristic equations.
𝐴

By solution we have two complex curves:

𝛼+𝛽 𝛼−𝛽
𝛼(𝑥, 𝑦) = 𝑐1 𝑎𝑛𝑑 𝛽(𝑥, 𝑦) = 𝑐2 𝜉=
𝑎𝑛𝑑 𝜂 =
2 2𝑖
For special case when all coefficients of the PDE are constant we set
𝐴𝑦−𝐵𝑥
𝜉(𝑥, 𝑦) = 𝑥 and 𝜂(𝑥, 𝑦) = √𝐴𝐶−𝐵2.

Step 3: Calculate 𝜉𝑥 , 𝜉𝑦 , 𝜉𝑥𝑦 , 𝜉𝑥𝑥 , 𝜉𝑦𝑦 and 𝜂𝑥 , 𝜂𝑦 , 𝜂𝑥𝑦 , 𝜂𝑥𝑥 , 𝜂𝑦𝑦

Step 4: Calculate 𝑢𝑥𝑥 , 𝑢𝑦𝑦 , 𝑢𝑥𝑦 , 𝑢𝑥 , 𝑢𝑦

Step 5: substitute 𝑢𝑥𝑥 , 𝑢𝑦𝑦 , 𝑢𝑥𝑦 , 𝑢𝑥 , 𝑢𝑦 into the main equation to reduce it to the canonical
(standard) form
• Hyperbolic Standard Form: 𝑢𝜉𝜂 = 𝐹1 (𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 )
• Parabolic PDEs Standard Form: 𝑢𝜂𝜂 = 𝐺(𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 )
22 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

• Elliptic PDEs Standard Form: 𝑢𝜉𝜉 + 𝑢𝜂𝜂 = 𝐻(𝜉, 𝜂, 𝑢, 𝑢𝜉 , 𝑢𝜂 )

Step 6: If it is possible, solve it based on 𝜉 and 𝜂.

Step 7: Now replace 𝜉 and 𝜂 by their values based on 𝑥 and 𝑦 and apply the IVBP conditions.

Example 6.6: Reduce to standard form

𝑢𝑥𝑥 + 8𝑢𝑥𝑦 + 7𝑢𝑦𝑦 + 𝑢𝑥 + 2𝑢𝑦 + 3𝑢 + 𝑦 = 0,
and find its characteristic equations and curves. Then solve it.
Solution:

Step 1: that 𝐴 = 1, 𝐵 = 4, 𝐶 = 7 and 𝑑 = 𝐵 2 − 𝐴𝐶 = 16 − 7 > 0. So the equation is hyperbolic.

dy 4+√9
= =7
Step 2: Characteristic equations are {dx 1
and the corresponding solutions, hence the
𝑑𝑦 4−√9
= 1 =1
𝑑𝑥
𝑦 = 7𝑥 + 𝑐1 ⇒ 𝑐1 = 𝑦 − 7𝑥
characteristic curves are {
𝑦 = 𝑥 + 𝑐2 ⇒ 𝑐2 = 𝑦 − 𝑥
So 𝜉(𝑥, 𝑦) = 𝑦 − 7𝑥 and 𝜂(𝑥, 𝑦) = 𝑦 − 𝑥 (*)

Step 3: 𝜉𝑥 = −7, 𝜉𝑦 = 1, 𝜉𝑥𝑦 = 𝜉𝑥𝑥 = 𝜉𝑦𝑦 = 0 and 𝜂𝑥 = −1, 𝜂𝑦 = 1, 𝜂𝑥𝑦 = 𝜂𝑥𝑥 = 𝜂𝑦𝑦 = 0

Step 4: 𝑢𝑥𝑥 = 49𝑢𝜉𝜉 + 𝑢𝜂𝜂 − 14𝑢𝜉𝜂 , 𝑢𝑦𝑦 = 𝑢𝜉𝜉 + 𝑢𝜂𝜂 + 2𝑢𝜉𝜂

𝑢𝑥𝑦 = −7𝑢𝜉𝜉 − 𝑢𝜂𝜂 − 8𝑢𝜉𝜂 , 𝑢𝑥 = −7𝑢𝜉 − 𝑢𝜂 , 𝑢𝑦 = 𝑢𝜉 + 𝑢𝜂

7𝜂−𝜉
Step 5: substituting 𝑢𝑥𝑥 , 𝑢𝑦𝑦 , 𝑢𝑥𝑦 , 𝑢𝑥 , 𝑢𝑦 and 𝑦 = 6 (we calculated it from (*) ) into the
equation
(49𝑢𝜉𝜉 + 𝑢𝜂𝜂 − 14𝑢𝜉𝜂 ) + 8(−7𝑢𝜉𝜉 − 𝑢𝜂𝜂 − 8𝑢𝜉𝜂 ) + 7(𝑢𝜉𝜉 + 𝑢𝜂𝜂 + 2𝑢𝜉𝜂 ) + (−7𝑢𝜉 − 𝑢𝜂 )
7𝜂 − 𝜉
+ 2(𝑢𝜉 + 𝑢𝜂 ) + 3𝑢 + ( ) = 0
6
7𝜂−𝜉
So we have −64𝑢𝜉𝜂 = −5𝑢𝜉 + 𝑢𝜂 + 3𝑢 + ( 6 )
Step 6: Solving this PDE is difficult, we may solve it later.

Example 6.7: Reduce the heat equation 𝑢𝑡 = 𝑐 2 𝑢𝑥𝑥 to canonical (standard) form.
Solution:
Step 1: We rearrange it: 𝑐 2 𝑢𝑥𝑥 − 𝑢𝑡 = 0. So 𝐴 = 𝑐 2 , 𝐵 = 0, 𝐶 = 0, hence 𝑑 = 0 , i.e. it is a
parabolic PDE.
dt 𝐵
Step 2: dx = 𝐴 = 0 is the characteristic equation and 𝑡 = 𝑘 is the characteristic curve for constant
number k. We set 𝜉(𝑥, 𝑡) = 𝑥 and 𝜂(𝑥, 𝑡) = 𝑡 − 𝑘.
Step 3: 𝜉𝑡 = 𝜉𝑥𝑡 = 𝜉𝑥𝑥 = 𝜉𝑡𝑡 = 𝜂𝑥 = 𝜂𝑥𝑡 = 𝜂𝑥𝑥 = 𝜂𝑡𝑡 = 0 and 𝜉𝑥 = 𝜂𝑡 = 1
Step 4: The equation only includes 𝑢𝑥𝑥 𝑎𝑛𝑑 𝑢𝑡 which 𝑢𝑥𝑥 = 𝑢𝜉𝜉 , 𝑢𝑡 = 𝑢𝜂
Step 5: Canonical form is 𝑐 2 𝑢𝜉𝜉 − 𝑢𝜂 = 0
As you can see again we have the same equation, because heat equation itself is of canonical form.
23 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Example 6.8 (Wave Equation):

Let us consider a uniform stretched linearly elastic string under a tension T that is displaced from
its equilibrium position and then released. This could, for example, represent the response of a
plucked violin string. One can derive the PDE governing the motion of the string after its release
the forces acting on the string at time t when it has been displaced through the string of length L in
the u-direction transverse to its equilibrium position along the x-axis which we show it by 𝑢(𝑥, 𝑡)
satisfying the PDE 𝑢𝑡𝑡 = 𝑐 2 𝑢𝑥𝑥 where 𝑐 2 = 𝑇/𝜌 and 𝜌 is the line density of the string.

Find the general solution of the wave equation after reducing it to the canonical form.
Solution:

Step 1: We rearrange it: 𝑢𝑡𝑡 − 𝑐 2 𝑢𝑥𝑥 = 0 (𝑡:=𝑥, 𝑥≔𝑦)

𝐴=1, 𝐵=0, 𝐶 = −𝑐 2 and 𝑑 = 𝐵 2 − 𝐴𝐶 = 𝑐 2 > 0. So it is of hyperbolic type.

dx 0+√𝑐 2
= =𝑐
Step 2: Characteristic equations are { dt 1
and the corresponding solutions, hence the
𝑑𝑥 0−√𝑐 2
=
= −𝑐
𝑑𝑡 1
𝑥 = 𝑐𝑡 + 𝑐1 ⇒ 𝑐1 =
𝑥 − 𝑐𝑡 𝜉 = 𝑥 − 𝑐𝑡
characteristic curves are {𝑥 = −𝑐𝑡 + 𝑐 ⇒ 𝑐 = 𝑥 + 𝑐𝑡 . We set {
2 2 𝜂 = 𝑥 + 𝑐𝑡

Step 3: 𝜉𝑥 = 1, 𝜉𝑡 = −𝑐, 𝜉𝑥𝑡 = 𝜉𝑥𝑥 = 𝜉𝑡𝑡 = 0 and 𝜂𝑥 = 1, 𝜂𝑡 = 𝑐, 𝜂𝑥𝑡 = 𝜂𝑥𝑥 = 𝜂𝑡𝑡 = 0

Step 4: We only need 𝑢𝑥𝑥 = 𝑢𝜉𝜉 + 𝑢𝜂𝜂 + 2𝑢𝜉𝜂 , 𝑢𝑡𝑡 = 𝑐 2 𝑢𝜉𝜉 + 𝑐 2 𝑢𝜂𝜂 − 2𝑐 2 𝑢𝜉𝜂 .

Step 5: substituting 𝑢𝑥𝑥 , 𝑎𝑛𝑑 𝑢𝑡𝑡 we have

𝑐 2 (𝑢𝜉𝜉 + 𝑢𝜂𝜂 + 2𝑢𝜉𝜂 ) = 𝑐 2 𝑢𝜉𝜉 + 𝑐 2 𝑢𝜂𝜂 − 2𝑐 2 𝑢𝜉𝜂
So we have 4𝑐 2 𝑢𝜉𝜂 = 0 or equivalently, 𝑢𝜉𝜂 = 0 is the canonical form.
Step 6: Now we solve 𝑢𝜉𝜂 = 0.
From special cases of section 3, we take integral from the both sides first with respect to 𝜂 then
with respect to 𝜉, hence
∫ 𝑢𝜉𝜂 𝑑𝜂 = ∫ 0 𝑑𝜂 ⇒ 𝑢𝜉 = 𝑐1 = 𝑓(𝜉)
Again we take integral
∫ 𝑢𝜉 𝑑𝜉 = ∫ 𝑓(𝜉) 𝑑𝜉 ⇒ 𝑢(𝜉, 𝜂) = ∫ 𝑓(𝜉) 𝑑𝜉 + 𝑐2

Since 𝑐2 is a function of 𝜂 and ∫ 𝑓(𝜉) 𝑑𝜉 is general, we can write the general solution of the wave
equation as
𝑢(𝜉, 𝜂) = 𝑓(𝜉) + 𝑔(𝜂)
Step 7: now we substitute again the values of 𝜉 and 𝜂, so
𝒖(𝒙, 𝒚) = 𝒇(𝒙 − 𝒄𝒕) + 𝒈(𝒙 + 𝒄𝒕)
is the general solution of the wave equation 𝑐 2 𝑢𝑥𝑥 = 𝑢𝑡𝑡 .
24 | P a g e Soran University-Faculty of Engineering
Engineering Analysis II: Partial Differential Equations

Section 7: Solution of PDEs by Laplace Transforms

Readers familiar with Laplace transforms may wonder whether they can also be used for solving
partial differential equations. The answer is yes, particularly if one of the independent variables
ranges over the positive axis. The steps to obtain a solution are similar to those in ODEs. For a
PDE in two variables they are as follows.
1. Take the Laplace transform with respect to one of the two variables, usually 𝑡. This gives an
ODE for the transform of the unknown function. This is so since the derivatives of this function
with respect to the other variable slip into the transformed equation. The latter also incorporates
the given boundary and initial conditions.
2. Solving that ODE, obtain the transform of the unknown function.
3. Taking the inverse transform, obtain the solution of the given problem.
If the coefficients of the given equation do not depend on t, the use of Laplace transforms will
simplify the problem.

Corresponding results are easily written down for mixed and higher order derivatives using the
results for the ordinary Laplace transform given in