Boundary Value Problems

Mathematical Formulation and Solution of Physical Problems

1. Mathematical Formulation: To achieve such formulation, we usually adopt
mathematical models which serve to approximate the real objects under
investigation. In the mathematical formulation, we use known physical laws to
set up equations describing the problem.
2. Mathematical Solution: Once a problem has been successfully formulated in
terms of equations, we need to solve them for the unknowns involved, subject to
the various conditions which are given or implied in the physical problem.
3. Physical Interpretation: After a solution has been obtained, it is useful to
interpret it physically. Such interpretations may be of value in suggesting other
kinds of problems, which could lead to new knowledge of a mathematical or
physical nature.
In this chapter, we shall be mainly concerned with the mathematical formulation of
physical problems in terms of partial differential equations and with the solution of such
equations by methods commonly called Fourier methods.

Definitions of Partial Differential Equations

A partial differential equation is an equation containing an unknown function of two or
more variables and its partial derivatives with respect to these variables. The order of
a partial differential equation is the order of the highest derivative present.
𝜕2 𝑢
Ex: = 2𝑥 − 𝑦 is a partial differential equation of order two, or second-order PDE.
𝜕𝑥𝜕𝑦

Here 𝑢 is the dependent variable while 𝑥 and 𝑦 are independent variables.

⚫ A solution of a PDE is any function which satisfies the equation identically.

⚫ The general solution is a solution which contains a number of arbitrary
independent functions equal to the order of the equation.
⚫ A particular solution is one which can be obtained from the general solution by
particular choice of the arbitrary functions.
1
Ex: As seen by substitution, 𝑢 = 𝑥 2 𝑦 − 𝑥𝑦 2 + 𝐹(𝑥) + 𝐺(𝑦) is a solution of the PDE
2

𝜕2𝑢
= 2𝑥 − 𝑦. Because it contains two arbitrary independent functions 𝐹(𝑥) and
𝜕𝑥𝜕𝑦

𝐺(𝑦), it is the general solution. If in particular 𝐹(𝑥) = 2 sin 𝑥, 𝐺(𝑦) = 3𝑦 4 − 5, we

1
obtain the particular solution: 𝑢 = 𝑥 2 𝑦 − 𝑥𝑦 2 + 2 sin 𝑥 + 3𝑦 4 − 5.
2
⚫ A singular solution is one which cannot be obtained from the general solution by
particular choice of the arbitrary functions.
𝜕𝑢 𝜕𝑢 2
Ex: If 𝑢 = 𝑥 −( ) where 𝑢 is a function of 𝑥 and 𝑦, we see by substitution
𝜕𝑥 𝜕𝑥

that both 𝑢 = 𝑥𝐹(𝑦) − [𝐹(𝑦)]2 and 𝑢 = 𝑥 2 /4 are solutions. The first is the
general solution involving one arbitrary function 𝐹(𝑦). The second, which cannot be
obtained from the general solution by any choice of 𝐹(𝑦), is a singular solution.

⚫ A boundary value problem involving a partial differential equation seeks all

solutions of the equation which satisfy conditions called boundary conditions.
Ex:
(a) Show that 𝑉 = 𝐹(𝑦 − 3𝑥), where 𝐹 is an arbitrary differentiable function, is a
𝜕𝑣 𝜕𝑣
general solution of the equation +3 =0
𝜕𝑥 𝜕𝑦

(b) Find the particular solution which satisfies the condition 𝑣(0, 𝑦) = 4𝑠𝑖𝑛𝑦.

<Sol>:
(a) Let 𝑦 − 3𝑥 = 𝑢, then 𝑣 = 𝐹(𝑢)
𝜕𝑣 𝜕𝑣 𝜕𝑢 𝜕𝑣 𝜕𝑣 𝜕𝑢
= = 𝐹 ′ (𝑢)(−3) = −3𝐹 ′ (𝑢), = = 𝐹 ′ (𝑢)(1) = 𝐹 ′ (𝑢)
𝜕𝑥 𝜕𝑢 𝜕𝑥 𝜕𝑦 𝜕𝑢 𝜕𝑦
𝜕𝑣 𝜕𝑣
∴ +3 =0
𝜕𝑥 𝜕𝑦
Since this equation is of order one, the solution 𝑣 = 𝐹(𝑢) = 𝐹(𝑦 − 3𝑥), which
involves one arbitrary function, is a general solution.
(b) 𝑣(𝑥, 𝑦) = 𝐹(𝑦 − 3𝑥), then 𝑣(0, 𝑦) = 𝐹(𝑦) = 4 𝑠𝑖𝑛 𝑦
∴ 𝑣(𝑥, 𝑦) = 𝐹(𝑦 − 3𝑥) = 4 sin(𝑦 − 3𝑥)

Linear Partial Differential Equations

The general linear partial differential equation of order two in two independent
variables has the form:
𝜕2𝑢 𝜕2𝑢 𝜕2𝑢 𝜕𝑢 𝜕𝑢
𝐴 2
+ 𝐵 + 𝑐 2
+𝐷 +𝐸 + 𝐹𝑢 = 𝐺
𝜕𝑥 𝜕𝑥𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑦
where 𝐴, 𝐵, … … , 𝐺 may depend on 𝑥 and 𝑦, but not on 𝑢.

A second-order equation with independent variables 𝑥 and 𝑦 which does not have
this form is called nonlinear.

- If 𝐺 = 0 identically the equation is called homogencens, while if 𝐺 ≠ 0 it is

called nonhomogeneous.
- Because of the nature of the solutions, the equation is often classified as elliptic,
hyperbolic, or parabolic according as 𝐵2 − 4𝐴𝐶 is less than, greater than, or
equal to zero, respectively.

Ex:
𝜕3 𝑅 𝜕2 𝑅
① 𝑥2 = 𝑦3 linear, order 3, dep. var. 𝑅, ind. var. 𝑥, 𝑦
𝜕𝑦 3 𝜕𝑥 2

𝜕2𝑊
② 𝑊 = 𝑟𝑠𝑡 nonlinear, order 2, dep. var. 𝑊, ind. var. 𝑟, 𝑠, 𝑡
𝜕𝑟 2

𝜕2𝜙 𝜕2𝜙 𝜕2𝜙

③ + + =0 linear, order 2, dep. var. 𝜙, ind. var. 𝑥, 𝑦, 𝑧
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑧 2

𝜕𝑧 2 𝜕𝑧 2
④ ( ) +( ) =1 nonlinear, order 1, dep. var. 𝑧, ind. var. 𝑢,𝑣
𝜕𝑢 𝜕𝑣

Ex:
𝜕2𝜙 𝜕2𝜙
① + =0
𝜕𝑥 2 𝜕𝑦 2

𝐴 = 1, 𝐵 = 0, 𝐶 = 1; 𝐵2 − 4𝐴𝐶 = −4 < 0
∴ The equation is elliptic.

𝜕𝑢 𝜕2𝑢
② =𝑘
𝜕𝑡 𝜕𝑥 2

𝐴 = 𝑘, 𝐵 = 0, 𝐶 = 0; 𝐵2 − 4𝐴𝐶 = 0 (Heat equation)

∴ The equation is parabolic.

𝜕2𝑦 𝜕2 𝑦
③ = 𝛼2
𝜕𝑡 2 𝜕𝑥 2

𝐴 = 𝛼 2 , 𝐵 = 0, 𝐶 = −1; 𝐵2 − 4𝐴𝐶 = 4𝛼 2 > 0 (Wave equation)

∴ The equation is hyperbolic.

𝜕2 𝑢 𝜕2𝑢 𝜕𝑢
④ 𝑥 +𝑦 + 3𝑦 2 =0
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑥

𝐴 = 𝑥, 𝐵 = 0, 𝑐 = 𝑦; 𝐵2 − 4𝐴𝐶 = −4𝑥𝑦
∴ 𝑥𝑦 > 0 → elliptic
𝑥𝑦 = 0 → parabolic
𝑥𝑦 < 0 → hyperbolic
Some Important Partial Differential Equations
𝜕2𝑦 𝜕2 𝑦
(1) Vibrating String Equation = 𝛼2
𝜕𝑡 2 𝜕𝑥 2

This equation is applicable to the small transverse vibrations of a flexible string, such
as a violin string, initially located on the 𝑥-axis and set into motion. The function
𝑦(𝑥, 𝑡) is the displacement of any point 𝑥 of the string at time 𝑡 . The constant
𝛼 2 = 𝜏/𝜇 , where 𝜏 is the tension in the string and 𝜇 is the mass per unit length of
the string. The equation can be easily generalized to higher dimensions, as for
example, the vibrations of a membrane or drumhead in two dimensions:
𝜕2𝑧 2
𝜕2𝑧 𝜕2𝑧
= 𝛼 ( 2 + 2)
𝜕𝑡 2 𝜕𝑥 𝜕𝑦

𝜕𝑢 𝜕𝑢
(2) Heat Conduction Equation = 𝑘∇2 𝑢 If = 0, ∇2 𝑢 = 0
𝜕𝑡 𝜕𝑡

𝑢(𝑥, 𝑦, 𝑧, 𝑡) is the temperature at position (𝑥, 𝑦, 𝑧) at time 𝑡. The constant 𝑘, called

the diffusivity, is equal to 𝐾/𝜎𝜇, where the thermal conductivity 𝐾, the specific heat
𝜎 and the density (mass per unit volume) 𝜇 are assumed constant. We call ∇2 𝑢 the
Laplacian of 𝑢; it is given in three-dimensional rectangular coordinates (𝑥, 𝑦, 𝑧) by:
𝜕2𝑢 𝜕2𝑢 𝜕2𝑢
∇2 𝑢 = + +
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑧 2

(3) Laplace's Equation ∇2 𝑢 = 0

This equation occurs in many fields. In the theory of heat conduction, 𝑢 is the
𝜕𝑢
steady-state temperature, whose equation is obtained by putting = 0 in the heat
𝜕𝑡

conduction equation above. In the theory of gravitation or electricity, 𝑢 represents

the gravitational or electric potential respectively. For this reason, the equation is
often called the potential equation.

𝜕2𝑢 𝜕2𝑢
(4) Longitudinal Vibration of a Beam = 𝑐2 
𝜕𝑡 2 𝜕𝑥 2

This equation describes the motion of a beam which can vibrate longitudinally (in the
𝑥 -direction) with the vibrations being assumed small. The variable 𝑢(𝑥, 𝑡) is the
longitudinal displacement from the equilibrium position of the cross section at 𝑥. The
𝐸
constant 𝑐 2 = , where 𝐸 is the modulus of elasticity and 𝜇 is the density (mass per
𝜇

unit volume).

<Note> This equation is the same as that for a vibrating string. (wave equation).
Methods of Solving Boundary Value Problems
There are many methods by which boundary value problems involving linear partial
differential equation can be solved. In this chapter, we shall be mainly concerned with
two methods which represent somewhat opposing points of view.
⚫ In the first method, we seek to find the general solution of the partial differential
equation and then particularize it to obtain the actual solution by using the
boundary conditions.
⚫ In the second method, we first find particular solutions of the partial differential
equation and then build up the actual solution by use of these particular solutions.
⚫ Of the two methods, the second will be found to be of far greater applicability than
the first.

Method 1 : General Solutions

In this method, we first find the general solution and then that particular solution which
satisfies the boundary conditions.
Theorem 1: (Superposition principle): If 𝑢1 , 𝑢2 , … , 𝑢𝑛 are solutions of a linear
homogeneous partial differential equation, then 𝑐1 𝑢1 + 𝑐2 𝑢2 + ⋯ ⋯ + 𝑐𝑛 𝑢𝑛 where
𝑐1 , 𝑐2 , ⋯ , 𝑐𝑛 are constants, is also a solution.
Theorem 2: The general solution of a linear nonhomogeneous partial differential
equation is obtained by adding a particular solution of the nonhomogeneous equation
to the general solution of the homogeneous equation.
We can sometimes find general solutions by using the methods of ordinary differential
equations.

Ex: (a) Solve the equation ∂2 𝑧/ ∂𝑥 ∂𝑦 = 𝑥 2 𝑦

Ex: (b) Find the particular solution for which
𝑧(𝑥, 0) = 𝑥 2 , 𝑧(1, 𝑦) = cos 𝑦
<Sol>: (a) Write the equation as
∂ ∂𝑧
( ) = 𝑥2𝑦
∂𝑥 ∂𝑦
Integrating with respect to 𝑥, we find
∂𝑧 1 3
= 𝑥 𝑦 + 𝐹(𝑦), where 𝐹(𝑦) is arbitrary
∂𝑦 3
Integrating with respect to 𝑦,
1
𝑧 = 𝑥 3 𝑦 2 + ∫ 𝐹(𝑦)𝑑𝑦 + 𝐺(𝑥), where 𝐺(𝑥) is arbitrary.
6

1 3 2
∴ 𝑧 = 𝑧(𝑥, 𝑦) = 𝑥 𝑦 + 𝐻(𝑦) + 𝐺(𝑥)
6
which has two arbitrary independent functions and is therefore a general solution.
(b) since 𝑧(𝑥, 0) = 𝑥 2 ⇒ 𝐻(0) + 𝐺(𝑥) = 𝑥 2 ⇒ 𝐺(𝑥) = 𝑥 2 − 𝐻(0)
1 2
𝑧(1, 𝑦) = cos 𝑦 ⇒ 𝑦 + 𝐻(𝑦) + 1 − 𝐻(0) = cos 𝑦
6
1 1
∴ 𝑧 = 𝑥 3 𝑦 2 + [cos 𝑦 − 𝑦 2 − 1 + 𝐻(0)] + [𝑥 2 − 𝐻(0)]
6 6
1 1
= 𝑥 3 𝑦 2 + cos 𝑦 − 𝑦 2 + 𝑥 2 − 1
6 6

If 𝐴, 𝐵, ⋯ ⋯ 𝐹 in the linear PDE are constants, then the general solution of the
homogeneous equation can be found by assuming that 𝑢 = 𝑒 𝑎𝑥+𝑏𝑦 , where 𝑎 and 𝑏
are constants to be determined.

Ex: Find a general solution of

∂𝑢 ∂𝑢 ∂2 𝑢 ∂2 𝑢 ∂2 𝑢
(a) 2 +3 = 2𝑢, (b) 4 2 − 4 + =0
∂𝑥 ∂𝑦 ∂𝑥 ∂𝑥 ∂𝑦 ∂𝑦 2
<Sol>:
(a) Let 𝑢 = 𝑐𝑒 𝑎𝑥+𝑏𝑦 , then
2 − 3𝑏
2𝑎 + 3𝑏 = 2 ⇒ 𝑎=
2
(2−3𝑏) 𝑏
∴ 𝑢 = 𝑐𝑒 [ 2 ]𝑥+𝑏𝑦 = 𝑐𝑒 𝑥 𝑒 (2)(2𝑦−3𝑥)
Thus, 𝑢 = 𝑒 𝑥 𝐹(2𝑦 − 3𝑥) is a general solution.

(b) Let 𝑢 = 𝑐𝑒 𝑎𝑥+𝑏𝑦 , then 4𝑎2 − 4𝑎𝑏 + 𝑏 2 = 0 ⇒ 𝑏 = 2𝑎, 2𝑎

∴ 𝑢 = 𝑐𝑒 𝑎(𝑥+2𝑦) , so 𝐹(𝑥 + 2𝑦) is a solution.
By analogy with repeated roots for ordinary differential equation, we might be led to
believe 𝑥𝐺(𝑥 + 2𝑦) or 𝑦𝐺(𝑥 + 2𝑦) to be another solution, and that this is in fact true.
Thus, a general solution is
𝑢 = 𝑐1 𝐹(𝑥 + 2𝑦) + 𝑐2 𝑥𝐺(𝑥 + 2𝑦) or 𝑢 = 𝑐1 𝐹(𝑥 + 2𝑦) + 𝑐2 𝑦𝐺(𝑥 + 2𝑦)