Partial Differential Equations

Module 2
By
Vallampati Ramachandra Prasad
2.1 INTRODUCTION
The reader is familiar with ordinary differential equations. These equations
involve functions of a single variable only and their derivatives. In many
problems that arise in geometry, physics, population dynamics, social sciences,
medicine and engineering, one has to deal with equations containing functions of
more than one variable and partial derivatives with respect to these independent
variables. Such equations are called partial differential equations. Thus a partial
differential equation is an equation of the form
 z z z  2 z  2z 
f  , , ..., 2 ,..., ,..., z, x, y, t   0 (2.1)
 t x y x tx 

containing independent variables t, x, y,… an unknown function z = z(x, y, …,t)

z z z  2 z  2z
and partial derivatives , , ..., 2 ,..., ,..., with respect to these variables
t x y x tx
t, x, y,…
2.2 ORDER, LINEARITY AND HOMOGENEITY OF A PARTIAL
DIFFERENTIAL EQUATION
2.2.1 Order
The order of a partial differential equation is the order of the highest derivative
appearing in it.
2.2.1 Linearity
As in the case of an ordinary differential equations, we say that a partial
differential equation is linear if it is of the first degree in dependent variable (the
unknown function) and its partial derivatives and are not multiplied together.
2.2.3 Homogeneity
A linear partial differential equation is called homogeneous if it contains no term
free from the unknown function and its derivatives; otherwise, it is called a
nonhomogeneous equation.

1
 Partial differential Order Linear/Nonlinear,
equations Homogeneous/Nonhomogeneous
1 z z One Linear, Homogeneous
 z
x y
2 z z One Linear, Homogeneous
x  y  2z
x y
3  2z  2z Two Linear, Homogeneous
 2  0 (Laplace’s
x 2
y
equation)
4  z   z  One Nonlinear, Nonhomogeneous
    4xy
 x   y 
5 z  2z Two Linear, Homogeneous
c 2
(Heat
t x 2
conduction equation)
6  2z  2z Two Linear, Homogeneous
 c2 (Wave
t 2
x 2
equation)
7   2z 2 Two Nonlinear, Nonhomogeneous
 2   xy
 x 
8   2 z  4  z 3 Two Nonlinear, Nonhomogeneous
    
 xy   x 
The following examples of partial differential equations with their order and
nature noted against each of them are meant to illustrate the points explained
above.

2.3 ORIGIN OF PARTIAL DIFFERENTIAL EQUATION

1. Consider the equation
x2+y2+(z-c)2=a2 (2.2)
where c and a are arbitrary constants. It represents the set of all spheres with their
centres on the z-axis.
Differentiating with respect to x and y we get
 z z 
x + p(z-c)=0, y + q(z-c)=0;  p  ,q   (2.3),(2.4)
 x y 

Eliminating the arbitrary constant c from (2.3) and (2.4) we obtain the first-order
linear partial differential equation

2
 y p – x q=0 (2.5)
which characterizes the set of all spheres with centres on the z – axis.

2. Consider again the equation

x2+y2=(z-c)2 tan2𝛼 (2.6)
where c and 𝛼 are arbitrary constants. Equation (2.6) represents the set of all right
circular cones whose axes coincide with the z – axis.
Differentiating (2.6) partially with respect to x and y we get
p (z-c) tan2 𝛼 = x, q (z-c) tan2 𝛼 = y (2.7) (2.8)
respectively. Eliminating the constants c and 𝛼 between these equations we again
obtain equation (2.5).
3. The spheres and cones are surfaces of revolution which have the line OZ as
their axis of symmetry. Now consider the equation
Z=f(x2+y2) (2.9)
where f is an arbitrary function. Equation (2.9) characterizes surfaces of
revolution having z – axis as the axis of symmetry.
Differentiating equation (2.9) partially with respect to x and y we get
p = 2x f ′ (u), q = 2y f ′(u) (2.10), (2.11)
where u = x2+y2 and f ′(u)= df/du. Eliminating f between equations (2.10) and
(2.11), we again obtain the first-order linear partial differential equation (2.5).
Formation of PDE Ordinary differential equations are formed by eliminating
arbitrary constants only, whereas partial differential equations are formed by
eliminating (a) arbitrary constants or (b) arbitrary functions.
We know that the order of an ordinary differential equations is equal to the
number of arbitrary constants to be eliminated from a relation.
In case of partial differential equations, if the number of arbitrary constants to
be eliminated is equal to the number of independent variables involved in a
relation, one obtains a first-order partial differential equations, and if the number
of arbitrary constants to be eliminated is more than the number of independent
variables, then one obtains a higher-order partial differential equation.

3
 Further, if one arbitrary is to be eliminated from a relation, then a first-order
partial differential equation is obtained; and if two arbitrary functions are to be
eliminated, then a second-order partial differential equation is obtained and so on.
A first-order partial differential equation is of the form
f(x, y, z, p, q) = 0 (2.12)
where x, y are independent variables, z is a dependent variable and p = 𝜕z/ 𝜕x
and q =𝜕 z/ 𝜕y.
2.4 FORMATION OF PARTIAL DIFFERENTIAL EQUATION BY
ELIMINATION OF TWO ARBITRARY CONSTANTS
Consider a relation of the type
F(x, y, z, a, b) = 0 (2.13)
where a and b are arbitrary constants.
Differentiating (2.13) partially with respect to x and y we have
F F F F
 p  0 and q 0 (2.14),(2.15)
x z y z

We can now eliminate the two arbitrary constants a and b between the equations
(2.13)- (2.15) and obtain a first order partial differential equation of the form
f (x, y, z, p, q) = 0 (2.16)
We now work out a few examples to show how partial differential equations are
formed by eliminating arbitrary constants.
Examples 3.1
Show by eliminating the arbitrary constants a and b from
(x-a)2+(y-b)2+z2 = 1 (1)
the partial differential equation
z2(p2+q2+1) = 1 (2)
is obtained, which is nonlinear.
Solution
Differentiating (1) partially with respect to x and y
(x-a)+ zp = 0, (y-b)+ zq = 0 (3), (4)
Eliminating a and b from (1), (3) and (4) we get
4
 z2(p2+q2+1) = 1
which is nonlinear.
Examples
Form the partial differential equation by eliminating the arbitrary constants a and
b from each of the relations given below:
Example 2.2
z = ax+ by
Solution We have
z = ax+ by (1)
Differentiating (1) partially with respect to x and y we have
z z
 p  a,  q  b (2), (3)
x y

Eliminating a and b from (1) - (3) we obtain the first order partial differential
equation
z = px + qy (4)
Example 2.3
z = (x2+a2) (y2+b2)
Solution
We have
z = (x2+a2) (y2+b2) (1)
Differentiating (1) partially with respect to x and y we have
z p
 p  2x(y 2  b 2 )  y 2  b 2  (2)
x 2x
z q
 q  2y(x 2  a 2 )  x 2  a 2  (3)
y 2y

Eliminating a and b from (1) - (3) we get

q p
z . or pq = 4xyz (4)
2y 2x

Example 2.4

5
 x 2 y2
z 
a 2 b2

Solution We have
x 2 y2
z  (1)
a 2 b2

Differentiating (1) partially with respect to x and y

z 2x z 2y
p 2 , q 2 (2), (3)
x a y b

Eliminating a and b from (1) - (3) we get

x y
z p  q  px  qy  2z (4)
2 2

Example 2.5
z = (x+a) (y+b)
Solution We have
z = (x+a) (y+b) (1)
Differentiating (1) partially with respect to x and y
z z
 p  y  b,  q  x  a (2), (3)
x y

Eliminating a and b from (1) - (3) we have

z = pq (4)
Example 2.6
2z = (ax+y)2+b
Solution We have
2z = (ax+y)2+b (1)
Differentiating (1) partially with respect to x and y
2p = 2a (ax+y), 2q = 2 (ax+y) (2), (3)
Eliminating a and b from (1) - (3) we have
px + qy = q2 (4)
which is nonlinear.

6
Example 2.7
ax2 + by2 +z2 = 1
Solution We have
ax2 + by2 +z2 = 1 (1)
Differentiating (1) partially with respect to x and y
zp + ax = 0, zq + by = 0 (2), (3)
Multiplying (2) by x and (3) by y and adding
z(px+qy) + (ax2 + by2) = 0
⇒ z(px+qy) = -(ax2 + by2) = - (1 – z2), by (1)
⇒ z(px+qy) = z2- 1 (4)
2.4 FORMATION OF PARTIAL DIFFERENTIAL EQUATION BY
ELIMINATION OF TWO ARBITRARY FUNCTIONS
(a) Elimination of one arbitrary function of the form z = f(u) where u
= u(x, y, z)
Let z = f(u) (2.17)
where f(u) is an arbitrary function of u where u = u(x, y, z) a known function of
x, y and z.
Differentiating (2.17) partially with respect to x and y
 u u   u u 
p  f    p,q  f    q  (2.18), (2.19)
 x z   y z 

where ′ on f denotes differentiation with respect to the argument u. Eliminating f

from equations (2.18)- (2.19) we obtain a first order partial differential equation.
(b). Elimination of two arbitrary function of the form z = f(x) . f(y)
Let z = f(x) . f(y) (2.20)
where f is a function of x and g is a function of g alone.
Differentiating (2.17) partially with respect to x and y, respectively, we have
z z  2z
 p  f g,  q  fg,  s  f g
x y xy (2.21)
pq  f g.fg  f g.fg  sz

7
or pq = sz
 2z
where s =
xy

(c) Formation of PDE by elimination of arbitrary function of the form F(u,

v)= 0
Consider a relation between x, y and z of the type
F(u, v) = 0 (2.21)
where u and v are known functions of x, y and z; and F is an arbitrary function of
u and v. Also, z is a function of x and y.
Differentiating (2.22) by chain rule, with respect to x and y we obtain
F  u u  F  v v 
  p    p  0 (2.23)
u  x z  v  x z 

F  u u  F  v v 
  q   q  0 (2.24)
u  y z  v  y z 

Eliminating F between (2.23) and (2.24) we get

u x  u zp vx  vzp
0 (2.25)
u y  u zq v y  vzq

u u
where u x  ,uy  and so on.
x y

Expanding the determinant in (2.25), collecting the terms and simplifying we

have
Pp + Qq = R (2.26)
(u, v) (u, v) (u, v)
where P  ,Q  ,R  which are functions of x, y and z and do
(y, z) (z, x) (x, y)
not contain p and q. This equation is called Lagrange’s linear equation. In fact, it
is a quasi-linear equation since the dependent variable may be present in P, Q and
R.
If P and Q are independent of z and R is linear in z the equation (2.26) is called
linear.
Equation (2.26) is of the type given by equation (2.12), which need not be
linear, as Example 2.1 above will illustrate.

8
Examples
Form the partial differential equation by eliminating the arbitrary f from each of
the following relations [Example 2.8 – 2.15 except (Example 2.13) in which
functions f and g have to be eliminated.]
Example 2.8
z= x f(x+y)
Solution We have
z= x f(x+y) (1)
Differentiating (1) partially with respect to x and y
p = 1f+x f ′, q = 0+x f ′ (2), (3)
Eliminating f between (1) - (3) we have
z
p  q  z = (p - q)x (4)
x
which is the required partial differential equation.
Example 2.9
z = x – y + f(xy)
Solution We have
z = x – y + f(xy) (1)
Differentiating (1) partially with respect to x and y
p = 1 + f ′ y, q = -1 + f ′ x (2), (3)
Eliminating f ′ between (1) and (2) we get the partial differential equation
px – qy = x + y (4)
Example 2.10
z = eax-by f(ax+by)
Solution We have
z = eax-by f(ax+by) (1)
Differentiating (1) partially with respect to x and y
p = a eax-by f + eax-by f ′ . a

9
 q = -b eax-by f + eax-by f ′ .b (2), (3)
Eliminating f between (1) - (3) we have
bp – aq = 2abz (4)
Example 2.11

z = xn f  
y
x

Solution We have

z = xn f  
y
(1)
x

Differentiating (1) partially with respect to x and y

 y  1
p  nx n 1f  x n f   2  , q  x n f . (2), (3)
x  x

We have from (2) and (3)

px+qy  nx n f  yx n 1f   yx n 1f   nz
⇒px+qy = nz, by (1) (4)
Example 2.12
 z 
yz + zx + xy= f  
xy

Solution We have
 z 
yz + zx + xy= f   (1)
xy

Differentiating (1) partially with respect to x and y

(x  y)p  z
y  z  (x  y)p  f . ,
(x  y) 2
(2), (3)
(x  y)q  z
z  x  (x  y)q  f .
(x  y) 2

Eliminating f ′ between (2) and (3) we have

y  z  (x  y)p (x  y)p  z y  2z 𝑑𝑖𝑓𝑓. 𝑜𝑓 𝑛𝑟𝑠.
  (= )
x  z  (x  y)q (x  y)q  z x  2z 𝑑𝑖𝑓𝑓. 𝑜𝑓 𝑑𝑟𝑠.

(x  y)[p(x  2z)  q(y  2z)]  z[x  2z  (y  2z)]

10
  px  qy  2z(p  q)  z(x  y) / (x  y)

Example 2.13
z = f(x+at) + g(x-at)
Solution We have
z = f(x+at) + g(x-at) (1)
Differentiating (1) partially with respect to x and t twice
z z
 f   g,  af   ag, (2)
x t

 2z  2z
 f   g,  a 2f   a 2g (3)
x 2
t 2

Eliminating the arbitrary functions from (3) we have

 2z 2  z
2
a (4)
t 2 x 2

which is a partial differential equation of the second order.

Example 2.14

z = y2+2f   log y 
1
x 

Solution We have

z = y2+2f   log y 
1
(1)
x 

Differentiating (1) with respect to x and y

 1  1
p  0  2f .   2  , q  2y  2f .   (2), (3)
 x  y

Eliminating f ′ between (2) and (3) we have

px2 + qy=2y2 (4)
which is the required partial differential equation.
Example 2.15
f(xy+z2, x+y+z) = 0
Solution We have

11
 f(u, v) = 0 (1)
where u = xy + z2, v = x+y+z (2)
We know that the partial differential equation is
(u, v) (u, v) (u, v)
p q
(y, z) (z, x) (x, y)

where
 (u, v) x 2z
  x  2z;
 (y, z) 1 1
 (u, v) 2z y
  2z  y;
 (z, x) 1 1

 (u, v) y x
  yx (3)
 (x, y) 1 1

Hence the required partial differential equation is

(x-2z)p – (y-2z)q = y-x (4)
2.6 CLASSIFICATION OF FIRST-ORDER PARTIAL DIFFERENTIAL
EQUATIONS
2.6.1 Linear Equation
Let z = z(x, y) be a function of two variables x and y. A first-order partial
differential equation is said to be linear if it is linear in p, q and z, i.e. if it is of
the form
P(x, y)p+ Q(x, y)q = R(x, y)z+ S(x, y)
𝜕𝑧 𝜕𝑧
where p = and q = (2.27)
𝜕𝑥 𝜕𝑦

yp − xq = xyz + x
E.g. } (2.28)
𝑦 2 𝑝 − 𝑥𝑦𝑞 = 𝑥 (𝑧 − 2𝑦)
2.6.2 Semi-Linear Equation
A first order partial differential equation is said to be semi - linear if it is linear in
p and q and the coefficients of p and q are functions of x and y alone, i.e. if it is
of the form
P(x, y)p + Q(x, y)q = R(x, y, z) (2.29)
E.g. 1. exp – yxq = xz2

12
 2. √𝑥 p + √𝑦 q = √𝑧 (2.30)
2.6.3 Quasi-Linear Equation
A first order partial differential equation is said to be quasi- linear if it is linear in
p and q i.e. if it is of the form
P(x, y, z)p + Q(x, y, z)q = R(x, y, z) (2.31)
E.g. 1. (x2-yz) p + (y2 - zx) q = z2 – xy
2. (x2+ y2)p – xyq = z3x+y2 (2.32)
Note that every semi-linear partial differential equation is quasi- linear.
2.6.4 Nonlinear Equation
A first order partial differential equation is said to be nonlinear if it does not fall
into any one of the above types.
2.7 CLASSIFICATION OF SOLUTIONS OF FIRST-ORDER PARTIAL
DIFFERENTIAL EQUATIONS
Consider a first-order partial differential equation of the form
f(x, y, z, p, q) = 0 (2.33)
Let z = z(x, y) (2.34)
be a continuously differentiable function of x and y in a region D.
𝜕𝑧 𝜕𝑧
Compute p = and q = from (2.34) and substitute in (2.33). If this reduces
𝜕𝑥 𝜕𝑦
(2.33) into an identity in x and y then (2.34) defines a solution of partial
differential equation..
Equation (2.34) defines a surface in three-dimensional space. It is called an
integral surface of (2.33).
There are different types of solutions of equation (2.33).
2.7.1 Complete Integral
A two-parameter family of solutions
z = F(x, y, z, a, b) (2.35)
is called a complete integral (or complete solution) of (2.33) if in the region D the
matrix

13
  F  2F  2F 
 a ax ay 
M=  (2.36)
 F  2F  2F 
 
 a bx by 

is of rank two.
2.7.2 General Integral
Let a and b be related and suppose
b =  (a) (2.37)
substitute this in (2.35) we get
z = F(x, y, a,  (a)) (2.38)
which is a one-parameter family of solutions of (2.33).
This is a sub-family of the two-parameter family given by (2.35). The envelope
of (2.38), if it exists, is obtained by eliminating a between (2.38) and
F F
 (a)  0 (2.39)
a b
If (2.39) can be solved for a then
a = a(x, y) (2.40)
Substituting for a in (2.38) we obtain an integral surface as
z = F(x, y, a(x, y),  (a(x, y))) (2.41)
If  is arbitrary then (2.41) is called a general integral (or general solution) of the
partial differential equation of the form (2.33).
2.7.3 Particular Integral
When a particular function  is used in (2.41) we obtain a particular integral or
particular solution of the partial differential equation of the form (2.33).
2.7.4. Singular Integral
In some cases we find another integral which satisfies the partial differential
equation (2.33) but is not a particular integral of (2.33). If it exists, it is obtained
by eliminating a and b from
F F
z = F(x, y, a, b),  0 and 0 (2.42), (2.43), (2.44)
a b

14
and it is the envelope of the two-parameter family of surfaces
z = F (x, y, a, b) (2.45)
and it is called a singular integral (or a singular solution).
In a sense, a general integral provides a much broader class of solutions of the
partial differential equation than does a complete integral.
However, it is possible to derive a general integral when a complete integral is
known.
Note that for the partial differential equation
z = px + qy (2.46)
The relation z = ax + by (2.47)
is a complete integral while the relation
y
z  yf   (2.47)
x

is a general integral.
2.8 EQUATIONS SOLVABLE BY DIRECT INTEGRATION
We now consider partial differential equations which can be solved by direct
integration. While carrying out integration with respect to a variable the other
variable is held fixed. So, in place of constant of integration we have to add an
arbitrary function of the variable held fixed.
Example 2.16
 2z
Solve 2  xy 2
x

Solution The given partial differential equation is

 2z
 xy 2 (1)
x 2

Integrating once partially with respect to x we get

z x 2 2
 .y  f (y) (2)
x 2

and integrating (2) partially with respect to x we get the solution as

15
 x3 2
z .y  xf (y)  g(y) (3)
6

where f and g are arbitrary functions of y.

Example 2.17
 2z
Solve  sin x tan y
xy

Solution The given partial differential equation is

 2z
 sin x tan y (1)
xy

Integrating (1) partially with respect to x

z = -cos x. log sec y + g(y) + h(x) (2)
where g(y) =  f (y)dy and h(x) are arbitrary functions of y and x, respectively.

Example 2.18
 2z
Solve  2xe y
yx

Solution The given partial differential equation is

 2z
 2xe y (1)
yx

Integrating (1) partially with respect to y we get

z
 2xe y  f (x) (2)
x

where f(x) is an arbitrary function of x. Integrating (2) partially with respect to x

we get
z  x 2 e y   f (x)dx  g(y) (3)

which is the solution of (1). We can write (3) as

z  x 2e y  h(x)  g(y) (4)
where h(x) and g(y) are arbitrary functions of x and y, respectively.

16
Example 2.19
 3z
Solve  cos(2x  3y)
x 2y

Solution The given partial differential equation is

 3z
 cos(2x  3y) (1)
x 2y

Integrating (1) partially with respect to x

 3z 1
 sin(2x  3y)  f (y) (2)
xy 2

where f(y) is an arbitrary function of y. Integrating (1) partially with respect to x

again
z 1
  cos(2x  3y)  xf (y)  g(y) (3)
y 4

where g is an arbitrary function of y. Integrating (3) partially with respect to y,

1
z sin(2x  3y)  x  f (y)dy   g(y)dy  h(x)
12
which can be put in the form
1
z  f1 (x)  f 2 (y)  xf 3 (y)  sin(2x  3y) (4)
12
where f1(x) is an arbitrary function of x and f2(y) and f3(y) are arbitrary functions
of y.
Example 2.20
 2z z z
Solve  a 2 z  0 . If  a sin y and  0 when x = 0.
x 2
x y

Solution The given partial differential equation is

 2z
 a 2z  0 (1)
x 2

We have to solve (1) under the conditions:

z z
When x = 0,  a sin y and 0 (2), (3)
x y

If we treat z as a function of x alone, the solution of (1) is

17
 z = A sinhx + B coshx (4)
where A and B are constants. Since z is a function of y also we can take z as
z = A(y) sinhx + B(y) coshx
z
Now  A(y) cosh x  B(y)sinh x x 0
x x 0

⇒A.1+B.0=a siny
z
 A(y)sin x  B(y) cosh x x 0
y x 0

⇒A′.0+B′.1 = 0
∴ A(y) = a siny and B(y) = constant.
Hence z = (a sin y) sinhx + b coshx,
where b is an arbitrary constant.
2.9 QUASI-LINEAR EQUATIONS OF FIRST ORDER
Quasi-linear partial differential equations of first order can be written as
Pp + Qq = R (2.49)
where P, Q, R are functions of x, y and z. They do not involve p or q. Equation
(2.49) is called Lagrange’s equation. Here, linear means that p and q appear to
the first degree only. This is in contrast to the situation in ordinary differential
equations where z must also be of first degree.
Note that partial differential equation
z z
x  y  z2  x 2 (2.50)
x y

is linear while the ordinary differential equation

dz
x  z2  x 2 (2.51)
dx
is nonlinear.
Theorem 2.1 The general solution of the quasi-linear partial differential equation
is
Pp + Qq = R (2.52)
F(u, v) = 0 (2.53)

18
where F is an arbitrary function and u(x, y, z) = c 1 and v(x, y, z) = c2 form a
solution of the equations
dx dy dz
  (2.54)
P Q R

Proof:
If u(x, y, z) = c1, v(x, y, z) = c2 (2.55)
Satisfy equations (2.54) then
ux dx + uy dy + uz dz = 0, (2.56)
and vx dx + vy dy + vz dz = 0 (2.57)
must be compatible with (2.54) so that we must have
P ux dx +Ruy dy + Ruz dz = 0, Pvx dx +Q vy dy + Rvz dz = 0 (2.58), (2.59)
From (2.58) and (2.59) we have
P Q R
  (2.60)
 (u, v)  (u, v) (u, v)
 (y, z)  (z, x) (x, y)

Now, differentiating (2.53) with respect to x and y

F  u u  F  v v 
  p    p  0 (2.61)
u  x z  v  x z 

F  u u  F  v v 
  q   q  0 (2.62)
u  y z  v  y z 

F F
Eliminating and from (2.61)- (2.62) we get
u v
u u v v
 p  p
x z x z
0 (or)
u u v v
 q  q
y z y z

(u, v) (u, v) (u, v)

p q
(y, z) (z, x) (x, y)

Pp + Qq = R (2.63)
(u, v) (u, v) (u, v)
where P  ,Q  ,R  are functions of x, y, z.
(y, z) (z, x) (x, y)

19
 Substituting from (2.60) into (2.63) we see that (2.53) is a solution of (2.49) if u
and v are given by (2.55).
2.10 SOLUTION OF LINEAR, SEMI-LINEAR AND QUASI-LINEAR
EQUATIONS
If the equation is linear or semi-linear or can be written in these forms then we
can solve them as follows.
2.10.1 All the Variables are Separable
Example 2.21
Solve: zx2p + zy2q = 1 – z2
Solution The equation can be thrown into the form
1 z2
x 2 p  y 2q  (1)
z

which is a semi-linear equation.

Lagrange’s auxiliary equations are
dx dy zdz
  (2)
x 2 y2 1  z2

Integrating the first and last equations we have

1 1 1 1
  c1 ,  log(1  z 2 )  log c 2
x y y 2

 1y 
 e 
or    c respectively (3), (4)
1 z2
2

General solution is
  1y  
  e  
1 1   
F  , 0 (5)
x y 1 z2 
 
 

2.10.2 Two Variables are Separable

Example 2.22
Solve: mp - lq=z tan (lx + my).

20
Solution This is a semi-linear equation.
Lagrange’s auxiliary equations are
dx dy dz
  (1)
m l z tan(lx  my)

From the first equation we have

lx + mdy = 0 ⇒ lx + my = c1 (2)
Using this in the last equation
dy dz
 (since lx + my =)
l z tan c1

ldz
 tan c1dy  0 (3), (4)
z
Integrating we have
y tan (lx+my) + l logz = log c2.
Or 𝑧 𝑙 . 𝑒 𝑦(tan(𝑙𝑥+𝑚𝑦)) = 𝑐2 (5)
𝐹(𝑙𝑥 + 𝑚𝑦, 𝑧 𝑙 . 𝑒 𝑦(tan(𝑙𝑥+𝑚𝑦))) = 0 (6)
Example:2.23
Solve: 𝒚𝒑 − 𝒙𝒒 = 𝒙𝒚𝒛 + 𝒙𝒚
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑦 −𝑥 𝑥𝑦𝑧+𝑥𝑦

From the first equation we have

𝑥𝑑𝑥 + 𝑦𝑑𝑦 = 0 (2)
Integrating,
𝑥 2 + 𝑦 2 = 𝐶1 (3)
From the last equation
𝑑𝑦 𝑑𝑧 𝑑𝑧
= ⇒ 𝑦𝑑𝑦 + =0 (4)
−1 𝑦(𝑧+1) 𝑧+1

Integrating,

21
𝑦2 𝑦 2⁄
+ log(𝑧 + 1) = log 𝐶2 ⇒ (𝑧 + 1)𝑒 2 = 𝐶2 (5)
2

General solution is
𝑦 2⁄
2 2
𝐹 (𝑥 + 𝑦 , (𝑧 + 1)𝑒 2) =0 (4)

Method of Multipliers
In addition to the above methods, we can apply the following method called
method of multipliers.
Example:2.24
Solve: 𝒙𝟐 (𝒚 − 𝒛)𝒑 + 𝒚𝟐 (𝒛 − 𝒙)𝒒 = 𝒛𝟐 (𝒙 − 𝒚)
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑥 2 (𝑦−𝑧) 𝑦 2 (𝑧−𝑥) 𝑧 2 (𝑥−𝑦)

1 1 1 1 1 1
Using multipliers ( , , ) and ( 2 , , 2 )in turn we get
𝑥 𝑦 𝑧 𝑥 𝑦2 𝑧
1 1 1
𝑥
𝑑𝑥+𝑦𝑑𝑦+𝑧 𝑑𝑧
Each ratio =
𝑥(𝑦−𝑧)+𝑦(𝑧−𝑥)+𝑧(𝑥−𝑦)=0

𝑑𝑥 𝑑𝑦 𝑑𝑧
⇒ + + =0 (2)
𝑥 𝑦 𝑧

⇒ log 𝑥 + log 𝑦 + log 𝑧 = log 𝐶1 (3)

⇒ 𝑥𝑦𝑧 = 𝐶1
1 1 1
𝑑𝑥+ 2 𝑑𝑦+ 2 𝑑𝑧
𝑥2 𝑦 𝑧
Each ratio =
𝑥(𝑦−𝑧)+𝑦(𝑧−𝑥)+𝑧(𝑥−𝑦)=0
1 1 1
𝑑𝑥 + 𝑑𝑦 + 𝑑𝑧 = 0 (4)
𝑥2 𝑦2 𝑧2

1 1 1
+ + = 𝐶2 (5)
𝑥 𝑦 𝑧

General solution is
1 1 1
𝐹 (𝑥𝑦𝑧, + + ) = 0 (6)
𝑥 𝑦 𝑧

Example: 2.25

22
Solve: 𝒛𝒙𝒑 + 𝒛𝒚𝒒 = 𝟏 + 𝒛𝟐
Solution: Lagrange’s auxiliary equation are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑧𝑥 𝑧𝑦 1+𝑧 2

𝑑𝑥 𝑑𝑦 𝑧𝑑𝑧
⇒ = =
𝑥 𝑦 1+𝑧 2

Taking the first two ratios,

𝑑𝑥 𝑑𝑦 𝑥
⇒ + = 0 ⇒ log 𝑥 = log 𝑦 + 𝑐𝑜𝑛𝑠𝑡. , on integration ⇒ = 𝐶1 (2)
𝑥 𝑦 𝑦

Taking the last two ratios,

𝑑𝑦 𝑧𝑑𝑧 1
= ⇒ log 𝑦 = log(1 + 𝑧 2) + constant(on integration)
𝑦 1+𝑧 2 2

𝑦
⇒ = 𝐶2 (3)
√1+𝑧 2

General solution is
𝑥 𝑦
𝐹( , )=0 (4)
𝑦 √1+𝑧 2

Example:2.26
Solve: 𝒚𝒑 + 𝒙𝒒 = (𝒙𝟐 − 𝒚𝟐 + 𝒛𝟐 )𝒙
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (𝑥 2 (1)
𝑦 𝑥 −𝑦 2 +𝑧 2 )𝑥

From the first two ratios we get

𝑥𝑑𝑥 − 𝑦𝑑𝑦 = 0 ⇒ 𝑥 2 − 𝑦 2 = 𝑎2 (𝑎2 is an arbitrary constant)
considering the last two ratios we have
𝑑𝑧
𝑑𝑦 = ∵ 𝑥 2 − 𝑦 2 = 𝑎2 (2)
𝑎2 +𝑧 2

Integrating,
1 𝑧 1 𝑧
𝑦 = tan−1 + 𝑏 = tan−1 + 𝑏 (b is an arbitrary constant) (3)
𝑎 𝑎 √𝑥 2 −𝑦 2 √𝑥 2 −𝑦 2

General solution is

23
 1 𝑧
𝐹 (𝑥 2 − 𝑦 2 , 𝑦 − tan−1 )=0 (4)
√𝑥 2 −𝑦 2 √𝑥 2 −𝑦 2

Example:2.27
Solve: 𝒚𝒑 + 𝒙𝒒 = (𝒙 + 𝒚)𝒛
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (𝑥+𝑦)𝑧 (1)
𝑦 𝑥

From the first two ratios we get

𝑥𝑑𝑥 − 𝑦𝑑𝑦 = 0 ⇒ 𝑥 2 − 𝑦 2 = 𝐶1
Again, each ratio
𝑑𝑥+𝑑𝑦 𝑑𝑧 𝑑𝑧
= = (𝑥+𝑦)𝑧
⇒ 𝑑 (𝑥 + 𝑦) = , cancelling (𝑥 + 𝑦)
𝑦+𝑥 𝑧

On integration we have
𝑥 + 𝑦 − log 𝑧 = 𝐶2 (3)
General solution is
𝐹 (𝑥 2 − 𝑦 2 , 𝑥 + 𝑦 − log 𝑧) = 0 (4)
Example:2.28
Solve: 𝒙𝒑 − 𝒚𝒒 = 𝒙𝒚𝒛
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑥 −𝑦 𝑥𝑦𝑧

From the first two ratios

𝑑𝑥 𝑑𝑦
= ⇒ 𝑥𝑑𝑦 + 𝑦𝑑𝑥 = 0 ⇒ 𝑥𝑦 = 𝐶1 (2)
𝑥 −𝑦

From the first two ratios

𝑑𝑦 𝑑𝑧
= = ∵ 𝑥𝑦 = 𝐶1
−𝑦 𝐶1 𝑧

Integrating
1 1
log 𝑧 + log 𝑦 =constant ⇒ log 𝑧 + log 𝑦 = 𝐶2 (3)
𝐶1 𝑥𝑦

24
General solution is
1
𝐹 (𝑥𝑦, log 𝑧 + log 𝑦) = 0 (4)
𝑥𝑦

Example: 2.29
Solve: 𝒛𝒑 + 𝒚𝒒 = 𝒙
Solution: Lagrange’s auxiliary equation are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑧 𝑦 𝑥

𝑑𝑥 𝑑𝑦 𝑧𝑑𝑧
⇒ = =
𝑥 𝑦 1+𝑧 2

Taking the first two ratios,

𝑑𝑥 𝑑𝑦
⇒ + ⇒ 𝑥𝑑𝑥 − 𝑧𝑑𝑧 = 0
𝑧 𝑥

Integrating we get
𝑥 2 − 𝑧 2 = 𝐶1 (2)
𝑑𝑦 𝑑𝑥+𝑑𝑦+𝑑𝑧
Each of the ratios(1) = =
𝑦 𝑧+𝑦+𝑥

Integrating we get
log 𝑦 = log(𝑧 + 𝑦 + 𝑥 ) + log 𝐶2
𝑦
⇒ (𝑥+𝑦+𝑧) = 𝐶2 (3)

General solution is
𝑦
𝐹 (𝑥 2 − 𝑧 2, (𝑥+𝑦+𝑧)) = 0 (4)

Example:2.30
Solve: (𝒚 − 𝒛)𝒑 + (𝒛 − 𝒙)𝒒 = (𝒙 − 𝒚)
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
(𝑦−𝑧)
= (𝑧−𝑥) = (𝑥−𝑦) (1)

Taking (1,1,1) as Lagrange’s multipliers, each ratio

𝑑𝑥+𝑑𝑦+𝑑𝑧
= (𝑦−𝑧)+(𝑧−𝑥)+(𝑥−𝑦)=0

25
 ⇒ 𝑑(𝑥 + 𝑦 + 𝑧) = 0 ⇒ 𝑥 + 𝑦 + 𝑧 = 𝐶1 (2)
Taking (𝑥, 𝑦, 𝑧) as Lagrange’s multipliers, each ratio
𝑥𝑑𝑥+𝑦𝑑𝑦+𝑧𝑑𝑧
=
𝑥(𝑦−𝑧)+𝑦(𝑧−𝑥)+𝑧(𝑥−𝑦)=0

⇒ 𝑑(𝑥 2 + 𝑦 2 +𝑧 2 ) = 0 ⇒ 𝑥 2 + 𝑦 2 +𝑧 2 = 𝐶2 (3)
General solution is
𝐹 (𝑥 + 𝑦 + 𝑧, 𝑥 2 + 𝑦 2 +𝑧 2 ) = 0 (4)
Example:2.31
Solve: (𝐬𝐞𝐜 𝒙)𝒑 + (𝐬𝐢𝐧 𝒙 − 𝒚 𝐬𝐞𝐜 𝒙 𝐭𝐚𝐧 𝒙 )𝒒 = (𝒂𝟐 − 𝒛𝟐 )
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
sec 𝑥 sin 𝑥−𝑦 sec 𝑥 tan 𝑥 𝑎2 −𝑧 2

From the first two ratios

𝑑𝑧 𝑑𝑧
cos 𝑥 𝑑𝑥 = 𝑜𝑟 (cos 𝑥 )𝑑𝑥 =
𝑎2 − 𝑧 2 𝑎2 − 𝑧 2
On integration, we get
1 𝑎+𝑧
sin 𝑥 = log + 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡
2𝑎 𝑎−𝑧
𝑎+𝑧
Or 2𝑎(sin 𝑥) − log [ ] = 𝐶1 (2)
𝑎−𝑧

From the first two ratios

𝑑𝑦
sec 𝑥 = sin 𝑥 − 𝑦 sec 𝑥 tan 𝑥
𝑑𝑥
𝑑𝑦
Or sec 𝑥 + 𝑦 sec 𝑥 tan 𝑥 = sin 𝑥
𝑑𝑥

Or d(ysec 𝑥) − sin 𝑥 𝑑𝑥 = 0
Integrating we have
y sec 𝑥 + cos 𝑥 = 𝐶2 (3)
General solution is
𝑎+𝑧
𝐹 (2𝑎(sin 𝑥) − log [ ] , y sec 𝑥 + cos 𝑥) = 0 (4)
𝑎−𝑧

26
Example:2.32
Solve: 𝒙(𝟏 − 𝒙𝒚)𝒑 − 𝒚(𝟏 + 𝒙𝒚)𝒒 = 𝒛(𝟏 − 𝒙𝒚)
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑥(1−𝑥𝑦) −𝑦(1+𝑥𝑦) 𝑧(1−𝑥𝑦)

𝑑𝑥 𝑑𝑧
From the first and last ratios = , cancelling(1 − 𝑥𝑦)
𝑥 𝑧
𝑥
Integrating log 𝑥 = log 𝑧 + log 𝐶1 ⇒ = 𝐶1 (2)
𝑧

On cross multiplication the first two ratios yield

−𝑦 𝑑𝑥 − 𝑥𝑦 2 𝑑𝑥 = 𝑥𝑑𝑦 − 𝑥 2𝑦𝑑𝑦
⇒ 𝑥 2𝑦𝑑𝑦 − 𝑥𝑦 2 𝑑𝑥 = 𝑦𝑑𝑥 + 𝑥𝑑𝑦, on transposition
𝑑𝑦 𝑑𝑥 𝑑(𝑥𝑦) 1
⇒ − = (𝑥𝑦 2 )
multiplying by (𝑥𝑦 2 )
𝑦 𝑥

On integration we get
1
log 𝑦 − log 𝑥 = − +constant
𝑥𝑦

1 𝑦
+ log ( ) = 𝐶2 (3)
𝑥𝑦 𝑥

General solution is
𝑥 1 𝑦
𝐹( , + log ( )) = 0 (4)
𝑧 𝑥𝑦 𝑥

Example:2.33
Solve: (𝒙𝟐 − 𝒚𝟐 − 𝒛𝟐 )𝒑 + 𝟐𝒙𝒚𝒒 = 𝟐𝒙𝒛
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
(𝑥 2 −𝑦 2 −𝑧 2 ) 2𝑥𝑦 2𝑥𝑧

𝑑𝑦 𝑑𝑧
Considering the last two ratios, we have = , On integration we get
𝑦 𝑧
𝑦
log 𝑦 = log 𝑧 +constant ⇒ = 𝐶1 (2)
𝑧

Using (𝑥, 𝑦, 𝑧) as Lagrange’s multipliers, we get that each ratio

27
 𝑥𝑑𝑥+𝑦𝑑𝑦+𝑧𝑑𝑧 𝑑𝑧
= =
𝑥(𝑥 2 +𝑦 2 +𝑧 2 ) 2𝑥𝑧

𝑑(𝑥 2 +𝑦 2 +𝑧 2 ) 𝑑𝑧
or = , on multiplication by 2𝑥
(𝑥 2 +𝑦 2 +𝑧 2 ) 𝑧

this gives, on integration,

log 𝑥(𝑥 2 + 𝑦 2 + 𝑧 2 ) − log 𝑧 = log 𝐶2
(𝑥 2 +𝑦 2 +𝑧 2 )
Or = 𝐶2 (3)
𝑧

General solution is
𝑦 (𝑥 2 +𝑦 2 +𝑧 2 )
𝐹( , )=0 (4)
𝑧 𝑧

Example:2.34
Solve: (𝒛𝟐 − 𝟐𝒚𝒛 − 𝒚𝟐 )𝒑 + (𝒙𝒚 + 𝒛𝒙)𝒒 = 𝒙𝒚 − 𝒛𝒙
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= (𝑥𝑦+𝑧𝑥) = (1)
(𝑧 2 −2𝑦𝑧−𝑦 2 ) 𝑥𝑦−𝑧𝑥

Using (𝑥, 𝑦, 𝑧) as Lagrange’s multipliers, each ratio

𝑥𝑑𝑥+𝑦𝑑𝑦+𝑧𝑑𝑧
=
𝑥𝑧 2 −2𝑥𝑦𝑧−𝑥𝑦 2 +𝑥𝑦 2 +𝑥𝑦𝑧+𝑥𝑦𝑧−𝑧 2 𝑥=0

⇒ 𝑥 2 + 𝑦 2 + 𝑧 2 = 𝐶1 (2)
Considering the last two ratios, we have
𝑑𝑦 𝑑𝑧 𝑑𝑦 − 𝑑𝑧
= = ⇒ 2𝑧𝑑𝑧 = (𝑦 − 𝑧)𝑑 (𝑦 − 𝑧)
𝑦+𝑧 𝑦−𝑧 2𝑧
On integrating and simplifying
𝑦 2 − 2𝑦𝑧 − 𝑧 2 = 𝐶2 (3)
The general solution is
𝐹 (𝑥 2 + 𝑦 2 + 𝑧 2, 𝑦 2 − 2𝑦𝑧 − 𝑧 2 ) = 0 (4)
Example:2.35
Solve: 𝒙𝒏 𝒑 + 𝒚𝒏 𝒒 = 𝒛𝒏
Solution: Lagrange’s auxiliary equations are

28
 𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑥𝑛 𝑦𝑛 𝑧𝑛

On integration we get two solutions

𝑥 −𝑛+1 − 𝑦 −𝑛+1 = 𝐶1 , 𝑥 −𝑛+1 − 𝑧 −𝑛+1 = 𝐶2(𝑛 ≠ 1) (2),(3)
𝑥 𝑥
⇒ = 𝐶1, ⇒ = 𝐶2 (4),(5)
𝑦 𝑧

General solution when 𝑛 ≠ 1 is

𝐹 (𝑥 −𝑛+1 − 𝑦 −𝑛+1 = 𝐶1 , 𝑥 −𝑛+1 − 𝑧 −𝑛+1) = 0 and (6)
𝑥 𝑥
When 𝑛 ≠ 1 is 𝐹 ( , ) = 0 (7)
𝑦 𝑧

Example:2.36
Solve: 𝒙(𝒚 − 𝒛)𝒑 + 𝒚(𝒛 − 𝒙)𝒒 = 𝒛(𝒙 − 𝒚)
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑥(𝑦−𝑧) 𝑦(𝑧−𝑥) 𝑧(𝑥−𝑦)

1 1 1 𝑑𝑥 𝑑𝑦 𝑑𝑧
𝑑𝑥 𝑑𝑦 𝑑𝑧 +𝑦+𝑧
𝑥 𝑦 𝑧 𝑥
Now, each ratio = (𝑦−𝑧) = (𝑧−𝑥) = (𝑥−𝑦) = ⇒ 𝑥𝑦𝑧 = 𝐶1 (2)
0

𝑑𝑥+𝑑𝑦+𝑑𝑧
Also each ratio =
0

⇒ 𝑥 + 𝑦 + 𝑧 = 𝐶2 (3)
General solution is
𝐹 (𝑥𝑦𝑧, 𝑥 + 𝑦 + 𝑧) = 0 (4)
Example:2.37
Solve: (𝒙𝟐 − 𝒚𝒛)𝒑 + (𝒚𝟐 − 𝒛𝒙)𝒒 = 𝒛𝟐 − 𝒙𝒚
Solution: Lagrange’s auxiliary equations are
𝒅𝒙 𝒅𝒚 𝒅𝒛
(𝒙𝟐 −𝒚𝒛)
= (𝒚𝟐 = (1)
−𝒛𝒙) 𝒛𝟐 −𝒙𝒚

𝑥𝑑𝑥+𝑦𝑑𝑦+𝑧𝑑𝑧
Each ratio=
𝑥 3 +𝑦 3 +𝑧 3 −3𝑥𝑦𝑧

𝑑𝑥+𝑑𝑦+𝑑𝑧
=
𝑥 2 +𝑦 2 +𝑧 2 −𝑥𝑦−𝑦𝑧−𝑧𝑥

29
 𝑥𝑑𝑥+𝑦𝑑𝑦+𝑧𝑑𝑧 𝑑𝑥+𝑑𝑦+𝑑𝑧
⇒ = ` (2)
𝑥+𝑦+𝑧 1

∵ 𝑥 3 + 𝑦 3 + 𝑧 3 − 3𝑥𝑦𝑧 = (𝑥 + 𝑦 + 𝑧)(𝑥 2 + 𝑦 2 + 𝑧 2 − 𝑥𝑦 − 𝑦𝑧 − 𝑧𝑥 )
⇒ 𝑑(𝑥 2 + 𝑦 2 + 𝑧 2 ) = 2(𝑥 + 𝑦 + 𝑧)𝑑(𝑥 + 𝑦 + 𝑧)
⇒ (𝑥 + 𝑦 + 𝑧)2 − (𝑥 2 + 𝑦 2 + 𝑧 2) =constant
⇒ 𝑥𝑦 + 𝑦𝑧 + 𝑧𝑥 = 𝐶1 (3)
𝑑𝑥−𝑑𝑦
Each ratio= (𝒙𝟐
−𝒚𝒛)−(𝒚𝟐 −𝒛𝒙)

𝑑𝑥−𝑑𝑧
= (𝑥 2
−𝑦𝑧)−(𝑧 2 −𝑥𝑦)

𝑑(𝑥−𝑦) 𝑑(𝑥−𝑧)
= (𝑥−𝑦)(𝑥+𝑦+𝑧) = (𝑥−𝑧)(𝑥+𝑦+𝑧)

𝑑(𝑥−𝑦) 𝑑(𝑥−𝑧) (𝑥−𝑦)

⇒ (𝑥−𝑦)
= (𝑥−𝑧)
⇒ (𝑥−𝑧)
= 𝐶2 (4)

From (3) and (4) we can write the general solution as

𝑥−𝑦
𝐹 (𝑥𝑦 + 𝑦𝑧 + 𝑧𝑥, )=0 (5)
𝑥−𝑧

Example:2.38
Solve: (𝒚 − 𝒛)𝒑 + (𝒙 − 𝒚)𝒒 = 𝒛 − 𝒙
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
(𝑦−𝑧)
= = (1)
𝑥−𝑦 𝑧−𝑥

𝑑𝑥+𝑑𝑦+𝑑𝑧
Now, each ratio = (𝑦−𝑧)+(𝑥−𝑦)+(𝑧−𝑥)=0

⇒ 𝑑 (𝑥 + 𝑦 + 𝑧) = 0 ⇒ 𝑥 + 𝑦 + 𝑧 = 𝐶1 (2)
−𝑥𝑑𝑥 𝑧𝑑𝑦+𝑦𝑑𝑧
each ratio = =
−𝑥(𝑦−𝑧) 𝑧(𝑥−𝑦)+𝑦(𝑧−𝑥)

𝑑𝑥 𝑑(𝑦𝑧) 𝑥2
= (𝑦−𝑧) = ⇒ + 𝑦𝑧 = 𝐶2 (3)
−𝑥(𝑦−𝑧) 2

General solution is
𝑥2
𝐹 (𝑥 + 𝑦 + 𝑧, + 𝑦𝑧) = 0 (4)
2

30
Example:2.39
Solve: 𝒙𝒑 + (𝟐𝒙 − 𝒚)𝒒 = (𝟏 − 𝒛)
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦 𝑑𝑧
= = (1)
𝑥 2𝑥−𝑦 1−𝑧

From the first and last ratios

𝑑𝑥 𝑑𝑧
=
𝑥 1−𝑧
Integrating we get
log 𝑥 = − log(1 − 𝑧) + log 𝐶1
⇒ 𝑥 (1 − 𝑧) = 𝐶1 (2)
From the first two ratios
𝑑𝑦 𝑑𝑦
= =
𝑥 2𝑥−𝑦

We have, on cross-multiplication,
2𝑥𝑑𝑥 − 𝑦𝑑𝑥 = 𝑥𝑑𝑦 or 2𝑥𝑑𝑥 = 𝑥𝑑𝑦 + 𝑦𝑑𝑥 or 𝑑 (𝑥 2 ) − 𝑑 (𝑥𝑦) = 0
Integrating
𝑥 2 − 𝑥𝑦 = 𝐶2 (3)
General solution is
𝐹 (𝑥 (1 − 𝑧), 𝑥 2 − 𝑥𝑦) = 0 (4)
Example:2.40
Solve: (𝒙 + 𝟐𝒚𝟐 )𝒑 + 𝒚𝒒 = − 𝐜𝐨𝐬 𝒛
Solution: Lagrange’s auxiliary equations are
𝑑𝑥 𝑑𝑦
= = −sec 𝑧 𝑑𝑧 (1)
𝑥+2𝑦 2 𝑦

From the last two ratios

𝑑𝑦 𝑑𝑧
=
𝑦 − cos 𝑧
Integrating we get

31
 log 𝑦 = − log(sec 𝑧 + tan 𝑧) + log 𝐶1
Or 𝑦(sec 𝑧 + tan 𝑧) = 𝐶1 (2)
From the first two ratios
𝑑𝑥 𝑑𝑦
= =
𝑥+2𝑦 2 𝑦

On cross-multiplication and transposition

𝑦𝑑𝑥 − 𝑥𝑑𝑦 = 2𝑦 2 𝑑𝑦
𝑦𝑑𝑥−𝑥𝑑𝑦 𝑥
= = 2𝑑𝑦 ⇒ 𝑑 ( ) = 2𝑑𝑦
𝑦2 𝑦

2𝑥𝑑𝑥 = 𝑥𝑑𝑦 + 𝑦𝑑𝑥 or 𝑑 (𝑥 2 ) − 𝑑 (𝑥𝑦) = 0

Integrating, we get
𝑥
− 2𝑦 = 𝐶2 (3)
𝑦

General solution is
𝑥
𝐹 (𝑦(sec 𝑧 + tan 𝑧), − 2𝑦) = 0 (4)
𝑦

2.11 Nonlinear Equations of First Order

The most general partial differential equation of the first order in two
independent variables is of the form
𝑓 (𝑥, 𝑦, 𝑧, 𝑝, 𝑞 ) = 0 (2.64)
Which may not be linear.
We do not propose to study any general method of integration of equation
(2.64) but consider only four standard forms which admit integration by very
short processes. Also, many equations can be reduced to one or the other of these
four forms.
The general integral, the singular integral and the complete integral must
be indicated in each case. Otherwise, the equation is not considered fully solved.
Standard Form I: 𝒑𝒒 -equation: Equation of the form 𝒇(𝒑, 𝒒) = 𝟎 where
𝒙, 𝒚, 𝒛 do not occur explicitly.
To solve such an equation we put 𝑝 = 𝑎 and find the value of 𝑞 in terms of
𝑎 i.e. 𝑞 = 𝑓 (𝑎) . Then we substitute these values in

32
 𝑑𝑧 = 𝑝 𝑑𝑥 + 𝑞 𝑑𝑦 (2.65)
So that we have 𝑑𝑧 = 𝑎 𝑑𝑥 + 𝑓 (𝑎) 𝑑𝑦
Integrating, we get the complete integral as
𝑧 = 𝑎𝑥 + ∫ 𝑓 (𝑎) 𝑑𝑦 + 𝑐 or 𝑧 = 𝑎𝑥 + 𝑓(𝑎 )𝑦 + 𝑐 (2.66)
Where 𝑎 and 𝑐 are arbitrary constants
Note 1 we can put instead of and proceed to obtain the complete integral.
Example 2.41
Solve 𝒑𝒒 = 𝒌
Solution: The complete integral is
𝑘
𝑧 = 𝑎𝑥 + 𝑦 + 𝑐 (1)
𝑎

Where 𝑎 and 𝑐 are arbitrary constants

The general integral is obtained by eliminating 𝑎 between the equations
𝑘
𝑧 = 𝑎𝑥 + 𝑦 + 𝜑(𝑎)1 (2)
𝑎

Where 𝐶 has been replaced by 𝜑(𝑎) and

𝑘
0=𝑥− 𝑦 + 𝜑′ (𝑎) (3)
𝑎2

Obtained by differentiating with respect to a.

The singular integral, if it exists, is determined from the equation (1)
𝑘
𝑧 = 𝑎𝑥 + 𝑦 + 𝑐
𝑎
𝑘
and 0 = 𝑥 − ,0 = 1 (4), (5)
𝑎2

Obtained from (1) on differentiating it with respect to 𝑎 and 𝑐, respectively.

The inconsistency of the last equation shows that singular integral does not
exist in this case.
Example 2.41
Solve 𝒑𝒒 = 𝒙𝜶 𝒚𝜷𝒛𝜸
Solution: The given equation is

33
𝜕𝑧 𝜕𝑧
= 𝑥𝛼 𝑦𝛽 𝑧𝛾 (1)
𝜕𝑥 𝜕𝑦

We can write (1) as

Case (i) 𝛼 ≠ 1, 𝛽 ≠ −1, 𝛾 ≠ −2
put
𝛾
𝛾 −( )+1
− 𝑧 2
𝑧 𝑑𝑧 = 𝑑𝑍 ⇒ 𝑍 =
2 𝛾 , 𝛾 ≠ −2 (2)
1−
2

𝛼
𝑥 𝛼+1
𝑥 𝑑𝑥 = 𝑑𝑋 ⇒ 𝑋 = , 𝛼 ≠ 1,
𝛼+1

𝛽
𝑦 𝛽+1
𝑦 𝑑𝑦 = 𝑑𝑌 ⇒ 𝑌 = , 𝛽 ≠ −1,
𝛽+1
Equation (2) now becomes,
𝜕𝑍 𝜕𝑍
=1 (3)
𝜕𝑋 𝜕𝑌

Whose complete integral is

𝑍 = 𝑎𝑋 + 𝑏𝑌 + 𝑐where 𝑎𝑏 = 1
𝛾
−( )+1
𝑧 2 𝑎 1 1
Or 𝛾 = 𝑥 𝛼+1 + 𝑦 𝛽+1 + 𝑐 (4)
1−2 𝛼+1 𝑎 𝛽+1

Case (ii) 𝛼 = 1, 𝛽 = −1, 𝛾 = −2 We have

𝜕𝑧 𝜕𝑧 1 1 1 𝑧𝜕𝑧 𝑧𝜕𝑧
= . . 2 ⇒ −1 . −1 = 1
𝜕𝑥 𝜕𝑦 𝑥 𝑦 𝑧 𝑥 𝜕𝑥 𝑦 𝜕𝑦
1
Put 𝑧 𝑑𝑧 = 𝑑𝑍 ⇒ 𝑍 = 𝑧
2

𝑥 −1𝑑𝑥 = 𝑑𝑋 ⇒ 𝑋 = 𝑙𝑜𝑔
𝑦 −1𝑑𝑦 = 𝑑𝑌 ⇒ 𝑌 = 𝑙𝑜𝑔
The CI is
1
𝑍 = 𝑎𝑋 + 𝑏𝑌 + 𝑐 −1 where 𝑎𝑏 = 1 or 𝑏 =
𝑎
1 1
⇒ 𝑧 2 = 𝑎 log 𝑥 + log 𝑦 + 𝑐 (5)
2 𝑎

34
To find the general Integral
Case(i) Writing 𝑐 = 𝜑(𝑎)in the complete integral at (4) we have
𝛾
−( )+1
𝑧 2 𝑎 1 1
𝛾 = 𝑥 𝛼+1 + 𝑦 𝛽+1 + 𝜑(𝑎) (6)
1− 𝛼+1 𝑎 𝛽+1
2

Differentiating with respect to a we get

𝛾
−( )+1
𝑧 2 𝑎 1 1
0= 𝛾 = 𝑥 𝛼+1 + 𝑦 𝛽+1 + 𝜑′(𝑎) (7)
1− 𝛼+1 𝑎 𝛽+1
2

The general integral is the eliminant of a between (6) and (7)

Case(ii) Writing 𝑐 = 𝜑(𝑎)in the complete Integral at (5)
1 2 1 1
𝑧 = 𝑎 log 𝑥 + log 𝑥 + log 𝑦 + 𝜑(𝑎) (8)
2 𝑎 𝑎

Differentiating (8) with respect to a

1
0 = log 𝑥 − log 𝑥 + 𝜑′(𝑎) (9)
𝑎2

Eliminant of a between (8) and (9) is the general integral.

To find the Singular Integral
Case(i) The singular integral, if it exists, is obtained from the following
equations:
𝛾
−( )+1
𝑧 2 𝑎 1 1
𝛾 = 𝑥 𝛼+1 + 𝑦 𝛽+1 + 𝑐
1−2 𝛼+1 𝑎 𝛽+1

(10)
Differentiating partially with respect to a and c, respectively, we get
1 1 1
0= 𝑥 𝛼+1 − 𝑦 𝛽+1 0 = 1 (11),(12)
𝛼+1 𝑎2 𝛽+1

The inconsistency in the last equation shows that the singular integral does not
exist in this case.
Case(ii) We can show that in this case also singular integral does not exist.
Example 2.43
Find a complete integral of 𝑝 + 𝑞 = 𝑝𝑞 (1)

35
 𝑎 𝑎
Solution: Put 𝑝 = 𝑎, then 𝑞 = and the equation 𝑑𝑧 = 𝑝𝑑𝑥 +
𝑎+1 𝛼+1
𝑞𝑑𝑦becomes
𝑎
𝑑𝑧 = 𝑎𝑑𝑥 + 𝑑𝑦 (2)
𝑎−1

Whose solution is
𝑎
𝑧 = 𝑎𝑥 + 𝑦+𝑐 (3)
𝑎−1

Which is the required complete integral

Standard Form II: 𝒛𝒑𝒒-equation: Equation of the form 𝒇(𝒛, 𝒑, 𝒒) = 𝟎 where
the independent variables do not appear explicitly.
In this case put
𝑞 = 𝑎𝑝 𝑜𝑟 (𝑝 = 𝑎𝑞) (2.67)

The given equation becomes 𝑓(𝑧, 𝑝, 𝑞) = 0. Solving for we have

𝑝 = 𝜙(𝑎, 𝑧) (2.68)
Substituting in 𝑑𝑧 = 𝑝𝑑𝑥 + 𝑞𝑑𝑦
We have
𝑑𝑧 = 𝑝𝑑𝑥 + 𝑎𝑝𝑑𝑦 = 𝑝𝑑(𝑥 + 𝑎𝑦) (2.69)
𝑑𝑧
⇒ 𝑑 (𝑥 + 𝑎𝑦) =
𝜙(𝑎, 𝑧)
𝑑𝑧
⇒ 𝑥 + 𝑎𝑦 + 𝑐 = ∫ (2.70)
𝜙(𝑎,𝑧)

Where c is a arbitrary constant

Example 2.44
Solve 𝑝2 = 𝑧 2(1 − 𝑝𝑞) (1)
Solution Put 𝑞 = 𝑎𝑝in the given partial differential equation (1) so that we have
𝑝2 = 𝑧 2 (1 − 𝑎𝑝2) (2)
1 1+𝑎𝑧 2
𝑝2 (1 + 𝑎𝑧 2 ) = 𝑧 2 or = (3)
𝑧2 𝑧2

Now, 𝑑𝑧 = 𝑝𝑑𝑥 + 𝑞𝑑𝑦

36
 = 𝑝𝑑𝑥 + 𝑞𝑑𝑦 = 𝑝𝑑(𝑥 + 𝑎𝑦)

𝑑𝑧
√1 + 𝑎𝑧 2 𝑑𝜃
𝑥 + 𝑎𝑦 + 𝑐 = ∫ =∫ 𝑑𝑧 = ∫
𝑝 𝑧 sin 𝜃 cos 2 𝜃

cos 2 𝜃 + sin2 𝜃
=∫ 𝑑𝜃 = ∫(sec 𝜃 + sec 𝜃 tan 𝜃 )𝑑𝜃
sin 𝜃 cos 2 𝜃
= log(csc 𝜃 − cot 𝜃 ) + c
2
= log [√1 + 𝑎𝑧 − 1⁄ ] + √1 + 𝑎𝑧 2
√𝑎𝑧
Put 𝑎𝑧 2 = tan2 𝜃
1
⇒ tan 𝜃
√𝑎
1 1 1
𝑑𝑧 = 𝑠𝑒𝑐 2 𝜃 𝑑𝜃 = 𝑑𝜃
√𝑎 √𝑎 cos 2 𝜃

√1 + 𝑎𝑧 2 sec 𝜃 √𝑎
∴ = =
𝑧 1
tan 𝜃 sin 𝜃
√𝑎
Where c is an arbitrary constant
Example 2.45
𝟏 𝟏
Solve 𝒛𝒑𝒒 = 𝒑𝟐 𝒒𝟐 (1)
Solution: Putting 𝑞 = 𝑎𝑝 in (1) we get
1 1
1 𝑎
𝑧𝑎𝑝2 = 𝑝2 (1 + 𝑞2 ) ⇒ = 𝑧 (2)
𝑝3/2 1+√𝑎

37
Now, 𝑑𝑧 = 𝑝𝑑𝑥 + 𝑞𝑑𝑦 = 𝑝𝑑(𝑥 + 𝑎𝑦)
𝑑𝑧 𝑎 2/3
𝑑 (𝑥 + 𝑎𝑦) = =( 𝑧) 𝑧 2/3𝑑𝑧 (3)
𝑝 1+√𝑎

On integration we get
𝑎 2/3 3
𝑥 + 𝑎𝑦 + 𝑐 = ( 𝑧) 𝑧 5/3 (4)
1+√𝑎 5

Where c is an arbitrary constant.

Standard Form III: Separable equation 𝒇(𝒙, 𝒑) = 𝒈(𝒚, 𝒒).
We can take 𝑓 (𝑥, 𝑝) = 𝑎 and 𝑔(𝑦, 𝑞 ) = 𝑎 (2.71)
Solving for 𝑝 and 𝑞 we get
𝑝 = 𝑓1(𝑎, 𝑥 ), 𝑞 = 𝑔1 (𝑎, 𝑦)
(2.72),(2.73)
Example 2.46
Solve 𝒑 + 𝒒 + 𝒙 + 𝒚
Solution: We can write the equation as
𝑝−𝑥 =𝑦−𝑞 =𝑎 (1)
So that
𝑝 = 𝑥 + 𝑎, 𝑞 = 𝑦 − 𝑎 (2)
Substituting in
𝑑𝑧 = 𝑝𝑑𝑥 + 𝑞𝑑𝑦 (3)
We have
𝑑𝑧 = (𝑥 + 𝑎)𝑑𝑥 + (𝑦 − 𝑎)𝑑𝑦 (4)
Integrating after multiplying throughout by 2, the complete integral is
2𝑧 = (𝑥 + 𝑎)2 + (𝑦 − 𝑎)2 + 𝑐 (5)
Where c is an arbitrary constant.
Example 2.47
Solve 𝒛𝒑𝒒 = 𝒙𝒚

38
Solution: The equation can be written as
𝑧 1/2 𝜕𝑧 𝑧 1/2 𝜕𝑧
= 𝑥𝑦 (1)
𝜕𝑥 𝜕𝑦

2
Put 𝑧 1/2𝑑𝑧 = 𝑑𝑍 ⇒ 𝑧 3/2 (2)
3
𝜕𝑧
𝜕𝑥 𝑦
= 𝜕𝑧 =𝑎 (3)
𝑥
𝜕𝑦

𝜕𝑧 𝜕𝑧 𝑦
∴ = 𝑎𝑥, = (4)
𝜕𝑥 𝜕𝑦 𝑎

Substituting in the equation

𝜕𝑧 𝜕𝑧
𝑑𝑍 = 𝑑𝑥 + 𝑑𝑦 (5)
𝜕𝑥 𝜕𝑦
𝑦
We have 𝑑𝑍 = 𝑎𝑥𝑑𝑥 + 𝑑𝑦 which on integrating yields
𝑎

2 3/2 𝑎 2 𝑦 2
𝑧 = 𝑥 + +𝑐
3 2 2𝑎
Where 𝑐 is an arbitrary constant.
The general integral and singular integral (if it exists) is obtained as explained in
2.7
Standard Form IV: Clairaut’s Equation 𝒛 = 𝒑𝒙 + 𝒒𝒚 + 𝒇(𝒑, 𝒒)
A partial differential equation of the type
𝑧 = 𝑝𝑥 + 𝑞𝑦 + 𝑓(𝑝, 𝑞) (2.76)
is called a Clairaut’s equation
Complete Integral A complete integral of an equation of type (2.76) is obtained
by replacing p and q by arbitrary constant a and b, respectively. Thus, a complete
integral of equation (2.76) is
𝑧 = 𝑎𝑥 + 𝑏𝑦 + 𝑓(𝑎, 𝑏) (2.77)
One can readily verify that (2.77) is a solution of equation (2.76)
Also writing (2.77) as
𝑧 = 𝐹 (𝑥, 𝑦, 𝑎, 𝑏) = 𝑎𝑥 + 𝑏𝑦 + 𝑓 (𝑎, 𝑏) (2.78)
We observe that the matrix

39
 𝜕𝐹 𝜕2 𝐹 𝜕2 𝐹 𝜕𝑓
𝜕𝑎 𝜕𝑥𝜕𝑎 𝜕𝑦𝜕𝑎
𝑥+ 1 0
𝜕𝑎
𝑀= [𝜕𝐹 𝜕2 𝐹 𝜕2 𝐹
] =[ 𝜕𝑓
] (2.79)
𝑦+ 0 1
𝜕𝑏 𝜕𝑥𝜕𝑏 𝜕𝑦𝜕𝑏 𝜕𝑏

is of rank two. Hence (2.78) is indeed a complete integral of equation.

General integral and singular integral (if it exists) are obtained as explained in
2.7.
Singular Integral The complete integral is
𝑧 = 𝑎𝑥 + 𝑏𝑦 + 𝑎𝑏 (2.80)
Differentiating (2.80) partially with respect to a and b we get
0=𝑥+𝑏 (2.81)
0 =𝑦+𝑏 (2.82)
respectively. Eliminating a and b between (2.81), (2.82) we have singular
solution of (2.76) as
𝑧 + 𝑥𝑦 = 0
Example
Find complete integral in each of the following cases:
Example:2.48

𝒛 = 𝒑𝒙 + 𝒒𝒚 − 𝟐√𝒑𝒒

Solution: The given partial differential equation

𝑧 = 𝑝𝑥 + 𝑞𝑦 − 2√𝑝𝑞 (1)

is an equation of Clairaut’s type

The complete integral of (1) is

𝑧 = 𝑎𝑥 + 𝑏𝑦 − 2√𝑎𝑏 (2)
Where a and b are arbitrary constants
Example 2.49
𝒑𝟒 + 𝒒𝟒
𝒛 = 𝒑𝒙 + 𝒒𝒚 +
𝒑𝒒

40
Solution: The given partial differential equation
𝑝4 +𝑞4
𝑧 = 𝑝𝑥 + 𝑞𝑦 + (1)
𝑝𝑞

is an equation of Clairaut’s type

The complete integral of this equation is
𝑎4 +𝑏4
𝑧 = 𝑎𝑥 + 𝑏𝑦 + (2)
𝑎𝑏

Example 2.50
𝟐𝒒(𝒛 − 𝒑𝒙 − 𝒒𝒚) = 𝟏 + 𝒒𝟐
Solution: The given partial differential equation
1+𝑞2
𝑧 = 𝑝𝑥 + 𝑞𝑦 + (1)
2𝑞

is an equation of Clairaut’s type

The complete integral of this equation is
1+𝑏2
𝑧 = 𝑎𝑥 + 𝑏𝑦 + (2)
2𝑏

Example 2.51
Find the singular integral in Example 2.48
Solution: The complete integral is

𝑧 = 𝑝𝑥 + 𝑞𝑦 − 2√𝑝𝑞 (1)

Differentiating this with respect to a and b, respectively, we get

𝑏 𝑎
0 = 𝑥 − √ ,0 = 𝑦 −√ (2),(3)
𝑎 𝑏

Eliminating a and b from these equations we get the singular integral as

𝑥𝑦 = 1 (4)

41
2.12 EULER'S METHOD OF SEPARATION OF VARIABLES
When we model scientific, engineering, biotechnological and other
processes there arise initial and boundary value problems involving partial
differential equations. The general solutions of these partial differential equations
are in the form of arbitrary functions which are not suitable for determining the
exact solutions satisfying the given initial and boundary conditions.
The method of separation of variables, due to Euler is a simple, yet
powerful technique in breaking up a partial differential equation into an ordinary
differential equation, which can be easily solved using the known methods.
For a partial differential equation in the function u  u( x, y) where x and y
are independent variables, we assume that the solution is separable, that is,
u ( x, y )  X ( x)Y ( y)

where X ( x) is a function of x alone and Y ( y) is a function of y alone. Substituting

in the given partial differential equation, separating the variables and assuming
each side must be equal to the same constant (say), we obtain two ordinary
differential equations, whose solution gives the solution of the problem. The
method is best illustrated through examples.
Example 2.52
u u
Solve  2  u by the method of separation of variables given that
x t
3 x
u ( x, 0)  6e

Solution
We assume the solution to be
u ( x, y )  X ( x)Y ( y) (1)
where X(x) is a function of x alone and 7() is a function of r alone. We have
u u
 X ' T and  X 'T (2), (3)
x t

Where denotes differentiation with respect to the argument.

Substituting in the given partial differential equation, we get
X ' T  2 XT ' XT and ( X ' X )T  2 XT ' (4)
Dividing both sides of (4) by 2XT we get

42
 X ' X T '
  (say)
2X T
(5),(6)
X ' X  2 X  0
X' T'
Or  1  2 and 
X T
Solving (5) we get
(1 2  ) x
log X  (1  2 ) x  log c1 or X  c1e (7)

and solving (6) we get

t
log T  t |  log c2 or T  c2e (8)

Combining (7) and (8) we may write the solution as

u ( x, y )  XT  c1 c2e(1 2  ) x e  x (9)

Now 6e3 x  u ( x, 0)  c1 c2e(1 2  ) x

 c1 c2  6 and 1  2  3

The latter equation gives   2

Substituting in the general solution we get

u ( x, y )  6e3 x e2 x  6e (3 x  2t )

Which is the required solution

Example 2.53
u u
Solve by the method of separation of variables 4   3u given that
x y
u  3e y  3e5 y when x  0
Solution
Let u  X ( x)Y ( y) (1)
Where X ( x) is a function of x alone and Y ( y) is a function of y alone.
Calculating the derivatives and substituting in (1) we have

43
 4 X 'Y  XY ' 3 XY  0
X' Y'
4 3     (say)
X Y
Thus , we have two ordinary differential equations.
4 X ' 3 X   X and Y ' Y  0 (2),(3)

Taking trial solution of the form e mx , e my for the above equation we obtain
auxiliary equation
3 
4m  3    0  m  , and m  k (4),(5)
4
respectively.
The solution for X is of the form

X  c1e((3 )/4) x (6)

The solution for Y is of the form

Y  c2e  y (7)

The solution for u can be taken as

u( x, y)  c1e((31 )/4) x .e1 y  c2e((32 )/4) x .e2 y  .... (8)

where we have used the principle of superposition of solutions since the equation
is linear.
Putting x=0 in (8) we get

u(0, y)  3e y  3e5 y  c1e((31 )/4)0 .e1 y  c2e((32 )/4)0 .e2 y  ....
and c3  c4  ....  0

Equating like terms on both sides

3e y  3e5 y  c1e1 y  c2e2 y

 c1  3, c2  3, 1  1, 2  5
and c3  c4  ....  0
The required solution is

44
 u ( x, y )  3e x  y  3e2 x 5 y
Example 2.54
 2u u u
Solve   2u subject to the conditions. u  0,  1  e3 y when x  0
x 2
y x

Solution
 2u u
Let u  X ( x)Y ( y) be a solution of   2u (1)
x 2 y

Substituting in the equation (1)

4 X '' Y  XY ' 2 XY
X '' Y'
 2   k (say) (2)
X Y
we obtain the ordinary differential equations.
X '' (2  k ) X  0, and Y '' kY  0

where k is a constant.

Taking the solutions as X  e mx ,Y  e my we get (3), (4)

Auxiliary equations m  (2  k )  0, m  k  0
2
(5), (6)

2 k x
The solution is of the form (e , e 2 k x
)e ky
If we take the solution as
2 k x
u  (c1e  c2e 2 k x
)eky
The condition

x  0  u  0 which gives c1  c2  0, (eky  0), or c2  c1 (7)

x
Since e  e
 x
 2 sinh  x,

and we have to satisfy another condition by the principle of superposition of

solutions, we may take

45
u ( x, y )  c1 sinh( 2  k x)e k1 y  c2 sinh( 2  k x)e k2 y  ...
(8)
u
 ( 2  k1 c1 cosh( 2  k1 x)ek1 y  2  k2 c2 cosh( 2  k2 x)e k2 y  ...) x 0
x x 0

 ( 2  k1 c1.1.ek1 y  2  k2 c2 .1.ek2 y  c3 ...  c4 ...)  1  e3 y

(9)
1
 k1  0, c1  , k2  3, (10)
2
1
 c2  , c3  c4  ....  0, (11)
i
1
Since 2  k2 c2  1 for k2  3,  c1  , the second solution is
i
1 eix  eix 3 y
c2 sinh 2  k2 xe k2 y
 e sin xe3 y
i 2
1
u ( x, y )  sinh 2 x  sin xe3 y
2
(13)
Example 2.55
 2u u u
Solve 2  0
x 2
x y

Solution
Let u  X ( x)Y ( y) (1)
Where X ( x) is a function of x alone and Y ( y) is a function of y alone.
Calculating the derivatives and substituting in (1) we have
X ''Y  2 XY ' XY '  0
X '' Y'
 2   k (say) (2)
X Y
we obtain the ordinary differential equations.
X '' 2 X ' kX  0, and Y '' kY  0 (3), (4)
where k is a constant.

46
Taking trial solutions, X  e mx ,Y  e my we get

Auxiliary equations m  2mk  0, m  k  0

2

2 4  4k
m  1  1  k ,m   k
2
1 k x
X  (c1e  c2 e  1 k x
)e x ,  ( e 2 k x
, e 2 k x
)Y  c3e  ky
1 k x
u ( x, y )  ( Ae  Be  1 k x
)e x  ky
where A and B are two arbitrary constants, consistant with the order of the

differential equation.

2.13 CLASSIFICATION OF SECOND-ORDER PARTIAL

DIFFERENTIAL EQUATIONS

2.13.1 Introduction

Many physical and engineering applications such as fluid flow, heat transfer and

wave motion involve second-order partial differential equations and hence we

take up a study of these equations and their solution by separation of variables

method. The general second-order partial differential equation may be written in

the form
 2u  2u  2u  u u 
L[u ]  A 2  B  C 2  H  x, y , u , ,   0 (2.84)
x xy y  x y 

Equation (2.84) is called semilinear if A, B and C are functions of the

independent variables x and y only. On the other hand, if A, B and C are functions
u u
of x, y, u, and then (2.84) is called quasi-linear. If A, B and C are functions
x y
u u
of x and y and H is a linear function u , and then equation (2.84) is called
x y
linear. The general second-order linear partial differential equation in two
independent variables x and y may be written as
47
  2u  2u  2u u u
A( x, y)  B ( x, y )  C ( x, y )  D ( x, y )  E ( x, y )  F ( x, y)u  G( x, y)  0
x 2 xy y 2 x y
(2.85)
If G  0 then equation (2.85) is called non-homogeneous and if G=0 then it is
called homogeneous.
2.13.2 Classification of Equations
We call the quantity   B2  4 AC the discriminant and classify equation (2.85) as
hyperbolic, parabolic or elliptic according as A>0, =0 or <0. The following are
well-known examples of these three types.
Hyperbolic Type
1. One-dimensional wave equation

 2u 2  u
2
a (2.86)
t 2 x 2
Here A  a 2 , B  0, C  1 and   B2  4 AC  4a 2  0 .
Parabolic Type
2. One-dimensional heat-flow equation

u 2  u
2
a (2.87)
t x 2
Here A  a 2 , B  C  0 and   B2  4 AC  0
Elliptic Type
3. (a) Two-dimensional Laplace equation

 2u  2u  2 2 2 
 2u    0      (2.88)
x 2 y 2  x 2 y 2 

Here A  1, B  0, C  1 and   B2  4 AC  4  0.
(b) Poisson's equation
 2u  2u
 2u    f ( x, y ) (2.89)
x 2 y 2

Equations (2.86)-(2.88) are homogeneous while equation (2.89) is non-

homogeneous.

48
2.13.3 Initial and boundary Value Problems and their Solution
The unique solution corresponding to particular physical problem obtained by
use additional information arising from physical situation. If this information is
given on the boundary as boundary conditions, the resulting problem is called a
boundary problem (BVP).If this information is given one instant initial
conditions, the resulting problem is called an initial value problem (IVP).
The principle of superposition of solutions is applicable as long as the
equation is linear and homogeneous. That is, if un is a solution for each n then
 n1 un , which is a linear combination of the solutions un  is also a solution of


the equation.
The hyperbolic and parabolic types of equations are either initial value
problems or initial and boundary value problems, whereas the elliptic-type
equation is always a boundary value problem. The boundary conditions may be
one of following three types:
1. Dirichlet problem (First boundary value problems )
The solution is prescribed along the boundary.
2. Neumann problem (Second boundary value problem)
The derivative of the solution is prescribed along the boundary.
3. Mixed problem (Third boundary value problem)
The solution and its derivative are prescribed along the boundary.

Any of the above conditions is called homogeneous if it is zero-condition and

non-homogeneous if it is nonzero condition.
2.13.4 Solution of One-dimensional Heat Equation (or diffusion equation)
Consider a long and thin wire, rod or bar OA length l and constant cross-section
and homogeneous heat-conducting material. Let the bar be placed along the x-
axis with one end O coinciding with the origin and the other end A at a distance
l from O (Figure 2.2).

Figure 2.2 Heat conduction along a bar

49
Suppose that the lateral surface of the bar is perfectly insulated, so that heat flow
is along the x-direction only. Therefore, the temperature u of the bar depends on
x and t only. The initial boundary value problem consists of one-dimensional heat
equation:

u 2  u
2
a (2.90)
t x 2
where a² is the thermal diffusivity. The boundary conditions at the ends O and A
are

u(0, t )  0  u(l , t ) (2.91, 2.92)

respectively. The initial temperature distribution in the bar is

u ( x,0)  f ( x) (2.93)

where f ( x) is a given function of x . The solution by the method of separation

of variables reduces the initial boundary value problem (IBVP) to that of two
ordinary differential equations.

Assume that the solution u ( x, t ) is separable

u ( x, t )  X ( x)T (t ) (2.94)

where X ( x) is a function of x only and T (t ) is a function of t only.

Differentiating (2.94) with respect to t and x we get
u  2u
 XT ', 2  X ''T (2.95)
t x
where ' denotes differentiation with respect to the corresponding independent
variable. Substituting into (2.90) we have
X '' T '
XT '  a 2 X '' T   (2.96)
X a 2T
Since the L..H. member is a function of x only and the R.H. member is a function
of t only, both sides must be equal to the same constant , say. So, we obtain
two ordinary differential equations
X ''  X  0, T ' a 2 T  0 (2.97, 2.98)

Three cases arise: Case 1.   0; Case 2.   0; Case 3.   0.

50
Case 1:   0. The general solution is
u ( x, t )  X ( x)T (t )  c1e  x
 c2e  x
 c3e a
2 t


(2.99)
Equation (2.99) shows that the solution has unbounded temperature for larger t
due to exponential growth, which is not physically possible.

Case 2:   0. The general solution is

u ( x, t )  X ( x)T (t )  (ax  b) (2.100)
which is independent of time. This is also not possible.

Case 3:   0. We write    p 2 where p is real. The general solution

in this case is

u ( x, t )  X ( x)T (t )  ( A cos px  B sin px)(Ce  a

2
p 2t
) (2.101)
The boundary condtion u (0, t )  0 implies that A=0 so that we have

u ( x, t )  BCe  a
2
p 2t
sin px (2.102)

The boundary condition u (l , t )  0 requires that

sin pl  0
which holds if

nP
pl  nP  p  pn  (n  1, 2,3,...) (2.103)
l
The constant pn is an eigenvalue and the function sin pn x is an eigen
function. Now, we can write the solution, using the principle of superposition, as
 
 n 
u ( x, t )   u ( x, t )   bn sin pn xe  a  pn  , n  1, 2, 3,... 
2
pn 2t

n 1 n 1  l 
(13.104)

51
The initial condition (2.93) will be satisfied at t=0 if

u ( x, 0)  f ( x)   bn sin pn x (2.105)
n 1

that is, if f ( x) can be expanded in a convergent half-range Fourier series in (0, l ).

The bn are given by

2 l
l 0
bn  f ( x)sin pn xdx (2.106)

Note 1 Solution of partial differential equations by separation of variables method

cannot be applied in all cases. It is only a certain special set of boundary
conditions that allows us to separate the variables.
Example 2.56
A long copper rod with insulated lateral surface has its left end maintained at a
temperature of 0°C and its right end at x  2m maintained at 100°C. Find the
temperature u ( x, t ) if the initial condition

100 x, 0  x  1
u ( x, 0)  t ( x)  
100,1  x  2
Solution We have to solve the partial differential equation for heat conduction

u 2  u
2
a (1)
t x 2
under the boundary conditions

u (0, t )  0 and u (l , t )  u (2, t )  100 (2),(3)

and the initial condition

100 x, 0  x  1
u ( x, 0)  f ( x)  
100,1  x  2
(4)
Assuming seperation of variables in the form

u ( x, t )  X ( x)T (t ) (5)
52
we get
X '' T '
  (6)
X a 2T
where λ is the separation constant. In this problem the eigenvalue λ = 0 is
important. The solution for λ=0 is
X ( x)  ax  b, T (t )  c (7)

u ( x, t )  c(ax  b) (8)
The boundary conditions u (0, t )  0 and u(2, t )  100 imply that b  0 and ac  50.

Then u( x, t )  50 x (9)

Now taking up the case of exponential decay of temperature name    p

2

where p is real we have the general solution as

u ( x, t )  ( A cos px  B sin px)(Ce  a

2
p 2t
)
(10)
Superimposing the above two solution we obtain a more general solution

u ( x, t )  50 x  ( A cos px  B sin px)(Ce  a

2
p 2t
)
(11)
The condition u (0, t )  0 requires that A=0 and the condition u(l , t )  u (2, t )  100
demands that

100  50  2  BC sin pl (e  a )  sin pl  0

2
p 2t

This will be satisfied if

n n
pl  n  p  pn   (n  1, 2,3,...) (12)
l 2
The solution, by the principle of superposition of solutions, is
 
u ( x, t )   un ( x, t )  50 x   bn sin pn xe  a
2
pn 2t
(13)
n 1 n 1

This must satisfy the initial condition (4) and hence

53
 
n x
f ( x)  50 x   bn sin (14)
n 1 2
Expanding [ f ( x)  50 x] in half-range Fourier sine series in [0, 2] we get
the solution. The Fourier coefficients bn are given by

n x n x n x
l 1 1
2 2 2
bn    f ( x)  50 x  sin dx   100 x  50 x  sin dx   100  50 x  sin dx
l 0 l 20 2 20 2
2
n x n x   n x  n x  
1 2
 2x 4 2   2  
 50   cos  2 2 sin   (100  50 x)   cos   50     sin 
 n 2 n 2  0  n  2   n   2  
1

n 200 n n n
2
100  2  400
 cos  2 2 sin  50    sin  2 2 sin
n 2 n 2  n  2 n 2
(15)
Finally, the solution is

400 
1 n n x  a2 n2 2t / 4
u ( x, t )  50 x  2 
n 1 n 2
sin
2
.sin
2
e
(16)
Example 2.57
An insulated rod of length l has its ends A and B maintained at 0°C and 100°C,
respectively, until steady-state conditions prevail. If B is suddenly reduced to 0°C
and maintained at 0°C, find the temperature at a distance x from A at time t .
[UNTU 2003 (Set 4)]
Solution Let u( x, t ) be the temperature at time t at a distance x from A. The
equation for the conduction of heat is:

u 2  u
2
a (1)
t x 2
where a 2 is the diffusivity of the material of the rod.
In the steady state when u depends only on x we get from (1)

 2u
0 (2)
x 2
whose general solution is

54
 u( x)  ax  b (3)
Boundary conditions are u(0)  0 and u(l )  100 . So we obtain b  0 and
a  100 / l This gives u( x)  (100 / l ) x at time t  0 . Thus, we have the initial
condition.
100
u ( x,0)  x (4)
l
Boundary conditions for unsteady flow are:
u (0, t )  0 and u(l , t )  0, for all t (5), (6)
Now, we have to solve equation (1) under the conditions (4)-(6). A solution of
(1) is of the form

u ( x, t )  ( A cos px  B sin px)(Ce  a

2
p 2t
) (7)

u (0, t )  0  Ae  a 0 A0

2
p 2t
(8)
Equation (7) becomes

u ( x, t )  B sin pxe  a
2
p 2t
(9)

n
u (l , t )  0  B sin p.e a 0 p  B  0)
2
p 2t
( (10)
l
By the principle of superposition of solutions we may write the solution as

u ( x, t )   Bn sin pn xe  a
2
pn 2t
n
where pn  (11)
n 1 l

Imposing the initial condition (4) on the solution (11) we have


100
u ( x, 0)   Bn sin pn x  x (12)
n 1 l
which is the half-range Fourier series expansion in (0, l ) for the function
(100 / l ) x . Therefore, Bn are given by

55
 l
2 100
Bn   x sin pn xdx
l 0 l

200   
cos pn x 
l
l
cos pn x 
 2  x       dx 
l  pn   
  0 0 pn


200  cos pn x 
l
l cos pnl   cos pn x  
 l
 2 0   x       dx  ( pnl  n )
l  pn 
  pn   0 0 pn 
 

200 (1) n 1 200(1) n 1

 2 .l 
l  n  n
 
 l  (13)

The solution for the problem is

 a 2n2 2t 
 
200 ( 1)
 n 1
n x  

t2
u ( x, t )  sin e  
 n 1 n l (14)

Example 2.58
A homogeneous rod of conducting material of length 100 cm has its ends kept at
zero temperature, and the temperature is initially

 x, 0  x  1
u ( x, 0)  
100  x, 50  x  100
Find the temperature u ( x, t ) at any time. [JNTU 2004s (Set 3)]
Solution We have to solve the differential equation for the conduction of heat

u  2u
 a2 2 (1)
t x
under the boundary and initial conditions
Boundary conditions:

56
 1. u(0, t )  0 for all t (2)
2. u(100, t )  0 for all t (3)

Initial condition:

 x, 0  x  1 
u ( x, 0)   
100  x, 50  x  100 
(4)

A solution of equation (1) may be taken as

u ( x, t )  ( A cos px  B sin px)e  a

2
p 2t
(5)

u (0, t )  0  A  0 equation (5) becomes

u ( x, t )  B sin px)e  a
2
p 2t
(6)

u (100, t )  0  B sin p100  0

 100 p  n ( B  0)

n
 p  pn  (7)
100
By the principle of superposition of solutions we may take the solution as

u ( x, t )   Bn sin pn xe  a
2
pn 2t

n 1 (8)
Imposition of the initial condition (4) on (8) yield,

 x, 0  x  1 
 n
B sin nx u ( x , 0)  
  

100 x , 50 x 100  (9)
We now expand u ( x, 0) in a half-range Fourier sine series in (0, 100) and
determine Bn

Now, Bn are given by

n x 2  
l 50 100
2
Bn   u ( x) sin
100  0 
dx   x sin pn xdx  (100  x ) sin p n xdx  (10)
l 0 l 50 

57
 2  
50 100
  cos pn x    sin pn x      cos pn x    sin pn x  
 x    1 

 

 (100  x )   
 
100   2 2

    0     50
pn pn  pn pn

= 0 at 0, 50 = 0 at 0 = 0 at 100, 50 = 0 at 100

 1, x  0
 n x 
cos pn x  cos 100  0, x  50
 (1) n , x  100
 
 0, x  0,100
sin p x  sin n x  
  n
, x  50
n
100 sin
  2

 n n 
sin sin
2  2  2    4 . 100 .sin n  400 sin n
2
  
100  pn 2 pn 2  100  n 2  2 n2
2 2
2
 

0, if " n " even


  400(1) n
 , if " n " odd
  2 n2

Finally, the required solution of the problem is

400 
(1) n (2n  1) x   2n  1  2 2 2 
u ( x, t ) 
2
 
sin exp   
   c t 
  
2
n 1 (2 n 1) 100 100 
Example 2.59
Find the temperature u( x, t ) in a homogeneous bar of heat conducting material of
length l cm with its ends kept at zero temperature and initial temperature given
by  x(l  x) / l 2 .
Solution The initial boundary value problem consists of the following:
1. Partial differential equation for conduction of heat:
u 2  u
2
a (1)
t x 2

2. Boundary conditions:
u(0, t )  0, u(l , t )  0 for all t (2),(3)

58
3. Initial condition:
u ( x, 0)   x(l  x) / l 2 , 0  x  1 (4)
A general solution of (1) is

u ( x, t )  ( A cos px  B sin px)e  a

2
p 2t
(5)
Boundary condition (2) is satisfied if we set A=0 and boundary condition (3) is
satisfied if

u (l , t )  B sin ple  a  0  pl  n
2
p 2t

n
 p  pn  ( B  0)
l (6)

By the principle of superposition of solutions we may write


u ( x, t )   Bn sin pn xe  a ( pn  n / l )
2
pn 2t
(7)
n 1

By the imposition of the initial condition (4) on (7) we get


u ( x, 0)   x(l  x) / l   Bn sin pn x
2
(8)
n 1

which is the half-range Fourier sine series in u (0, l ) for f ( x)   x(l  x) / l . The
2

constants Bn, are given

2  x(l  x)
l l
2
Bn   f ( x) sin pn xdx   sin pn xdx
l 0 l 0 l2
2
l
 3  (lx  x 2 ) sin pn xdx
l 0
l
2   cos pn x    sin pn x   cos pn x  
 3 (lx  x 2 )     (l  2 x )  2   ( 2)  3 
l   p n   pn   p n 0

59
 2  2l 3  8
 3  0  0  (( 1) n
 1)   . , ( n : odd )
l  n 
3 3
 n 
3 3

Hence the temperature distribution in the bar is given by

8 1 
(2n  1) x  (2n  1) 2 2 2 
u ( x, t )  3  sin exp    a t
 n 1 (2n  1) 3
l  l 2


Example 2.60
Find the temperature in thin metal rod of length l with both ends insulated and
with initial temperature the rod sin( x / l ) .
Solution The initial boundary value problem consists of the following
1.Partial differential equation for conduction of heat
u  2u
 a2 2 (1)
t x

2. Boundary conditions: (Insulation at both ends)

u u
 0;  0; (2), (3)
x x 0
x x l

3.Initial condition:

x 
u ( x, 0)  sin   (4)
 l 
The general solution of equation (1) is of the form
A0
u ( x, t )   ( A cos px  B sin px)e a p t
2 2

(5)
2
(Note: we have added the constant A0/2 since in this case we have half-range
Fourier cosine series expansion)
Differentiating (5) we have
u
 ( pA sin px  pB cos px)e  a p t
2 2

(6)
x
Boundary condition (2) is satisfied if we set B=0. Also, boundary condition (3) is
satisfied if
60
 n
sin pl  0( A  0)  p  pn  (7)
l
Therefore, by the principle of superposition, the solution may be taken as

A0
u ( x, t )    An cos pn x.e  a pn t
2 2

(8)
2 n 1

Hence
l
 cos  x 
2 sin  x 2 
l l
2
A0   f ( x)dx   dx   l 
l 0 l 0 l l x   (9)
   
  l  0
4l 4
 
l 

2
1
x 2 x 1
x
A1   f ( x) cos dx   sin cos dx
l 0 l l 0 l l

x
l
1
  sin 2  0 (10)
 l 0

n≠ 1
n x x n x
1 1
2 2
An   f ( x) cos dx   sin cos dx
l 0 l l 0 l l

1  (n  1) x (n  1) x 
1 1
    sin
l 0 0 l
 sin
l
 dx

l
1  l (n  1) x x
1
l
  cos  cos(n  1) 
l 0  (n  1) l (n  1) l 0

1  1 
1
  
 0  (n  1)
1  (1)n 1  
l
(n  1)
1  (1) n 1  


2 1 1  4
   
  n  1 n  1    n 2  1

4
 (n even) (11)
  n2  1

The temperature distribution in the rod is

61
 2 4 
1  2n x   4n 2 a 2 2 
u ( x, t ) 

 
 n 1 4n 2  1  l 
cos exp  2
t (12)
 l 

Example 2.61
A homogeneous rod of conduction material of length l has its ends kept at zero
temperature. The temperature at the centre is T and it falls uniformly to zero at
the two ends. Find the temperature at any time t.
Solution The initial boundary value problem consists of solving the partial
differential equation for heat conduction:
u  2u
 a2 2 (1)
t x

under the boundary conditions.

u(0, t )  0  u(l , t ) (2), (3)
and the initial condition that u ( x, 0) falls uniformly to zero at the ends.
d 2u
To find this condition we have to solve the steady-state equation: whose
dx 2
general solution is
u( x)  ax  b (4)
If C= l /2 is the centre of the rod then for the portion AC of the rod we have
at A : u(0)  a.0  b  0  b  0

and at C : u    a  T  a 
l l 2T
2 2 l

2T l
So that at u ( x)  x,0  x  (5)
l 2
For the portion CB of the rod we have
At B : u(l )  al  b  0  b  al

And at C : u    T  a  al   al  a  
l l 1 2T
2 2 2 l

2T l
So that u ( x)   x  2T ,  x  l (6)
l 2

Consequently the initial condition is

62
  2T l
 l x, 0  x  2
u ( x, 0)   (7)
 2T (l  x), l  x  l
 l 2

By the method of separation of variables the solution of (1) may be put in the
form

u ( x, t )  ( A cos px  B sin px)e  a

2
p 2t
(8)
Boundary condition (2) is satisfied if we set A=0, and boundary conditon (3) is
satisfied if we set sin pl  0

n
 p  pn  (9)
l
Thus, by the principle of superposition of solutions, we may write the
general solution as

u ( x, t )   Bn sin pn xe  c
2
pn 2t
(10)
n 1

Imposing the initial condition (7) we must have

 2T l
  l x , 0  x 
u ( x, 0)   Bn sin pn x  
2
n 1  2T (l  x), l  x  l
 l 2
(11)
Equation (11) is a half-range Fourier sine series expansion in (0, l ) for the function
( x, 0) and so the constants Bn, are given by

 n 
l l /2 l
2 2 2T 2 2T
Bn   u ( x)sin pn xdx   x sin pn xdx   (l  x)sin pn xdx  pn  
l0 l 0
l l l /2 l  l 

l /2 l
4T   cos pn x    sin pn x   4T   cos pn x    sin pn x  
 2  
x  
  1 2   2 
(l  x )   2 
l   pn   pn  0 l   p n   pn   l /2

63
 sin pn l  sin n  sin 0  0
4T (1) n l 2 4T ( 1) n l 2 8T ( 1) n
 2 . 2 2  2 . 2 2  pl n
l n l n n 2 2 sin n  sin ( 1) n
2 2

(13)

cos pnl  cos n  sin 0  0(1)n

pnl n
cos  cos 0
2 2
cos 0  1
Finally, the solution of the problem
8T (1) n

n x  n 2 2 c 2t 
u ( x, t )  2  sin exp   
 n 1 n 2
l  l 2

(14)

2.13.5 One-dimensional Wave Equation

Consider the vibrations of an elastic string placed along the x-axis, stretched to
length l between two fixed points x  0 and x  l First we consider the problem
when there is an initial displacement but no initial velocity (string released from
rest). Next we consider motion of a string with an initial velocity but no initial
displacement (string given an initial blow, but from its horizontal stretched
position). Finally, we consider the case of both initial velocity and initial
displacement.
2. 13.6 Vibrating String with Zero Initial Velocity
Consider an elastic string of length l , fastened at its ends on the x-axis at x  0
and x  l The string is displaced, then released from rest to vibrate in the xy-plane.
We want to find the displacement function y( x, t ) whose graph is a curve in the
xy-plane showing the shape of the string at time t. The boundary value problem
for the displacement function y( x, t ) consist in the solution of the partial
differential equation
2 y 2  u
2
 a for 0  x  l , t  0 (2.107)
t 2 x 2

Under the boundary conditions:

64
 y(0, t )  y(l , t )  0 for t  0 (2.108,2.109)
And the initial conditions:
y ( x, 0)  f ( x) 

y  for 0  x  l (2.110,2.111)
( x, 0)  0 
t 

The graph of f ( x) is the position of the string before release.

The separation of variables method consists of attempting a solution of the form
y( x, t )  X ( x)T (t ) where X ( x) is a function of x only and T (t ) is a function of t
only. Substituting into the wave equation we obtain
XT ''  a 2 X '' T (2.112)
where ' denotes differentiation with respect to the respective independent variable.
Then
X '' T ''
 2 (2.113)
X aT

The left-hand member is a function of x only and the right-hand member is a

function of t only. The equality is possible only if both the quantities are equal
to the same constant, which we take as  2 for convenience. This is called the
separation constant. . (Taking the separation constant as 0 or positive real number
leads to trivial solutions.)
We now have
X '' T ''
 2   2 (2.114)
X aT

we obtain two ordinary differential equations

X ''  2 X  0 and T '' a 2 2T  0 (2.115,2.116)
The boundary conditions y(0, t )  0 , and y(1, t )  0 yield X (0)  0 and X (l )  0
respectively. The general solutions of equations are
X ( x)  A cos  x  B sin  x (2.117)
T (t )  C cos at  D sin at (2.118)
The condition X (0)  0 implies that A=0 and the condition X (l )  0 implies that
l  0( B  0)

Therefore, l  n or λ= λn  n / l , which are the eigenvalues of the problem. The

corresponding eigenfunctions are

65
 X n ( x)  Bn sin n x (n=1,2,3,…) (2.119)

The initial condition (the string is released from rest)

y
( x,0)  0 gives X ( x)T '(0)  0  T '(0)  0 (2.120)
t
Hence
an (C sin ant  D cos ant t 0  0  D  0

Therefore, we obtain

𝑇(𝑡) = 𝑇𝑛 (𝑡) = 𝐶𝑛 cos 𝑎𝜆𝑛 𝑡 (𝑛 = 1,2,3, … ) (2.121)

Now we take the solutions for the problem as

𝑛𝜋
𝑦𝑛 (𝑥, 𝑡) = 𝐶𝑛 sin 𝜆𝑛 𝑥 cos 𝜆𝑛 at- (𝜆𝑛 = ; 𝑛 = 1,2,3, … ) (2.122)
𝑙

Each of these functions satisfies the wave equation, both boundary conditions and the initial
𝜕𝑦
condition 𝜕𝑡 (𝑥, 0) = 0. We need to satisfy the condition
𝑦(𝑥, 0) = 𝑓(𝑥 ). This is achieved by an infinite super- position of solutions in the form

𝑛𝜋
𝑦𝑛 (𝑥, 𝑡) = ∑∞ ∞
𝑛=1 𝑦𝑛 (𝑥, 𝑡 ) = ∑𝑛=1 𝐶𝑛 𝑠𝑖𝑛 𝜆𝑛 𝑥 cos 𝜆𝑛 𝑎𝑡 (𝜆𝑛 = ; 𝑛 = 1,2,3, … ) (2.123)
𝑙

we must choose the 𝐶𝑛 ’s to satisfy

𝑦(𝑥, 0) = ∑∞ ∞
𝑛=1 𝑦𝑛 (𝑥, 0) = ∑𝑛=1 𝐶𝑛 sin 𝜆𝑛 𝑥 = 𝑓(𝑥) (2.124)
This series is the half-range Fourier sine series of 𝑓(𝑥) in [0, 𝑙]. The Fourier constants are
given by
2 𝑙 𝑛𝜋𝑥
𝐶𝑛 = 𝑙 ∫0 𝑓(𝑥 ) sin 𝑙 𝑑𝑥 (2.125)
Example 10.62
A string 𝐴𝐵 of length 𝑙 is fastened at both ends 𝐴 and 𝐵 At a distance ′𝑏′ from the end
𝐴, the string is transeversely displaced to a distance ′𝑑′ and released from rest when it is in
this position. Find the solution for the initial displacement function and zero initial velocity.
Solution: Let 𝑦(𝑥, 𝑡) be the displacement of the string. The initial displacement is given by
𝐴𝑃𝐵

66
 Figure 2.3 String with transverse displacement d at a point M(b,0)
Equation of 𝐴𝑃 is 𝑦 = (𝑑/𝑏)𝑥 (1)
𝑑(𝑥−𝑙)
Equation of 𝑃𝐵 is 𝑦 = (2)
𝑏−𝑙

The problem is to solve one-dimensional wave equation

𝜕2𝑦 𝜕2𝑦
= 𝑎2 𝜕𝑥 2 (3)
𝜕𝑡 2

with boundary conditions

𝑦(0, 𝑡) = 0 = 𝑦(𝑙, 𝑡) (4)(5)

and initial conditions

𝑑
𝑥 for 0≤𝑥≤𝑏
𝑦(𝑥, 0) = 𝑓 (𝑥 ) = {𝑏 𝑑(𝑥−𝑙)
} (6)
for 𝑏 ≤ 𝑥 ≤ 𝑙
𝑏−𝑙

The solution is given by

𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = ∑∞
𝑛=1 𝐶𝑛 sin cos (7)
𝑙 𝑙

Where 𝐶𝑛 are given by

2 𝑙 𝑛𝜋𝑥 2 𝑏𝑑 𝑛𝜋𝑥 𝑙
𝑑(𝑥 − 𝑙) 𝑛𝜋𝑥
( )
𝐶𝑛 = ∫ 𝑓 𝑥 sin 𝑑𝑥 = [∫ 𝑥 sin 𝑑𝑥 + ∫ sin 𝑑𝑥 ]
𝑙 0 𝑙 𝑙 0 𝑏 𝑙 𝑏 𝑏−𝑙 𝑙
𝑏
2𝑑 −𝑙 𝑛𝜋𝑥 −𝑙2 𝑛𝜋𝑥 2𝑑 −𝑙 𝑛𝜋𝑥
= [𝑥 ( ) cos − (𝑛2𝜋2 ) sin ] + [(𝑥 − 𝑙 ) ( ) cos −
𝑏𝑙 𝑛𝜋 𝑙 𝑙 0 𝑙(𝑏−𝑙) 𝑛𝜋 𝑙
𝑙
−𝑙2 𝑛𝜋𝑥
(𝑛2𝜋2 ) sin ]
𝑙 𝑏

−2𝑑𝑏𝑙 𝑛𝜋𝑏 2𝑑𝑙 2 𝑛𝜋𝑏 2𝑑 𝑛𝜋𝑏 2𝑑𝑙 2 𝑛𝜋𝑏

= cos + 2 2
𝑠𝑖𝑛 + cos − 2 2
sin
𝑏𝑙𝑛𝜋 𝑙 𝑏𝑙𝑛 𝜋 𝑙 𝑛𝜋 𝑙 𝑙 (𝑏 − 𝑙 )𝑛 𝜋 𝑙

67

P(b d )
 2𝑑𝑙2 𝑛𝜋𝑏
= 𝑏(𝑙−𝑏)𝑛2𝜋2 sin (8)
𝑙

Hence, the displacement of the string at any point and time is given by

2𝑑𝑙2 1 𝑛𝜋𝑏 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡

𝑦(𝑥, 𝑡) = 𝑏(𝑙−𝑏)𝜋2 ∑∞
𝑛=1 𝑛 2 sin ∙ sin ∙ cos (9)
𝑙 𝑙 𝑙

Example 2.63
A string of length 𝑙 fastened at both ends 𝐴 = (0,0) and 𝐵 (𝑙, 0) undergoes initially a
transversal displacement given by

𝑥 for 0 ≤ 𝑥 ≤ 𝑙/2
𝑓 (𝑥 ) = {
𝑙−𝑥 for 𝑙/2 ≤ 𝑥 ≤ 𝑙
and is released at rest when it is in this position. Find the displacement function 𝑦(𝑥, 𝑡) for the
subsequent motion.
Solution The problem consists of solving the wave equation
𝜕2𝑦 𝜕2𝑦
= 𝑎2 𝜕𝑥 2 (1)
𝜕𝑡 2

under the boundary conditions:

𝑦(0, 𝑙 ) = 0 = 𝑦(𝑙, 𝑡) (2),(3)
and the initial conditions:
Initial displacement:
𝑥 for 0 ≤ 𝑥 ≤ 𝑙/2
𝑦(𝑥, 0) = 𝑓(𝑥) = { (4)
𝑙−𝑥 for 𝑙/2 ≤ 𝑥 ≤ 𝑙

𝜕𝑦
Initial velocity: (𝑥, 0) = 𝑔(𝑥 ) = 0 (5)
𝜕𝑡

The solution of equation (1) under the boundary conditions (2) and (3) and the zero initial
velocity is
𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = ∑∞ 𝑛=1 𝐶𝑛 sin 𝑙 cos 𝑙 (6)

Where

68
 Figure 2.4 : String with transverse displacement at the mid point

2 𝑙 𝑛𝜋𝑥
𝐶𝑛 = ∫ 𝑓(𝑥 ) sin 𝑑𝑥
𝑙 0 𝑙

2 𝑙/2 𝑛𝜋𝑥 2 𝑙 𝑛𝜋𝑥

= ∫ 𝑥 sin 𝑑𝑥 + ∫ (𝑙 − 𝑥) sin 𝑑𝑥
𝑙 0 𝑙 𝑙 𝑙/2 𝑙
𝑙/2
2 −𝑙 𝑛𝜋𝑥 −𝑙 2 𝑛𝜋𝑥
= [𝑥 ( ) cos − ( 2 2 ) sin ]
𝑙 𝑛𝜋 𝑙 𝑛 𝜋 𝑙 0

= 0 at 0 = 0 at 0
𝑙
2 −𝑙 𝑛𝜋𝑥 −𝑙 2 𝑛𝜋𝑥
+ [(𝑙 − 𝑥) ( ) cos − (−1) ( 2 2 ) sin ]
𝑙 𝑛𝜋 𝑙 𝑛 𝜋 𝑙 𝑙/2

= 0 at 𝑙 = 0 at 𝑙
2 𝑙 1 𝑛𝜋 𝑙2 𝑛𝜋 𝑙 2 𝑙 𝑛𝜋 𝑙2 𝑛𝜋
= [ (− ) cos + 2 2 sin ] + [ ∙ cos + 2 2 sin ]
𝑙 2 𝑛𝜋 2 𝑛 𝜋 2 2 𝑙 𝑛𝜋 2 𝑛 𝜋 2

4𝑙 𝑛𝜋
0 𝑖𝑓 𝑛 is even
= 𝑛2 𝜋2 sin ={ 4𝑙 (𝑛−1)/2 (7)
2 (−1) if 𝑛 is odd
𝑛 2 𝜋2

The solution for the initial displacement given by (4) and zero initial velocity is
4 (−1)𝑚−1 (2𝑚−1)𝑛𝜋 (2𝑚−1)𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = 𝜋 ∑∞
𝑛=1 (2𝑚−1)2 sin cos (8)
𝑙 𝑙

2.13.7 Vibrating String with Given Initial Velocity and Zero Initial Displacement
Next, we consider the case when the string is released from its horizontal position with zero
initial displacement but with an initial velocity given at 𝑥 by 𝑔(𝑥). The boundary value
problem for the displacement function is
𝜕2𝑦 𝜕2 𝑦
𝜕𝑡 2
= 𝑎2 𝜕𝑥 2 for 0 ≤ 𝑥 ≤ 𝑙, 𝑡 > 0 (2.126)

69
 𝑦(0, 𝑙 ) = 0 = 𝑦(𝑙, 𝑡) for 𝑡 > 0 (2.127),(2.128)
zero initial displacement:𝑦(𝑥, 0) = 0 for 0 ≤ 𝑥 ≤ 𝑙 (2.129)
𝜕𝑦
prescribed initial velocity: (𝑥, 0) = 𝑔(𝑥 ) for 0 ≤ 𝑥 ≤ 𝑙 (2.130)
𝜕𝑡
By the method of separation of variables we set 𝑦(𝑥, 𝑡) = 𝑥(𝑡)𝑇(𝑡) and obtain ordinary
differential equations
𝑋 ′′ + 𝜆2 𝑋 = 0 and 𝑇 ′′ + 𝑎2 𝜆2 𝑇 = 0 (2.131),(2.132)
The boundary conditions are same as before and hence we obtain eigenvalues

𝑛 2𝜋2
𝜆2𝑛 = (2.133)
𝑙2

and the corresponding eigen functions are constant multiples of

𝑛𝜋𝑥
𝑋𝑛 = sin 𝑙 (2.134)

with the values of 𝜆 as 𝜆 = 𝜆𝑛 = 𝑛𝜋/𝑙 the differential equation for 𝑇 becomes

𝑛 2 𝜋2
𝑇 ′′ + 𝑎2 𝑇=0 (2.135)
𝑙2

𝑎𝑡 𝑎𝑡
whose general solution is 𝑇(𝑡) = 𝐴 cos (𝑛𝜋 𝑙 ) + 𝐵 sin(𝑛𝜋 𝑙 )

The initial condition of zero initial displacement gives

𝑦(𝑥, 0) = 0 = 𝑋(𝑥 )𝑇(0) ⇒ 𝑇(0) = 0 (2.136)
Since T (0) = 𝐴 = 0 solutions for T𝑇(𝑡) are constant multiples of sin 𝑛𝜋𝑎𝑡/𝑙. Thus, for 𝑛 =
1,2,3, … we have functions
𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦𝑛 (𝑥, 𝑡) = 𝐶𝑛 sin 𝑙 sin 𝑙 (2.137)

Each of these functions satisfies the wave equation, the boundary conditions and the zero
initial dis- placement condition. In order to satisfy the initial velocity condition
𝜕𝑦
( ) (𝑥, 0) = 𝑔(𝑥) we invoke the superposition principle and write
𝜕𝑡
𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = ∑∞ ∞
𝑛=1 𝑦𝑛 (𝑥, 𝑡) = ∑𝑛=1 𝐶𝑛 sin sin (2.138)
𝑙 𝑙

We assume that the series admits of term-by-term differentiation. So, we get

𝜕𝑦 𝑛𝜋𝑎 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
(𝑥, 𝑡) = ∑∞
𝑛=1 𝐶𝑛 ( ) sin ( ) sin ( ) (2.139)
𝜕𝑡 𝑙 𝑙 𝑙

Now, the initial velocity condition yields

𝜕𝑦 𝑛𝜋𝑎 𝑛𝜋𝑥
(𝑥, 𝑡) = ∑∞
𝑛=1 𝐶𝑛 ( ) sin ( ) ∙ 1 = 𝑔(𝑥) (2.140)
𝜕𝑡 𝑙 𝑙

This is the half-range Fourier series expansion of 𝑔(𝑥) on [0, 𝑙 ]. Here, the entire coefficient
of sin 𝑛𝜋𝑥/𝑙 is the Fourier sine coefficient of 𝑔(𝑥 ) on [0, 𝑙 ] so that we have
𝑛𝜋𝑎 2 1 𝑛𝜋𝑥
𝐶𝑛 ∙ = ∫ 𝑔(𝑥 ) sin 𝑑𝑥
𝑙 𝑙 0 𝑙
2 1 𝑛𝜋𝑥
or 𝐶𝑛 = 𝑛𝜋𝑎 ∫0 𝑔(𝑥 ) sin 𝑑𝑥 (2.141)
𝑙

70
 Example 2.64
Find the displacement 𝑦(𝑥, 𝑡) of a string stretched between two fixed points at a distance
2𝑙 apart when the string is initially at rest in equilibrium position and the points of the string
are given initial velocity 𝑔 (𝑥 ), given by
𝑥
for 0≤𝑥≤𝑙
𝑙
( )
𝑔 𝑥 = {2𝑙 𝑥
for 𝑙 ≤ 𝑥 ≤ 2𝑙
𝑙
Solution
We have to solve the wave equation
𝜕2𝑦 𝜕2𝑦
= 𝑎2 𝜕𝑥 2 (1)
𝜕𝑡 2

under the boundary conditions:

𝑦(0, 𝑡) = 0 = 𝑦(2𝑙, 𝑡) (2)(3)
And the initial conditions:
Initial displacement: 𝑦(𝑥, 0) = 𝑓(𝑥 ) = 0 (4)
and initial velocity:
𝑥/𝑙 for 0≤𝑥≤𝑙
𝑔(𝑥, 0) = { } (5)
2𝑙 − 𝑥/𝑙 for 𝑙 ≤ 𝑥 ≤ 2𝑙
The solution is
2𝑙 𝑏𝑛 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = 𝜋𝑎 ∑∞
𝑛=1 𝑠𝑖𝑛 𝑠𝑖𝑛 (6)
𝑛 2𝑙 2𝑙

Where

2 2𝑙 𝑛𝜋𝑥
𝑏𝑛 = ∫ 𝑔(𝑥)𝑠𝑖𝑛 𝑑𝑥
2𝑙 0 2𝑙

1 𝑙𝑥 𝑛𝜋𝑥 1 2𝑙 2𝑙 − 𝑥 𝑛𝜋𝑥
= ∫ 𝑠𝑖𝑛 𝑑𝑥 + ∫ 𝑠𝑖𝑛 𝑑𝑥
𝑙 0 𝑙 2𝑙 𝑙 𝑙 𝑙 2𝑙
𝑙
1 𝑥 2𝑙 𝑛𝜋𝑥 1 4𝑙2 𝑛𝜋𝑥
= 𝑙 [ 𝑙 (−1) (𝑛𝜋) 𝑐𝑜𝑠 − 𝑙 (−1) 𝑛2𝜋2 𝑠𝑖𝑛 ]
2𝑙 2𝑙 0
2𝑙
1 2𝑙−𝑥 −2𝑙 𝑛𝜋𝑥 −1 4𝑙2 𝑛𝜋𝑥
=𝑙[ ∙ ( 𝑛𝜋 ) 𝑐𝑜𝑠 − ( 𝑙 ) ∙ 𝑛2𝜋2 𝑠𝑖𝑛 ]
𝑙 2𝑙 2𝑙 𝑙

2 𝑛𝜋𝑙 4 𝑛𝜋𝑙 2 𝑛𝜋 4 𝑛𝜋
= − 𝑛𝜋 𝑐𝑜𝑠 + 𝑛2 𝜋2 𝑠𝑖𝑛 + 𝑛𝜋 𝑐𝑜𝑠 + 𝑛2𝜋2 𝑠𝑖𝑛
2𝑙 2𝑙 2𝑙 2𝑙
8 𝑛𝜋
= 𝑛2 𝜋2 𝑠𝑖𝑛 (7)
2

Hence, the displacement function is given by

71
 2𝑙 8 1 1 𝑛𝜋 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = 𝜋𝑎 ∙ 𝜋2 ∑∞
𝑛=1 𝑛 ∙ 𝑛 2 𝑠𝑖𝑛 ∙ 𝑠𝑖𝑛 ∙ 𝑠𝑖𝑛
2 2𝑙 2𝑙

16𝑙 1 𝑛𝜋 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡

𝑦(𝑥, 𝑡) = 𝜋3 𝑎 ∙ ∑∞
𝑛=1 𝑛 3 𝑠𝑖𝑛 ∙ 𝑠𝑖𝑛 ∙ 𝑠𝑖𝑛 (8)
2 2𝑙 2𝑙

Example 2.65
A tightly stretched string with fixed end points is initially at rest in its equilibrium position, and
each of its points is given a velocity 𝑣, which is given by
𝑙
𝐶𝑥 for 0≤x≤
( )
𝑣 𝑥 ={ 2
𝑙
𝐶(𝑙 − 𝑥) for ≤x≤𝑙
2
Find the displacement 𝑦(𝑥, 𝑡).

Solution We have to solve the wave equation

𝜕2𝑦 𝜕2𝑦
= 𝑎2 𝜕𝑥 2 (1)
𝜕𝑡 2

under the boundary conditions:

𝑦(0, 𝑡) = 0 = 𝑦(𝑙, 𝑡) for t≥0 (2), (3)

and the initial conditions initial displacement:

𝑦(𝑥, 0) = 0 (0 ≤ 𝑥 ≤ 𝑙) (4)

Prescribed initial velocity:

𝑙
𝜕𝑦
𝐶𝑥 for 0≤x≤ 2
(𝑥, 0) = 𝑣(𝑥) = { 𝑙
} (5)
𝜕𝑡
𝐶(𝑙 − 𝑥) for ≤x≤𝑙
2

Solving equation (1) by the method of separation of variables under the conditions is given by
𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = ∑∞ ∞
𝑛=1 𝑦𝑛 (𝑥, 𝑡 ) = ∑𝑛=1 𝐶𝑛 sin sin (6)
𝑙 𝑙

Differentiating (6) partially with respect to ‘t’ we obtain

𝜕𝑦 𝑛𝜋𝑎 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
(𝑥, 0) = ∑∞
𝑛=1 𝐶𝑛 sin cos (7)
𝜕𝑡 𝑙 𝑙 𝑙

𝜕𝑦 𝑛𝜋𝑎 𝑛𝜋𝑥
⇒ (𝑥, 0) = ∑∞
𝑛=1 ( 𝐶𝑛 ) sin ∙ 1 = 𝑣 (𝑥 ). (8)
𝜕𝑡 𝑙 𝑙

This is the half-range Fourier sine series expansion for 𝑣(𝑥) in [0, 𝑙]. Therefore, we have
𝑛𝜋𝑎 2 𝑙 𝑛𝜋𝑥
∙ 𝐶𝑛 = 𝑙 ∫0 𝑣(𝑥) sin 𝑑𝑥
𝑙 𝑙

2 𝑙/2 𝑛𝜋𝑥 2 𝑙 𝑛𝜋𝑥

= ∫ 𝐶𝑥 sin 𝑑𝑥 + ∫ 𝐶(𝑙 − 𝑥) sin 𝑑𝑥
𝑙 0 𝑙 𝑙 𝑙/2 𝑙

72
 𝑙/2
2𝐶 𝑙 𝑛𝜋𝑥 𝑙2 𝑛𝜋𝑥
= [𝑥 (− ) cos − 1 ∙ (− 2 2 ) sin ]
𝑙 𝑛𝜋 𝑙 𝑛 𝜋 𝑙 0

= 0 𝑎𝑡 0 = 0 𝑎𝑡 0

𝑙
2𝐶 𝑙 𝑛𝜋𝑥 𝑙2 𝑛𝜋𝑥
+ [(𝑙 − 𝑥) (− ) cos − (−1) ∙ (− 2 2 ) sin ]
𝑙 𝑛𝜋 𝑙 𝑛 𝜋 𝑙 𝑙/2

= 0 𝑎𝑡 𝑙 = 0 𝑎𝑡 𝑙
2𝐶 𝑙2 𝑛𝜋 𝑙2 𝑛𝜋 2𝐶 𝑙2 𝑛𝜋 𝑙2 𝑛𝜋
= [− 𝑐𝑜𝑠 + 2 2 𝑠𝑖𝑛 ] − [− 𝑐𝑜𝑠 + − 2 2 𝑠𝑖𝑛 ]
𝑙 2𝑛𝜋 2 𝑛 𝜋 2 𝑙 2𝑛𝜋 2 𝑛 𝜋 2
4𝐶 𝑙2 𝑛𝜋 4𝐶𝑙 𝑛𝜋
= 𝑠𝑖𝑛 = 𝑛2𝜋2 𝑠𝑖𝑛
𝑙 𝑛 2 𝜋2 2 2

1 4𝐶𝑙 𝑛𝜋 4𝐶𝑙2 𝑛𝜋
⇒ 𝐶𝑛 = 𝑛𝜋𝑎 𝑛2𝜋2 𝑠𝑖𝑛 = 𝑎𝑛3 𝜋3 𝑠𝑖𝑛 (9)
2 2

Hence, the solution of the problem is

𝑛𝜋
4𝐶𝑙2 𝑠𝑖𝑛 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = ∑∞ 2
𝑠𝑖𝑛 𝑐𝑜𝑠 (10)
𝑎𝜋3 𝑛=1 𝑛 3 𝑙 𝑙

Example 2.66
Find the solution for the above problem when the string is released from its horizontal position
with an initial velocity given by 𝑔(𝑥 ) = 𝑥(𝑙 + cos 𝜋𝑥/𝑙).
Deduce the result for 𝑎 = 1 and l=π.
Solution
Following the procedure of the above problem, the solution in the present case is obtained as
𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑌(𝑥, 𝑡) = ∑∞ ∞
𝑛=1 𝑦𝑛 (𝑥, 𝑡 ) = ∑𝑛=1 𝐶𝑛 𝑠𝑖𝑛 𝑠𝑖𝑛 (1)
𝑙 𝑙

The prescribed initial velocity in this case is

𝜕𝑦 𝑛𝜋
(𝑥, 0) = 𝑔(𝑥 ) = 𝑥 (1 + 𝑐𝑜𝑠 ) (2)
𝜕𝑡 𝑙

Differentiating (1) partially with respect to ‘𝑡’ we have

𝜕𝑦 𝑛𝜋𝑎 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
(𝑥, 𝑡) = ∑∞
𝑛=1 ( ) 𝐶𝑛 𝑠𝑖𝑛 𝑐𝑜𝑠 (3)
𝜕𝑡 𝑙 𝑙 𝑙

Imposing the initial velocity condition (2) we obtain

𝜕𝑦 𝑛𝜋𝑎 𝑛𝜋𝑥 𝜋𝑥
(𝑥, 0) = ∑∞
𝑛=1 ( 𝐶𝑛 ) 𝑠𝑖𝑛 = 𝑔(𝑥 ) = 𝑥 + 𝑥𝑐𝑜𝑠 (4)
𝜕𝑡 𝑙 𝑙 𝑙

73
This is the half-range Fourier series expansion for 𝑔(𝑥 ) in [0, 𝑙 ]. Therefore, we have
𝑛𝜋𝑥 2 1 𝑛𝜋𝑥
𝐶𝑛 = ∫0 𝑔(𝑥 ) 𝑠𝑖𝑛 𝑑𝑥
𝑙 𝑙 𝑙

2 1 𝑛𝜋𝑥 𝑛𝜋𝑥
= ∫ 𝑥 (1 + 𝑐𝑜𝑠 ) 𝑠𝑖𝑛 𝑑𝑥
𝑙 0 𝑙 𝑙
2 1 𝑛𝜋𝑥 2 1 𝑛𝜋𝑥 𝑛𝜋𝑥
= 𝑙 ∫0 𝑥 𝑠𝑖𝑛 𝑑𝑥 + 𝑙 ∫0 𝑥 𝑠𝑖𝑛 𝑐𝑜𝑠 dx (5)
𝑙 𝑙 𝑙

For evaluation of the second integral, we have to separate the cases.

Case(1) 𝑛 = 1 and Case (2) 𝑛 ≠ 1
Case(1) 𝑛 = 1
𝜋𝑥 2 1 𝜋𝑥 2 1 𝜋𝑥 𝜋𝑥
𝐶1 = ∫0 𝑥 𝑠𝑖𝑛 𝑑𝑥 + ∫0 𝑥 𝑠𝑖𝑛 𝑐𝑜𝑠 dx
𝑙 𝑙 𝑙 𝑙 𝑙 𝑙
2 1 𝜋𝑥 1 1 2𝜋𝑥
= ∫0 𝑥 𝑠𝑖𝑛 𝑑𝑥 + ∫0 𝑥 𝑠𝑖𝑛 𝑑𝑥
𝑙 𝑙 𝑙 𝑙
2 1 𝜋𝑥 𝑙2 𝜋𝑥 1
= [𝑥 (− ) 𝑐𝑜𝑠 − 1 ∙ (− ) 𝑠𝑖𝑛 ]
𝑙 𝜋 𝑙 𝜋2 𝑙 0

= 0 𝑎𝑡 0 = 0 𝑎𝑡 𝑙, 0
1
1 1 2𝜋𝑥 𝑙2 2𝜋𝑥
+ [𝑥 (− ) 𝑐𝑜𝑠 − 1 ∙ (− 2 ) 𝑠𝑖𝑛 ]
𝑙 2𝜋 𝑙 4𝜋 𝑙 0

= 0 𝑎𝑡 0 = 0 𝑎𝑡 𝑙, 0
2 𝑙2 1 𝑙2 2𝑙 𝑙 3𝑙
= [− (−1)] + [− ∙ 1] = − =
𝑙 𝜋 𝑙 2𝜋 𝜋 2𝜋 2𝜋
3𝑙2
⇒ 𝐶1 = 2𝑎𝜋2 (6)

Case (2) 𝑛 ≠ 1
𝑛𝜋𝑥 2 1 𝑛𝜋𝑥 2 1 𝑛𝜋𝑥 𝜋𝑥
𝐶𝑛 = ∫0 𝑥 𝑠𝑖𝑛 𝑑𝑥 + ∫0 𝑥 𝑠𝑖𝑛 𝑐𝑜𝑠 dx
𝑙 𝑙 𝑙 𝑙 𝑙 𝑙
2 1 𝑛𝜋𝑥 1 1 (𝑛+1)𝜋𝑥
= ∫0 𝑥 𝑠𝑖𝑛 𝑑𝑥 + ∫0 [𝑥 𝑠𝑖𝑛 𝑑𝑥 +
𝑙 𝑙 𝑙 𝑙
(𝑥−1)𝜋𝑥
𝑥𝑠𝑖𝑛 𝑑𝑥]
𝑙
𝑙
2 𝑙 𝑛𝜋𝑥 𝑙2 𝑛𝜋𝑥
= [𝑥 (− ) . cos − 1 ∙ (− 2 2 ) 𝑠𝑖𝑛 ]
𝑙 𝑛𝜋 𝑙 𝑛 𝜋 𝑙 0

= 0 𝑎𝑡 0 = 0 𝑎𝑡 𝑙, 0
𝑙
1 𝑙 (𝑛 + 1)𝜋𝑥 𝑙2 (𝑛 + 1)𝜋𝑥
+ [𝑥 (− ) . cos − 1 ∙ (− 2 2
) 𝑠𝑖𝑛 ]
𝑙 (𝑛 + 1)𝜋 𝑙 (𝑛 + 1) 𝜋 𝑙 0

74
 = 0 𝑎𝑡 0 = 0 𝑎𝑡 𝑙, 0
𝑙
1 𝑙 (𝑛 − 1)𝜋𝑥 𝑙2 (𝑛 − 1)𝜋𝑥
+ [𝑥 (− ) . cos − 1 ∙ (− 2 2
) 𝑠𝑖𝑛 ]
𝑙 (𝑛 − 1)𝜋 𝑙 (𝑛 − 1) 𝜋 𝑙 0

= 0 𝑎𝑡 0 = 0 𝑎𝑡 𝑙, 0
2 𝑙2 1 𝑙2 1 𝑙2
= [− ∙ (−1)𝑛 ] + [− ∙ (−1)𝑛+1 ] + [− ∙ (−1)𝑛−1 ]
𝑙 𝑛𝜋 𝑙 (𝑛 + 1)𝜋 𝑙 (𝑛 − 1)𝜋
2𝑙 (−1)𝑛 𝑙 (−1)𝑛 1 1 2𝑙 (−1)𝑛 1
=− + ( + )= ∙ 2
𝑛𝜋 𝜋 𝑛+1 𝑛−1 𝜋𝑛 𝑛 −1
2𝑙2 (−1)𝑛
⇒ 𝐶𝑛 = 𝑎𝜋2 𝑛2(𝑛2−1) (𝑛 ≠ 1) (7)

Therefore, the solution for 𝑔(𝑥) = 𝑥(1 + cos 𝜋𝑥/𝑙) as the initial velocity function is
3𝑙2 𝜋𝑥 𝜋𝑎𝑡 2𝑙2 (−1)𝑛 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = 2𝑎𝜋2 sin sin + 𝑎𝜋2 ∑∞
𝑛=2 𝑛 2 (𝑛 2−1) sin sin (8)
𝑙 𝑙 𝑙 𝑙

By taking 𝑎 = 1 and 𝑙 = 𝜋 in the above result, we obtain the solution for this particular case
as
3 (−1)𝑛
𝑦(𝑥, 𝑡) = 2 sin 𝑥 sin 𝑡 + 2 ∑∞
𝑛=2 𝑛 2 (𝑛 2−1) sin 𝑛𝑥 sin 𝑛𝑡 (9)

2.13.8 Vibrating String with Initial Displacement and Initial Velocity

Consider the motion of the string with both initial displacement given by 𝑓(𝑥) and initial
velocity given by 𝑔(𝑥). We have to now solve two separate problems, one with initial
displacement 𝑓(𝑥) and zero initial velocity and the other with zero displacement and initial
velocity 𝑔(𝑥). Let 𝑦1 (𝑥, 𝑡) and 𝑦2 (𝑥, 𝑡) be the respective solutions of the two problems and
let
𝑦(𝑥, 𝑡) = 𝑦1 (𝑥, 𝑡) + 𝑦2 (𝑥, 𝑡) (2.142)
Then y satisfies the wave equation and the boundary conditions. Further,

𝑦(𝑥, 0) = 𝑦1 (𝑥, 0) + 𝑦2 (𝑥, 0) = 𝑓(𝑥 ) + 0 = 𝑓(𝑥) (2.143)

and
𝜕𝑦 𝜕𝑦1 𝜕𝑦2
(𝑥, 0) = (𝑥, 0) + (𝑥, 0) = 0 + 𝑔(𝑥 ) = 𝑔(𝑥) (2.144)
𝜕𝑡 𝜕𝑡 𝜕𝑡

Thus, 𝑦(𝑥, 𝑡) is the solution in the case of nonzero initial displacement and velocity
functions.
Example 2.67

An elastic string of length 𝑙, fastened at its ends on the 𝑥-axis at 𝑥 = 0 and 𝑥 = 𝜆, is given
initial displacement 𝑓(𝑥) and initial velocity 𝑔(𝑥). Find the displacement function 𝑦(𝑥, 𝑡) by
solving the wave equation 𝑦𝑡𝑡 = 𝑎2 𝑦𝑥𝑥 under the conditions: 𝑦(0, 𝑡) = 𝑦(𝑙, 𝑡) =
0, 𝑦(𝑥, 0) = 𝑓 (𝑥 ) = 𝑥 for

75
 𝑙
𝑥 for 0≤x≤
( )
𝑦 𝑥, 0 = { 2}
𝑙
𝑙−𝑥 for ≤x≤𝑙
2
and
𝜋𝑥
y𝑡 (𝑥 ) = 𝑔(𝑥 ) = {𝑥 (1 + 𝑐𝑜𝑠 ).
𝑙

solution
from equation (2.143) and (2.144), we obtain the solution for the present problem as
4𝑙 (−1)𝑚−1 (2𝑚−1)𝜋𝑥 (2𝑚−1)𝜋𝑎𝑡
𝑦(𝑥, 𝑡) = ∑∞
𝑚=1 ∙ sin cos
𝜋2 (2𝑚−1)2 𝑙 𝑙

3𝑙 2 𝜋𝑥 𝜋𝑎𝑡 2𝑙 2 (−1)2 𝑛𝜋𝑥 𝑛𝜋𝑎𝑡

+ 𝑎𝜋2 sin 𝑙
sin 𝑙
+ 𝑎𝜋2 ∑∞
𝑛=2 𝑛2 (𝑛2 −1) sin 𝑙
𝑠𝑖𝑛 𝑙
(2.145)

2.13.9 Laplace’s Equation or Potential Equation or Two Two-dimensional Steady-State

Heat Flow Equation
The two dimensional heat conduction equation is given by
𝜕𝑢 𝜕2 𝜕2
= 𝑎2 ∇2 𝑢, where ∇2 = 𝜕𝑥 2 + 𝜕𝑦 2 is the laplacian operator in the case of steady-state heat
𝜕𝑡
𝜕𝑢
flow = 0 and equation reduce to
𝜕𝑡

2
𝜕2𝑢 𝜕2𝑢
∇ 𝑢= 2+ 2=0
𝜕𝑥 𝜕𝑦
The solution 𝑢(𝑥, 𝑦) of the above equation can be obtained by the method of separation of
variable in a rectangular region both in the dirichlet problem as well as in Neumann’s problem.
A rectangular thin plate with its two faces insulated is considered so that the heat flow is two-
dimensional. The boundary conditions are prescribed on the four edges of the edges of the
plate.
Example: 2.68
𝜕 2𝑢 𝜕 2𝑢
Solve Laplace’s equation 𝜕𝑥 2 + 𝜕𝑦 2 = 0 (1)

In the rectangle; 0 < 𝑥 < 𝑎, 0 < 𝑦 < 𝑏 in the 𝑥𝑦 –plane, with the boundary conditions
𝑢(𝑥, 0) = 0 on 𝑂𝐴; 𝑢(𝑥, 𝑏) = 0 on 𝐵𝐶 (2), (3)

𝑢(0, 𝑦) = 0 on 𝑂𝐶; 𝑢(𝑎, 𝑦) = 𝑓(𝑦) on 𝐴𝐵 (4)(5)

76
 Figure: 2.5 Two-dimentional heat flow in a rectangular plane
Solution
Let 𝑢(𝑥, 𝑦) = 𝑋(𝑥)𝑌(𝑦) (6)
Substituting in(1) we get
𝒀′′ 𝑿′′
= = −𝝀𝟐 (7)
𝒀 𝑿

where the separation constant is taken as negative to get non-trivial solutions. The boundary
value problem reduces to solution of the ordinary differential equations
𝑌 ′′ + 𝜆2 𝑌 = 0, 𝑋 ′′ − 𝜆2 𝑋 = 0 (8)(9)
Under the conditions
0 = 𝑢(𝑥, 0) = 𝑋(𝑥 )𝑌(0) ⇒ 𝑌(0) = 0 (10)
0 = 𝑢(𝑥, 𝑏) = 𝑋 (𝑥 )𝑌(𝑏) ⇒ 𝑌 (𝑏) = 0 (11)
0 = 𝑢(0, 𝑦) = 𝑋(0)𝑌(𝑦) ⇒ 𝑋(0) = 0 (12)
The general solution of equation (8) is

𝑌(𝑦) = 𝐴 𝑐𝑜𝑠 𝜆𝑦 + 𝐵 𝑠𝑖𝑛 𝜆𝑦

By (10) we get 𝐴 = 0 and by (11) we get 𝑠𝑖𝑛 𝜆𝑏 = 0
𝑛𝜋
⇒ 𝜆𝑛 = (𝑛 = 1,2,3, … ) (13)
𝑏

𝜆𝑛 = 𝑛𝜋/𝑏 are the eigenvalues and the corresponding eigenfunctions are

𝑛𝜋𝑦
𝑌𝑛 (𝑦) = 𝑠𝑖𝑛 (with B=1) (14)
𝑏

Now the general solution of (9) is

77
 𝑛𝜋𝑥 𝑛𝜋𝑥
𝑋𝑛 (𝑥 ) = 𝐶 𝑐𝑜𝑠 ℎ +𝐷 𝑠𝑖𝑛 ℎ (15)
𝑏 𝑏

using (12) we get 𝐶 = 0 so that

𝑛𝜋𝑥
𝑋𝑛 (𝑥) = 𝐷 𝑠𝑖𝑛
𝑏
Therefore, the solution of equation (1) satisfying the boundary conditions (2), (3) and (4) is
𝑛𝜋𝑥 𝑛𝜋𝑦
𝑢𝑛 (𝑥, 𝑦) = 𝐶𝑛 𝑠𝑖𝑛 ℎ 𝑠𝑖𝑛
𝑏 𝑏
where we have replaced 𝐷 by 𝐶
By the principle of superposition we write the solution as
𝑛𝜋𝑥 𝑛𝜋𝑦
𝑢(𝑥, 𝑦) = ∑∞ ∞
𝑛=1 𝑢𝑛 (𝑥, 𝑦) = ∑𝑛=1 𝐶𝑛 𝑠𝑖𝑛 ℎ 𝑠𝑖𝑛 (16)
𝑏 𝑏

Lastly we have condition (5) namely (𝑦) 𝑓(𝑦) to be satisfied. This gives
𝑛𝜋𝑎 𝑛𝜋𝑦
𝑢(𝑎, 𝑦) = ∑∞
𝑛=1 𝐶𝑛 𝑠𝑖𝑛 ℎ 𝑠𝑖𝑛 = 𝑓(𝑦) (17)
𝑏 𝑏

This is a half-range Fourier sine series expansion of f ( y) in (0, 𝑏) and the constants 𝐶 are
given by
𝑛𝜋𝑎 2 𝑏 𝑛𝜋𝑦
𝐶𝑛 𝑠𝑖𝑛 ℎ 𝑏 = 𝑏 ∫0 𝑓(𝑦)𝑠𝑖𝑛 𝑏 𝑑𝑦
2 𝑏 𝑛𝜋𝑦
𝐶𝑛 =
𝑏𝑠𝑖𝑛 ℎ
𝑛𝜋𝑎 ∫0 𝑓 (𝑦)𝑠𝑖𝑛 𝑏
𝑑𝑦 (18)
𝑏

Thus, this harmonic function 𝑢(𝑥, 𝑦) satisfying Laplace’s equation (1) and the boundary
conditions
(2) − (5) is given by (16) where the constants 𝐶𝑛 are determined by (18) for any specific
function 𝑓 (𝑦).
Example 2.69
A rectangular plate is bounded by the lines 𝑥 = 0, 𝑦 = 0, 𝑥 = 𝑎, and 𝑦 = 𝑏, and the edge
5𝜋𝑥
temperatures are 𝑢(0, 𝑦) = 𝑢(𝑥, 𝑏) = 𝑢(𝑎, 𝑦) and 𝑢(𝑥, 0) = 5 sin ( ) + 3 sin(3𝜋𝑥/𝑎).
𝑎
Find the steady-state temperature at any point of the plate.
Solution Let 𝑢(𝑥, 𝑦) be the steady-state temperature at any point 𝑝(𝑥, 𝑦) of the rectangular
plate. We have to solve Laplace’s equation

𝜕2𝑢 𝜕 2𝑢
+ 𝜕𝑦 2 = 0 … (1)
𝜕𝑥 2

under the boundary conditions

1. 𝑢(0, 𝑦) = 0 for 0 < 𝑦 < 𝑏 (2)
2. 𝑢(𝑎, 𝑦) = 0 for 0 < 𝑦 < 𝑏 (3)

78
3. 𝑢(𝑥, 𝑏) = 0 for 0 < 𝑦 < 𝑎 (4)
5𝜋𝑥 3𝜋𝑥
4. 𝑢(𝑥, 0) = 5 sin + 3 sin for 0 < 𝑥 < 𝑎 (5)
𝑎 𝑎

A suitable solution of equation (1), by the method of separation of variables, satisfying

boundary conditions (2) − (4) for each 𝑛 = 1, 2, 3, . .. is
𝑛𝜋𝑥 𝑛𝜋(𝑦−𝑏)
𝑢𝑛 (𝑥, 𝑦) = 𝐶𝑛 𝑠𝑖𝑛 𝑠𝑖𝑛 ℎ (6)
𝑏 𝑏

By the principle of superposition of solution we may write the general solution as

𝑛𝜋𝑥 𝑛𝜋(𝑦−𝑏)
𝑢(𝑥, 𝑦) = ∑∞ ∞
𝑛=1 𝑢𝑛 (𝑥, 𝑦) = ∑𝑛=1 𝐶𝑛 𝑠𝑖𝑛 𝑠𝑖𝑛 ℎ (7)
𝑏 𝑏

Imposing condition (5) we get

𝑛𝜋𝑥 𝑛𝜋𝑏 5𝜋𝑥 3𝜋𝑥

𝑢(𝑥, 0) = − ∑∞
𝑛=1 𝐶𝑛 𝑠𝑖𝑛 𝑠𝑖𝑛 ℎ = 5 sin + 3 sin (8)
𝑎 𝑎 𝑎 𝑎

Equating the coefficients of like terms on either side we get

5𝜋𝑏 3𝜋𝑏
−𝐶5 sin h = 5 and −𝐶5 sin h = 3 and 𝐶𝑛 = 0 for all 𝑛 ≠ 3 or 5 (9)
𝑎 𝑎

Hence, the required solution of steady-state temperature is

3 3𝜋𝑏 3𝜋(𝑏−𝑦) 5 5𝜋𝑏 5𝜋(𝑏−𝑦)
𝑢(𝑥, 𝑦) = 3𝜋𝑏 sin sin ℎ + 5𝜋𝑏 sin sin ℎ (10)
sin ℎ 𝑎 𝑎 sin ℎ 𝑎 𝑎
𝑎 𝑎

Example 2.70
An infinitely long plane uniform plate is bounded by two parallel edges and an end at right
angles to them. The breadth in 𝜋. This end is maintained at a temperature 𝑢0 at all points
and the other edges are at zero temperature. Determine the temperature at any point of the
plate in the steady state.

79
 Figure 2.6 Temperature in an infinitely long plate
Solution
Let 𝑢(𝑥, 𝑦) be the temperature at any point 𝑝(𝑥, 𝑦) of the plate. Then the steady state
temperature distribution is given by Laplace’s equation
𝜕2𝑢 𝜕 2𝑢
+ 𝜕𝑦 2 = 0 … (1)
𝜕𝑥 2

with the boundary conditions

1. 𝑢(0, 𝑦) = 0; 2. 𝑢(𝜋, 𝑦) = 0 for all 𝑦

3. 𝑢(𝑥, 0) = 𝑢0 ; 4. 𝑢(𝑥, ∞) = 0 for 0 < 𝑥 < 𝜋
The general solution of equation (1) is

𝑢(𝑥, 𝑦) = (𝐴 cos 𝜆𝑥 + 𝐵 sin 𝜆𝑥)(𝐶𝑒 𝜆𝑥 + 𝐷𝑒 −𝜆𝑦 ) (2)

By condition (1): 𝑢(0, 𝑦) we get 𝐴 = 0. Now equation (2) becomes

𝑢(𝑥, 𝑦) = 𝐵 sin 𝜆𝑥 (𝐶𝑒 𝜆𝑥 + 𝐷𝑒 −𝜆𝑦 ) (3)

By condition (2): 𝑢(𝜋, 𝑦) = 0 we get
sin 𝜆𝜋 = 0 ⇒ 𝜆 = 𝜆𝑛 = 𝑛 (4)
so that the solution becomes 𝑢(𝑥, 𝑦) = (𝑐𝑛 𝑒 𝑛𝑦 + 𝑑𝑛 𝑒 −𝑛𝑦 ) sin 𝑛𝑥 n where 𝐵𝐶 and 𝐵𝐷 have
been replaced by 𝑐𝑛 𝑎𝑛𝑑 𝑑𝑛, respectively.
By condition (4):
𝑢(𝑥, ∞) = 0 we have 𝑐𝑛 = 0 for all n (5)
Finally, the general solution satisfying the conditions (1), (2) and (4) is (by the principle of
superposition)
𝑢(𝑥, 𝑦) = ∑∞ ∞
𝑛=1 𝑢𝑛 (𝑥, 𝑦) = ∑𝑛=1 𝑑𝑛 sin 𝑛𝑥 𝑒
−𝑛𝑦
(6)
Now applying boundary condition (3):

𝑢(𝑥, 0) = ∑∞
𝑛=1 𝑑𝑛 sin 𝑛𝑥 = 𝑢0 (7)

80
 Equation (7) is a half-range Fourier sine series expansion for 𝑢0 in (0, 𝜋) and hence the
Fourier constants 𝑑𝑛 are given by

2 𝜋 −2𝑢0 cos 𝑛𝑥 𝜋
𝑑𝑛 = ∫ 𝑢0 sin 𝑛𝑥 𝑑𝑥 = ( )
𝜋 0 𝜋 𝑛 0

−2𝑢0 0 if 𝑛 is even
= (cos 𝑛𝜋 − 1) = { 4𝑢0
if 𝑛 is odd (8)
𝑛𝜋
𝑛𝜋

Now (6) reduces to

4𝑢0
𝑢(𝑥, 𝑦) = ∑𝑛=1,3,5 sin 𝑛𝑥 𝑒 −𝑛𝑦 (9)
𝑛𝜋

which is the required solution.

Example 2.71
The temperature 𝑢(𝑥, 𝑦) is maintained at 0°𝐶 along three edges of a square plate of side
100 𝑐𝑚 and the fourth edge is maintained at a constant temperature 𝑢0 until steady-state
condition’s prevail. Find the temperature at any point 𝑢(𝑥, 𝑦) of the plate and also at the centre
of the plate.

Figure 2.7 Temperature in a square plate

Solution : Let the side of the plate be 𝑎 = 100 𝑐𝑚. The temperature function 𝑢(𝑥, 𝑦) satisfies
Laplace’s equation
𝜕2𝑢 𝜕 2𝑢
+ 𝜕𝑦 2 = 0 … (1)
𝜕𝑥 2

with the boundary conditons:

1. 𝑢(0, 𝑦) = 0; 2. 𝑢(𝑎, 𝑦) = 0 for 0≤𝑦 ≤ 𝑎
3. 𝑢(𝑥, 0) = 0; 4. 𝑢(𝑥, 𝑎) = 𝑢0 for 0 ≤ 𝑥 ≤ 𝑎.
The general solution of equation (1) is

81
 𝑢(𝑥, 𝑦) = (𝐶1 cos 𝑝𝑥 + 𝐶2 sin 𝑝𝑥)(𝐶3 𝑒 𝑝𝑥 + 𝐶4 𝑒 −𝑝𝑦 ) (2)

By condition (1) we get 𝐶1 = 0 so that equation (2) becomes

𝑢(𝑥, 𝑦) = 𝐶2 sin 𝑝𝑥 (𝐶3 𝑒 𝑝𝑥 + 𝐶4 𝑒 −𝑝𝑦 ) (3)

𝑛𝜋
By condition (2)we get 𝑢(𝑎, 𝑦) = 0 = 𝐶2 sin 𝑝𝑎 (𝐶3 𝑒 𝑝𝑥 + 𝐶4 𝑒 −𝑝𝑦 ) ⇒ 𝑝 = ( 𝑎 ) (𝑛 = 1,2,3),
so that we have
𝑛𝜋𝑥
𝑢(𝑥, 𝑦) = 𝐶2 sin (𝐶3 𝑒 𝑛𝜋𝑦/𝑎 + 𝐶4 𝑒 −𝑛𝜋𝑦/𝑎 ) (4)
𝑎

Now, condition (3) implies that 𝐶4 = −𝐶3 (5)

The solution may therefore be written as
𝑛𝜋𝑥 𝑛𝜋𝑦
𝑢(𝑥, 𝑦) = 𝐶𝑛 sin sin ℎ (𝑛 = 1,2,3, … ) (6)
𝑎 𝑎

For each 𝑛. By the principle of superposition we may write

𝑛𝜋𝑥 𝑛𝜋𝑦
𝑢(𝑥, 𝑦) = ∑∞ ∞
𝑛=1 𝑢𝑛 (𝑥, 𝑦) = ∑𝑛=1 𝐶𝑛 sin sin ℎ (7)
𝑎 𝑎

Imposing the condition (4) we have

𝑛𝜋𝑦
𝑢(𝑥, 𝑎) = 𝑢0 = ∑∞
𝑛=1 𝐶𝑛 sinh 𝑛𝜋, sin (8)
𝑎

Equation (8) is a half-range function sine series expansion of 𝑢0 in (0, 𝑎), and the Fourier
coefficients are obtained by
2 𝑎 𝑛𝜋𝑥
𝐶𝑛 sinh 𝑛𝜋 = ∫ 𝑢0 sin 𝑑𝑥
𝑎 0 𝑎
𝑎
2 𝑛𝜋𝑥
C𝑛 = ∫ 𝑢0 sin 𝑑𝑥
𝑎 sinh 𝑛𝜋 0 𝑎
2 −𝑎 𝑛𝜋𝑥 𝑎
⇒ C𝑛 = ∙ ( ) [cos ]
𝑎 sinh 𝑛𝜋 𝑛𝜋 𝑎 0
2 −𝑎
= ∙ ( ) ∙ (cos 𝑛𝜋 − 1)
𝑎 sinh 𝑛𝜋 𝑛𝜋
0 if 𝑛 is even
={ 4𝑢0
if 𝑛 is odd (9)
𝑛𝜋 sinh 𝑛𝑥

Substituting for 𝐶𝑛 in (7) we have

∞
4𝑢0 𝑛𝜋𝑥 𝑛𝜋𝑦
𝑢(𝑥, 𝑦) = ∑ sin sinh
𝑛𝜋 sinh 𝑛𝑥 𝑎 𝑎
𝑛=1,3,5,…

82
 Where 𝑎 = 100 𝑐𝑚 (10)

Temperature at the centre = (50,50) is

4𝑢 𝑛𝜋 𝑛𝜋
𝑢(50,50) = ∑∞ 0
𝑛=1,3,5,… 𝑛𝜋 sinh 𝑛𝑥 sin sinh (11)
𝑎 𝑎

Example 2.72

Find the steady-state temperature in a rectangular plate 0 ≤ 𝑥 ≤ 𝑎, 0 ≤ 𝑦 ≤ 𝑏 when the side

𝑐 𝜋𝑥
𝑥 = 0, 𝑥 = 𝑎, and 𝑦 = 𝑏 are insulated while the edge 𝑦 = 0 is kept at temperature (𝑎) cos .
𝑎

Figure 2.8 Steady state temperature in a rectangular plate

Solution we have to solve Laplace’s equation

𝜕2𝑢 𝜕 2𝑢
+ 𝜕𝑦 2 = 0 (1)
𝜕𝑥 2

under the boundary condition

𝑢𝑥 |𝑥=0 = 𝑢𝑥 |𝑥=𝑎 = 𝑢𝑦 |𝑦=𝑏 = 0 (2), (3), (4)

𝜋𝑥
And 𝑢(𝑥, 0) = 𝑓(𝑥) = 𝑐 𝑐𝑜𝑠 (5)
𝑎

By separating the variables we get the following two ordinary differential equations:
𝑋 ′′ + 𝜆2 𝑋 = 0 and 𝑌 ′′ − 𝜆2 𝑌 = 0, (6), (7)
these solution are
𝑋(𝑥 ) = 𝐴 𝑐𝑜𝑠 𝜆𝑥 + 𝐵 𝑠𝑖𝑛 𝜆𝑥; 𝑌(𝑦) = 𝐶𝑒 𝜆𝑥 + 𝐷𝑒 −𝜆𝑥 (8), (9)

83
The above boundary conditions imply that

𝑋 ′ (0) = (−𝐴𝜆 sin 𝜆𝑥 + 𝐵𝜆 cos 𝜆𝑥)𝑥=0 = 0 ⇒ 𝐵 = 0

𝑛𝜋
𝑋 ′ (𝑎) = 0 = −𝐴𝜆 𝑠𝑖𝑛𝜆𝑎 ⇒ 𝜆 = (𝑛 = 0,1,2)
𝑎
𝑛𝜋𝑥
𝑋𝑛 (𝑥) = 𝐴𝑛 cos 𝑛 = 0,1,2, …
𝑎

′( 𝜆𝑏 −𝜆𝑏 )
𝐷𝑒 −𝜆𝑏
𝑌 𝑎) = 𝜆(𝐶𝑒 − 𝐷𝑒 = 0 ⇒ 𝐶 = 𝜆𝑏
𝑒
𝑛𝜋(𝑏−𝑦)
Thus 𝑌(𝑦) = 𝐶𝑛 cosh 𝑎

Hence, the required solution is

𝑛𝜋𝑥 𝑛𝜋(𝑏 − 𝑦)
𝑢(𝑥, 𝑦) = ∑ 𝐴𝑛 cos cosh
𝑎 𝑎

where
𝑎
2 𝑛𝜋𝑥
𝐴𝑛 = ∫ 𝑓(𝑥) cos 𝑑𝑥
𝑎 cosh(𝑛𝜋𝑏/𝑎) 0 𝑎
𝑎
2𝑐 𝑛𝑥 𝑛𝜋𝑥
= ∫ cos cos 𝑑𝑥
𝑎 cosh(𝑛𝜋𝑏/𝑎) 0 𝑎 𝑎

2𝑐 𝑎
𝐴1 = ( )
𝑎 cosh(𝜋𝑏/𝑎) 2

𝐴𝑛 = 0 for 𝑛 = 0 and 𝑛 ≥ 2

𝜋𝑏 𝜋𝑥 𝜋(𝑏 − 𝑦)
∴ 𝑢(𝑥, 𝑦) = 𝑐 sech cos 𝑐𝑜𝑠ℎ
𝑎 𝑎 𝑎

2.14 TWO-DIMENSIONAL WAVE EQUATION

Vibrations of a rectangular membrane

To solve the problem of a vibrating, we have to determine a solution 𝑢(𝑥, 𝑦, 𝑡) of the two-
dimensional wave equation

𝜕2𝑢 𝜕 2𝑢 𝜕2𝑢 𝑇
= 𝑐 2 ( 𝜕𝑥 2 + 𝜕𝑦 2 ) (𝑐 2 = 𝑃) (1)
𝜕𝑡 2

with the boundary conditions 𝑢 = 0, (2)

on the boundary of the membrane for all 𝑡 ≥ 0

84
and two initial conditions

𝑢(𝑥, 𝑦, 𝑜) = 𝑓(𝑥, 𝑦) (3)

and
𝜕𝑢
| = 𝑔(𝑥, 𝑦) (4)
𝜕𝑡 𝑡=0

Where 𝑓(𝑥, 𝑦) is the given initial displacement and 𝑔(𝑥, 𝑦) is the given initial velocity.

𝑢(𝑥, 𝑦) gives the displacement of the point (𝑥, 𝑦) of the membrane from rest (𝑢 = 0) at time
𝑡 = 0. We note that the conditions (2) − (4) are similar to those for the vibrating string.

As a first important case, we consider the rectangular membrane 𝑅 shown in Fig.

Figure 2.9 Rectangular membrane

By the method of separation of variables we first find solutions of (1) satisfying 𝑏. 𝑐. (2).

𝑢(𝑥, 𝑦, 𝑡) = 𝐹 (𝑥, 𝑦)𝑇(𝑡) (5)

By substituting this into equation (1) we have

𝐹𝑇̈ = 𝑐 2 (𝐹𝑥𝑥 𝑇 + 𝐹𝑦𝑦 𝑇)

Where subscripts denote partial derivatives and dots denote derivatives w. r. t. 𝑡.

We divide both sides by 𝑐 2 𝐹𝑇 and obtain

𝑇̈ 1
2
= (𝐹𝑥𝑥 + 𝐹𝑦𝑦 )
𝑐 𝑇 𝐹

The LHS depends only on 𝑡, while the RHS is independent of 𝑡. so, each side must be equal to
a constant, −𝑘 2 say.Therefore,

𝑇̈ 1
2
= (𝐹𝑥𝑥 + 𝐹𝑦𝑦 ) = −𝑘 2
𝑐 𝑇 𝐹
This gives two equations. For the time function 𝑇(𝑡), We have the ordinary differential

85
 equation

𝑇̈ + 𝜆2 𝑇 = 0 (6)

where 𝜆2 = 𝑐 2 𝑘 2 and for the amplitude function 𝐹(𝑥, 𝑦),

we have the partial differential equation

𝐹𝑥𝑥 + 𝐹𝑦𝑦 + 𝑘 2 𝐹 = 0 (7)

This is known as the two-dimensional Helmholtz equation,

To separate this equation we set

𝐹 (𝑥, 𝑦) = 𝜙(𝑥)𝜓(𝑦) (8)

Substituting this into (7) we get

𝜙 ′′ 𝜓 + 𝜙𝜓 ′′ + 𝑘 2 𝜙𝜓 = 0
Where prime denotes differentiation w. r. t the argument
Hermann Von Helmholtz (1821–1894) German physicist known for his basic work in
thermo- dynamics, fluid flow and acoustics.

𝜙 ′′ 1
= − 𝜓 (𝜓 ′′ + 𝑣 2 𝜓) = −𝑝2 (9)
𝜙

Thus we get two ordinary linear differential equations for 𝜙 and 𝜓

𝜙 ′′ + 𝑝2 𝜙 = 0, 𝜓 ′′ + 𝑝2 𝜓 = 0, 𝑞 2 = 𝐾 2 − 𝑝2 (10)(𝑎), (𝑏)

The general solution of 10(𝑎), (𝑏) are

𝜙 (𝑥) = 𝐴 cos 𝑝𝑥 + 𝐵 sin 𝑝𝑥, 𝜓(𝑦) = 𝐶 cos 𝑞𝑦 + 𝐷 sin 𝑞𝑦,

Where A, B, C and D are constants. Since 𝐹 (𝑥, 𝑦) = 𝜑(𝑥 )𝜓(𝑥 ) must be zero on the boundary
i.e. where 𝑥 = 0, 𝑥 = 𝑎, 𝑦 = 0 𝑎𝑛𝑑 𝑦 = 𝑏 therefore we get
𝜑(0) = 0, 𝜑(𝑎) = 0, 𝜓(0) = 0, 𝜓(𝑏) = 0
Now,
𝜑(0) = 0 ⇒ 𝐴 = 0 and 𝜑(𝑎) = 0 ⇒ 𝐵 sin 𝑝𝑎 = 0
⇒ 𝑝𝑎 = 𝑚𝜋
𝑚𝜋
⇒𝑝= (𝑚 ∈ 𝑧 + )
𝑎
𝑛𝜋
Similarly 𝜓(0) = 0 ⇒ 𝐶 = 0 and 𝜓(𝑏) = 0 ⇒ 𝑞 = (𝑛 ∈ 𝑧 + )
𝑏

We, therefore obtain the solutions

𝑚𝜋𝑥 𝑚𝜋𝑦
𝜑𝑚 (𝑥 ) = sin and 𝜓𝑛 (𝑦) = sin 𝑚, 𝑛 = 1,2,3, … …
𝑎 𝑏
𝑚𝜋𝑥 𝑛𝜋𝑦
𝐹𝑚𝑛 (𝑛, 𝑦) = 𝜑𝑚 (𝑥 )𝜑𝑛 (𝑦) = sin sin 𝑚, 𝑛 = 1,2,3, … … (11)
𝑎 𝑏

86
Eigenfunctions and eigenvalues
Since 𝑞 2 = 𝑘 2 − 𝑝2 and λ= 𝑐𝑘 we have

λ = c√𝑝2 + 𝑞 2
𝑚𝜋 𝑛𝜋
Hence to 𝑝 = and 𝑞 = there corresponds the value
𝑎 𝑏

𝑚2 𝑛2
λ = λ𝑚 = 𝑐𝜋√ 𝑎2 + 𝑏2 𝑚, 𝑛 = 1,2,3, … …
(12)
Now, the general solution of
𝑇𝑚𝑛 (𝑡) = 𝐵𝑚𝑛 cos λ𝑚𝑛 𝑡 + 𝐵′ 𝑚𝑛 sin λ𝑚𝑛 𝑡
Finally the general solution of equation is
𝑚𝜋𝑥 𝑛𝜋𝑦
𝑢𝑚𝑛 (𝑥, 𝑦, 𝑏) = (𝐵𝑚𝑛 cos λ𝑚𝑛 𝑡 + 𝐵′ 𝑚𝑛 sin λ𝑚𝑛 𝑡) × sin sin
𝑎 𝑏
(13)
These functions are called the eigenfunctions and the numbers, λ𝑚𝑛 are called the eigenvalues
λ𝑚𝑛
of the vibrating membrane. The frequency of 𝑢𝑚𝑛 is .
2𝜋

Solution of the Entire problem and Double Fourier series

By the principle of superposition of solutions the general solution of the problem may be
written as
∞ ∞

𝑢𝑚𝑛 (𝑥, 𝑦, 𝑏) = ∑ ∑ 𝑢𝑚𝑛 (𝑥, 𝑦, 𝑏)

𝑚=1 𝑛=1
∞ ∞
𝑚𝜋𝑥 𝑛𝜋𝑦
= ∑ ∑(𝐵𝑚𝑛 cos λ𝑚𝑛 𝑡 + 𝐵′ 𝑚𝑛 sin λ𝑚𝑛 𝑡) sin sin (14)
𝑎 𝑏
𝑚=1 𝑛=1

Imposing condition (3) we obtain

∞ ∞
𝑚𝜋𝑥 𝑛𝜋𝑦
𝑢(𝑥, 𝑦, 0) = ∑ ∑ 𝐵𝑚𝑛 sin sin sin = 𝑓(𝑥, 𝑦) (15)
𝑎 𝑏
𝑚=1 𝑛=1
This series is called a double Fourier series. The constants in (15) are determined by
setting
∞
𝑥𝜋𝑦
𝐾𝑚 (𝑦) = ∑ 𝐵𝑚𝑛 sin (16)
𝑏
𝑛=1

And writing (15) in the form

∞
𝑥𝜋𝑦
𝑓(𝑥, 𝑦) = ∑ 𝐾𝑚 (𝑦)sin (17)
𝑏
𝑛=1

87
For fixed y this is the Fourier since series of 𝑓(𝑥, 𝑦), considered as a function of x. we see that
the coefficients of this expansion are
2 𝑎 𝑥𝜋𝑦
𝐾𝑚 (𝑦) = ∫ 𝑓(𝑥, 𝑦) sin 𝑑𝑥 (18)
𝑎 0 𝑏
So, the coefficients are
2 𝑎 𝑛𝜋𝑦
𝐵𝑚𝑛 = ∫ 𝐾𝑚 (𝑦)sin 𝑑𝑦 (19)
𝑏 0 𝑏
From (18) and (19) we obtain the generalized Euler Formula
4 𝑏 𝑎 𝑚𝜋𝑥 𝑛𝜋𝑦
𝐵𝑚𝑛 = ∫ ∫ 𝑓(𝑥, 𝑦) sin sin 𝑑𝑥𝑑𝑦 𝑚, 𝑥 = 1,2,3, … … .. (20)
𝑎𝑏 0 0 𝑎 𝑏
For the Fourier coefficient of on the double Fourier series (15).
The constant 𝐵𝑚𝑛 in (14) are now found in terms of 𝑓(𝑥, 𝑦) , To determine the 𝐵′𝑚𝑛
we differentiate (14) termwise w.r.t. t and use the condition (4); we obtain
∞ ∞
𝜕𝑢 𝑚𝜋𝑥 𝑛𝜋𝑦
| = ∑ ∑ 𝐵′𝑚𝑛 sin sin sin = 𝑔(𝑥, 𝑦) (21)
𝜕𝑡 𝑡=0 𝑎 𝑏
𝑚=1 𝑛=1

Assuming that can be expanded in the double Fourier series we find the coefficients as
𝑏 𝑎
4 𝑚𝜋𝑥 𝑛𝜋𝑦
𝐵′𝑚𝑛 = ∫ ∫ 𝑔(𝑥, 𝑦) sin sin 𝑑𝑥𝑑𝑦 𝑚, 𝑥 = 1,2,3, … … .. (22)
𝑎𝑏𝜆𝑚𝑛 0 0 𝑎 𝑏
Now, for (14) to satisfy the initial condition, the coefficients 𝐵𝑚𝑛 and 𝐵′𝑚𝑛 must be chosen
according to (21) and (22)
𝜕2𝑢
The vibration of a membrane are given by the wave equation = 𝑐 2 ∇2 𝑢 where
𝜕𝑡 2

2
𝜕2𝑢 𝜕2𝑢
∇ = 2+ 2
𝜕𝑥 𝜕𝑦
Example 2.73
Find the deflection of a square membrane if a=b=1 and c=1 if the initial velocity is zero and
the initial deflection 𝑓 (𝑥, 𝑦) = 𝐴 sin 𝜋𝑥 sin 2𝜋𝑦
Solution: Putting a=b=1 and 𝑓(𝑥, 𝑦) = 𝐴 sin 𝜋𝑥 sin 2𝜋𝑦 in equation (20) we get
1 1
𝐵𝑚𝑛 = 4𝐴 ∫ ∫ sin 𝜋𝑥 sin 2𝜋𝑦 sin 𝑚𝜋𝑥 sin 𝑛𝜋𝑦 𝑑𝑦𝑑𝑥
0 0
1 1
= 4𝐴 ∫ sin 𝜋𝑥 sin 𝑚𝜋𝑥 𝑑𝑥 (∫ sin 2𝜋𝑦 sin 𝑛𝜋𝑦 𝑑𝑦 )
0 0

= 0 𝑓𝑜𝑟 𝑚 ≠ 1

88
 1 1
= 4𝐴 ( ) ∫ sin 2𝜋𝑦 sin 𝑛𝜋𝑦 𝑑𝑦 𝑓𝑜𝑟 𝑚 = 1
2 0
1
1
∵ ∫ sin 𝜋𝑥 sin 𝜋𝑥 𝑑𝑥 =
0 2
1
⇒ 𝐵𝑚𝑛 = 2𝐴 ∫ sin 2𝜋𝑦 sin 𝑛𝜋𝑦 𝑑𝑦 = 0 𝑓𝑜𝑟 𝑛 ≠ 2
0

1
= 2𝐴 ( ) = 𝐴 𝑓𝑜𝑟 𝑛 = 2
2
∴ 𝐵12 = 𝐴 Also, from (12)

𝑚2 𝑛2
λ𝑚𝑛 = 𝑐𝜋√ 2 + 2
𝑎 𝑏

12 22
λ12 = 1. 𝜋√ + 2 = √5𝜋
12 1

𝑐 = 1, 𝑎 = 1, ℎ = 1
𝑚 = 1, 𝑛 = 2
From equation (13), the required solution is therefore,

𝑢(𝑥, 𝑦, 𝑡) = 𝐴 sin 𝜋𝑥 sin 2𝜋𝑦 cos(√5𝜋𝑡)

89