LECTURE NOTES

Engineering Analysis and Design

(Differential Equations and Applications)

Course Code: MC207

Jamkhongam Touthang
Assistant Professor
Department of Applied Mathematics
Delhi Technological University
Delhi-110042
 2

Syllabus: MC207 Differential Equations and Applications

Unit-I. Ordinary Differential Equations (ODE): Matrix method for homoge-

neous linear system with constant coefficients, Boundary value problems for
second order equations-Sturm Liouville’s problem, Eigen values and Eigen vec-
tors, Saddle points, Attractor, Repellant.

Unit-II. Partial Differential Equations (PDE): Formation, Solution of linear

PDE of first order (Lagrange’s method), Solution of first order non-linear PDE:
standard forms, Charpit’s method.

Unit-III. Linear PDE with constant coefficients-homogeneous and non-homogeneous,

Classification of second order PDE, Method of separation of variables, Cauchy’s
problem, Boundary value problems.

Unit-IV. Initial boundary value problems involving wave equation, heat equa-
tion and Laplace equation. Method of separation of variables and application
of Fourier transformations for Laplace, wave and heat equations.

Email: [email protected]
 UNIT-II
PARTIAL DIFFERENTIAL EQUATIONS
 4

Notations: Assume that z is a function of x and y.

∂z ∂z ∂2z ∂2z ∂2z
p= , q= , r= , s = , t=
∂x ∂y ∂x2 ∂x∂y ∂y 2
For n independent variables,
∂z ∂z ∂z ∂z
p1 = , p2 = , p3 = , .... . . . ..., pn = .
∂x1 ∂x2 ∂x2 ∂xn
By using suffixes,
∂u ∂u ∂2u ∂2u ∂2u
ux = , uy = , uxx = , uxy = , uyy =
∂x ∂y ∂x2 ∂y∂x ∂y 2
Formation of partial differential equations can be achieved by eliminating arbi-
trary constants or by eliminating arbitrary functions.
I. Elimination of arbitrary constants
1. Form a partial differential equation by eliminating the arbitrary constants α
and β.
(i) z = αx+βy+α2 +β 2 (ii) z = (x2 +α)(y 2 +β) (iii) z = (z−α)2 +(y−β)2

(i) We have

z = αx + βy + α2 + β 2 (1)
∂z ∂z
Then = α and = β.
∂x ∂y
Substituting these values in (1), we obtain
     2  2
∂z ∂z ∂z ∂z
z=x +y + +
∂x ∂y ∂x ∂y
which is the required partial differential equation.
(ii) We have

z = (x2 + α)(y 2 + β) (2)

Then
∂z 1 ∂z
= 2x(y 2 + β) or = (y 2 + β) (3)
∂x 2x ∂x
and
∂z 1 ∂z
= 2y(x2 + α) or = (x2 + α) (4)
∂y 2y ∂y
Using (3) and (4) in (2), we obtain
1 ∂z 1 ∂z
z= ×
2y ∂y 2x ∂x

Email: [email protected]
 5

Required partial differential equation is

  
∂z ∂z
4xyz =
∂x ∂y
2. Find the differential equation of the family of all right circular cones whose
axes coincide with z-axis.
General equation of the family of all right circular cones whose axes coincide
with z-axis, having semi-vertical angle α and vertex at (0, 0, c) is

x2 + y 2 = (z − c)2 tan2 α (5)

where c and α are arbitrary constants.

Differentiating (5) partially with respect to x and y, we get

∂z
2x = 2(z − c) tan2 α
∂x
∂z
or xy = y(z − c) tan2 α (6)
∂x
and
∂z
2y = 2(z − c) tan2 α
∂y
∂z
or xy = x(z − c) tan2 α (7)
∂y

From (6) and (7), we obtain

∂z ∂z
y(z − c) tan2 α = x(z − c) tan2 α
∂x ∂y
Hence, required partial differential equation is
   
∂z ∂z
y =x .
∂x ∂y

3. Find a partial differential equation by eliminating the arbitrary constants α,

β and γ from

x2 y2 z2
2
+ 2 + 2 = 1.
α β γ
We have
x2 y2 z2
2
+ 2 + 2 = 1. (8)
α β γ

Email: [email protected]
 6

Differentiating (8) partially with respect to x and y, we get respectively

2x 2z ∂z
+ 2 =0
α2 γ ∂x
∂z
or, γ 2 x + α2 z =0 (9)
∂x
and
2y 2x ∂z
+ 2 =0
β2 γ ∂y
∂z
or, γ 2 y + β 2 z =0 (10)
∂y
Differentiating (9) partially with respect to x and (10) partially with respect to
y, we obtain
 2
∂z ∂2z
γ 2 + α2 + α2 z 2 = 0 (11)
∂x ∂x
 2
2 2 ∂z ∂2z
and γ + β + β2z 2 = 0 (12)
∂y ∂y
From (9),
α2 z
 
2 ∂z
γ =− (13)
x ∂x
Substituting this value of γ 2 in (11) and dividing by α2 , we obtain
 2
z ∂z ∂z ∂2z
− + +z 2 =0
x ∂x ∂x ∂x
2
 2
∂ z ∂z ∂z
zx 2 + x −z =0 (14)
∂x ∂x ∂x
Similarly, from (10) and (12),
2
∂2z

∂z ∂z
zy 2 + y −z =0 (15)
∂y ∂y ∂y
Either (14) or (15) gives the required partial differential equation.
4. Form the partial differential equation by eliminating the arbitrary constants
from z = αeβt sin βx .
We have,

∂z ∂2z
= βαeβt cos βx, = −β 2 αeβt sin βx
∂x ∂x2
∂z ∂2z
= αβeβt sin βx, = αβ 2 eβt sin βx
∂t ∂t2
∂2z ∂2z
Clearly, + = 0.
∂x2 ∂t2

Email: [email protected]
 7

II. Elimination of arbitrary functions

1. Form a partial differential equation by eliminating f from

f (x + y + z, x2 + y 2 − z 2 ) = 0 . (16)

We write u = x + y + z and v = x2 + y 2 − z 2 . Then (16) reduces to

f (u, v) = 0 . (17)

Differentiating (17) partially with respect to x, we obtain

   
∂f ∂u ∂u ∂f ∂v ∂v
+p + +q =0 (18)
∂u ∂x ∂z ∂v ∂x ∂z

Here,
∂u ∂u ∂v ∂v

= 1, = 1, = 2x, = −2z 

∂x ∂z ∂x ∂z (19)
∂u ∂v
= 1, = 2y 

∂y ∂y
Using (19)in (18),

∂f ∂f
(1 + p) + 2(x − pz) =0
∂u  ∂v 
∂f
∂u 2(x − pz)
or   =− (20)
∂f p+1
∂v

Again, differentiating (17) partially with respect to y, we obtain

   
∂f ∂u ∂u ∂f ∂v ∂v
+q + +q =0 (21)
∂u ∂y ∂z ∂v ∂y ∂z

Using (19) in (21),

∂f ∂f
(1 + q) + 2(y − qz) =0
∂u  ∂v 
∂f
∂u 2(y − qz)
or   =− (22)
∂f q+1
∂v

Eliminating f from (20) and (22),

x − pz y − qz
=
p+1 q+1
which is the required partial differential equation.

Email: [email protected]
 8

5. Form a partial differential equation by eliminating f from

f (x2 + y 2 + z 2 , z 2 − 2xy) = 0 .

Solution: Left as an exercise.

6. Form a partial differential equation by eliminating f and g from

z = f (x2 − y) + g(x2 + y) . (23)

Differentiating (23) partially with respect to x and y respectively, we obtain

∂z
= 2xf 0 (x2 − y) + 2xg 0 (x2 + y) = 2x{f 0 (x2 − y) + g 0 (x2 + y)} (24)
∂x
∂z
= −f 0 (x2 − y) + g 0 (x2 + y) . (25)
∂y

Differentiating (24) and (25) partially with respect to x and y respectively, we

obtain
∂2z
= 2{f 0 (x2 − y) + g 0 (x2 + y)} + 4x2 {f 00 (x2 − y) + g 00 (x2 + y)} (26)
∂x2
∂2z
and = f 00 (x2 − y) + g 00 (x2 + y) . (27)
∂y 2

Using (24) and (27) in (26), we obtain the required partial differential equation

∂2z 2
 
1 ∂z 2∂ z
= 2 + 4x
∂x2 2x ∂x ∂y 2
2 2
∂ z ∂z ∂ z
or, x 2 = + 4x3 2 .
∂x ∂x ∂y
7. Form a partial differential equation by eliminating ψ from
y
(i) z = ψ(x2 + y 2 ) (ii) z = x2 ψ
x
 
1
(iii) z = x2 + 2ψ + logx (iv) x + y + z = ψ(x2 + y 2 + z 2 )
y
8. Form a partial differential equation by eliminating f and g from

z = f (x − αy) + g(x + αy) .

Solution: Left as an exercise.

Email: [email protected]
 9

Solution of a linear partial differential equation of first order

(Lagrange’s method)
General form of linear PDE of first order is

P p + Qq = R , (28)

where P , Q and R are functions of x, y and z. The general solution of (28) is

f (u, v) = 0 , (29)

where f is an arbitrary function, and u(x, y, z) = c1 and v(x, y, z) = c2 are two

independent solutions of
dx dy dz
= = . (30)
P Q R
Solution by Lagrange’s method
Step 1. Write the given PDE in the form (28).
Step 2. Obtain the auxiliary (subsidiary) equations of (28):

dx dy dz
= = .
P Q R

Step 3. Solve the auxiliary equations (in Step 2) to obtain two independent
solutions, say, u(x, y, z) = c1 and v(x, y, z) = c2 .
Step 4. The general solution (integral) of (28) is written in one of the three
equivalent forms

f (u, v) = 0, u = f (v), v = f (u) .

Problems
1. Solve:
y2 z
 
(i) xp + yq = z (ii) x2 p + y 2 q + z 2 = 0 (iii) p + xzq = y 2
x
(iv) y 2 p − xyq = x(z − 2y) (v) αp + βq = 1

(i) We have

xp + yq = z (31)

This is of the form P p + Qq = R where P = x, Q = y and R = z.

Auxiliary equations:
dx dy dz
= = (32)
x y z

Email: [email protected]
 10

dx dz
Integrating the first and third fractions of (32), namely, = , we obtain
x x z
logx = logz + logc1 . This gives log = logc1 . Therefore,
z
x
= c1 (33)
z
dy dz
Integrating the second and third fractions of (32), namely, = , we obtain
y z
logy = logz + logc2 . This gives
y
= c2 (34)
z
From (33) and (34), the required general solution is
x y x y
f , = 0 or, =f
z z z z
where f is an arbitrary function.
(iv) We have

y 2 p − xyq = x(z − 2y). (35)

This is of the form P p + Qq = R where P = y 2 , Q = −xy and R = x(z − 2y).

Auxiliary equations:
dx dy dz
= = (36)
y2 −xy x(z − 2y)
From the first two fractions of (36), we have

2xdx + 2ydy = 0.

Integrating, we obtain

x2 + y 2 = c1 (37)

From the second and third fractions of (36), we have

dz z − 2y
=−
dy y
dz 1
or, + z=2 (38)
dy y
R
(1/y)dy
which is a linear differential equation in z and y. Integrating factor (I.F.)=e =
elogy = y. Therefore, solution of (38) is
Z
zy = 2ydy + c2

or, zy − y 2 = c2 . (39)

Email: [email protected]
 11

Hence, from (37) and (39) the general solution of (35) is obtained as

f (x2 + y 2 , zy − y 2 ) = 0

where f is an arbitrary function.

Note. In Problems 1, one of the variables is either absent or cancels out from
any two fractions of the auxiliary equations and then one proceed to obtain the
integral.

2. Solve:
(i) py + qx = xyz 2 (x2 − y 2 ) (ii) (x2 − y 2 − z 2 )p + 2xyq = 2xz
(iii) z(xy + z 2 )(px − qy) = x4 (iv) xzp + yzq = xy
(i) We have

py + qx = xyz 2 (x2 − y 2 ) (40)

Comparing with P p + Qq = R, we have

P = y, Q = x and R = xyz 2 (x2 − y 2 ).

Auxiliary equations:
dx dy dz
= = (41)
y x xyz 2 (x2 − y 2 )
From the first and second fractions of (41), we have

2xdx − 2ydy = 0.

Integrating, we get

x2 − y 2 = c1 . (42)

Using (42), from the second and third fractions of (41), we have
1
c1 ydy − dz = 0. (43)
z2
Integrating both sides of (43) and using (42), we obtain
 2  
y 1
(x2 − y 2 ) + = c2 (44)
2 z
From (42) and (44), we obtain the general solution of (40) as
  2   
y 1
f x2 − y 2 , (x2 − y 2 ) + =0
2 z
 2  
y 1
or, (x2 − y 2 ) + = f (x2 − y 2 )
2 z

Email: [email protected]
 12

where f is an arbitrary function.

(ii) Ans. (x2 + y 2 + z 2 )/z = f (y/z), where f is an arbitrary function.
(iii) Ans. x4 − z 4 − 2xyz 2 = f (xy), where f is an arbitrary function.
(iv) We have

xzp + yzq = xy. (45)

Comparing this with P p + Qq = R, we have P = xz, Q = yz and R = xy.

Auxiliary equations:
dx dy dz
= = (46)
xz yz xy

From the first and second fractions of (46),

dx dy
=
x y
Integrating, we get

logx − logy = logc1

x
or, = c1 (47)
y

Using x = c1 y, from the second and third fractions of (46), we obtain

dy dz
=
yz c1 y 2
or, 2c1 ydy − 2zdz = 0

Integrating both sides, we get

c1 y 2 − z 2 = c2
or, xy − z 2 = c2 (since c1 = x/y). (48)

From (47) and (48), the required general solution is

f (x/y, xy − z 2 ) = 0
or, xy − z 2 = f (x/y)
or, x/y = f (xy − z 2 )

where f is an arbitrary function.

Note. In Problems 2, after computing one integral of the given PDE from the
auxiliary equations, this known integral is used to find the second integral.

Email: [email protected]
 13

The auxiliary equations corresponding to the PDE P p + Qq = R are

dx dy dz
= = .
P Q R
It is a well-known result from algebra that each of the fractions in the auxiliary
equations is equal to
Ldx + M dy + N dz
.
LP + M Q + N R
If LP +M Q+N R = 0, then we must have the numerator Ldx+M dy +N dz = 0
which on integration can give us the first integral u(x, y, z) = c1 . The other
integral, say v(x, y, z) = c2 can be obtained by repeating the method. Here, we
call L, M and N the multipliers. In the following we solve some PDEs using
this technique.
3. Solve:
(i) (x + 2z)p + (4xz − y)q = 2x2 + y (ii) x(y 2 + z)p − y(x2 + z)q = z(x2 − y 2 )
(iii) (y 2 + z 2 )p − xyq = −zx (iv) (y + zx)p − (x + yz)q = x2 + y 2
(v) (mz − ny)p + (nx − lz)q = ly − mx (vi) x(y − z)p + y(z − x)q = z(x − y)
(i) We have

(x + 2z)p + (4xz − y)q = 2x2 + y. (49)

Comparing with P p + Qq = R, we have

P = x + 2z, Q = 4xz − y and R = 2x2 + y.
Auxiliary equations:
dx dy dz
= = 2 . (50)
x + 2z 4xz − y 2x + y
Choosing 2x, −1, −1 as multipliers, each fraction of (50) is equal to
2xdx − dy − dz 2xdx − dy − dz
=
2x(x + 2z) − (4xz − y) − (2x2 + y) 0
It follows that 2xdx − dy − dz = 0. Integrating this, we obtain

x 2 − y − z = c1 . (51)

Choosing y, x, −2z as multipliers, each fraction of (50) is equal to

ydx + xdy − 2zdz d(xy) − 2zdz
=
y(x + 2z) + x(4xz − y) − 2z(2x2 + y) 0
It follows that d(xy) − 2zdz = 0. Integrating this, we obtain

xy − z 2 = c2 . (52)

Email: [email protected]
 14

From (51) and (52), the required solution is

f (x2 − y − z, xy − z 2 ) = 0
or, x2 − y − z = f (xy − z 2 )
or, xy − z 2 = f (x2 − y − z) .

where f is an arbitrary function.

(ii) Ans. x2 + y 2 − 2z = f (xyz), where f is an arbitrary function.
(iii) Ans. x2 + y 2 + z 2 = f (y/z), where f is an arbitrary function.
(iv) Ans. x2 + y 2 − z 2 = f (xy + z), where f is an arbitrary function.
(v) Ans. f (x2 + y 2 + z 2 , lx + my + nz) = 0, where f is an arbitrary function.
(vi) We have

x(y − z)p + y(z − x)q = z(x − y). (53)

Comparing with P p + Qq = R, we have

P = x(y − z), Q = y(z − x) and R = z(x − y).
Auxiliary equations:
dx dy dz
= = . (54)
x(y − z) y(z − x) z(x − y)

Choosing 1, 1, 1 as multipliers, each fraction of (54) is equal to

dx + dy + dz dx + dy + dz
=
x(y − z) + y(z − x) + z(x − y) 0

It follows that dx + dy + dz = 0. Integrating this, we obtain

x + y + z = c1 . (55)

Choosing yz, zx, xy as multipliers, each fraction of (54) is equal to

yzdx + xzdy + xydz d(xyz)

=
xyz(y − z) + xyz(z − x) + xyz(x − y) 0

It follows that d(xyz) = 0. Integrating this, we obtain

xyz = c2 . (56)

From (55) and (56), the required general solution is

f (x + y + z, xyz) = 0
or, x + y + z = f (xyz) or, xyz = f (x + y + z)

Email: [email protected]
 15

where f is an arbitrary function.

4. Solve:
(i) y 2 (x−y)p−x2 (x−y)q = z(x2 +y 2 ) (ii) (xz+y 2 )p+(yz−2x2 )q+2xy+z 2 = 0

(iii) (x3 +3xy 2 )p+(y 3 +3x2 y)q = 2z(x2 +y 2 ) (iv) (x2 −y 2 −z 2 )p+2xyq = 2xz

(v) xyp + y(2x − y)q = 2xz (vi) (y + z)p + (z + x)q = x + y

(i) We have

y 2 (x − y)p − x2 (x − y)q = z(x2 + y 2 ). (57)

Comparing with P p + Qq = R, we have

P = y 2 (x − y), Q = −x2 (x − y) and R = z(x2 + y 2 ).
Auxiliary equations:
dx dy dz
= = . (58)
y 2 (x − y) −x2 (x − y) z(x2 + y 2 )
Choosing 1, −1, 0 as multipliers, each fraction of (58) is equal to
dx − dy dx − dy
= . (59)
y 2 (x − y) + x2 (x − y) (x − y)(x2 + y 2 )
Combining (59) and the third fraction of (58),
d(x − y) dz
− = 0.
x−y z
Integrating, we get

log(x − y) − logz = logc1

(x − y)
or, = c1 (60)
z
From the first and second fractions of (58), we have

x2 dx = −y 2 dy
or, 3x2 dx + 3y 2 dy = 0 .

Integrating, we obtain

x3 + y 3 = c2 . (61)

From (60) and (61), the required general solution is

x−y
= f (x3 + y 3 ),
z

Email: [email protected]
 16

where f is an arbitrary constant.

(ii) Ans. f (2xz − y 2 , yz + x2 ) = 0, f is an arbitrary function.
1 1  xy 
(iii) Ans. − = f , f is an arbitrary function.
(x − y)2 (x + y)2 z2
x2 + y 2 + z 2
(iv) Ans. = f (y/z), f is an arbitrary function.
z
(v) Ans. xy − x2 = f (z/xy), f is an arbitrary function.
(vi) We have

(y + z)p + (z + x)q = x + y. (62)

Comparing with P p + Qq = R, we have

P = (y + z), Q = (z + x) and R = x + y.
Auxiliary equations:
dx dy dz
= = . (63)
y+z z+x x+y
Choosing 1, 1, 1 as multipliers, each fraction of (63) is equal to
dx + dy + dz d(x + y + z)
= . (64)
(y + z) + (z + x) + (x + y) 2(x + y + z)
Choosing 1, −1, 0 as multipliers, each fraction of (63) is equal to
dx − dy d(x − y)
= . (65)
(y + z) − (z + x) −(x − y)
Choosing 0, 1, −1 as multipliers, each fraction of (63) is equal to
dy − dz d(y − z)
= . (66)
(z + x) − (x + y) −(y − z)
From (64), (65) and (66),
d(x + y + z) d(x − y) d(y − z)
= = (67)
2(x + y + z) −(x − y) −(y − z)
From the first two fractions of (67), we have
d(x + y + z) d(x − y)
+2 =0
(x + y + z) (x − y)
Integrating, we get

log(x + y + z) + 2log(x − y) = logc1

or, (x − y)2 (x + y + z) = c1 . (68)

Email: [email protected]
 17

From the last two fractions of (67), we have

d(x − y) d(y − z)
=
(x − y) (y − z)
Integrating, we get

log(x − y) = log(y − z) + logc2

 
x−y
or, = c2 . (69)
y−z

From (68) and (69), the required general solution is

 
x−y
(x − y)2 (x + y + z) = f ,
y−z

where f is an arbitrary constant.

Solution of first order non-linear PDE

A solution z = z(x, y) of a PDE f (x, y, z, p, q) = 0 when interpreted as a surface
in 3D space will be called an integral surface.
Types of solutions
(i) Complete integral
(ii) General integral
(iii) Singular integral

Complete integral. An expression of the type

φ(x, y, z, a, b) = 0 (70)

which is a solution of a PDE f (x, y, z, p, q) = 0, where a and b are arbitrary

constants, is said to be a complete integral of the first order PDE

f (x, y, z, p, q) = 0

The expression

φ(x, y, z, a, b) = z − (ax + by + a2 + b2 ) = 0 (71)

has two arbitrary constants a and b. And it is also a solution of the PDE

f (x, y, z, p, q) = z − px − qy − p2 − q 2 = 0. (72)

So, (71) is a complete integral of (72).

Particular integral of a PDE f (x, y, z, p, q) = 0 is obtained by assigning partic-
ular values to the arbitrary constants a and b in the complete integral.

Email: [email protected]
 18

General integral. Let the complete integral of f (x, y, z, p, q) = 0 be

φ(x, y, z, a, b) = 0. (73)

The general integral is obtained by eliminating a between (73) and the equations

b = ψ(a)

and
∂φ ∂φ
+ ψ 0 (a) = 0,
∂a ∂b
where ψ is an arbitrary function.

Singular integral. The singular integral is obtained by eliminating a and b

between φ(x, y, z, a, b) and the equations
∂φ ∂φ
=0 and = 0.
∂a ∂b
Charpit’s Method.
Step 1. Write the given equation in the form f (x, y, z, p, q) = 0.
Step 2. Charpit’s auxiliary equations
dp dq dz dx dy
= = = =
∂f ∂f ∂f ∂f ∂f ∂f ∂f ∂f
+p +q −p −q − −
∂x ∂z ∂y ∂z ∂p ∂q ∂p ∂q
∂f ∂f ∂f
Find the values of , , , etc and substitute these values in the auxiliary
∂x ∂y ∂z
equations. Simplify.
Step 3. Choose two fractions in Step 2 to obtain an integral involving at least
one of p and q (get the simplest relation).
Step 4. Solve the relation (obtained in Step 3) along with the given equation in
Step 1 to determine p and q. Substitute these values of p and q in dz = pdx+qdy
which on integration yields the complete integral of the given equation.
1. Find a complete integral of px + qy = pq.
We have,

f (x, y, z, p, q) = px + qy − pq = 0. (74)

Charpit’s auxiliary equations

dp dq dz dx dy
= = = =
∂f ∂f ∂f ∂f ∂f ∂f ∂f ∂f
+p +q −p −q − −
∂x ∂z ∂y ∂z ∂p ∂q ∂p ∂q

Email: [email protected]
 19

or
dp dq dz dx dy
= = = = (75)
p+p×0 q+q×0 −p(x − q) − q(y − p) −(x − q) −(y − q)
dp dq
Taking the first two fractions of (75), = , which on integration gives
p q

log p = log q + log α

or, p = αq (76)

Substituting this value of p in (74), we get

αqx + qy − αq 2 = 0
αq = αx + y as q 6= 0. (77)
αx + y
From (76) and (77), q = and p = αx + y.
α
Substituting these values of p and q in dz = pdx + qdy, we obtain
αx + y
dz = (αx + y)dx + dy
α
or, αdz = (αx + y)(αdx + dy)
or, αdz = (αx + y)d(αx + y)

(αx + y)2
Integrating, αz = + β, which is the required complete integral, where
2
α and β are arbitrary constants.
2. Find complete integral of z 2 (p2 z 2 + q 2 ) = 1.
3. Find a complete, singular and general integrals of (p2 + q 2 )y = qz.
We have,

f (x, y, z, p, q) = (p2 + q 2 )y − qz = 0. (78)

Charpit’s auxiliary equations

dp dq dz dx dy
= = = =
∂f ∂f ∂f ∂f ∂f ∂f ∂f ∂f
+p +q −p −q − −
∂x ∂z ∂y ∂z ∂p ∂q ∂p ∂q
or
dp dq dz dx dy
= 2 = 2 2
= = (79)
−pq p −2p y + qz − 2q y −2py −2qy + z
Taking the first two fractions of (79), pdp + qdq = 0, which on integration gives

p2 + q 2 = α 2 (80)

Email: [email protected]
 20

α2 y
Substituting this value in (78), we get α2 y = qz or q = . Putting this value
z
of q in (80),
s
α4 y 2

p αp 2
p= (α2 − q2 ) = α2 − 2 = (z − α2 y 2 ).
z z

Substituting these values of p and q in dz = pdx + qdy, we obtain

αp 2 α2 y
dz = (z − α2 y 2 )dx + dy
z z
zdz − α2 ydy
or, p = αdx.
(z 2 − α2 y 2 )
Integrating, we obtain,
p
(z 2 − α2 y 2 ) = αx + β
or, z 2 − α2 y 2 = (αx + β)2 . (81)
which is the required complete integral, where α and β are arbitrary constants.

Singular integral. Differentiating (81) partially with respect to α and β, we

have
0 = 2αy 2 + 2(αx + β)x (82)
0 = 2(αx + β) (83)
Eliminating α and β between (81), (82) and (83), we get z = 0 which clearly
satisfies (78)and hence it is the required singular integral.
General integral. Replacing β by ψ(α) in (81), we get
z 2 − α2 y 2 = {αx + ψ(α)}2 . (84)
Differentiating (84) partially with respect to α, we get
−2αy 2 = 2{αx + ψ(α)}.{x + ψ 0 (α)}. (85)
General integral is obtained by eliminating α from (84) and (85).
4. Find a complete integral of p2 + q 2 − 2px − 2qy + 1 = 0.
We have,
f (x, y, z, p, q) = p2 + q 2 − 2px − 2qy + 1 = 0. (86)
Charpit’s auxiliary equations
dp dq dz dx dy
= = = =
∂f ∂f ∂f ∂f ∂f ∂f ∂f ∂f
+p +q −p −q − −
∂x ∂z ∂y ∂z ∂p ∂q ∂p ∂q

Email: [email protected]
 21

or,
dp dq dz dx dy
= = = = (87)
−2p −2q −p(2p − 2x) − q(2q − 2y) −(2p − 2x) −(2q − 2y)
dp dq
From the first two fractions of (87), = , which on integration gives
p q
log p = log q + log α
or, p = αq (88)
Substituting this value of p in (86), we get
α2 q 2 + q 2 − 2αqx − 2qy + 1 = 0
or, (α2 + 1)q 2 − 2(αx + y)q + 1 = 0

Therefore,
p
4(αx + y)2 − 4(α2 + 1)
2(αx + y) ±
q=
2(α2 + 1)
p
2(αx + y) ± 4(αx + y)2 − 4(α2 + 1)
and p = αq = α .
2(α2 + 1)
Substituting these values of p and q in dz = pdx + qdy, we get
p
(αx + y) ± (αx + y)2 − (α2 + 1)
dz = (αdx + dy). (89)
(α2 + 1)
Putting αx + y = u (so that αdx + dy = du) in (89), we obtain
h p i
(α2 + 1)dz = u ± u2 − (α2 + 1) du.

Integrating, we obtain the complete integral

u2 (α2 + 1)
 
up 2  p
(α2 +1)z = ± u − (α2 + 1) − log u + u2 − (α2 + 1) +β,
2 2 2
where u = αx + y and α, β are arbitrary constants.
5. Find a complete integral of px + qy = z(1 + pq)1/2 .
We have,
f (x, y, z, p, q) = px + qy − z(1 + pq)1/2 = 0. (90)
Charpit’s auxiliary equations
dp dq dz dx dy
= = = =
∂f ∂f ∂f ∂f ∂f ∂f ∂f ∂f
+p +q −p −q − −
∂x ∂z ∂y ∂z ∂p ∂q ∂p ∂q

Email: [email protected]
 22

From (90),
∂f ∂f ∂f
= p, = q, = −(1 + pq)1/2
∂x ∂y ∂z
∂f zq ∂f zp
=x− 1/2
, =y−
∂p 2(1 + pq) ∂q 2(1 + pq)1/2
Substituting these values in the auxiliary equations, we obtain
dp dq
1/2
=
p − p(1 + pq) q − q(1 + pq)1/2
dz
=    
zq zp
−p x − − q y −
(1 + pq)1/2 (1 + pq)1/2
dx dy
=  =   (91)
zq zp
− x− − y −
2(1 + pq)1/2 2(1 + pq)1/2
dp dq
From the first two fractions of (91), = , which on integration gives
p q
log p = log q + log α
or, p = αq
Substituting the value of p in (90), we get
q(αx + y) = z(1 + αq 2 )1/2
or, q 2 [(αx + y)2 − αz 2 ] = z 2 .
Therefore,
z
q=p
(αx + y)2 − αz 2
αz
and p = αq = p .
(αx + y)2 − αz 2
Now, substituting the values of p and q in dz = pdx + qdy, we have
z
dz = p (αdx + dy)
(αx + y)2 − αz 2
dz 1
or, =p (αdx + dy).
z (αx + y)2 − αz 2
√ √
Put αx + y = αv so that αdx + dy = αdv. Therefore
√
dz αdv
=√
z αv 2 − αz 2
√ r 
dv v2 − z2 v 2
or, = = − 1,
dz z z

Email: [email protected]
 23

v
which is a homogeneous linear differential equation. To solve this, we let = w
z
dv dw
or v = wz so that =w+z . Therefore
dz dz
dw p 2
w+z = w −1
dz
dz dw
or, =√
z w2 − 1 − w
dz p 
or, =− w2 − 1 + w dw.
z
On integration, we obtain the complete integral
 p
w 1 p  w 2
logz = − w2 − 1 − log w2 − 1 + w − + β,
2 2 2
v 1
where w = = √ (αx + y).
z αz

6. Find a complete integral of z = px + qy + p2 + q 2 .

Ans. z = αx + βy + α2 + β 2 , where α and β are arbitrary constants.
7. Find a complete integral of p2 x + q 2 y = z.
p √ √
Ans. z(1 + α) = αx + y + β.
8. Find a complete integral of (i) pq = px + qy (ii) (p2 + q 2 )x = pz
√
Ans. (i) z = xy + y x2 − α + β (ii) z 2 = α2 x2 + (αy + β)2

Standard forms
Standard form I. PDE involving only p and q.
PDE is of the form f (p, q) = 0. Solution is given by
z = αx + βy + c (92)
where
f (α, β) = 0 (93)
and γ is an arbitrary constant. Suppose we get β = ψ(α) by solving (93) for β.
Then the complete integral of the given PDE f (p, q) = 0 is
z = αx + ψ(α)y + γ (94)
where α and γ are arbitrary constants.
Differentiating (94) partially with respect to α and γ, we obtain
0 = x + ψ 0 (α)y and 0 = 1. (95)

Email: [email protected]
 24

The singular integral of the given PDE f (p, q) = 0 is obtained by eliminating

the arbitrary constants α and γ between the complete integral (94) and the
equations (95). But the second equation in (95) is absurd. Hence, PDE in
standard form I has no singular solution.
To find the general integral of f (p, q) = 0, we take γ = ψ(a) in (94) where ψ
is an arbitrary constant,

z = αx + ψ(α) + φ(α). (96)

Differentiating (96) partially with respect to α, we get

0 = x + ψ 0 (α) + φ0 (α) . (97)

Now, the general integral of the given PDE is obtained by eliminating α between
(96) and (97).
1. Solve:
(i) pq = γ, where γ is a constant
(ii) p2 + q 2 = 1
(iii) p2 + q 2 = γpq, where γ is a constant
√ √
(iv) p + q = 1
(i) We have

pq = γ (98)

Clearly, (98) is of the form f (p, q) = 0 and thus, its solution is

z = αx + βy + δ (99)

where αβ = γ or β = γ/α, obtained by taking p = α and q = β in (98).

Substituting β = γ/α in (99), the complete integral is
γy
z = αx + +δ (100)
α
where α and δ are arbitrary constants. Differentiating (100) partially with
respect to α and δ, we get
γy
0=x− and 0 = 1.
α2
But 0 = 1 is absurd. Hence, there is no singular solution of (98).
Taking δ = ψ(α) in (100), we get
γy
z = αx + + ψ(α). (101)
α

Email: [email protected]
 25

Differentiating (101) partially with respect to α, we get

γy
0=x− + ψ 0 (α) (102)
α2
The general solution is obtained by eliminating a from (101) and (102).
(iii) We have
p2 + q 2 = γpq (103)
which is of the form f (p, q) = 0. Solution is of the form
z = αx + βy + δ (104)
where α2 + β 2 = γαβ, obtained by taking p = α and q = β in (103). Solving
α p
for β, we get β = (γ ± γ 2 − 4).
2
Substituting the value of β in (104), the complete integral is
α p
z = αx + (γ ± γ 2 − 4)y + δ (105)
2
where α and δ are arbitrary constants.
Differentiating (105) partially with respect to α and δ, we obtain
1 p
0 = x + (γ ± γ 2 − 4)y and 0 = 1.
2
But 0 = 1 is absurd. Hence, there is no singular solution.
To find the general integral, put δ = ψ(α) in (105). Then
α p
z = αx + (γ ± γ 2 − 4)y + ψ(α). (106)
2
Differentiating (106) partially with respect to α, we obtain
1 p
0 = x + (γ ± γ 2 − 4)y + ψ 0 (α). (107)
2
The general solution is obtained by eliminating α between (106) and (107).
Equations reducible to standard form I
2. Solve:
(i) x2 p2 + y 2 q 2 = z (ii) pq = xm y n z l
(iii) (y − x)(qy − px) = (p − q)2 (iv) (x + y)(p + q)2 + (x − y)(p − q)2 = 1
(i) We have
x2 p 2 + y 2 q 2 = z
 2  2
x ∂z y ∂z
or, √ × + √ × =1 (108)
z ∂x z ∂y

Email: [email protected]
 26

1 1 1
Put dx = dX, dy = dY and √ dz = dZ so that log x = X, log y = Y and
√ x y z
2 z = Z.
Then (108) becomes
 2  2
∂Z ∂Z
+ =1
∂X ∂Y
or, P 2 + Q2 = 1 (109)
∂Z ∂Z
where P ≡ and Q ≡ .
∂X ∂Y
Now, (109) is of the form f (P, Q) = 0. Therefore, solution of (109) is
Z = αX + βY + γ
√
where α2 + β 2 = 1 or β = 1 − α2 which is obtained by replacing P by α and
Q by β in (109).
The complete integral is
p
Z = αX + 1 − α2 Y + γ
√ p
or, 2 z = α log x + 1 − α2 log y + γ (substituting the values of X, Y, Z)
√ √
2
or, 2 z = log xα + log y 1−α − log K, taking γ = log K
√ !
2
√ xα y 1−α
or, 2 z = log
K
√ √
1−α2
or, xα y = Ke2 z

where α and K are arbitrary constants.

Standard form II. Clairaut’s form. z=px+qy+f(p,q)
Suppose the given equation is of the form
z = px + qy + f (p, q). (110)
Using Charpit’s method, we can show that the complete integral is
z = αx + βy + f (α, β) (111)
where α and β are arbitrary functions. (Note that the complete integral in this
form is obtained by simply replacing p by α and q by β).
Differentiating (111) partially with respect to α and β, we get
∂f
x+=0 (112)
∂α
∂f
and y + = 0. (113)
∂β

Email: [email protected]
 27

Eliminating α and β from (111), (112) and (113), the singular integral of (110)
is obtained.
Taking β = ψ(α), (111) becomes

z = αx + ψ(α)y + f (α, ψ(α)). (114)

Differentiating (114) partially with respect to α, we get

0 = x + ψ 0 (α)y + f 0 (α, ψ(α)). (115)

Eliminating α between (114) and (115), we get the general integral of (110).
1. Solve z = px + qy + p2 − q 2 .
We have

z = px + qy + p2 − q 2

which is in Clauraut’s form z = px + qy + f (p, q). Therefore, complete solution

is given by

z = αx + βy + α2 − β 2 . (116)

To find singular integral, differentiating (116) with respect to α and β,

0 = x + 2α (117)
0 = y − 2β (118)
x y
∴ α=− and β = (119)
2 2
Substituting the values of α and β in (116),

x2 y2 x2 − y 2
z=− + +
2 2 4
∴ 4z = y 2 − x2

which is the required singular integral.

Putting β = ψ(α) in (116),

z = αx + ψ(α)y + α2 − [ψ(α)]2 (120)

Differentiating (120) with respect to α,

0 = x + ψ 0 (α)y + 2α − 2ψ(α)ψ 0 (α). (121)

The general solution is now obtained by eliminating α between (120) and (121).

Email: [email protected]
 28

p
2. Solve z = px + qy + 1 + p2 + q 2 .
We have
p
z = px + qy + 1 + p2 + q 2

which is in Clauraut’s form z = px + qy + f (p, q). Therefore, complete solution

is given by
p
z = αx + βy + 1 + α2 + β 2 . (122)

To find singular integral, differentiating (122) with respect to α and β,

α
0=x+ p
1 + α2 + β 2
β
and 0 = y + p
1 + α2 + β 2
α
∴ x = −p (123)
1 + α2 + β 2
β
and y = − p (124)
1 + α2 + β 2
From (123) and (124),

α2 + β 2
x2 + y 2 = p
1 + α2 + β 2
α2 + β 2 1
1 − x2 − y 2 = 1 − p =
1 + α2 + β 2 1 + α2 + β 2
p 1
∴ 1 + α2 + β 2 = p .
1 − x2 − y 2
Using these relations in (123) and (124), we get
p p
x = −α 1 − x2 − y 2 and y = −β 1 − x2 − y 2

or,
x y
α = −p and β = − p
1− x2 − y2 1 − x2 − y 2
Substituting these values of α and β in (124),

x2 y2 1
z = −p −p +p
1 − x2 − y 2 1 − x2 − y 2 1 − x2 − y 2
1 − x2 − y 2 p
=p = 1 − x2 − y 2
2
1−x −y 2

∴ z 2 = 1 − x2 − y 2

Email: [email protected]
 29

Thus, x2 + y 2 + z 2 = 1 is the singular solution. Find the general solution as

usual.
3. Solve z = px + qy + p2 q 2 .
We have

z = px + qy + p2 q 2

which is in Clairaut’s form z = px + qy + f (p, q).

The complete integral is

z = αx + βy + α2 β 2 (125)

Differentiating (125) partially with respect to α and β, we get respectively

0 = x + 2αβ 2 and 0 = y + 2α2 β.

Therefore,

x = −2αβ 2 (126)
2
and y = −2α β (127)

It follows that
x y 1
= = −2αβ = (say)
β α k
⇒ α = ky and β = kx

1
From (126), x = −2k 3 yx2 or k 3 = − .
2xy
Substituting the values of α and β in (125),
 
1
z = kxy + kxy + k 4 x2 y 2 = 2kxy + kx2 y 2 −
2xy
k 3
= 2kxy − xy = kxy
2
 2
3 27 3 3 1 27
z = x y − = − x2 y 2
8 2xy 16

Therefore, the singular solution is 16z 3 + 27x2 y 2 = 0.

Putting β = ψ(α) in (125),

z = αx + ψ(α)y + α2 [ψ(α)]2 (128)

Differentiating (128) partially with respect to α,

0 = x + ψ 0 (α)y + 2[ψ(α) + αψ 0 (α)] (129)

Email: [email protected]
 30

The general solution is obtained by eliminating α between (128) and (129).

√
4. Solve (i) z = px + qy + log(pq) (ii) z = px + qy − 2 pq.
Ans. (i) Complete integral is z = αx + βy + log(αβ), Singular integral is
z = −2 − log (xy)
√
(ii) Complete integral is z = αx + βy − 2 αβ, Singular integral is (x − y)(y −
z) = 1.
Standard form III. PDE involving only p, q and z. It is of the form

f (p, q, z) = 0 . (130)

To solve such PDEs we apply the following steps:

Step 1. Let u = x + αy.
dz dz
Step 2. Replace p by and replace q by α in (??). Solve the resulting
du du
first order ODE by known methods.
Step 3. Replace u by x + αy in the solution obtained in Step 2.
Remarks. Singular and generals are obtained by usual methods.
1. Find a complete integral of p3 = qz.
We have

p3 = qz (131)

which is of the form f (p, q, z) = 0. Let u = x + αy, where α is an arbitrary

dz dz
constant. Replacing p by and q by α in (131), we get
du du
 2
dz
z=α
du
√
dz z
or = ±√
du α
√
or ± αz −1/2 dz = du.
√
Integrating, we get ±2 αz = u + β or 4(αz) = (x + αy + β)2 , since u = x + αy,
α and β are arbtrary constants.
2. Find a complete integral of p3 + q 3 − 3pqz = 0.
Ans. (1 + α3 )log z = 3α(x + αy) + β, α and β are arbitrary constants.
3. Find a complete integral of p(1 + q 2 ) = q(z − k).
Ans. (x + αy + β)2 = 4(α(z − k) − 1)2 , α and β are arbitrary constants.
4. Find a complete integral of 4(1 + z 3 ) = 9z 4 pq.
Ans. Complete integral is α(1 + z 3 ) = (x + αy + β)2 , α and β are arbitrary

Email: [email protected]
 31

constants. Singular integral is 1 + z 3 = 0.

5. Find a complete integral of z 2 (p2 + q 2 + 1) = n2 .
Ans. (1 + α2 )(n2 − z 2 ) = (x + αy + β)2 , α and β are arbitrary constants.
6. Find the complete and singular integrals of q 2 = z 2 p2 (1 − p2 ).
We have

q 2 = z 2 p2 (1 − p2 ) (132)

which is of the form f (p, q, z) = 0. Let u = x + αy, α is an arbitrary constant.

dz dz
Replacing p by and q by α in (132), we get
du du
 2  2 "  2 #
dz 2 dz dz
α =z 1−
du du du
"  2 #
dz
or, α2 = z 2 1 −
du
 2
dz z 2 − a2
or, =
du z2
zdz
or, du = ± 2 .
(z − a2 )1/2

Integrating, we get
p
± z 2 − α2 = u + β
or (z 2 − α2 ) = (x + αy + β)2 , since u = x + αy, (133)

where α and β are arbtrary constants.

To find the singular integral, differentiating (133) partially with respect to α
and β, we obtain respectively

−2α = 2y(x + αy + β) (134)

and 0 = 2(x + αy + β). (135)

From (134) and (135), x + αy + β = 0 and α = 0. Using these values, (133) gives
z = 0 which is free from α and β. It follows that p = q = 0. Also, z = p = q = 0
satisfies the given equation and hence the required singular solution is z = 0.
Standard form IV. PDE of the form

f (x, p) = g(y, q) (136)

where z is absent, the terms containing p and x appear on one side and the
terms containing q and y appear on another side. A PDE which can be written

Email: [email protected]
 32

in this form is also said to be separable.

Equating each of the expressions in (136) to an arbitrary constant α (say)

f (x, p) = g(y, q) = α. (137)

Solving for p and q, we get

p = f1 (x, α) and q = g1 (y, α). (138)

We know that dz = pdx + qdy. It follows that

dz = f1 (x, α)dx + g1 (y, α)dy

Z Z
∴ z = f1 (x, α)dx + g1 (y, α)dy + β.

The last equation contains two arbitrary constants and hence it is the complete
integral of the given PDE. The singular and general integrals are obtained as
usual.

1. Solve p2 + q 2 = x + y.
We have

p2 + q 2 = x + y (139)

which is clearly separable as

p2 − x = y − q 2 .

Equating each expression to an arbitrary constant α,

p2 − x = α and y − q 2 = α. (140)

Solving (140) for p and q,

√ √
p= x+α and q = y − α. (141)

Now

dz = pdx + qdy
√ √
= x + αdx + y − αdy
2 2
∴ z = (x + α)3/2 + (y − α)3/2 + β (142)
3 3
which is the complete integral of the given equation. Differentiating (142) par-
tially with respect to β, we get 0 = 1 which is absurd. Hence, there is no
singular solution. (You may find the general integral as usual).

Email: [email protected]
 33

2. Solve py + qx + pq = 0.
We have

py + qx + pq = 0 (143)

which is clearly separable as

p q
=− .
p+x y
Equating each expression to an arbitrary constant α,
p q
=α and − = α.
p+x y
Solving for p and q
αx
p= = and q = −αy. (144)
1−α
Substituting these values of p and q in dz = pdx + qdy, we get
αx
dz = dx − αydy
1−α
On integration, we get

αx2 αy 2 β
z= − +
2(1 − α) 2 2
2
αx
or, 2z = − αy 2 + β
1−α
which is the required complete integral consisting of two arbitrary constants α
and β.(The singular and general integrals are obtained as usual).
3. Solve p2 y(1 + x2 ) = qx2 .
We have

p2 y(1 + x2 ) = qx2 . (145)

This equation can be rewritten as

1 + x2
 
2 q
p = .
x2 y

Equating each side to an arbitrary constant α,

1 + x2
 
q
p2 = α and = α.
x2 y

Email: [email protected]
 34

Solving for p and q, we get

√
x α
p= √ and q = αy.
1 + x2
Now
dz = pdx + qdy
√
x α
=√ dx + αydy.
1 + x2
Taking integration,
√
Z Z
x
z= α √ dx + α ydy
1 + x2
p αy 2
= α(1 + x2 ) + +β
2
This is the complete integral of the given PDE consisting of two arbitrary con-
stants α and β.
4. Solve p − x2 = q + y 2 .
x3 y3
Ans. z = αx + + αy − + β, where α and β are arbitrary constants.
3 3
5. Solve p2 + q 2 = x2 + y 2 .
We have
p2 + q 2 = x2 + y 2 or p2 − x2 = y 2 − q 2 .
Let p2 − x2 = y 2 − q 2 = α2 , an arbitrary constant.
Solving for p and q
p p
p= α2 + x2 and q = y 2 − α2 .
Now
dz = pdx + qdy
p p
= α2 + x2 dx + y 2 − α2 dy
Integrating,
Z p Z p
z= α2 + x2 dx + y 2 − α2 dy

Therefore
α2 x xp 2 yp 2 α2 y
z= sinh−1 + α + x2 + y − α2 − cosh−1 + β.
2 α 2 2 2 α
This is the complete integral consisting of two arbitrary constants α and β.
There is no singular integral. Obtain the general integral as usual.

Email: [email protected]