9 LINEAR DIFFERENTIAL EQUATIONS OF

SECOND AND HIGHER ORDER

aaaaa

9.1 INTRODUCTION
dy d2 y
A differential equation in which the dependent variable, y(x) and its derivatives, say, ,
dx dx2
etc. occur in the first degree and are not multiplied together is called linear differential
equation. Then, we classify them as linear differential equation with constant co-efficients
and the other with variable coefficients.
The linear differential equations with constant coefficients generally arises in practical problems
related to the study of mechanical, acoustical and electrical vibrations, whereas linear
differential equations with variable coefficients arise generally in mathematical modeling of
physical problems. Some of the important linear differential equations with variable coefficients
are Bessel equation, Legendre’s equation, Chebyshev equation etc.
The solution of linear differential equations with constant coefficients are generally found
in terms of known standard functions while there exists no such procedure in case of
differential equations with variable coefficients and their solutions many a times results in
the form of an infinite series.
The general form of the nth order linear differential with constant coefficients is
dn y dn −1y dy
k0 n
+ k1 n −1
+ … + kn −1 + kn y = X ( x ) …(1)
dx dx dx
where k0, k1, k2,…, kn are constants and X is a function of ‘x’ only.
The general linear differential equation with variable coefficients is written as
dn y dn −1y dn − 2 y dy
P0 n + P1 n −1 + P2 n − 2 + … + Pn −1 + Pn y = X ( x) … (2)
dx dx dx dx
where P0 (≠ 0), P1, P2,…, Pn and X are function of ‘x’ only.
If X(x) = 0 in (1) and (2), then they are called linear homogeneous differential equations
with constant coefficients and variable coefficients respectively.

9.2 SOLUTION OF LINEAR DIFFERENTIAL EQUATIONS

In equation (1), x varies on some interval of definition, say I, which may be open, semi open,
closed or infinite and the differential equation may be valid for all x ∈ (0, ∞), (–∞, 0), (–∞, ∞).
If y1(x) is a solution of the equation (1), then it must satisfy the equation identically and
577
578 Engineering Mathematics through Applications

whence y1 (x) must be continuously differentiable (n – 1) times and dn

y1(x) must be
dxn
continuous in that interval.
Further, if coefficients P0(x), P1(x) ,…….., Pn(x), P0(x) ≠ 0, in the linear homogeneous
equation (2) are continuous on some interval of def, say I, then this equation has n linearly
independent solutions. If y1(x), y2(x) ,…, yn(x) are n linearly independent solutions, then the
general solution is their combination.
i.e. y(x) = c1y1 + c2y2 + … + cnyn … (3)

Theorem: If y1(x), y2(x) ,…, yn(x) be n linearly independent solutions of

dn y dn −1y dy
k0 n
+ k1 n −1
+ … + kn −1 + kn y = 0,
dx dx dx
where k0, k1 ,…, kn are all constants, then y = c1y1 + c2y2 + … + cnyn is also a solution of (1). This
is called the Principle of Superposition or Principle of linearity.
Proof: Putting y = c1y1 + c2y2 + … + cnyn into left hand side of equation (1), we get
dn dn −1
k0  c y + c2 y2 + … + cn yn  + k1 n −1 c1 y1 + c2 y2 + … + cn yn 
n  1 1
dx dx
d
+ … + kn −1 c1y1 + c2 y2 + … + cn yn  + kn c1y1 + c2 y2 + … + cn yn 
dx 
 dny dn−1y dy 
= c1 k0 n1 + k1 n−11 + … + kn−1 1 + kny1 
 dx dx dx 
 dn y dn −1y dy 
+ c2 k0 n2 + k1 n −12 + … + kn −1 2 + kn y2 
 dx dx dx 
 dny dn −1y dy 
+…+ cn  k0 nn + k1 n −1n + … + kn −1 n + kn yn 
 dx dx dx 
= c1 [0] + c2 [0] + … + cn [0] = 0, since y1(x), y2(x), …, yn(x)
are solution of the linear equation. This proves the theorem.
Remarks: The above n linearly independent solutions y1(x), y2(x),…,yn(x) are called the fundamental solutions
of equation (1) and the set comprising them forms a basis of the nth order linear homogeneous equations.

The solution (3), y = c1y1 + c2y2 + … + cnyn = u(x) is a combination of n linearly independent
solutions containing n arbitrary constants c1, c2,…, cn. It is called the general solution of
equation (1). It is also known as the complementary function (C.F.).
Further, if y = v(x) be a solution of the non–homogeneous equation containing no arbitrary
constant is called it’s particular solution (P.I.). Therefore, y = u(x) + v(x) = C.F. + P.I. is called
the complete solution of equation (1). Hence, in order to solve equation (1), first find the
general solution (C.F.) and then find the particular solution (P.I.).
 Linear Differential Equations of Second and Higher Order 579

Linear Independence and Dependence of Solutions

Functions y1(x), y2(x) ,…, yn(x) are said to linearly independent on some interval of definition,
say I, if the relation (3) viz. c1y1 + c2y2 + … + cnyn = 0. Implies c1 = 0 = c2 = … = cn. This system
(or set) of linearly independent solutions is called fundamental system (or set) of solution
(or integrals).
However, these functions are said to be dependent on the interval of definition, say I, if
relation (3) holds for c1, c2 ,…, cn not all zero. In this case, one or more functions can be
expressed as a linear combination of the remaining functions.
1
if c1 ≠ 0, then y1 = − c2 y2 + c3 y3 + … + cn yn 
c1 
e.g. …(4)

Conversely, if any of yi's can be expressed as the linear combination of the remaining
functions y1, y2 ,…, yi – 1, yi + 1 ,…, yn then the given set of functions are linearly dependent.
Theorem: The necessary and sufficient condition that n integrals y1, y2, …, yn of the linear
dn y dn −1y
differential equation (2) viz. P0
n
+ P1 n −1 + … + Pn y = 0 , where P0, P1, P2 ,…, Pn (P0 ≠ 0)
dx dx
are continuous functions of x on a common interval I or constants, be linearly independent is
that the determinant, W* (Wronskian),

y1 y2 … … … … … yn
y1' y2' … … … … … yn'
… … … … … … … … does not vanish identically on I.
… … … … … … … …
n −1
y1 y2n −1 … … … … … ynn −1

Proof: Necessary Condition: If y1(x), y2(x) ,…, yn(x) are not linearly independent then there
are constants c1, c2 ,…, cn not all zero such that c1y1 + c2y2 … + cnyn = 0.
Also c1y1(i) + c2y2(i) + … + cnyn(i) = 0, i = 1, 2 ,…, n – 1.
It follows the determinant

y1 y2 … … … … … yn
y1' y2' … … … … … yn'
W * ( x) = … … … … … … … …
… … … … … … … …
y1n −1 y2n −1 … … … … … ynn −1

c1y1 + c2 y2 + ………… + cn yn y2 … … … yn
c1y1´+ c2 y2´ + ………… + cn yn´ y2 ´ … … … yn´
= …………………………… … ………
…………………………… …………
c1y1n −1 + c2 y2n −1 + … + cn ynn −1 y2n −1 … … … ynn −1
580 Engineering Mathematics through Applications

0 y2 … … … … … yn
0 y2' … … … … … yn'
1
= … … … … … … … … =0
c1
… … … … … … … …
0 y2n −1 … … … … … ynn −1

Thus, if y1(x), y2(x) ,…, yn(x) are not linearly independent, then the determinant W*
(Wronskian) will be identically zero.
Sufficient Condition: If W*(x) = 0, then there can be found constants c1, c2, …, cn not all
zero such that c1y1 + c2y2 + … + cnyn = 0;
whence c1y1(i) + c2y2(i) + … + cnyn(i) = 0, i = 1, 2, ,…, n – 1.
Now W* = 0 implies that the determinant must be reducible to the form wherein all the
elements of one row (or one column) are zero or in other words there must be certain
multiplier λ1, λ2 ,…, λn such that

λ1y1 + λ2 y2 + … … … … … + λnyn = 0 …(i) 

λ1y1´+ λ2 y2´+ … … … … … + λnyn´= 0 …(ii) 

… … … … … … … …  …(5)
… … … … … … … … 

λ1y1n −1 + λ2 y2n −1 + … … … … … + λnynn −1 = 0 …(n) 

Differentiating each of these equations and subtracting from the next following equations,
we get

λ1´y1 + λ2´y2 + … … … … … + λn´yn = 0 …(i) 

λ1´y1´+ λ2´y2´+ … … … … … + λn´yn´= 0 …(ii)

… … … … … … … …  …(6)
… … … … … … … … 

λ1´y1n −1 + λ2´y2n −1 + … … … … … + λn´ynn −1 = 0 …(n) 
Ist operation is explained as: d(λ1y1 + λ2y2 + … + λnyn) = 0 implies
(λ1y'1 + λ2y'2 + … + λny'n) + (λ1' y1 + λ2'y2 + … + λn'yn) = 0
and on subtraction this from (λ1y1 + λ2y2 + … + λnyn) = 0 results in
(l1y1 + l2y2 + … + lnyn) = 0 and so on.

y1 y2 … … … … … yn
y1´ y2´ … … … … … yn´
Equation (6) in matrix form is equivalent to … … … … … … … … =0
… … … … … … … …
y1n−1 y2n−1 … … … … … ynn−1

If one of the determinant, say formed by omitting the rth column, then there will be a
relation, a1y1 + a2y2 + – – – – + ar – 1yr – 1 + ar + 1yr + 1 + – – – + anyn = 0
 Linear Differential Equations of Second and Higher Order 581

In general, if one of these determinants vanishes, then from (6) and the first (n – 1) equations
in (5), we get
λ1´ λ2´ λ´
= = … = n = µ (say),
λ1 λ2 λn

λ1´ d(λ1)
implying = µ or = µ or log λ1 = ∫ µ dx or λ1 = c1e∫ µdx
λ1 λ1

Similarly, λ2 = c2e∫ µdx , …, λn = cn e∫ µdx ; c1 , c2 ,… , cn being constants. On substituting these

λi's in the equation (5(i)), we get c1y1 + c2y2 + … + cnyn = 0.
Hence, if the condition holds for (n – 1) functions, it also holds for n. Whence the necessary
and sufficient condition that y1, y2, y3, …, yn forms a system of linearly independent integral
is that determinant W* does not vanish indentically.

9.3 OPERATOR ‘D’ AND COMPLEMENTARY FUNCTION

To solve the equation

dn y dn−1y dy
n
+ K1 n −1
+ … + Kn −1 + Kn y = 0 …(1)
dx dx dx
We first define the operator ‘D’

dy d
Operator D: If we write = Dy so that D = is an operator which when applied to a
dx dx
dn y
function of ‘x’, differentiates that function with respect to ‘x’. Likewise, = Dn y and so
dxn
dn  dm y  dm+ n y
on. Further, = or DmDn = DnDm = Dm+ n Means the differential operator,
dxn  dxm  dxm + n
‘D’ fully obeys the laws of algebra.
Whence, the linear differential equation (1) in symbolic form may be written as
Dny + k1Dn – 1y + … + kn – 1Dy + kny = X(x) …(2)
Or more precisely, f(D)y = X where in f(D) can be treated much the same as an algebraic
expression in D.
Further, (Dn + K1Dn – 1 + … + kn – 1D + kn) = 0 or f(D) = 0 … (3)
is called auxiliary equation to (2) with m1, m2, --------, mn as its n roots, means further
discussion on solution of (1), depends on the nature of these n roots.

Case1: When all the roots are real and distinct:

In this case, equation (2) will become
(D – m1) (D – m2) … (D – mn)y = 0 … (4)
It will be satisfied by the solutions of
(D – m1)y = 0; (D – m2)y = 0; ……; (D – mn)y = 0,
582 Engineering Mathematics through Applications

dy
where (D – m1)y = 0 means − m1y = 0 which is a Leibnitz linear equation
dx
Here, I.F. = e−m1x and y.e −m1x = c1 or y = c1em1x . Like wise others.
Therefore, general solution becomes
y(x) = c1em1x + c2 em2x + … + cnemnx … (5)

Case 2: When two roots are real and equal:

In this case, equation (2) corresponding to the two repeated roots becomes
(D – m1) (D – m1)y = 0
Let (D – m1)y = z, so that above equation reduces to
(D – m1)z = 0 or Dz – m1z = 0
Again a Leibnitz Linear, resulting in z = c1e−m1x so that (D – m 1 )y = z becomes
dy
(D − m1 )y = c1em1x or − m1y = c1em1x , a Leibnitz Linear with P = −m1 , Q = c1em1x and its I.F. is
dx
e−m1x and solution becomes ye 1 = ∫ ( c1e 1 ) e 1 dx = c1x + c2 or y = ( c1x + c2 ) e 1
−m x m x −m x mx

Thus, the general solution of equation (2), corresponding to two repeated roots becomes
y = (c1x + c2 ) em1x + c3 em3 x + … + cnemn x … (6)

Case 3: (i) When one pair of roots is imaginary (complex):

Let m1 = α + iβ and m2 = α – iβ and the remaining roots are real, then the general solution of
equation (1) becomes
y = c1e(α+ iβ )x + c2e(α −iβ)x + c3em3x + … + cnemnx

= eαx ( c1eiβx + c2 e−iβx ) + c3 em3x + … + cn emnx

= eαx  c1 ( cos β x + i sin βx ) + c2 ( cos β x − i sin β x ) + c3 em3 x + … + cnemn x

= eαx ( c1 + c2 ) cos βx + i ( c1 − c2 ) sin βx  + c3em3x + … + Cnemmx

= eαx [C1 cos βx + C2 sin βx ] + C3 em3x + … + Cnemnx … (7)

where (c1 + c2) = C1, i(c1 – c2) = C2, c3 = C3 etc.
Case 3: (ii) When one pair of roots is equal and imaginary:
(complex roots equal means m1 = α + iβ = m2, m3 = α – iβ = m4)
In this case general solution becomes.
{ }
y = eα x (c1x + c2 ) cos βx + (c3 x + c4 ) sin βx + c5 em5x + … + cnemnx … (8)

Example 1: Find the solution of the differential equation 4y”’ + 4y” + y’ = 0.

Solution: The given equation in symbolic form is written as (4D3 + 4D2 + D)y = 0 and its
auxiliary equation is 4D3 + 4D2 + D = 0
 Linear Differential Equations of Second and Higher Order 583

D(4D2 + 4D + 1) = 0 D(2D + 1)2 = 0 i.e. D = 0, − 1 , − 1

2 2
x
−
Whence y (C.F.) = c1e0 + (c2 x + c3 ) e 2 .

d3 y
Example 2: Solve +y=0
dx3

Solution: The given equation in symbolic form is given by (D3 + 1)y = 0 and its auxiliary

1 3
equation is D3 + 1 = 0 i.e. (D + 1) (D2 – D + 1) = 0 or D = − 1, ±i
2 2
x
 3 3 
Whence C.F. = c1e−x + e 2  c2 cos x + c3 sin x
 2 2 

d4 x
Example 3: = m4 x , show that x = c1 cos mt + c2 sin mt + c3 cosh mt + c4 sinh mt .
dt 4
Solution: The given equation in symbolic form is (D4 – m4)x = 0
The auxiliary equation becomes (D2 – m2) (D2 + m2) = 0
Which implies either D2 – m2 = 0 or D2 + m2 = 0
i.e. D = ±m and D = ±im
Hence the required solution is
x =  aemt + be−mt  +  ceimt + de−imt 

 ( cosh mt + sinh mt )   ( cosh mt − sinh mt ) 

= a  + b 
 2   2 
 ( cos mt + i sin mt ) (cos mt − i sin mt) 
+ c +d 
 2 2 
(a + b) (a − b)   (c + d) (c − d)
=  cosh mt + sinh mt  +  cos mt + i

sin mt 
 2 2   2 2 
⇒ x = [C1 cosh mt + C2 sinh mt ] + [C3 cos mt + C4 sin mt ]

a+b a−b c+d i(c − d)

where C1 = , C2 = , C3 = and C4 =
2 2 2 2

d4 y
Example 4: Solve + a4 y = 0 [NIT Kurukshetra, 2007; NIT Jalandhar, 2007]
dx4

Solution: Symbolic form of the given equation is

584 Engineering Mathematics through Applications

(D4 + a4)y = 0 …. (1)

Therefore, the auxiliary equation becomes
D4 + a4 = 0 or D4 + a4 + 2a2D2 – 2a2D2 = 0

(D2 + a2 ) − ( 2 aD ) = 0
2 2
or

(D2 + a2 − )(
2 aD D2 + a2 + 2 aD = 0 )
Which implies, either D2 + a2 − 2 aD = 0 …. (2)
or D2 + a2 + 2 aD = 0 ….(3)
On solving the two quadratic equations (2) and (3), we get
a a −a a
D= ±i and D= ±i respectively
2 2 2 2
Hence the required solution is

x) + e−
a a
y=e x + c2 sin x + c4 sin
x a a x a a
2 (c1 cos 2 2
2 (c3 cos 2 2
x) … (4)

1 (2n + 1)π
Alternately: D4 = − a4 i.e. D = (−1)4 a = a cis , n = 0, 1, 2, 3
4
Using, D. Moivre`s Theorem,

π π  1 1 
D1 = a  cos + i sin  = a 
a
n = 0, +i  = (1 + i),
 4 4  2 2 2

3π 3π   1 1 
D2 = a  cos
a
n = 1, + i sin  = a− +i =− (1 − i),
 4 4   2 2 2

5π 5π  1 1 
D3 = a  cos + i sin  = a  −
a
n = 2, + −i  =− (1 + i),
 4 4   2 2 2

7π 7π  1 1 
D4 = a  cos + i sin  = a 
a
n = 3, −i  = (1 − i)
 4 4  2 2 2

Now roots are comparable to (α ± iβ), (−α ± iβ) , so that complementary function becomes,

y = eα xc1 (cos β x + i sin β x) + e−α x c2 (cos β x − i sin β x)

−a
 a   a 
a
x a x a
=e 2
 c1 cos x + c2 sin x + e 2
 c3 cos x + c4 sin x
2 2  2 2 
 Linear Differential Equations of Second and Higher Order 585

ASSIGNMENT 1

d2 y dy
Solve (i)
dx 2
−6
dx
+ 9y = 0, (ii) (D2 + D + 1) y = 0,
d3 y d4 y d2 y
(iii) − y = 0, (iv) − 8 + 16y = 0
dx3 dx4 dx2

d2 y dy
(v) 2
− 2y + 10y = 0, y (0 ) = 4; y' (0 ) = 1
dx dx

(vi) 4y ′′′ + 4y ′′ + y ′ = 0 (vii) (D4 + 2n2D2 + n4 ) y = 0

(viii) l
d2θ
dt2
+ gθ = 0, with
θ=α
θ=0 } at t=0

9.4 INVERSE OPERATOR

1
(a) Defn: X ( x ) is that function of 'x' independent of arbitrary constants which when
f (D)

1
operated on by f(D) gives X(x). Hence is the inverse operator of f(D).
f (D)

1
Further, X ( x ) is the particular integral of the equation f ( D ) y = X ( x ) .
f (D)

1
(b) X = ∫ Xdx, no arbitrary constant is being added to.
D
1 1 dy
Let X(x) = y ⇒ D. X(x) = Dy or X = or dy = Xdx
D D dx
Integrating both sides, y = ∫ Xdx.
1
(c) X ( x ) = eax ∫ e−axX dx
D−a
1 1
X (x ) = y ⇒ (D − a ) X ( x ) = (D − a ) y
Let, D−a (D − a )
dy
X = Dy − ay or − ay = X
dx
which is a Leibnitz linear equation whose, integrating factor is e–ax and its solution becomes
y.e−ax = ∫ e−axX dx or y = eax ∫ e−axX dx
586 Engineering Mathematics through Applications

Problems based on use of Inverse operator in finding particular integral.

d2 y
Example 5: Solve the differential equation + a2y = sec ax
dx2

Solution: Equation in its symbolic form: (D2 + a2)y = sec ax … (1)

Thus auxiliary equation, D2 + a2 = 0 i.e. D = ± ia … (2)
∴ C.F. = eox (c1cos ax + c2sin ax … (3)
1 1 1
P.I. = sec ax = 2 sec ax = sec ax
D2 + a2 D − ( − a2 ) (D + ia )(D − ia )
1  1 1 
=  −  sec ax, (By Partial Fractions) … (4)
2ia D − ia D + ia 
 
∫ ∫
1 1
Now, sec ax = eiax sec ax e−iax dx,  using X(x) = eax Xe−ax dx
D − ia D−a 

∫
= eiax sec ax ( cos ax − i sin ax ) dx = eiax ∫ (1 − i tan ax ) dx
 i ( − log cos ax )   log cos ax 
= eiax  x −  = eiax  x + i  … (5)
 a   a
On the same lines, changing i to –i, we get

1  log cos ax 
sec ax = e−iax  x − i  … (6)
D + ia  a
Using (5) and (6), (4) becomes,

P.I. =
1  iax

2ia 
e x+i{log cos ax
a }
− e−iax x − i {
log cos ax 
a 

}
1  iax
∴ =
2ia 
x ( e − e−iax ) + i ( e + e−iax )
log cos ax iax
a 
x log cos ax
∴ = .sin ax + .cos ax … (7)
a a2
 eiax − e−iax eiax + eiax 
 since = sin ax and = cos ax 
 2i 2 
Hence the complete solution is
x log (cos ax ) 
y = (c1 cos ax + c2 sin ax ) +  sin ax + 2
cos ax 
a a 
d2 y
Note: If a = 1, then differential equation becomes + y = secx and its solution is
dx2
(c1cos x + c2sin x) + (x sin x + cosx log cos x)
 Linear Differential Equations of Second and Higher Order 587

d2 y
Example 6: Solve the differential equations + a2 y = cosec ax
dx2
Solution: The given equation in its symbolic form is (D2 + a2)y = cosec ax
Thus, the auxiliary equation becomes D2 + a2 = 0 i.e. D = ±ia
∴ C.F. = (c1cos ax + c2sin ax)
1
For P.I. = cosec ax
D2 + a2
1
= cos ec ax
D − (−a2 )
2

1
= cos ec ax
(D + ia )( D − ia )
1  1 1 
=  −  cos ec ax (By Partial Fractions)
2ia  D − ia D + ia 

 
∫
1 1
Now, cos ec ax = eiax e−iax cos ec ax dx  u sin g X = eax ∫ e−ax X dx
D − ia D−a 

∫
= eiax (cos ax − i sin ax)cosec ax dx

∫
= eiax (cot ax − i) dx

∫
eiax (cot ax − i) dx

 log(sin ax) 
= eiax  − ix  … (1)
 a 
On the same lines, changing i to –i, we get
1  log(sin ax) 
cos ec ax = e−iax  + ix  … (2)
D + ia  a 
Using (1) and (2),

P.I. =
1  iax
2ia  e {
x+i
log cos ax
a }
− e−iax x − { log cos ax 
a 

}
1  log(sin ax) 
= (eiax − e−iax ) − ix(eiax + e−iax )
2ia  a 
1  log sin ax 
= sin ax − x cos ax 
a 
… (3)
a 

 eiax − e−iax eiax + e−iax 

 since = sin ax and = cos ax 
 2i 2 
588 Engineering Mathematics through Applications

Hence the complete solution is

1  log(sin ax) 
y = (c1 cos ax + c2 sin ax ) +  sin ax − x cos ax 
a a 
2
Note: If a = 1, then differential equation reduces to d y + y = sec x and the corresponding solution
dx2
becomes y = (c1 cosx + c2 sinx) + sinx(log sinx) – x cosx. However, these above two problems are easier if
followed under the method of variation of parameter, discussed in subsequent articles.

9.5 GENERAL PROCEDURE FOR FINDING PARTICULAR INTEGRAL

Consider the equation,
dn y dn −1y dy
+ k1 n −1
+ … + kn −1 + kny = X ( x ) … (1)
dx n
dx dx
Which in symbolic form becomes
(Dn + k1Dn – 1 + … + kn – 1D + kn)y = X(x) or f(D)y = X(x) …(2)
1
Thus the particular integral, y(x) = X(x)
f (D )

Case 1: when X(x) = eax, then

1 ax eax
P.I. = e = , provided f (a) ≠ 0 … (3)
f (D) f (a)

Deax = aeax 
D2 eax = a2 eax 

Since, .....................
.....................
Dneax = aneax 

∴ (Dn + k1Dn – 1 + … + kn – 1D + kn)eax = (an + k1an – 1 + … + kn – 1a + kn)eax

f(D)eax = f(a)eax
1
Operating on both sides f (D) ,

1 1 ax 1 ax eax
⋅ f (D)eax = f (a) e or e = , f (a) ≠ 0
f (D) f (D) f (D) f (a)
If f(a) = 0, it is a case of failure and then we proceed as

1 ax 1 eax
e =x eax = x , provided f ′(a) ≠ 0
f (D) d
 f (D) f ′(a) … (3a)
dD
 Linear Differential Equations of Second and Higher Order 589

If f(a) = 0, then (D – a) must be a factor of f(D)

d
f (D) = (D − a) φ (D) and  f (D) = (D − a)φ′(D) + 1.φ(D),
dD 
i.e.

d
which means, f ′(a) =  f (D) at D = a = φ(a), provided φ(a) ≠ 0
dD 
1 ax 1 1 1 1 ax ax −ax
So that e = eax = eax = e ∫ e e dx
f (D) (D − a) φ(D) φ(a) (D − a) φ(a)
eax eax
= x=x
φ(a) f ′(a) , provided f’(a) ≠ 0
Further if f’(a) = 0 then again it is case of failure and
1 ax eax
e = x2 , f ′′(a) ≠ 0 … (3b)
f (D) f ′′(a)

Example7: Solve the differential equation (D2 – 2D – 3)y = 3e2x

Solution: The given equation (D2 – 2D – 3)y = 3 e2x … (1)

has auxiliary equation as: D2 – 2D – 3 = 0 or D = 3, – 1 .…(2)
Thus the complementary function, C.F. = c1e3x + c2e–x … (3)
For particular integral,

1  1 ax eax 
P.I. = 3 e2x
 comparable to e = , f (a) ≠ 0 
D − 2D − 3
2
 f (D) f (a) 
3
= e2 x = −e2 x … (4)
(2)2 − 2(2) − 3
Whence the complete solution, y = (c1e3x + c2 e– x) – e– 2x

Example 8: Solve the differential equation (D3 – 3D2 + 4D – 2)y = ex

Solution: Auxiliary equation for the given equation becomes

D3 – 3D2 + 4D – 2 = 0 or (D – 1)(D2 – 2D + 2) = 0
2+ 4−8
Implying D = 1, i.e. 1, (1 ± i)
2
Whence yC.F.(x) = c1ex + ex(c2 cosx + c3 sinx)
1 1
And yP.I .(x) = 3 ex = ex
D − 3D + 4D − 2
2
1− 3 + 4 − 2
A case of failure as, f(a) = 0.
 
1 1
yPI = x ex , x eax , at D = a provide f ´(a) ≠ 0
3D − 6D + 4
2  f ( D)′ 

1
=x ex = x ex
3(1) − 6(1) + 4
2
590 Engineering Mathematics through Applications

d2 y
Example 9: Solve − 4y = cosh ( 2x − 1 ) + 3x [VTU, 2000; KUK, 2003–04]
dx2
Solution: Symbolic form of the equation: (D2 – 4)y = cosh (2x – 1) + 3x
Its auxiliary equation becomes,
(D2 – 4) = 0 ⇒ D2 = 4 or D = ±2
2x
So that yC.F.(x) = c1e + c2e – 2x

1 1
And y(x)P.I = 2 cos h ( 2x − 1) + 2 3x
D −4 D −4
1  e( 2x −1) + e−( 2x − 1)  1
yP.I =  + D2 − 4 3
x
D − 4 
implying

2
2

e−1 1 e 1 ex log 3
= e2x + e− 2x + 2 , [3x is comparable to az = ez.loga]
2 D −4
2
2D −4
2
D −4
1 1 2x e 1 − 2x e (log 3)x
= x e + x e + ,
(log 3)2 − 4 [3 = e = e
x x.log3 (log 3)x]
2e 2D 2 2D

x e2x x.e e− 2x 3x
= − + ,
2e 4 2. 4 (log 3 )2 − 4 (Rewrite, e(log 3)x = 3x)

x  e2 x −1 − e− (2 x −1)  3x
=   +
 ( log 3) − 4
2
4 2

x 3x
= sinh(2x − 1) +
4 (log 3)2 − 4

x 3x
∴ Complete Solution, y = c1e2x + c2 e−2x + sinh ( 2x − 1) +
4 (log 3 )2 − 4
Case 2: When X(x) = sin(ax + b) or cos(ax + b), then
1 sin(ax + b)
yP.I. = sin(ax + b) = , f (− a2 ) ≠ 0
2
f (D ) f (− a2 )
Since Dsin(ax + b) = a cos(ax + b)
D2sin(ax + b) = – a2sin(ax + b)
D3sin(ax + b) = – a3cos(ax + b)
D4sin(ax + b) = a4sin(ax + b)
or (D2)2sin(ax + b) = (– a2)2sin(ax + b)
………………………………
………………………………
In general, (D2)rsin(ax + b) = (– a2)rsin(ax + b)
Adding all the above expressions,
 Linear Differential Equations of Second and Higher Order 591

((D2 )n + k1(D2 )n−1 + … + kn−1D2 + kn ) sin(ax + b)

= ((− a2 )n + k1(− a2 )n −1 + … + kn −1(− a2 ) + kn ) sin(ax + b)

or f(D2) sin (ax + b) = f(–a2) sin (ax + b)

1
Operating on both sides the inverse operator
f (D2 )

1 1
f (D2 )sin(ax + b) = f (− a2 )sin(ax + b)
f (D2 ) f (D2 )

sin(ax + b)
, f ( −a2 ) ≠ 0
1
or sin(ax + b) =
f (D2 ) f (− a2 )
In case of failure where f(–a2) = 0, we proceed as:
1 sin(ax + b)
sin(ax + b) = x , provided f ’(–a2) ≠ 0 … (4a)
2
f (D ) f ′(−a2 )
Similarly, if f ’(–a2) = 0 then
1 sin(ax + b)
sin(ax + b) = x2 , f ′′(−a2 ) ≠ 0 and so on. …(4b)
2
f (D ) f ′′(−a2 )

Example 10: Solve the equation (2D2 + D – 1)y = 16 cos 2x [Jammu Univ. 2002]

Solution: Given equation (2D2 + D – 1)y = 16 cos 2x … (1)

1
Its auxiliary equation as: 2D2 + D – 1 = 0 i.e. D = , −1 …(2)
2
(C.F.) = c1e 2 + c2e− x
x
Thus complementary function. … (3)

1  1 cos(ax + b) 
P.I. = 16 cos 2x Q cos(ax + b) = ; f (− a2 ) ≠ 0
(2D + D − 1)
2
 f (D )
2
f (− a )
2


16
= cos 2x (as here a = 2)
2(−4) + D − 1
16 16 (D + 9)
= cos 2x = cos 2x
D−9 (D − 9)(D + 9)
16
= (D + 9)cos 2x
D − 81
2

16
= [D cos 2x + 9cos 2x ] = 16 (2 sin 2x − 9cos 2x) … (4)
− −
4 81 85
592 Engineering Mathematics through Applications

Hence, the complete solution becomes

 x  16
y =  c1e 2 + c2 e−x  + ( 2 sin 2x − 9cos 2x )
  85
d2 y
Example 11: Solve the equation + 4y = sin 2x.
dx2
Solution: Symbolic form of the given equation, (D2 + 4)y = sin 2x
Corresponding auxiliary equation, D2 + 4 = 0 i.e. D = ±2i
Thus, y(C.F) = (c1 cos2x + c2 sin2x)
1
y ( P.I.) = sin 2x; Replace D2 = – a2 = – 4, but here f(–a2) = 0
D2 + 4
 1 1 
In case of failure, sin ( ax + b ) = x sin ( ax + b )

2
f (D ) f ′(−a )
2

x − cos 2x 
sin 2x = 
1 x cos 2x
Implying y ( P.I.) = x =− .
2.D 2  2  4
Hence the complete solution, y = (c1 cos 2x + c2 sin 2x ) −
x cos 2x
.
4

1
Case 3: When X(x) = xm then, yP.I. = f(D) x = [1 ± φ (D) ] x , where m is a positive integer.
m –1 m

From f(D), take the lowest degree term outside so that the remaining expression in f(D)
becomes [1 ±φ(D)]. Now take [1 ±φ(D)] to the numerator and expand it by binomial theorem
upto Dm so that (m + 1)th and higher order derivatives of xm are zero.
Note: Some useful results:
(i) (1 – D)–1 = 1 + D + D2 + ...…………… ∞
(ii) (1 + D)–1 = 1 – D + D2 + ...…………… ∞
(iii) (1 – D)–2 = 1 + 2D + 3D2 + …………… ∞
(iv) (1 – D)–3 = 1 + 3D + 6D2 + …………… ∞

d3 y d2 y dy
Example 12: Solve − −6 = 1 + x2
dx3 dx2 dx

Solution: Symbolic form becomes (D3 – D2 – 6D)y = 1 + x2

Auxiliary Equation, D3 – D2 – 6D = 0
Implying D(D + 2)(D – 3) = 0 or D = 0, 3, – 2
Whence yC.F. = (c1 + c2 e3x + c3e– 2x)
1 1
yP.I . = (1 + x2 ) = (1 + x2 )
D − D − 6D
3 2
 D D2

And − 6D  1 + − 
 6 6 
 Linear Differential Equations of Second and Higher Order 593

−1
1   D D2  
=− 1+  −  (1 + x )
2
6D   6 6  

1   D D2   D D2  
2
=−  
1 − −  +  −  + − − − +  (1 + x2 )
6D   6 6  6 6  

1   D D2  D2 
=− 1− −
6D   6
+
6  36 
(1 + x2 )

1 
(1 + x2 ) − (1 + x2 ) + D (1 + x2 )
D 7 2
=− 
6D  6 36 

1 
(2) = −
1  2 x 25 
=− 
6D 
(1 + x2 ) − +
x 7
3 36  6D 
x − + 
3 18 

1  x3 1 x2 25 
=−  − +
6  3 3 2 18 
x = −
1
108
(6x3 − 3x2 + 25x)

y(x) = ( c1 + c2 e3 x + c3 e−2 x ) − (6x3 − 3x2 + 25x )

1
Therefore
108

Example 13: Solve the differential equation (D4 + 16D2)y = x2 + 5.

[Jammu Univ, 2002; NIT Kurukshetra, 2006]

Solution: Given (D4 + 16D2)y = x2 + 5. …(1)

A.E., D4 + 16D2 = 0 i.e. D2(D2 + 16) = 0 or D = 0, 0, ±4i … (2)
Thus, the complementary function,
C.F. = (c1 + c2x) + (c3 cos 4x + c4 sin 4x) … (3)
For particular integral,

 
( x2 + 5)
1 1 m
 comparable to f (D) x = f (D) x 
−1 m
P.I. =
D + 16D2
4

=
1
D2 ( D2 + 16 )
( x2 + 5) = D12  1 1  ( x2 + 5)
16 1 + D2
 16 
−1
1  1 2
= 1 +
16D2 
D 
16 
( x2 + 5 )
1  1 2 2
= 1 − D  ( x + 5)
16D2  16 
594 Engineering Mathematics through Applications

1  2 1  1 1  2 39 
=  x + 5 − .2 =
16D2  16
 x +  dx
16 D  8 ∫
1 1  x3 39  1  x4 39x2 
=
 + x = +
16 D 3 8  16  12 16 
Hence complete solution becomes,
x2  x2 39 
y = (c1 + c2 x ) + (c3 cos 4x + c4 sin 4x ) + + 
64  3 4

1 ax 1
Case 4: When X(x) = eax V(x), then, e V (x) = eax V (x)
f (D) f (D + a)
For some u(x),
D(eaxu) = eaxDu + aeaxu = eax(D + a)u
D2(eaxu) = eaxD2u + 2aeaxDu + a2eaxu = eax(D + a)2u
………
………
Dn(eaxu) = eax(D + a)nu
Adding all,
Dn(eaxu) + k1Dn – 1(eaxu) + … + kn – 1D(eaxu) + kn(eaxu)
= eax(D + a)nu + k1eax(D + a)n – 1u + … + kn – 1 (D + a)u + knu
Implying f(D)(eaxu) = eax f(D + a)u

Operating f(D) on both sides, f (D) f (D) ( e u ) = f (D) e f (D + a)u 

1 ax 1 ax

1
eax u = eax f (D + a)u
f (D) 
or

1
Now put f(D + a)u = V i.e. u= V
f (D + a)

So that eax
1
f(D + a)
V =
1
f(D)
(
eax V )
Example 14: Solve (D4 – 1)y = ex cos x

Solution: The corresponding Auxiliary Equation is

D4 – 1 = 0 i.e. (D2 – 1)(D2 + 1) = 0 or D = ± 1, ±i
Hence the complementary function
y(C.F.) = (c1ex + c2 e– x) + (c3 cos x + c4 sin x)
For particular Integral,
1 1
y ( P.I.) = ex cos x = ex cos x
(D4 − 1) ( D + 1)4 − 1
 Linear Differential Equations of Second and Higher Order 595

1
= ex cos x
(D4 + 4D3 + 6D2 + 4D + 1) − 1
1
= ex cos x
(D ) 2 2
+ 4D (D ) + 6 ( D2 ) + 4D
2

1
= ex cos x (replace D2 = – a2 = – 1)
1 + 4D(−1) + 6(−1) + 4D
cos x
= ex
−5
Hence the complete solution is

y = ( c1ex + c2e− x ) + (c3 cos x + c4 sin x ) −

1 x
e cos x
5

Note: For problem (D4 – 1)y = cosx.coshx, rewrite, (D4 − 1) y = 1 (ex cos x + e−x cos x )
2

ex cos x e−x cos x

Here, PI1 = and PI2 =
−5 −5

d2 y dy
+3 + 2y = ee
x
Example 15: Solve the differential equation
dx2 dx
[KUK, 2003; NIT Kurukshetra 2010]

Solution: Symbolic form of the given equation, (D2 + 3D + 2)y = ee

x

Auxiliary equation becomes D2 + 3D + 2 = 0 or D = –1, –2.

C.F. = c1e– x + c2 e– 2x

And y(P.I.) =
1
D + 3D + 2
2
ee = 2
x 1
D + 3D + 2
e− x ex ee
x
( )
 1 ax 1 x

 Comparable to f (D) e V(x) = e f (D + a) V(x) with a = −1 and V(x) = ex ee 

ax
 

= e− x
1
(D − 1)2 + 3(D − 1) + 2
(e e )
x ex

= e−x
1
D +D
2
x
D ee , ( ) as
 ( ) = e D (e ) = e
D ee
x ex x ex ex 


= e−x
1 1
(D + 1) D
D ee
x
( )
 1 x 
= e−x  ∫
e(e )  = e−x  e−x ex .ee dx  ,  using 1 X(x) = eax e−ax Xdx
∫
x

 (D + 1)  
   D−a 
596 Engineering Mathematics through Applications

∫ ( ) (
= e−x  e−x d ee dx  = e−x  e−x ee  = e−2x ee )
x x x

   
Alternately:

yP.I . =
1
D + 3D + 2
2
ee =
x 1
(D + 1)(D + 2)
ee
x
( )
1 ex 1 ex
= e − e (by partial fraction )
D+1 D+2

= e−x ∫ ex ee dx − e− 2x ∫ e2x ee dx
x x

x
( x x
)
= e−x ee − e− 2x ex ee − ∫ ee ex dx = e− 2x ∫ d ee dx ( )
x

= e − 2x e e .
x

Example 16: Solve the differential equation (D3 + 2D2 + D)y = x2e2x + sin2x
[PTU, 2003; NIT Kurukshetra, 2008]

Solution: The auxiliary equation given by D3 + 2D2 + D = 0 i.e. D = 0, –1, –1

∴ C.F. = c1 + (c2x + c3)e–x

For P.I. =
1
D3 + 2D2 + D
( x2e2x + sin2 x ) = P.I.1 + P.I.2
1
P.I.1 = x2 e2 x
D3 + 2D2 + D
1
= e2 x x2
(D + 2 ) 3
+ 2 (D + 2 ) + ( D + 2 )
2

1
= e2 x x2
  21 8 2 1 3 
18 1 + D+ D + D
  18 18 18  
−1
e2 x  7 4 2  2
= 1 +  6 D + 9 D + ………  x
18  

e2x  7 4 2  +  7 D  + ……… x2
2
=  
1 − D + D + ………    
18   6 9  6  

=
e2x 1 − 7 + 33 2  2 = e2x  2 − 7 2 + 33 .2
 D D x x x
18 6 36  18  6 36 

=
e2x  2 − 7 + 11 
18  x 3
x
6 
 Linear Differential Equations of Second and Higher Order 597

1
P.I.2 = sin2 x
D + 2D2 + D
3

1 1 − cos 2x
=
D + 2D + D
3 2
2

1 1 1 1
= e0x − cos 2x
2D 1 + (2D + D2 ) 2 D(D2 ) + 2D2 + D

1 1 cos 2x
= −
2D 2 (−4D − 8 + D)

=
1 1 1
+
1 ( 3D − 8 ) cos 2x
2 D 2 ( 3D + 8 ) ( 3D − 8 )

1 1 3D − 8 x 1 (3D − 8)cos 2x
= x+ cos 2x = +
2 2 9D2 − 64 2 2 ( −36 − 64 )

x 1
= − ( 3D cos 2x − 8 cos 2x )
2 200

x 3 sin 2x + 4 cos 2x
= +
2 100
Hence the complete solution is
x 3 sin 2x + 4 cos 2x e2 x  2 − 7 + 11 
y = c1 + (c2 x + c3 ) e− x + + +  x x
2 100 18 3 6 

Case 5: When X = xV, V being a function of x, then

1 1  d 1 
(xV ) = x V+ V
f (D) f (D)  dD f (D) 
By Leibnitz’s Theorem on successive differentiation, we have
Dn(Xx) = nc0(DnX).x + nc1Dn – 1X.1
⇒ Dn(xX) = x DnX + n Dn – 1X

 d n ,  d n 
⇒ Dn ( xX ) = x DnX +  D X Q D = nDn −1  … (1)
 dD  dD 

 d 
or f (D)(xX) = x f (D)X +  f (D)  X … (2)
 dD 
(if Dn represents a polynomial of nth degree in D)
598 Engineering Mathematics through Applications

1 1 1
Putting f(D)X = V, we get ( f (D)X) = V or X = V … (3)
f (D) f (D) f (D)
With the help of (3), (2) becomes

 1   1 
f (D)  x. V  = xV
. + f ′(D)  V
 f (D)   f (D) 

1
Operating on both sides of above equation, we get
f (D)

 1  1 1 f ′(D)
 x. f (D) V  = f (D) (xV
. )+
f (D) f (D)
V

1 f ′(D) 1
or x. V− 2V = (xV
. )
f (D)  f (D) f (D)

 d 1   f ′(D) 
or x.
1
V −  V=
1
(xV
. ),  since

d
( f (D)) = −
−1
2
f (D)  dD f (D)  f (D) dD  f (D) 

d2 y
Example 17: Solve + 4y = x sin x [Raipur, 2004]
dx2

Solution: Symbolic form, (D2 + 4)y = x sin x

A.E., D2 + 4 = 0 implying D = ± 2i so that
yC.F.(x) = (c1cos 2x + c2 sin 2x)

1
y(P.I.) = (x sin x) (Comparable to case 5.)
D2 + 4
1 2D
=x sin x − sin x
D +4
2
(D + 4)2
2

1 2D
=x sin x − sin x Replace f(D2) = f(– a2)
(− 1) + 4 ( + 4 )2
− 1
x sin x 2 1 2
= − D (sin x ) = x sin x − cos x.
3 9 3 9

y(x) = yC.F. + yP.I. = (c1 cos 2x + c2 sin 2x ) +

1
∴ (3x sin x − 2 cos x )
9
 Linear Differential Equations of Second and Higher Order 599

d2 y dy
Example 18: Solve +3 + 2y = xex sin x [VTU, 2001; Osmania, 2003]
dx2 dx

Solution: Here A.E. is D2 + 3D + 2 = 0 ⇒ D = –1, –2

y(C.F.) = c1e–x + c2e–2x
1 1
y( P.I .) = x ex sin x = ex ( x sin x )
D2 + 3D + 2 (D + 1)2 + 3 (D + 1) + 2
1
= ex ( x sin x )
D + 5D + 6
2

  sin x 
sin x + 
1 d 1
= ex  x 2  ,
 D + 5 D + 6  dD D2
+ 5 D + 6  

 
1 2D + 5
= ex  x sin x − sin x 
 −1 + 5D + 6 ( ) 
2
 D2
+ 5D + 6 

 1 2D + 5 
= ex  x sin x − 2 sin x 
 5 (D + 1) ( −1 + 5D + 6) 

x D + 1 2D + 5 
= ex  sin x − 2 sin x 
 5 D − 1
2
25 ( D + 1) 

 x
= ex  − (D sin x + sin x ) −
1 ( 2D + 5 ) sin x 

 10 25 ( −1 + 2D + 1) 

1 2D + 5
= ex  − (D sin x + sin x ) − sin x 
x
 10 25 2D 

= ex  − (cos x + sin x ) − sin x)

x 1 5 1
(sin x +
 10 25 2D 

= ex  − (sin x + cos x ) − (2 sin x − 5 cos x )

x 1
 10 50 

−2 x  x
Complete solution, y(x) = (c1e + c2 e ) − e  (sin x + cos x ) +
−x −2 x 1
(2 sin x − 5 cos x )
 10 50 

Example 19: Solve (D2 – 4D + 4)y = 8x2e2xsin2x

[ Delhi, 2002; JNTU, 2006; NIT Kurukshetra, 2007; UP Tech, 2007]

Solution: A.E. D2 – 4D + 4 = 0 or D = 2, 2
600 Engineering Mathematics through Applications

Whence, yC.F.(x) = (c1 + c2x)e2x … (1)

1
yPI (x) = 8x2 e2 x sin 2x
D2
− 4D+ 4

1
= 8 e2 x x2 sin 2 x
(D + 2 ) 2
− 4 (D + 2 ) + 4

= 8 e2 x
1 2
D2
x sin 2x = 8e2 x
DD
( x sin 2x )
1 1 2
… (2)

∫
1
= 8 e2x x2 sin 2x dx,
D

1  x2 x 1 
= 8 e2 x − cos 2x + sin 2x + cos 2x  , integration by parts
D  2 2 4 

= 8 e2 x  − ∫ ∫ cos 2x dx 
∫
1 2 1 1
x cos 2x dx + x sin 2x dx +
 2 2 4 

 1
= 8 e2 x  − x2
 2 {
sin 2x
2
− 2x
sin 2x
2 ∫
dx +
1
2 }
x sin 2x dx +
1
4 ∫

cos 2xdx 
 ∫
 x2 sin 2 x 
∫ ∫
1
= 8 e2 x  − + x sin 2 x dx + cos 2 x dx 
 4 4 

 x2 sin 2x ( − cos 2x ) 1 ( − cos 2x ) 

∫ ∫
1
= 8 e2 x  − +x − dx + cos 2x dx 
 4 2 2 4 

 x2 
∫
x 3
= 8 e2 x  − sin 2 x − cos 2x + cos 2 x dx 
 4 2 4 

 x2 x 3 
= 8 e2x  − sin 2 x − cos 2 x + sin 2x  … (3)
 4 2 8 
Therefore,

 x2 3 x 
y(x) = yCF + yPI = (c1 + c2 x ) e2x + 8 e2x  − sin 2x + sin 2x − cos 2x 
 4 8 2 
= [c1 + c2x – (2x2 – 3)sin2x – 4x cos2x] e2x

Alternately: From equation (2) onwards,

1 2 2ix 1
y.P.I. = Imag. Part of 8 e2 x x e = I.P. 8 e2 x e2ix x2
D2 (D + 2i )2
 Linear Differential Equations of Second and Higher Order 601

−2
1 1  iD 
= imag 8 e2x e2ix x2 = imag. 8 e2x e2ix 1− x2
  D 
2
−4  2 
2i  1 + 2i  

= imag.
8 2 x 2ix 
e e 1 + 2
iD
+3
( iD)2 + … x2

−4  2 4 

= imag. − 2 e2x e2ix  x2 + 2ix − 

3
 2 
= imag. e2x (cos 2x + i sin 2x) (3 − 2x2 ) − 4ix 

 x2 x 3 
= 8 e2x  − sin 2x − cos 2x + sin 2x 
 4 2 8 

ASSIGNMENT 2

d2 y d2 y
Solve 1. + a2 y = x cos ax, 2. − y = x2 cos x,
dx2 dx2
d2 y d4 y
3. + y = cos ec x, [MDU, 2002] 4. − y = x sin x, [KUK, 2005-06]
dx2 dx4
5. (D2 + a2)y = tan ax, [MDU, 2002; VTU, 2005]
6. (D2 + 4)y = ex sin2x 7. (D + 4D + 3)y = e– x sinx + x,
2

8. (D2 – 1)y = x sinx + (1 + x2)ex, [KUK, 2010]

9. (D2 – 4)y = x sinhx, [Madras, 2002; KUK, 2007]
10. (D4 + 1)y = x2 cosx, [Osmania, 2002]
d4 y
11. − y = cos x cosh x,
dx4
12. (D2 – 4D + 3)y = sin3x cos2x, [Madras, 2000]

d2 y dy
13*. −2 + y = x ex sin x, 14. (D2 + n2)y = x2ex, [*UP Tech; 2009]
dx2 dx
[VTU, 2001; Osmania, 2003; UPTech., 2005; JNTU, 2006; SVTU, 2007; PTU, 2009]
 1 x 1 1 
Hint: yPI = e V(x) = ex V(x) = ex 2 (x sin x); integration by parts
 f (0) f (D + 1) D 

9.6 SPECIAL METHODS FOR FINDING PARTICULAR INTEGRAL

I. Method of Variation of Parameters
The method of variation of parameters which is due to J.L Langrange (1736–1813), enables
us to find the general solution of any second order non homogeneous linear differential
602 Engineering Mathematics through Applications

equation (including both, with variable coefficients as well as with constant coefficients)
whose complementary function is known. Under this method, complementary function of
the reduced equation with no X(x) is used by replacing the arbitrary constants with functions
so chosen that a particular integral is obtained.
d2 y dy
Consider the equation, +P + Qy = X(x) … (1)
dx2 dx
y1 y2
If y1 and y2 are two linearly independent solutions of the above equation and W * =
y1' y2'
is the Wronskian of y1 and y2 (named after the noted Polish Mathematician and Philosopher
Hoene Wronsky (1778–1853) ), then particular integral.
yP.I . = u1y1 + u2 y2 = y1 −y2  ∗  dx + y2 y1  ∗  dx
∫ ∫
X X
… (2)
W  W 
Proof: Let general solution of the equation (1) be,
yC.F. = c1y1 + c2y2, … (3)
where c1 & c2 are arbitrary constants. The idea behind this method is that assume the arbitrary
constants of the general solution c1 & c2 as functions of x i.e. c1 = u1(x) & c2 = u2(x), so that
yP.I. = u1(x)y1 + u2(x)y2 … (4)

Differentiating (4),
dy
dx
( ) (
= u1y1′ + u2 y2′ + u1′ y1 + u2′ y2 )
On assumption that u’1 y1 + u’2 y2 = 0 …(6)

We get
dy
dx
(
= u1y1′ + u2 y2′ ) … (5)

Further, differentiating (6),

d2 y
dx2
( ) (
= u1y1″ + u2 y2″ + u1′ y1′ + u2′ y2′ ) … (7)

dy d2 y
On substituting values of y, , from (4), (6), (7) respectively into (1), we see
dx dx2
u1’y1’ + u2’y2’ = X … (8)
Now equations (5) and (8) constitute two linear equations in u1’ and u2’, which on solving,
results in
 X  X ,
u1′ = −y2 and u2′ = y1
 W∗   W∗  where W* = y1y2’ – y1’y2
From above u1(x) and u2(x) are obtained by integration and, whence the particular integral
and complete solution, y = yC.F. + yP.I. .
Observations:
1. Here constant of integration is not required, since we find particular integral, and any function u1(x) and
u2(x) satisfying the required conditions will take care of it.
2. Though this method is commonly applied for 2nd order equations but it can easily be extended to
equations of any order.
 Linear Differential Equations of Second and Higher Order 603

d3 y d2 y dy
e.g. Consider 3rd order equation P0(x)
3
+ P1(x) 2 + P2(x) + P3(x) y = R(x)
dx dx dx
with yC.F.(x) = c1y1(x) + c2y2(x) + c3y3(x), where y1(x), y2(x) and y3(x) are three linearly independent
solutions corresponding to homogeneous equation and c1, c2, c3 are arbitrary constants. In this case,
yP.I.(x) = u1(x) y1 + u2(x) y2 + u3(x) y3:
Here also we follow the same procedure as discussed earlier and obtain the required corresponding
equations for solving u1’, u2’, u3’ as

u1′y1 + u2′y2 + u3′y3 = 0 

u1′y1′ + u2′y2′ + u3′y3′ = 0 
R(x) 
u1′y1″ + u2′y2″ + u3′y3″ = = X(x)
P0(x) 
y1 y2 y3
Here, Wronskian (a non-zero entity) becomes, W * = y1' y2' y3'
y1" y2" y3"

d2 y
Example 20: Solve + y = tan x by the method of variation of parameters.
dx2
[KUK, 2004–05; PU, 2003-04; Raipur, 2004; NIT Kurukshetra, 2008; UP Tech, 2009]

Solution: Rewrite the given equation as (D2 + 1)y = tanx

Here A.E D2 + 1 = 0 or D = ±i
Thus yC.F.(x) = c1y1 + c2y2 = c1cosx + c2 sinx
Assume yP.I.(x) = u1(x) y1 + u2(x) y2 = u1cosx + u2 sinx
By method of variation of parameter, we have

X 
u1′(x) = −y2  ∗  
 W   and W * = y1 y2
=
cos x sin x
= 1; X = tan x

u2′(x) = y1  ∗ 
X y1′ y2′ − sin x cos x
 W 
On integration,
cos2 x − 1
∫ ∫ ∫
sin x
u1(x) = − sin x tan x dx = − sin x dx = dx
cos x cos x

= ∫ ( cos x − sec x ) dx = sin x − log (sec x + tan x )

And u2(x) = ∫ cos x tan x dx = ∫ sin x dx = − cos x

Whence yP.I.(x) = cosx[sinx – log(sec x + tan x)] + sin x(– cos x)
= – cos x log(secx + tan x)
And therefore the complete solution becomes,
y = (c1cos x + c2sin x) – (cos x) log (sec x + tan x)
604 Engineering Mathematics through Applications

d2 y
Example 21: Solve + y = x sin x , using method of variation of parameter.
dx2
[SVTU 2007; JNTU, 2005]

Solution: Symbolic form of the given equation, (D2 + 1) y = x sinx

Auxiliary equation, D2 + 1 = 0 i.e. D = ±i
Thus C.F. = c1y1 + c2y2 = c1cosx + c2 sinx … (1)
Let P.I. = u1(x)y1 + u2(x)y2;
Now by method of variation of parameter,

X 
u1′ ( x ) = −y2  ∗  
 W   and W* = y1 y2 = cos x sin x
= 1;
 −
u2′ ( x ) = y1  ∗  
X y1' y2' sin x cos x …(3)
 W  

(1 − cos 2x ) dx
∫ ∫ ∫
X
So that u1(x) = − y2 dx = − sin x ( x sin x ) dx = − x
W∗ 2
 x2 x sin 2x cos 2x 
= − + +  … (4)
 4 4 8 

∫ ∫ ∫
X 1
u2 (x) = y1 dx = cos x (x sin x) dx = x sin 2x dx … (5)
W∗ 2

=  − cos 2x +
x sin 2x 
 4 
8 

 x2 x 
P.I. = cos x  − + sin 2x + cos 2x + sin x  − cos 2x +
1 x sin 2x 
Thus,  4 
 4 4 8  8 

x2 x 1
=− cos x + ( sin 2x cos x − sin x cos 2x ) + ( cos x cos 2x + sin x sin 2x )
4 4 8

x2 x 1
=− cos x + sin x + cos x … (6)
4 4 8
Therefore complete solution
x2
y = ( c1 cos x + c2 sin x ) +
1 x
cos x + sin x − cos x
8 4 4

Example 22: Solve y” – 2y’ + 2y = ex tan x by variation of parameter. [Burdwan, 2003]

Solution: Rewrite the given equation, (D2 – 2D + 2)y = ex tanx.

Here A.E becomes, D2 – 2D + 2 = 0 D = 1 ± i
 Linear Differential Equations of Second and Higher Order 605

Thus, yC.F.(x) = c1y1 + c2y2 = ex [c1cosx + c2 sinx] … (1)

Assume: yP.I.(x) = u1(x)y1 + u2(x)y2 … (2)
By method of variation of parameter,

X 
u1´(x) = − y2  ∗  , u2´(x) = y1  ∗   ;
X
W   W 

y1 y2 ex cos x ex sin x
W* = =
y1’ y2’ − e sin x + e cos x ex sin x + ex cos x
x x

ex cos x ex sin x
=
e (cos x − sin x) ex (cos x + sin x)
x

Thereby on integration of u1´ & u2´, we get

∫ ∫
X X
yP.I .(x) = y1 −y2 dx + y2 y1 ∗ dx
W∗ W

∫ ∫
ex tan x ex tan x
= y1 −ex sin x dx + y2 e x
cos x dx
e2 x e2 x

∫ ∫
= −y1 sin x tan x dx + y2 cos x tan x dx = −y1I1 + y2 I2 … (3)

∫ ∫
I1 = tan x sin x dx = − tan x cos x + sec2 x cos x dx , integration by parts.

= – sin x + log(sec x + tan x) … (4)

I2 = ∫ sin x dx = − cos x … (5)

On putting the values of integrals, (4) and (5) again into (3),
yPI = – ex cos x[– sin x + log(sec x + tan x)] – ex sin x cos x
= – ex cos x log (sec x + tan x)
Hence the complete solution is
y = ex(c1cos x + c2sin x) – excos x log (sec x + tan x).

Example 23: Solve (D2 – 1)y = e–2x sin e–x using method of variation of parameter.

Solution: The given equation (D2 – 1)y = e– 2x sin e–x … (1)

Auxiliary form D2 – 1 = 0 or D = ±1
So that yC.F.(x) = c1y1 + c2y2 = (c1ex + c2e–x) … (2)
Assume yPI(x) = u1(x) y1 + u2(x) y2 … (3)
Now by the method of variation of parameter,

u1' = −y2  ∗  , u2' = −y1  ∗  ; where X = e2x sin e−x

X X
W  W 
606 Engineering Mathematics through Applications

y1 y2 ex e−x
And W* = = x = −2
y1´ y2´ e −e−x
Thereby giving
e−2x sin e−x t . t2 sin t dt
∫
u1 = − e−x
−2
dx = − ∫ −2 −t
(taking e–x = t)

−1  2
∫ ∫
t ( − cos t ) − 2t ( − cos t ) dt 
1 2
=− t (sin t ) dt =
2 2  

∫
= −  −t2 cos t + 2t (sin t ) − 2 (sin t ) dt 
1
2  
t2
= cos t − t sin t − cos t
2
e− 2 x
= cos e−x − e−x sin e−x − cos e−x … (4)
2
e−2x sin e−x
∫ ∫
1 −x
And u2 = − ex dx = e sin e−x dx
−2 2
−1 −1
∫
1
=sin t dt = cos t = cos e−x … (5)
2 2 2
Whence using (4) and (5) in (3), we get

 e − 2x  −1
yP.I .(x) = ex  cos e−x − e−x sin e−x − cos e−x  + e−x  cos e−x 
 2   2 

= –(sin e–x + excos e–x) … (6)

And therefore complete solution becomes
y(x) = yCF(x) + yPI(x) = (c1ex + c2e–x) – (sin e–x + excos e–x) … (7)

2
Example 24: Solve by the method of variation of parameter, d y − y = 2 .
dx2 1 + ex
[VTU 2005; Hisar, 2005]

Solution: Here auxiliary equation becomes D2 – 1 = 0, means D = ±1.

Thus yC.F.(x) = c1y1 + c2y2 = c1ex + c2e–x … (1)
By method of variation of parameter, we assume the particular integral as
yPI(x) = u1y1 + u2y2 = u1ex + u2e–x, … (2)
y1 y2 ex e−x
W* = = x = −2
y1´ y2´ e −e−x

 X  , ′( ) =  X  ,
and u1′ (x) = −y2 u x y1 …(3)
 W∗  2  W∗ 
 Linear Differential Equations of Second and Higher Order 607

 2 ⋅ 1 e−x
Implying ∫
u1(x) = −e−x  
 1 + e −2 
x
dx =
1 + ex
dx ∫ (Taking ex = t, exdx = dt)

∫ t (1 + t) dt =∫ t ∫ t dt +∫ 1 + t dt
1 1 1 1
= 2 2
dt − (By partial fractions)

1  1 + ex 
= − − log t + log (1 + t ) = log  x  − e− x
t  e 

f ′(x)
∫1+ e ∫ dx = − log (1 + ex )
ex
and u2 = − x
dx = −
f (x)

  1 + ex  
Whence, yP.I.(x) = log  x  − e−x  ex +  − log (1 + ex ) e−x
  e  
And the complete solution becomes

 1 + ex 
y(x) = yC.F. + yP.I. = ( c1ex + c2 e−x ) + ex log  x  − e−x log (1 + ex ) − 1.
 e 

1
Example 25: It is given that y1 = x and y2 = are two linear independent solution of
x
d2 y dy
the associated homogeneous equation x2 2
+x − y = x, x ≠ 0. Find particular integral
dx dx
and the general solution. [NIT Kurukshetra, 2010]

d2 y 1 dy y 1
Solution: Rewrite the given equation as + − = , x ≠ 0. … (1)
dx2 x dx x2 x
1
Further, we are given yC.F.(x) = c1y1 + c2 y2 = c1x + c2 … (2)
x
By method of Variation of Parameter,
1
yP.I .(x) = u1(x) y1 + u2 (x) y2 = u1x + u2 … (3)
x

1
1 x
The Wronskian of y1 = x and y2 = is W * = x =−2, x≠0
x 1 x
1 − 2
x
u1′ (x) = −y2  ∗  and u2′ (x) = y1  ∗  with X =
X X 1
W  W  x
11 x
∫ x  x ⋅ −2  dx = 2 log x
1
Implying u1(x) = −
608 Engineering Mathematics through Applications

u2 (x) = x  ⋅
1 x  
∫
x2
dx = −

 x −2  4  … (4)

 −x2   1 
yP.I .(x) =  log x x +  
1
Whence 2   4   x 
And complete solution,

y(x) =  c1x + c2  +

1
x
x
2{log x −
x
4 }
= c1∗x + 2 + log x, c1 * =  c1 −  .
c
x 2
x

1
4

II Method Of Undetermined Coefficients:

Under this method which is applicable only to equations with constant coefficients, we find yP.I.(x)
of f(D)y = X(x) by assuming a trial solution containing unknown constants which are determined
by substituting it and its derivatives in the given equation. Here, assumed yP.I.(x) depends
on the form of X(x) in each case. In cases, when the right hand side of the equation viz. X(x)
is of the form containing exponential, polynomial, cosine and sine functions, sums or product of
these functions, then particular integral can easily be found by this method. The accompanying
table illustrates the procedure as:

X(x) Assumed yP.I.(x)

eax A eax
cos βx or sin βx (A sin βx + B cos βx)
xn (A0xn + A1xn – 1 + …… + An – 1x + An)
eaxxn eax(A0xn + A1xn – 1 + …… + An – 1x + An)
eax cos βx or eax sin βx eax(A sin βx + B cos βx)

This method holds so long as no term of the assumed Particular Integral appears in
Complementary Function. If any term of the assumed expression (P.I.) is present in the
complementary function, then multiply the assumed particular integral by x repeatedly until
no terms of product is present in the complementary function. It fails, when X(x) = tanx,
sec x, cot x, since the differentiation of X(x) results in infinite number of terms.

Example 26: Using the method of undetermined coefficients, solve the differential
equation: (D2 – 2D + 3)y = x3 + cos x

Solution: The corresponding auxiliary equation is

2 ± 4 − 12
D2 − 2D + 3 = 0 ⇒ D = i.e. D = 1 ± i 2
2

∴ C.F. = ex (c1 cos 2 x + c2 sin 2 x ) … (1)

Supposing the particular integral
y = ( a1x3 + a2 x2 + a3 x + a4 ) + ( a5 cos x + a6 sin x ) … (2)

= ( 3a1x2 + 2a2 x + a3 ) + ( − a5 sin x + a6 cos x ) ,

dy
So that … (3)
dx
 Linear Differential Equations of Second and Higher Order 609

d2 y
= ( 6a1x + 2a2 ) + ( − a5 cos x − a6 sin x ) , … (4)
dx2
dy d2 y
On substituting above values of y, and 2 in the given equation, we get
dx dx

( 6a1x + 2a2 ) + ( −a5 cos x − a6 sin x ) − 2 ( 3a1x2 + 2a2 x + a3 ) + ( −a5 sin x + a6 cos x )

+3 ( a1x3 + a2 x2 + a3 x + a4 ) + ( a5 cos x + a6 sin x ) = x3 + cos x … (5)

Comparing the coefficients of x3, x2, x, constant term, cosx and that of sinx on both sides of
equation (5) we get

1
3a1 = 3a2– 6a1 = 0 6a1 – 4a2 + 3a3 = 0 2a2 – 2a3 + 3a4 = 0 2a5 – 2 a6 = 1,
3 

1 2 6 8 4 4
⇒ a1 = ⇒ a2 = ⇒ − + 3a3 = 0 ⇒ − + 3a4 = 0 a5 + a6 = 0

3 3 3 3 3 9
2 8 1 1
i.e. a3 = i.e. a4 = − i.e. a5 = , a6 = −
9 27 4 4

With the above values of a1, a2, a3, a4, a5 and a6, particular integral becomes

 x2 2 2 2 8
+ x + x −  +  cos x − sin x 
1 1
y ( P.I.) = 
 3 3 9 27   4 4 

=
1
27
( 9x2 + 18x2 − 6x − 8) + 14 (cos x − sin x ) … (6)

Hence the complete solution,

y = ex (c1 cos 2x + c2 sin 2 x) +

1
27
( 9x2 + 18x2 − 6x − 8) + 41 (cos x − sin x)

Example 27: Solve by Method of undetermined, the equation (D2 – 2D)y = ex sinx
[VTU, 2006; NIT Kurukshetra, 2009]

Solution: Here yC.F.(x) = (c1 + c2 e2x) … (1)

Assume yPI(x) = ex(A cos x + B sin x) … (2)

dy
Dy = = ex ( −A sin x + B cos x ) + ex ( A cos x + B sin x ) … (3)
dx

= ex ( A + B) cos x + ( B − A) sin x 

d2 y
D2 y = = ex  −(A + B)sin x + (B − A) cos x + ex ( A + B) cos x + (B − A) sin x
dx2
610 Engineering Mathematics through Applications

= ex(2B cos x – 2A sin x) … (4)

On substituting values D2y, Dy and y in the given equation, we get
ex ( 2B cos x − 2A sin x ) − 2ex ( A + B) cos x + (B − A) sin x  = ex sin x … (5)
Comparing the coefficient of excosx, ex sin x on both sides,
2B − 2 ( A + B ) = 0 ⇒ A=0 

1
and − 2 A − 2 (B − A ) = 1 ⇒ B=−  … (6)
2
1
Whence, yPI (x) = − ex sin x and complete solution is
2

y(x) = ( c1 + c2 e2 x ) −
1 x
e sin x
2

Example 28: Solve (D2 – 2D + 1)y = ex by method of undetermined coefficients.

[NIT Kurukshetra, 2006]

Solution: Here yC.F. = c1ex + c2 xex Since the right hand member of the complete equation is
X(x) = ex, therefore the first possible assumption for particular integral would be yP.I. = Aex,
but this solution has already occurred in yC.F. We, therefore, yP.I.(x) = A x ex. Still this solution
also has appeared in yC.F. We, therefore, again multiply by x and try yP.I.(x) = Ax2ex
dy
Thus = A  2xex + x2 ex  = A(2x + x2 )ex
dx
d2 y
And = A ( 2ex + 2xex ) + ( 2xex + x2 ex ) = A ( x2 + 4x + 2 ) ex
dx2
On substituting values of D2y, Dy and y in the given equation, we get
A(x2 + 4x + 2)ex – 2A(2x + x2)ex + Ax2ex = ex

or 2A ex = ex implying A = 1 (ex ≠ 0)
2
1
Whence, yP.I .(x) = x2 ex
2
1 2 x
and complete solution becomes y(x) = c1ex + c2e−x + xe .
2

III The Complete Solution in Terms of a known Integral

OR
Solution: By Method of Reducing the Order of the Equation
If an integral included in the complementary function of a second order equation,
d2 y dy
2
+ P(x) + Q(x) y = X … (1)
dx dx
 Linear Differential Equations of Second and Higher Order 611

be known, then the complete solution can be found.

Let y = u(x) be a known solution (i.e. complementary function) of the homogeneous
equation corresponding to (1).
d2 u du
∴ + P(x) + Q(x) u = 0 … (2)
dx2 dx
Now, let y = u(x) v(x) be the general solution of (1), so that

dy dv du d2 y d2 v d2 u du dv
=u +v and 2
= u 2 + v 2 +2 … (3)
dx dx dx dx dx dx dx dx
dy d2 y
On substituting these values of , into equation (1) and re-arranging the terms,
dx dx2

d2 v  du  dv  d2 u du 
we get. u +  Pu + 2  + 2 +P + Qu v = X
dx 2  dx dx  dx dx 

d2 v 
+  Pu + 2 
du dv
or u = X , using (2)
dx2  dx  dx
d2 v  2 du  dv X
or + P +  = … (4)
dx 2  u dx  dx u
dv
If we take = p, equation (4) reduces to
dx
dp  2 du  X
+ P +  p = … (5)
dx  u dx u
Clearly, this equation being a Leibnitz Linear differential equation of 1st order can be
dv
solved for ‘p’, where p = which on integration w.r.t. x results in an expression for v.
dx
∴ y = uv gives a solution of equation (1).

d2 y dy
Note: Complete solution corresponding to equation + P(x) + Q(x)y = X(x) can be found by the
dx2 dx
following rules:
1. y = x is the part of the C.F. if P + xQ = 0
2. y = x2 is the part of the C.F. if 2 + 2xP + x2Q = 0
3. y = ex is the part of the C.F. if 1+P+Q=0
4. y = e–x is the part of the C.F. if 1–P+Q=0
P Q
5. y = eax is the part of the C.F. if 1 + + 2 = 0
a a
Example 29: Solve y” – 2y’ + y = ex sin x given that y = ex is a solution corresponding to
homogeneous equation.
612 Engineering Mathematics through Applications

d2 y dy
Solution: The given equation is comparable to + P( x) + Q(x)y = X with P= – 2, Q = 1
dx2 dx
so that 1 + P + Q = 0, means y = ex is a part of the complementary function. Let y = v ex be a
solution of the given equation.
dy dv d2 y x dv
2
x d v
Thus, = v ex + ex , = v e x
+ 2 e + e
dx dx dx2 dx dx2
dy d2 y
On using these values of and into the given equation,
dx dx2

 x d2 v 
x dv
+ ex 2  − 2  vex + ex  + vex = ex sin x
dv
 ve + 2e 
dx dx  dx 
d2 v d2 v
On simplification, = ex
e x
sin x or = sin x (ex ≠ 0)
dx2 dx2
dv
Which on integration implies, = − cos x + c1
dx
Further, v = – sin x + c1x + c2
∴ The complete solution, y = ex(– sin x + c1x + c2)

Example 30: Apply Method of Variation of Parameter to solve the equation

d2 y
− y = (1 − x2 )
dy
(1 − x ) 2
+x
dx dx

Solution: Rewrite the given equation as

d2y x dy 1
+ − y = (1 + x )
dx2
(1 − x ) dx (1 − x ) … (1)

x 1
Here P + xQ = 0, with P = and Q = − ; means y = x is a part of the
1−x 1−x
complementary function. Thus for finding complementary function of the equation (1),
Put y = vx so that equation (1), reduces to

d2 v  x 2  dv
+ +  =0
 (1 − x ) x  dx
dx 2 … (2)

dp  −x
−  p = 0,
2 dv
or − where p=
dx  1 − x x  dx

dp 
−  dx
dp 
−  p = 0
1 2 1 2
− 1 + ⇒ = 1 +
or
dx  1 − x x p  x − 1 x
(case of variable-separable)
On integration, log p = x + log(x – 1) – 2log x + log c1
 Linear Differential Equations of Second and Higher Order 613

c1 ( x − 1) ex
implying p= …(3)
x2

dv  ex ex 
i.e. = c1  − 2  again a case of variable-separable, …(4)
dx x x 

 ex ex  c 
v = c1 
 x ∫ x ∫
dx − 2 dx + c2 =  1 ex + c2 
x  … (5)

∴ y(C.F) = vx =(c1ex + c2x) … (6)

Now let, y(P.I.) = u1y1 + u2y2,
where u1 and u2 are functions of x and y1 = ex, y2 = x.
 X  X 
Now under this method u1′ = −y2 and u2 ′ = y1
 W∗   W∗ 

y1 y2 ex x
where W* = = x = ex(1 − x); and X = (1 − x)
y1' y2' e 1

(1 − x ) dx = xe−xdx = x + 1 e−x
u1 = − y2  ∗  dx = − x ⋅ x
∫ ∫ ∫
X
∴ ( )
W  e (1 − x ) … (7)

(1 − x ) dx = x
u2 = y1  ∗  dx = ex ⋅ x
∫ ∫
X
And W  e (1 − x ) … (8)

∴ y(P.I.) = u1y1 + u2y2 = (x + 1)e–xex + x2 = (1 + x + x2)

And complete solution y = (c1ex + c2x) + (1 + x + x2)

d2 y dy
Example 31: Solve 2
+ ( 1 − cot x ) − y cot x = sin2 x.
dx dx

Solution: The given equation is comparable to

d2 y dy
2
+P + Qy = sin2 x … (1)
dx dx
Here, P = 1 – cot x, Q = – cot x and 1 – P + Q = 0.
Therefore, y = ve – x is complementary solution of the equation,
d2 y dy
2
+ (1 − cot x ) − y cot x = 0 … (2)
dx dx


= e− x  − v
dy dv
Take, y = ve– x so that 
dx  dx 

−x  d v 
2 2
dy dv
= e  − 2 + v
dx2  dx2 dx  
614 Engineering Mathematics through Applications

By this substitution, the given equation reduces to

d2 v dv dz dv
= (1 + cot x ) or = (1 + cot x ) z for =z … (3)
dx2 dx dx dx

∫z = ∫ (1 + cot x ) dx or log (c1z ) = x + log sin x

dz
Implying
dv
or c1 = (ex sin x) ... (4)
dx
ex c
Integrating, c1 ∫ dv = ∫ ex sin x dx or v = (sin x − cos x ) + 2 … (5)
2c1 c1

y  ex c 
∴ −x
= (sin x − cos x) + 2  , as y = ve– x
e  2c1 c1 

1 c
or yC.F. = a1(sin x – cos x) + a2 e – x; a1 = , a2 = 2 …(6)
2c c1
Let yP.I.(x) = u1y1 + u2y2, where u1 and u2 are functions of 'x'
(i.e. replace arbitrary constants in C.F by u1(x) and u2(x).)
Then by method of variation of parameter,
 X ; ′ =  X ,
u1′ = −y2 u y1 ... (7)
 W∗  2  W∗ 

sin x − cos x e−x

W∗ = = − 2 e−x sin x and X = sin2x …(8)
cos x + sin x −e−x

∫ ∫ ∫
X −x sin2 x 1 cos x
Now u1 = − y2 ∗
dx = − e −x
dx = sin x dx = … (9)
W − 2 e sin x 2 2

∫ ∫ (sin x − cos x ) −2 e
X sin 2 x
And, u2 = y1 dx = dx
W∗ − x sin
x

e ( sin2 x − sin x cos x ) dx

∫
1 x
=−
2
(1 − cos 2x ) ex dx + 1
∫ ∫ e sin 2x dx
1
=− x
2 2 4

ex
(cos 2x + 2 sin 2x ) − e (2 cos 2 x − sin 2x) − −e
x x
= …(10)
20 20 4
On substituting values of u1, u2, and y1, y2 particular integral becomes.

( sin x − cos x ) + e−x  e (cos 2x + 2sin 2x ) − ( 2 cos 2x − sin 2x ) − −e 

cos x x
ex x
yP.I . =
2  20 20 4 
and, y(complete) = yC.F. + yP.I.
 Linear Differential Equations of Second and Higher Order 615

ASSIGNMENT 3
1. Using Method of Variation of Parameter, solve the following equations
d2 y
(i) + y = cos ec x2 ; (ii) (D2 – 2D + 2)y = ex cosec x;
dx2
(iii) (D2 − 4D + 4) y = 8x2ex sin 2x; (iv) y” – 2y’ + 2y = ex tan x;

x + 3x) e3x
(v) (D2 – D – 2)y = e(e ; (vi) y" – 6y' + 9y = ; [PTU, 2006]
x2
2. Apply Method of variation of parameter to the following equations:
(i) x2y” – 4xy’ + 6y = x2 log x (ii) x2y2 + xy1 – y = x2 ex
[Hint: Either follow example 27 or reduce the given Cauchy equation with variable
coefficient to equations with constant coefficients by putting x = et and then apply
Method of variation].
3. Solve by Method of variation of parameter, the equation
d2 y dy
x2 2
− 2x (1 + x ) + 2 (1 + x ) y = x3
dx dx
[Hint: y = x is a part of C.F., take y = vx. as y (C.F.)]
4. Using method of undetermined coefficients, solve the following equations:
d2 y
(i) (D2 – 3D + 2)y = 10; (ii) + y = x;
dx2
d2 y dy d2 y dy
(iii) 2
− 2 + y = − 4 ex
; (iv) 2
+4 + 4y = xex + sin x;
dx dx dx dx
5. Use Method of undetermined coefficient to find solution of
(i) (D2 – 3D + 2)y = x2 + ex (ii) (D2 – 1)y = x sin 3x + cos x
[KUK, 2008]
2
dy dy
6. Solve − cot x − (1 − cot x ) y = ex sin x; by reducing the order of integration .
dx2 dx
[Type III, complete solution in terms of known integral.]

9.7 SOLUTION OF EULER–CAUCHY AND LEGENDRE: LINEAR EQUATIONS

These are two special type of homogeneous linear differential equations with variable coefficients
which on application of certain transformation reduce to homogeneous linear differential
equations with constant coefficients. Cauchy–Euler equation is more commonly known as
Cauchy’s equation. The details of the two are as follows:

I Cauchy’s Linear Equation

The differential equation of the form
n −1
dn y n −1 d y dy
K0 xn n
+ K1 x n −1
+ …… + Kn −1x + Kn y = X … (1)
dx dx dx
616 Engineering Mathematics through Applications

where K0, K1, ……, Kn are constants and X is a function of x only, known as Cauchy's linear
equation. Such equations can be reduced to linear differential equation with constant
coefficients by putting.
dt 1
x = et or t = logx so that =
dx x
dy dy dt dy 1 dy dy d
= = or x = = Dy. if = D.
dx dt dx dt x dx dt dt

d2 y d dy d  1 dy  1 dy 1 d dt  dy  1 dy 1 d2 y
Again, = =  =− 2 +  =− 2 +
dx 2
dx dx dx  x dt  x dt x dt dx  dt  x dt x2 dt2

d2 y
or x2 = D2 y − Dy = D (D − 1) y and so on.
dx2
By substituting all these values in (1), we obtain linear equation with constant coefficients,
which has already been discussed.

Example 32: Solve the differential equation [Raipur, 2004]

d3 y d2 y
+ 8y = 65 cos (log x ) ,
dy
x3 3
+ 3x2 2 + x x>0
dx dx dx

d3 y 2
2 d y
+ 8y = 65 cos ( log x ) , x > 0
dy
Solution: The equation 3
+ 3 x x3 2
+x represents
dx dx dx
Cauchy's Homogeneous Linear Equation. In this case, we substitute, x = et
dy d2 d
so that x= Dy ; x2 2 y = D (D − 1) y and so on. for = D;
dx dx dt
The given equation becomes,
D(D – 1)(D – 2)y + 3D(D – 1)y + Dy + 8y = 65cos t
⇒ [D3 + 8]y = 65 cos t
A.E. D3 + 8 = 0 ⇒ D = – 2, 1 ± 3i

(
Hence the complementary function (C.F) = c1e−2t + et c2 cos 3t + c3 sin 3t )
1 1
Particular Integral, P.I. = 65 cos t = 65 cos t
(D + 8 )
3
−D + 8

1 8+D
= 65 cos t = 65 cos t
8−D 64 − D2
(8 + D)cos t
= 65 = (8 cos t − sin t), t = log x
64 + 1
 Linear Differential Equations of Second and Higher Order 617

d2 y dy
+ 4y = ( 1 + x ) .
2
Example 33: Solve x2 − 3x [SVTU, 2007]
dx2 dx

Solution: Here we substitute x = et so that log x = t dt 1

and =
dx x
d dy d2 y
If = D; x = Dy, x2 2 = D (D − 1) y and so on.
dt dx dx
Therefore the given equation reduces to [D(D – 1) – 3D + 4]y = (1 + et)2
or (D2 – 4D + 4)y = 1 + 2et + e2t. …(1)
A.E. D2 – 4D + 4 = 0 i.e. D = 2, 2 so that
yC.F.(x) = (c1 + c2t)e2t = (c1 + c2 log x)x2 …(2)

and yPI (x) =

1
D − 4D + 4
2
(1 + 2et + e2t ) = PI1 + PI2 + PI3
1 1
PI1 = e0t = , (replace D by 0)
D2 − 4D + 4 4
1 1
PI2 = 2 2 e = 2
t
et = 2 et = 2x.
(D − 2) (1 − 2)2
t2 e2t ( log x ) x
2
2
1 1
PI3 = e 2t
= t e 2t
= = .
(D − 2 )2 2 (D − 2 ) 2 1 2
Whence complete solution becomes,

y(x) = yC.F. + yP.I. = ( c1 + c2 log x ) x2 + + 2x + ( log x )

1 x2 2

4 2
d2 y
+ y = ( log x ) sin(log x), x > 0
dy
Example 34: Solve x2 2
+x
dx dx
[KUK 2006, Madras 2006; Kerala, 2005; Raipur, 2005]

Solution: This is a Cauchy's linear homogeneous equation, where

dy d2 y d
x = Dy, x2 2
= D (D − 1) y, … with D =
dx dx dt
Given equation reduces to
[D(D – 1) + D + 1]y = t sin t or (D2 + 1)y = t sin t …(1)
∴ yC.F.(x) = (c1 cost + c2 sint) …(2)
1 1 1
yP.I .(x) = t sin t = Imaginary 2 teit = Imaginary eit t
And, D2 + 1 D +1 ( i )2 + 1
D +
618 Engineering Mathematics through Applications

−1
1 eit 1  D
= Img. eit t = Img.  1 +  t
2iD  1 + 
D 2i D  2i
 2i 

eit 1  1 i eit  t − 1 1
= Img.  t −  = Img.  
2i D 2i −2 D D 2i 

i eit  t2 i2 t  i eit  t2 it 
= Img  +  = Img  + 
− 2 2 2i −2  2 2 
i ( cos t + i sin t ) 2
= Img
−4
(t + it )
t
( sin t − t cos t )
= …(3)
4
Therefore, complete solution
y(x) = yCF + yPI = (c1 cos t + c2 sin t ) +
t
(sin t − t cos t ) , t = log x …(4)
4

II Legendre's Linear Differential Equation:

The differential equations of the form
n −1
dn y n−1 d y dy
K0 ( ax + b ) ( ) + ……… + Kn −1 ( ax + b)
n
n
+ K1 ax + b n −1
+ Kn y = X
dx dx dx
where K0, K1, …, Kn are constants and X is a function of x only, known as Legendre's Equation.
Such equation can be reduced to linear differential equation with constant coefficients by
putting
dt a
(ax + b) = et or t = log(ax + b) so that = .
dx ax + b
dy dy dt dy a dy dy d
= = or ( ax + b) =a = aDy, if =D
dx dt dx dt ax + b dx dt dt
d2 y d dy d  a dy  a2 dy a d dt  dy 
Again, = =   = − +  
dx 2
dx dx dx ax + b dt ( ax + b) dt ax + b dt dx  dt 
2

a2 dy a2 d2 y
=− +
( ax + b) dt ( ax + b ) dt2
2 2

d2 y 2d y dy 
2
or ( ax + b)2 = a  −  = a2D(D – 1)y and so on
dx 2  dt 2
dt 
By substituting all these values in (2), we obtain linear equation with constant coefficients
which has already been discussed.
 Linear Differential Equations of Second and Higher Order 619

d2 y
+ y = 4 cos ( log ( 1 + x ) )
dy
Example 35: Solve ( 1 + x )2 2
+ (1 + x )
dx dx
[MDU, 2005; NIT Kurukshetra, 2010]

Solution: As the given equation is Legendre's Linear equation. Here we take (1 + x) = et

d 
(1 + x ) y = Dy,

dx
2 d
2 
(1 + x ) 2 y = D (D − 1) y,  where D=
d
dx  dt
3
(1 + x )2 d 3 y = (D )( D − 1)(D − 2 ) y, … so on 
dx 

The given equation reduces to

D(D – 1)y + Dy + y = 4 cost ⇒ (D2 + 1)y = 4 cos t, t = log(1+ x)
A.E. D2 + 1 = 0 ⇒ D=±i
∴ C.F. = (c1cos t + c2sin t),
1 1 D D
P.I. = 4cos t = 4t cos t = 2t 2 cos t = 2t cos t = 2t sin t
D2 + 1 2D D −1
Complete solution, y = (c1cos t + c2 sin t) + 2t sin t, t = log(1 + x).

Example 36: Solve [(3x + 2)2D2 + 3(3x + 2)D – 36]y = (3x2 + 4x + 1) [Sambalpur, 2002]

Solution: Legendre’s equation ((3x + 2)2D2 + 3(3x + 2)D – 36)y = (3x2 + 4x + 1) … (1)
dt 3
Take (3x + 2) = et so that t = log (3x + 2) and = … (2)
dx 3x + 2
dy dy dt 3 dy 3 d
∴ = = = δy, where δ =
dx dt dx 3x + 2 dt 3x + 2 dt
dy
or ( 3x + 2) = 3δ y … (3)
dx
d2 y d dy d  3 dy   − 9  dy 9 d2 y
= =   = +
dx2 dx dx dx  3x + 2 dt   ( 3 x + 2 )2  dt ( 3x + 2 ) dt2
Similarly .

d2 y  d2 y dy 
or ( 3x + 2 )2 = 9  2 −  = 9δ ( δ − 1) y … (4)
dx 2
 dt dt 
Therefore, the given equation reduces to

9δ (δ − 1) y + 3(3δy) − 36y =
1
3
( 9x2 + 12x + 3)

(9δ2 − 36 ) y = (3
x + 2) − 1
2
Implying
3
620 Engineering Mathematics through Applications

e2t − 1
9 ( δ2 − 4) y = ; as ( 3x + 2 ) = et … (5)
3
A.E. δ2 – 4 = 0 or δ = ± 2 and thus
yC.F.(t) = c1 y1(t) + c2 y2(t) = c1e2t + c2 e–2t

1 e2t − 1 1 1  1 1  1  e2t e0t 

yPI (t) = = .  2 e2t − 2 e0t  = t −
And f (δ) 3 3 9  ( δ − 4) δ − 4  27  2δ −4 

1  t e2t 1  te2t + 1
= + = , t = log(3x + 2)
27  4 4 108

te2t + 1
Complete solution, y(t) = ( c1e + c2 e− ) +
2t 2t
108

d2 y dy
Example 37: Solve ( 2 x + 3 )2 2
− ( 2x + 3 ) − 12y = 6x
dx dx
[VTU, 2003, 2007; Kerala, 2005; KUK, 2005; NIT Kurukshetra, 2008]
Solution: Take 2x + 3 = et, t = log(2x + 3) so that the given equation reduced to
[4D(D – 1) – 2D – 12]y = 3(et – 3) as 6x = 3(2x) = 3(et – 3)
or 2(2D2 – 3D – 6)y = 3(et – 3)
3 ± 57
A.E 2D2 − 3D − 6 = 0 or D= = m1 , m2
4
3 + 57 3 − 57
yCF (t) = c1 em1t + c2 em2t = c1 (et )m1 + c2 (et )m2 = c1(et ) 4 + c2 (et ) 4

3 ( et − 3 )
1
P.I. =
4D − 6D − 12
2

1 1 3 3
=3 et − 9 2 e0t = − et +
4D − 6D − 12
2
4D − 6D − 12 14 4
3 3
y(x) = yCF (x) + yPI (x) = c1 ( 2x + 3 ) 1 + c2 ( 2x + 3 ) 2 − (2 x + 3 ) + .
m m
Whence
14 4

ASSIGNMENT 4
Solve the following equations
d2 y dy d2 y dy
1. (i) x2 2 − 4x + 6y = x2 ; (ii) x2 2
− 2x − 4y = x4 ;
dx dx dx dx
d2 y 2y 1 d2 y 1 dy 12 log x
(iii) x2 2
− =x+ 2; (iv) + = , x > 0;
dx x x dx2 x dx x2

+ 2x2 2 + 2y = 10  x +  ;
d3 y d2 y 1
(v) x3 [SVTU, 2006; PTU, 03]
dx 3
dx  x
 Linear Differential Equations of Second and Higher Order 621

d2 y dy
(vi) x2 − 2x − 4y = x2 + 2 log x ; x > o [Bhopal, 2003; KUK, 2010]
dx2 dx
(vii) The radial displacement u in a rotating disc at a distance r from the axis is given
d2 u du
by r2 2
+r − u + kr3 = 0, where k is a constant. Solve the equation under
dr dr
the conditions u = 0, when r = 0; u = 0 when r = a.

d2 y
+ y = 2 sin (log (1 + x ))
dy
(viii) (1 + x )2 + (1 + x )
dx2 dx
(ix) ((1 + 2x)2D2 – 6(1 + 2x)D + 16)y = 8(1 + 2x)2
(x) ((x + 1)2D2 + (x + 1)D)y = (2x + 3)(2x + 4)
2. Establish the Euler`s Cauchy equation of IIIrd order whose general solution is
y = Ax + Bx2 + Cx3.
3. Establish the Euler-Cauchy equation of IIIrd order whose general solution is
y = Ax + Bx(log x) + Cx(log x)2

9.8 SIMULTANEOUS LINEAR DIFFERENTIAL EQUATIONS WITH CONSTANT

COEFFICIENTS
Linear differential equations consisting of two or more dependent variables and single
independent variable say x(t), y(t) etc. are called simultaneous differential equations. To
solve these type of equations, eliminate one of the dependent variable and then solve the
resulting equation which gives the solution of the dependent variable in terms of the
independent variable. Repeat the process for 2nd dependent variable. In this section, we
shall restrict our discussion to solutions of linear equations with constant coefficients only.

dx dy
Example 38: Solve + 2y = et , − 2 x = e− t .
dt dt

Solution: The given equation in symbolic form can be written as

Dx + 2y = et …(1)
Dy – 2x = e–t …(2)
Operate D on (2) and add to it 2 time of (1), we get
(D2 + 4)y = 2et – e–t … (3)
Here A.E. is D2 + 4 = 0 i.e. D = ± 2i
∴ yC.F.(t) = (c1cos 2t + c2 sin 2t) … (4)
et e−t
And yP.I .(t) =
1
D +4
2
(2 e t − e −t ) = 2 4
1
D +4
et − 2
1
D +4
e −t = 2 −
5 5
… (5)

2 e−t 
Hence y(t) = ( c1 cos 2t + c2 sin t ) +  et −  …(6)
5 5
622 Engineering Mathematics through Applications

1 2 e −t  
Dy − e−t  = D ( c1 cos 2t + c2 sin 2t ) + D  et −  − et 
1
Now from (2), x =
2 2  5 5  

1 2 t e−t 
=  −2c1 sin 2t + 2c2 cos 2t + e + − e−t 
2 5 5 
t
e 2
= −c1 sin 2t + c2 cos 2t + − e −t … (7)
5 5
Example 39: Solve (D – 1)x + Dy = 2t + 1, (2D + 1)x + 2Dy = t

Solution: For elimination of y, take difference of 2 time of 1st from 2nd i.e.
((2D + 1)x + 2Dy) – 2((D – 1)x + Dy) = t – 2(2t + 1)
2 dx
or 3x = – 3t – 2 or x(t) = − t − Implying = −1.
3 dt
dx
Now using above values of x(t) and in Ist equation, we get
dt

− 1 −  −t −  + Dy = 2t + 1
Dx – x + Dy = 2t + 1 or
2
 3
4 t2 4
Implying Dy = t + i.e. y(t) = + t + c
3 2 3
where c is a constant of integration.

dx dy
Example 40: Solve − 7x + y = 0, − 2x − 5y = 0
dt dt
Solution: The given equations in symbolic form are written as:
(D – 7)x + y = 0 …(1)
–2x + (D – 5)y = 0 …(2)
To eliminate y, operate (D–5) on (1) and add the two equations to get
(D – 5)(D – 7)x + 2x = 0 or (D2 – 12D + 37)x = 0 … (3)
So that A.E. is D2 – 12D + 37 = 0 or D = 6 ± i
∴ xC.F.(t) = e6t(c1cost + c2 sin t) … (4)

= e6t ( −c1 sin t + c2 cos t ) + 6 e6t ( c1 cos t + c2 sin t )

dx
Implying … (5)
dt
On substituting values of x(t) and Dx from (4) and (5) respectively in equation (1), we get
e6t(–c1sin t + c2 cos t) + 6 e6t(c1cos t + c2 sin t) – 7e6t(c1cos t + c2 sin t) + y = 0
or y = e6t[(c1 – c2)cos t + (c1 + c2)sin t]
= e6t[C cost + D sin t] … (6)
where, C = c1 – c2 and D = c1 + c2
 Linear Differential Equations of Second and Higher Order 623

Example 41: Solve the simultaneous differential equations

dx dy
t + y = 0, t + x = 0 with conditions: x(1) = 1 and y(–1) = 0
dt dt
dx
Solution: t +y=0 … (1)
dt
dy
t +x=0 … (2)
dt
Differentiating (1) with recept to ‘t’,
d2 x dx dy d2 x dx dy
t 2
+ + = 0 or t2 2 + t +t =0 … (3)
dt dt dt dt dt dt
dy
Substituting the value of t from (2), equation (3) reduces to
dt
d2 x dx
t2 +t −x=0 …(4)
dt2 dt
Which is Cauchy’s homogeneous linear differential equation in t.
dx dx dp 1 dx dx dx
Put t = ep or log t = p so that = = or t = = Dx
dt dp dt t dp dt dp

d2 x
Likewise t2 = D (D − 1) x and so on
dt2
Therefore (4) reduces to, D(D – 1)x + Dx – x = 0 or (D2 – 1)x = 0 …(5)
 1
Whence xC.F.(p) = c1ep + c2 e−p or x(t) =  c1t + c2  … (6)
 t
dx c
Implying = c1 − 22 … (7)
dt t

dx  c   c 
From (1), we get y = −t = −t  c1 − 22  =  −c1t + 2  … (8)
dt  t   t
1
On using the conditions, x (1) = 1 and y (–1) = 0 in (6) & (7), we get c1 = = c2
2

Example 42: To small Oscillations of a certain system with two degrees of freedom are
given by the equations

D2 x + 3x − 2y = 0  d
 , where D =
D x + D y − 3x + 5y = 0
2 2
dt
If x = 0, y = 0; Dx = 0, Dy = 2 when t = 0, find x and y when t = 1/2.

Solution: Re–write the given set of simultaneous equations as

(D2 + 3)x – 2y = 0 …(1)
624 Engineering Mathematics through Applications

(D2 – 3)x + (D2 + 5)y = 0 …(2)

In order to make above equations separately in x(t) and y(t) i.e. to eliminate x, first operate
these equations by (D2 – 3) and (D2 + 3) respectively and then subtract (1) from (2), we get
[(D2 + 3)(D2 + 5) + 2(D2 – 3)]y = 0
On simplification, (D4 + 10D2 + 9)y = 0 … (3)
Corresponding auxiliary equation becomes
D4 + 10D2 + 9 = 0 or (D2 + 1)(D2 + 9) = 0 i.e. D = ± i, ± 3i
Thus, y(C.F.) = (a1cos t + a2 sin t) + (a3 cos 3t + a4 sin 3t) … (4)
To find x, eliminate y from (1) and (2).
Operate (1) by (D2 + 5) and multiply (2) by 2 and then add the two,
(D4 + 10D2 + 9)x = 0 (an equation identical to (3)) i.e. D = ± i, ± 3i ...(5)
Thus, x(C.F.) = (b1 cos t + b2 sint) + (b3cos 3t + b4sin 3t) …(6)
To find the relation between constants involved in (4) and (6)
Substitute values of x and y in either of the given equations, say in (1), we get
2(b1 – a1) cost + 2(b2 – a2) sin t – 2(3b3 + a3) cos 3t – 2(3b4 + a4)sin 3t = 0 …(7)
Which must holds for all t.
On equating co–efficient of cost, sin t, cos 3t, sin 3t, we get
a3 a4
b1 = a1, b2 = a2, b3 = − , b4 = − …(8)
3 3
Thus on replacing bi‘s by respective ai´s, equation (6) becomes

x = ( a1 cos t + a2 sin t ) − (a3 cos 3t + a4 sin 3t )

1
…(9)
3
Clearly, relations (4) and (9) are the solutions of the set of given equation.
In order to find out these constants, make use of the initial conditions.
Using x = y = 0 when t = 0, equation (4) and (9) gives,

a1 + a3 = 0 

implying a1 = a3 = 0.
a1 − a3 = 0
1
3 

With above values of a1 and a3, equation (9) and (4) reduces to

sin 3t 
}
a4
x = a2 sin t − Dx = a2 cos t − a4 cos 3t
3  and Dy = a2 cos t + 3a4 cos 3t …(10)
y = a2 sin t + a4 sin 3t 
With the conditions Dx = 3 and Dy = 2 when t = 0, equations (10) give
3 = a2 − a4
2 = a2 + 3a4 } implying a2 =
11
4
1
, a4 = − .
4
Hence from equation (10),
 Linear Differential Equations of Second and Higher Order 625

x=
1 1 
 11sin t + sin 3t  
4 3

1
y = (11sin t − sin 3t ) 
4 

1
11sin ( 0.5) + sin (1.5 ) = 1.4015 
1 1
t= , x=
4  
Further when
2 3 


1
y = 11sin ( 0.5) − sin (1.5) = 1.069 
4 


ASSIGNMENT 5
Solve the following simultaneous equations:
dx dx
(i) + 5x − 2y = t, [UP Tech, 2008] (ii) + 2x + 3y = 0,
dt dt
dy [KUK, 2005] dy [Delhi, 2002]
+ 2x + y = 0; 3x + + 2y = 2e2t ;
dt dt
dx
(iv) t + y = 0,
(iii) ( D + 1) x + ( 2D + 1) y = e ,
t dt x (1) = 1 
y ( − 1) = 0
dy given
( D − 1) x + (D + 1) y = 1; t + x = 0;
dt
d2 x d2 y
(v) + y = sin t , + x = cos t; [NIT Jalandhar, 2006]
dt2 dt2
dx dy dz
(vi) = 2y, = 2z, = 2x
dt dt dt
(vii) A mechanical system with two degrees of freedom satisfies the equation
d2 y dy d2 y dx
2 2
+ 3 = 4, 2 2
−3 =0
dt dt dt dt
dy dy
Obtain expression for x and y in terms of t, given x, , all vanish at t = 0.
dx dt

ANSWERS
Assignment 1
x  3 3 
(i) y = (c1 + c2x)e3x, (ii) y = e− 2  c1 cos x + c2 sin x
 2 2 

−x  3 3 
(iii) y = c1e + e 2  c1 cos 2 x + c2 sin 2 x
x
626 Engineering Mathematics through Applications

(iv) y = (c1 + c2x) e2x + (c3 + c4x) e–2x

x c1 = 4
(v) y = e ( c1 cos 3x + c2 sin 3x ) , c = −3 ,
2
} (vi) y = c1 + (c2 + c3x)e–x/2

(vii) y = (c1 + c2x) cos nx + (c3 + c4x) sin nx, (viii) θ = α cos g / l

Assignment 2

1. y = (c1 cos ax + c2 sin ax ) +

x
4a2
( cos ax + ax sin ax )

2. (c1ex + c2 e−x ) − 12 ( x2 − 1) cos x + x sin x

3. (c1cosx + c2sinx) + [sin log (sinx) – xcosx]

4. y = (c1 cos x + c2 sin x ) + (c3e + c4 e ) +

x
x −x
( x cos x − 3x sin x )
8

5. y = (c1 cos ax + c2 sin ax ) −

1
cos ax log ( sec ax + tan ax )
a2

ex  1
6. y = ( c1 cos 2x + c2 sin 2x ) + ( 4 sin 2x + cos 2x )
1
 −
2  5 17 

e−x
−x − 3x
7. y = c1e + c2 e − (2 cos x + sin x ) + 1  x − 4 
5 3 3

8. y = ( c1e + c2e− ) − ( x sin x + cos x ) + xe ( x2 − x + 3 )

x
x x 1
2 4

9. y = ( c1e + c2 e ) −
− 2x x 2
2x
sinh x − cosh x
3 7
x x
 x x  −  x x  x2 − 2
10. y = e 2
 c1 cos 2 + c2 sin 2  + e
2
 c3 cos 2 + c4 sin 2  + cos x − 2x sin x
    2

11. y = ( c1e + c2 e ) + (c3 cos x + c4 sin x ) −

x −x 1
cos x cosh x
5

12. y = ( c1e + c2 e ) +
1
x 3x
(10 cos 5x − 11sin 5x ) + 1 (sin x + 2 cos x )
884 20
13. y = (c1 + c2x)ex – ex(x sinx + cosx)

 
ex  2 4x − 2
14. y = (c1 cos nx + c2 sin nx ) +
8
 x + + .
1 + n2  1 + n2 (1 + n2 )2 
 
 Linear Differential Equations of Second and Higher Order 627

Assignment 3
1. (i) y = (c1cos x + c2 sin x) – x cos x + sin x (log sinx)
(ii) y = (c1cosx + c2sinx)ex – xexcosx + ex sin (log sinx)
(iii) y = (c1 + c2x)e2x – e2x[4x cos2x + (2x2 – 3)sin x]
(iv) y = ex(c1cos x + c2sin x) – ex(cos x) log (sec x + tan x)
y = ( c1e2x + c2 e−x ) + ee ( 3ex − 6 + e− x )
x
(v)
(vi) y = (c1 + c2x)e3x – e3x log x

 c2   x ex 
(c1x2 + c2 x3 ) − x2 (log x )2 − x2 (log x)
2
2. (i) (ii)  c1x +  +  e − 
x x

y = ( c1x e2x + c2 x ) −
x2
3.
2
4. (i) y = (c1ex + c2e2x) + 5; (ii) y = (c1cosx + c2sinx) + x;
(iii) y = (c1 + c2x)ex – 2x2λex;

(iv) (c1 + c2x ) e−2x +  19 x − 27

2 x  3
 e +  sin x −
25
4
25
cos x 


1 2 
y = ( c1ex + c2e2x ) +
7
5. (i) x + 3x + − 2xex
2  2 
1  3  cos x
(ii) y=−  x sin 3x + cos 3x −
10 5 2
ex cos x c1e−x
6. y=− − ( 2 sin x + cos x ) + c2ex.
2 5

Assignment 4
c2 x4
1. (i) y = c1x2 + c2x3 – x2 logx; (ii) y = c1x 4 + + log x
x 5
c2 1  2 1 
(iii) y = c1x + +  x −  log x;
2
(iv) y = 2(log x)3 + c1 log x + c2.
x 3 x

+ x c2 cos ( log x ) + c3 sin ( log x ) + 5x + 10

c1 log x
(v) y = ;
x x
c1 x2 1 3
(vi) y = + c2 x4 − − log ( x ) + ;
x 6 2 8

(vii) u = ( a − r )
kr 2 2
8
(viii) y = c1 cos (log(1 + x)) + c2 sin ( log(1 + x)) − log (1 + x ) cos (log (1 + x ))
628 Engineering Mathematics through Applications

(ix) y = (1 + 2x )
2
{ log (1 + 2x) 2
+ (c1 log (1 + 2x )) + c2 }
(x) y = c1 + c2 log ( x + 1) + (log ( x + 1)) + x + 8x 
2 2

2. x3 y ′′′ − 3x2 y ′′ + 6xy ′ − 6y = 0

3. x3 y ′′′ + xy '− y = 0

Assignment 5
1
(i) x = − (1 + 6t ) e−3t + 1 (1 + 3t ) , y = − 2 (2 + 3t ) e−3t + 2 ( 2 − 3t )
27 27 27 27
(ii) x = et + e−t , y = e−t − et + sin t
(iii) x = c1et + c2 e−2t + 2 e−t , y = 3c1 et + 2c2 e−2t + 3 e−t

1 1 1 1
(iv) x = t + , y=  −t + 
2 t 2 t

(v) x = ( c1e + c2e ) + (c3 cos t + c4 sin t ) −

−t t t
t
cos t + sin t,
4 4

y = ( −c1et − c2 e−t ) + (c3 cos t + c4 sin t ) +

1
( 2 + t )(sin t − cos t )
4
2π
(vi) x = c1 e2t + c2 e−t cos ( 3t − c3 ) , y = c1 e2t + c2 e−t cos( 3t − c3 +
3
)

4π
z = c1 e2t + c2 e−t cos( 3t − c3 + )
3
8 3  4 8 3
(vii) x = 
1 − cos t , y = t − sin t

9 2 3 9 2