5

Laplace Transforms
5.1 Introduction................................................................................................................................... 5-1
5.2 Laplace Transform of Some Typical Functions .................................................................... 5-2
5.3 Properties of the Laplace Transform ....................................................................................... 5-3
5.4 The Inverse Laplace Transform .............................................................................................. 5-10
5.5 Solution of Ordinary Linear Equations with Constant Coefficients .............................. 5-13
Alexander D. Poularikas 5.6 The Inversion Integral............................................................................................................... 5-17
University of Alabama in Huntsville 5.7 Applications to Partial Differential Equations..................................................................... 5-20
5.8 The Bilateral or Two-Sided Laplace Transform.................................................................. 5-27
y
Samuel Seely References ................................................................................................................................................ 5-43

5.1 Introduction* is called the Laplace transformation of f(t). Here s can be either a
real variable or a complex quantity. Observe the shorthand nota-
The Laplace transform has been introduced into the mathemat- tion +{f(t)} to denote the Laplace transformation of f(t). Observe
ical literature by a variety of procedures. Among these are: (a) in also that only ordinary integration is involved in this integral.
its relation to the Heaviside operational calculus, (b) as an exten- To amplify the meaning of condition (2), we consider piece-
sion of the Fourier integral, (c) by the selection of a particular wise continuous functions, defined for all positive values of the
form for the kernel in the general Integral transform, (d) by a variable t, for which
direct definition of the Laplace transform, and (e) as a math-
ematical procedure that involves multiplying the function f(t) by lim f (t)ect ¼ 0, c ¼ real constant:
est dt and integrating over the limits 0 to 1. We will adopt this t!1

latter procedure.
Not all functions f(t), where t is any variable, are Laplace Functions of this type are known as functions of exponential
transformable. For a function f(t) to be Laplace transformable, order. Functions occurring in the solution for the time response
it must satisfy the Dirichlet conditions—a set of sufficient but not of stable linear systems areÐ of exponential order zero. Now we
1
necessary conditions. These are can recall that the integral 0 f (t)est dt converges if

1. f(t) must be piecewise continuous; that is, it must be single ð

1
valued but can have a finite number of finite isolated j f (t)est jdt < 1, s ¼ s þ jv
discontinuities for t > 0.
0
2. f(t) must be of exponential order; that is, f(t) must remain
less than Meao t as t approaches 1, where M is a positive
If our function is of exponential order, we can write this integral as
constant and ao is a real positive number.
2
For example, such functions as: tan bt, cot bt, et are not Laplace ð
1
transformable. Given a function f(t) that satisfies the Dirichlet j f (t)ject e(sc)t dt:
conditions, then 0

ð
1
This shows that for s in the range s > 0 (s is the abscissa of
F(s) ¼ f (t)est dt written +{f (t)} (5:1) convergence) the integral converges; that is
0
ð
1

j f (t)est jdt < 1, Re(s) > c:

0

* All the contour integrations in the complex plane are counterclockwise.

5-1

© 2010 by Taylor and Francis Group, LLC

 5-2 Transforms and Applications Handbook

1
Abscissa of convergence To carry out the integration, define the quantity x ¼ t 2 , then
dx ¼ 12 t 2 dt, from which dt ¼ 2t 2 dx ¼ 2x dx. Then
1 1
jω
σ + jω
ð
1
4
x2 esx dx:
2

Region of absolute F(s) ¼ pffiffiffiffi

p
S-plane convergence 0

But the integral

σ
0 c σ
ð
1 pffiffiffiffi
p
x2 esx dx ¼
2
:
4s3=2
0
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Thus, finally,
σ – jω

1
F(s) ¼ : (5:4)
s3=2
FIGURE 5.1 Path of integration for exponential order function.

The restriction in this equation, namely, Re(s) ¼ c, indicates that Example 5.3
we must choose the path of integration in the complex plane as
shown in Figure 5.1. Find the Laplace transform of f (t) ¼ erfc 2pk ffit, where the error
function, erf t, and the complementary error function, erfc t,
are defined by
5.2 Laplace Transform of Some
Typical Functions ðt ð
1
2 2
eu du, eu du,
2 2
erf t ¼ pffiffiffiffi erfc t ¼ pffiffiffiffi
We illustrate the procedure in finding the Laplace transform of a p p
0 t
given function f(t). In all cases it is assumed that the function f(t)
satisfies the conditions of Laplace transformability.
SOLUTION
Example 5.1
Consider the integral
Find the Laplace transform of the unit step function f(t) ¼ u(t), 2 3
where u(t) ¼ 1, t > 0, u(t) ¼ 0, t < 0. ð
1 ð
1
2 6 u2 7 k
I ¼ pffiffiffiffi est 6 7
4 e du5dt where l ¼ : (5:5)
p 2
SOLUTION 0 plffi
t

By Equation 5.1 we write

Change the order of integration, noting that u ¼ plffit , t ¼ lu2
2

ð
1 ð
1 1
st est  1 2 3
+{u(t)} ¼ u(t)e dt ¼ est dt ¼ ¼ : (5:2)
s 0 s 2
ð
1 ð
1
6 st 7 2
ð
1 
l2 s

u2 6 7
0 0 I ¼ pffiffiffiffi e 4 e 5dtdu ¼ spffiffiffi
ffi exp u2  2 du
p p u
0 l2 0
The
Ð 1 region ofÐ convergence is found from the expression u2
st 1
0 je jdt ¼ 0 est dt < 1, which is the entire right half-
plane, s > 0. The value of this integral is known
pffiffiffiffi
2 p 2lpffis
Example 5.2 ¼ p ffiffiffi
ffi  e ,
s p 2
qffiffiffi
Find the Laplace transform of the function f (t) ¼ 2 t
p which leads to
ð
1  
2 k 1 pffiffi
t 2 est dt:
1
F(s) ¼ pffiffiffiffi (5:3) + erfc pffiffi ¼ exp {k s}: (5:6)
p 2 t s
0

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-3

Example 5.4 Re(s) > sg :

Find the Laplace transform of the function f(t) ¼ sinh at. Thus,

SOLUTION +{ f (t) þ g(t)} ¼ F(s) þ G(s):

Express the function sinh at in its exponential form
As a direct extension of this result, for K1 and K2 constants,
at
e e
at
sinh at ¼ : +{K1 f (t) þ K2 g(t)} ¼ K1 F(s) þ K2 G(s): (5:10)
2

The Laplace transform becomes

ð
1
1  (sa)t
 e(sþa)t dt THEOREM 5.2 Differentiation
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

+{ sinh at} ¼ e
2
0
a Let the function f(t) be piecewise continuous with sectionally
¼ : (5:7)
s2  a2 continuous derivatives df(t)=dt in every interval 0  t  T. Also
let f(t) be of exponential order ect as t ! 1. Then when Re(s) > c,
A moderate listing of functions f(t) and their Laplace trans-
the transform of df(t)=dt exists and
forms F(s) ¼ +{f(t)} are given in Table A.5.1.
 
df (t)
5.3 Properties of the Laplace Transform + ¼ s+{ f (t)}  f (0þ) ¼ sF(s)  f (0þ): (5:11)
dt
We now develop a number of useful properties of the Laplace
transform; these follow directly from Equation 5.1. Important in Proof Begin with Equation 5.8 and write
developing certain properties is the definition of f(t) at t ¼ 0, a
quantity written f(0þ) to denote the limit of f(t) at t approaches   ðT
zero, assumed from the positive direction. This designation is df (t) df (t) st
+ ¼ lim e dt:
consistent with the choice of function response for t > 0. This dt T!1 dt
0
means that f(0þ) denotes the initial condition. Correspondingly,
f (n) (0þ) denotes the value of the nth derivative at time t ¼ 0þ, Write the integral as the sum of integrals in each interval in
and f (n) (0þ) denotes the nth time integral at time t ¼ 0þ. This which the integrand is continuous. Thus, we write
means that the direct Laplace transform can be written
ðT ðt1 ðt2 ðT
ðR est f (1) (t)dt ¼ [ ] þ [ ] þ    þ [ ]:
F(s) ¼ lim f (t)est dt, R > 0, a > 0: (5:8)
R!1 0 0 t1 tn1
a!0þ a

Each of these integrals is integrated by parts by writing

We proceed with a number of theorems.
u ¼ est du ¼ sest dt
df
THEOREM 5.1 Linearity dv ¼ dt v ¼ f
dt

with the result

The Laplace transform of the linear sum of two Laplace trans-
formable functions f(t) þ g(t) with respective abscissas of conver-
t1 t 2 T ðT
gence sf and sg, with sg > sf, is   
est f (t) þ est f (t) þ    þ est f (t) þ s est f (t)dt:
0 t1 tn1
0
+{ f (t) þ g(t)} ¼ F(s) þ G(s): (5:9)
But f(t) is continuous so that f(t1  0) ¼ f(t1 þ 0), and so forth,
Proof From Equation 5.8 we write hence

ð
1 ð
1 ð
1 ðT ðT
st (1) sT
+{ f (t) þ g(t)} ¼ [f (t) þ g(t)]e st
dt ¼ f (t)e st
dt þ g(t)e st
dt, e f (t)dt ¼ f (0þ) þ e f (T) þ s est f (t)dt:
0 0 0 0 0

© 2010 by Taylor and Francis Group, LLC

 5-4 Transforms and Applications Handbook

However, with limt!1 f(t)est ¼ 0 (otherwise the transform Proof Because f(t) is Laplace transformable, its integral is written
would not exist), then the theorem is established. 8 t 9 12 t 3
< ð = ð ð
+ f (j)dj ¼ 4 f (j)dj5est dt:
: ;
1 0 1
THEOREM 5.3 Differentiation
This is integrated by parts by writing
Let the function f(t) be piecewise continuous, have a continuous
derivative f (n1)(t) of order n  1 and a sectionally continuous ðt
derivative f (n)(t) in every finite interval 0  t  T. Also, let f(t) u¼ f (j)dj du ¼ f (j)dj ¼ f (t)dt
and all its derivatives through f (n1)(t) be of exponential order ect 1
as t ! 1. Then the transform of f (n)(t) exists when Re(s) > c and 1
dv ¼ est dt v ¼  est :
it has the following form: s
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Then
+{ f (n) (t)} ¼ sn F(s)  sn1 f (0þ)  sn2 f (1) f (0þ)
8 t 9 2 31
     snn f (n1) (0þ): (5:12) < ð = st ð
t  ð
1
e  1
+ f (j)dj ¼ 4 f (j)dj5 þ f (t)est dt
: ; s  s
1 1 0 0
Proof The proof follows as a direct extension of the proof of
ð
1 ð0
Theorem 5.2. 1 1
¼ f (t)est dt þ f (j)dj
s s
0 1
Example 5.5
from which
Find +{tm} where m is any positive integer.
8t 9
<ð = 1 1
SOLUTION + f (j)dj ¼ F(s) þ f (1) (0þ)
: ; s s
The function f(t) ¼ tm satisfies all the conditions of Theorem 0

5.3 for any positive c. Thus,

where [ f (1)(0þ)=s] is the initial value of the integral of f(t) at
t ¼ 0þ. The negative number in the bracketed exponent indicates
f (0þ) ¼ f (1) f (0þ) ¼    ¼ f (m1) (0þ) ¼ 0 integration.
f (m) (t) ¼ m!, f (mþ1) (t) ¼ 0:
Example 5.6
By Equation 5.12 with n ¼ m þ 1 we have
Deduce the value of +{sin at} from +{cos at} by employing
Theorem 5.4.
+{ f (mþ1)
(t)} ¼ 0 ¼ s mþ1
+{t }  m!:
m

SOLUTION
It follows, therefore, that
By ordinary integration
m!
+{t m } ¼ : ðt
smþ1 sin at
cos ax dx ¼ :
a
0

From Theorem 5.4 we can write, knowing that

THEOREM 5.4 Integration +{ cos at} ¼ s2 þa
s
2.

 
If f(t) is sectionally continuous and has a Laplace transform, then sin at 1
Ðt + ¼ 2 :
the function 0 f (j)dj has the Laplace transform given by a s þ a2

8t 9 so that
<ð =
F(s) 1 (1)
+ f (j)dj ¼ þ f (0þ): (5:13)
: ; s s +{ sin at} ¼
a
:
0 s2 þ a2

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-5

Proof This is a restatement of Theorem 5.6. This theorem is

THEOREM 5.5
often useful for evaluating some types of integrals, and can be
used to extend the table of transforms.
Division of the transform of a function by s corresponds to
integration of the function between the limits 0 and t
Example 5.7
  ðt
1 F(s) Employ Theorem 5.7 to evaluate qF(s)=qs for the function
+ ¼ f (j)dj
s f(t) ¼ sinh at.
0
(5:14)
  ðt ðj
1 F(s) SOLUTION
+ ¼ f (l)dldj
s2
0 0 Initially we establish sinh at

and so forth, for division by sn, provided that f(t) is Laplace ð

1

Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

e at  eat a
transformable. +{ sinh at} ¼ est dt ¼ 2 ¼ F(s):
2 s  a2
0
Proof The proof of this theorem follows from Theorem 5.4.
By Theorem 5.7

ð
1
THEOREM 5.6 Multiplication by t qF(s) qh a i 2as
¼ (t) sinh at est dt ¼ ¼ 2
qs qs s2  a2 (s  a2 )2
0
If f(t) is piecewise continuous and of exponential order, then each
of the Laplace transforms: +{ f(t), +{tf(t), +{t2f(t), . . . is uni- from which
formly convergent with respect to s when s ¼ c, where s > c, and
ð
1
2as
dn F(s) est t sinh at dt ¼ +{t sinh at} ¼ :
+{t f (t)} ¼ (1)
n
: n
(5:15) (s2  a2 )2
dsn 0

Further We can, of course, differentiate F(s) with respect to a. In this

case, Theorem 5.7 does not apply. However, the result is
dn F(s) significant and is
lim ¼ 0, +s!1 {t n f (t)} ¼ 0, n ¼ 1, 2, 3, . . .
s!1 dsn
ð
1
qF(s) q h a i
Proof If follows from Equation 5.8 when this integral is uni- ¼ est (t cosh at)dt ¼ +{t cosh at} ¼
qa qa s2  a2
formly convergent and the integral converges, that 0

s2 þ a2
ð
1 ¼ :
qF(s) (s2  a2 )2
¼ est (t)f (t)dt ¼ +{tf (t)}:
qs
0

Further, it follows that THEOREM 5.8 Complex Integration

ð
1
q2 F(s) f (t)
¼ est (t)2 f (t)dt ¼ +{t 2 f (t)}: If f(t) is Laplace transformable andÐ 1provided that limt!0þ t
qs2 exists, the integral of the function s F(s)ds corresponds to the
0
Laplace transform of the division of the function f(t) by t,
Similar procedures follow for derivatives of higher order.
  ð
1
f (t)
+ ¼ F(s)ds: (5:17)
t
0
THEOREM 5.7 Differentiation of a Transform
Proof Let F(s) be piecewise continuous in each finite interval
Differentiation of the transform of a function f(t) corresponds to and of exponential order. Then
the multiplication of the function by t; thus
ð
1
n
d F(s) F(s) ¼ est f (t)dt
¼ F (n) (s) ¼ +{(t)n f (t)}, n ¼ 1, 2, 3, . . . (5:16)
dsn 0

© 2010 by Taylor and Francis Group, LLC

 5-6 Transforms and Applications Handbook

is uniformly convergent with respect to s. Consequently, we can Now introduce a new variable t ¼ t  l. This converts this equa-
write for Re(s) > c and any a > c tion to the form

ða ða 1
ð ð
1 ð
1

F(s)ds ¼ est f (t)dtds: +{ f (t)u(t)} ¼ esl f (t)u(t)est dt ¼ esl f (t)est dt

s s 0 l 0
sl
¼e F(s)
Express this in the form
because u(t) ¼ 0 for l  t  0.
ð
1 ða ð
1
st f (t) st We would similarly find that
¼ f (t) e dsdt ¼ (e  eat )dt:
t
0 s 0
+{ f (t þ l)u(t þ l)} ¼ esl F(s): (5:19)
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Now if f(t)=t has a limit as t ! 0, then the latter function is

piecewise continuous and of exponential order. Therefore, the Example 5.8
last integral is uniformly convergent with respect to a. Thus, as a
tends to infinity Find the Laplace transform of the pulse function shown in
Figure 5.3.
ð
1  
f (t)
F(s)ds ¼ + : SOLUTION
t
s
Because the pulse function can be decomposed into step func-
tions, as shown in Figure 5.3, its Laplace transform is given by


THEOREM 5.9 Time Delay; Real Translation 1 1 1:5s 2
+f2½u(t)  u(t  1:5)g ¼ 2  e ¼ (1  e1:5s )
s s s
The substitution of t  l for the variable t in the transform
where the translation property has been used.
+{ f(t)} corresponds to the multiplication of the function F(s)
by els; that is,

+{ f (t  l)} ¼ esl F(s): (5:18) THEOREM 5.10 Complex Translation

The substitution of s þ a for s, where a is a real or complex, in the

Proof Refer to Figure 5.2, which shows a function f(t) u(t) and
function F(s þ a), corresponds to the Laplace transform of the
the same function delayed by the time t ¼ l, where l is a positive
product eatf(t).
constant.
We write directly Proof We write

ð
1 ð
1 ð
1
st at st
+{ f (t  l)u(t  l)} ¼ f (t  l)u(t  l)e dt: e f (t)e dt ¼ f (t)e(sþa)t dt for Re(s) > c  Re(a),
0 0 0

f(t) f (t)
2
2u(t)

f(t) f (t – λ)
f(t)

2 1.5
0 t

λ
t –2u(t – 1.5)
–2
t 1.5

FIGURE 5.2 A function f(t) at the time t ¼ 0 and delayed time t ¼ l. FIGURE 5.3 Pulse function and its equivalent representation.

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-7

which is Example 5.9

F(s þ a) ¼ +{eat f (t)}: (5:20) Given f1(t) ¼ t and f2(t) ¼ eat, deduce the Laplace transform of
the convolution t * eat by the use of Theorem 5.11.
In a similar way we find
SOLUTION
F(s  a) ¼ +{eat f (t)}: (5:21)
Begin with the convolution

ðt 

teat t teat eat t
THEOREM 5.11 Convolution t * eat ¼ (t  t)eat dt ¼  
a 0 a a2 0
0

The multiplication of the transforms of two sectionally continu- 1 at

¼ (e  at  1):
ous functions f1(t) (¼F1(s)) and f2(t) (¼F2(s)) corresponds to the a2
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Laplace transform of the convolution of f1(t) and f2(t). Then

 
F1 (s)F2 (s) ¼ +{ f1 (t) * f2 (t)} (5:22) 1 1 1 1 1 1
+{t * e } ¼ 2
at
  ¼ 2 :
a s  a s2 s s (s  a)
where the asterisk * is the shorthand designation for convolution.
By Theorem 5.11 we have
Proof By definition, the convolution of two functions f1(t)
and f2(t) is 1
F1 (s) ¼ +{ f1 (t)} ¼ +{t} ¼ ,
s2
ð
1 ð
1 1
F2 (s) ¼ +{ f2 (t)} ¼ +{eat } ¼ :
f1 (t) * f2 (t) ¼ f1 (t  t)f2 (t)dt ¼ f1 (t)f2 (t  t)dt: (5:23) sa
0 0 and

Thus, 1 1
+{t * eat } ¼ :
81 9 s2 (s  a)
<ð =
+{ f1 (t) * f2 (t)} ¼ + f1 (t  t)f2 (t)dt
: ;
0
21 3 THEOREM 5.12
ð
1 ð
¼ 4 f1 (t  t)f2 (t)dt5est dt
The multiplication of the transforms of three sectionally continu-
0 0
ous functions f1(t), f2(t), and f3(t) corresponds to the Laplace
ð
1 ð
1 transform of the convolution of the three functions
¼ f2 (t)dt f1 (t  t)est dt:
0 0
+{ f1 (t) * f2 (t) * f3 (t)} ¼ F1 (s)F2 (s)F3 (s): (5:24)

Now effect a change of variable, writing t  t ¼ j and therefore Proof This is an extension of Theorem 5.11. The result is
dt ¼ dj, then obvious if we write
ð
1 ð
1
F1 (s)F2 (s)F3 (s) ¼ +{ f1 (t) * +1 {F2 (s)F3 (s)}}:
¼ f2 (t)dt f1 (j)es(jþt) dj:
0 t
Example 5.10
But for positive time functions f1(j) ¼ 0 for j < 0, which permits
changing the lower limit of the second integral to zero, and so Deduce the values of the convolution products: 1 * f(t); 1 * 1 * f(t).

ð
1 ð
1
SOLUTION
¼ f2 (t)est dt f1 (j)esj dj,
By Equations 5.21 and 5.23 we write directly
0 0 nÐ o
t
a. For f1(t) ¼ 1, f2(t) ¼ f(t), +{1 * f (t)} ¼ F(s)
s ¼ + 0 f (j)dj
which is by Equation 5.14
b. FornÐ f1(t) ¼ 1, f2(t) ¼
o 1, f3(t) ¼ f(t), +{1 * 1 * f (t)} ¼ s2 ¼
F(s)
t Ðj
+{ f1 (t) * f2 (t)} ¼ F1 (s)F2 (s): + 0 0 f (l)dl dj

© 2010 by Taylor and Francis Group, LLC

 5-8 Transforms and Applications Handbook

The Laplace transform of f1(t), the integral on the right, con-

THEOREM 5.13 Frequency Convolution—s-Plane
verges in the range Re(s  z) > s1, were s1 is the abscissa of
convergence of f1(t). In addition, Re(z) ¼ s2 for the z-plane
The Laplace transform of the product of two piecewise and integration involved in Equation 5.25. Thus, the abscissa of
sectionally continuous functions f1(t) and f2(t) corresponds to convergence of f1(t) f2(t) is specified by
the convolution of their transforms, with

1 Re(s) > s1 þ s2 : (5:27)

+{ f1 (t)f2 (t)} ¼ [F1 (s) * F2 (s)]: (5:25)
2pj
This situation is portrayed graphically in Figure 5.4 for the case
Proof Begin by considering the following line integral in the
when both s1 and s2 are positive. As far as the integration in the
z-plane:
complex plane is concerned, the semicircle can be closed either
ð to the left or to the right just so long as F1(s) and F2(s) go to zero
1
f2 (t) ¼ F2 (z)ezt dz, s2 ¼ axis of the convergence: as s ! 1.
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

2pj
C2 Based on the foregoing, we observed the following:

This means that the contour intersects the x-axis at x1 > s2

. Poles of F1(s  z) are contained in the region Re(s  z) < s1
(see Figure 5.4). Then we have
. Poles of F2(z) are contained in the region Re(z) < s2
. From (a) and Equation 5.27 Re(z) > Re(s  s1) > s2
ð
1 ð
1 ð . Poles of F1(s  z) lie to the right of the path of integration
st 1
f1 (t)f2 (t)e dt ¼ f1 (t)dt F2 (z)e (zs)t
dz: . Poles of F2(z) are to the left of the path of integration
2pj . Poles of F1(s  z) are functions of s whereas poles of F2(z)
0 0 C2
are fixed in relation to s
Assume that the integral of F2(z) is convergent over the path of
integration. This equation is now written in the form Example 5.11

ð
1 ð
s2 þj1 ð
1 Find the Laplace transform of the function f(t) ¼ f1(t) f2(t) ¼
1
f1 (t)f2 (t)e st
dt ¼ F2 (z)dz f1 (t)e(sz)t dt et e2t u(t).
2pj
0 s2 j1 0
SOLUTION
s2 þj1
ð
1 D From Theorem 5.13 and the absolute convergence region for
¼ F2 (z)F1 (s  z)dz ¼ +{ f1 (t)f2 (t)}:
2pj each function, we have
s2 j1

(5:26)
1
F1 (s) ¼ , s1 > 1
sþ1
C2
jy 1
F2 (s) ¼ , s2 > 2:
sþ2
z-plane

Further, f(t) ¼ exp [(2 þ 1)t] u(t) implies that sf ¼ s1 þ s2 ¼ 3.

We now write
Region of
values of s 1 1
F2 (z)F1 (s  z) ¼
σ1 zþ2 szþ1
1 1 1 1
x ¼  :
σ2 x1 σ > σ2 + σ1 3 þ s z  (1 þ s) 3 þ s z þ 2

To carry out the integration dictated by Equation 5.26 we use

the contour shown in Figure 5.5. If we select contour C1 and
use the residue theorem, we obtain
þ
1
F(s) ¼ F2 (z)F1 (s  z)dz ¼ 2pj Re s½F2 (z)F1 (s  z)jz¼2
2pj
C1

1
¼ :
FIGURE 5.4 The contour C2 and the allowed range of s. sþ3

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-9

Imaginary If f(t) has a discontinuity at the origin, this expression specifies

the value of the impulse f(0þ). If f(t) contains an impulse term,
then the left-hand side does not exist, and the initial value
C1 C2 property does not exist.
s-plane

THEOREM 5.15 Final Value Theorem

Real
–2 –1
Let f(t) and f (1)(t) be Laplace transformable functions, then for
t!1

lim f (t) ¼ lim sF(s): (5:29)

t!1 s!0
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Proof Begin with Equation 5.13 and Let s ! 0. Thus, the

FIGURE 5.5 The contour for Example 5.11. expression

ð
1
The inverse of this transform is exp(3t). If we had selected df st
contour C2, the residue theorem gives lim e dt ¼ lim ½sF(s)  f (0þ):
s!0 dt s!0
0
þ
1
F(s) ¼ F2 (z)F1 (s  z)dz ¼ 2pj Re s[F2 (z)F1 (s  z)]jz¼1þs Consider the quantity on the left. Because s and t are independ-
2pj ent and because est ! 1 as s ! 0, then the integral on the left
C2

 becomes, in the limit
1 1
¼  ¼ :
sþ3 sþ3 ð
1
df
dt ¼ lim f (t)  f (0þ):
The inverse transform of this is also exp(3t), as to be dt t!1
0
expected.
Combine the latter two equations to get

lim f (t)  f (0þ) ¼ lim sF(s)  f (0þ):

THEOREM 5.14 Initial Value Theorem t!1 s!1

It follows from this that the final value of f(t) is given by

Let f(t) and f (1)(t) be Laplace transformable functions, then for
case when lim sF(s) as s ! 1 exists, lim f (t) ¼ lim sF(s):
t!1 s!0

lim sF(s) ¼ lim f (t): (5:28)

s!1 t!0þ This result applies F(s) possesses a simple pole at the origin, but it
does not apply if F(s) has imaginary axis poles, poles in the right
Proof Begin with Equation 5.13 and consider half plane, or higher order poles at the origin.

ð
1
df st Example 5.12
lim e dt ¼ lim ½sF(s)  f (0þ):
s!1 dt s!1
0 Apply the final value theorem to the following two functions:

Because f(0þ) is independent of s, and because the integral sþa s

F1 (s) ¼ , F2 (s) ¼ :
vanishes for s ! 1, then (s þ a)2 þ b2 s2 þ b2

lim ½sF(s)  f (0þ) ¼ 0:

s!1 SOLUTION

Furthermore, f(0þ) ¼ limt!0þ f(t) so that For the first function from sF1(s),

s(s þ a)
lim sF(s) ¼ lim f (t): lim ¼ 0:
s!1 t!0þ s!0 (s þ a)2 þ b2

© 2010 by Taylor and Francis Group, LLC

 5-10 Transforms and Applications Handbook

For the second function, SOLUTION

Observe that the denominator can be factored into the form
s2 (s þ 2) (s þ 3). Thus, F(s) can be written in partial fraction
sF(s) ¼ :
s þ b2
2
form as

However, this function has singularities on the imaginary axis s3 A B

at s ¼ jb, and the final value theorem does not apply. F(s) ¼ ¼ þ : (5:32)
(s þ 2)(s þ 3) s þ 2 s þ 3

The important properties of the Laplace transform are con- where A and B are constants that must be determined.
tained in Table A.5.2. To evaluate A, multiply both sides of Equation 5.32 by
(s þ 2) and then set s ¼ 2. This gives

5.4 The Inverse Laplace Transform s  3
A ¼ F(s)(s þ 2)js¼2 ¼ ¼ 5
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

s þ 3s¼2
We employ the symbol +1{F(s)}, corresponding to the direct
Laplace transform defined in Equation 5.1, to denote a function
and B(s þ 2)=(s þ 3)js ¼2 is identically zero. In the same manner,
f(t) whose Laplace transform is F(s). Thus, we have the Laplace pair to find the value of B we multiply both sides of Equation 5.32
by (s þ 3) and get
F(s) ¼ +{ f (t)}, f (t) ¼ +1 {F(s)}: (5:30)

s  3
B ¼ F(s)(s þ 3)js¼3 ¼ ¼ 6:
This correspondence between F(s) and f(t) is called the inverse s þ 3s¼3
Laplace transformation of f(t).
Reference to Table A.5.1 shows that F(s) is a rational function The partial fraction form of Equation 5.32 is
in s if f(t) is a polynomial or a sum of exponentials. Further, it
appears that the product of a polynomial and an exponential 5 6
F(s) ¼ þ :
might also yield a rational F(s). If the square root of t appears on sþ2 sþ3
f(t), we do not get a rational function in s. Note also that a
continuous function f(t) may not have a continuous inverse The inverse transform is given by
transform.    
1 1
Observe that the F(s) functions have been uniquely deter- f (t) ¼ +1 {F(s)} ¼ 5+1 þ 6+1
mined for the given f(t) function by Equation 5.1. A logical sþ2 sþ3
question is whether a given time function in Table A.5.1 is the ¼ 5e2t þ 6e3t
only t-function that will give the corresponding F(s). Clearly,
Table A.5.1 is more useful if there is a unique f(t) for each F(s). where entry 8 in Table A.5.1, is used.
This is an important consideration because the solution of prac-
tical problems usually provides a known F(s) from which f(t)
Example 5.14
must be found. This uniqueness condition can be established
using the inversion integral. This means that there is a one- Find the inverse Laplace transform of the function
to-one correspondence between the direct and the inverse trans-
form. This means that if a given problem yields a function F(s), sþ1
the corresponding f(t) from Table A.5.1 is the unique result. In F(s) ¼ :
[(s þ 2)2 þ 1](s þ 3)
the event that the available tables do not include a given F(s), we
would seek to resolve the given F(s) into forms that are listed in
Table A.5.1. This resolution of F(s) is often accomplished in SOLUTION
terms of a partial fraction expansion.
A few examples will show the use of the partial fraction form This function is written in the form
in deducing the f(t) for a given F(s).
A Bs þ C sþ1
F(s) ¼ þ ¼ :
s þ 3 [(s þ 2)2 þ 1] [(s þ 2)2 þ 1](s þ 3)
Example 5.13
The value of A is deduced by multiplying both sides of this
Find the inverse Laplace transform of the function equation by (s þ 3) and then setting s ¼ 3. This gives

s3 3 þ 1
F(s) ¼ : (5:31) A ¼ (s þ 3)F(s)js¼3 ¼ ¼ 1:
s2 þ 5s þ 6 (3 þ 2)2 þ 1

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-11

To evaluate B and C, combine the two fractions and equate the where
coefficients of the powers of s in the numerators. This yields
A1 A2 Ak Ap1 Ap2
F(s) ¼ þ þ  þ þ þ þ 
1[(s þ 2)2 þ 1] þ (s þ 3)(Bs þ C) sþ1 s  s1 s  s2 s  sk s  sp (s  sp )2
¼
[(s þ 2)2 þ 1](s þ 3) [(s þ 2)2 þ 1](s þ 3) Apr
þ :
(s  sp )r
from which it follows that

To find the constants Ak that are the residues of the function F(s)
(s2 þ 4s þ 5) þ Bs2 þ (C þ 3B)s þ 2C ¼ s þ 1:
at the simple poles sk, it is only necessary to note that as s ! sk
the term Ak(s  sk) will become large compared with all other
Combine like-powered terms to write
terms. In the limit

(1 þ B)s2 þ (4 þ C þ 3B)s þ (5 þ 3C) ¼ s þ 1:

Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Ak ¼ lim (s  sk )F(s): (5:35)

s!sk

Therefore,
Upon taking the inverse transform for each simple pole, the
1 þ B ¼ 0, 4 þ C þ 3B ¼ 1, 5 þ 3C ¼ 1: result will be a simple exponential of the form

 
From these equations we obtain Ak
+1 ¼ Ak esk t : (5:36)
s  sk
B ¼ 1, C ¼ 2:
Note also that because F(s) contains only real coefficients, if sk is a
The function F(s) is written in the equivalent form
complex pole with residue Ak, there will also be a conjugate pole
sk* with residue Ak* . For such complex poles
1 sþ2
F(s) ¼ þ :
s þ 3 (s þ 2)2 þ 1  
Ak Ak*
+1
*
þ ¼ Ak esk t þ Ak* esk t :
Now using Table A.5.1, the result is s  sk s  sk*

These can be combined in the following way:

f (t) ¼ e3t þ e2t cos t, t > 0:
response ¼ (ak þ jbk )e(sk þjvk )t þ (ak  jbk )e(sk jvk )t
In many cases, F(s) is the quotient of two polynomials with real
coefficients. If the numerator polynomials is of the same or ¼ esk t ½(ak þ jbk )( cos vk t þ j sin vk t)
higher degree than the denominator polynomial, first divide the þ (ak  jbk )( cos vk t þ j sin vk t)
numerator polynomial by the denominator polynomial; the div- ¼ 2esk t (ak cos vk t  bk sin vk t)
ision is carried forward until the numerator polynomial of the ¼ 2Ak esk t cos (vk t þ uk ) (5:37)
remainder is one degree less than the denominator. This results
in a polynomial in s plus a proper fraction. The proper fraction where uk ¼ tan1 (bk=ak) and Ak ¼ ak=cos uk.
can be expanded into a partial fraction expansion. The result of When the proper fraction contains a multiple pole of order r,
such an expansion is an expression of the form the coefficients in the partial-fraction expansion Ap1, Ap2, . . . Apr
that are involved in the terms
A1 A2 Ap1
F 0 (s) ¼ B0 þ B1 (s) þ    þ þ þ  þ Ap1 Ap2 Apr
s  s1 s  s2 s  sp þ þ  þ
(s  sp ) (s  sp )2 (s  sp )r
Ap2 Apr
þ 2 þ  þ : (5:33)
(s  sp ) (s  sp )r
must be evaluated. A simple application of Equation 5.35 is not
adequate. Now the procedure is to multiply both sides of Equa-
This expression has been written in a form to show three types of
tion 5.34 by (s  sp)r, which gives
terms; polynomial, simple partial fraction including all terms with
distinct roots, and partial fraction appropriate to multiple roots.

A1 A2 Ak
To find the constants A1, A2, . . . the polynomial terms are (s  sp )r F(s) ¼ (s  sp )r þ þ  þ
s  s1 s  s2 s  sk
removed, leaving the proper fraction
þ Ap1 (s  sp )r1 þ   
F 0 (s)  (B0 þ B1 s þ    ) ¼ F(s) (5:34) þ Ap(r1) (s  sp ) þ Apr (5:38)

© 2010 by Taylor and Francis Group, LLC

 5-12 Transforms and Applications Handbook

In the limit as s ¼ sp all terms on the right vanish with the From Table A.5.1 the inverse transform is
exception of Apr. Suppose now that this equation is differentiated
once with respect to s. The constant Apr will vanish in the f (t) ¼ d(t) þ 4 þ t  et , for t  0:
differentiation but Ap(r1) will be determined by setting s ¼ sp.
This procedure will be continued to find each of the coefficients
If the function F(s) exists in proper fractional form as the quo-
Apk. Specifically, the procedure is specified by
tient of two polynomials, we can employ the Heaviside expansion
 rk  theorem in the determination of f(t) from F(s). This theorem is
1 d an efficient method for finding the residues of F(s). Let
Apk ¼ F(s)(s  s p )r
, k ¼ 1, 2, . . . , r:
(r  k)! dsrk s¼sp

(5:39) P(s) A1 A2 Ak
F(s) ¼ ¼ þ þ  þ
Q(s) s  s1 s  s2 s  sk
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Example 5.15 where P(s) and Q(s) are polynomials with no common factors
and with the degree of P(s) less than the degree of Q(s).
Find the inverse transform of the following function: Suppose that the factors of Q(s) are distinct constants. Then, as
in Equation 5.35 we find
s3 þ 2s2 þ 3s þ 1

F(s) ¼ :
s2 (s þ 1) s  sk
Ak ¼ lim P(s) :
s!sk Q(s)

SOLUTION Also, the limit P(s) is P(sk). Now, because

This is not a proper fraction. The numerator polynomial is
divided by the denominator polynomial by simple long div- s  sk 1 1
ision. The result is
lim ¼ lim (1) ¼ (1) ,
s!sk Q(s) s!sk Q (s) Q (sk )

s2 þ 3s þ 1 then
F(s) ¼ 1 þ :
s2 (s þ 1)
P(sk )
The proper fraction is expanded into partial fraction form Ak ¼ :
Q(1) (sk )

s2 þ 3s þ 1 A11 A12 A2 Thus,

Fp (s) ¼ ¼ þ 2 þ :
s2 (s þ 1) s s sþ1

P(s) Xk
P(sn ) 1
The value of A2 is deduced using Equation 5.35 F(s) ¼ ¼  : (5:40)
Q(s) n¼1 Q(1) (sn ) (s  sn )

s2 þ 3s þ 1
A2 ¼ [(s þ 1)Fp (s)]s¼1 ¼  ¼ 1: From this, the inverse transformation becomes
s2 s¼1

  X
To find A11 and A12 we proceed as specified in Equation 5.39 P(s) k
P(sn ) sn t
f (t) ¼ +1 ¼ e :
Q(s) Q(1) (sn )
 n¼1
s2 þ 3s þ 1
A12 ¼ [s2 Fp (s)]s¼0 ¼ ¼1
s þ 1 s¼0
This is the Heaviside expansion theorem. It can be written in
 

1 d 2 d s2 þ 3s þ 1 formal form.
A11 ¼ s Fp (s) ¼
1! ds s¼0 ds sþ1 s¼0

s2 þ 3s þ 1 2s þ 3
¼ þ ¼ 4:
(s þ 1)2 s þ 1 s¼0 THEOREM 5.16: Heaviside Expansion Theorem

Therefore, If F(s) is the quotient P(s)=Q(s) of two polynomials in s such that

Q(s) has the higher degree and contains simple poles the factor
4 1 1 s  sk, which are not repeated, then the term in f(t) corresponding
F(s) ¼ 1 þ þ 2  :
s s sþ1 to this factor can be written QP(s k) s t
(1) (s ) e k .
k

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-13

Example 5.16 Then

pffiffiffi pffiffiffi
Repeat Example 5.13 employing the Heaviside expansion 1  j2 3 (2j2pffiffi3)t 1 þ j2 3 (2þj2pffiffi3)t
theorem. f (t) ¼ pffiffiffi e þ pffiffiffi e
j2 3 j2 3

pffiffiffi pffiffiffi 
SOLUTION 1  j2 3 j2 3t 1 þ j2 3 j2pffiffi3t
p ffiffi
¼ e2t pffiffiffi e þ pffiffiffi e
j2 3 j2 3
We write Equation 5.31 in the form " pffiffi pffiffi #
j2 3t
 ej2 3t ) pffiffi pffiffi
2t (e j2 3t
¼e pffiffiffi þ (e þe j2 3t
)
P(s) s3 s3 j2 3
F(s) ¼ ¼ ¼ :  
Q(s) s2 þ 5s þ 6 (s þ 2)(s þ 3) pffiffiffiffi 1 pffiffiffiffi
¼ e2t 2 cos 2 3t  pffiffiffi sin 2 3t
3
The derivative of the denominator is
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Q(1) (s) ¼ 2s þ 5 5.5 Solution of Ordinary Linear

Equations with Constant
from which, for the roots of this equation,
Coefficients
Q(1) (2) ¼ 1, Q(1) (3) ¼ 1: The Laplace transform is used to solve homogeneous and non-
homogeneous ordinary differential equations or systems of such
Hence, equations. To understand the procedure, we consider a number
of examples.
P(2) ¼ 5, P(3) ¼ 6:
Example 5.18
The final value for f(t) is
Find the solution to the following differential equation subject
f (t) ¼ 5e2t þ 6e3t : to prescribed initial conditions: y(0þ); (dy=dt) þ ay ¼ x(t).

SOLUTION
Example 5.17 Laplace transform this differential equation. This is accom-
plished by multiplying each term by estdt and integrating
Find the inverse Laplace transform of the following function from 0 to 1. The result of this operation is
using the Heaviside expansion theorem:
sY(s)  y(0þ) þ aY(s) ¼ X(s),
 
2s þ 3
+1 : from which
s2 þ 4s þ 7

X(s) y(0þ)
Y(s) ¼ þ :
sþa sþa
SOLUTION
The roots of the denominator are If the input x(t) is the unit step function u(t), then X(s) ¼ 1=s
and the final expression for Y(s) is
pffiffiffi pffiffiffi
s2 þ 4s þ 7 ¼ (s þ 2 þ j 3)(s þ 2  j 3):
1 y(0þ)
Y(s) ¼ þ :
s(s þ a) s þ a
That is, the roots of the denominator are complex. The deriva-
tive of the denominator is Upon taking the inverse transform of this expression

 
Q(1) (s) ¼ 2s þ 4: 1 1 1 y(0þ)
y(t) ¼ +1 {Y(s)} ¼ +1  þ
a s sþa sþa
We deduce the values P(s)=Q(1)(s) for each root
pffiffiffi pffiffiffi pffiffiffi with the result
For s1 ¼ 2  j 3 Q(1) (s1 ) ¼ j2 3 P(s1 ) ¼ 1  j2 3
pffiffiffi pffiffiffi 1
For s2 ¼ 2 þ j=3 Q(1) (s2 ) ¼ þj2 3 P(s2 ) ¼ 1 þ j2 3: y(t) ¼ (1  eat ) þ y(0þ)eat :
a

© 2010 by Taylor and Francis Group, LLC

 5-14 Transforms and Applications Handbook

Example 5.19 Example 5.20

Find the general solution to the differential equation Find the velocity of the system shown in Figure 5.6a when the
applied force is f(t) ¼ etu(t). Assume zero initial conditions.
d2 y dy Solve the same problem using convolution techniques. The
þ 5 þ 4y ¼ 10 input is the force and the output is the velocity.
dt 2 dt

subject to zero initial conditions. SOLUTION

SOLUTION The controlling equation is, from Figure 5.6b,

Laplace transform this differential equation. The result is ðt

dv
þ 5v þ 4 v dt ¼ et u(t):
dt
10
s2 Y(s) þ 5sY(s) þ 5Y(s) ¼ : 0
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

s
Laplace transform this equation and then solve for F(s).
Solving for Y(s), we get We obtain

10 10 s s
Y(s) ¼ ¼ : V(s) ¼ ¼ :
s(s2 þ 5s þ 4) s(s þ 1)(s þ 4) (s þ 1)(s2 þ 5s þ 4) (s þ 1)2 (s þ 4)

Write this expression in the form

Expand this into partial-fraction form, thus
A B C
A B C V(s) ¼ þ þ
Y(s) ¼ þ þ : s þ 4 s þ 1 (s þ 1)2
sþ1 sþ4 s
where
Then

 s  4
10  A¼ ¼
A ¼ Y(s)(s þ 1)js¼1 ¼ ¼
10 (s þ 1)2 s¼4 9
s(s þ 4)s¼1 3  
 1 d s  4
B¼  ¼
10  10 1! ds s þ 4 s¼1 9
B ¼ Y(s)(s þ 4)js¼4 ¼ ¼
s(s þ 1)s¼4 12 
 s  1
10  C¼  ¼ :
C ¼ sY(s)js¼0 ¼  ¼ 10 s þ 4 s¼1 3
(s þ 1)(s þ 4)s¼0 4
The inverse transform of V(s) is given by
and
4 4 1

 v(t) ¼  e4t þ et  tet , t  0:
1 1 1 9 9 3
Y(s) ¼ 10  þ þ :
3(s þ 1) 12(s þ 4) 4s To find v(t) by the use of the convolution integral, we first find
h(t), the impulse response of the system. The quantity h(t) is
The inverse transform is specified by

 ð
1 1 1 dh
x(t) ¼ 10  et þ e4t þ : þ 5h þ 4 h dt ¼ d(t)
3 12 4 dt

V
+
D=5

f M=1
f K=4 V
M=1
K=4 D=5
(a) (b)

FIGURE 5.6 The mechanical system and its network equivalent.

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-15

where the system is assumed to be initially relaxed. The All terms in these equations are Laplace transformed. The
Laplace transform of this equation yields result is the set of equations

s s 4 1 1 1 (3 þ 2s)I1 (s)  (I þ 2s)I2 (s) ¼ V1 (s) þ 2[i1 (0þ)  i2 (0þ)]

H(s) ¼ ¼ ¼  :  
s2 þ 5s þ 4 (s þ 4)(s þ 1) 3 s þ 4 3 s þ 1 1 q2 (0þ)
(1 þ 2s)I1 (s) þ 3 þ 2s þ I2 (s) ¼ 2[  i1 (0þ) þ i2 (0þ)] 
s s
The inverse transform of this expression is easily found to be
V2 (s) ¼ 2I2 (s):
4 1
h(t) ¼ e4t  et , t  0: The current through the inductor is
3 3
iL (t) ¼ i1 (t)  i2 (t):
The output of the system to the input etu(t) is written
At the instant t ¼ 0þ
ð
1

ðt 
4 1
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

v(t) ¼ h(t)f (t  t)dt ¼ e(tt) e4t  et dt iL (0þ) ¼ i1 (0þ)  i2 (0þ):

3 3
1 0
2 3 Also, because
ðt ðt
  t 
4 1 t 4 1  1 :
3t 
¼ et 4 e3t dt  5
dt ¼ e e  t ðt
3 3 3 3 0 3 1 1
0 0 q2 (t) ¼ i2 (t)dt
C C
4 4 1 1
¼  e4t þ et  tet , t0
9 9 3 ðt ð0
1 1
¼ lim i2 (t)dt þ i2 (t)dt ¼ 0 þ vc (0),
This result is identical with that found using the Laplace C t!0þ C
0 1
transform technique.
then
Example 5.21
q2 (0þ) D q2 (0þ)
¼ vc (0þ) ¼ vc (0) ¼ i2(1) (0) ¼ :
C 1
Find an expression for the voltage v2(t) for t > 0 in the circuit of
Figure 5.7. The source v1(t), the current iL(0) through L ¼ 2 H, The equation set is solved for I2(s), which is written by
and the voltage vc(0) across the capacitor C ¼ 1 F at the Cramer’s rule
switching instant are all assumed to be known.
 
 3 þ 2s V1 (s) þ 2iL (0þ) 

SOLUTION  
 (1 þ 2s) 2iL (0þ)  vc (0þ) 
I2 (s) ¼  
s
After the switch is closed, the circuit is described by the loop  3 þ 2s (1 þ 2s) 

equations  
 (1 þ 2s) 3 þ 2s þ 1 
s
    
2di1 2di2 (3 þ 2s) 2iL (0þ)  vc (0þ) þ (1 þ 2s)[V1 (s) þ 2iL (0þ)]
3i1 þ  1i2 þ ¼ v2 (t) ¼ 2s2 þ3sþ1
s
dt dt (3 þ 2s)  (1 þ 2s)2
   ð  s
2di1 2di2
 1i1 þ þ 3i2 þ þ i2 dt ¼ 0 (2s2 þ 3s)vc (0þ)  4siL (0þ) þ (2s2 þ s)V1 (s)
dt dt ¼ :
8s2 þ 10s þ 3
v2 (t) ¼ 2i2 (t):
Further

V2 (s) ¼ 2I2 (s):

1F
s 2Ω Then, upon taking the inverse transform
+
+ υc v1 (t) ¼ 2+1 {I2 (s)}:
1Ω
+
v1 (t) i1 i2 2Ω v2 (t) If the circuit contains no stored energy at t ¼ 0, then iL(0þ) ¼
vc(0þ) ¼ 0 and now
iL 2H
 
(2s2 þ s)V1 (s)
v2 (t) ¼ 2+1 :
8s2 þ 10s þ 3
FIGURE 5.7 The circuit for Example 5.21.

© 2010 by Taylor and Francis Group, LLC

 5-16 Transforms and Applications Handbook

For the particular case when vl ¼ u(t) so that V1(s) ¼ 1=s For zero initial current through the inductor, the Laplace
( ) transform of the equation is
 
1 2s þ 1 1 2s þ 1
v2 (t) ¼ 2+ ¼ 2+ 
8s2 þ 10s þ 3 8 s þ 12 (s þ 3=4) (s þ 1)I(s) ¼ V(s):
( )
1 1 1 1 Now, from the fact that +{d(t)} ¼ 1 and the shifting property
¼ + ¼ e3t=4 , t  0:
2 s þ 34 2 of Laplace transforms, we can write the explicit form for V(s),
which is
The validity of this result is readily confirmed because at the
instant t ¼ 0þ the inductor behaves as an open circuit and V(s) ¼ 2 þ es þ 2e2s þ e3s þ 2e4s þ   
the capacitor behaves as a short circuit. Thus, at this instant,
¼ (2 þ es )(1 þ e2s þ e4s þ    )
the circuit appears as two equal resistors in a simple series
circuit and the voltage is shared equally. 2 þ es
¼ :
1  e2s
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Example 5.22 Thus, we must evaluate i(t) from

The input to the RL circuit shown in Figure 5.8a is the recurrent

2 þ es 1 2 es
series of impulse functions shown in Figure 5.8b. Find the I(s) ¼ 2s
¼ 2s
þ 2s
:
output current. 1e s þ 1 (1  e )(s þ 1) (1  e )(s þ 1)

Expand these expressions into

SOLUTION
The differential equation that characterizes the system is 2 
I(s) ¼ 1 þ e2s þ e4s þ e6s þ   
sþ1
di(t) 1  s
þ i(t) ¼ v(t): þ e þ e3s þ e5s þ e7s þ    :
dt sþ1

R=1 Ω
v(t)

2
+
i(t) L=1 H
v(t)
1

t
0 1 2 3 4 5
(a) (b)

FIGURE 5.8 (a) The circuit, (b) the input pulse train.

2
i(t)

0
1 2 3 4 5 6

FIGURE 5.9 The response of the RL circuit to the pulse train.

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-17

The inverse transform of these expressions yields transform, the regions of convergence for functions of time that are
zero for t > 0, zero for t < 0, or in neither category, must be
i(t) ¼ 2et u(t) þ 2e(t2) u(t  2) þ 2e(t4) u(t  4) þ    distinguished. For the one-sided transform, the region of conver-
þ e(t1) u(t  1) þ e(t3) u(t  3) þ e(t5) u(t  5) þ    gence is given by s, where s is the abscissa of absolute convergence.
The path of integration in Equation 5.42 is usually taken as
The result has been sketched in Figure 5.9. shown in Figure 5.10 and consists of the straight line ABC
displayed to the right of the origin by s and extending in the
limit from j1 to þj1 with connecting semicircles. The evalu-
5.6 The Inversion Integral ation of the integral usually proceeds by using the Cauchy inte-
The discussion in Section 5.4 related the inverse Laplace trans- gral theorem, which specifies that
form to the direct Laplace transform by the expressions þ
1
f (t) ¼ lim F(s)est ds
F(s) ¼ +{ f (t)} (5:41a) 2pj R!1
G1
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

1 X
f (t) ¼ + {F(s)}: (5:41b) ¼ residues of F(s)est at the singularities

The subsequent discussion indicated that the use of Equation 5.41b to the left of ABC; t > 0: (5:43)
suggested that the f(t) so deduced was unique; that there was no
other f(t) that yielded the specified F(s). We found that although f(t) But the contribution to the integral around the circular path with
represents a real function of the positive real variable t, the trans- R ! 1 is zero, leaving the desired integral along the path ABC, and
form F(s) can assume a complex variable form. What this means, of þ
1
course, is that a mathematical form for the inverse Laplace trans- f (t) ¼ lim F(s)est ds
form was not essential for linear functions that satisfied the Dirich- 2pj R!1
G2
let conditions. In some cases, Table A.5.1 is not adequate for many X
functions when s is a complex variable and an analytic form for the ¼ residues of F(s)est at the singularities
inversion process of Equation 5.41b is required. to the right of ABC; t < 0: (5:44)
To deduce the complex inversion integral, we begin with the
Cauchy second integral theorem, which is written We will present a number of examples involving these equations.
þ
F(z)
dz ¼ j2pF(s) Example 5.23
sz

where the contour encloses the singularity as s. The function F(s) Use the inversion integral to find f(t) for the function
is analytic in the half plane Re(s)  c. If we apply the inverse
1
Laplace transformation to the function s on both sides of this F(s) ¼ :
s2 þ w 2
equation, we can write
Note that by entry 15 of Table A.5.1, this is sin wt=w.
ð
sþjv  
1
j2p+1 {F(s)} ¼ lim F(z)+1 dz:
v!1 sz
sjv jω
t>0 C t<0

But F(s) is the Laplace transform of f(t); also, the inverse trans-
Г1 Г1 Г2
form of 1=(s  z) is ezt. Then it follows that
R Г2
ð
sþjv ð
sþj1
1 1
f (t) ¼ lim ezt F(z)dz ¼ ezt F(z)dz: (5:42)
2pj v!1 2pj σ
sjv sj1 B
σ
This equation applies equally well to both the one-sided and the
two-sided transforms.
It was pointed out in Section 5.1 that the path of integration
(Equation 5.42) is restricted to value of s for which the direct
A
transform formula converges. In fact, for the two-sided Laplace
transform, the region of convergence must be specified in order to
determine uniquely the inverse transform. That is, for the two-sided FIGURE 5.10 The path of integration in the s-plane.

© 2010 by Taylor and Francis Group, LLC

 5-18 Transforms and Applications Handbook

SOLUTION branch point. Because a branch cut can never be crossed, this
essentially ensures that F(s) is single valued. Now, however, the
The inversion integral is written in a form that shows the poles inversion integral (Equation 5.43) becomes for t > 0
of the integrand.
þ ð ð
sþj1
1 est 1 1
f (t) ¼ ds: f (t) ¼ lim F(s)e ds ¼st
F(s)est ds
2pj (s þ jw)(s  jw) R!1 2pj 2pj
GAB sj1
2 3
The path chosen is G1 in Figure 5.10. Evaluate the residues ð ð ð ð ð ð ð
1 6 7

  ¼ 4 þ þ þ þ þ þ 5, (5:45)
est est  e jwt 2pj
Res (s  jw) ¼  ¼ BC G2 ‘ g ‘þ G3 FG
s2 þ w 2 s¼jw s þ jw s¼jw 2wj

 
est est  ejwt which does not include any singularity.
Res (s þ jw) ¼ ¼ : First we will show that for t > 0 the integrals over theÐ con-
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

s þ w2
2 
s  jw s¼jw 2wj
s¼jw
tours BC and CD vanish as R ! 1, from which G2 ¼
Ð Ð 1
Therefore, G3 ¼ GBC ¼ FG ¼ 0. Note from Figure 5.11 that b ¼ cos (s=R)
so that the integral over the arc BC is, because je j ¼ 1,
ju

X e jwt  ejwt sin wt

f (t) ¼ Res ¼ ¼ :
2jw w ð  st jvt 
 pffiffiffi ð
p=2
e e 
jIj   jRe du ¼ e
ju st
R du
 12 ju=2 
BC
R e b
Example 5.24 pffiffiffip s pffiffiffi s
 pffiffi ¼ est R  cos1 ¼ est R sin1
2 R R
Evaluate + 1= s .

But for small arguments sin1(s=R) ¼ s=R, and in the limit as

SOLUTION
R ! 1, I ! 0. By a similar approach, we find that the integral
pffiffi
The function F(s) ¼ s is a double-valued function because of over CD is zero. Thus, the integrals over the contours G2 and
the square root operation. That is, if s is represented in polar G3 are also zero as R ! 1.
form p re j(u þ 2p)
byffiffi re jup, ffiffiffiffiffiffiffiffiffiffiffiffiffiffi For evaluating the integral over g, let s ¼ reju ¼ r(cos u þ
ffi is apsecond
ffiffiffiffiffiffiffi acceptable representation,
and s ¼ pffiffirej(uþ2p) ¼  reju , thus showing two different j sin u) and
values for s. But a double-valued function is not analytic
and requires a special procedure in its solution. ð ðp
The procedure is to make the function analytic by restrict- er( cos uþj sin u)t ju
F(s)est ds ¼ pffiffiffiffi ju=2 jre du ¼ 0 as r ! 0:
ing the angle of s to the range p < u < p and by excluding re
g p
the point s ¼ 0. This is done by constructing a branch cut
along the negative real axis, as shown in Figure 5.11. The end
of the branch cut, which is the origin in this case, is called a The remaining integrals in Equation 5.45 are written

2 3
ð ð
1 4
jω f (t) ¼  F(s)est ds þ F(s)est ds5: (5:46)
2pj
C B ‘ ‘þ

pffiffi pffiffiffi
pffiffiffilet s ¼ ue ¼ u; s ¼ j u, and ds ¼ du,
jp
Г2 R Along path l,
Г1 where u and u are real positive quantities. Then

β
l– H ð ð0 ð
1
D A eut 1 eut
σ F(s)e ds ¼ 
st
pffiffiffi du ¼ pffiffiffi du:
E γ j u j j u
l+ ‘ 1 0
σ
pffiffi pffiffiffi pffiffiffi
Along path lþ, s ¼ ue j2p ¼ u, s ¼ j u (not þ j u),
Г3
and ds ¼ du: Then

F G ð ð
1 ð
1
eut 1 eut
F(s)est ds ¼  pffiffiffi du ¼ pffiffiffi du:
j u j j u
pffiffi ‘þ
FIGURE 5.11 The integration contour for +1 {1= s}. 0 0

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-19

Combine these results to find The problem we now face in this evaluation is that

2 3  
ð
1 ð
1 n(s)  0
Res (s  a) ¼
d(s) s¼a 0
1 42 1
u2 eut du5 ¼ u2 eut du,
1 1
f (t) ¼ 
2pj j p
0 0
where the roots of d(s) are such that s ¼ a cannot be factored.
which is a standard form integral with the value However, we know from complex function theory that

rffiffiffiffi d[d(s)] d(s)  d(a) d(s)
1 p 1 ¼ lim ¼ lim
f (t) ¼ ¼ pffiffiffiffiffi , t > 0: ds s¼a s!a s  a s!a s  a
p t pt
because d(a) ¼ 0. Combine this result with the above equation
to obtain
Example 5.25 
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

 
n(s)  n(s) 
Res (s  a) ¼  : (5:48)
Find the inverse Laplace transform of the function d(s) s¼a dsd [d(s)]
s¼a

1 By combining Equation 5.48 with Equation 5.47, we obtain

F(s) ¼ :
s(1 þ es )
( )
est  e jnpt

Res d  ¼ n odd:
s ds (1 þ es )  jnp
SOLUTION s¼jnp

st
e
The integrand in the inversion integral s(1þe s ) possesses We obtain, by adding all of residues,
simple poles at: s ¼ 0 and s ¼ jnp, n ¼ 1, 3, 1 . . . (odd
values). These are illustrated in Figure 5.12. We see that the 1 X1
e jnpt
function est=s(1 þ es) is analytic in the s-plane except at the f (t) ¼ þ :
2 n¼1 jnp
simple poles at s ¼ 0 and s ¼ jnp. Hence, the integral is
specified in terms of the residues in the various poles. We
have, specifically This can be rewritten as follows

 

sest  1 ej3pt ejpt e jpt e j3pt
Res  ¼1 for s ¼ 0 f (t) ¼ þ    þ þ þ þ þ 
s(1 þ es ) s¼0 2 2 j3p jp jp j3p
 
(s  jn)est 
(5:47) 1 X1
2j sin npt
Res ¼
0
for s ¼ jnp: ¼ þ :
s(1 þ es ) s¼jn 0 2 n¼1
jnp
n odd

This assumes the form

1 2 X1
sin (2k  1)pt
jω f (t) ¼ þ : (5:49)
2 p k¼1 2k  1
B (k + 1)th pole
k th pole As a second approach to a solution to this problem, we will
show the details in carrying out the contour integration for
this problem. We choose the path shown in Figure 5.12 that
includes semicircular hooks around each pole, the vertical
s-plane
R connecting line from hook to hook, and the semicircular
C path at R ! 1. Thus, we examine
σ
þ
1 est
f (t) ¼ ds
2pj s(1 þ es )
2 3
ð ð X ð X
1 66 þ
7
–k th pole ¼ þ  Res7
5: (5:50)
(–k + 1)th pole 2pj 4
A BCA vertical connecting lines Hooks
I1 I2 I3

FIGURE 5.12 The pole distribution of the given function. We consider the several integrals in this equation.

© 2010 by Taylor and Francis Group, LLC

 5-20 Transforms and Applications Handbook

Integral I1. By setting s ¼ re ju and taking into consideration In addition, the function f(t) is continuous for t > 0 and f(0) ¼ 0,
that cos u ¼ cos u for u > p=2, the integral I1 ! 0 as r ! 1. and f(t) is of the order O(ect) for all t > 0.
Integral I2. Along the Y-axis, s ¼ jy and Suppose that there are two transformable functions f1(t) and
f2(t) that have the same transforms
ð
1
e jyt
I2 ¼ j dy:
jy(1 þ ejy ) +{ f1 (t)} ¼ +{ f2 (t)} ¼ F(s):
1
r!0

The difference between the two functions is written f(t)

Note that the integrand is an odd function, whence I2 ¼ 0.
Integral I3. Consider a typical hook at s ¼ jnp. The result is
f(t) ¼ f1 (t)  f2 (t)


(s  jn)e
st
0
lim ¼  where f(t) is a transformable function. Thus,
r!0
s!jnp
s(1 þ es ) 0
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

This expression is evaluated (as for Equation 5.47) and yields +{f(t)} ¼ F(s)  F(s) ¼ 0:
e jnpt=jnp. Thus, for all poles
Additionally,
2 3
ð
p

jp 4 X
2
1
1 est e jnpt 15
I3 ¼ ds ¼ þ f(t) ¼ +1
t {0} ¼ 0, t > 0:
2pj s
s(1 þ e ) 2pj n¼1 jnp 2
p2 n odd
2 3 Therefore, this requires that f1(t) ¼ f2(t). The result shows that it
1 61 2 X
1
sin npt 7 is not possible to find two different functions by using two
¼ 4 þ 5: different values of s in the inversion integral. This conclusion
2 2 p n¼1 n
n odd can be expressed as follows:

Finally, the residues enclosed within the contour are

est 1 X1
ejnpt 1 2 X
1
sin npt THEOREM 5.18
Res ¼ þ ¼ þ ,
s(1 þ es ) 2 n¼1 jnp 2 p n¼1 n
n odd n odd
Only a single function f(t) that is sectionally continuous, of expo-
which is seen to be twice the value around the hooks. Then nential order, and with a mean value at each point of discontinuity,
when all terms are included in Equation 5.50, the final result is corresponds to a given transform F(s).

1 2 X
1
sin npt 1 2 X1
sin (2k  1)pt 5.7 Applications to Partial Differential
f (t) ¼ þ ¼ þ 
2 p n¼1 n 2 p k¼1 2k  1
n odd Equations
We now shall show that the direct and inverse transforms The Laplace transformations can be very useful in the solution
specified by Equation 5.30 and listed in Table A.5.1 constitute of partial differential equations. A basic class of partial differ-
unique pairs. In this connection, we see that Equation 5.42 can ential equations is applicable to a wide range of problems.
be considered as proof of the following theorem: However, the form of the solution in a given case is critically
dependent on the boundary conditions that apply in any par-
ticular case. In consequence, the steps in the solution often will
call on many different mathematical techniques. Generally, in
THEOREM 5.17
such problems the resulting inverse transforms of more com-
plicated functions of s occur than those for most linear systems
Let F(s) be a function of a complex variable s that is analytic and problems. Often the inversion integral is useful in the solution
of order O(sk) in the half-plane Re(s)  c, where c and k are real of such problems. The following examples will demonstrate the
constants, with k > 1. The inversion integral (Equation 5.42) approach to typical problems.
written +1t {F{(s)} along any line x ¼ s, with s  c converges
to the function f(t) that is independent of s,
Example 5.26
f (t) ¼ +1
t {F(s)}
Solve the typical heat conduction equation
whose Laplace transform is F(s),
q2 w qw
¼ , 0 < x < 1, t  0 (5:51)
F(s) ¼ +{ f (t)}, Re(s)  c: qx 2 qt

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-21

subject to the conditions Also write

C-1: w(x, 0) ¼ f (x), t ¼ 0 pffiffi (x  l)

qw s t  pffiffi ¼ u:
C-2: ¼ 0, w(x, t) ¼ 0 x ¼ 0: 2 t
qx

Then
SOLUTION
Multiply both sides of Equation 5.51 by esx dx and integrate ð
1
 1ð
from 0 to 1. 1 (x  l)2 2 du
w(x, t) ¼ f (l) exp  dl eu pffiffi :
2pj 4t t
1
ð
1 0

F(s, t) ¼ esx w(x, t)dx:

0 But the integral
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Also ð
1
pffiffiffiffi
eu du ¼
2

ð
1 p:
q w sx
2
qw
e dx ¼ s2 F(s, t)  sw(0þ)  (0þ): 0
qx 2 qx
0
Thus, the final solution is
Equation 5.51 thus transforms, subject to C-2 and zero bound-
ary conditions, to
ð
1
1 (xl)2

dF w(x, t) ¼ pffiffiffiffiffi f (l)e 4t dl:

 s2 F ¼ 0: 2 pt
1
dt

The solution to this equation is

Example 5.27
s2 t
F ¼ Ae :
A semi-infinite medium, initially at temperature w ¼ 0
throughout the medium, has the face x ¼ 0 maintained at
By an application of condition C-1, in transformed form, we have
temperature w0. Determine the temperature at any point of
the medium at any subsequent time.
ð
1

F¼A¼ f (l)esl dl:

SOLUTION
0
The controlling equation for this problem is
The solution, subject to C-1, is then
q2 w 1 qw
ð
1 ¼ (5:52)
qx 2 K qt
F(s, t) ¼ eþs t f (l)esl dl:
2

0
with the boundary conditions:
Now apply the inversion integral to write the function in terms a. w ¼ w0 at x ¼ 0, t > 0
of x from s, b. w ¼ 0 at t ¼ 0, x > 0.
21 3 To proceed, multiply both sides of Equation 5.52 by est dt
ð
1 ð
1 þs2 t 4 and integrate from 0 to 1. The transformed form of Equa-
w(x, t) ¼ e f (l)e dl5esx ds
sl
2pj tion 5.52 is
1 0
ð
1 1ð
¼
1
f (l)dl
2
es tslþsx ds: d2 F  s 
2pj  F ¼ 0, K > 0:
1 0 dx 2 K

Note that we can write The solution of this differential equation is

  pffiffiffiffiffi pffiffiffiffiffi
pffiffi (x  l) 2 (x  l)2
s2 t  s(x  l) ¼ s t  pffiffi  : F ¼ Aex s=k
þ Bex s=k
:
2 t 4t

© 2010 by Taylor and Francis Group, LLC

 5-22 Transforms and Applications Handbook

But F must be finite or zero for infinite x; therefore, B ¼ 0 and As in Example 5.24

pffis ð ð ð ð
F(s, x) ¼ Ae K x : ¼ ¼ ¼ ¼ 0:
G2 G3 BC FG
Apply boundary condition (a) in transformed form, namely
For the segments
ð
1
w0 ð ð
F(0, s) ¼ est w0 dt ¼ for x ¼ 0:
s , let s ¼ rejp and for , let s ¼ rejp :
0
‘ ‘þ
Therefore,
Then for ‘ and ‘þ, writing this sum I‘,
w0
A¼
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

s ð
1
1 h pffiffiffiffiffi pffiffiffiffiffi i ds
I‘ ¼ est ejx s=K  ejx s=K
and the solution in Laplace transformed form is 2pj s
0
w pffis ð
1 rffiffiffi
F(s, x) ¼ 0 e K x : (5:53) 1 s ds
s ¼ est sin x :
p K s
0
To find w(x, t) requires that we find the inverse transform of
this expression. This requires evaluating the inversion integral
Write

ð
sþj1 pffis rffiffiffi
w ex K est s
w(x, t) ¼ 0 ds: (5:54) u¼ , s ¼ ku2 , ds ¼ 2ku du:
2pj s K
sj1

Then we have
This integral has a branch point at the origin (see Figure 5.13).
To carry out the integration, we select a path such as that
shown (see also Figure 5.11). The integral in Equation 5.54 is ð
1
2 du
eKu t sin ux
2
written I‘ ¼  :
p u
0
2 3
ð ð ð ð ð ð ð
w0 6 7
w(x, t) ¼ 4 þ þ þ þ þ þ 5: This is a known integral that can be written
2pj
BC G2 l g lþ G3 FG
ffiffi
px
ð
2 Kt
2
eu du:
2
Il ¼  pffiffiffiffi
p
0
jω
C B
Finally, consider the integral over the hook,
Г2
ð pffiffiffiffiffi
Г1 s=K
1 ex
Iy ¼ e st
ds:
2pj s
γ g
D l–
A
σ
E Let us write
l+
σ
ds
Г3 s ¼ re ju , ds ¼ jre ju du, ¼ ju,
s

F G then

ð pffiffiffiffiffi
j ju
r =K e ju=2
FIGURE 5.13 The path of integration. Ig ¼ etre ex du:
2pj

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-23

For r ! 0, Ig ¼ j2pj
2p
¼ 2pj
2pj , then Ig ¼ 1. Hence, the sum of the By condition c
integrals in Equation 5.53 becomes

2 3 dF
ffiffi
px ¼0 x ¼ 0 t > 0:
ð
2 Kt
 dx
6 2 x 7
eu
2
w(t) ¼ w0 41  pffiffiffiffi du ¼ w0 1  erf pffiffiffiffi 5:
p 2 Kt This imposes the requirement that B ¼ 0, so that
0

(5:55) rffiffiffi
s
F ¼ A cosh x :
k

Example 5.28
Now condition b is imposed. This requires that
A finite medium of length l is at initial temperature w0. There is rffiffiffi
no heat flow across the boundary at x ¼ 0, and the face at x ¼ l w1 s
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

¼ A cosh :
is then kept at w1 (see Figure 5.14). Determine the tempera- s k
ture w(t).
Thus, by b and c
SOLUTION
pffiffi
Here we have to solve cosh x ks
F ¼ w1 pffiffis :
s cosh k
q2 w 1 qw
¼
qx 2 k qt Now, to satisfy c we have

subject to the boundary conditions: pffi

w0 w0 cosh x ks
F¼  pffiffis :
a: w ¼ w0 t¼0 0xl s s cosh k

b: w ¼ w1 t>0 x¼1
Thus, the final form of the Laplace transformed equation that
qw satisfies all conditions of the problem is
c: ¼0 t>0 x ¼ 0:
qx
pffi
w0 w1  w0 cosh x ks
Upon Laplace transforming the controlling differential equa- F¼ þ pffis :
tion, we obtain s s cosh k

d2 F s To find the expression for w(x, t), we must invert this expres-
 F ¼ 0: sion. That is,
dx 2 k

The solution is ð
sþj1 pffiffi
w1  w0 cosh x ks ds
w(x, t) ¼ w0 þ est pffis : (5:56)
pffis pffis rffiffiffi rffiffiffi 2pj cosh k
s
s s sj1
F ¼ A0 ex k þ B0 ex k ¼ A cosh x þ B sinh x :
k k
The integrand is a single
2 pvalued function of s with poles
2
at s ¼ 0 and s ¼ k 2n1
2 l2 , n ¼ 1, 2, . . . :
Insulator
We select the path of integration that is shown in Figure
5.15. But the inversion integral over the path BCA( ¼ G) ¼ 0.
Thus, the inversion integral becomes

ð
sþj1 pffiffi
Insulator

1 cosh x ks
(t) 1 est pffis ds
k s :
2pj cosh
sj1

x By an application of the Cauchy integral theorem, we require

0 l
the residues of the integrand at its poles. There results
Insulator

FIGURE 5.14 Details for Example 5.28. Resjs¼0 ¼ 1

© 2010 by Taylor and Francis Group, LLC

 5-24 Transforms and Applications Handbook

jω Then Equation 5.57 transforms to

B  2 
d F 1 dF
k þ  sF ¼ 0,
dr 2 r dr
Г

R which we write in the form

rffiffiffi
C d2 F 1 dF s
σ þ  mF ¼ 0, m¼ :
dr 2 r dr k
σ
This is the Bessel equation of order 0 and has the solution

F ¼ AI0 ðmr Þ þ BN0 ðmr Þ:

Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

A However, the Laplace transformed form of C-1 when z ¼ 0

imposes the condition B ¼ 0 because N0(0) is not zero. Thus,
FIGURE 5.15 The path of integration for Example 5.28.
F ¼ AI0 (mr):

 The boundary condition C-2 requires F(r, a) ¼ ws0 when r ¼ a,

ekðn2Þ l2 cosh j n  12 pxl
1 2 p2

Resjs¼k ¼
: hence,
ðn12Þ pl22  pffiffi
2

s dsd cosh l ks s¼k n1 2 p2

ð 2Þ l 2
w0 1
A¼
s I(ma)
Combine these with Equation 5.55 to write finally
so that
4(w1  w0 ) X1
(1)n kðn12Þ2 p2 =l2
w(x, t) ¼ w0 þ l
p n¼1
2n  1 w0 I0 (mr)

   F¼ :
1 s I0 (ma)
 cos n  px=l : (5:57)
2
To find the function w(r, t) requires that we invert this function.
By an application of the inversion integral, we write

Example 5.29 ð
sþj1 rffiffiffi
w lt I0 (jr)
dl l
w(r, t) ¼ 0 e , j¼ : (5:59)
A circular cylinder of radius a is initially at temperature zero. 2pj I0 (ja) l k
sj1
The surface is then maintained at temperature w0. Determine
the temperature of the cylinder at any subsequent time t.
Note that I0(jr)=I0(ja) is a single-valued function of l. To evalu-
ate this integral, we choose as the path for this integration that
SOLUTION shown in Figure 5.16. The poles of this function are at l ¼ 0
The heat conduction equation in radial form is and at the roots of the Bessel function J0(ja) (¼ I0(jja)); these
occur when J0(ja) ¼ 0, with the roots for J0(ja) ¼ 0, namely
l ¼ kj21 ,  kj22 , . . .. The approximations for I0(jr) and I0(ja)
q2 w 1 qw 1 qw show that when n ! 1 the integral over the path BCA tends
þ ¼ , 0  r < a, t > 0: (5:58)
qr 2 r qr k qt to zero. The resultant value of the integral is written in terms of
the residues at zero and when l ¼ kj2n . These are
And for this problem the system is subject to the boundary
conditions Resj¼0 ¼ 1
C-1. w ¼ 0 t ¼ 0 0  r < a  
 ldI0 (ja)
C-2. w ¼ w0 t > 0 r ¼ a. 
Reskj2n ¼ :
dl kj2n
To proceed, we multiply each term in the partial differen-
tial equation by est dt and integrate. We write Therefore,
" #
ð
1
X J0 (jn r)
kj2n t
we st
dt ¼ F(r, s) w(r, t) ¼ w0 1 þ e d
:
n l dl I0 (ja)jl¼kj2n
0

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-25

jω Apply this transformation to both members of Equation 5.62

B
subject to F(0, s) ¼ 0. The result is


qF
s2 N(z, s)  sF(z) ¼ a2 z 2 N(z, s)  (0, s) , F(z) ¼ +{w0 }:
qx

C We denote qF
qx (0, s) by C. Then the solution of this equation is
σ
0
C s 1
N(z, s) ¼ 2  F(z) 2 s2 :
z 2  as 2 a2 z  a2

The inverse transformation with respect to z is, employing

A convolution.
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

ðx
FIGURE 5.16 The path of integration for Example 5.29. aC sx 1 s
F(x, s) ¼ sinh  w(j) sinh (x  j)dj:
s a a a
0

Further, d
l dl I0 (jn a) ¼ 1 (1)
2 jaI0 (ja). Hence, finally,
To satisfy the condition limx!1 F(x, s) ¼ 0 requires that the
" #
sinh terms be replaced by their exponential forms. Thus, the
2X 1
kj2n t J0 (jn r)
w(t) ¼ w0 1 þ e : (5:60) factors
a n¼1 jn J0(1) (jn a)

sx 1 s esj=a
sinh ! , sinh (x  j) ! , x ! 1:
Example 5.30 a 2 a 2

A semi-infinite stretched string is fixed at each end. It is given Then we have the expression
an initial transverse displacement and then released. Deter-
mine the subsequent motion of the string.
ðx
aC 1
F(x, s) ¼  w(j)esj=a dj:
SOLUTION 2s 2a
0
This requires solving the wave equation
But for this function to be zero for x ! 1 requires that
q2 w q2 w
a2 ¼ 2 (5:61)
qx 2 qt
ð
1
aC 1
subject to the conditions ¼ w(j)esj=a dj, x ! 1:
s a
0
C-1. w(x, 0) ¼ f(x) t ¼ 0, w(0, t) ¼ 0 t>0
C-2. limx!1 w(x, t) ¼ 0.
Combine this result with F(x, s) to get
To proceed, multiply both sides of Equation 5.61 by est dt
and integrate. The result is the Laplace-transformed equation
ð
1 ð
1
s(jx)=a
d2 F 2aF(x, s) ¼ w(j)e dj  w(j)es(xþj)=a dj
a2 ¼ s2 F  sw(0þ), x > 0: (5:62)
dx 2 0 0
ðx
C-1. F(0, s) ¼ 0 þ w(j)es(xj)=a dj:
C-2. limx!1 F(x, s) ¼ 0. 0

To solve Equation 5.62 we will carry out a second Laplace

transform, but this with respect to x, that is +{F (x, s)} ¼ N(z, s). Each integral in this expression is integrated by parts. Here
Thus, we write

ð
1
s(jx) a
N(z, s) ¼ F(x, s)ezx dx: u ¼ w(j), du ¼ w(1) (j)dj; dv ¼ e a dj, v ¼ es(jk)=a :
s
0

© 2010 by Taylor and Francis Group, LLC

 5-26 Transforms and Applications Handbook

The resulting integrations lead to Example 5.31

ð
1 ð
1 A stretched string of length l is fixed at each end as shown
1 1 s(jx) 1 s(xþj) in Figure 5.17. It is plucked at the midpoint and then released
F(x, s) ¼ w(x) þ (1)
w (j)e a dj  (1)
w (j)e a dj
s 2s 2s at t ¼ 0. The displacement is b. Find the subsequent motion.
x 0
ð
1
1 s(xj) SOLUTION
 w(1) (j)e a dj:
2s
0 This problem requires the solution of

We note by entry 61, Table A.5.1 that q2 y 2q y

2
¼ c , 0 < y < l, t > 0 (5:64)
qy 2 qt 2
 
1 s(jx)
+1 e a ¼1 when at > j  x
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

s subject to
¼0 when at < j  x:
9
2bx
1: y¼ 0 < x < l=2 >
>
=
l
This function of j vanishes except when j  x þ at. Thus, t¼0
2b 1 >
>
2: y¼ (l  x) <x<l ;
8 1 9 l 2
<1 ð =1
ð
xþat
qy
1 s(jþx)=a ¼0 0<x<l t¼0
+ (1)
w (j)e dj ¼ w(1) (j)dj 3:
:2 ; 2 qt
x x
4: y¼0 x ¼ 0; x¼l t > 0:
1 1
¼ w(x þ at)  w(x):
2 2
To proceed, multiply Equation 5.63 by estdt and integrate in t.
This yields
Proceed in the same way for the term

8 1 9 d2 Y
<1 ð = s2 Y  sy(0) ¼ c2
dx 2
+1 w(1) (j)es(xþj)=a dj ¼1 when at > x þ j
:2 ;
0
or
¼0 when at < x þ j:

d2 Y
Thus, the second term becomes c2  s2 Y ¼ sy(0) ¼ sf (x) (5:65)
dx 2

8 1 9
<1 ð = ð
xþat subject to Y(0, s) ¼ Y(l, s) ¼ 0. To solve this equation, we pro-
1 s(xþj)=a 1 ceed as in Example 5.30; that is, we apply a transformation
+ (1)
w (j)e dj ¼ w(1) (j)dj
:2 ; 2 on x, namely +{Y(x, s)} ¼ N(z, s). Thus,
x x
1

¼ w(x  at): y(0, s)
2 s N(z, s)  sY(0) ¼ c z N(z, s) 
2 2 2
:
x
The final term becomes
This equation yields, writing sY(0) as F(x, s),

 12 w(x) when at > x F(z, s)

N(z, s) ¼
z 2  cs 2
2
 12 w(x) þ 1
2 w(x  at) when at < x:

The final result is

1 b
w(x, t) ¼ [f (at þ x)  f (at  x)] when t > x =a
2
0 l/2 l
1
¼ [f (x þ at) þ f (x  at)] when t < x =a: (5:63)
2 FIGURE 5.17 A stretched string plucked at its midpoint.

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-27

The inverse transform is l 2n  1 p 2n  1

¼j or l ¼ j pc:
2c l 2 l
F(x, s)
Y(x, s) ¼ +1 {N(x, s)} ¼ Thus, we have, by the theory of residues
c sinh cs

x
where Resjl¼0 ¼
c
ðx pcx sinh j(2n  1) pxl
(l  x)s js Resjj2n1pc ¼ e j(2n1)  2 :
F(x, s) ¼ sinh y(j) sinh dj l d ll
dl l cosh 2c j2n1pc
l
c c l
0
ðl These poles lead to
xs (l  j)s
þ sinh y(j) sinh dj:
c c
0 2l sin (2n  1) pxl j(2n1)pct
¼ (1)n e l :
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

p2 c (2n  1)2
Combine these integrals with the known form of f(x) in C-1
and C-2. Upon carrying out the integrations, the resulting
Thus, the poles at j(2n  1)pcl lead to
forms become, with k ¼ cs,

4l sin (2n  1) pxl cos (2n  1) pcl t

lY x c sinh kx ¼ (1)n :
¼  , 0  x  l=2 p2 c (2n  1)2
2b s s2 cosh kl2

lY (l  x) c sinh k(l  x) l Then we have

¼  2 ,  x  l:
2b s s cosh kl2 2
‘y
To find y(t), we must invert these expressions. Note that ¼x
2b
symmetry exists and so we need consider only the first term. ( )
We use the inversion integral on the term s12 sinh kx
. Thus, we 4l X
1
(1)n px pct
coshkl2  xþ 2 sin (2n  1) cos (2n  1)
consider the integral p n¼1 (2n  1)2 l l

ð
sþj1
so that finally
1 sinh xlc dl
I¼ elb 2
:
2pj cosh xl
2c l
8b X
sj1 1
(1)n1 px pct
y¼ sin (2n  1) cos (2n  1) ,
We choose the path in the l-plane as shown in Figure 5.18. The p n¼1 (2n  1)2
2 l l
value of the integral over path G is zero. Thus, the value of the l
integral is given in terms of the residues. These occur at l ¼ 0 0x :
2
and at the values for which cosh 2cl ¼ 0, which exist where
For the string for which 2l  x < l, the corresponding expres-
jω sion is the same except that (l  x) replaces x.
B Note that this equation can be written, with h ¼ (2n  1) pl ,

Г sin h(x  ct) þ sin h(x þ ct)

λ-plane sin hx cos hct ¼ ,
2

which shows the traveling wave nature of the solution.

C
σ

σ
5.8 The Bilateral or Two-Sided
Laplace Transform
A In Section 5.1 we discussed the fact that the region of absolute
convergence of the unilateral or one-sided Laplace transform is
the region to the left of the abscissa of convergence. The situation
FIGURE 5.18 The path of integration for Example 5.31. for the two-sided Laplace transform is rather different; the region

© 2010 by Taylor and Francis Group, LLC

 5-28 Transforms and Applications Handbook

of convergence must be specified if we wish to invert a function For the second signal,
F(s) that was obtained using the bilateral Laplace transform. This
requirement is necessary because different time signals might ð
1 ð0
have the same Laplace transform but different regions of absolute 1
F2 (s) ¼ eat u(t)est dt ¼  e(sþa)t dt ¼
convergence. sþa
1 1
To establish the region of convergence, write the bilateral
Laplace transform in the form
and its region of convergence is Re(s) < a.
Clearly, the knowledge of the region of convergence is
ð
1 ð
1 ð0 necessary to find the time functions unambiguously.
F2 (s) ¼ est f (t)dt ¼ est f (t)dt þ est f (t)dt: (5:66)
1 0 1
Example 5.33
If the function f(t) is of exponential order (es1 t), then the region
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

of convergence for t > 0 is Re(s) > s1. If the function f(t) for t < 0 Find the function f(t), if its Laplace transform is given by
is of exponential order exp(s2t), then the region of convergence
is Re(s) < s2. Hence, the function F2(s) exists and its analytic in 3
F2 (s) ¼ , 2 < Re(s) < 1:
the vertical strip defined by s1 < Re(s) < s2, provided, of course, (s  4)(s þ 1)(s þ 2)
that s1 < s2. If s2 > s1, no region of convergence would exist
and the inversion process could not be performed. This region of
SOLUTION
convergence is shown in Figure 5.19.
The region of convergence and the paths of integration are
shown in Figure 5.20. For t > 0 we close the contour to the left
Example 5.32 and we obtain

Find the bilateral Laplace transform of the signals f(t) ¼ eatu(t) 

and f(t) ¼ eatu(t) and specify their regions of convergence. 3est  1
f (t) ¼  ¼ e2t , t > 0:
(s  4)(s þ 1)s¼2 2
SOLUTION
Using the basic definition of the transform (Equation 5.66), we For t < 0, the contour closes to the right and now
obtain
 
3est  3est  4t
ð
1 ð
1 f (t) ¼  þ  ¼  3 et þ e ,
1 (s  4)(s þ 2)s¼1 (s þ 1)(s þ 2)s¼4 5 10
F2 (s) ¼ eat u(t)est dt ¼ e(sþa)t dt ¼
sþa t < 0:
1 0

and its region of convergence is Re(s) > a.

jω
jω

σ + jω
t>0 t<0

s-Plane

σ
σ –3 –2 –1 1 2 3 4
σ1 0 σ σ2

σ – jω

FIGURE 5.19 Region of convergence for the bilateral transform. FIGURE 5.20 Illustrating Example 5.33.

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-29

Appendix A

TABLE A.5.1 Laplace Transform Pairs

F(s) f(t)
n (n)
1 s d (t) nth derivative of the delta function
dd(t)
2 s
dt
3 1 d(t)
4 1 1
s
1
5 t
s2
t n1
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

1
6 (n ¼ 1, 2, . . . )
sn (n  1)
1 1
7 pffiffi pffiffiffiffiffi
s pt
rffiffiffiffi
2 t
8 s3=2
p
2n t n(1=2)
9 s[nþ(1=2)] (n ¼ 1, 2, . . . ) pffiffiffiffi
1  3  5    (2n  1) p
G(k)
10 (k  0) t k1
sk
1
11 eat
sa
1
12 teat
(s  a)2
1 1
13 (n ¼ 1, 2, . . . ) t n1 eat
(s  a)n (n  1)
G(k)
14 (k  0) t k1 eat
(s  a)k
1 1
15 (eat  ebt )
(s  a)(s  b) (a  b)
s 1
16 (aeat  bebt )
(s  a)(s  b) (a  b)
1 (b  c)eat þ (c  a)ebt þ (a  b)ect
17 
(s  a)(s  b)(s  c) (a  b)(b  c)(c  a)
1
18 eat valid for complex a
(s þ a)
1 1
19 (1  eat )
s(s þ a) a
1 1 at
20 ðe þ at  1Þ
s2 (s þ a) a2


1 1 1 at 2 1 at
21  t þ  e
s3 (s þ a) a2 a 2 a
1 1  at
22 e  ebt
(s þ a)(s þ b) (b  a)


1 1 1  at
23 1þ be  aebt
s(s þ a)(s þ b) ab (a  b)


1 1 1  2 bt 2 at

24 a e  b e þ abt  a  b
s2 (s þ a)(s þ b) (ab)2 (a  b)

3  
1 1 a  b3 1 2 (a þ b) 1 b at a bt
25 þ t  t þ e  e
s3 (s þ a)(s þ b) (ab) (ab)2 (a  b) 2 ab (a  b) a2 b2
1 1 1 1
26 eat þ ebt ect
(s þ a)(s þ b)(s þ c) (b  a)(c  a) (a  b)(c  b) (a  c)(b  c)
1 1 1 1 1
27  eat  ebt  ect
s(s þ a)(s þ b)(s þ c) abc a(b  a)(c  a) b(a  b)(c  b) c(a  c)(b  c)

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-30 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
8
> ab(ct  1)  ac  bc 1
>
< þ 2 eat
28 1 (abc)2 a (b  a)(c  a)
s2 (s þ a)(s þ b)(s þ c) >
> þ 1 1
: ebt þ 2 ect
b2 (a  b)(c  b) c (a  c)(b  c)
8
> 1  ab þ ac þ bc 1 2
>
< (ab þ ac þ bc)2  abc(a þ b þ c)  tþ t
29 1 (abc)3 (abc)2 2abc
s3 (s þ a)(s þ b)(s þ c) >
>  1 1 1
: eat  3 ebt  3 ect
a3 (b  a)(c  a) b (a  b)(c  b) c (a  c)(b  c)
1 1
30 sin at
s2 þ a2 a
s
31 cos at
s2 þ a2
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

1 1
32 sinh at
s2  a2 a
s
33 cosh at
s2  a2
1 1
34 (1  cos at)
s(s2 þ a2 ) a2
1 1
35 (at  sin at)
s2 (s2 þ a2 ) a3
1 1
36 ( sin at  at cos at)
(s2 þ a2 )2 2a3
s t
37 sin at
(s2 þ a2 )2 2a
s2 1
38 ( sin at þ at cos at)
(s2 þ a2 )2 2a
s2  a2
39 t cos at
(s2 þ a2 )2
s cos at  cos bt
40 (a2 6¼ b2 )
(s2 þ a2 )(s2 þ b2 ) b2  a2
1 1 at
41 e sin bt
(s  a)2 þ b2 b
sa
42 eat cos bt
(s  a)2 þ b2
n  
1 eat X 2n  r  1 dr
43 (2t)r1 r [ cos (bt)]
[(s þ a)2 þ b2 ]n 4n1 b2n r¼1 n1 dt
8 ( !
>
> eat X n 2n  r  1 dr
>
> (2t)r1 r [a cos (bt) þ b sin (bt)]
>
<4 bn1 2n
n1 dt
s r¼1
44  n ! )
(s þ a)2 þ b2 >
> Xn1 2n  r  2 r
>
> r1 d
>
: 2b r (2t) [ sin (bt)]
r¼1 n1 dt r
 pffiffiffi pffiffiffi
3a2 at 3 pffiffiffi at 3
45 eat  e(at)=2 cos  3 sin
s3 þ a3 2 2
4a3
46 sin at cosh at  cos at sinh at
s4 þ 4a4
s 1
47 ( sin at sinh at)
s4 þ 4a4 2a2
1 1
48 ( sinh at  sin at)
s4  a4 2a3
s 1
49 ( cosh at  cos at)
s4  a4 2a2
8a3 s2
50 (1 þ a2 t 2 ) sin at  cos at
(s2 þ a2 )3
 
1 s1 n et dn n t
51 Ln (t) ¼ (t e )
s s n! dt n

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-31

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
[Ln (t) is the Laguerre polynomial of degree n]
1 t (n1) eat
52 where n is a positive integer
(s þ a)n (n  1)!
1 1
53 [1  eat  ateat ]
s(s þ a)2 a2
1 1
54 [at  2 þ ateat þ 2eat ]
s2 (s þ a)2 a3

  
1 1 1 2 2 at
55 1  a t þ at þ 1 e
s(s þ a)3 a3 2
1 1  at 
56 e þ ½(a  b)t  1ebt
(s þ a)(s þ b)2 (a  b)2


a  2b
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

1 1 1 at 1
57  e  t þ ebt
s(s þ a)(s þ b)2 ab2 a(a  b)2 b(a  b) b2 (a  b)2
 

1 1 at 1 1 2 1 2(a  b)  b bt
58 e þ t   þ t þ e
s2 (s þ a)(s þ b)2 a2 (a  b)2 ab2 a b b2 (a  b) b3 (a  b)2
8

> 1 2c  a  b
>
< (c  b)(c  a) t þ ect
1 (c  a)2 (c  b)2
59
(s þ a)(s þ b)(s þ c)2 >
> þ 1 1
: eat þ ebt :
(b  a)(c  a)2 (a  b)(c  b)2
1 1 1 v
60 eat þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sin (vt  f); f ¼ tan1
(s þ a)(s2 þ v2 ) a2 þ v2 v a þv 2 2 a
 
1 1 1 1 a 1
61  sin vt þ 2 cos vt þ eat
s(s þ a)(s2 þ v2 ) av2 a2 þ v2 v v a
8
> 1 1 1 at
>
< av2 t  a2 v2 þ a2 (a2 þ v2 ) e
1
62
> a
s2 (s þ a)(s2 þ v2 ) > 1
: þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cos (vt þ f); f ¼ tan1
v a þv
3 2 2 v
1 1 at
63  2 e [ sin vt  vt cos vt]
(s þ a)2 þ v2 2v3
1 1
64 sinh at
s2  a2 a
1 1 1
65 sinh at  2 t
s2 (s2  a2 ) a3 a
1 1 1
66 ( cosh at  1)  2 t 2
s3 (s2  a2 ) a4 2a

 pffiffiffi pffiffiffi 
1 1 at a 3 pffiffiffi 3
67 e  e 2t cos at  3 sin at
s3 þ a3 3a2 2 2
1 1
68 ( sin at cosh at  cos at sinh at)
s4 þ 4a4 4a3
1 1
69 ( sinh at  sin at)
s4  a4 2a3
1 1 at
70 e sinh vt
[(s þ a)2  v2 ] v
8 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
>
> (a  b)2 þ v2 bt
> a
< 1
sþa  þ e sin (vt þ f);
71 b2 þ v2 v b2 þ v2
s[(s þ a)2 þ v2 ] >
>    v 
>
: f ¼ tan1 v þ tan1
b ab
8 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
>
> 1 2ab (a  b)2 þ v2 bt
sþa < [1 þ at]  þ e sin (vt þ f)
72 b2 þ v2 (b þ v )
2 2 2
v(b2 þ v2 )
s2 [(s þ b)2 þ v2 ] >
>  v  v 
: f ¼ tan1 þ 2 tan1
ab b

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-32 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
8 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
>
> ac (a  b)2 þ v2 bt
>
< ect þ þ
1
e sin (vt þ f)
sþa (c  b) þ v
2 2 v (c  b)2 þ v2
73
(s þ c)[(s þ b)2 þ v2 ] >
>    
>
: f ¼ tan1 v v
 tan1
ab cb
8
>
> a (c  a)
>
> þ ect
>
> c(b2 þ v2 ) c[(b  c)2 þ v2 ]
>
> sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sþa <
74 1 (a  b)2 þ v2 bt
s(s þ c)[(s þ b)2 þ v2 ] >  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi e sin (vt þ f)
>
> v b þv
2 2 (b  c)2 þ v2
>
>
>
>    v   v 
>
: f ¼ tan1
v
þ tan1  tan1
b ab cb


sþa b  3a 3a  b a  b 2 2a  b bt
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

a
75 tþ þ þ t þ t e
s2 (s þ b)3 b3 b4 b4 2b2 b3


sþa a  c ct ab 2 ca a  c bt
76 3e þ t þ 2tþ 3 e
(s þ c)(s þ b)3 (b  c) 2(c  b) (c  b) (c  b)
s2 a2 b2 c2
77 eat þ ebt þ ect
(s þ a)(s þ b)(s þ c) (b  a)(c  a) (a  b)(c  b) (a  c)(b  c)

2 
s2 a2 at b b2  2ab bt
78 2e þ tþ 2 e
(s þ a)(s þ b)2 (b  a) (a  b) (a  b)


s2 a2 2 at
79 2  2at þ t e
(s þ a)3 2
s2 a 2
v v
80 eat  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sin (vt þ f); f ¼ tan1
(s þ a)(s2 þ v2 ) (a2 þ v2 ) a þv
2 2 a
8

>
< a2 2av2 v
s2 t 2 eat  2 sin (vt þ f);
81 (a2 þ v2 ) (a þ v2 )2 (a þ v2 )
(s þ a) (s2 þ v2 )
2 >
: 
f ¼ 2 tan1 va
8
> a2 b2
s2 < eat þ ebt
82 (b  a)(a þ v )
2 2 (a  b)(b2 þ v2 ) h v vi
(s þ a)(s þ b)(s2 þ v2 ) > v
:  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sin (vt þ f); f ¼  tan1 þ tan1
(a2 þ v2 )(b2 þ v2 ) a b
s2 a v
83  sin (at)  2 sin (vt)
(s2 þ a2 )(s2 þ v2 ) (v2  a2 ) (a  v2 )
s2 1
84 ( sin vt þ vt cos vt)
(s2 þ v2 )2 2v
8 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
>
> a2 1 (b2  v2 )2 þ 4b2 v2 bt
>
> at
þ e sin (vt þ f)
s2 < e
85 (a  b)2 þ v2 v (a  b)2 þ v2
(s þ a)[(s þ b)2 þ v2 )] >    v 
>
> 2bv
>
: f ¼ tan1 2  tan1
b v 2 ab
8

>
> a 2
at a[(b  a) 2
þ v2 ] þ a2 (b  a) at
>
> te  2 e
> (a  b) þ v2
> 2
[(b  a) þ v2 ]2 2
>
>
>
> s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s2 <
(b2  v2 )2 þ 4b2 v2 bt
> þ e sin (vt þ f)
86
(s þ a) [(s þ b)2 þ v2 ]
2
>
> v[(a  b)2 þ v2 ]
>
>
>
>    v 
>
> 2bv
>
: f ¼ tan
1
 2 tan1
b v
2 2 ab
s2 þ a b2 þ a bt a a
87 e þ t 2
þ b)
s2 (s b2 b b
s2 þ a a 2 a 1
88 t  2 t þ 3 b2 þ a  (a þ b2 )ebt
þ b)
s3 (s 2b b b
s2 þ a a (b2 þ a) bt (c2 þ a) ct
89 þ e  e
s(s þ b)(s þ c) bc b(b  c) c(b  c)

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-33

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
s þa2
b2 þ a bt c2 þ a ct a a(b þ c)
90 e þ 2 e þ t 2 2
s2 (s þ b)(s þ c) b2 (c  b) c (b  c) bc bc
s2 þ a b2 þ a c2 þ a d2 þ a
91 ebt þ ect þ edt
(s þ b)(s þ c)(s þ d) (c  b)(d  b) (b  c)(d  c) (b  d)(c  d)
s2 þ a a b2 þ a c2 þ a d2 þ a
92 þ ebt þ ect þ edt
s(s þ b)(s þ c)(s þ d) bcd b(b  c)(d  b) c(b  c)(c  d) d(b  d)(d  c)
8
>
> a a b2 þ a
>
< t  2 2 2 (bc þ cd þ db) þ 2 ebt
s2 þ a bcd bc d b (b  c)(b  d)
93
s2 (s þ b)(s þ c)(s þ d) >
> c2 þ a d2 þ a
>
: þ 2 ect þ 2 edt
c (c  b)(c  d) d (d  b)(d  c)
s2 þ a 1 1
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

94 (a þ v2 ) sin vt  2 (a  v2 )t cos vt
(s2þ v2 )2 2v3 2v
s2  v2
95 t cos vt
(s2 þ v2 )2
s þa
2
a (a  v2 ) a
96  t sin vt  4 cos vt
s(s2 þ v2 )2 v4 2v3 v

2 
s(s þ a) b2  ab bt c  ac c2  2bc þ ab ct
97 2e þ tþ e
(s þ b)(s þ c)2 (c  b) bc (b  c)2
8
>
> b2  ab c2  ac d2  ad
>
< ebt þ ect þ tedt
s(s þ a) (c  b)(d  b)2 (b  c)(d  c)2 (b  d)(c  d)
98
(s þ b)(s þ c)(s þ d)2 >
> þ a(bc  d ) þ d(db þ dc  2bc) edt :
2
>
:
(b  d)2 (c  d)2
s2 þ a1 s þ a0 b2  a1 b þ a0 bt a0 a1 b  a0
99 e þ tþ
s2 (s þ b) b2 b b2
s2 þ a1 s þ a0 a1 b  b2  a0 bt a0 2 a1 b  a0 b2  a1 b þ a0
100 e þ t þ tþ
s3 (s þ b) b3 2b b2 b3
s2 þ a1 s þ a0 a0 b2  a1 b þ a0 bt c2  a1 c þ a0 ct
101 þ e þ e
s(s þ b)(s þ c) bc b(b  c) c(c  b)
s2 þ a1 s þ a0 a0 a1 bc  a0 (b þ c) b2  a1 b þ a0 bt c2  a1 c þ a0 ct
102 tþ þ e þ 2 e
þ b)(s þ c)
s2 (s bc b2 c2 b2 (c  b) c (b  c)
s2 þ a1 s þ a0 b2  a1 b þ a0 bt c2  a1 c þ a0 ct d 2  a1 d þ a0 dt
103 e þ e þ e
(s þ b)(s þ c)(s þ d) (c  b)(d  b) (b  c)(d  c) (b  d)(c  d)
s2 þ a1 s þ a0 a0 b2  a1 b þ a0 bt c2  a1 c þ a0 ct d2  a1 d þ a0 dt
104  e  e  e
s(s þ b)(s þ c)(s þ d) bcd b(c  b)(d  b) c(b  c)(d  c) d(b  d)(c  d)
s2 þ a1 s þ a0 a0 b2  a1 b þ a0 bt b2  a0 bt
105  te þ e
s(s þ b)2 b2 b b2
s2 þ a1 s þ a0 a0 a1 b  2a0 b2  a1 b þ a0 bt 2a0  a1 b bt
106 tþ þ te þ e
s2 (s þ b)2 b2 b3 b2 b3
s2 þ a1 s þ a0 b2  a1 b þ a0 bt c2  a1 c þ a0 ct c2  2bc þ a1 b  a0 ct
107 e þ te þ e
(s þ b)(s þ c)2 (c  b)2 (b  c) (b  c)2
8
>
> b3 bt c3 ct d3 dt
>
< (b  c)(d  b)2 e þ (c  b)(d  c)2 e þ (d  b)(c  d) te
s3
108
(s þ b)(s þ c)(s þ d)2 >
> d2 ½d 2  2d(b þ c) þ 3bc dt
>
: þ e
(b  d)2 (c  d)2
8
> b3 c3
>
> 2e
bt
þ ect
>
>      c)(f  c)2
>
>
(b c)(d b)(f b) (c b)(d
>
>
>
> d3 f3
> dt ft
s3 < þ (d  b)(c  d)(f  d)2 e þ (f  b)(c  f )(d  f ) te
>
109

(s þ b)(s þ c)(s þ d)(s þ f )2 >

> 3f 2
>
> þ
>
>   f )(d  f )
>
> (b f )(c
>
> 
>
> f 3
½(b  f )(c  f ) þ (b  f )(d  f ) þ (c  f )(d  f ) dt
>
: þ e :
(b  f ) (c  f ) (d  f )
2 2 2

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-34 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
3
s b3 b2 (3c  b) bt c3 c2 (3b  c) ct
110  bt
þ  ct
þ
(s þ b)2 (s þ c)2 2 te 3 e 2 te e
(c  b) (c  b) (b  c) (b  c)3
8
>
> d3 b3
>
>  edt þ tebt
>
> (b  d) (c  d)
2 2
(c  b)2 (b  d)
>
>

s3 < 3b2 b3 (c þ 2d  3b) bt c3
111 þ þ 2 e þ tect
(s þ d)(s þ b)2 (s þ c)2 >
> (c  b) 2
(d  b) (c  b) 3
(d  b) (b  c)2
(c  d)
>
>

>
> 3c2 c3 (b þ 2d  3c) ct
>
>
: þ þ e
(b  c)2 (d  c) (b  c)3 (d  c)2
8
>
> b3 c3
>
> ebt þ ect
> (b  c)(b2 þ v2 )
> (c  b)(c 2 þ v2 )
>
<
s3 v2
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

112  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi sin (vt þ f)
(s þ b)(s þ c)(s2 þ v2 ) >
> (b þ v2 )(c2 þ v2 )
2
>
>
>
>   v
>
: f ¼ tan1
c
 tan1
v b
8
>
> b 3
bt c3 ct
>
> (b  c)(d  b)(b2 þ v2 ) e þ (c  b)(d  c)(c2 þ v2 ) e
>
>
>
>
>
> d3
>
>
s3 < þ edt
113 (d  b)(c  d)(d 2 þ v2 )
(s þ b)(s þ c)(s þ d)(s2 þ v2 ) >
> v2
>
>  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi cos (vt  f)
>
>
>
> (b2 þ v2 )(c2 þ v2 )(d2 þ v2 )
>
> v v v
>
>
: f ¼ tan1 þ tan1 þ tan1
b c d
8
>
> b 3
b 2 2
(b þ 3v 2
) v2
> 2
< t ebt þ ebt  2 sin (vt þ f)
s3 b þ v2 (b2 þ v2 )2 (b þ v2 )
114    
(s þ b)2 (s2 þ v2 ) >
> b v
>
: f ¼ tan1  tan1
v b
s3
115 cos (vt) cosh (vt)
s4 þ 4v4
s3 1
116 [ cosh (vt) þ cos (vt)]
s4  v4 2
8
>
> a a (b þ c)  a1 bc b3 þ a2 b2  a1 b þ a0 bt
> 0t 0
< þ e
s3 þ a2 s2 þ a1 s þ a0 bc b2 c2 b2 (c  b)
117
s2 (s þ b)(s þ c) >
> c þ a2 c  a1 c þ a0 ct
3 2
>
: þ e
c2 (b  c)
8
>
> a0 b3 þ a2 b2  a1 b þ a0 bt c3 þ a2 c2  a1 c þ a0 ct
>
< bcd  e  e
s3 þ a2 s2 þ a1 s þ a0 b(c  b)(d  b) c(b  c)(d  c)
118
s(s þ b)(s þ c)(s þ d) >
> d3 þ a2 d2  a1 d þ a0 dt
>
:  e
d(b  d)(c  d)
8

>
> a a a (bc þ bd þ cd) b3 þ a2 b2  a1 b þ a0 bt
> 0 tþ 1  0
< þ e
s3 þ a2 s2 þ a1 s þ a0 bcd bcd 2
bcd2 2 b2 (c  b)(d  b)
119
þ b)(s þ c)(s þ d)
s2 (s >
> c þ a2 c  a1 c þ a0 ct d þ a2 d  a1 d þ a0 dt
3 2 3 2
>
: þ e þ e
c2 (b  c)(d  c) d2 (b  d)(c  d)
8 3
>
> b þ a2 b2  a1 b þ a0 bt c3 þ a2 c2  a1 c þ a0 ct
>
< (c  b)(d  b)(f  b) e þ (b  c)(d  c)(f  c) e
s3 þ a2 s2 þ a1 s þ a0
120
(s þ b)(s þ c)(s þ d)(s þ f ) >
> d3 þ a2 d2  a1 d þ a0 dt f 3 þ a2 f 2  a1 f þ a0 ft
>
: þ e þ e
(b  d)(c  d)(f  d) (b  f )(c  f )(d  f )
8
>
> a0 b3 þ a2 b2  a1 b þ a0 bt c3 þ a2 c2  a1 c þ a0 ct
>
<  e  e
s3 þ a2 s2 þ a1 s þ a0 bcdf b(c  b)(d  b)(f  b) c(b  c)(d  c)(f  c)
121
s(s þ b)(s þ c)(s þ d)(s þ f ) >
> d þ a2 d  a1 d þ a0 dt f þ a2 f  a1 f þ a0 ft
3 2 3 2
>
:  e  e
d(b  d)(c  d)(f  d) f (b  f )(c  f )(d  f )

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-35

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
8
122 s3 þ a2 s2 þ a1 s þ a0 >
> b þ a2 b  a1 b þ a0
3 2
c3 þ a2 c2  a1 c þ a0
>
> ebt þ ect
(s þ b)(s þ c)(s þ d)(s þ f )(s þ g) >
> (c  b)(d  b)(f  b)(g  b) (b  c)(d  c)(f  c)(g  c)
>
>
< d3 þ a2 d2  a1 d þ a0 f 3 þ a2 f 2  a1 f þ a0
þ edt þ eft
>
> (b  d)(c  d)(f  d)(g  d) (b  f )(c  f )(d  f )(g  f )
>
>
>
> g 3 þ a2 g 2  a1 g þ a0
>
>
: þ egt
(b  g)(c  g)(d  g)(f  g)
8 3
> b þ a2 b2  a1 b þ a0 bt c3 þ a2 c2  a1 c þ a0 ct
>
> e þ e
>
> (c  b)(d  b)2 (b  c)(d  c)2
>
>
>
>
>
< þ d þ a2 d  a1 d þ a0 tedt
3 2
s3 þ a2 s2 þ a1 s þ a0
123 (b  d)(c  d)
(s þ b)(s þ c)(s þ d)2 >
>
>
> a0 (2d  b  c) þ a1 (bc  d2 )
>
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

>
>
>
> þa d(db þ dc  2bc) þ d2 (d2  2db  2dc þ 3bc) dt
>
: þ 2 e
(b  d)2 (c  d)2
8
> a0 b3 þ a2 b2  a1 b þ a0 bt c3 þ a2 c2  a1 c þ a0 ct
>
>  e  e
>
> bcd 2
b(c  b)(d  b)2 c(b  c)(d  c)2
>
<
s3 þ a2 s2 þ a1 s þ a0 d þ a2 d  a1 d þ a0 dt 3d  2a2 d þ a1 dt
3 2 2
124  te  e
s(s þ b)(s þ c)(s þ d)2 >
> d(b  d)(c  d) d(b  d)(c  d)
>
>
>
> (d þ a2 d  a1 d þ a0 )½(b  d)(c  d)  d(b  d)  d(c  d) dt
3 2
:  e
d 2 (b  d)2 (c  d)2
8 3
>
> b þ a2 b2  a1 b þ a0 bt c3 þ a2 c2  a1 c þ a0 ct
> (c  b)(d  b)(f  b)2 e þ (b  c)(d  c)(f  c)2 e
>
>
>
>
>
>
>
< þ d þ a2 d  a1 d þ a0 edt þ f þ a2 f  a1 f þ a0 teft
> 3 2 3 2
s3 þ a2 s2 þ a1 s þ a0
125 (b  d)(c  d)(f  d)2 (b  f )(c  f )(d  f )
(s þ b)(s þ c)(s þ d)(s þ f )2 >
>
>
> (f 3 þ a2 f 2  a1 f þ a0 )[(b  f )(c  f ):
>
>
>
>
>
> 3f 2  2a2 f þ a1 þ(b  f )(d  f ) þ (c  f )(d  f )] ft
>
: þ eft  e
(b  f )(c  f )(d  f ) (b  f )2 (c  f )2 (d  f )2
s 1
126 pffiffiffiffiffi eat (1 þ 2at)
(s  a)3=2 pt
pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffi 1
127 sa sb pffiffiffiffiffiffiffi (ebt  eat )
2 pt 3
1 1 2 pffiffi
128 pffiffi pffiffiffiffiffi  aea t erfc(a t )
sþa pt
pffiffi
s 1 2  pffiffi
129 pffiffiffiffiffi þ aea t erf a t
s  a2 pt pffi
pffiffi að t
s 1 2a a2 t 2
130 pffiffiffiffiffi  pffiffiffiffi e el dl
s þ a2 pt p
0
1 1 a2 t  pffiffi
131 pffiffi e erf a t
sðs  a2 Þ a pffi
að t
1 2 a2 t 2
132 pffiffi pffiffiffiffi e el dp
sðs þ a2 Þ a p
0
b2  a2 pffiffi pffiffi
pffiffi
2 2
133 ea t [b  a erf (a t )]  beb t erfc(b t )
(s  a2 )(b þ s)
1 pffiffi
pffiffi pffiffi
2
134 ea t erfc(a t )
sð s þ aÞ
1 1 pffiffiffiffiffiffiffiffiffiffiffipffiffi
135 pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffi eat erf b  a t
(s þ a) s þ b ba


b2  a2 a2 t b
pffiffi 2 pffiffi
136 pffiffi pffiffi e erf (a t )  1 þ eb t erfc(b t )
s(s  a2 )( s þ b) a
8
> n! pffiffi
>
> pffiffiffiffiffi H2n t
(1  s)n < (2n)! pt
137

snþ(1=2) >
> 2 d
n
>
: Hn (t) ¼ Hermite polynomial ¼ ex (ex )
2

dx n

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-36 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
(1  s) n n! pffiffi
138  pffiffiffiffi H2nþ1 t
snþ(3=2) p(2n þ 1)!
pffiffiffiffiffiffiffiffiffiffiffiffiffi (
s þ 2a aeat [I1 (at) þ I0 (at)]
139 pffiffi  1
s [In (t) ¼ jn Jn (jt) where Jn is Bessel0 s function of the first kind]
 
1 ab
140 pffiffiffiffiffiffiffiffiffiffipffiffiffiffiffiffiffiffiffiffi e(1=2)(aþb)t I0 t
sþa sþb 2
   
G(k) pffiffiffiffi t k(1=2) ab
141 (k  0) p e(1=2)(aþb)t Ik(1=2) t
(s þ a)k (s þ b)k ab 2

   
1 ab ab
142 te(1=2)(aþb)t I0 t þ I1 t
(s þ a)1=2 (s þ b)3=2 2 2
pffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffi
s þ 2a  s 1 at
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

143 pffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffi e I1 (at)

s þ 2a þ s t
 
(a  b)k k (1=2)(aþb)t ab
144 pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffi 2k (k > 0) t
e Ik
2
t
sþaþ sþb
pffiffiffiffiffiffiffiffiffiffi pffiffi 2n  
sþaþ s 1 (1=2)(at) 1
145 pffiffipffiffiffiffiffiffiffiffiffiffi e In at
s sþa an 2
1
146 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi J0 (at)
s2 þ a2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi n
s2 þ a2  s
147 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi (n > 1) an Jn (at)
s2 þ a2
pffiffiffiffi  k(1=2)
1 p t
148 (k > 0) Jk(1=2) (at)
(s2 þ a2 )k G(k) 2a
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi k kak
149 s2 þ a2  s (k > 0) Jk (at)
 pffiffiffiffiffiffiffiffiffiffiffiffiffiffin t
s  s2  a 2
150 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi (n > 1) an In (at)
s2  a2
pffiffiffiffi  k(1=2)
1 p t
151 (k > 0) Ik(1=2) (at)
ðs2  a2 Þk G(k) 2a
ðy
1 pffiffi 2
eu du
2
152 pffiffiffiffiffiffiffiffiffiffi erf t ; erf (y)D the error function ¼ pffiffiffiffi
s sþ1 p
0
1
153 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi J0 (at); Bessel function of 1st kind, zero order
s2 þ a2
1 J1 (at)
154 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; J1 is the Bessel function of 1st kind, 1st order
s2 þ a2 þ s at
1 N JN (at)
155 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi N ; N ¼ 1, 2, 3, . . . , JN is the Bessel function of 1st kind, Nth order
s2 þ a2 þ s aN t
ð
t
1 N JN (au)
156 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi N N
du; N ¼ 1, 2, 3, . . . , JN is the Bessel function of 1st kind, Nth order
s s2 þ a 2 þ s a u
0
1 1
157 pffiffiffiffiffiffiffiffiffiffiffiffiffiffipffiffiffiffiffiffiffiffiffiffiffiffiffiffi J1 (at); J1 is the Bessel function of 1st kind, 1st order
s2 þ a2 s2 þ a2 þ s a
1 1
158 pffiffiffiffiffiffiffiffiffiffiffiffiffiffipffiffiffiffiffiffiffiffiffiffiffiffiffiffi N JN (at); N ¼ 1, 2, 3, . . . , JN is the Bessel function of 1st kind, Nth order
s2 þ a2 s2 þ a2 þ s aN
1
159 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi I0 (at); I0 is the modified Bessel function of 1st kind, zero order
s2  a2

eks 0 when 0 < t < k
160 Sk (t) ¼
s 1 when t > k

eks 0 when 0 < t < k
161
s2 tk when t > k
8
< 0 when 0 < t < k
eks
162 (m > 0) (t  k)m1
sm : when t > k
G(m)

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-37

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
ks

1e 1 when 0 < t < k
163
s 0 when t > k
1 1 þ coth 12 ks
164 ¼ S(k, t) ¼ {n when (n  1)k < t < nk (n ¼ 1, 2, . . . ):
s(1  e )
ks 2s 8
1 < 0 when 0 < t < k
>
165 Sk (t) ¼ 1 þ a þ a2 þ    þ an1
sðeþks  aÞ >
:
when nk < t < (n þ 1)k (n ¼ 1, 2, . . . )
8
>
< M(2k, t) ¼ (1)n1
1
166 tanh ks when 2k(n  1) < t < 2nk
s >
:
(n ¼ 1, 2, . . . )

1 1  (1)n
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

1 1
167 M(k, t) þ ¼ when (n  1)k < t < nk
s(1 þ eks ) 2 2 2

1 H(2k, t) [H(2k, t) ¼ k þ (r  k)(1)n where t ¼ 2kn þ r;
168 tanh ks
s2 0  r  2k; n ¼ 0, 1, 2, . . . ]
1
169 f2S(2k, t þ k)  2 ¼ 2(n  1) when (2n  3)k < t < (2n  1)k (t > 0)
s sinh ks
1
170 fM(2k, t þ 3k) þ 1 ¼ 1 þ (1)n when (2n  3)k < t < (2n  1)k (t > 0)
s cosh ks
1
171 coth ks f2S(2k, t)  1 ¼ 2n  1 when 2k(n  1) < t < 2kn
s
k ps
172 coth jsin kt j
s 2 þ k2 2k

1 sin t when (2n  2)p < t < (2n  1)p
173
(s2 þ 1)(1  eps ) 0 when (2n  1)p < t < 2np
1 k=s  pffiffiffiffi
174 e J0 2 kt
s
1 1 pffiffiffiffi
175 pffiffi ek=s pffiffiffiffiffi cos 2 kt
s pt
1 k=s 1 pffiffiffiffi
176 pffiffi e pffiffiffiffiffi cosh 2 kt
s pt
1 k=s 1 pffiffiffiffi
177 e pffiffiffiffiffiffi sin 2 kt
s3=2 pk
1 k=s 1 pffiffiffiffi
178 e pffiffiffiffiffiffi sinh 2 kt
s3=2 pk
1 k=s  t (m1)=2  pffiffiffiffi
179 e (m > 0) Jm1 2 kt
sm k
1 k=s  t (m1)=2  pffiffiffiffi
180 e (m > 0) Im1 2 kt
sm k
 2
pffi k k
181 ek s (k > 0) pffiffiffiffiffiffiffi exp 
2 pt 3 4t
1 kpffis  
182 e (k  0) erfc pk ffi
2 t
s
pffi
 2
1 1 k
183 pffiffi ek s (k  0) pffiffiffiffiffi exp 
s pt 4t
rffiffiffiffi  2  
pffi t k k
184 s3=2 ek s (k  0) 2 exp   k erfc pffiffi
p 4t 2 t
pffi
aek s  pffiffi   
kffi
pffiffi þ erfc 2pk ffit
2
185 (k  0) eak ea t erfc a t þ 2p
s(a þ s) t
pffi
ek s  pffiffi 
pffiffi pffiffi
2
kffi
186 (k  0) eak ea t erfc a t þ 2p
sða þ sÞ t

pffiffiffiffiffiffiffiffiffi (
ek s(sþa) 0  pffiffiffiffiffiffiffiffiffiffiffiffiffiffi when 0 < t < k
187 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s(s þ a) e(1=2)at I0 2a t  k
1 2 2 when t > k

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-38 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
pffiffiffiffiffiffiffiffiffi (
k s2 þa2 0 when 0 < t < k
188 e
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
(s2 þ a2 ) J0 a t 2  k2 when t > k
pffiffiffiffiffiffiffiffiffi (
ek s a
2 2 0 when 0 < t < k
189 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
(s  a )
2 2 I0 a t 2  k2 when t > k
pffiffiffiffiffiffiffiffiffi
ekð s þa sÞ  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2

190 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (k  0) J0 a t 2 þ 2kt

(s2 þ a2 ) 8
pffiffiffiffiffiffiffiffiffi <0 when 0 < t < k
 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
eks  ek s þa
2 2
191 ak
: pffiffiffiffiffiffiffiffiffiffiffiffiffiffi J1 a t  k2 2 when t > k
t 2  k2
8
pffiffiffiffiffiffiffiffiffi <0 when 0 < t < k
 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ek s2 þa2
 eks ak
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

192
: 2p ffiffiffiffiffiffiffiffiffiffiffiffiffiffi I1 a t 2  k 2 when t > k
t  k2
pffiffiffiffiffiffiffiffiffi
8
< 0 when 0 < t < k
an ek s a 
2 2

193 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffipffiffiffiffiffiffiffiffiffiffiffiffiffiffi n (v > 1) t  k (1=2)n  pffiffiffiffiffiffiffiffiffiffiffiffiffiffi 

(s2 þ a2 ) s2 þ a2 þ s : Jn a t 2  k2 when t > k
tþk
1
194 log s G0 (1)  log t [G0 (1) ¼ 0:5772]
s
 
1 G0 (k) log t
195 log s (k > 0) t k1
sk [G(k)]2 G(k)
log s
196 (a > 0) eat ½log a  Ei(at)
sa
log s
197 cos tSi(t)  sin tCi(t)
s2 þ 1
s log s
198 sin t Si(t)  cos t Ci(t)
s2 þ 1
1  t
199 log (1 þ ks) (k > 0) Ei 
s k
sa 1 bt
200 log (e  eat )
sb t
1 t 
201 log (1 þ k2 s2 ) 2Ci
s k
1
202 log (s2 þ a2 ) (a > 0) 2 log a  2Ci(at)
s
1 2
203 log (s2 þ a2 ) (a > 0) ½at log a þ sin at  atCi(at)
s2 a
s2 þ a2 2
204 log (1  cos at)
s2 t
s2  a2 2
205 log (1  cosh at)
s2 t
k 1
206 arctan sin kt
s t
1 k
207 arctan Si(kt)
s s
 
1 t2
pffiffiffiffi exp  2
2 2
208 ek s erfc(ks) (k > 0)
k p 4k
1 k2 s2 t
209 e erfc(ks) (k > 0) erf
s 2k
pffiffiffiffi pffiffiffi
k
210 eks erfc ks (k > 0) pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
p t(t þ k)

1 pffiffiffiffi 0 when 0 < t < k
211 pffiffi erfc( ks)
s (pt)1=2 when t > k
1 pffiffiffiffi 1
212 pffiffi eks erfc( ks) (k > 0) pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s p(t þ k)
 
k 1  pffiffi
213 erf pffiffi sin 2k t
s pt

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-39

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
  pffi
1 2 k 1
214 pffiffi ek =s erfc pffiffi pffiffiffiffiffi e2k t
s s pt
1
215 eas Ei(as) ; (a > 0)
tþa
1 1
216 þ seas Ei(as) ; (a > 0)
a (t þ a)2
hp i 1
217  Si(s) cos s þ Ci(s) sin s
2 t2 þ 1

0 when 0 < t < k [Kn (t)is Bessel function of the
218 K0 (ks)
(t 2  k2 )1=2 when t > k second kind of imaginary argument]
 2
pffiffi 1 k
K0 ðk sÞ exp 
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

219
2t 4t
1 ks 1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
220 e K1 (ks) t(t þ 2k)
s k
 2
1 pffiffi 1 k
221 pffiffi K1 (k s) exp 
s k 4t
 
1 k 2 pffiffiffi
222 pffiffi ek=s K0 pffiffiffiffiffi K0 (2 2kt)
s s pt
(
½t(2k  t)1=2 when 0 < t < 2k
223 peks I0 (ks)
0 when t > 2k
(
pkt
ffiffiffiffiffiffiffiffiffiffiffi when 0 < t < 2k
ks pk t(2kt)
224 e I1 (ks)
0 when t > 2k
1 P
1
225 2 u[t  (2k þ 1)a]
s sinh (as) k¼0

8 f (t)
6
4
2
0
0 a 3a 5a 7a t
1 P
1
226 2 (1) u(t  2k  1)
k
s cosh s k¼0

f(t)
2

0
0 1 2 3 4 5 6 7 t
1 as P1
227 tanh u(t) þ 2 (1) u(t  ak)
k
s 2 k¼1

f (t) Square wave

1

0
a 2a 3a 4a 5a t
–1

1 as P
1
228 1 þ coth u(t  ak)
2s 2 k¼0

f(t) Stepped function

4
3
2
1
0
0 a 2a 3a 4a t

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-40 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
m ma  as  P
1
229  coth  1 mt  ma u(t  ka)
s2 2s 2 k¼1

Sawtooth function
f(t) SLOPE = m

0
0 a 2a 3a t
" #
1 as 1 X
1
230 tanh tþ2 (1)k (t  ka)  u(t  ka)
s2 2 a k¼1
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Triangular wave
f (t)

0
0 a 2a 3a 4a 5a 6a t
1 P
1
231 (1) u(t  k)
k
s(1 þ es ) k¼0

f (t)

0
0 1 2 3 4 5 6 7 t
1 h   i 
a X p p
232 sin a t  k u tk
(s2 þ a2 )(1  e a s )
p

k¼0
a a
Half-wave rectification of sine wave
f (t)

0
0 π 2π 3π 4π t
a a a a

ps 1 h 
a X pi  p
233 coth ½sin (at)  u(t) þ 2 sin a t  k u tk
(s þ a )
2 2 2a k¼1
a a
Full-wave rectification of sine wave
f (t)

0
0 π 2π 3π 4π t
1 as a a a a
234 e u(t  a)
s
f (t)

1 ∞

0
0 a t

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-41

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
1 as u(t  a)  u(t  b)
235 (e  ebs )
s
f (t)
1

0
0 a b t
m as
236 e m  (t  a)  u(t  a)
s2
f (t)

SLOPE = m
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

0
0 a t
mt  u(t  a)
hma Or
mi as
237 þ e [ma þ m(t  a)]  u(t  a)
s s2
f(t)

E SLOPE = m

0
0 a t
2 as
238 e (t  a)2  u(t  a)
s3
f (t)

0

 0 a t
2
2 2a a as
239 3
þ 2 þ e t 2  u(t  a)
s s s
f (t)
t2
a2

0
0 a t
m m as
240  e mt  u(t)  m(t  a)  u(t  a)
s2 s2
f (t)
ma

SLOPE = m
0
0 a t
m 2m as m 2as
241  2 e þ 2e mt  2m(t  a)  u(t  a) þ m(t  2a)  u(t  2a)
s2 s s
f (t)
ma SLOPE SLOPE
=m =–m

0
0 a 2a t

(continued)

© 2010 by Taylor and Francis Group, LLC

 5-42 Transforms and Applications Handbook

TABLE A.5.1 (continued) Laplace Transform Pairs

F(s) f(t)
m ma m as mt  [ma þ m(t  a)]  u(t  a)
242  þ 2 e
s2 s s
f (t)
ma
SlOPE = m

0
0 a t
(1  es )2
243 0.5t2 for 0  t < 1
s3
1  0.5(t  2)2 for 0  t < 2
1 for 2  t
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

f(t)
1

0
0 1 2 3 t


(1  es ) 3
244 0.5t for 0  t < 1
2
s
0.75  (t  1.5)2 for 1  t < 2
0.5(t  3)2 for 2  t < 3
0 for 3 < t

f(t)
1

0
0 1 2 3 t
b b(ta)
þ (eba  1) (e  1)  u(t)  (e  1)  u(t  a) þ Ke
bt bt
 u(t  a)
s(s  b)
" #
1 s þ ebab1 as
245  e where K ¼ (eba  1)
s þ b s(s  b)
f (t)

0
0 a t

TABLE A.5.2 Properties of Laplace Transforms

F(s) f(t)
1 Ð
1
f(t)
est f (t)dt
0
2 AF(s) þ BG(s) Af(t) þ Bg(t)
3 sF(s)  f(þ0) f 0 (t)
4 s F(s)  s
n n1
f (þ0)  s n2 (1)
f (þ0)      f (n1)
(þ0) f (n)(t)
1 Ðt
5 F(s) f (t)dt
s 0
1 Ðt Ðt
6 F(s) f (l)dldt
s2 0 0

© 2010 by Taylor and Francis Group, LLC

 Laplace Transforms 5-43

TABLE A.5.2 Properties of Laplace Transforms

F(s) f(t)
7 F1(s)F2(s) Ðt
f1 (t  t)f2 (t)dt ¼ f1 * f2
0
8 F 0 (s) tf (t)
9 (1)nF(n)(s) tnf(t)
Ð
1 1
10 F(x)dx f (t)
s t
11 F(s  a) at
e f(t)
12 ebsF(s) f(t  b), where f(t) ¼ 0; t < 0
1 t 
13 F(cs) f
c c
1 (bt)=c t 
14 F(cs  b) e f
c c
Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

Ða
est f (t)dt
15 0
f(t þ a) ¼ f(t) periodic signal
1  eas
Ða
est f (t)dt
16 0
f(t þ a) ¼ f(t)
1 þ eas
F(s)
17 f1(t), the half-wave rectification of f(t) in No. 16.
1  eas
as
18 F(s) coth f2(t), the full-wave rectification of f(t) in No. 16.
2
p(s) Xm
p(an ) an t
19 , q(s) ¼ (s  a1 )(s  a2 )    (s  am ) 0 (a )
e
q(s) 1
q n

p(s) f(s) Xfr (rn)

(a) t n1
20 ¼ eat þ 
q(s) (s  a)r n¼1
(r  n) (n  1)

Sources: Campbell, G.A. and Foster, R.M., Fourier Integrals for Practical Applications, Van Nostrand,
Princeton, NY, 1948; McLachlan, N.W. and Humbert, P., Formulaire pour le calcul symbolique, Gauthier-
Villars, Paris, TX, 1947; A. Erdélyi and W. Magnus, Eds., Tables of Integral Transforms, Bateman Manuscript
Project, California Institute of Technology, McGraw-Hill, New York, 1954; based on notes left by Harry
Bateman.
Note: In these tables, only those entries containing the condition 0 < g or k < g, where g is our t, are Laplace
transforms. Several additional transforms, especially those involving other Bessel functions, can be found in
sources.

References 6. A.D. Poularikas and S. Seeley, Signals and Systems, cor-

rected 2nd edn., Krieger Publishing Co., Melbourne, FL,
1. R.V. Churchill, Modern Operational Mathematics in Engin- 1994.
eering, McGraw-Hill, New York, 1944. 7. G.A. Campbell and R.M. Foster, Fourier Integrals for
2. J. Irving and N. Mullineux, Mathematics in Physics and Practical Applications, Van Nostrand, Princeton, NJ,
Engineering, Academic Press, New York, 1959. 1948.
3. H.S. Carslaw and J.C. Jaeger, Operational Methods in 8. N.W. McLachlan and P. Humbert, Formulaire pour le
Applied Mathematics, Dover Publications, Dover, NH, 1963. calcul symbolique, Gauthier–Villars, Paris, 1947.
4. W.R. LePage, Complex Variables and the Laplace Trans- 9. A. Erdélyi and W. Magnus, Eds., Tables of Integral Trans-
form for Engineers, McGraw-Hill, New York, 1961. forms, Bateman Manuscript Project, California Institute of
5. R.E. Bolz and G.L. Turve, Eds., CRC Handbook of Tables for Technology, McGraw-Hill, New York, 1954; based on
Applied Engineering Science, 2nd edn., CRC Press, Boca notes left by Harry Bateman.
Raton, FL, 1973.

© 2010 by Taylor and Francis Group, LLC

 Downloaded by [University of Illinois at Urbana-Champaign] at 15:24 12 August 2017

© 2010 by Taylor and Francis Group, LLC