Appendix A: Laplace Transforms

Important analytical method for solving linear ordinary

differential equations.

Definition
The Laplace transform of a function, f(t), is defined as:

F ( s ) = L [ f (t ) ] = f ( t ) e st dt

- Application to nonlinear ODEs? Must linearize first.

Laplace transforms play a key role in important process
control concepts and techniques.
- Examples:
Transfer functions

(3-1)

where F(s) is the symbol for the Laplace transform, L is the

Laplace transform operator, and f(t) is some function of time, t.
Note: The L operator transforms a time domain function f(t)
into an s domain function, F(s). s is a complex variable:
s = a + bj, j 1

Frequency response
Control system design
Stability analysis

Inverse Laplace Transform

where:

By definition, the inverse Laplace transform operator, L-1,

converts an s-domain function back to the corresponding time
domain function:

f (t ) = L 1 "$ F ( s )#%

Important Properties:

- a and b are constants

- X s L !" x t #$ and Y s L !" y t #$

()

()

()

()

Similarly,

Both L and L-1 are linear operators. Thus,

L !# ax ( t ) + by ( t )"$ = aL !# x (t )"$ + bL !# y (t )"$

= aX ( s ) + bY ( s )

- x(t) and y(t) are arbitrary functions

L 1 "$aX ( s ) + bY ( s )#% = ax (t ) + b y (t )
(3-3)

Laplace Transforms of Common Functions

2. Step Function
The unit step function is widely used in the analysis of process
control problems. It is defined as:

1. Constant Function

"$ 0 for t < 0

S t #
$% 1 for t 0

Let f(t) = a (a constant). Then from the definition of the

Laplace transform in (3-1),

L ( a ) = ae
0

st

a
dt = e st
s

# a$ a
= 0% & =
' s( s

(3-4)

()

(3-5)

Because the step function is a special case of a constant, it

follows from (3-4) that

1
L !# S (t )"$ =
s

(3-6)

2nd order derivatives

3. Derivatives
This is a very important transform because derivatives appear
in the ODEs we wish to solve. In the text (p.53), it is shown
that
! df "
L $ % = sF ( s ) f ( 0 )
(3-9)
& dt '
initial condition at t = 0
Proof:

d (e st )
L[ f ' (t )] = f (t )e st 0 f (t )
dt
dt
0

L(

f (t )( s )e st dt
0

Where

= f ' (t )

use the result for the 1st-order derivative

L(

d
) = s( s) (0)
dt

( s) = sF ( s) f (0)

= f (t )e st

d2 f
d
) = L( )
dt
dt 2

L(

d2 f
) = s[ sF ( s) f (0)] (0)
dt 2

L(

d2 f
) = s 2 F ( s) sf (0) f ' (0)
dt 2

= f (t )e st

+ s f (t )e st dt
0

= 0 f (0) + sL[ f (t )]

Similarly, for higher order derivatives:

4. Exponential Functions

"dn f #
1
L $ n % = s n F ( s ) s n1 f ( 0 ) s n2 f ( ) ( 0 )
&$ dt '%
n2
n 1
... sf ( ) 0 f ( ) 0
(3-14)

( )

Consider f (t ) = ebt where b > 0. Then,

( )

1 # (b+ s )t $
1
e
=
&0
b+s %
s+b
5. Trigonometric Functions

where: - n is an arbitrary positive integer

k
d f
f- ( ) 0
dt k t=0
Special Case: When all initial Conditions are Zero, then:

()

cos( wt ) =

!dn f "
L # n $ = sn F ( s )
%# dt $&

!0 for t < 0
"
f ( t ) = $h for 0 t < tw
"0 for t t
w
&

L[cos( wt )] =

f (t )

tw

tw

Then,

Time, t

1
L[ f (t )] = h e dt + 0 e dt = h e dt = h( )e st
s
0
tw
0
st

h
L[ f (t )] = [1 e t w s ]
s

s
s2 + w2

Find L[sin(wt)] by yourself

()

st

st

L "$ (t )#% = 1

Solution of ODEs by Laplace Transforms

tw

7. Impulse Function (or Dirac Delta Function)

The impulse function is obtained by taking the limit of the
rectangular pulse as its width, tw, goes to zero but holding
1
the area under the pulse constant at one. (i.e., let h = )
tw
t impulse function
Let,

The Laplace transform of the rectangular pulse is given by

h
F ( s ) = 1 e t w s
(3-22)
s

1 jwt
(e + e jwt )
2

It is defined by:

(3-16)

1
L[cos( wt )] = [ e jwt e st dt + e jwt e st dt
2 0
0
1
1
1
1 2s
= [
+
]=
2 s jw s + jw 2 s 2 + w2

In process control problems, we often assume zero

initial conditions. Reason: This corresponds to the
nominal steady state when deviation variables are
used, as shown in Ch. 4.

6. Rectangular Pulse Function

b+ s t
L #ebt $ = ebt e st dt = e ( ) dt
%
& 0
0

Procedure:
tw
0

1. Take the L of both sides of the ODE.

2. Rearrange the resulting algebraic equation in the s domain to
solve for the L of the output variable, e.g., Y(s).
3. Perform a partial fraction expansion.
4. Use the L-1 to find y(t) from the expression for Y(s).

Table&A.1.&Laplace&Transforms&

Example 3.1
Solve the ODE,

dy
+ 4y = 2
y (0) = 1
dt
First, take L of both sides of (3-26),
2
5 ( sY ( s ) 1) + 4Y ( s ) =
s
Rearrange,
5s + 2
Y (s) =
s ( 5s + 4 )
5

See&page&A53&of&the&textbook.&

Take L-1,

y (t ) = 0.5 + 0.5e0.8t

Example:

Perform a partial fraction expansion (PFE)

s+5
= 1 + 2
( s + 1)( s + 4 ) s + 1 s + 4

(3-37)

To find 1 : Multiply both sides by s + 1 and let s = -1

Basic idea: Expand a complex expression for Y(s) into

simpler terms, each of which appears in the Laplace
Transform table. Then you can take the L-1 of both sides of
the equation to obtain y(t).

s+5
Y (s) =
( s + 1)( s + 4 )

(3-34)

" 5s + 2 #
y (t ) = L 1 $
%
& s ( 5s + 4 ) '

From Table 3.1,

Partial Fraction Expansions

(3-26)

1 =

4
3

=
s =1

To find 2 : Multiply both sides by s + 4 and let s = -4

2 =

(3-41)

s+5
s+4

s+5
s +1

=
s =4

1
3

A General PFE
Consider a general expression,

(3-42)

where coefficients 1 and 2 have to be determined.

Y (s) =

N (s)
D(s)

N (s)
n

( s + bi )

i =1

(3-46a)

Heaviside expansion method

Example 3.2 (continued)

Here D(s) is an n-th order polynomial with the roots ( s = bi )

all being real numbers which are distinct so there are no repeated
roots.

Recall that the ODE,

y + +6
y +11 y + 6 y = 1, with zero initial
conditions resulted in the expression

The PFE is:

Y (s) =

N (s)
n

( s + bi )

=
i =1

Y (s) =

i
s + bi

(3-46b)

i =1

(3-40)

The denominator can be factored as

(3-50)

Note: Normally, numerical techniques are required in order to

calculate the roots.
The PFE for (3-40) is

s = bi

Y (s) =

Solve for coefficients to get

(For example, find , by multiplying both sides by s and then

setting s = 0.)
Substitute numerical values into (3-51):
1/ 6 1/ 2 1/ 2 1/ 6
Y ( s) =

+
+
s
s +1 s + 2 s + 3
Take L-1 of both sides:
"1/ 6 # 1 " 1/ 2 # 1 " 1/ 2 # 1 " 1/ 6 #
L 1 $"Y ( s )%# = L 1 &
L &
+L &
+L &
$ s '%
$ s + 1'%
$ s + 2 '%
$ s + 3 '%
From Table 3.1,

1 1 t 1 2t 1 3t
e + e e
6 2
2
6

1
= 1 + 2 + 3 + 4 (3-51)
s ( s + 1)( s + 2 )( s + 3) s s + 1 s + 2 s + 3

General Procedure for Solving Differential Equations

1
1
1
1
1 = , 2 = , 3 = , 4 =
6
2
2
6

y (t ) =

s s + 6s + 11s + 6

Where:

N ( s)
D( s)

s s3 + 6s 2 + 11s + 6 = s ( s + 1)( s + 2 )( s + 3)

Note: D(s) is called the characteristic polynomial.

i = ( s + bi )

1
3

(3-52)

Special Situations:

Repeated Factors (different from the text)

Two other types of situations commonly occur when D(s) has:

When the denominator D(S) has a term (s+b) occurring r

times, r number of terms must be included in the expansion.

i) Complex roots: e.g., bi = 3 4 j

ii) Repeated roots (e.g., b1 = b2 = 3 )

( j @ 1)

N ( s)

Y ( s) =

Where: n=k+r

[ ( s + bi )]( s + b) r
i =1

Y ( s) =
i =1

i
( s + bi )

1
s+b

2
( s + b) 2

+ +

r
( s + b) r

i can be obtained by Heaviside method as before.

Heaviside rule has to be modified to determine j
r
Let both sides multiplied by ( s + b)

Take derivative with respect to s to both sides of the above eqn.

k
N ( s)
i
( s + b) r = ( s + b) r
D( s )
i =1 ( s + bi )

k
(
i %
d '( s + b) r [
]$
(
s
+ bi ) #
dQ ( s )
i =1
&
=
ds
ds
+ ( r 1) 1 ( s + b) r 2 + 2 ( r 2)( s + b) r 3 + + r 1 + 0

+ 1 ( s + b) r 1 + 2 ( s + b) r 2 + + r

Let Q( s) =

N ( s)
( s + b) r
D( s)
k

Q ( s ) = ( s + b) r [
i =1

i
( s + bi )

+ 1 ( s + b) r 1 + 2 ( s + b) r 2 + + r 1 ( s + b) + r

Let s = b

r = Q( s) s=b =

N ( s)
( s + b) r
D( s)

s = b

Let s = b
dQ( s)
ds s=b
1 d ( l )Q( s)
Similarly: r l =
l ! ds ( l ) s = b
Therefore j can be determined by:

r 1 =

j =

1 d ( r j )Q( s )
j = 1r
( r j )! ds ( r j ) s = b

Repeated factor - example

( s + 1)
Y ( s) =
s( s + 2) 2
Y ( s) =

1 = s
2 =

1 =

1
s+2

( s + 1)
s( s + 2) 2

2
( s + 2) 2

s=0

( s + 1)
( s + 2) 2
s( s + 2) 2

# ( s + 1)
$
d&
( s + 2) 2 '
2
( s( s + 2)
)
ds

1
4

s =2

s =2

=1

y (t ) =

1 1 2t 1 2 t
e + te
4 4
2

# s + 1$
d&
( s )'
ds

Y ( s) =
s =2

= (

1
)
s2

s =2

1
4

( s 2 + d1 s + d 0 ) = ( s + a + jw)( s + a jw)

where

1 1 2t t 21e 2t 1
e +

4 4
(2 1)! 2

Complex Factors
Another possibility is that the factor of the characteristic
polynomial involves complex numbers.

1
1

1
2
Y (s) = 4 + 4 +
s s + 2 ( s + 2) 2

d
a = 1 and
2

y(t ) =

d2
w = d0 1
4

N ( s)
N ( s)
= k
D( s)
( s + bi ) ( s + b) r ( s2 + d1s + d 0 )

Where

d12
< d0
4

Multiply (s+a-jw) for the both sides

k
r
j

N ( s)
( s + a jw) = ( s + a jw)[ i +
]
D ( s)
s
+
b
(
s
+
b) j
i =1
j =1
i

Y(s) can be expanded into:

N ( s)
Y ( s) =
D ( s)
k

=
i =1

i
s + bi

i =1

cr + jci
( s + a jw) + cr jci
( s + a + jw)

Let s = a + jw
,

1
2
r
+
+ +
s + b ( s + b) 2
( s + b) r
c + jci
c jci
+ r
+ r
s + a + jw s + a jw
+ 1

cr and ci can be determined by the modified

Heaviside rule as well.

cr jci =

N ( s)
( s + a jw)
D( s)

s = a + jw

Therefore
" N ( s)
%
cr = Re#[
( s + a jw)] s= a + jw &
$ D( s)
'

" N ( s)
%
ci = Im#[
( s + a jw)] s= a + jw &
$ D( s)
'

Alternative: (complex factor)

Example: find y(t)

N ( s)
Y ( s) =
D ( s)
=

Y ( s) =

N ( s)
k
!
$
r
2
2
#"i =1( s + bi ) &%[( s + b) ][( s + a ) + w ]

s=2

1
2

N ( s)
1
( s + 1 i ) s =1+i = [
] s =1+i
D( s)
( s + 2)( s + 1 + i )
1
1
1
1 1
1 1+ i
=
=
=
=
=
(1 + i )( 1 + i + 1 + i ) (1 + i )(2i ) 2i 2
2 1 i
2 1 (i ) 2
1 i
1 1
=
= i
4
4 4
= cr ici

Alternative method
Y ( s) =

1
1
1
(1 i ) (1 + i )
Y ( s) = 2 + 4
+ 4
s+2
s +1+ i
s +1 i
1
" 1
#
" 1
#
$ (1 i ) % 1 $ 4 (1 + i ) %
y (t ) = L1 ( 2 ) + L1 $ 4
%+L $
%
s+2
$ s +1+ i %
$ s +1 i %
&
'
&
'
1 2 t 1
1
1 2 t 1
1
(1+i ) t
(1i ) t
e (1 i )e
(1 + i )e
= e (1 i )e t e it (1 + i )e t eit
2
4
4
2
4
4
1 2 t 1 t
1
1
e e (*(1 i )e it + (1 + i )eit )+ = e 2 t e t (*(e it + eit ) i (e it eit ) )+
2
4
2
4
1 2 t 1 t
e e [cos(t ) sin(t )]
2
2

cr + ici cr ici
+
s +1+ i s +1 i

cr ici =

! c (s + a)
N ( s)
c1s + c2
c2 ac1 "
= ... +
= ... + $ 1 2
+
%
2
D( s )
( s + a ) 2 + w2
(
s
+
a
)
+
w
(
s
+
a ) 2 + w2 '
&

# N ( s)
&
s+2
( s + 2) ( s=2 =
D
(
s
)
(
s
+
2
)[(
s + 1) 2 + 1]
$
'

(" k
+
( s + bi ) %'[( s + b) r ][( s + a ) 2 + w2 ],
)$i
&
*# =1
-

Therefore: cr = 1 and ci = 1
4

s+2

1 = %

collect all the terms with the same order, and solve for
all the coefficients
Y ( s) =

Y ( s) =

'k
i * ' 1
2
r *
= (
+
++(
+
2 +...+
( s + b)r ,
) i =1 s + bi , ) s + b ( s + b)
c1s + c2
+
( s + a ) 2 + w2

Multiply both sides by

1
( s + 2)[( s + 1) 2 + 1]

c s + c2
= 1 + 1
( s + 2)[( s + 1) 2 + 1] s + 2 ( s + 1) 2 + 1

1[ s 2 + 2 s + 2] + ( s + 2)( c1s + c2 )
( s + 2)[( s + 1) 2 + 1]

s2 :
s:

1 + c1 = 0
21 + 2c1 + c2 = 0
21 + 2c2 = 1

s0:
Y ( s) =

1 =

1
2

c1 =

1
2

c2 = 0

1 1
1
s +1
1
1

+
2 s + 2 2 ( s + 1) 2 + 1 2 ( s + 1) 2 + 1

y (t ) =

1 2t 1 t
1
e e cos t + e t sin t
2
2
2

Same result as before

Important Properties of Laplace Transforms

Proof:
Laplace transform for the 1st-order derivative

1.Final Value Theorem

It can be used to find the steady-state value of a closed loop
system (providing that a steady-state value exists.
Statement of FVT:

dy
dy
] = e st dt = sY ( s ) y (0)
dt
dt
0
\______/
(1)

\_______/
(2)

Take limit for terms (1) and (2), as s! 0

dy
lim e st dt = lim[ sY ( s) y(0)]
s 0
s 0
dt
0

lim y ( t ) = lim #% sY ( s )$&

t

L[

s 0

y (t ) 0 = lim [sY ( S ) y (0)]

providing that the limit exists (is finite) for all

Re ( s ) 0, where Re (s) denotes the real part of complex
variable, s.

s >0

y () y (0) = lim sY ( s) y(0)

s 0

y ( ) = lim sY ( s)
s0

2. Initial Value Theorem (IVT)

Similar to the final value theorem, initial value theorem (IVT)
can be stated as:

Example:
Suppose,

Y (s) =

5s + 2
s ( 5s + 4 )

(3-34)

Then,

y (0) = lim[ sy ( s)]

s

Prof.:

dy

dt e

st

dt = sy ( s) y (0)

# 5s + 2 $
y ( ) = lim y (t ) = lim %
& = 0.5
t
s 0 ' 5s + 4 (

take the limit for s

lim

You may use FVT to final steady value without doing the
inverse Laplace transform; also, it may be used to check if
an inversion is correctly performed.

dy st
e dt = lim[ sy( s) y(0)]
s
dt

y(0) = lim[ sY ( s)]

s

Example:

y( s) =

5s + 2
s(5s + 4)

y (0) = lim[ s
s

Find y(0)?

5s + 2
5s + 2
] = lim
=1
s(5s + 4) s 5s + 4

3. Transform of an integral

4. Time Delay

It is also some times necessary to find Laplace transform of

an integral.
t

L[ f (t * )dt * ] = { f (t * )dt * }e st dt
0

use integration-by-part
t

udv = uv vdu

st

1
d (e )
{ f (t * )dt * }
s 0 0
dt

t
$
1#
1
= &[( f (t * )dt * )e st ] ' +
0
s( 0
) s

Time delays occur due to fluid flow, time required to do an

analysis (e.g., gas chromatograph). The delayed signal can be
represented as
y ( t ) = time delay
Also,
L "$ y (t )#% = esY ( s )
At time t = 0,
TC
TC
Motor
Ti=0
heater turns on
w
delay may also
V
be represented
Q
by td.
1

t
#
$
d f (t * )dt * %
%
% st 0
%
& e
'
dt
%0
%
%(
%)

Heater

$ 1
1#
= %0 0 e st f (t )dt & = f (t )e st dt
s'
0
( s0

1
= F ( s)
s

# f (t td ) t td
f d (t ) = $
t < td
% 0

t s
How to find fd(t) from Fd ( s ) = e F ( s )
d

or

Two steps:

f d (t ) = f (t td ) s(t td )
td

st

(1) Find f(t)

st

st

L[( f d (t )] = f d (t )e dt = 0 e dt + f (t td )e dt
0

td

(2) Replace f(t) by f d (t ) = f (t td ) s(t td )

Example exercise

= f (t t d )e st dt
td

Y ( s) =

Let t = t td
*

L[( f d (t )] = f (t * )e st e std dt * = e std f (t * )e st dt * = e std F ( s )

0

Fd ( s ) = L[ f (t td ) s(t td )] = e

td s

1 + e2 s
(4s + 1)(3s + 1)

F ( s)

Find y(t)? (do it by yourself)