CHE160 MODULE 4

LAPLACE TRANSFORM AND

TRANSFER FUNCTION

Amierson C. Tilendo, MSc.

Asst. Professor IV
Chemical Engineering Department
Mindanao State University-Main

CHE160 Process Dynamics and Control

 LAPLACE TRANSFORM FOR LINEAR ODE
AND PDE
• Laplace Transform
– Not in time domain, rather in frequency domain
– Derivatives and integral become some operators.
– ODE is converted into algebraic equation
– PDE is converted into ODE in spatial coordinate
– Need inverse transform to recover time-domain solution
(D.E. calculation)
u(t) y(t)
ODE or PDE

L L -1
L L -1
L L -1

Transfer
U(s) Function Y(s)

(Algebraic calculation)
CHE160 Process Dynamics and Control
 DEFINITION OF LAPLACE TRANSFORM

• Definition

F (s) = L  f (t) =  f (t )e − st dt
0
– F(s) is called Laplace transform of f(t).
– f(t) must be piecewise continuous.
– F(s) contains no information on f(t) for t < 0.
– The past information on f(t) (for t < 0) is irrelevant.
– The s is a complex variable called “Laplace transform variable”

• Inverse Laplace transform

f (t) = L -1
F (s)
– L and L-1 are linear. L af1 (t) + bf 2 (t)= aF1 (s) + bF2 (s)
CHE160 Process Dynamics and Control
 LAPLACE TRANSFORM OF FUNCTIONS

• Constant function, a f(t)


 a  a a
L a= 
a
ae − st dt = − e−st = 0−−  =
0 s 0  s s t

• Step function, S(t)

1 for t  0
f (t) = S(t) =  f(t)
0 for t  0 1

  1 1
L S(t)=  e− st dt = − e− st = 0 −  −  =
1 t
0 s 0  s s

• Exponential function, e-bt

f(t) b<0

−1 −(b+s )t
 
 1
L e−bt =  e−bt e −st dt =
1
e = b>0
0 s +b 0 s +b t

CHE160 Process Dynamics and Control

 • Trigonometric functions sin(t)
– Euler’s Identity:e jt = cos t + j sin t 1

(e ) sin t = 1 (e jt − e − jt )

1 j t + e − jt 2
cos t =
t
2j 
2
1  1  1  1 1  
L sin t = L  e jt  − L  e− jt  =  −  =
 2 j  2 j  s − j s + j  s + 
2 2
2 j
1  1  1 1 1  s
L cos t = L  e jt + L  e− jt  =  + =
 2
2  2  2  s − j s + j  s + 
2

• Rectangular pulse, P(t)

0 for t  tw
f (t) = P(t) =  h for t w  t  0
f(t)
h
0 for t  0
 tw t

(1 − e )
tw
h h
L P(t) =  he − st dt = − e − st =
w t −tw s

0 s 0 s

CHE160 Process Dynamics and Control

 • Impulse function,  (t)
0 for t  tw
 f(t)
f (t) =  (t) = lim  1/ tw for tw  t  0 1/tw
t w →0
0 for t  0
 tw t

(1 − e−t ws )= 1
tw 1 1
L  (t) = lim  e− st dt = lim
tw →0 0 t tw →0 t s
w w

 f (t ) f (t) 
 L'Hospital's rule: lim = lim 
t →0 g (t)
 t →0 g(t)

• Ramp function, t
 −st
L t=  te dt f(t)
0
 1
e − st
−
t − st  1  − st 1
= e dt =  e dt = 2
−s −s
1 t
0
0 s 0 s

(Integration by part: 


0

f ' gdt = f  g 0 −


0
f  gdt )
• Refer the Table 3.1 (Seborg et al.) for other functions
CHE160 Process Dynamics and Control
CHE160 Process Dynamics and Control
CHE160 Process Dynamics and Control
 PROPERTIES OF LAPLACE TRANSFORM

• Differentiation
 df  
L   =  f  e −st dt = f (t )e −st

− 
f (−s)e −st dt (by i.b.p.)
 dt  0 0 0


= s  f  e−st dt − f (0) = sF (s) − f (0)
0

d 2 f 
L  2  =  f  e −st dt = f (t)e −st − f   (−s)e −st dt = s  f  e −st dt − f (0)
   

 dt  0 0 0 0

= s (sF (s) − f (0) )− f (0) = s 2 F (s) − sf (0) − f (0)


dn f    
L  n  =  f (n)  e −st dt = f (t) (n−1) e −st −  f ( n−1)  (−s)e −st dt
 dt  0 0 0

   d n−1 f 
= s 0 f (n−1) −st
 e dt − f (n−1)
(0) = s L  n−1  − f
(n−1)
(0)

  dt 
= s n F (s) − s n−1 f (0) − − sf (n−2) (0) − f (n−1) (0)
CHE160 Process Dynamics and Control
 • If f (0) = f ’ (0) = f ”(0) = …= f (n-1)(0) = 0, L   = sF(s)
df
– Initial condition effects are vanished.  dt 
– It is very convenient to use deviation d2 f  2
L  2  = s F(s)
variables so that all the effects of  dt 
initial condition vanish. 
dn f  n
L  n  = s F(s)
 dt 
• Transforms of linear differential equations.
y(t) ⎯L⎯ →Y (s), u(t) ⎯L⎯ →U (s)
dy(t) L
⎯⎯→sY (s) (if y(0) = 0)
dt
dy(t)
 →( s + 1)Y (s) = KU (s)
= − y(t ) + Ku(t) ( y(0) = 0) ⎯L⎯
dt
TL TL 1 TL(s) s +1)T (s) = T (s)
= −v + L
(Tw −TL ) ⎯⎯ → HL v + (
t z  HL z
HL L w

CHE160 Process Dynamics and Control

 • Integration
L  f ( )d =  ( f ( )d )e
t

0 0
 t

0
−st
dt
0 
−st
e 1  F (s)
 s 0
t
= f ( )d + f e − st
dt = (by i.b.p.)
−s 0
0
s
 Leibniz rule: d b(t) f ( )d = f (b( t)) db(t) − f (a(t)) da(t) 

 dt a(t ) dt dt 


• Time delay (Translation in time) f(t)

f (t) ⎯+⎯ ⎯→ f (t −  )S(t −  )

 in t
 t
 
L f (t −  )S(t −  )=  f (t − )e dt =  f ( )e − s( + ) d (let  = t −  )
− st
 0

 f ( )e − s d = e − s F(s)
− s
=e
0

• Derivative of Laplace transform

dF (s) d  d − st
e dt =  (−t  f )e − st dt = L −t  f (t) 
 

ds
=  f  e − st dt =
ds 0  0
f
ds 0

CHE160 Process Dynamics and Control

 • Final value theorem
– From the LT of differentiation, as s approaches to zero
 df
lim   e− st dt = lim sF (s) − f (0) 
s→0 0 dt s→0

 df
0 dt dt = f () − f (0) = lim s →0
sF (s) − f (0)  f () = lim sF (s)
s →0

– Limitation: f () has to exist. If it diverges or oscillates,

this theorem is not valid.
• Initial value theorem
– From the LT of differentiation, as s approaches to infinity
 df − st
lim   e dt = lim sF (s) − f (0) 
s→ 0 dt s→

 df
lim  e− st dt = 0 = lim sF (s) − f (0)  f (0) = lim sF (s)
s→ 0 dt s→ s→

CHE160 Process Dynamics and Control

 EXAMPLE ON LAPLACE TRANSFORM (1)
1.5t for 0  t  2
• 3
f(t)
 for 2  t  6
3 f (t) = 
0 for 6  t
2 6 t
0 for t  0
f (t) = 1.5t S(t) −1.5(t − 2)S(t − 2) − 3S(t − 6)
 F (s) = L  f (t)= 2 (1− e−2s ) − e −6s
1.5 3
s s
2
• For F (s) = , find f (0) and f ().
s−5
– Using the initial and final value theorems
2s 2s
f (0) = lim sF(s) = lim = 2 f () = lim sF (s) = lim =0
s→ s→ s − 5 s→0 s→0 s − 5

– But the final value theorem is not valid because

lim f (t) = lim2e5t = 
t → t →
CHE160 Process Dynamics and Control
 EXAMPLE ON LAPLACE TRANSFORM (2)

• What is the final value of the following system?

x  + x  + x = sin t; x(0) = x(0) = 0
1 1
 s X (s) + sX (s) + X = 2
2
 x(s)= 2
s +1 (s +1)(s 2 + s +1)
s
x()=lim 2 =0
s→0 (s +1)(s 2 + s +1)

– Actually, x() cannot be defined due to sin t term.

• Find the Laplace transform for (t sin  t ) ?
= L −t  f (t) 
dF(s)
From
ds
  2 s
L t sin t = −  2
d =
ds  s +  2  (s 2 +  2 ) 2

CHE160 Process Dynamics and Control

 INVERSE LAPLACE TRANSFORM

• Used to recover the solution in time domain

L -1 F (s)= f (t)
– From the table
– By partial fraction expansion
– By inversion using contour integral
f (t) = L -1 F(s) =
1


2 j C
est F (s)ds

• Partial fraction expansion

– After the partial fraction expansion, it requires to know some
simple formula of inverse Laplace transform such as
1 s (n −1)! e− s
, , , 2 2 , etc.
( s +1) (s + b) 2 +  2 sn  s + 2 s +1
CHE160 Process Dynamics and Control
 PARTIAL FRACTION EXPANSION
N (s) N (s) 1 n
F (s) = = = + +
D(s) (s + p1 ) (s + p n ) (s + p1 ) (s + p n )

• Case I: All pi’s are distinct and real

– By a root-finding technique, find all roots (time-consuming)
– Find the coefficients for each fraction
• Comparison of the coefficients after multiplying the denominator
• Replace some values for s and solve linear algebraic equation
• Use of Heaviside expansion
– Multiply both side by a factor, (s+pi), and replace s with –pi.

 i = (s + pi ) N (s)
D(s) s=− pi
– Inverse LT:
f (t) =  1e− p1t + 2e − p2t + + n e− pnt
CHE160 Process Dynamics and Control
 • Case II: Some roots are repeated
N (s) N (s) br −1s r −1 +  + b0 1 r
F (s) = = = = +  +
D(s) (s + p)r (s + p)r (s + p) (s + p) r

– Each repeated factors have to be separated first.

– Same methods as Case I can be applied.
– Heaviside expansion for repeated factors
1 d ( i)  N (s) r
=
 r −i (i ) 
(s + p)  (i = 0, , r −1)
i! ds  D(s)  s=− p

– Inverse LT
f (t) =  e− pt +  te − pt + + r
t r−1e− pt
1 2
(r −1)!

CHE160 Process Dynamics and Control

 • Case III: Some roots are complex
N(s) c s + c0  (s + b) + 1
= 1
F (s) = = 2 1
D(s) s + d1s + d 0 (s + b) 2 +  2

– Each repeated factors have to be separated first.

– Then,

1(s + b) + 1 = (s + b) 
 + 1
(s + b) + 
2 2 1
(s + b) + 
2 2
(s + b)2 +  2

where b = d1 / 2,  = d 0 − d12 / 4
1 = c1, 1 = (c0 − 1b ) / 
– Inverse LT
f (t) =  e−bt cos t +  e−bt sin t
1 1

CHE160 Process Dynamics and Control

 EXAMPLES ON INVERSE LAPLACE
TRANSFORM
(s + 5) A B C D
• F (s) = = + + + (distinct)
s(s +1)(s + 2)(s + 3) s s +1 s + 2 s + 3
– Multiply each factor and insert the zero value
(s + 5) D 
=  A + s
B C
+s +s  A = 5/6
(s +1)( s +2)(s + 3) s=0
 s +1 s +2 s + 3  s=0
(s + 5) A(s +1) C(s +1) D(s +1) 
=  +B+ +   B = −2
s(s + 2)(s + 3) s=−1
 s s+2 s + 3  s=−1

(s + 5) A(s + 2) B(s + 2) D(s + 2) 

=  + +C +   C = 3/2
s(s +1)(s + 3) s =−2  s s +1 s + 3 s =−2

(s + 5) A(s + 3) B(s + 3) C(s + 3)

=  + + + D   D = −1/3
s(s +1)(s + 2) s=−3  s s +1 s + 2  s =−3

 f (t) = L -1 F(s) =
5 3 1
− 2e− t + e−2t − e−3t
6 2 3
CHE160 Process Dynamics and Control
 • 1 As 2 + Bs + C D
F (s) = = + (repeated)
(s + 1) (s + 2)
3
(s +1) 3
(s + 2)
1 = (As 2 + Bs +C)(s + 2) + D(s +1) 3
= (A + D)s3 + (2A+ B + 3D)s2 + (2B + C + 3D)s + (2C + D)
 A = −D, 2 A + B + 3D = 0, 2B + C + 3D = 0, 2C + D = 1
 A = 1, B =1, C = 1, D = −1
1 d (i)  N (s)
= r 
– Use of Heaviside expansion  r −i (i)  (s + p)  (i = 0, ,r −1)
i! ds  D(s)  s=− p
s2 + s +1
=
1 +  2 +  3
(s + 1)3 (s + 1) (s + 1) 2 (s +1)3
(
(i = 0) :  3 = s 2 + s +1 ) =1
s=−1

(i = 1) :  2 =
1 d 2
1! ds
(
s + s +1 = −1)
s=−1

(i = 2):  1 =
1 d2 2
2! ds 2
(s + s +1) = 1
s=−1

 f (t) = L -1 F (s) = e−t − te−t + t 2e−t − e−2t

1
2
CHE160 Process Dynamics and Control
 (s +1) A(s + 2) + B Cs + D
• F (s) = 2 2 = + (complex)
s (s + 4s + 5) ( s + 2) +1
2
s2
s + 1 = A(s + 2)s 2 + Bs 2 + (Cs + D)(s 2 + 4s + 5)
= ( A + C)s 3 + (2 A + B + 4C + D)s 2 + (5C + 4D)s + 5D
 A = −C, 2 A + B + 4C + D = 0, 5C + 4D = 1, 5D = 1
 A = −1/25, B = −7/ 25, C =1/25, D =1/5

A(s + 2) + B 1 (s + 2) 7 B
= − −
(s + 2) 2 +1 25 (s + 2) 2 +1 25 (s + 2) 2 +1
Cs + D 1 1 1 1
2
= +
s 25 s 5 s2

 f (t) = L -1 F(s)= −
1 7 1 1
e−2t cos t − e−2t sin t + + t
25 25 25 5

CHE160 Process Dynamics and Control

 −2s
= 1+ e =  A B  −2s
• F (s)  +  (1+ e ) (Time delay)
+1
(4s +1)(3s +1)  4s +1 3s 
A = 1/(3s + 1) = 4, B = 1/(4s + 1) = −3
s=−1/4 s=−1/3

 4 3  −1  4e−2s 3e−2s 
 f (t) = L F(s) = L 
−1
−−1
+L  − 
 4s + 1 3s +1   4s + 1 3s +1 
= e−t / 4 − e−t / 3 + (e− (t−2 ) /4 − e− (t −2) / 3 )S(t − 2)

It is a brain teaser!!!
But you have to live with it.

Hang on!!!
You are half way there.

CHE160 Process Dynamics and Control

 SOLVING ODE BY LAPLACE TRANSFORM

• Procedure
1. Given linear ODE with initial condition,
2. Take Laplace transform and solve for output
3. Inverse Laplace transform

• Example: Solve for 5 dy + 4y = 2; y(0) = 1

dt

L 5  + L 4 y=L 2  5(sY(s) − y(0)) + 4Y (s) =

dy 2
 dt  s
2 5s + 2
(5s + 4)Y (s) = + 5  Y (s) =
s s(5s + 4)

 y(t) = L−1 Y (s)= L−1  +

0.5 2.5 
 = 0.5 + 0.5e−0.8t
 s 5s + 4 
CHE160 Process Dynamics and Control